SCREENING FOR FOOD INSECURITY IN PRIMARY CARE TO ENHANCE THE
MANAGEMENT OF DYSGLYCEMIA IN INDIVIDUALS WITH TYPE 2 DIABETES
by
Stephanie L. Gyra
B.H.K., Trinity Western University, 2007
BSc Nursing, University of British Columbia, 2011

PROJECT SUBMITTED IN PARTIAL FULFILLMENT OF
THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF NURSING
IN
NURSING - FAMILY NURSE PRACTIONER

UNIVERSITY OF NORTHERN BRITISH COLUMBIA
July 2019
© Stephanie Gyra, 2019

ii

Abstract
Type 2 Diabetes (DM 2) is increasingly prevalent worldwide. Its potential for debilitating
long-term sequelae and subsequent burden on healthcare systems highlight the importance of
adequate diabetes management. Glucose control remains central to treatment and often includes
nutritional therapy, pharmacotherapy and self-management strategies. Individuals with DM 2
who experience food insecurity (FI) are at an increased risk of poorly managed diabetes. Nurse
practitioners in primary care are specifically skilled at identifying patient difficulties in making,
adopting and adhering to lifestyle changes, thus are ideally positioned to address barriers to
chronic disease management. However, it remains unclear how FI influences DM 2 and how it is
accurately identified in the primary care setting. An integrative literature review was completed
to identify which strategies nurse practitioners can employ in primary care to identify and thus
enhance the management of DM 2 among patients experiencing FI. With the guidance of
Whittemore and Knalf’s approach to integrative reviews and the incorporation of Maslow’s
Theory of Motivation (MTM), a literature analysis was conducted revealing a complex
relationship between FI and the management of DM 2. The review highlights how FI interferes
with an individual’s capacity to self-manage their DM 2 by influencing food behaviors,
medication use and mental health. MTM highlights prerequisite needs necessary for diabetes
self-management amongst individuals experiencing FI. Challenges in identifying FI in practice
are identified, along with strategies to incorporate into practice. Important interventions and
clinical recommendations are discussed including: assessing for dysglycemia, assessing for
depression and distress, pharmacological management, social support and resource allocation.
All are important approaches that nurse practitioners can incorporate in the primary care setting.
Finally, limitations and areas for future research are discussed.

iii

TABLE OF CONTENTS
Abstract

ii

Acknowledgements

v

Glossary of Terms

vi

List of Tables and Figures

ix

Chapter 1 Introduction
Research Question

1
4

Chapter 2 Background

6

Overview of Type 2 Diabetes Mellitus
Incidence and Prevalence in Canada
Pathophysiology
Assessment, Diagnosis and Monitoring
Complications
Non-pharmacological Management
Pharmacological Management
Pharmacare and Diabetes
Overview of Food Insecurity
Food Insecurity in Type 2 Diabetes
Identifying Food Insecurity
Benefits of Screening Food Insecurity in Practice
Maslow’s Theory of Motivation
The Nurse Practitioner (NP)
NPs in Primary Care

6
6
7
8
8
10
14
19
21
22
23
24
25
26
26

Chapter 3 Methods
Literature Search Strategy
Preliminary Search
Primary Search Strategy
Inclusion and Exclusion Criteria
Secondary Search Strategy
Data Evaluation and Analysis

28
28
29
30
33
34
35

Chapter 4 Findings
Underlying Mechanisms of Influence
Changes in Food Behaviors
Hierarchy of Eating Patterns
Clinical Approach
Changes in Mental Health
Hierarchy of Social Need

37
38
38
42
42
44
45

iv

Clinical Approach
Changes in Medication Use
Clinical Approach
Reduced Self-Efficacy
Clinical Approach
Food Insecurity Screening
Clinical Approach
Definition
Attributes & Tools

47
48
50
51
52
52
53
54
57

Chapter 5 Discussion
Theme 1: Mechanisms of Influence
Changes in Food Behaviors
Recommendations for Practice
Changes in Mental Health
Recommendations for Practice
Changes in Medication Use
Recommendations for Practice
Theme 2: Hierarchy of Need and Motivation
Recommendations for Practice
Theme 3: Food Insecurity Screening
Selecting a Definition
Identifying Attributes
Screening Tool Selection
Recommendations for Practice

60
62
62
63
64
65
67
69
72
73
73
74
75
75
76

Limitations and Future Research
Gaps in the Literature
Other Limitations and Considerations

76
77
79

Conclusions

80

References

84

Appendices
Appendix A: Search Strategy Diagrams
Appendix B: Literature Review Matrix
Appendix C: Table of Food Insecurity Tools

98
98
100
124

v

Acknowledgements
I would like to express my gratitude and appreciation for my advisory committee
members Sylvia Barton, Linda Van Pelt and Shanda Rojas. Thank you for your guidance,
encouragement, ongoing support and contributions to this paper. Thank you, Shanda Rojas, for
the perspective you brought to this project, for your passion in diabetes care, and for sharing your
wealth of clinical expertise.
To my wonderful husband, John Gyra: Thank you for inspiring and motivating me to
follow my passion. Thank you for always believing in me in my moments of greatest doubt.
Thank you for just listening when I contemplated an alternate career in landscaping over the last
year and for providing unwavering belief in my abilities. You inspired me when it was needed
the most. LOML. Thank you, Lynne and Carol Sylvain, for your steadfast support in my
academic pursuits and in life. Your strong work ethic, beliefs about commitment, simple words
of wisdom and overall encouragement provided me with a foundation supportive of success.
I would like to dedicate this paper to Frances Hodgson. Your nursing stories from abroad, fierce
independence, insatiable desire for learning, and tenacity in recovering inspire me beyond
measure. I love you, Grandma.

vi

Glossary of Terms
Canadian Community Health Survey (CCHS): National survey comprised of both pan-Canadian
annual surveys and focused surveys directed at 35,000 Canadians every 3 years. The
purpose is to extract population-level information on the health determinants, health
status and health system utilization of Canadians. The CCHS is a collaborative effort
between Health Canada, the Public Health Agency of Canada (PHAC), Statistics Canada
(SC), and the Canadian Institute for Health Information (CIHI) (Government of Canada,
2017).
Diabetes Canada (DC): The national association for diabetes in Canada that advocates on behalf
of individuals with diabetes and supports Canadians, their care providers and research,
which is aimed at living healthy lives, preventing diabetes-associated complications
and finding a cure. (Diabetes Canada, 2019).
Diabetes Distress (DD): Negative emotions and feelings such as despondency and turmoil, and the
perceived burden of living with and managing diabetes. Particularly, diabetes distress usually
refers to the pressure felt to monitor, treat, and worry about complications on a
continuous basis (Robinson et al., 2018)

Depression: Refers to the experience of individuals who present with persistent depressive
feelings and feelings of low mood, which interferes with daily living, as well as of
individuals meeting the DSM-V criteria for major depressive disorder (MDD).
Fair Pharmacare: The only Pharmacare plan eligible to all residents of B.C. with active Medical
Service Plan (MSP) B.C. accounts. Fair Pharmacare is income-based, meaning families
with lesser income receive more help with the cost of Pharmacare-approved drugs,
supplies and dispensing fees (Government of British Columbia, 2018a).

vii

Food Insecurity: Defined differently throughout the literature. With respect to diabetes it
generally refers to uncertain access and consumption of nutritionally adequate food in
sufficient quantities, and usually the result of many complex sociopolitical interactions.
Glycated Hemoglobin (A1C): HB A1C is used as an estimate of mean plasma glucose (or
glucose control) over the previous 8-12 weeks. Fifty percent of the value is derived from
the preceding 30 days, while only 10% of the value is derived from the prior 90 to 120
days. The A1C is used for diabetes diagnosis, and monitoring of treatment effectiveness.
DC provides evidence-based guidelines for A1C monitoring (Berard, Siemens, & Woo,
2018).
Medical Services Plan (MSP): The name of publicly funded health insurance in British
Columbia, which covers the costs of all medically-necessary care provider services. All
residents of B.C. are eligible (Government of British Columbia, 2018f).
Nurse Practitioner: A graduate-prepared registered nurse who obtained advanced knowledge for
autonomous practice by completing either an accredited family or acute care nurse
practitioner (NP) master’s program. NPs possess competencies in health management
provision, ordering and interpreting diagnostic tests, diagnosing and treating, providing
disease prevention, health promotion and health management (Canadian Nurses
Association [CNA], 2010).
Pharmacare: A publicly funded medication and supply program through the government of B.C.,
assisting its residents with active Medical Services Plan of B.C. (MSP) coverage, with the
costs of prescription medications and supplies that are eligible within the program.
Residents are eligible under one of several plans (Government of British Columbia,
2018a).

viii

Primary Care Provider: Health care provider (physician, nurse practitioner or) who is most
responsible for the care of a patient in the community setting, usually over an extended
period of time.
Special Authority (SA) Coverage: Refers to when Pharmacare provides full or partial benefits to
a drug or medical supplies previously not covered, and when the drug’s eligibility criteria
has been met and SA forms have been sent to Pharmacare. The level of reimbursement is
still dependent upon annual deductible requirements (Government of British Columbia,
2018g).

ix

List of Tables and Figures

Table 1: Search Terms Utilized in Specified Databases to Extract Articles for Aim 1 & 2 31
Table 2: Search Terms Utilized in Specified Databases to Extract Articles for Aim 3

32

Table 3: Sequential Search Strategy Utilized for the Database Searches

33

Table 4: Inclusion and Exclusion Eligibility Criteria for the Selected Literature

33

Diagram 1: Concept Map Driven by PubMed MeSH and Keywords

30

Diagram 2: The Impact of FI on DM 2 and Strategies to Improve Management

61

1

Chapter 1
Introduction
Type 2 diabetes mellitus (DM 2) is a chronic metabolic disease with increasing
worldwide prevalence (Vas et al., 2017). In Canada, the prevalence of DM 2 is projected to
reach 10.8% by 2020 (Canadian Diabetes Association and Diabetes Quebec, 2010). Individuals
with diabetes access primary care at significantly higher rates and have higher rates of
comorbidities (Greiver et al., 2014). Vasculopathies, retinopathies, neuropathies and
nephropathies are often the long-term sequelae of DM 2 (Nield et al., 2017), meaning the
potential for burden on healthcare systems is substantial. Managing DM 2 is lifelong and
requires ongoing monitoring, interventions and follow-up from both the patient and healthcare
provider (Vas et al., 2017). Current evidence establishes that individualized, provider-supported,
patient-focused strategies are key to DM 2 treatment and management (Ferguson, Swan,
Swaldone, 2015; Vas et al, 2017).
According to the Canadian Guidelines on Diabetes (DCG), pharmacological and lifestyle
interventions remain the cornerstone of treatment (Lipscombe et al., 2018). The significance of
nutrition in management of DM 2, for example, cannot be overstated. The DCG endorse a
variety of nutritional eating patterns, and recommends setting nutritional goals best suited to the
individual’s preferences (Sievenpiper, Chan, Dworatzek, Freeze, & Williams, 2018). A poor diet
can have detrimental effects on the management of DM 2 and the prevention of its complications
(Seligman, Jacobs, López, Tschann, & Fernandez, 2012).
Definitions of food insecurity (FI) are varied in the literature. FI may arise from several
causes, although common antecedents include uncertain access to food (Essien, Shahid, &
Berkowitz, 2016; Seligman et al., 2012; Silverman et al., 2015) resulting in inadequate
consumption (Essien et al., 2016; Gucciardi, Vogt, DeMelo, & Stewart, 2009). In Canada, FI

2

presents a significant social and health burden. It also is closely linked to financial constraints
and poverty. In fact, poverty is the single most important predictor of FI in Canada (Tarasuk,
Mitchell, & Dachner, 2016). While poverty may co-occur commonly with FI, other possible
determinants of FI remain unknown. Antecedents and other determinants are examined in the
literature review in order to suggest possible solutions.
An estimated one third of individuals with diabetes experience FI (Montgomery, Lu,
Ratliff, & Mezuk, 2017). The impact of FI on the management and control of DM 2 has been
documented (Heerman et al., 2016). Food insecurity is an independent predictor of glycemic
control (Patil, Craven, & Kolasa, 2017; Seligman et al., 2012) and presents a significantly
heightened risk of dysglycemia (Essien et al., 2016). For example, individuals with DM 2
experience an increase in hypoglycemic events, subsequent hospitalizations and also higher
glycated hemoglobin (A1C) levels at disproportionate rates (Gucciardi, Vahabi, Norris, Del
Monte, & Farnum, 2014). What is challenging is that the mechanisms through which FI
influences DM 2 are unknown and not commonly identified in practice. Thus, its impact has
remained invisible to the practitioner. Establishing management strategies for individuals with
DM 2 without a clear understanding of the patient’s underlying struggle with FI may result in
both ineffective and unsafe treatment plans (Essien et al., 2016). In order to enhance the
management of DM 2 in individuals experiencing FI, it is imperative that providers understand
how food insecurity influences the management of DM 2, and how it can be correctly identified
in practice.
Primary care providers (PCP) are well-situated to address intimate issues such as FI, due
to the trust and respect that frequently develops between patients and PCP. The chronicity of the
professional relationship provides opportunities for the PCP to identify the patient with DM 2

3

experiencing food insecurity. Further, PCPs primarily manage the individual with DM 2, thereby
allowing the PCP to monitor the effects and outcomes of management plans. Nurse practitioners
(NPs) are positioned to address barriers to patient health. NPs have the ability to grasp crucial
aspects of a patient’s perspective, motivations, issues, barriers and goals (Bartol, 2012). In
British Columbia (B.C), NPs are trained to work with diverse populations, identify disparities in
health outcomes, and address challenges to making, adopting and adhering to lifestyle changes
(British Columbia College of Nursing Professionals [BCCNP], 2018a). B.C-trained NPs possess
the competencies necessary to play a comprehensive role in healthcare, working with patients,
families, communities and policymakers in order to improve health outcomes (BCCNP, 2018a).
By identifying FI as a barrier to well-managed diabetes, NPs can adapt their expectations and comanagement strategies to safely meet the needs of the patient to improve dysglycemia and
overall outcomes.
Addressing FI in its entirety is complex, requiring consideration of socioeconomical and
political contexts. Certain primary care interventions, however, can offer fairly immediate relief
to an otherwise highly complex situation. A comprehensive strategy to improve all aspects of FI
requires a multipronged approach consisting of individual, community, policy and political
interventions, which are beyond the scope of this paper.
Project Purpose
This integrative literature review arose from graduate research on FI in diabetic
populations and key conversations with NPs managing diabetes in primary care. Shanda Rojas, a
primary care NP who specializes in diabetes care, highlighted a population of individuals with
DM 2 who appeared to suffer additional difficulties in managing their diabetes. Subsequently,
these difficulties led to additional risks and consequences to the patient’s overall health. This

4

trend was evidenced in my graduate research throughout my NP program. Regrettably, what the
literature supported was the providers’ overall unawareness of the patient’s struggle with FI and
thereby an inability to create an effective management plan consisting of safe, patient-centered
pharmacological and non-pharmacological recommendations.
Although there are clear guidelines for the treatment and management of DM 2 (Imran,
Agarwal, Bajaj, & Ross, 2018; Lipscombe et al., 2018; Sievenpiper et al., 2018; Sigal et al.,
2018), it is unclear how providers can identify and safely manage individuals with DM 2
experiencing FI without understanding its impact on food consumption and other selfmanagement behaviors. Diabetes Canada recognizes the influence of FI on diabetes and has
subsequently published a position statement (Diabetes Canada, 2017a). However, it does not
provide any recommendations or guidance for the practitioner managing individuals’ FI. This
project aims to answer the question: “How can nurse practitioners use the identification of FI in
primary care settings to enhance the management of DM 2 among adults experiencing
dysglycemia?”. In order to establish this aim, this integrative review examined key background
areas through the following three questions: (1) How is food insecurity defined in the context of
DM 2?, (2) How does food insecurity influence the management of DM 2?, and, (3) What tools
presently exist to identify food insecurity and how can they successfully be applied to the
Canadian healthcare context?
An integrative literature review methodology was used to explore each subtheme of the
research question. For each theme, a preliminary search was conducted, and followed the
establishment of inclusion and exclusion criteria to narrow each search strategy. Finally, a
strategical analysis was conducted of the retrieved literature. A literature review matrix presents
key information and findings, as well as strengths and limitations. Practice recommendations

5

elicited through an analysis and synthesis of the findings are described. Clinical implications and
potential areas for future research are highlighted. By outlining the impact of food insecurity on
DM 2 and emphasizing how screening can be used to modify DM 2 management, the findings of
this integrative review will assist future PCNPs in managing patients with DM 2 experiencing FI.

6

Chapter 2
Background
Food insecurity (FI) can significantly affect the management of type 2 diabetes (DM 2).
Poorly managed diabetes is a significant problem in healthcare and leads to several long-term
consequences. PCP can help mediate some of the burden and risk associated with FI to improve
the management of DM 2. In this chapter, the incidence and prevalence of DM 2 in Canada is
reviewed, followed by the pathophysiology, assessment, diagnosis of DM 2, along with its
complications. Following this, the pharmacological and non-pharmacological clinical guidelines
from Diabetes Canada are discussed, proceeded by a short discussion on Pharmacare and the
coverage of antihyperglycemic medications. This is followed by a discussion on FI, its influence
on health and the management of DM 2, along with barriers to identifying FI in practice. Further,
Maslow’s Theory of Motivation (MTM) is discussed along with its relevance to patient
behaviors, motivation and managing DM 2 in individuals with FI. Lastly, the role of the NP as a
PCP is discussed, and their capacity to identify and mediate the risks associated with FI in DM 2.
Overview of Type 2 Diabetes Mellitus
Diabetes Mellitus 2, or Type 2 Diabetes (DM 2) is a chronic metabolic disease affecting
the entire body. Traditional clinical indicators of DM 2 include: (a) insulin resistance in the liver,
muscle and adipose tissue; (b) decreased insulin production in the pancreas secondary to beta cell
dysfunction and loss; which (c) results in inappropriate metabolism of glucose throughout the
body (Brashers, Jones & Huether, 2014b). Recently, multiple organs and hormones have been
implicated in the development of DM 2 (DeFronzo, 2009). And, the onset of DM 2, historically
between 40-70 years, has shifted to much younger populations in recent years (Brashers, Jones &
Huether, 2014b).

7

Incidence and prevalence.
Over the last three decades, the incidence of DM 2 has doubled across all age groups
(Brashers, Jones & Huether, 2014). DM 2 accounts for 90-95% of incidence rates of diabetes
(Canada Institution for Health Information, 2015; Centre for Disease Control and Prevention,
2017; Diabetes Canada, 2017; International Diabetes Federation, 2014 & PHAC, 2011). In 2016,
the estimated incidence of DM 2 in Canada was approximately 6.3%-6.65%, and occurred more
frequently in males than females (Statistics Canada, 2016).
Despite this uniform increase in incidence, some age groups have experienced a sharp
rise in prevalence. Between 1999-2009, the prevalence of DM 2 doubled in individuals aged 3549 years (Public Health Agency of Canada [PHAC], 2011), meaning that DM 2 is becoming
more prevalent in younger age groups. In 2015, the estimated prevalence of DM 2 in Canada was
between 8.4-8.8% (Diabetes Canada, 2017a). In 2017, the prevalence of DM 2 in British
Columbia (B.C.) was 9.5% (Diabetes Canada, 2017b). A steady increase in prevalence is
projected for Canada (Diabetes Canada, 2017b) to an estimated 10.8% by 2020 (Canadian
Diabetes Association and Diabetes Quebec, 2010). It is estimated that treating DM 2, its
complications, and associated productivity loss will cost greater than $19 billion a year by 2020
(Canadian Diabetes Association, 2009).
The incidence and prevalence of DM 2 varies by age, gender, income, ethnicity and
geographical region (PHAC, 2011; Shah, 2013). The incidence and prevalence of DM 2 is likely
underrepresented as a third of cases are estimated to be undiagnosed (Diabetes Canada, 2017a),
and up to a third of people with known diabetes do not feel comfortable disclosing their
diagnosis (Canadian Diabetes Association, 2015).

8

Pathophysiology.
DM 2 involves complex interactions between the environment and individual genetics,
resulting in dysfunction of multiple organs and hormones which cause: (a) insulin resistance
secondary to liver (DeFronzo, 2009), muscle cell (Brasher, Jones & Huether, 2014b) and adipose
tissue dysfunction (Yazıcı & Sezer, 2017); (b) reduced insulin production secondary to beta cell
dysfunction and loss in the pancreas (Brasher, Jones & Huether, 2014b); (c) dysregulated
signalling of insulin secretion by the incretins in the gastrointestinal tract (Tasyurek, Altunbas,
Balci, & Sanlioglu, 2014) and brain (Kleinridders, Ferris, Cai, & Kahn, 2014); (d) dysregulation
of blood glucose levels by the kidneys (Brasher, Jones & Huether, 2014b); (e) dysregulated
signalling of glucose production by adipose tissue (Stinkens, Goossens, Jocken, & Blaak, 2015),
dysregulated signalling of incretins in the gastrointestinal tract (Tasyurek, Altunbas, Balci, &
Sanlioglu, 2014) and the pancreas (Brasher, Jones & Huether, 2014b); (f); dysregulated glucose
production in the liver (Petersen, Vatner, & Shulman, 2017); (g) dysregulated glucose
homeostasis from disrupted glucose sensing in brain (Brashers, Jones & Huether, 2014a, p.734);
and (h) dysregulated leptin (Meek & Morton, 2016) and ghrelin secretion (Chabot, Caron,
Laplante, & St-Pierre, 2014).
Assessment, diagnosis and monitoring.
Diabetes Canada (DC) provides national guidelines for the assessment, diagnosis and
management of individuals with DM 2. According to the DC guidelines, diabetes is diagnosed
when two of the following criteria are met: (a) The glycated hemoglobin (A1C) is greater than
6.5 %, (b) a fasting plasma glucose is greater than 7mmol/L, (c) a random plasma glucose is
greater than 11 mmol/L, or (d) a 2-hour plasma glucose with a 75 gram oral glucose tolerance
test is greater than 11.1 mmol/L (Punthakee, Goldenberg, & Katz, 2018). Only one positive test

9

is required for the individual presenting with symptomatic hyperglycemia (Punthakee et al.,
2018).
Glycated hemoglobin (A1C).
The A1C is a reliable surrogate marker of glycemic control, representing the mean
plasma glucose for the previous 8-12 weeks (Berard et al., 2018). Chronically elevated A1C’s
are associated with an increased risk of diabetic complications, such as retinopathies,
nephropathies, neuropathies and cardiovascular disease (Seligman & Schillinger, 2010), and
therefore the guidelines recommend measuring the A1C every 3 months until target A1C levels
are reached (Berard et al., 2018). The target A1C for most individuals is less than 7%, and when
clinically indicated, select individuals target A1C levels between 7.1 to 8.5%, or greater than
8.5% (Imran et al., 2018). While A1C is a reliable mean, it does not capture individual episodes
of hyperglycemia and hypoglycemia.
Self monitoring of blood glucose (SMBG).
Using a glucose meter to SMBG provides isolated blood glucose values, and therefore
can accurately identify individual episodes of dysglycemia and provide immediate information
on the effects of pharmacological and non-pharmacological interventions (Berard et al., 2018).
The DC guidelines recommend an individualized approach to SMBG considerate of diabetic
type, pharmacological regimen, treatment changes, illnesses, hypoglycemia awareness or risks
and occupational requirements (Berard et al., 2018). SMBG is most successful at promoting
behavior change, improving blood glucose levels and preventing hypoglycemia when paired with
structured education (Berard et al., 2018).

10

Complications.
Several complications are associated with diabetes: (a) macrovascular complications
including hypertension, coronary artery disease, peripheral vascular disease and cerebral vascular
disease; (b) microvascular complications including retinopathies, nephropathies, peripheral
neuropathies and gastroparesis; (c) infections; and (d) amputations (Brashers, Jones & Huether,
2014a). Diabetes Canada (2017b) estimates that 30% of cerebral vascular events, 40% of
myocardial infarctions, 50% of end-stage renal failure and 70% of non-traumatic amputations are
attributable to diabetes (p.3). Optimal diabetes management, including maintaining target A1C
levels, is key in the prevention of complications.
Management.
The Canadian diabetes guidelines for the treatment and management of DM 2
recommend a multifaceted pharmacological and non-pharmacological approach that: (a)
manages and prevents hyperglycemia and its associated symptoms, and (b) decreases the risks of
microvascular and macrovascular complications (Lipscombe et al., 2018). Even with optimal
management and strict adherence to successful treatment strategies, DM 2 usually progresses
over time resulting in increased use of non-pharmacological and pharmacological interventions
(Lipscombe et al., 2018).
Non-pharmacological management.
Non-pharmacological strategies are the first-line treatment in the management of DM 2,
and recommended as monotherapy for the first 3 to 6 months when an individual’s A1C is less
than 8.5% at diagnosis and they present with no signs of metabolic decompensation (Lipscombe
et al., 2018). Some individuals achieve target A1C levels using lifestyle interventions only.

11

DM 2 Self-management education (SME) and support (SMS).
DM 2 SME and SMS use patient-provider collaboration and clinical support that focuses
on educating, motivating and empowering individuals to problem-solve and use goal setting
strategies to modify behaviors and self-monitor their chronic disease (Sherifali, Rabi, et al.,
2018). SME and SMS improves glycemic control, self-efficacy, SMBG and healthy eating while
decreasing diabetes-related distress and complications (Sherifali, Rabi, et al., 2018, p. S36).
Weight management.
An increased body-mass index (BMI) and abdominal adiposity impairs glycemic control,
contributes to hypertension and dyslipidemia, and increases the risk of diabetic complications
(Wharton, Pedersen, Lau, & Sharma, 2018). An estimated 80 to 90% of individuals with DM 2
are overweight (Wharton et al., 2018), but substantial improvements in insulin sensitivity,
cellular glucose uptake, and hepatic glucose regulation are noted with as little as 5% to 10% loss
of initial body weight (Wharton et al., 2018). The DC guidelines recommend a combination of
physical activity and nutritional interventions to target a minimum weight loss of 5 to 10% in
individuals who are overweight (Wharton et al., 2018). Addressing barriers to weight loss
including depression, food behaviors or physical impediments to exercise is key to creating a
successful weight loss strategy (Wharton et al., 2018).
Nutrition.
Healthy dietary strategies can significantly lower an A1C approximately 1 to 2%
(Sievenpiper et al., 2018). The DC guidelines endorses healthy food choices from 4 major food
groups, based on Eating Well with Canada’s Food Guide: (1) fruits and vegetables, (2) grains,
(3) milk and alternatives, and (4) meat and alternatives (Health Canada, 2007; Sievenpiper et al.,

12

2018). Meals should be a mix of macronutrients and distribution should be flexible to allow for
individual preferences (Sievenpiper et al., 2018)
Moderate and consistent intake of carbohydrates (CHO) comprised of varied, low
glycemic, high fiber and whole food options are key in lowering A1C levels, controlling
dysglycemia, managing weight, controlling lipids, improving clinical and metabolic outcomes
and preventing micro and macrovascular complications such as cardiovascular events
(Sievenpiper, Chan, Dworatzek, Freeze, & Williams, 2018, p. S64). Fruits, vegetables, legumes
and whole grains are preferred choices (Sievenpiper et al., 2018). Maintaining CHO at 45% to
60% of daily intake appears to demonstrate the greatest benefit in glycemic control and the
prevention of diabetic complications (Sievenpiper et al., 2018). Added sugars, in particular highfructose corn syrup, sucrose and fructose at greater than 10% of daily intake, should be avoided
as much as possible due to their impact on fasting blood glucose levels and triglycerides
(Sievenpiper et al., 2018).
Preferred selections exist for several non-CHO macronutrients: Unsaturated fat is
recommended over saturated fat to reduce risk of cardiovascular events (Sievenpiper et al.,
2018); Vegetable protein is recommended over animal protein, as vegetable protein substitutions
have demonstrated reduced A1C levels, fasting plasma glucose (FPG) and improved fasting
insulin (Viguiliouk et al., 2015).
Other dietary strategies include: (a) reducing caloric intake and increasing daily protein
intake to target weight loss and increase satiety (Sievenpiper et al., 2018); (b) carbohydrate
counting to regulate the timing, constancy and quantity of CHO (Sievenpiper et al., 2018); (c)
limiting sugar-sweetened foods and beverages (SSFB) to less than 10% of total daily intake to
maintain weight and glycemic control (Sievenpiper et al., 2018); (d) minimizing refined sugars

13

and fast foods due to dysglycemia, weight gain and diabetic complications (Sievenpiper et al.,
2018); and (e) regular consumption of legumes, beans and lentils to lower A1C, promote
glycemic control and reduce various serum markers associated with cardiovascular disease
(Sievenpiper et al., 2018). Dietary patterns most closely aligning with these recommendations
include the Mediterranean diet, DASH diet and a vegetarian diet (Sievenpiper et al., 2018).
Consideration to values, preferences and goals helps a person adhere to dietary strategies over
the long term (Sievenpiper et al., 2018).
Research examining the effects of intermittent fasting and energy restriction on weight
loss has become prevalent in the general population, but its effects in on glycemic control is
limited. In a recent pilot trial, Carter, Clifton and Keogh (2016) compared the effects of
intermittent severe energy restriction versus a moderate continuous calorie restricted weight loss
diet and found glycemic control and weight reduction were similar (Carter, Clifton, & Keogh,
2016). However, the individuals who severely restricted their intake required significant
medication adjustments or discontinuation, particularly with insulin and sulfonylureas, to prevent
hypoglycemic events (Carter et al., 2016). De-prescribed sulfonylureas produced inconsistent
results, with some participants experiencing rebound hyperglycemia (Carter et al., 2016).
Diabetes SMS and SME should work to identify and mediate barriers to healthy food behaviors
(Sievenpiper et al., 2018) for individuals experiencing involuntary restriction of nutritional
intake.
Physical activity.
Aerobic exercise, i.e. continuous movement of large muscle groups for a minimum of 10
minutes (Sigal et al., 2018, p. S55) and resistance training, i.e. repetitive weighted movements to
increase muscle strength or mass (Sigal et al., 2018, p. 55) assist in the regulation of blood

14

glucose, insulin resistance, lipid regulation, blood pressure, weight loss and reduction of
macrovascular complications (Sigal et al., 2018). The DC guidelines recommend at least 150
minutes of aerobic activity a week, a minimum of 2 days per week of resistance exercise (Sigal
et al., 2018), and physical movement every 20 to 30 minutes (Sigal et al., 2018).
Pharmacological management.
The guidelines recommend that pharmacological management be commenced when nonpharmacological options are insufficient in reducing the A1C to target levels within the first 3 to
6 months, or if the individual’s A1C is greater than 8.5% at diagnosis (Lipscombe et al., 2018).
Metformin is preferred as first line treatment in the pharmacological management of DM 2
because it is safe, inexpensive, effective at reducing A1Cs, and is associated with a low risk of
hypoglycemia, weight gain and adverse effects (Lipscombe et al., 2018). Metformin is a
biguanide that decreases hepatic glucose production and improves hepatic insulin sensitivity by
activating AMP-activated protein kinase (Lipscombe et al., 2018). However, when symptomatic
hyperglycemia or signs of metabolic decompensation such as weight loss or ketosis are present at
diagnosis, insulin should be initiated, irrespective of A1C (Lipscombe et al., 2018).
Second line agents are added to metformin when: (a) initial A1C or symptomatology at
diagnosis indicate the need for more aggressive therapy or (b) the A1C remains above the
patient-specific target after 3-6 months of monotherapy (Lipscombe et al., 2018). The DC
guidelines recommend adding an agent every 3 to 6 months until target A1C levels are reached
(Lipscombe et al., 2018). Traditionally, insulins and insulin secretagogues, such as sulfonylureas,
have been first-line treatment for add-on therapy. However, the risks of hypoglycemia and
weight gain are greatest in these classes and should not be the first choice in populations at
increased risk of hypoglycemia (Lipscombe et al., 2018).

15

Over the last decade, newer pharmacological options have become available, meaning
there is greater flexibility in individualizing treatment plans to the needs of a patient (Diabetes
Canada, 2017b). Further, these newer agents have fewer side effects (Diabetes Canada, 2017b),
provide superior glucose control and A1C stability (Diabetes Canada, 2017b), and have a
superior safety profile (Lipscombe et al., 2018). Approach and selection of antihyperglycemic
therapy should consider clinical presentation, patient risk factors, values and preferences in order
to provide a treatment regimen most appropriate for the individual (Lipscombe et al., 2018). Key
to this process includes understanding the individual’s disposition, their social needs and their
ability to afford or observe a particular pharmacological regimen (Lipscombe et al., 2018).
Treatment should be promptly modified when adherence issues are present (Lipscombe et al.,
2018).
The DC guidelines recommend dipeptidyl peptidase-4 (DPP-4) inhibitors, glucagon-like
peptide-1 (GLP-1) receptor agonists or sodium-glucose linked transporter-2 (SGLT-2) inhibitors
over traditional agents due to their negligible risk of hypoglycemia and weight gain, and noninferior A1C-lowering capacity (Lipscombe et al., 2018). Additionally, select GLP-1 agonists
and SGLT-2 inhibitors are associated with reduced microvascular and cardiovascular
complications in individuals with identified cardiovascular risks (Lipscombe et al., 2018).
Insulins.
Insulins are an exogenous source of insulin, and when injected, stimulate insulin cell
receptors to normalize the metabolism of carbohydrate, fat, and protein (Lipscombe et al., 2018)
through cellular uptake. Insulins offer the greatest potential for a reduction in A1C; however, all
pose some risk of hypoglycemia when food intake is inconsistent or reduced (Lipscombe et al.,
2018).

16

When commencing insulins, long-acting or intermediate-acting insulins are generally
commenced first, and given at night to offer basal or glycemic control and improve fasting
morning blood sugars (Lipscombe et al., 2018). Long-acting insulins are preferred over
intermediate-acting insulins because they do not have a peak effect, i.e. there is a decreased risk
of symptomatic or nocturnal hypoglycemia (Lipscombe et al., 2018). When bolus or mealtime
insulin is commenced, sulfonylureas should be de-prescribed due to concomitant risks of
hypoglycemia (Lipscombe et al., 2018). Rapid acting insulins pose less risk of postprandial
hypoglycemia than short-acting insulins and should be preferentially prescribed wherever
affordable and appropriate (Lipscombe et al., 2018). Concomitant prescribing of other
antihyperglycemics is encouraged: Regimens consisting of combination insulin and non-insulin
agents reduce the level of weight gain and risks of hypoglycemia associated with insulin only
regimens (Lipscombe et al., 2018).
Sulfonylureas.
Sulfonylureas are an insulin secretagogue which acts on β-cell sulfonylurea receptors and
stimulates the release of endogenous insulin (Lipscombe et al., 2018). Sulfonylurea work fairly
rapidly and do not work in a glucose-dependent fashion, meaning they can pose a moderate risk
of hypoglycemia (Lipscombe et al., 2018) especially with reduced glucose intake. Certain
sulfonylureas are more problematic over others (Lipscombe et al., 2018). Glyburide offers the
greatest risk of hypoglycemia; however, prescribing habits are likely influenced by Pharmacare
coverage, as glyburide remains the only unrestricted sulfonylurea in the public formulary
(Government of Canada, 2018b). Gliclazide poses the least risk of hypoglycemia, cardiovascular
events and overall mortality and is the preferred choice of sulfonylurea (Lipscombe et al., 2018).

17

Weight gain is a common side effect due to their insulin-releasing properties (Lipscombe et al.,
2018).
Dipeptidyl peptidase-4 (DPP-4) inhibitors and glucagon-like peptide-1 (GLP-1) receptor
agonists.
Incretins exert their effect only when glucose is ingested, meaning they possess a near
negligible risk of hypoglycemia. Both DPP-4 inhibitors and GLP-1 receptor agonists increase the
signalling and release of insulin when glucose is sensed in the gastro-intestinal tract. They also
slow gastric emptying and inhibit glucagon release (Lipscombe et al., 2018). All GLP-1 receptor
agonists are associated with weight loss; however, additional weight loss is seen at higher doses
of liraglutide (Wharton et al., 2018). Liraglutide and semaglutide are the only cardioprotective
GLP-1 receptor agonists (Lipscombe et al., 2018). Pharmacological complications are rare,
however, gastrointestinal side effects such as nausea, or diarrhea, are common especially in
GLP-1 receptor agonists (Lipscombe et al., 2018).
Sodium-glucose linked transporter-2 (SGLT-2) inhibitors.
SGLT-2 inhibitors block the sodium-dependent glucose cotransporters responsible for
renal glucose reabsorption back into the bloodstream (Lipscombe et al., 2018). The net action is
increased glucosuria and decreased blood-glucose levels (Lipscombe et al., 2018). SGLT-2
inhibitors possess nearly a negligible risk of hypoglycemia, and a potential reduction of 2 to 3 kg
of body weight (Lipscombe et al., 2018). This class’ diuretic-like properties offer an additional
benefit of a reduction in systolic blood pressure, and a reduced or slowed progression of
nephropathy (Lipscombe et al., 2018). In individuals with known cardiovascular disease, they are
associated with: (a) reduced heart failure; (b) with empagliflozin and canagliflozin only, a
reduction in major adverse cardiovascular events; and (c) with empagliflozin only, a decrease in

18

cardiovascular death (Lipscombe et al., 2018). There is a significant increase in the occurrence of
genital mycotic and urinary tract infections, hypotension and volume depletion (Lipscombe et
al., 2018). There are some adverse outcomes attributed to canagliflozin in particular, which
include increased fractures, lower extremity amputations and acute renal injury (Lipscombe et
al., 2018).
Thiazolidinediones (TZDs).
TZDs target insulin resistance by increasing peripheral tissue and hepatic insulin
sensitivity in a variety of cell receptors (Lipscombe et al., 2018). Subsequently they are
associated with weight gain, but a negligible risk of hypoglycemia (Lipscombe et al., 2018). The
use of TZDs has substantially diminished over the past decade due to an increased incidence of
edema, heart failure, significant risk of fractures, and possible risk of myocardial infarctions with
rosiglitazone (Lipscombe et al., 2018). Pioglitazone remains the only TZD available in B.C.
(Government of British Columbia, 2018b).
Alpha glucosidase inhibitors (acarbose).
Acarbose works by preventing pancreatic α-amylase and intestinal α-glucosidase from
breaking down and absorbing carbohydrates (Lipscombe et al., 2018), which reduces and delays
the absorption of glucose into the bloodstream. Gastrointestinal side effects such as bloating, gas
and diarrhea are common and require three times a day dosing (Lipscombe et al., 2018). B.C. is
the only province where Acarbose is de-listed and no longer available (Diabetes Canada, 2017b).
Orlistat.
Orlistat is used as a weight loss agent by preventing lipase from breaking down fats.
By assisting with weight loss, Orlistat improves glycemic control and subsequently results in
dose reduction of antihyperglycemic medications (Wharton et al., 2018). It is associated with

19

gastrointestinal side effects such as nausea and diarrhea. However, due to its mechanism of
action, it produces a negligible risk of hypoglycemia (Lipscombe et al., 2018).
Pharmacare
Pharmacare is British Columbia’s public drug insurance program. Pharmacare is
comprised of several publicly-funded drug plans for whom select individuals are eligible
depending on each plan’s criteria (Government of British Columbia, 2018a). Select populations
have 100% of the drugs in the Pharmacare’s formulary covered. This includes permanent
residents of licensed residential care facilities (Plan B), recipients of B.C. income assistance
(Plan C) and First Nations (Plan W). Fair Pharmacare is B.C’ s only public drug plan eligible to
all B.C. residents with active Medical Services Plan of B.C. (MSP). Fair Pharmacare is
dependent upon household income (Government of British Columbia, 2018a). An annual
minimum deductible must be reached before 70% of the drug cost will be covered, and a family
maximum deductible must be reached before drugs are 100% covered (Government of British
Columbia, 2018a). While Plan C offers 100% coverage to British Columbians living below the
poverty line and who are receiving income assistance, many Canadians not receiving income
assistance live below the poverty line and must spend money before they receive assistance from
Pharmacare. For example, a family of four in B.C. living on wages equivalent to the average
poverty line in B.C1, must spend approximately $950.00-$1100.00 to reach their deductible and
$1350-1500.00 to have medications fully covered (Government of British Columbia, 2018c).
While eligibility and deductibles vary from person to person, the list of
antihyperglycemics available through Pharmacare’s formulary is the same across all plans:

Calculated using Statistics Canada’s (2018a) estimated Market Basket Measure (MBM) for a family of four living
in B.C. ($37,153- $39,951 depending on rural versus urban living). Canada has recently structured and developed an
official poverty line called the Market Basket Measure (MBM) and is currently being validated throughout Canada
(Statistics Canada, 2018b).
1

20

Metformin, glyburide and select short and intermediate-acting insulins are listed restriction-free
on the formulary (Government of Canada, 2018b). Rapid-acting insulins are available to
individuals capable of paying the cost difference between short and rapid-acting insulins
(Government of Canada, 2018b).
Presently, B.C. remains the only province where none of the newer more expensive
antihyperglycemic medications such as SGLT-2 inhibitors, GLP-1 agonists, or DPP-4 inhibitors
are listed on the public formulary without Special Authority (Diabetes Canada, 2017b). Further,
Special Authority (SA) coverage only exists for linagliptin and saxagliptin2 when maximum
tolerated doses of metformin and sulfonylureas have been tried and insulin NPH was found to be
ineffective or inappropriate (Government of British Columbia, 2018d). Subsequently primary
care providers have had to find creative strategies to provide the pharmacotherapy most suitable
for the patient while working in conjunction with Pharmacare and patient income.
Therefore, individuals with financial constraints and without additional private coverage
are more likely to rely on Pharmacare-covered medications due to the prohibitive costs of newer
medications (Diabetes Canada, 2017b). They may experience significant restrictions in
pharmacological options, subsequently resulting in greater difficulty and risks in managing his or
her DM 2. Lastly, Pharmacare funds blood glucose strips solely based on pharmacological
regimen (Government of British Columbia, 2018b), which presents a challenge to individuals
who cannot afford additional strips, possess additional challenges in monitoring their blood
glucose, and are at increased risk for dysglycemia. Under special circumstances, specific
healthcare providers may apply for additional strips if the patient is not on insulin and: (a) is not
meeting target glycemic levels after 3 months, (b) has present or concomitant illnesses which

2

Includes combinations with metformin

21

interfere with glycemic control, (c) is undergoing medication changes which affect glycemic
control, (d) has a vocation where dysglycemia presents a significant safety risk, or (e) the patient
is a gestational diabetic (Government of Canada, 2019). If SA is approved, individuals are
eligible for 100 extra strips annually.
Overview of Food Insecurity (FI)
FI exists when individuals have uncertain access to adequate amounts of nutritious food
(Essien et al., 2016; Montgomery et al., 2017; Seligman et al., 2012) It is a worldwide epidemic,
reaching all corners of society (WHO, 2018). Several Canadian studies have identified poverty
and financial constraints as antecedents to FI in Canada (Beryl Pilkington et al., 2010; Chan,
DeMelo, Gingras, & Gucciardi, 2015; Li, Dachner, & Tarasuk, 2016; Tarasuk et al., 2016;
Thomas, Fitzpatrick, Sidani, & Gucciardi, 2018). In Canada, the prevalence and severity of FI
does not exist exclusively among the impoverished (Patil et al., 2017): FI increases as income
decreases in a near linear fashion (Li et al., 2016). The “Research to Identify Policy Options to
Reduce Food Insecurity” (PROOF) is funded by the federal Canadian government to monitor the
prevalence of income-associated food insecurity (IAFI) in Canada (2018), and defined FI as “the
inadequate or insecure access to adequate food due to financial constraints (Tarasuk et al., 2016,
p. 6)”. Since 2005, FI has continued to increase in Canada and in British Columbia (Li et al.,
2016; Tarasuk et al., 2016). Approximately 1 in 8 households in Canada are classified as “food
insecure”, which accounts for approximately 4 million Canadians (Tarasuk, Mitchell & Dachner,
2014); In B.C. approximately 1 in 10 households, or half a million British Columbians, are food
insecure (Li et al., 2016). It remains unclear, however, what the prevalence rates and causes of
other forms of FI in Canada are, but it is speculated to encompass significantly more of the
Canadian population. Examples of other documented causes of FI include physical barriers to

22

access, geography, inability to safely prepare food, labile food prices, and variability to food
crops (Strickhouser et al., 2014). Further, some populations appear to be at greater risk than the
general population, including women, single parents, elderly and immigrants (Gucciardi et al.,
2014).
FI affects health, though the relationship is complex and not necessarily causational
(Berkowitz & Fabreau, 2015). FI impairs the prevention, treatment and management of several
chronic diseases (Tarasuk et al., 2015), Though the mechanisms are not entirely understood, FI is
independently associated with increased healthcare utilization costs, and both poorer physical
and mental health (Tarasuk et al., 2015). FI interferes with self-care and functional health, and
overall worse chronic disease outcomes (Tarasuk et al., 2015).
Food insecurity in type 2 diabetes.
The prevalence of food insecurity (FI) among Canadians with DM 2 occurs at
disproportionate rates (Chan et al., 2015; Galesloot, McIntyre, Fenton, & Tyminski, 2012;
Gucciardi et al., 2009). FI in individuals with DM 2 is associated with inferior glycemic control,
and worsening A1C values (Ippolito et al., 2016). There is evidence that FI interferes with DM 2
self-management strategies (Ippolito et al., 2016), but the mechanism is not yet understood.
Diabetic complications also occur at significantly higher rates (Montgomery et al., 2017;
Seligman et al., 2012). There are no best-practice strategies for the management of individuals
with DM 2 experiencing FI (Thomas et al., 2018), therefore the interface between FI and
successful management of DM 2 remains unclear and is at the epicentre of this integrative
review. Thus, an integrative review exploring how FI interferes with the management of DM 2 is
necessary before an effective management approach can be established for the food-insecure

23

diabetic. Additionally, further exploration of the literature is required to examine how FI may
best be identified in primary care.
Identifying food insecurity.
Presently, there is no standardized process or method for identifying all forms and
measures of FI (Jones, Ngure, Pelto, & Young, 2013), nor is there a specific method for
identifying FI amongst individuals with DM 2 (Thomas et al., 2018). Identifying food insecurity
can be challenging for several reasons. Firstly, people who are food insecure do not represent a
homogenous group (Strickhouser, Wright, & Donley, 2014). For example, based on Canadian
data that monitors IAFI, 65% of the FIFC in British Columbia are employed (Li et al., 2016),
16% live on social assistance, 12 % are seniors relying on seniors’ income, and the remaining
rely on Employment Insurance, workers compensation or other sources as their main source of
income (Li et al., 2016). Unattached individuals and individuals with no children represent over
40% of IAFI in B.C., but families with children and women have higher rates of FI overall (Li et
al., 2016). Caucasians, Asians, and other ethnic groups in B.C. share similar rates of FI, but
nearly a third of B.C.’s IAFI self-identify as Aboriginal (Li et al., 2016).
Secondly, FI screening is not routinely completed or well integrated into practice. There
is no systematic evidence-based method for determining how FI screening should unfold in
healthcare settings. In a recent systematic review examining the validity, acceptability and
implementation of FI screening in healthcare, De Marchis, Torres, Fichtenberg and Gottlieb
(2019) found that there is limited direction in healthcare settings on how to implement FI
screening in practice.
Thirdly, while De Marchis et al., (2019) identified several FI screening tools validated in
US healthcare settings, it is unclear whether Canadian providers have access to peer-reviewed,

24

standardized and validated tools to identify FI in the Canadian healthcare context. Furthermore,
it is unclear how FI is defined in each screening tool: in a recent systematic review examining
available tools for measuring FI, Ashby, Kleve, McKechnie and Palermo, (2016) found that most
current tools were subjective in nature, had a limited scope in identifying FI and primarily
identified access issues in FI. Identifying a standardized FI screening tool validated for use in
Canadian healthcare is an essential aspect of this integrative review.
In Canada, IAFI has been routinely monitored through research grants by Policy Options
to Reduce Food Insecurity (PROOF, 2018b), using a validated and standardized measurement
called the Household Food Security Survey Module (HFSSM). The HFSSM is part of the
Canadian Community Health Survey (CCHS), which Statistics Canada uses to collect health
information on a sample of Canadians every two years (PROOF, 2018b). While the HFSSM was
originally developed and used in the Unites States (US) by the US Department of Agriculture, it
was later adapted and approved by Health Canada for its use in Canada (Health Canada Office of
Nutrition Policy and Promotion 2007). It remains unclear if the Canadian-specific HFSSM has
been validated for use in healthcare settings to identify FI at the individual level or whether it has
been validated to identify other forms of FI.
Benefits of screening food insecurity in practice.
Clinicians have questioned whether routine screening for FI in clinical settings should be
completed (Berkowitz & Fabreau, 2015). There is little evidence that screening and identifying
food insecurity improves health outcomes (Berkowitz & Fabreau, 2015), but screening is
supported by some clinical experts when it results in adjustments to clinical management and
improves safety (Berkowitz & Fabreau, 2015). For example, Berkowitz and Fabreau (2015)
indicate that identifying FI in healthcare is clinically significant for individuals with DM 2,

25

because it can directly support important pharmacological and clinical decisions when food
behaviors may be causing dysglycemia. In this circumstance, identifying FI assists the provider
with specific and appropriate SME and SMS, including self-titration of medication during low
food periods (Berkowitz & Fabreau, 2015), or de-prescribing specific medications which pose a
specific risk to the individual.
Maslow’s Theory of Motivation (MTM)
It is important to consider an individual’s capacity to be motivated to make health
behavior changes and self-actualize in the face of long-standing FI. Maslow’s Theory of
Motivation (MTM) was developed in the 1940’s, and describes motivation for human behavior
(Maslow, 1943). While our understanding of human motivation and behavior has evolved over
time, MTM continues to be influential in practice and offers insight to human behaviors with
respect to health behaviors. Maslow’s theory (1943) is based on 5 basic tenets: (1) We have five
sets of basic needs, namely physiological, safety, love, esteem and self-actualization. Humans
are motivated by a desire to satisfy these needs (Maslow, 1943, p. 394); (2) These 5 needs exist
in a hierarchical, but not mutually exclusive manner: when the more basic needs are unmet, they
dominate the conscious mind, while other ‘higher’ needs become irrelevant, minimized and
sequestered from awareness (Maslow, 1943, p. 394). Only when a more basic need is adequately
satisfied, can a higher need begin to dominate an individual’s consciousness (Maslow, 1943, p.
394); (3) Any real or perceived threat to the security of these needs is perceived as a
psychological threat (Maslow, 1943, p. 394); (4) these threats generate emergency or
compensatory reactions in behavior (Maslow, 1943, p. 394); (5) some motivators of human
behavior are not yet or cannot be explained by this model (Maslow, 1943, p. 394).

26

MTM postulates that food is one of the most basic physiological human needs (Maslow,
1943); Therefore, MTM may provide further insight into why individuals with FI may struggle
with managing DM 2. Further, MTM may highlight how recommending self-management
approaches in DM 2 when FI is present may be inadequate and even inappropriate. Lastly, MTM
may highlight the enhanced value of individualized, FI-informed management strategies when
patients are significantly limited in their capacity to self-manage their DM 2.
The Nurse Practitioner (NP)
In Canada, NPs are graduate level, master degree prepared advanced practice nurses
(BCCNP, 2018a) working within a legislated scope of practice under the B.C. Health Professions
Act (RSBC, 1996), and regulated by the British Columbia College of Nursing Professionals
(BCCNP, 2018a). The NP scope includes diagnosing, ordering and interpreting diagnostic tests,
prescribing pharmaceuticals, and providing holistic health management, health promotion and
disease prevention strategies (BCCNP, 2018a).
Nurse Practitioners (NPs) in primary care.
NPs are trained as family nurse practitioner in B.C. As of 2018, over 500 NPs are
registered in B.C, and approximately 60% work in primary care (Nurse and Nurse Practitioners
of British Columbia [NNPBC], 2018) offering unique strategies in addressing barriers to patient
health. Fundamental to this capacity is the NP’s ability to grasp crucial aspects of a patient’s
perspective, motivations, issues, barriers and goals in order to effectively manage health (Bartol,
2012). NPs are specifically educated to identify difficulties in making, adopting and adhering to
lifestyle changes and can partner with patients, families, communities and organizations to
implement useful techniques (BCCNP, 2018a).

27

A recent evaluation of the integration of NPs into the B.C. healthcare system revealed
that the majority of NPs are providing direct patient care in office settings for chronic conditions,
including diabetes (Sangster-Gormley et al., 2015). Many of these patients have complex health
conditions and several social issues (Sangster-Gormley et al., 2015). The review of data from
B.C.’s NP demonstrate NPs are spending the majority of time managing these chronic health
conditions by screening, providing pharmacotherapy, engaging in health education, and
integrating preventive health approaches (Sangster-Gormley et al., 2015).
NPs are seen as an added benefit because they increase access to care, are effective at
enacting behavior change, use a collaborative, comprehensive, team-based and patient-centered
approach and deliver high levels of patient satisfaction (Sangster-Gormley et al., 2015). In B.C.,
NPs are delivering a high-quality care with patient outcomes similar to that of general
practitioners (Prodan-Bhalla & Scott, 2016). The NP practice approach could facilitate the
implementation of formal FI screening and modified patient-centered treatment strategies in
practice. This integrative literature review will contribute to the growing interest in NPs and their
unique contributions as PCPs by exploring how NPs can identify FI in the PC setting to enhance
the management of individuals with DM 2 experiencing FI.

28

Chapter 3
Methods
As part of my prior scholarly work, an extensive initial literature search was executed to
examine the best evidence and recommendations for the dietetic management of type 2 diabetes
(DM 2). Several practice guidelines, systematic reviews and meta analyses were found
describing nutritional recommendations for the management for DM 2. A follow-up literature
search examined how food insecurity (FI) influenced a person’s capacity to follow evidencedbased dietary strategies. However, there was a lack of specific literature guiding primary care
providers (PCPs) in identifying and supporting individuals with DM 2 experiencing FI. Further,
while it appeared clear that FI had a profound effect on glycemic control and subsequent
complications, the effect of FI on food only partly explained the relationship. Other mechanisms
of influence remained unclear. In order for providers to aid in the management of DM 2 amongst
individuals experiencing FI, all mechanisms of influence must be understood and strategies to
identify FI, elicited.
Through these prior findings, the following question was produced: “What strategies can
nurse practitioners employ to identify FI in primary care to enhance the management of DM 2
amongst individuals experiencing FI?” An integrative review methodology (IRM) was selected
since IRMs identify and amalgamate a vast array of literature to provide greater insight into
clinical questions (Whittemore & Knalf, 2005). Whittemore and Knalf’s (2005) approach to an
integrative review was used to guide this paper.
Literature Search
A systematic approach was employed, using databases to search available literature.
Afterwards, an additional search strategy was conducted for individual selection of key articles.

29

Inclusion and exclusion criteria were established that guaranteed a breadth of articles were
selected that also focused the literature relevant to the topic. A descriptive analysis of the
literature search strategy is provided followed by data evaluation and analysis using Whitemore
and Knalf’s (2005) approach to integrative reviews.
Preliminary Search Strategies
An initial scoping of the literature was conducted in PubMed to review pertinent
keywords and medical subject heading (MeSH) terms used in articles discussing DM 2, FI and FI
screening methodologies. Terms frequently used in the articles from prior scholarly work were
also examined. Terms returning irrelevant literature were first cross-examined against other
popular terms to ensure no relevant articles were missed. Some duplicate articles arose, but no
new articles; therefore, terms returning a high volume of irrelevant literature were eliminated.
Scope notes were examined for each MeSH heading. Reviewing keywords within MeSH terms
and scope notes ensured terms were accurate, relevant to the project aim, suitable for the search
strategy and not limiting to the search strategy by erroneous keyword selection or omission. A
concept map with important keywords was used to guide the literature search strategy for each
database. The concept map is presented in Diagram 1.

30

Diagram 1
Concept Map Driven by PubMed MeSH and Keywords3

Food Securit*
Food Insecurit*
Food Supplies
Food Supply

Adult-onset Diabetes
Mellitus
Type 2 Diabetes
Non-insulin dependent
diabetes mellitus
NIDDM

MeSH: Food Supply

MeSH: Diabetes Mellitus,
Type 2
Research Question
MeSH: Patient Health
Questionnaire AND Mass
Screening

Primary Search Strategies

Mass Screenings
Screening, Mass
Screenings, Mass
Screening
Screenings
Questionnaires
Survey
Medical questionnaire

For each aim of the integrative review, a combination of MeSH terms and keywords were
used. Their appropriateness was verified for accuracy by comparing the scope note and crossexamining the terms and notes against the PubMed terms initially established. For each database,
missing terms under the MeSH heading were added as keyword search terms. New terms
discovered under MeSH headings were independently searched to establish if additional relevant
articles were returned. If no extra articles were found, the term was eliminated. Combining
search strategies for all aims returned zero articles. Therefore, the search strategy was divided as
follows: For aim 1 and 2, the Cumulative Index of Nursing and Allied Health.

3

“*” indicates word is truncated and includes all possible endings to the truncated word.

31

Literature (CINAHL), Medline (Ovid), PubMed, Cochrane Databases (EBMR), and Academic
Search Complete (ASC) were searched. For aims 1 and 2, the following search terms were used
in each database and verified with the university librarian. They are presented in Table 1.
Table 1
Search Terms Utilized in Specified Databases to Extract Articles for Aim 1 and 2
CINAHL

Medline OVID

PubMed

EBMR

ASC

MeSH terms

MeSH terms

MeSH terms

MeSH terms

Food Security
Diabetes
Mellitus, Type 2

Food Supply
Diabetes
Mellitus, Type 2

Food Supply
Diabetes Mellitus,
Type 2

No MeSH terms
available

Keywords

Keywords

No Keywords

Keywords

No keywords

Food Insecurity

Food Security
Non-insulin
Dependent
Diabetes

Food securit*
Food Insecurit*
Food Supply
Adult-onset
Diabetes
Mellitus
Type 2 Diabetes
Non-insulin
Dependent
Diabetes
mellitus
NIDDM
Diabetes
Mellitus Type 2

For aim number 3, PubMed, ASC and Web of Science (WOS) were used, following a
discussion with the university librarian. Several adjustments were made in ASC due to the
number of results returned, bearing no relevance to the review. Appropriate terms were identified
searching MeSH headings and comparing scope notes with PubMed. For aim number 3, the
results for the search strategy are presented in Table 2.

32

Table 2
Search Terms Utilized in Specified Databases to Extract Articles for Aim 3
PubMed
MeSH terms

WOS
No MeSH terms

ASC
MeSH terms

Food supply
Patient health questionnaire
Mass Screening
No Keywords

Food security
Medical Screening
Patient Screening
Questionnaire
Keywords

No Keywords

Patient health questionnaire
Screening
Questionnaire
Food security
Food insecurity
Terms were then combined with the Boolean operators “AND” and “OR” when
appropriate. The search strategy for each database was verified with the university librarian and
is presented in Table 3.
Table 3
Sequential Search Strategy Utilized for the Database Searches
Search

Search Terms-Keyword and MeSH

#1

“Food Security” OR “Food Insecurity” OR “Food Supply”

#2

“Diabetes mellitus, Type 2” OR “Non-insulin dependent diabetes” OR
“NIDDM” OR “Adult-onset Diabetes Mellitus” OR “Type 2 Diabetes”

#3

#1 AND #2

#4

“Patient Health Questionnaire” OR “Mass screening” OR “Medical
Screening” OR “Patient Screening” OR “Screening” OR “Questionnaire”

#5

#1 AND #4

33

Inclusion and exclusion criteria.
Inclusion and exclusion criteria assisted in isolating and identifying relevant literature
that addressed: (a) methods available to identify FI, and (b) how FI influences the management
of DM 2. Inclusion and exclusion criteria are presented in Table 4.
Table 4
Inclusion and Exclusion Eligibility Criteria for the Selected Literature
Inclusion Criteria
The study was published in the last 10 years

Exclusion Criteria
The study addressed age-specific or culturalspecific populations

The study critically examined questionnaires,
screening tools & surveys that identify FI

The study did not critically assess the FI
screening method

Study examined the influence of FI on type 2
diabetes

The study examined the influence of FI on the
development of DM 2

The study contained any definition or concept
of food insecurity (including individual and
household food insecurity)

The study exclusively focused on screening
and identifying food insecure pediatric
patients

Study was in English

Study was in a language other than English

Articles included in this review examined either one of the following: (a) the influence of
FI on the management of DM 2 or (b) methods of identifying FI. Articles examining methods of
identifying FI were not restricted to individuals with DM 2, since there was a paucity of
literature noted during preliminary search strategies when terms were combined. Furthermore,
there was a dearth of literature examining the identification of FI in primary care in Canada.
Therefore, the search strategy was expanded to include all methods and environments where
food insecurity was identified. Similarly, preliminary search strategies revealed that methods for
identifying FI were varied and often contingent upon the study’s specific interest in FI. For
example, some studies examined food insecurity secondary to financial constraints using tools

34

only validated to identify the same. Therefore, all tools, methods, questionnaires and surveys
were considered.
While some consensus on the definition of food insecurity has been established
internationally (WHO, 2018), there is no established consensus in the literature and definitions
remain quite varied. Subsequently, no limitations were placed on the basis of definitions used.
Furthermore, part of the review aimed to conceptualize the definition of food insecurity in the
context of DM 2, therefore a breadth of literature examining the relationship of FI and DM 2 was
sought. No limitation was placed on age as DM 2 is almost exclusively adult-onset (Brasher,
Jones & Huether, 2014b). For this reason, any study examining methods of identifying FI among
pediatric populations were excluded. To avoid difficulty in generalizing and applying results to a
Canadian context, articles discussing either DM 2 or FI within a specific cultural or social
context were eliminated. The relationship between DM 2 and FI is complex (Essien et al., 2016),
interacting at various levels of disease. In order to extract data pertaining to the influence of FI
on the management of DM 2, articles examining the risk of developing DM 2 with FI were
excluded.
Only articles published in English and within the last 10 years were included. Our
understanding of DM 2 has significantly changed over the last 20 years (Defronzo, 2009).
Subsequently, management strategies have also changed, especially over the last decade
(Diabetes Canada, 2017b; Lipscombe et al., 2018). Literature was limited to reflect our current
understanding and management of DM 2.
Secondary Search Strategies
The cumulative search returned a total of 532 articles following the exclusion of articles
related to the date and language exclusion criteria. Following the removal of all duplicate

35

articles, a total of 411 remained. Next, each title and abstract were screened to determine their
relevancy to the review, using the eligibility criteria from Table 4. This process yielded 55
articles that were eligible for full text review.
Several hand searches were completed. Firstly, Pubmed database was searched by
observing “related searches” in the side bar. Next, the UNBC library general search database was
searched using various combinations of the keywords from Tables 1 and 2. Lastly, google
scholar was searched using the same method as with the UNBC library database. The cumulative
hand search efforts yielded three more articles.
A review of the references of the remaining articles was completed to establish whether
any relevant articles were missed in the database search. The reference search did not elicit any
additional articles. The remaining 60 full articles were carefully examined using the full
inclusion and exclusion criteria. Further, the references of review articles were screened to
ensure that articles included in the full text review had not been previously included, evaluated
and summarized in one of the extracted review articles. Fifteen articles in total were extracted for
the literature review. Two search diagrams are presented in Appendix A. Combined, they
represent the results of this literature search.
Data Evaluation and Analysis
A review matrix was used to assess and classify data from the final literature set of
articles. An extensive appraisal of literature is required for a robust integrative literature review
(Whitemore & Knalf, 2005). The Critical Appraisal Skills Programme (CASP) systematic
review, case control, and qualitative study checklists were utilized to evaluate the quality of the
literature (CASP, 2017). As there was no cross-sectional analysis checklist, the case control
checklist was utilized in its stead to evaluate cross-sectional studies. Each article was evaluated

36

and assigned a quality rating using the guidelines and prompts provided by the CASP tools. The
literature review matrix and ratings are found in Appendix B.
In order to guarantee bias is not introduced by the evaluator, data analysis requires both a
synthesis of literature and systematic review (Whitemore & Knalf, 2005). The first step in
Whitemore and Knalf’s (2005) data analysis is data reduction. Data reduction was accomplished
in this literature review by classifying the methodologies used in each article.
Details extracted from the literature review matrix were included under the headings: “aim”,
“study design and sampling”, “methods and analysis” and “strengths and limitations”. Findings
particular to each study aim were displayed in a separate column of the review matrix titled
“findings”. Findings included in the matrix were: (a) how FI is defined in the context of DM2,
(b) how FI influences the management of DM 2, and (c) methods used to identify FI. Whitemore
and Knalf (2005) indicate that validating and drawing conclusions from condensed data sets can
assist in interpreting findings. Data comparison, the final step of analysis, highlighted patterns
within the literature. This was accomplished by creating an additional table, whereby all study
findings were re-arranged into common themes and subthemes. The following chapter discusses
the findings and data synthesis from the literature.

37

Chapter 4
Findings
Through a comprehensive search, nine articles were identified that addressed food
insecurity (FI) in connection with Type 2 Diabetes (DM 2), and six were identified that
addressed methods of identifying food insecurity. The final set of articles included: 5 reviews (3
systematic, 2 focused literature reviews), 1 combined systematic review with a qualitative study,
1 longitudinal observational study, 1 case control study, and 6 cross sectional studies.
This literature review sought to explore food insecurity with relevance to a Canadian
context. As such, Canadian and countries of similar context were sought wherever possible.
Thirteen percent of articles were conducted in Canada, 53% of the selected articles were
conducted in the United States (US) and the remaining 34% were conducted in various other
countries. Of the remaining 34%, all but one of these articles were reviews covering articles from
multiple countries. Combined, the reviews extracted accounted for most of the literature
available from Canada and the US. The methods, key findings, strengths, and limitations were
extracted from each article and are displayed in the literature review matrix in Appendix B. The
CASP tool was used and modified as necessary to assist in critiquing and appraising the content
and quality of the selected articles. Four themes arose from the literature review highlighting
how DM 2 is affected by FI, and how FI screening can be used to improve the management of
DM 2. These themes were:
1. Underlying mechanisms between FI and the management of DM 2
2. Hierarchies of need and motivation
3. Strategies to identify FI
4. Clinical approaches

38

The research question was indirectly addressed by highlighting three underlying
mechanisms of influence (food behaviors, medication use, mental health) and three established
hierarchical needs (food consistency, social support, self-efficacy). The research question was
directly answered by focusing on strategies that effectively identify food insecurity (selecting
corresponding definitions, measurements, and tools), and discussing clinical approaches. The
identified themes are discussed in the following sections.
Underlying Mechanisms of Influence
This review highlights that the influence of FI on the management of DM 2 is significant:
There is a near independent effect of FI on diabetes control (Shalowitz et al., 2017), which
occurs in an increasing dose-response manner (Ippolito et al., 2016). As a result, individuals are
at a significant increased risk for both hyperglycemia, hypoglycemia and overall poor diabetes
control (Essien et al., 2016). Ippolito et al (2016) found that hypoglycemic episodes are more
frequent and severe with FI, and Shalowitz et al. (2017) found that glycemic control and A1C are
more challenging to manage as the severity of FI increases. Understanding the mechanisms
through which FI influences DM 2 is critical to its optimal management. Common underlying
mechanisms arose from the literature regarding FI and its influence on the management of DM 2.
These mechanisms include changes in food behaviors, medication adherence and mental health.
Changes in food behaviors.
The literature highlighted that DM 2 was more likely to be poorly controlled as a result of
changes in food behaviors secondary to FI. Two separate studies attempted to examine food
behavior changes amongst individuals with DM 2 experiencing FI (Bawadi et al., 2012; Essien et
al., 2016). Essien et al., (2016) examined the influence of FI on DM 2 an extensive focused
literature review. They defined FI as uncertain or limited access to adequate and safe amounts of

39

food” (Essien et al., 2016, p. 78). They found that when food affordability was the main driver
for FI, food behavior changes may occur as a compensatory strategy. For example, individuals
may frequently engage in binge-fast cycles coinciding with pay cycles, subsequently
experiencing widely varied dysglycemia throughout the month (Essien et al., 2016).
Hypoglycemic episodes occur more frequently when individuals fast in times of low income, but
their medication regimens remain unchanged (Essien et al., 2016). Several studies examined by
Essien et al. (2016) found that hypoglycemia occurred significantly more often individuals with
income-associated FI (AIFI). In one study, Essien et al. (2016) found hospital visits secondary to
hypoglycemic episodes increased by 27% at the end of the month amongst low-income
individuals, compared to the first week of the month. Furthermore, hyperglycemic events also
appeared to occur according to pay-cycles, resulting in food binging when it is available. The
exact cause for binging remains unclear in the literature, but appears partly driven by FIassociated stress and anxiety and sudden availability of food following a fast (Essien et al.,
2016). It is unclear whether binge-fast cycles occur in other FI groups who are financially secure.
Essien et al. (2016) also found overall poor dietary choices are associated with FI:
Individuals with AIFI consumed more saturated fats, simple carbohydrates, highly processed
foods and less fruits and vegetables contributing to overall inferior glycemic control (Essien et
al., 2016). Most studies were cross-sectional designs; therefore, the relationship is associative.
However, it is speculated that these changes in food behaviors occur in IAFI because they are
cheaper and consumed more readily (Essien et al., 2016). Essien et al., (2016) do not describe the
food behaviors of income secure-food insecure individuals. Distress associated with FI and
interferes with the management of DM 2 (Essien et al., 2016). However, Essien et al., (2016) do
not comment on whether poor self-management interferes with food behaviors.

40

While this focused literature review covered a substantial amount of material, the
strength of the findings is limited by several factors: This literature review did not explain
individual study strengths and limitations, therefore it is difficult to comment on the strength of
the findings. Most studies were of cross-sectional or longitudinal design, meaning no cause and
affect can be established, nor can the complex relationship between FI, food behaviors and DM 2
be further explained. Further, it is difficult to generalize the results because studies came from
varying countries. Each nation’s healthcare model, coverage, insurance and circumstance
influence overall diabetes control. Lastly, these particular findings are limited to the IAFI
subgroup. No conclusions can be drawn about food behaviors amongst income-secure
individuals with DM 2 experiencing FI.
Bawadi et al. (2011) found similar results in their cross-sectional study examining the
relationship of FI to glycemic control. Nearly 100% of individuals with severe FI engaged in
compensatory food behaviors to avoid running out of food at the end of the month including
meal skipping or cutting portion sizes, and that there was a corresponding worsening of glycemic
control with increasing FI severity ([p <0.04], Bawadi et al., 2011). The authors also found that
worsening FI was associated with increasing difficulties in consuming a balanced meal:
approximately 70% of individuals who identified as food insecure could not afford balanced
meals (Bawadi et al., 2011). Additionally, increasing FI severity was associated with greater
sugar intake, despite caloric intake remaining similar across food security groups (Bawadi et al.,
2011). These findings suggest that the ratio of carbohydrates among FI groups is greater than
food secure counterparts. This may account, in part, for worsening glycemic control.
Interestingly, Bawadi et al., 2011 defined FI as “limited or uncertain availability of nutritionally
adequate and safe foods in socially acceptable ways (p.250).” However, this study used the

41

HFSSM-sf questionnaire. While it is validated to use in multiple countries, it is only validated to
measure AIFI. It is then, unclear, if the results were then intended to be generalized to all
individuals experiencing FI. Irrespective of intent, the results of this study can be only applied to
a subgroup of IAFI.
Unfortunately, several limitations exist: Most importantly, this study used cross sectional
design, therefore the results are associative, not causational. Based on this study’s parameters, no
further understanding can be gleaned surrounding relationship between food behaviors, FI, and
DM 2 management. Additionally, there was no power calculation. Power analyses are usually
completed before a study begins, in order to establish the sample size required for the statistical
analysis to reliably detect what it is aiming to measure (Sutherland, 2017b). Similar to other
studies, the use of questionnaires is always subject to recall bias. It is also difficult to generalize
the result from a Jordanian study as the healthcare system is a variable not controlled for across
studies.
Changing food behaviors appears to occur primarily amongst individuals with IAFI, and
can affect overall glycemic control and safety. Care should be taken to identify this subset
population of FI individuals to ensure their unique risks are identified and mitigated. It is
speculated that income influences the quantity, quality and temporal aspect of food consumption.
Due to the nature of the studies, however, this relationship is not well understood and the
research does not clarify how much less food is consumed, which specific foods and how
frequently. Further the magnitude of impact of these food behaviors compared with FS
counterparts is not well understood.

42

Hierarchy of eating patterns.
The compensatory food behaviors discussed by Essien et al., (2016) highlight a pattern of
eating motivated by basic physiological needs rather than from a desire to manage DM 2.
Consuming food to managing DM 2 appears to be superseded by the immediacy of insufficient
access to consistent food (Essien et al., 2016). Essien et al. (2016) found that when consumption
of food is uncertain due to finances, eating what is accessible, available and affordable appears to
take precedence (Essien et al., 2016).
FI may also re-prioritize an individual’s motivations for eating in other ways: When
hypoglycemia occurs, Essien et al. (2016) found that individuals experiencing FI may be forced
to eat whatever is available to them to correct hypoglycemia or to prevent hypoglycemic risks
associated with their medications, instead of eating to maintain proper glycemic control.
Subsequently, this may increase the risk of reactionary hyperglycemia, poorer glycemic control
over time, and more prescriptions (Essien et al., 2016).
Clinical approach.
A literature review by Thomas, Fitzpatrick, Sidani, & Gucciardi (2018) examined the
management of individuals with DM 2 experiencing FI and suggest several important clinical
approaches: (1) each individual who screens positive for FI be assessed for hypoglycemia risk
and occurrence; (2) nutrition recommendations be individualized to respect the individual’s
capacity to manage a diabetic diet; (3) information on low-cost, but effective methods of treating
hypoglycemia be provided; and (4) referrals to dietitians and social workers be considered as
they may play a supportive role in managing FI (Thomas et al., 2018).
Thomas et al. (2018) recommend connecting identified patients with low cost food
options within the community, including known local community food banks, kitchens, and low-

43

cost grocery stores. Depending on jurisdictions, application for Special Diet Allowance funding
may be appropriate (Thomas et al., 2018). Unfortunately, this systematic review had several
limitations: Firstly, the authors did not discuss their critical appraisal of the evidence, what level
of evidence informed each recommendation, how the findings were synthesized and whether
they were based on one or more studies. Therefore, the strength and level of evidence of the
recommendations are unclear. Additionally, Thomas et al., (2018) defined FI as “inadequate
access to food because of financial constraints (p.258)”, but it is unclear whether the studies
included in the systematic review measured and identified FI similarly. Lastly, if Thomas et al.,
(2018) restricted their studies to IAFI, then it is challenging to generalize these recommendations
to the general population experiencing FI.
In support of these findings, the review by Essien et al. (2016) found that dietary quality
and A1C levels improved with healthy supplemental food assistance programs (SNAP) and selfmanagement education support programs both improve glycemic control amongst individuals
with IAFI. These findings are limited by a lack of qualifying information describing the nature of
the SNAP program (i.e. which foods) and which self-management education strategies result in
glycemic improvements. These results are also limited to the subset population of individuals
experiencing IAFI.
In contrast, Ippolito et al. (2016) examined the association between food security (FS)
and diabetes self-management in individuals already accessing food pantries. Their findings
suggest that food pantries and banks alone do not appear to mitigate food insecurity or its effect
on the management of DM 2. Only 16% of individuals accessing the food bank self-rated as food
secure, while the majority, 42%, continued to report very low food security (Ippolito et al.,
2017). Further, compared to the food secure group, the very low food secure group continued to

44

report greater diabetes distress (p < 0.001), decreased diabetes self-efficacy (p < 0.001),
medication adherence (p < 0.001), severe hypoglycemic episodes (p < 0.001; OR 2.63), and
depressive symptoms (p < 0.001; OR 4.56], Ippolito et al., 2016). These findings suggest that
access to food banks and food supply alone is insufficient in mitigating FI. The results of this
study demonstrated strong statistical significance, and odds ratios demonstrates strong clinical
significance. Similar to other studies, these results were limited by cross-sectional design, and
the levels of food insecurity (food secure, low food secure and very low food secure) were
decided by convention.
Changes in mental health.
While depression is common among chronic diseases, this review of the literature
suggests that rates of distress and depression are elevated and are further compounded by chronic
FI (Essien et al., 2016; Ippolito et al., 2016). These changes in mental health influence the
management of DM 2 (Essien et al., 2016; Walker et al., 2018): In a cross-sectional study,
Walker, Williams and Egede, (2018) examined direct and indirect pathways through which FI
influences diabetes self care behaviors and glycemic control. Diabetic self-care behaviors were
defined as healthy eating, exercising, self-monitoring of blood glucose and foot checks and
glycemic control was measured by A1C levels (Walker et al., 2018). They found FI that is
significantly associated with stress (p <0.001) and increased A1C levels (p< 0.03). Further the
higher participants perceived their level of stress, the poorer their capacity to engage in self-care
behaviours (p < 0.001). This study did not demonstrate statistically significant worsening of A1C
from poorer self-care behaviors (Walker et al., 2018). However, poor self-care behaviors do not
have immediate effects on glycemic control or diabetic complications, therefore this relationship
may be difficult to capture in cross-sectional design. Similar findings were found by Essien et al.,

45

(2016) in their systematic review. One study in Essien’s review found that depressive symptoms
were associated with lower self-efficacy leading to poorer glycemic control. It is speculated that
the superimposed stress of diabetes on FI creates feelings of powerlessness (Essien et al., 2016).
Self-efficacy is not defined; therefore, it is challenging to understand how mental health
specifically influence diabetes management in individuals with FI.
In another cross-sectional study, Montgomery, Lu, Ratliff and Mezuk (2017) found that a
dose-response relationship exists between FI and depression (p > 0.001) Mild FI is associated
with 2.5-fold increase in mild depression and severe FI with a 3.5-fold increase in depression
(Montgomery et al., 2017). Unfortunately, this study does not draw connections between
depression and to diabetes control, therefore its clinical significant to diabetes management
remain unclear.
Hierarchy of social need.
Social support emerged as a requisite need amongst individuals with DM 2 experiencing
FI. Kollannoor-Samuel et al., (2011) found increased depressive symptoms were associated with
an increased perceived interference of FI on the management of DM 2 (p <0.001), and increased
diabetes-related clinical symptoms (p <0.001). Social support, however, mitigated depressive
symptoms (p <0.01). Social support was defined by Kollannoor-Samuel et al., (2011) as the
participants’ perceptions of consistent support when needed from friends and family, and was
captured using validated 11-item questionnaire called the Multi-Dimensional Diabetes
Questionnaire. FI was defined as “limited access, availability, or intake of nutritionally adequate
foods” (Kollannoor-Samuel et al., 2011, p. 986). In their study, Kollannoor-Samuel et al., (2011)
found consistent informal support from friends and family provided a buffering effect against
depression and distress amongst individuals with FI. An inverse relationship between social

46

support and depression levels were noted (Kollannoor-Samuel et al., 2011). Further, as FI
intensified, the influence of social support on depression was more profound (KollannoorSamuel et al., 2011). These results suggest that social support is an important buffer amongst the
most food insecure individuals. An important strength to the findings was that a post-hoc alpha
Cronbach demonstrated internal consistency and reliability of the surveys used. Important
limitations to this study were its cross-sectional design, and an inability to control for underlying
genetic predispositions for depression. Further, perceived social support is limited by recall bias
and a limited generalizability of study findings because the study population consisted of mostly
Puerto Rican female participants. Further, there was no power calculation, therefore it is difficult
to ascertain if the study was sufficiently powered to demonstrate small variations for its sample
size. Importantly, Kollanoor-Samuel et al., (2011) used the short-form Household Food Security
Supplement Module (HFSSM-sf) as a FI screening tool. This tool is validated only to capture
income-associated FI (IAFI), therefore these findings are limited to this subset population of FI
individuals.
Walker et al. (2018) also found social support mitigated the effects of stress and
depression on DM 2 amongst individuals with FI. In this study, Walker et al. (2018) defined FI
as an “inability or limitation to accessing nutritionally adequate foods, or a dependence on
emergency food supplies (p.3237)”. Walker et al. (2018) examined the pathways through which
FI influences glycemic control in individuals with DM 2, and found that FI is directly related to
an increased glycated hemoglobin (Hb A1C) through increased stress. Further, Walker et al.
(2018) found FI is indirectly related to an increased Hb A1C through poor self-care behaviors
driven by stress and low social support. Social support appeared to mitigate these pathways (p <
0.001). Social support was defined as positive social interaction using a 3-item survey, and

47

included (1) when individuals could identify others with whom they could have a good time
with, (2) with whom they could get together regularly and (3) with whom they could relax or do
something enjoyable (Walker et al., 2018). Individuals scoring lower on this social interaction
scale experienced a significantly higher level of stress directly interfering with diabetes
management ([p < 0.001]; Walker et al., 2018). These findings suggest that facilitating social
support can improve glycemic control in individuals with FI by increasing an individual’s
capacity to engage in diabetes self-care behaviors and by directly lowering stress levels.
The results by Walker et al., (2018) are more reliable. A calculated power indicates that
the sample size was large enough to accurately capture the associations between FI, stress, and
DM 2 management. The results of the findings were highly statistically significant (0.05-0.0001)
within narrow confidence intervals. Further, well-validated questionnaires were used. Similar to
Kollanoor-Samuel et al., (2011), the results are limited by cross-sectional design, the use of selfreport questionnaires and are subject to recall bias.
Overall, both studies highlight how addressing nutrition choices alone in individuals with
DM 2 experiencing FI is insufficient strategy and requires a multi-prong approach, considerate of
strategies that mitigate stress and depression. The findings suggest social support as a requisite to
well-controlled diabetes amongst individuals with DM 2 experiencing FI.
Clinical approach.
Walker et al., (2018) and Thomas et al. (2018) concluded that the benefits of social
support should be considered in addition to dietary strategies when managing individuals with
DM 2 who are food insecure. According to Walker et al., (2018) and Kollanoor-Samuel et al.,
(2011) even promoting informal social support amongst individuals with DM 2 experiencing FI
may decrease an individual’s overall risk of depression and therefore improve the individual’s

48

capacity to perform diabetes self-care behaviours. Thomas et al. (2018), suggest that all
individuals with FI should be assessed for social support, and that family members should be
encouraged, wherever possible, to be included in education and management as a means of both
social and logistical support. It is unclear what evidence informs this recommendation.
Lastly, income-associated (IAFI) or financially-driven FI is common in developed
countries (Essien et al. 2016). Kollannoor-Samuel et al., (2011) highlighted that individuals with
DM 2 with lower household income were especially vulnerable to increased rates of depression,
in particular, because limited income impairs an individual’s capacity to manage stress (p <
0.03). Individuals experiencing IAFI may be additionally vulnerable to the effects of stress and
depression, since they may be juggling the cost of basic needs such as rent, electricity, and
clothing on top of food (Kollannoor-Samuel et al., 2011). Identifying when finances are at the
core of FI may be necessary to address additional barriers this population may be facing, and
consider social work involvement, recommended by Thomas et al., (2018).
Changes in medication use.
FI appears to interfere with changes in overall medication use and safety. Several studies
noted that medication underuse was significantly more likely to occur IAFI (Essien et al., 2016;
Ippolito et al., 2016). “Food-medicine-health-supply trade-offs” (FMS trade-off) is a
phenomenon that occurs when people have less money, and results in having to choose between
purchasing food, medicine or health supplies, because the individual cannot not afford all three
(Ippolito et al., 2016). In their study examining FI and diabetes self-management in individuals
accessing food pantries, Ippolito et al. (2016) found that FMS trade offs were almost 7 times
times more likely to occur in the very low food secure group, and medication affordability 4
times more likely compared with the more food secure group of individuals with DM 2 accessing

49

food banks (p < 0.001). The focused literature review by Essien et al. (2016) revealed similar
findings: some included studies reported up to half of their participants with IAFI either stretched
out medication prescriptions, delayed filling prescriptions or did not use the medications at all
(Essien et al., 2016). Other studies in their review reported that individuals with FI were six
times more likely to stretch out their medication prescriptions (Essien et al, 2016). Caution,
however, must also be taken when comparing medication and supply affordability in comparison
with American studies. For example, glucose monitors are free when tests strips are purchased in
Canada, and test strips are part of Fair Pharmacare; meaning, low income patients will either
have the majority or all of their strips covered based on their diabetic medication regimen
(Government of British Columbia, 2018b). A select list of medications, previously discussed, is
also covered under Fair Pharmacare for individuals with low income once deductibles are met
(Government of British Columbia, 2018c).
Interestingly, in a cross-sectional study examining whether food insecurity is
independently associated with delays in filling prescriptions, Bilimek and Sorkin (2012) found
individuals with FI are two times more likely delay or not to fill their prescriptions even when all
other variables are adjusted for (p <0.01). FI was defined as “limited or uncertain access to food
due to inadequate financial resources (Bilimek and Sorkin, 2012, p.2160). However, the cause
medication adherence issues without a financial drive remain unknown and are limited by study
design. One important limitation to this study is that assume individuals below poverty line are
only classified with IAFI, and does consider other populations of IAFI. The study also does not
perform a power analysis to ensure enough participants were recruited to power the findings.
Lastly, even though multivariable logistic regression analysis shows, medication adherence

50

issues is statistically significant for reasons other than finances, this has yet to be demonstrated
as a primary outcome among all individuals with FI.
In contrast, Shalowitz et al. (2017), sought to determine whether FI is associated with
poor glucose control over time, and found that insulin use increases over time (p < 0.010). The
authors comment that insulin use and its initiation, however, is associated with and is a predictor
of increasing A1C over time (Shalowitz et al., 2017). Therefore, increased insulin use in
individuals with FI is a consequence of dysglycemia, rather than a demonstration of medication
adherence (Shalowitz et al., 2017). Longitudinal study design does not prove causality of
findings the findings. Further no power analysis was performed. Irrespective of reason, increased
use of insulin over time, brings into question medication adherence amongst FI populations.
An important finding that emerged from the literature review by Essien et al., (2016): In
considering changes to food behaviors, a significant risk of medication-glycemia mismatch exists
when: (1) food becomes scarce but medication regimens remain unchanged, or (2) food
consumption is maintained at the expense of discontinuing medications (Essien et al., 2016). A
high risk of hypoglycemia can also contribute when the provider does not understand food
consumption behaviors These findings highlight important safety concerns and the importance of
addressing medication regimens and adherence in the presence of FI.
Clinical approach.
Thomas et al., (2018) suggest the following pharmacological approach of individuals
with DM 2 experiencing FI. Firstly, understand a client’s medication coverage, and then consider
replacing antihyperglycemics with the greatest risk of hypoglycemia with safer alternatives
(Thomas et al., 2018). If taking insulin, consider a regimen that is more flexible (Thomas et al.,
2018). Further, consider scheduling medications with meals wherever appropriate, and

51

withholding when there is no food intake (Thomas et al., 2018). Lastly, Thomas et al. (2018)
suggest that less stringent glycemic targets should be considered. Unfortunately, a lacking
critical analysis of the evidence limits the strength of these findings.
Reduced self-efficacy.
Within the literature a theme emerged related to FI and an overall decreased self-efficacy.
Ippolito et al., (2016) define self-efficacy as an individual’s belief in his or her own capacity to
manage his or her diabetes. Self-efficacy is a key aspect in chronic disease self-management
(Essien et al., 2016; Ippolito et al., 2016). Decreased self-efficacy impairs self-management
behaviors such as following a diet, resulting in inferior diabetes control (Essien et al., 2016). In
their cross-sectional design examining the association between FI and diabetes self-management,
Ippolito et al., (2016) used an eight-item instrument validated to measure self-efficacy and used
Likert scale to establish mean result for each subset group of food security: Food secure, low
food secure and very low food secure (Ippolito et all., 2016). Lower self-efficacy scores are
usually correlate with worse A1C scores (Ippolito et all., 2016). This study found overall
decreased self-efficacy interfered with diabetes self-management, on average, 36% more often in
severe FI individuals, relative to food secure participants (p < 0.001). This, however, did not
translate into inferior A1C control and the underlying mechanism is unclear due to the nature of
cross-sectional studies. Further, it is unclear how self-management behaviors are affected.
These trends in self-care behaviors were also noted by Gucciardi et al., (2009), who
found that FI was associated with decreased self-efficacy and individuals were more likely to be
physically inactive, be overweight, eat less fruits and vegetables, be twice as likely to be a
smoker, be twice as likely to report stress, and report lower general health, mental health and
perceived stress. Essien et al. (2016) offer one mechanism through which self-efficacy may be

52

impaired: competing demands between food and managing DM 2 may precipitate chronic stress,
distress, and depression contributing to a reduced mental capacity, and elicit feelings of
powerlessness that diminish one’s self-efficacy (Essien et al., 2016). Self-efficacy is not defined
by Essien et al. (2016), affecting the clarity of findings.
Clinical approach.
Evidence is limited in providing strategies to improve self-efficacy; however, it appears
that if depression, distress and stress are mechanisms through which self-efficacy is hampered
(Essien et al., 2016), then social support as described by (Kollannoor-Samuel et al., 2011) may
be one clinical approach to improve self-efficacy. Caution must be taken in these
recommendations due to a lack of supporting evidence.
Screening for Food Insecurity
The findings from this literature suggest screening for FI can be successfully
implemented in the healthcare setting. Studies demonstrated most patients and care providers
involved in the FI screening felt the healthcare setting was both beneficial and appropriate (De
Marchis et al., 2019; Thomas et al. 2018). In a systematic review examining the validity and
acceptability of food insecurity (FI) screening tools and their implementation in health care
settings, De Marchis et al. (2019) found that between 66-88% of patients screened for FI found it
appropriate to be asked about FI in the healthcare setting. De Marchis et al. (2019) also found
that between 80-89% of HCP felt FI screening was appropriate during healthcare visits. The
strength of these findings were limited by low quality to very-low quality findings from two
qualitive and four cross sectional studies (De Marchis et al., 2019). The generalizability of the
findings are limited to income-associated food insecurity (IAFI), because all FI screening tools
were validated against the United States Department of Agriculture Food Security Scale (USDA-

53

FSS) which is only validated to measure. Lastly, generalizability was also limited by inclusion of
some pediatric studies in the systematic review.
The findings by De Marchis et al., (2019) were also in keeping with the results from a
pilot project from Thomas et al. (2018), who examined the acceptability and feasibility of an
implemented FI screening in healthcare using a qualitative pilot project based in a Toronto
healthcare setting. Thomas et al. (2018) found 71% of participants were comfortable discussing
their personal experiences of FI with their healthcare providers (HCPs). The HCPs found FI
screening helpful in guiding care and support for patients. Further, the HCPs reported that FI
screening brought meaning and understanding as to why patients may not follow diabetes
management recommendations, how it impacts their ability to self-manage their diabetes and
how they cope with FI (Thomas et al., 2018). Both the HCPs and patients found FI screening
both relevant and non-intrusive in the healthcare setting. None of the patients reported negative
experiences (Thomas, et al, 2018). In fact, strong rapport with their provider increased the
likelihood of engaging in FI screening (Thomas et al., 2018).
Important caveats exist: De Marchis et al., (2019) found time required to complete FI
screening tools was between 30 seconds to 10 minutes. Four out of 10 studies reported time as a
barrier to FI screening (De Marchis et al., 2019). It is unclear whether an optimal screening time,
or threshold existed. It should also be noted that not everyone who screens positive for FI wants
care provider assistance, therefore judicious use of screening should be considered.
Clinical approach.
The literature review did not support one preferred method of identifying food insecurity.
Instead, strategies emerged that focus on context-specific, individualized approaches to
identifying FI. This strategy includes: (1) selecting a well-defined, context-specific definition of

54

food insecurity useful to the specific purpose, which (2) highlights the attributes of FI intended
on being measured and studied, and lastly, (3) selecting a corresponding tool that measures these
attributes.
Selecting a definition of food insecurity (FI).
A context specific definition was preferred over a comprehensive and universal definition
of FI for two reasons: (1) A diversity of FI determinants or attributes exist unique to each
circumstance, and (2) very few attributes are antecedents. These findings are corroborated by
Jones et al. (2013). Their systematic review of the literature attempted to conceptualize a
universal and comprehensive definition food (in)security and found that a universally acceptable
definition of FI unlikely to be appropriate and applicable in all circumstances (Jones et al., 2013).
Lack of antecedents.
FI is not universally defined in the literature: Twelve out of fifteen articles provided a
definition of food insecurity. Of the twelve definitions, seven different variations emerged and
only two antecedents. The two antecedents included: (1) uncertain or inconsistent access, and (2)
an inadequate amount of food (Bawadi et al., 2012; Billimek & Sorkin, 2012; De Marchis et al.,
2019; Essien et al., 2016; Gucciardi et al., 2009; Ippolito et al., 2016; Jones, Ngure, Pelto, &
Young, 2013; Kollannoor-Samuel et al., 2011; Makelarski, Abramsohn, Benjamin, Du, &
Lindau, 2017; Montgomery et al., 2017; Shalowitz et al., 2017; Walker et al., 2018).
One antecedent emerged with respect to FI and DM 2: universal acknowledgement that
FI interferes with proper diabetes management and control (Bawadi et al., 2012; Billimek &
Sorkin, 2012; Essien et al., 2016; Gucciardi et al., 2009; Ippolito et al., 2016; Kollannoor-Samuel
et al., 2011; Montgomery et al., 2017; Shalowitz et al., 2017; Walker et al., 2018). A final
compilation of antecedents, “uncertain access to sufficient amounts of food which interferes with

55

diabetes management”, has several limitations: (1) The terms “food”, “access”, and “inadequate
amounts” are not adequately explored or explained (Jones et al., 2013), which limits applicability
of results. More specifically, what foods are specifically limited in access and in which amounts?
None of the literature explored these concepts in detail. Essien et al., (2016) discussed some
changes to food group choices such as fruits and vegetables, as well as changes to food
behaviors; however, these are non-specific. (2) How FI interferes with diabetes management is
multifactorial, and specific to the factors causing FI. For example, income-associated FI (AIFI) is
specifically associated with binge-fast cycles resulting in dysglycemia, but this trend has not
been explored amongst the general FI population. Lastly, (3) based on this review, the literature
suggests that diabetes management is also influenced by FI in several indirect ways including
depression, distress, diminished self-efficacy and medication behaviors. These also have not
been explored in the literature in ways that allows for sufficient understanding and concise
definition. Study design is a significant limiting factor to understanding the complex relationship
and details required to be capable of defining these terms.
Diversity of attributes.
Several FI attributes emerged from the literature review, but were not universal
antecedents. Instead, these attributes were relevant and specific to the study to which it applied.
Attributes (or determinants) are features inherent to, or associated with FI (Jones et al., 2013).
Attributes differ from antecedents in that they do not need to be universal in all cases of FI.
These determinants included: (1) lack of access to nutritionally appropriate food (Bawadi et al.,
2012; Gucciardi et al., 2009; Jones et al., 2013; Kollannoor-Samuel et al., 2011; Montgomery et
al., 2017; Shalowitz et al., 2017; Walker et al., 2018); (2) lack of access to safe foods (Bawadi et
al., 2012; Essien et al., 2016; Gucciardi et al., 2009; Makelarski et al., 2017; Montgomery et al.,

56

2017; Shalowitz et al., 2017); (3) access to food in socially acceptable ways (Bawadi et al.,
2012; Jones et al., 2013); (4) FI secondary to financial constraints (Billimek & Sorkin, 2012; De
Marchis et al., 2019a; Ippolito et al., 2016; Jones et al., 2013; Shalowitz et al., 2017); and (5)
having insufficient food to meet dietary needs to maintain a healthy and active life (Gucciardi et
al., 2009). Similar to the antecedents, efforts to understand and adequately define these attributes
are not found in the literature.
Context-specific approach.
Jones et al. (2013) suggest an individualized approach to FI definitions focused on
pertinent attributes that can be defined by the clinic or provider, and managed by a particular
program, clinic, or individual’s resources (Jones et al., 2013). Jones et al. (2013) suggest
disassembling the definition of FI into smaller, more manageable concepts. For example,
income-associated FI (IAFI) or FI secondary to financial constraints occurs very frequently, but
does not occur in all cases of food insecurity (De Marchis et al., 2019; Ippolito et al., 2016; Jones
et al., 2013; Shalowitz et al., 2017). Identifying IAFI is important because there are several IAFIspecific diabetic consequences that can arise (Essien et al., 2016). Medication underuse, for
example, appears to be a problem primarily among individuals with IAFI (Essien et al., 2016).
Using a context-specific definition focused on particular attributes of interest helps identify
specific issues related to FI (Jones et al., 2013). Further, using a context-specific definition for FI
assists in selecting an appropriate FI screening tool: Marques et al., (2015) and Jones et al.
(2013) found screening tools are validated to identify and measure specific attributes of FI, rather
than FI as comprehensive concept. For example, several tools exist that are specifically validated
only to identify IAFI (Marques et al., 2015). Defining FI as secondary to financial constraints
(IAFI) points to which screening tools are more appropriate for selection.

57

Selecting attributes and screening tools.
According to Jones et al. (2013), definitions, attributes, and screening tools should all
align in order adequately and accurately capture what is intended on being measured. The
literature highlighted significant limitations to this process: In their systematic review examining
24 different FI tools from 184 different articles, Marques et al., (2015) found all included
screening tools measured income-associated FI (IAFI) and that most key attributes of FI remain
unanalyzed in the literature (75% of evidence rated positive or neutral). Similarly, another
systematic review by Ashby et al., (2016) examined the attributes of eight different FI screening
tools used in developed countries and found physical and economic access to food were only
measured (Ashby et al., 2016).
Few tools attempt to measure other attributes of FI (Marques et al., 2015): The
Radimer/Cornell tool attempts to measure food utilization, and enquires about anxiety secondary
to required assistance in preparing food (Ashby et al., 2016). The Kuyper tool attempted
to measure food stability or insecurity over time (Ashby et al., 2016), and only one tool
attempted to measure both household level FI, and individual level FI (Ashby et al., 2016). A
final systematic review by Jones et al., (2013) examined the attributes of 17 different FI
screening tools (evidence rated low) and also found most attributes of FI were not captured in
screening tools. Jones et al. (2013) also found food measures are not always clear about what
attributes they are measuring, nor are the attributes consistently defined. For example, access to
food can be further broken down to access, acquisition, consumption (Jones et a., 2013).
However, this is often not well articulated in screening tools. It could be speculated that the lack
of clarity surrounding attributes in screening tools may be, in part, a consequence of the lack of
clarity surrounding FI definitions and concepts.

58

Another limitation to FI screening is that most tools have not been significantly analyzed
(Marques et al., 2015). Marques et al. (2015) found the Core Food Security
Measurement/Household Food Security Survey Module (CFSM/HFSSM) is one of the few FI
screening tools significantly analyzed in literature, while the Self-Perceived Household Food
Security Scale (SPHFSS) and the short-form HFSSM (HFSSM-6SF) have been moderately
analyzed in the literature. The CFSM/HFSSM and HFSSM-6SF have been studied and validated
in many countries worldwide (Marques et al., 2016). Half of the other tools examined by
Marques et al. (2015) had only been tested or studied in the literature one or two times (Marques
et al., 20165).
In general, several challenges remain in accurately capturing individuals with FI in
practice: screening questionnaires will always be subject to a response or recall bias (Jones et al.,
2013). Well-validated tools such as the Household Food Security Survey Module (HFSSM) and
its short form have an internal consistency of 0.82-0.94 and have found to be valid and reliable
(Marques et al., 2014); however, few screening tools have been so well analyzed. It remains a
challenge that the only tools well analyzed only measure IAFI. Validating non-IAFI screening
tools can also be challenging when they are so diverse in what they measure (Jones et al., 2013).
FI cut-offs are also not standardized, with some studies establishing levels of FI by convention
(Ippolito et al., 2016). Additionally, an important trade-off to consider when selecting FI tools is
whether to use comprehensive versus short form (1 or 2 question) tools in the setting of time
constraints. Comprehensibility and contextual detail need to be weighed against simplicity and
comparability (Jones et al., 2013). Thus, it remains a challenge to ensure true capture of the FI
population representative of what is intended on being identified. Based on these systematic
reviews, there are no comprehensive tools identifying all attributes or domains of FI (Ashby et

59

al., 2016; Jones et al., 2013). While an individualized approach is certainly preferred (Jones et
al., 2013), these findings suggest that selecting attributes of interest are significantly limited by a
lack of diverse, well-analyzed, validated screening tools.

60

Chapter 5
Discussion
Through this integrative review, several themes were elicited that highlighted the
mechanisms of influence through which FI impacts the management of DM 2. These included
changes in food behaviors, mental health and medication use. Strategies to identify food
insecurity were presented. Collectively, the findings addressed the integrative review question:
“What strategies can NPs employ to identify FI and enhance the management of DM 2 in the
primary care setting among adults experiencing dysglycemia?”
The literature highlighted unique motivations for behavior and hierarchies of need which
appeared to take precedence in the lives of individuals with DM 2 experiencing FI. Such
mechanisms of influence also reflect the chief tenets of the MTM. As discussed, MTM suggests
that needs occur in a hierarchical, but not mutually exclusive, manner (Maslow, 1943), and that
humans are motivated by a desire to satisfy these needs (Maslow, 1943, p. 394). Further, real or
perceived threats to the security of these needs can generate either emergency or compensatory
reactions in behavior (Maslow, 1943, p. 394).
In MTM, food is one of the most basic physiological human needs (Maslow, 1943),
which provides further insight into why individuals with FI struggle with higher-level needs,
such as self-management. MTM may highlight perspectives on how NPs can provide realistic,
FI-informed strategies that can enhance the management of DM 2. Lastly, MTM may highlight
the enhanced value of individualized, FI-informed management strategies when patients are
significantly limited in their capacity to self-manage their DM 2.
The findings from the literature review are discussed in the following section, drawing
connections to the significance for NPs in the PC setting. Included in the discussion are

61

recommendations for NP clinical practice, followed by a discussion of the limitations and future
research recommendations. A diagram highlighting the review findings is presented below.
Diagram II: The impact of FI on DM 2 and strategies to improve management.
WHAT IS
AFFECTED

UNDERLYING
MECHANISMS

HOW?

Self-Management
SMBG
◦ Healthy eating
◦ Exercise
◦ Smoking
◦ Stress

Diabetes
Control

Medication practices

Food Behaviors
◦ Eating what is
available
◦ Binge-fast
eating
◦Treating
hypoglycemia
by any means
available
◦ Food-medicine
mismatch

UNMET
NEEDS

Adequate Food

Social Support

Self-Efficacy
Mental Health
◦ Depression
◦ Distress
Medication
◦ Use
◦ Safety

APPROACHES

SCREENING
No comprehensive tool
exists. Instead, select a FI
tool by focusing on:
 Definition used
 Attributes measured
 Attributes & definition
match population
intended to be measured
 Attributes match
intervention/available
resources

RESEARCH

CLINICAL





Pharmacotherapeutics
Adjusting targets
Mental health screening
Referral to care
providers, community
resources and supports






Diverse screening tools
usable in primary care
Relationship links
between underlying
mechanisms, prepotent
needs and FI
FI-specific management
approaches

62

Theme One: Mechanisms of Influence
Amongst the literature there is a dearth of evidence directly informing the management of
DM 2 amongst individuals who are FI (Thomas et al, 2018). Findings from the literature set
predominantly examined the underlying mechanisms through which FI influences the
management of DM 2, and recommendations for practice were drawn from these findings. These
mechanisms are discussed in detail in the following sections.
Changes in food behaviors.
The literature highlighted that DM 2 was more likely to be poorly controlled as a result of
poor food behaviors secondary to FI (Bawadi et al., 2012; Essien et al., 2016). Specifically,
affordability was a major driver to changes in food behavior and often were compensatory in
nature (Essien et al., 2016). Individuals were found to binge and fast (Essien et al., 2016), cut
meal sizes, and skip meals (Bawadi et al., 2012) according to pay cycles, subsequently placing
them at increased risk for dysglycemia (Essien et al., 2016). In particular, a review highlighted a
significant increase in hypoglycemia risk, and occurrence and severity secondary to decreased
consumption of food while maintaining the same medication regimens (Essien et al., 2016).
Food behaviors also appeared to influence dietary choices. Individuals with FI appeared
to consume more saturated fats, simple carbohydrates, highly processed foods and less fruits and
vegetables, possibly linking food behaviors to overall inferior glycemic control. While the
relationship is associative, several authors within an examined review speculate that these foods
are cheaper and easier to access, therefore are consumed more readily (Bawadi et al., 2012;
Essien et al., 2016).
Several important clinical approaches were suggested in the literature, including
assessing for hypoglycemia risk and occurrence amongst individuals with FI; individualizing

63

nutrition recommendations based on the individual’s resources and abilities; providing
information on low-cost, but effective methods of treating hypoglycemia; and considering
referrals to dietitians and social workers for additional support (Thomas et al., 2018).
Additionally, connecting identified patients with low cost food options within the community has
been recommend (Thomas et al., 2018). In particular, healthy supplemental food assistance
programs paired with self-management support demonstrates evidence for improved dietary
quality and A1C levels (Essien et al., 2016).
Importantly, increasing access to food may not be a sufficient strategy alone in managing
FI. Studies examining individuals with DM 2 accessing food banks found the majority of
individuals still identified as food insecure and reported most, if not all, issues highlighted in this
integrative review, including diabetes distress, decreased diabetes self-efficacy, medication
adherence, severe hypoglycemic episodes, medication affordability issues, medicine-supplytrade-offs, and depressive symptoms (Ippolito et al., 2016). It remains unclear whether these
findings reflect the inadequacy of this particular food bank in providing sufficient food, in
providing diabetic-friendly options, or whether access to food is insufficient in ameliorating FI
amongst individuals with DM 2.
Recommendations for Practice.
These findings suggest that a multi-faceted approach is necessary in food behaviors
amongst individuals with DM 2. Primary care providers should consider their familiarity with
resources in the community, including food banks and low-cost food stores, and determine
whether supplemental funding exists in their jurisdiction, especially for individuals experiencing
IAFI. Thomas et al. (2018) found that establishing a list of community resources was useful in

64

the implementation of FI screening and management in diabetes clinics, and helped increase
access to meaningful resources.
PCPs should also routinely assess for hypoglycemia risk and occurrence, and consider
pharmacological strategies to prevent this from occurring4. The literature does not suggest a
frequency for assessment, but the Canadian Diabetes Guidelines (CDG) do suggest assessing for
hypoglycemia amongst all individuals every visit (Yale, Paty, & Senior, 2018).
There are no recommendations or suggestions in the literature regarding individualized
nutritional strategies. PCPs may consider co-management of an individualized dietary strategy
with social work and dieticians. When individualizing a nutritional strategy, consideration should
be taken in relation to the patient’s food resources and also the individual’s level of agency and
self-efficacy. The CDG suggests self-management support (SMS) strategies as they focus
directly on building a trusting and collaborative relationship, with consideration of the
individual’s abilities and capacity (Sherifali, Berard, Gucciardi, MacDonald, & MacNeill, 2018).
It remains unclear, however, how SMS may be used for individuals with FI who experience a
decreased capacity for self-management (Essien et al., 2016). Future studies should examine how
PCPs may motivate and partner with individuals with FI who face difficulties with food
consumption, self-management and self-efficacy.
Changes in mental health.
Studies found that rates of distress and depression are elevated by chronic FI (Essien et
al., 2016; Ippolito et al., 2016) and influence the management of DM 2 (Essien et al., 2016;
Montgomery et al., 2017; Walker et al., 2018). Stress directly interferes with an individual’s
ability to perform self-care behaviours such as healthy eating, exercising, self-monitoring of

4

See discussion section on medication use

65

blood glucose and foot checks, subsequently resulting in worsening A1C levels (Ippolito et al.,
2016). The results from the integrative review suggest a close relationship between FI, mental
health and an individual’s capacity to self-manage their diabetes, although the exact mechanism
remains unclear.
Some studies highlighted that individuals with DM 2 and lower household income are
especially vulnerable to increased rates of depression, in particular, because limited income
impairs an individual’s capacity to manage stress. Therefore, individuals experiencing incomeassociated FI (IAFI) may be additionally vulnerable to the effects of stress and depression.
Identifying when finances are at the core of FI may be necessary to address additional barriers
this population may be facing, such as the cost of rent, electricity, and clothing on top of food
(Kollannoor-Samuel et al., 2011).
Strategies in addressing mental health amongst individuals with DM 2 experiencing FI is
limited amongst the studies. Thomas et al. (2018) recommend assessing the mental health status
of individuals experiencing FI and considering social work referrals, although the literature does
not provide any specific strategies.
Recommendations for practice.
Identifying mental health concerns such as depression and distress amongst individuals
with DM 2 is recommended by the CDG (Robinson et al., 2018). The CDG provide
recommendations as outlined below.
Identifying diabetes distress and major depression.
The CDG recommends screening for mental health in individuals with diabetes
throughout the lifespan (Robinson et al., 2018). There is no preferred screening tool in measuring
depression or distress, although it is important to distinguish between them (Robinson et al.,

66

2018). All tools recommended by the CDG for measuring depression and distress are roughly
equivocal in terms of sensitivity and specificity in the diabetic population (Robinson et al.,
2018). In particular, the Diabetes Distress Scale (DDS) can be used to identify individuals with
DM 2 at risk for, or experiencing, distress and a PHQ-9 can be used to identify individuals at risk
of, or experiencing symptoms of, a major depressive disorder (Robinson et al., 2018). Both
scales can assist in differentiating between distress and depression, as well as provide guidance
in practice (Robinson et al., 2018). There are no recommendations for the frequency of screening
in the literature review, nor are there recommendations in the guidelines. Future research should
examine whether similar strategies for identifying depression and distress are equally as affective
amongst individuals experiencing depression and distress secondary to FI.
Interventions for diabetes distress and major depression
The CDG recommends several non-pharmacological strategies that PCPs can incorporate
into practice that have evidence for improved mental health and diabetes control (Robinson et al.,
2018). This includes one-on-one individualized therapeutic strategies, cognitive behavioral
therapy, motivational interviewing, family interventions, increasing self-management and coping
skills, improving self-efficacy and stress management (Robinson et al., 2018). Further, there is
some evidence for the use of pharmacotherapy in treating depression amongst individuals with
DM 2, although caution must be weighed against the risk of glucose dysregulation. While
evidence exists for mental health interventions amongst the general DM 2 population, the
literature review found insufficient evidence promoting any specific therapeutic interventions to
ameliorate depression and distress secondary to FI in the DM 2 population. Future studies should
examine the efficacy of the interventions endorsed by the CDG in improving distress and
depression amongst individuals who are FI.

67

Changes in medication use.
Changes in overall medication use and safety was described in the literature set. Studies
noted medication underuse was significantly more likely to occur when FI was secondary to
finances (Essien et al., 2016; Ippolito et al., 2016). However, there is evidence to support that all
individuals with FI are significantly more likely not to fill their prescriptions, even when cost and
all other variables are adjusted for (Ippolito et al., 2016). Stretching out medication prescriptions,
delaying filling prescriptions or not using the medications at all have all been attributed to FI
(Billimek & Sorkin, 2012). While cost is implicated in medication underuse in individuals with
FIFC, the underlying mechanisms influencing all medication underuse behaviors remain
unknown.
Medication safety issues may also arise if certain medication regimens remain the same,
despite ongoing FI (Essien et al. 2016). FI is associated with a significantly increased risk of
hypoglycemia secondary to medication-glycemia mismatch (Essien et al., 2016). These findings
suggest that practitioners should be addressing medication use with all individuals who identify
as food insecure, and that patterns of medication use may be complex and interfering with
overall glycemic control.
Other unique patterns of medication use may arise when FI is secondary to financial
constraints (FIFC): Ippolito et al. (2016) found that food-medicine-supply trade offs may occur
when individuals cannot afford all three in order to manage their diabetes. As a result,
dysglycemia may occur if food consumption is maintained at the expense of discontinuing
medications. These findings highlight the importance of medicine reconciliation and addressing
medication regimens and adherence in the presence of any individual identified as FI.

68

While not a trend, one longitudinal study demonstrated that individuals with FI use more
insulin over time (Shalowitz et al., 2017). Insulin use is a predictor of poorly controlled diabetes
and increasing A1C over time. As a result, insulin use among individuals with FI was seen as a
consequence of poorly-controlled dysglycemia. These findings highlight the importance of FI
identification in practice, as there are significant consequences to increasing insulin use amongst
a population with inconsistent or insecure food consumption, or without understanding food
consumption behaviors.
Assessing a client’s medication regime, medication coverage, adherence to the regime,
and individual risk for and occurrence of hypoglycemia, can lead to basic patient-centered
pharmaco-education which may help guide diabetes pharmacotherapy (Thomas et al., 2018). By
assessing for hypoglycemia that is conscious of coverage, providers can replace
antihyperglycemics that pose the greatest risk of hypoglycemia with safer alternatives (Thomas
et al., 2018). Further, by assessing an individual’s medication regime, providers can encourage
insulin regimens which are more flexible, safe and patient-centered (Thomas et al., 2018).
Simple educational strategies can help promote safety, such as promoting medications with
meals wherever appropriate and withholding certain medications when there is no food intake
(Thomas et al., 2018). Lastly, it is prudent to consider less stringent glycemic targets that are
safer and conscious of an individual’s capacity to maintain glycemic control (Thomas et al.,
2018).
PCPs, including NPs, are ideally situated to address pharmacological strategies amongst
individuals with DM 2 experiencing FI. PCPs are primarily responsible for overseeing the
medical management of their patients, are familiar with Pharmacare coverage, and can
accommodate routine follow up and monitoring. While no recommendations specific to NPs

69

were retrieved from the literature set, NPs are specifically skilled at addressing difficult
conversations such as medication adherence, and patient difficulties in adopting behavior change
(Bartol, 2012).
Recommendations for Practice.
Medication reconciliation should ideally occur at every patient appointment as per CDG
(Lipscombe et al., 2018). Care should be taken to consider the dates medications were prescribed
via Pharmanet versus the amount of prescription remaining. Since the relationship of medication
use amongst individuals with FI is complex and may accompany feelings of distress, depression
and powerlessness, an open-ended, non-confrontational approach should be used to further
understand the unique patterns of behavior related to medication use.
FI appears to be closely associated with financial constraints in the Canadian context
(Tarasuk et al., 2016). With approximately 1 in 8 Canadians experiencing income-associated FI
([IAFI], Tarasuk et al., 2016), PCPs need to be cognizant of the interface between Pharmacarecovered medication reliance (Diabetes Canada, 201b), and safe prescribing practices in the
context of FI. Specific pharmacotherapy strategies to prevent hypoglycemia are included below,
are thoughtful of individual regimens, and are based on the recommendations from Thomas et al.
(2018).
Adjusting sulfonylureas.
If an individual with FI is presently taking sulfonylureas, the following considerations are
considered prudent: educate and reinforce that sulfonylureas work fairly rapidly and do not work
in a glucose-dependent method (Lipscombe et al., 2018). As recommended by Thomas et al.
(2018), if meals are skipped, encourage individuals to hold their sulfonylurea dose and to
consume only with meals. If hypoglycemia has been identified on at least one occasion, the PCP

70

should consider one of three interventions: (1) if the individual is taking glyburide, the PCP
should apply for Special Authority (SA) to substitute glyburide for an alternate sulfonylurea
associated with less risk of hypoglycemia, such as gliclazide (Lipscombe et al., 2018); gliclazide
is the preferred sulfonylurea due to its superior safety profile (Lipscombe et al., 2018); (2)
consider the A1C and whether targets needs to be softened, as sulfonylureas may need to be
discontinued, and whether increasing metformin to maximum tolerated doses is sufficient; (3)
apply for SA-approved DPP4 inhibitors, linagliptin or saxagliptin, in place of sulfonylureas
(Government of British Columbia, 2018b). Importantly, the SA for DPP4 inhibitors mandates
that NPH must be tried and found either ineffective or inappropriate (Government of British
Columbia, 2018e). Declaring NPH inappropriate on the application for DPP4 inhibitors is
reasonable given the evidence surrounding FI and inconsistent food intake (Essien et al., 2016), a
subsequent significant increased risk of hypoglycemia (Essien et al., 2016; Ippolito et al., 2016),
significant harms associated with hypoglycemia (Yale et al., 2018), and the potential for
nocturnal hypoglycemia with NPH (Lipscombe et al., 2018). This is supported by the Canadian
Diabetes Guidelines (CDG), which recommends that medications with potential for
hypoglycemia be avoided amongst individuals with high risk of hypoglycemia (Yale et al.,
2018). Furthermore, the CDG supports the use of newer agents such as DPP4 inhibitors due to
their safety profile and non-inferior A1C-lowering abilities (Lipscombe et al., 2018).
Changing insulin regimes.
If patients with FI are already taking insulins or require insulins, it would be prudent to
consider one of the following options: (1) if the treatment of an individual with DM 2 has
progressed in that they require exogenous intermediate acting insulin such as NPH, assess
whether nocturnal hypoglycemia has been identified on at least one occasion within the last

71

month. SA for long-acting insulin Detemir should be considered as per SA criteria (Government
of British Columbia, 2017f) for individuals experiencing hypoglycemia. As above, NPH presents
challenges to the individual with DM 2 experiencing FI, but Detemir does not have a peak effect
(Lipscombe et al., 2018); (2) if the treatment of an individual with DM 2 has progressed in that
they require exogenous short acting insulin, consider if rapid acting insulin is both affordable and
appropriate, as it poses less risk of postprandial hypoglycemia compared to short-acting insulins
(Lipscombe et al., 2018). Other considerations in keeping with Thomas et al.’s (2018) findings
include decreasing overall postprandial targets and insulin dosing, and educating on dose-holding
when meals are skipped.
Other strategies that enhance pharmacotherapy.
Evidence is limited for increasing SMBG frequency beyond what is recommended by the
CDG for each individual class of pharmacotherapy (Berard et al., 2018; Government of British
Columbia, 2019). However, the CDG does endorse SMBG as a strategy to identify, monitor and
prevent hypoglycemia, as well as promote the modification of healthy behaviors or adjust
medications (Berard et al., 2018). Unfortunately, diabetes self-management behaviors are
impaired as a result of FI (Essien et al., 2016; Ippolito et al., 2016; Walker et al., 2018).
Therefore, it remains unclear whether increased SMBG may help prevent hypoglycemia in light
of impaired capacity for self-management. If individuals are agreeable, willing and able to
increase SMBG, an individualized strategy could be implemented which assists the individual in
monitoring their glycemic levels where appropriate, in order to identify periods of hypoglycemia.
In consideration of the cost of blood glucose strips, PCPs encouraging this strategy should apply
for additional strips through Pharmacare as per criteria for additional test strips (Government of
British Columbia, 2019).

72

Theme Two: Hierarchy of Need and Motivation
A hierarchy of need and motivation unique to individuals experiencing FI was noted
throughout the literature. What is concerning amongst individuals with FI is that, unlike their
food secure counterparts, individuals with FI do not necessarily derive benefit from access to
comprehensive management by health care providers (Shalowitz et al., 2017). While the
mechanism is not entirely clear, authors Shalowitz et al. (2017) suggest there are social
determinants of diabetes control required for the successful management of individuals with FI
(Shalowitz et al. 2017). Inadequate food, social support and self-efficacy appear to interfere with
the individual’s capacity to manage his or her diabetes. As a result, these determinants contribute
to a hierarchy of need, each uniquely influencing an individual’s motivation and ability to
prioritize diabetes management.
When FI is present, eating patterns appear driven by physiological needs (Ippolito et al.,
2016), availability, and accessibility (Essien et al., 2016), rather than a desire for diabetes
control. These findings highlight the importance of understanding underlying motivators for food
consumption in the presence of FI, and a hierarchy of eating patterns driven by the level of food
security.
Importantly, stable access to food appears insufficient in supporting adequate selfmanagement (Ippolito et al., 2016). Social support emerged as an important requisite to diabetes
management amongst individuals with FI. Individuals with FI experience increased rates of
depression and distress, but social support appears to act as a buffer against the detrimental
effects of depression and distress on diabetes self-management (Walker et al., 2018). An
individual’s perception of consistent support from friends and family (Kollannoor-Samuel et al.,
2011), as well as positive social interactions while engaging in enjoyable activities with others

73

(Walker et al., 2018), was sufficient in mitigating depression and distress symptoms. General
recommendations for practitioners include assessing for, promoting, and fostering social support
(Thomas et al., 2018; Walker et al., 2018).
Individuals with FI have an overall decreased self-efficacy leading to poorer diabetes
self-management and A1C control (Gucciardi et al., 2009; Ippolito et al., 2016). The connection
between FI and decreased self-efficacy remains unclear, however, it is speculated that repeated
exposure to stress, distress, and depression arising from chronic FI promotes feelings of
diminished control and self-efficacy (Essien et al., 2016).
Recommendations for Practice.
While practical recommendations for managing DM 2 amongst individuals who are food
insecure were elicited from the literature, it remains unclear whether these recommendations
may mitigate the hierarchical structure of needs present in individuals experiencing FI. Future
research should focus on further understanding the relationship between food, social support and
self-efficacy, and examine whether other needs or social determinants are necessary for effective
diabetes management amongst individuals experiencing FI. The role of MTM as an underlying
mechanism may provide further insight into these relationships and serves as a framework for
understanding individual motivations and behaviors.
Theme Three: Food Insecurity Screening
Screening for FI can be successfully managed and implemented in the healthcare setting,
and has been found to be both appropriate and acceptable by patients and care providers (De
Marchis et al., 2019; Thomas et al. 2018). While evidence is limited to one pilot study, FI
screening has been shown to assist in guiding care, providing support, and bringing meaning and
understanding to patient self-management behaviors.

74

A specific and individualized approach for FI screening was preferred over any particular
screening method (Jones et al., 2013). This individualized approach should include: selecting a
clear definition of food insecurity useful to the specific purpose, identifying which attributes of
FI subsequently need to be measured, and finally, selecting a corresponding tool that measures
these elements (Jones et al., 2013).
Selecting a definition of food insecurity.
There was universal acknowledgement that FI results in interference of proper diabetes
management and control (Bawadi et al., 2012; Billimek & Sorkin, 2012; Essien et al., 2016;
Gucciardi et al., 2009; Ippolito et al., 2016; Kollannoor-Samuel et al., 2011; Montgomery et al.,
2017; Shalowitz et al., 2017; Walker et al., 2018). However, few antecedents could be identified
across various studies. Subsequently, it is unlikely that a universally acceptable definition of FI
exists. Rather the literature suggests that selecting a definition of FI which corresponds with
what is intended on being measured is preferred (Jones et al., 2013). For example, the Canadian
government has operationalized FI screening nationwide, but defines FI as income associated
(IAFI). They have operationalized this definition in order to identify a specific subpopulation
experiencing poverty for whom policy changes could be developed to reduce IAFI (Tarasuk et
al., 2016). How food insecurity is defined in a context sets the tone for how it will be measured
and identified, what problems are elicited, and which solutions are appropriate. A significant
limitation to this approach is that the majority of tools presently used measure IAFI, only thus
validated to measure IAFI (Marques et al., 2015). No singular way to validate a screening toolthe process is varied depending on what the tool is trying to measure

75

Identifying the attributes.
Based on the literature review, there are no comprehensive tools that measure all
attributes of FI (Ashby et al., 2016; Jones et al., 2013). This is likely in part to a lack of a
universal definition of FI. Instead, the literature highlighted that successful FI screening requires
the selection of target attributes intended on being examined (Jones et al., 2013). A focused
attribute approach prevents the wrong attribute from being measured, confounding attributes
from interfering with results, irrelevant information from being gathered that does not inform
management, and identifying attributes of FI that cannot be managed in the healthcare setting
(Jones et al., 2013). What is problematic with this approach is that the majority of tools are
validated to measure access to food as a primary attribute for FI (Ashby et al., 2016; Jones et al.,
2013). Future research should examine and develop new tools that measure other important
attributes particularly important to DM 2 management, such as nutritionally adequate food.
Screening tool selection.
While an individual context-specific approach is favored, several of the most widely used
tools such as the Household Food Security Scale are only validated to identify and measure IAFI
and very few other attributes (Marques et al., 2015). Further, Marques et al., (2015) highlight
that most FI screening tools have not been substantially analyzed or heavily validated (Marques
et al., 2015). Several consequences may occur if a FI screening method is incorrectly selected:
(1) the wrong determinant of FI is measured and the information gathered is not pertinent to the
cause; (2) the tool is validated for the wrong level of data being gathered, i.e. a national FI
screening tool is selected to identify individual FI; (3) repeat screening of FI is indicated but the
screening tool selected cannot be used to re-measure FI; and (4) the FI screening tool selected

76

identifies issues beyond available resources for sufficient collection, analysis and intervention
(Jones et al., 2013).
The literature review identified that one screening tool is unlikely to be able to address all
domains of FI and that FI tools are not universally transferrable across circumstances (Ashby et
al., 2016; Jones et al., 2013) Further, the majority of the tools have not been substantially
analyzed (Marques et al., 2015), therefore, selecting a tool appropriate to a healthcare setting
may be challenging. The recommended approach is to select a tool suited to a particular
program’s resources (Jones et al., 2013). For example, if practitioners agree to measure access
and poverty as key attributes to FI, then resources must be in place to support practitioners and
patients in increasing access to food, and diminishing financial burden.
Recommendations for Practice. 5
Ideally, clinics should identify a definition of FI and corresponding attributes that
resounds with their patient population and results in the provision of effective management
strategies. Practitioner limitations and capacity in managing certain attributes of FI should be
first identified before selecting certain FI tools. While it is important to select tools that are well
matched to the patient population, ethical issues may arise in identifying aspects of FI that cannot
be remedied. Additional research is needed to identify how clinics and primary care providers
operationalize effective identification and screening of FI.
Limitations and Future Research
There were several limitations to the integrative literature review findings. They are
discussed in the following sections: gaps in literature, and other limitations.

5

For a list of all available screening tools and their properties, please see Appendix C

77

Gaps in the Literature
The majority of FI screening methods available and acceptable for use in the healthcare
setting are only validated to identify FI secondary to financial constraints. For example, in a
systematic review examining the validity and acceptability of food insecurity (FI) screening tools
and their implementation in health care settings, De Marchis et al. (2019) found that all brief 1-3
question FI screening instruments examined, were all derived from and validated against the
United States Department of Agriculture-Food Security Survey (USDA-FSS). The USDA-FSS,
also known as the Household Food Security Survey (HFSS), is the most widely used, accepted
and validated tool in the United States and other countries, but it is validated only to identify FI
secondary to financial constraints (Marques et al., 2015). These findings are corroborated by
Makelarski et al., (2017), who examined the diagnostic accuracy and validity of two other shortform FI tools used in healthcare; and Ashby et al. (2016) whose systematic review concluded all
tools function to identify food access issues secondary to financial constraints. What is equally
problematic is that it is challenging to validate other attributes in FI surveys, because they are so
varied in what they measure (Jones et al., 2013). Therefore, the scope of these tools is
significantly narrowed to identify a subpopulation of individuals experiencing FI, and the
applicability of the tools in practice is significantly limited. Future research should identify or
establish new screening strategies that identify multiple determinants of food insecurity within
the healthcare setting.
Another limitation regarding the feasibility of implementing FI screening in primary care
is that the majority of measures or tools have not been rigorously appraised. In a recent
systematic review examining the psychometric properties of all available household FI measures,
Marques et al., (2015) found that there was limited appraisal for most tools used. Half of the

78

tools searched had only been tested or appraised in the literature 1-2 times (Marques et al.,
2015). Well-appraised tools were limited to identifying food insecurity secondary to financial
constraints (Marques et al., 2015). Additional research should critically appraise existing
screening tools available that have limited research.
The literature review supports the integration of FI screening in healthcare settings (De
Marchis et al., 2019; Thomas et al., 2018). However, the evidence is limited to qualitative
findings from one study (Thomas et al., 2018) and one systematic review (De Marchis et al.,
2019). A significant limitation of this integrative review is that there is no available literature
measuring the influence of FI screening on the management of DM 2 (Thomas et al., 2018) to
support its implementation. While the majority of participating patients and healthcare providers
supported screening implementation (De Marchis et al., 2019; Thomas et al., 2018), the benefits
of screening for FI in primary care are inferred and are, at best, speculative in nature.
There is a dearth of literature available examining the effects of specific interventions on
the management of DM 2 in individuals experiencing FI. The present practice recommendations
are based on the discussion section of cross-sectional studies, one focused literature review, and
one systematic review. The systematic review offered a limited discussion of the findings.
Therefore, the strength of findings and recommendations are significantly limited and should, at
present, be considered suggestions for practice. Several recommendations are based off of the
Canadian Diabetes Guidelines. Therefore, many of the interventions suggested are based on
general recommendations for the diabetic population. Future research should directly examine
the impact of both pharmacological and non-pharmacological interventions on the management
of DM 2 in individuals experiencing FI.

79

Other limitations in the literature identified by Jones et al. (2013) include discourse on
how to establish accurate “cut-offs” for a positive FI screen, and how often to screen for FI
(Jones et al., 2013). Since the definitions and causes for FI can be so diverse (Jones et al., 2013),
it remains unclear how implemented tools can integrate accurate cut-offs for identifying FI and
provide recommendations for repeat screening. Future research should focus on exploring these
tenets so that PCPs can confidently implement FI screening methods into practice.
Other Limitations and Considerations
An A1C, or glycemic control, is a surrogate marker and its value does not translate into
absolute risk. However, hard outcomes such as cardiovascular events, mortality and end-organ
damage require longitudinal studies and extensive funding. As a result, most studies rely on AIC
as a proxy measure for overall diabetes management. Previous studies caution estimating risk
using the A1C (Yudkin, Lipska, & Montori, 2011). However, evidence suggests that lowering
and controlling an A1C remains an important strategy in lowering the risk of micro and
macrovascular complications (Gorst et al., 2015; Mitsios, Ekinci, Mitsios, Churilov, & Thijs,
2018; O’Sullivan & Dinneen, 2008; Roussel, Steg, Mohammedi, Marre, & Potier, 2018). To
establish with certainty whether the current proposed strategies improve the management of DM
2 in individuals experiencing FI, future research is warranted that measures hard outcomes.
Time has been cited as a barrier to FI screening in healthcare by care providers.
Short form questionnaires (SFQ) can be used readily as they require little time to complete in
healthcare (De Marchis et al. 2019) and are among the few that have been rigorously analyzed.
An important trade-off to consider with SFQ are simplicity and comparability over the
comprehensive and contextual detail of longer questionnaires (Jones et al. 2013). Lastly, it is
important to note that most SFQ are validated to measure almost exclusively FI secondary to

80

financial constraints (Ashby et al., 2016a; Makelarski et al., 2017; Marques et al., 2015). Future
research should explore primary care providers’ experiences of time constraints and establish
acceptable time-appropriate strategies for identifying more diverse forms of FI.
Research is lacking into how PCPs can best use the hierarchy of needs and motivations
identified in the literature to support and develop an individualized self-management plan
cognizant of a patient’s capacities and motivations for behavior. Future research should examine
the perspectives and lived experiences of individuals with DM 2 and FI, which may help elicit
further detail about underlying needs and motivations, and viable management strategies.
Conclusion
The relationship between FI and DM 2 is complex. FI has profound effects on the
management of DM 2 but is difficult to identify, presenting a challenge for providers in
managing DM 2 in individuals experiencing FI. The harmful effects of poorly managed diabetes
secondary to FI emphasize the importance of identifying FI to enhance the management of, and
to prevent complications associated with, DM 2.
This integrative review examined how NPs can use FI screening in primary care to
enhance the management of individuals with DM 2 experiencing FI. A comprehensive literature
search highlighted relevant articles. Fifteen articles in total were selected for final review. The
selected articles were analyzed and the findings provided insight into how screening can be most
effective in identifying FI, and which management strategies in primary care may be most
effective for individuals with DM 2.
A synthesis of the findings from this review highlighted the importance of understanding
the influence of FI on food behaviors, mental health, and medication use. Through the support of
Maslow’s Theory of Motivation (MTM), a better understanding was achieved with respect to

81

motivations for behavior and diabetes self-management amongst individuals with FI. Key
findings from the literature review highlighted how FI reduces one’s relationship to food to basic
physiological need, superseding their capacity to use food for the dietary management of DM 2.
As a result, compensatory food behaviors develop that are destructive to glycemic control. While
efforts to supply stable food through low-cost alternatives or food banks are important, this
strategy alone is not sufficient. Strategies to manage FI-mediated changes in food behaviors
include individualized nutrition approaches cognizant of an individual’s personal limitations,
assessments for the risk and occurrence of hypoglycemia, and consideration of referrals to social
work and dieticians for additional support.
Another key theme was the influence of FI on mental health and the subsequent
challenges it presented in managing DM 2. The literature revealed that FI created a persistent
level of stress, depression and distress, ultimately affecting an individual’s capacity to engage in
self-care behaviors such as healthy eating, exercising, and self-monitoring of blood glucose,
which ultimately influenced the level of glycemic control. Furthermore, there was evidence that
distress and depression resulted in an increase of some unhealthy behaviors such as smoking.
This relationship appeared to be mitigated by social support. In particular, the presence of
informal social support of friends and family in social contexts was particularly effective at
mitigating depression. Strategies to manage the influence of mental health on the management of
DM 2 include assessing an individual’s level of social support, providing referrals or resources if
necessary and creating a patient centered treatment strategy that acknowledges the individual’s
mental health and their capacity to self-manage their diabetes. Lastly, the findings also indicated
that particular attention should be paid when financial constraints underlie FI, as this

82

subpopulation appears to suffer additional consequences and experience additional stressors
hampering their ability to self-manage their diabetes.
FI screening in primary care is both feasible and acceptable. At present, there exists no
comprehensive FI tools capturing all attributes of FI. Instead, the literature review suggests: (a)
carefully examining the population of interest, available resources and feasible interventions; (b)
selecting a definition and attributes of FI that reflect these key components; and (c) selecting a
screening tool validated to capture these attributes. Challenges in this process were also
identified. Several reviews indicated that, despite a diversity in FI conceptualization, the majority
of tools were only validated to measure food access secondary to financial constraints, and that
the majority of available tools also lacked rigorous appraisal.
Recommendations for future research were also identified in this integrative review.
Additional research should establish validated strategies that identify multiple attributes of FI
within the healthcare setting. Although the literature indicated that identifying FI in healthcare
may be feasible and appropriate, FI identification methods are almost exclusively limited to
identify FI secondary to financial constraints. Further, there is a dearth of literature examining
whether FI screening improves the management of DM 2. Similarly, no studies to date have
examined the benefit of specific interventions on the management of DM 2. Future research
should examine how FI screening directly benefits the management of DM 2 in primary care and
which interventions provide the most benefit. Additionally, few articles examined the
perspectives of individuals with DM 2 experiencing FI.
While a hierarchy of needs and underlying motivations for behavior were identified in the
literature, there is a lack of research into how PCPs can best use this knowledge to support
individuals with DM 2 experiencing FI, and how to develop an individualized self-management

83

plan, cognoscente of a patient’s capacities and motivations for behavior. These areas for future
research which help support PCPs in how best to identify FI in the PC setting and to enhance the
care they provide to individuals with DM 2 experiencing FI.
This integrative review highlighted how NPs as PCPs can play influential roles in
screening for FI and managing DM 2 among individuals with FI. The literature findings
highlight how PCPs can be key in identifying FI among their patients based on the trust and
rapport that develops. Further, NPs are undertaking difficult conversations with patients
surrounding the more vulnerable aspects of FI such as unhealthy food behaviors, stress,
depression and medication adherence. NPs are also managing pharmacological and nonpharmacological aspects of care, considering the social determinants of health, including food
insecurity, when providing care.
As the incidence and prevalence of DM 2 in Canada is projected to increase and FI
remains pervasive in our Canadian population NPs must strive to identify the unique struggles of
this subpopulation in primary care. The findings of this integrative literature review highlight
how FI influences the management of DM 2. By screening for FI in primary care, the findings of
this review can support NPs in identifying individuals with DM 2 experiencing FI. Ultimately,
the findings assist NPs better managing DM 2 and glycemic control amongst individuals who are
food insecure and contribute to the work in preventing diabetic complications.

84

References
Ashby, S., Kleve, S., McKechnie, R., & Palermo, C. (2016). Measurement of the dimensions of
food insecurity in developed countries: a systematic literature review. Public Health
Nutrition, 19(16), 2887–2896. doi.10.1017/S1368980016001166
Bartol, T. (2012). Improving the treatment experience for patients with type 2 diabetes: Role of
the nurse practitioner. Journal of the American Academy of Nurse Practitioners,
24(supplement), 270-276. doi: 10.1111/j.1745-7599.2012.00722.x
Bawadi, H. A., Ammari, F., Abu-Jamous, D., Khader, Y. S., Bataineh, S., & Tayyem, R. F.
(2012). Food insecurity is related to glycemic control deterioration in patients with type 2
diabetes. Clinical Nutrition, 31(2), 250–254. doi.10.1016/j.clnu.2011.09.014
Berard, L. D., Siemens, R., & Woo, V. (2018). Monitoring Glycemic Control. Canadian Journal
of Diabetes, 42, S47–S53. doi.10.1016/j.jcjd.2017.10.007
Berkowitz, S. A., & Fabreau, G. E. (2015). Food insecurity: What is the clinician’s role?
Canadian Medical Association. Journal: CMAJ; Ottawa, 187(14), 1031–1032.
Berkowitz, S. A., Karter, A. J., Corbie-Smith, G., Seligman, H. K., Ackroyd, S. A., Barnard, L.
S., … Wexler, D. J. (2018). Food insecurity, food “deserts,” and glycemic control in
patients with diabetes: a longitudinal analysis. Diabetes Care, 41(6), 1188–1195. doi: 10
.2337/dc17-1981
Beryl Pilkington, F., Daiski, I., Bryant, T., Dinca-panaitescu, M., Dinca-panaitescu, S., &
Raphael, D. (2010). The experience of living with diabetes for low-income Canadians.
Canadian Journal of Diabetes, 34(2), 119–126. doi.10.1016/S1499-2671(10)42008-0
Billimek, J., & Sorkin, D. H. (2012). Food insecurity, processes of care, and self-reported
medication underuse in patients with type 2 diabetes: results from the California Health

85

Interview Survey. Health Services Research, 47(6), 2159–2168. doi.10.1111/j.1475
-6773.2012.01463.x
Brashers, V. L., Jones, R. E., & Huether, S. E. (2014a). Alterations of hormonal regulation. In
K. McCance & S. Huether (Eds.), Pathophysiology: The biological basis for disease in
adults and children (7th ed., pp. 717-759). St. Louis, MO: Mosby Elsevier.
Brashers, V. L., Jones, R. E., & Huether, S. E. (2014b). Mechanisms of hormonal regulation. In
K. McCance & S. Huether (Eds.), Pathophysiology: The biological basis for disease in
adults and children (7th ed.) (pp. 689-712). St. Louis, MO: Mosby Elsevier.
British Columbia College of Nursing Professionals. (2018a). Entry-level competencies for Nurse
Practitioners in Canada (Pub. No. 806). Retrieved from https://www.bccnp.ca/
British Columbia College of Nursing Professionals. (2018b). Scope of Practice for Nurse
Practitioners (Pub. No. 688). Retrieved from https://www.bccnp.ca/
Canada Institute for Health Information. (2015). International comparisons: A focus on diabetes
(CIHI no 11116-0915). Retrieved from https://secure.cihi.ca/
Canadian Diabetes Association. (2009). An economic tsumani: the cost of diabetes in Canada.
Retrieved from https://www.diabetes.ca/
Canadian Diabetes Association and Diabetes Quebec. (2010). Diabetes: Canada at the tipping
point. Charting a new path. Retrieved from https://www.diabetes.ca/
Canadian Nurses Association. (2010). Canadian nurse practitioner: Core competency framework.
Retrieved from http://www.cnaaiic.ca/
Canadian Nurses Association. (2016). Position statement: The nurse practitioner. Retrieved from
https://www.cna-aiic.ca/

86

Carter, S., Clifton, P. M., & Keogh, J. B. (2016). The effects of intermittent compared to
continuous energy restriction on glycaemic control in type 2 diabetes; a pragmatic pilot
trial. Diabetes Research and Clinical Practice, 122, 106–112. doi.10.1016/j.diabres.2016
.10.010
Carryer, J., & Yarwood, J. (2015). The nurse practitioner role: solution or servant in improving
primary health care service delivery. The Collegian. doi: 10.1016/j.colegn.2015.02.004
Centre for Disease Control and Prevention. (2017). National Diabetes Statistics Report, 2017
(CDC Publication No CS279910-A). Retrieved from https: //www.cdc.gov/
Chabot, F., Caron, A., Laplante, M., & St-Pierre, D. H. (2014). Interrelationships between
ghrelin, insulin and glucose homeostasis: Physiological relevance. World Journal of
Diabetes, 5(3), 328–341. doi.10.4239/wjd.v5.i3.328
Chan, J., DeMelo, M., Gingras, J., & Gucciardi, E. (2015). Challenges of diabetes selfmanagement in adults affected by food insecurity in a large urban centre of Ontario,
Canada. International Journal of Endocrinology, 2015.doi.10.1155/2015/903468
DeFronzo, R. A. (2009). From the triumvirate to the ominous octet: a new paradigm for the
treatment of type 2 diabetes mellitus. Diabetes, 58, 773-795. doi:10.2337/db09-9028
De Marchis, E. H., Torres, J. M., Fichtenberg, C., & Gottlieb, L. M. (2019a). Identifying food
insecurity in health care settings: A systematic scoping review of the evidence. Family &
Community Health, 42(1), 20–29. doi.10.1097/FCH.0000000000000208
Diabetes Canada. (2019). About. Retrieved from https://www.diabetes.ca/
Diabetes Canada. (2018a). Diabetes Canada’s position on food insecurity. Retrieved from
https://www.diabetes.ca/
Diabetes Canada. (2018b). Formulary listings for diabetes medications in Canada. Retrieved

87

from https://www.diabetes.ca/
Diabetes Canada. (2018c). 2018 clinical practice guidelines quick reference guide (4165669-18).
Retrieved from http://guidelines.diabetes.ca/
Diabetes Canada. (2017a). Diabetes statistics in Canada. Retrieved from http://www.diabetes.ca/
Diabetes Canada. (2017b). 2017 Report on Diabetes in British Columbia. Retrieved from
http://www.diabetes.ca/
DiCenso, A., Bourgeault, I., Abelson, J., Martin-Misener, R., Kaasalainen, S., Carter, N., …
Kilpatrick, K. (2010). Utilization of nurse practitioners to increase patient access to
primary healthcare in Canada-thinking outside the box. Nursing Leadership (Toronto,
Ont.), 23 Spec No 2010, 239–259.
Donald, F., Martin-Misener, R., Carter, N., Donald, E. E., Kaasalainen, S., Wickson-Griffiths,
A., . . . DiCenso, A. (2013). A systematic review of the effectiveness of advanced
practice nurses in long-term care. Journal of Advanced Nursing, 69, 2148-2161. doi:10
.1111/ jan.12140
El-Masri, M. M. (2013). Recall bias. Canadian Nurse, 109(9), 14–14.
Essien, U. R., Shahid, N. N., & Berkowitz, S. A. (2016). Food insecurity and diabetes in
developed societies. Current Diabetes Reports, 16(9), 79-87. doi:10.1007/s11892-0160774-y
Ferguson, S., Swan, M., & Smaldone, A. (2015). Does diabetes self-management education in
conjunction with primary care improve glycemic control in Hispanic patients? A
systematic review and meta-analysis The Diabetes Educator, 41(4), 472-484. doi:
10.1177/0145721715584404

88

Galesloot, S., McIntyre, L., Fenton, T., & Tyminski, S. (2012). Food insecurity in Canadian
adults: Receiving diabetes care. Canadian Journal of Dietetic Practice and Research,
73(3), e261–e266. doi.10.3148/73.3.2012.e261
Government of British Columbia. (2018a). About Pharmacare. Retrieved from https://www2
.gov.bc.ca/gov/
Government of British Columbia. (2018b). BC Pharmacare formulary search. Retrieved from
https://pharmacareformularysearch.gov.bc.ca/
Government of British Columbia. (2018c). Fair Pharmacare calculator. Retrieved from http:
//www.health.gov.bc.ca/
Government of British Columbia. (2018d). Special authority request for linagliptin, pioglitazone,
saxagliptin. Retrieved from https://www2.gov.bc.ca/
Government of British Columbia. (2018e). Special authority request for insulin detemir.
Retrieved from https://www2.gov.bc.ca/
Government of British Columbia. (2018f). Medical services plan (MSP). Retrieved from https:
//www2.gov.bc.ca/
Government of British Columbia. (2018g). Special authority coverage. Retrieved from https:/
/www.2.gov.bc.ca/
Government of British Columbia. (2019). Blood glucose testing and Pharmacare annual
quantity limits. Retrieved from https://www2.gov.bc.ca/
Government of Canada. (2018). Social determinants of health and health inequities. Retrieved
from https://www.canada.ca/
Government of Canada. (2017). Canadian community health survey. Retrieved from https://
www.canada.ca/

89

Gray, J. R. (2017). Interpreting research outcomes. In J. Gray, S. Grove, & S. Sutherland,
(Eds.), Burns and Grove’s the practice of Nursing Research: Appraisals, Synthesis, and
Generation of Evidence (8th ed., pp.192-216). St. Louis: MO: Elsevier
Greiver, M., Williamson, T., Barber, D., Birtwhistle, R., Aliarzadeh, B., Khan, S., … Katz, A.
(2014). Prevalence and epidemiology of diabetes in Canadian primary care practices: a
report from the Canadian Primary Care Sentinel Surveillance Network. Canadian
Journal of Diabetes, 38(3), 179–185. doi.10.1016/j.jcjd.2014.02.030
Gucciardi, E., Vahabi, M., Norris, N., Del Monte, J. P., & Farnum, C. (2014). The intersection
between food insecurity and diabetes: a review. Current Nutrition Reports, 3(4), 324–
332. doi.10.1007/s13668-014-0104-4
Gucciardi, E., Vogt, J. A., DeMelo, M., & Stewart, D. E. (2009). Exploration of the relationship
between household food insecurity and diabetes in Canada. Diabetes Care, 32(12), 2218–
2224. doi.10.2337/dc09-0823
Hamm, M. W., & Bellows, A. C. (2003). Community food security and nutrition educators.
Journal of Nutrition Education and Behavior, 35(1), 37–43. doi:10.1016/S1499-4046
(06)60325-4
Health Canada. (2007). Eating well with Canada’s food guide. Retrieved from
https://www.canada.ca/
Health Canada Office of Nutrition Policy and Promotion. (2007). Canadian Community Health
Survey Cycle 2.2, Nutrition (2004): Income-related household food security in Canada.
Ottawa, ON: Ministry of Health
Health Professions Act, c. 183 of Public Statutes and Regulations, (RSBC, 1996).

90

Heerman, W. J., Wallston, K. A., Osborn, C.Y., Bian, A., Schlundt, D.G., Barto S.D., &
Rothman, R.L. (2016). Food insecurity is associated with diabetes self-care behaviours
and glycaemic control. Diabetic Medicine, 33(6), 844-850. doi: 10.1111/dme.12896.
Imran, S. A., Agarwal, G., Bajaj, H. S., & Ross, S. (2018). Targets for Glycemic Control.
Canadian Journal of Diabetes, 42, S42–S46. doi.10.1016/j.jcjd.2017.10.030
Ippolito, M., Lyles, C., Prendergast, K., Berger Marshall, M., Waxman, E., & Kessler Seligman,
H. (2016). Food insecurity and diabetes self-management among food pantry clients.
Public Health Nutrition, 1, 1–7. doi.10.1017/S1368980016001786
Jones, A. D., Ngure, F. M., Pelto, G., & Young, S. L. (2013). What are we assessing when we
measure food security? A compendium and review of current metrics. Advances in
Nutrition, 4(5), 481–505. doi.10.3945/an.113.004119
Kleinridders, A., Ferris, H. A., Cai, W., & Kahn, C. R. (2014). Insulin action in brain regulates
systemic metabolism and brain function. Diabetes, 63(7), 2232–2243. doi.10.2337/db14
-0568
Kollannoor-Samuel, G., Wagner, J., Damio, G., Segura-Pérez, S., Chhabra, J., Vega-López, S.,
& Pérez-Escamilla, R. (2011). Social support modifies the association between household
food insecurity and depression among latinos with uncontrolled type 2 diabetes. Journal
of Immigrant and Minority Health, 13(6), 982–989. doi.10.1007/s10903-011-9499-9
Li, N., Dachner, N., & Tarasuk, V. (2016). Priority health equity indicators for British
Columbia: Household food insecurity indicator report. 44.
Lipscombe, L., Booth, G., Butalia, S., Dasgupta, K., Eurich, D. T., Goldenberg, R., … Simpson,
S. (2018). Pharmacologic glycemic management of type 2 diabetes in adults. Canadian
Journal of Diabetes, 42, S88–S103. doi.10.1016/j.jcjd.2017.10.034

91

Mackey, S., Hatcher, D., Happell, B., & Cleary, M. (2013). Primary health care as a
philosophical and practical framework for nursing education: Rhetoric or reality?
Contemporary Nurse, 45(1), 79–84. doi.10.5172/conu.2013.45.1.79
Makelarski, J. A., Abramsohn, E., Benjamin, J. H., Du, S., & Lindau, S. T. (2017). Diagnostic
accuracy of two food insecurity screeners recommended for use in health care settings.
American Journal of Public Health, 107(11), 1812–1817. doi.10.2105/AJPH.2017
.304033
Marques, E. S., Reichenheim, M. E., de Moraes, C. L., Antunes, M. M. L., & Salles-Costa, R.
(2015). Household food insecurity: a systematic review of the measuring instruments
used in epidemiological studies. Public Health Nutrition, 18(5), 877–892. doi.10.1017
/S1368980014001050
Maslow, A. H. (1943). A theory of human motivation. Psychological Review, 50(4), 370–396.
doi.10.1037/h0054346
Meek, T. H., & Morton, G. J. (2016). The role of leptin in diabetes: metabolic effects.
Diabetologia, 59(5), 928–932. doi.10.1007/s00125-016-3898-3
Montgomery, J., Lu, J., Ratliff, S., & Mezuk, B. (2017). Food insecurity and depression among
adults with diabetes: Results from the National Health and Nutrition Examination Survey
(NHANES). The Diabetes Educator, 43(3), 260–271. doi.10.1177/0145721717699890
Nield, L., Moore, H., Hooper, L., Cruickshank, K., Vyas A., Whittaker, V., & Summerbell, C. D.
(200). Dietary advice for the treatment of type 2 diabetes mellitus in adults (Article.
No. CD004097). Cochrane Database of Systematic Reviews, (3). doi: 10.1002/14651858
.CD004097.pub4.

92

Nurse and Nurse Practitioners of British Columbia (NNPBC, 2018). About Nurse Practitioners.
Retrieved from https://www.nnpbc.com/
Patil, S. P., Craven, K., & Kolasa, K. M. (2017). Food insecurity: It is more common than you
think, recognizing it Can improve the care you give. Nutrition Today, 52(5), 248–257.
doi.10.1097/NT.0000000000000232
Petersen, M. C., Vatner, D. F., & Shulman, G. I. (2017). Regulation of hepatic glucose
metabolism in health and disease. Nature Reviews. Endocrinology, 13(10), 572–587.
doi.0.1038/nrendo.2017.80
Prodan-Bhalla, N. & Scott, L. (2016). Primary care transformation in British Columbia: A new
model to integrate Nurse Practitioners. British Columbia Nurse Practitioner Association.
Discussion Paper. Retrieved from https://bcnnp.org/
PROOF. (2018a). Household food insecurity in Canada. Retrieved from https://proof.utoronto
.ca/
PROOF. (2018b). Household food insecurity in Canada: A guide to measurement and
interpretation. Retrieved from https://proof.utoronto.ca/
Public Health Agency of Canada (PHAC). (2011). Diabetes Canada: Facts and figures from a
public health perspective (online category no. HP35-25/2011E-PDF). Retrieved from
https://www.canada.ca/
Punthakee, Z., Goldenberg, R., & Katz, P. (2018). Definition, classification and diagnosis of
diabetes, prediabetes and metabolic syndrome. Canadian Journal of Diabetes, 42, S10–
S15. doi.10.1016/j.jcjd.2017.10.003
Robinson, D. J., Coons, M., Haensel, H., Vallis, M., & Yale, J.-F. (2018). Diabetes and mental
health. Canadian Journal of Diabetes, 42, S130–S141. doi.10.1016/j.jcjd.2017.10.031

93

Russell, G. M., Dahrouge, S., Hogg, W., Geneau, R., Muldoon, L., & Tuna, M. (2009).
Managing chronic disease in Ontario primary care: The impact of organizational factors.
Annals of Family Medicine, 4, 309-318. doi:10.1370/afm.982
Sangster-Gormley, E., Canitz, B., Schreiber, R., Borycki, E., Frisch, N., Kuo, A., ...Feddema, A.
(2015). An Evaluation of the Integration of Nurse Practitioners into the British Columbia
Healthcare System (pp.1-35). Retrieved from https://www.msfhr.org/
Seligman, H. K., Jacobs, E. A., López, A., Tschann, J., & Fernandez, A. (2012). Food insecurity
and glycemic control among low-income patients with type 2 diabetes. Diabetes Care,
35(2), 233–238. doi.10.2337/dc11-1627
Seligman, H. K., & Schillinger, D. (2010). Hunger and socioeconomic disparities in chronic
disease. New England Journal of Medicine, 363(1), 6–9. doi.10.1056/NEJMp1000072
Shalowitz, M. U., Eng, J. S., McKinney, C. O., Krohn, J., Lapin, B., Wang, C.-H., & Nodine, E.
(2017). Food security is related to adult type 2 diabetes control over time in a United
States safety net primary care clinic population. Nutrition & Diabetes, 7(5), e277.
doi.10.1038/nutd.2017.18
Sherifali, D., Berard, L. D., Gucciardi, E., MacDonald, B., & MacNeill, G. (2018). Selfmanagement education and support. Canadian Journal of Diabetes, 42, S36–S41.
doi.10.1016/j.jcjd.2017.10.006
Sherifali, D., Rabi, D. M., McDonald, C. G., Butalia, S., Campbell, D. J. T., Hunt, D., …
Houlden, R. L. (2018). Methods. Canadian Journal of Diabetes, 42, S6–S9.
doi.10.1016/j.jcjd.2017.10.002

94

Sievenpiper, J. L., Chan, C. B., Dworatzek, P. D., Freeze, C., & Williams, S. L. (2018). 2018
Clinical practice guidelines: nutrition therapy. Canadian Journal of Diabetes, 42(2018),
S64-S79. doi:10.1016/j.jcjd.2017.10.009
Shah, B. R. (2013). Diabetes in visible minority populations in Ontario. Healthcare Quarterly,
16(4), 14-17. doi:10.12927/hcq.2014.23660
Sigal, R. J., Armstrong, M. J., Bacon, S. L., Boulé, N. G., Dasgupta, K., Kenny, G. P., & Riddell,
M. C. (2018). Physical activity and diabetes. Canadian Journal of Diabetes, 42, S54–
S63. doi.10.1016/j.jcjd.2017.10.008
Silverman, J., Krieger, J., Kiefer, M., Hebert, P., Robinson, J., & Nelson, K. (2015). The
relationship between food insecurity and depression, diabetes distress and medication
adherence among low-income patients with poorly-controlled diabetes. Journal of
General Internal Medicine, 30(10), 1476–1480. doi.10.1007/s11606-015-3351-1
Stinkens, R., Goossens, G. H., Jocken, J. W. E., & Blaak, E. E. (2015). Targeting fatty acid
metabolism to improve glucose metabolism. Obesity Reviews, 16(9), 715–757.
doi.10.1111/obr.12298
Statistics Canada. (2016). Diabetes in Canada. Retrieved from http://www.statcan.gc.ca/
Statistics Canada. (2017). Health fact sheet: Diabetes, 2016. Retrieved from
https://www150.statcan.gc.ca/
Statistics Canada. (2018a). Market basket measures (MBM) thresholds for economic families
and persons not in economic families, 2015. Retrieved from https://www12.statcan.gc.ca/
Statistics Canada. (2018b). Measuring low income and Canada’s official poverty line. Retrieved
from https://www.statcan.gc.ca/

95

Strickhouser, S., Wright, J.D., & Donley, A.M. (2014). Food insecurity among older adults: Full
report. Retrieved from the American Association for Retired Persons website: https:
//www.aarp.org/
Sutherland, S. (2017a). Introduction to quantitative research. In J. Gray, S. Grove, & S.
Sutherland, (Eds.), Burns and Grove’s The practice of Nursing Research: Appraisals,
Synthesis, and Generation of Evidence (8th ed., pp.37-61). St. Louis: MO: Elsevier
Sutherland, S. (2017b). Quantitative methodology: noninterventional designs and methods. In J.
Gray, S. Grove, & S. Sutherland, (Eds.), Burns and Grove’s The practice of Nursing
Research: Appraisals, Synthesis, and Generation of Evidence (8th ed., pp.192-216). St.
Louis: MO: Elsevier
Tait, C. A., L’Abbé, M. R., Smith, P. M., & Rosella, L. C. (2018). The association between food
insecurity and incident type 2 diabetes in Canada: A population-based cohort study.
PLOS ONE, 13(5), e0195962. doi:10.1371/journal.pone.0195962
Tarasuk, V., Cheng, J., de Oliveira, C., Dachner, N., Gundersen, C., & Kurdyak, P. (2015).
Association between household food insecurity and annual health care costs. CMAJ :
Canadian Medical Association Journal, 187(14), E429–E436. doi.10.1503/cmaj.150234
Tarasuk, V., Mitchell, A., & Dachner, N. (2016). Household food insecurity in Canada: Research
to Identify Policy Options to Reduce Food Insecurity (PROOF), 20(4), 299–312.
doi.10.1097/00008486-200510000-00003
Tasyurek, H. M., Altunbas, H. A., Balci, M. K., & Sanlioglu, S. (2014). Incretins: Their
physiology and application in the treatment of diabetes mellitus. Diabetes/Metabolism
Research and Reviews, 30(5), 354–371. doi.org/10.1002/dmrr.2501

96

The National Rural Health Alliance. (2016). Food security and health in rural and remote
Australia (Pub No 16/053). Retrieved from the Rural Industries Research and
Development Cooperation at: https://www.agrifutures.com.au/
The World Health Organization. (2018). The state of food security and nutrition in the world
(pp.1-151). Retrieved from the World Health Organization website: http://www.who.int/
Thomas, B., Fitzpatrick, S., Sidani, S., & Gucciardi, E. (2018). Developing and implementing a
food insecurity screening initiative for adult patients living with type 2 diabetes.
Canadian Journal of Diabetes, 42(3), 257–262. doi.10.1016/j.jcjd.2017.06.004
Vas, A., Devi, E.S., Vidyasagar, S., Acharya, R., Rau, N.R., George A., … & Nayak, B.
(2017). Effectiveness of self‐management programmes in diabetes management: A
systematic review (e12571). International Journal of Nursing Practice, 23(5), 1-8. doi:10
.1111 /ijn.12571.
Viguiliouk, E., Stewart, S. E., Jayalath, V. H., Ng, A. P., Mirrahimi, A., de Souza, R. J., …
Sievenpiper, J. L. (2015). Effect of replacing animal protein with plant protein on
glycemic control in diabetes: A systematic review and meta-analysis of randomized
controlled trials. Nutrients, 7(12), 9804–9824. doi.10.3390/nu7125509
Walker, R. J., Williams, J. S., & Egede, L. E. (2018). Pathways between food insecurity and
glycaemic control in individuals with type 2 diabetes. Public Health Nutrition, 21(17),
3237–3244. doi.0.1017/S1368980018001908
Wharton, S., Pedersen, S. D., Lau, D. C. W., & Sharma, A. M. (2018). Weight Management in
Diabetes. Canadian Journal of Diabetes, 42, S124–S129. doi.10.1016/j.jcjd.2017.10.015
Yale, J.-F., Paty, B., & Senior, P. A. (2018). Hypoglycemia. Canadian Journal of Diabetes, 42,
S104–S108. doi.10.1016/j.jcjd.2017.10.010

97

Yazıcı, D., & Sezer, H. (2017). Insulin Resistance, Obesity and Lipotoxicity. In A. B. Engin &
A. Engin (Eds.), Obesity and Lipotoxicity (pp. 277–304). doi.10.1007/978-3-319-483825_12

Appendix A
Search Strategy: Aim One and Two

Adapted from Prisma Flow Diagram Generator (2019)

99

Search Strategy: Aim 3

Adapted from Prisma Flow Diagram Generator (2019)

Appendix B
Literature Review Matrix
Authors, Date
& Title
Essien, Shahid
& Berkowitz
(2016)
Food Insecurity
and Diabetes in
Developed
Societies

Aim
Examine
evidence
surrounding
food insecurity
and its
connection to
diabetes and
determine what
connections
exist.

Study Design &
Location
Focused literature
review
Unknown
No limitations on
location

Methods & Analysis

Key Findings:

Strengths & Limitations

Definition used:
“Food insecurity refers
to the uncertain or
limited access to
adequate and safe
foods”.

More likely to have poorly
controlled diabetes due to:
(1) Poor diet choices and
compensatory changes in dietary
habits:
• Binge-fast eating depending on
pay cycles or food availability/
affordability
• Incentive to consume saturated
fatty foods, simple
carbohydrates, highly
processed foods since they are
cheaper and consumed more
readily
• Less consumption of fruits and
vegetables
• place individuals at risk for
both hyper and hypoglycemia
(2) When FI is secondary to
finances, significantly increased
likelihood of medication underuse
due to costs (3) Significantly
increased risk of hypoglycemic
events when food becomes scarce
and medication regimens remain
unchanged

Strengths

Articles published on
or later than January
1st, 2013 were included
Cross-sectional,
longitudinal, and
interventional designs
were included
PubMed and Medical
Subject Headings
(MeSH)
terms were used.
Search strategy:
diabetes mellitus OR
prediabetic state OR
abdominal obesity
AND Food
Insecurity OR Food
Insufficiency OR Food
Supply OR
Hunger OR Poverty.
Relevant studies from
the reference section of

(4) Other proposed mechanisms
• Mental health (stress,
depression and distress) from
competing demands between
food and managing disease
with little income may

Based on references, covered a
significant amount of relevant
articles
Limitations
How FI is defined creates
classification bias
No description of
inclusion/exclusion criteria, how
many studies included or how
many articles informed each
conclusion
Cross-sectional, longitudinal,
and interventional designs
reviewed only


Cross-sectional study
designs not causational
cannot help explain
complex relationship
between food insecurity,
diabetes and other elements

Difficult to generalize some
results based on:
 specific population subgroups

101
included articles
included

Authors, Date
& Title

Aim

Walker,
Williams &
Egede (2018)

recognize direct
and indirect
pathways that
food insecurity
influences
diabetes self
care behaviors
and glycemic
control in
individuals with
diabetes

Pathways
between food
insecurity and
glycaemic
control in
individuals with
type 2 diabetes

Study Design,
Sample &
Location
Cross-sectional
survey
N= 615 Englishspeaking adults
with type 2
diabetes (DM 2)
Southeast USA

Methods & Analysis
Definition of FI used:
“Inability to or
limitation in accessing
nutritionally adequate
foods, or dependence
on emergency food
supplies”
Food insecurity
measured using the US
Household
Food Security Survey
Module short form
scale
The Perceived Stress
Scale (PSS) was used
to measure stress and is
highly correlated with
stress, depression and
anxiety.
Social support
measured using
Medical Outcomes
Study (MOS) Social
Support Survey

decrease self-efficacy and selfmanagement of diabetes:
(a) Reduced mental capacity
from the stress of managing
diabetes in the face of FI
(b) creates significant distress
(due feelings of to
powerlessness)
Decreased likelihood of
accessing healthcare
Key Findings:
FI is directly
related to increased HB A1C
through:
 increased stress
 other mechanisms independent
of stress
FI indirectly related increased A1C
through poorer self-care mediated
by
 increased stress
 lower social support.
Findings suggest
that addressing stress and providing
support for individuals with DM 2
experiencing FI may improve the
individual’s capacity to perform
diabetes
self-care behaviours.



healthcare coverage,
insurance or circumstance in
each state or country.

Recall bias for studies suing
questionnaires
CASP rating:
Low
Strengths & Limitations
Strengths
Calculated power based on
specific model used N=615
appropriate sample size.
All exposures (stress, support, FI
clearly defined and measured
using well-validated
questionnaires
Statistical significance between
0.05-0.0001
Limitations
Cross sectional survey usedcannot draw causational results
Recall and reporting bias for
questionnaires

There are direct pathways
independent of self-care by which
FI impacts glycemic control.

Difficult to generalize from
study population

Results suggest

CASP rating:
High

102

Summary of Diabetes
Self-Care Activities
(SDSCA) scale
measured selfmanagement/self-care
activities.
Morisky
Medication Adherence
Scale (MMAS)
measured medication
adherence






a multipronged approach to FI
in individuals with DM2.
Addressing nutrition alone in
individuals with DM 2
experiencing FI is insufficient
recommends additional
pathways be targeted
social support acts as buffer in
poor diabetes self-management

Glycemic control was
measured using A1C

Authors, Date
& Title

Aim

Montgomery,
Lu, Ratliff &
Mezuk
(2017)

(1) estimate
prevalence of
food
insecurity (FI)
among US
adults with
normoglycemia,
prediabetes,
and diabetes

Food Insecurity
and Depression
Among Adults
with Diabetes:
Results From

Study Design,
Sample &
Location
Cross sectional
design
Adults ≥20 years
old with diabetes
or prediabetes:
 as per fasting
plasma
glucose
parameters in
the American

Structural equation
modelling (SEM) used
to examine
mechanisms through
which food insecurity
influences
diabetes self-care
behaviours and
glycaemic control
Methods & Analysis
Four years of data
taken from 2011-2014
NHANES. (National
Health and Nutrition
Examination
Survey).
FI defined as “Food
insecurity
refers to the lack of a
dependable

Key Findings:

Strengths & Limitations

Approximately 1 in 3 adults with
diabetes experiences food
insecurity
 10.2% mild FI and 20.3%
severe FI
Dose-response relationship exists
between FI and depression

Strengths

(1) mild food insecurity associated
with 2.5-fold increase risk of
depression

Selection bias minimized based
on selection strategy (using
NHANES)

Validated questionnaires used
C-statistic values used to
determine predictive capability
and study met baseline criteria

103
the National
Health and
Nutrition
Examination
Survey
(NHANES)

(2) determine
whether a
difference
existed in the
prevalence of
FI and major
depression
Across groups.



Diabetes
Guidelines, or
self-report

N=1724 for
diabetes
group
N=2004 for
prediabetes group
Control: N=7713
depression among
adults ≥20 with
normoglycemia
United States of
America

means of obtaining
safe, sufficient, and
nutritious food”
Depressive symptoms
measured using PHQ-9
Global Physical
Activity Questionnaire
used to measure
physical activity
Smoking measured
using self-reporting
(never, previous,
current)

(2) severe food insecurity
associated with 3.5-fold increase in
depression.
Increased smoking rates and
decreased physical activity only
partly explains depression amongst
adults with DM 2 experiencing FI

Recall bias attenuated by
including subjective and
objective measures. I.e. not
100% based on questionnaires
and recall
NHANES gathers data using
mobile clinic-i.e. same
measurement tools and
personnel=minimize
measurement bias
Results clinically significant
based on OR for depression
amongst mild and severe FI

Multiple logistic
regression models used
to evaluate relationship
between FI and
depression.

P-value 0.001

Covariates
measured/controlled
for by using multiple
regression analysis
 includes age,
gender,
race/ethnicity,
educational
attainment,
household income,
smoking, physical
activity

Generalizability restricted by
sample limitations
 Certain ethnic groups
excluded
 Limited to the United States

Limitations
Cross sectional survey usedcannot draw causational results

Pre-diabetes and diabetes groups
were combined affecting validity
of effect on DM2:
Recall bias with PHQ- 9
questionnaires
CASP rating:
High

104
Authors, Date
& Title

Aim

Bawadi,
Ammari AbuJamous,
Khader,
Bataineh,
Tayyeme,
(2011)

Determine
prevalence of
food insecurity
(FI)among type
2 diabetics
(DM2) in a
major Jordanian
hospital and to
investigate its
relation to
glycemic
control.

Food insecurity
is related to
glycemic
control
deterioration in
patients with
type 2
diabetes

Study Design,
Sample &
Location
Cross-sectional
design
843 patients (327
male and 516
female) between
22-84 years old
with confirmed
diagnosis of type 2
diabetes who were
patients in the
endocrinology
clinic at King
Abdullah
University
Hospital (KAUH)
in Al-Ramtha.
Northern Jordan

Methods & Analysis

Key Findings:

Definition of FI used:
“food insecurity is
defined as limited or
uncertain
availability of
nutritionally adequate
and safe foods in
socially
acceptable ways”

(1) gender, age, education,
and income are all related to food
security status

Interview-based
questionnaire used to
gather socioeconomic
and health data
Weight and height
measured by a trained
nutritionist.
food frequency
questionnaire (FFQ)
used for dietary
assessment.
food processor
software used to
process dietary data.
Short form of the U.S.
food security survey
used to identify FI
Food insecurity
categorized as
moderately food
insecure (MFIS) and
severely food

(2) Glycemic control worse in
MFIS and SFIS
(3) 70% of SFIS stated they often
could not afford to eat a balanced
meal.
(4) 99% of the SFIS run out of food
by the end of the month, and either
meal skip, or cut meal size as a
compensatory mechanism for
running out of food.
(5) SFIS associated with
significantly higher average body
mass index (2 BMI units)
(6) No difference in overall intake
of calories between all groups but
SFIS consumed more sugar, on
average, than MFIS and FS
counterparts

Strengths & Limitations
Strengths
Previous studies confirm that the
US short-form questionnaire has
been validated to be used in
developing countries
FFQ locally validated by asking
20 volunteer diabetic patients to
review and with multiple day
diaries.
200 Jordanian-specific items
were added
Limitations
No power calculation
Food frequency questionnaires
are not 100% correlative with
true diets
Important covariates not
controlled for:
alcohol intake and smoking
.
Difficult to generalize based on
Jordanian population, food costs,
economical circumstances
Self-report creates recall bias
Cross-sectional surveys cannot
determine causality
The general questionnaire was

105
insecure (SFIS) using
standardized protocol

not standardized or validated. It
was developed by the research
team

Glycosylated
hemoglobin (HbA1c)
used to measure
glycemic control

Selection bias:
 screening patients in
hospital versus diabetics in
general population seeking
primary care

Chi-square, and posthoc analysis of
variance was used as to
analyze the data
Authors, Date
& Title

Aim

Bilimek &
Sorkin, (2012)

To establish
whether there is
an independent
association of
food insecurity
to processes of
care or delays
in filling
prescriptions

Food
Insecurity,
Processes of
Care, and
Self-Reported
Medication
Underuse in
Patients with
Type 2
Diabetes:
Results
from the
California
Health
Interview
Survey

Study Design,
Sample &
Location
Cross sectional
design
N = 3,401
Adults selected
from the 2007
Public Use File of
the California
Health Interview
Survey (CHIS), (a
randomized
process) and
(1) had a diagnosis
of type 2 diabetes
(2) saw a care
provider over the
last year (3) were
presently taking
diabetes
medications
California

CASP Rating:
Moderate/low

Methods & Analysis

Key Findings:

Strengths & Limitations

Multivariable logistic
regression analyses that
adjusted for age,
gender, education,
race/ethnicity, factors
related to access to
care: insurance status,
number of visits to the
doctor in the previous
year, and access to an
automobile), duration
of diabetes, general
health condition, and
psychological distress.

(1) Individuals with food insecurity
are significantly more likely not to
fill their prescriptions even when
other variables are adjusted for

Strengths
Multivariable logistic regression
analyses adjusted for several
factors

(2) food insecurity may
independently interfere with
optimal diabetes management

Limitations

Process of care defined
by completing
hemoglobin A1c, a
dilated eye exam,
and a foot exam over
the last year

(3) Food insecurity does not
interfere with process of care

How FI is defined creates
classification bias
-Assumes below poverty line are
the only people with food
insecurity. Does not take into
account other reasons for not
being ale to afford medications.
Does not address other
medication behaviors connected
with food
insecurity such as medication
underuse
skipping doses, and
discontinuing medications
Does not measure power

106
Timing of prescription
pick up measured
using:
(1) the Medication
Expenditures Panel
Survey (Agency for
Healthcare
Research and Quality
[AHRQ] 2004) and a
questionnaire asking
about the nature of
delays in filling scripts

Authors, Date
& Title

Aim

Gucciardi,
Vogt, Demelo,
& Stewart
(2009)

(1) determine
the
prevalence of
food insecurity
amongst
Canadians with
diabetes
(2)
examine the
relationship
between food
insecurity
and diabetes
management

Exploration of
the
Relationship
Between
Household
Food Insecurity
and Diabetes in
Canada

Study Design,
Sample &
Location
Cross-sectional
design
N= 2,523
Ontarians to
examine the
relationship of FI
to management,
health status, and
self-care
(management)
practices
N= 6, 237 from
throughout Canada
to examine the

Binary scoring used to
determine the presence
of food insecurity
amongst individuals
with household income
less than 200 percent of
the federal poverty
level.
Methods & Analysis
Definition of food
insecurity: “exists
when all people, at all
times, have physical
and economic access to
sufficient, safe, and
nutritious food to meet
their dietary needs and
food preferences for an
active and healthy life”
Data was analyzed
from
A Canadian-validated
survey (2005 Canadian
Community Health

Possible selection bias only
including individuals with
diabetes who had sought
healthcare in the last year.
Generalizability limited due to
population
Recall bias possible in
questionnaire.
Observational studies don’t
provide robust enough evidence
to recommend changes to
clinical practice
No power measure

Key Findings:
(1) Process of care
was the same in individuals with
food insecurity and food security
(2) Individuals with diabetes who
are food insecure are more likely to
be physically inactive, be
overweight, eat less fruits and
vegetables, twice as likely to be a
smoker, twice as likely to report
stress, report lower general health,
mental health and perceived stress
(3) No impact on accessing medical
care.

CASP rating: Moderate
quality
Strengths & Limitations
Strengths
OR=Results appear clinically
significant
HFSS and CCHS are Canadian
validated surveys
Limitations
No p-values provided
How FI is defined creates
classification bias

107
prevalence of food
insecurity in
individuals with
type 2 diabetes
versus the rest of
Canada.
Canada

Survey [CCHS]) was
used to compare
the prevalence of food
insecurity in Canada
compared to the
prevalence of FI
Canadians with
diabetes.
Ontario CCHS surveys
contained both FI and
diabetes modules and
were used to examine
the relationship
between food
insecurity
self-care and
management and health
status.
Used Canadianvalidated Household
Food Security Survey
(HFSS)
to identify incomeassociated food
insecurity
Process of care
measured by
completing hemoglobin
A1c, a dilated eye
exam, a urinalysis
and a foot exam over
the last year and taking
ASA and cholesterol
medication

Cross sectional analysis-difficult
to draw causational relationships
Only Ontario examined for the
impact of FI on diabetes and
some populations were excluded
e.g. First Nations. i.e.
Possibly not generalizable to all
of the Canadian population, or
may result in an information bias
Cannot generalize that overall
diabetes management is is
similar for FI and food secure
counterparts based on the study’s
specific identifiers.
Ontario coverage may influence
generalizability of results of
SMBG and medication
adherence
Possible measurement bias:
Surveys any missing information
on diabetes or food insecurity
were excluded
Questionnaires are subject to
recall bias
Questionnaire not validated to
identify difference between DM
1, DM 2 and gestational.
No power calculation
CASP rating:
Moderate-low

108
Authors, Date
& Title

Aim

Ippolito, Lyles,
Prendergast,
Marshall,
Waxman5 &
Seligman
(2016)

To observe the
association
between food
security (FI)
and diabetes
selfmanagement of
patients
accessing food
pantries

Food insecurity
and diabetes
selfmanagement
among food
pantry clients

Study Design,
Sample &
Location
Cross-sectional
design
Adults with
diagnosed diabetes
(via HbA1C or
self-report)
Using diabetes
medications
Speak English or
Spanish
accessing
designated food
pantries
Sonoma County,
Columbus, &
Corpus Christi,
USA

Methods & Analysis

Key Findings:

Strengths & Limitations

Food insecurity defined
as “limited or uncertain
access to adequate food
at the level of the
household related to
affordability

Very low food secure participants
had
 inferior diabetes self-efficacy
 more medication nonadherence
 more frequent severe
hypoglycaemic episodes

Strengths

Low food secure and very low food
secure both had
 Increased prevalence of
depression and distress
 Greater issues affording
medication
 More food- medicine-health
supply trade-offs.

Validated questionnaires used

Baseline surveys for
diabetes selfmanagement
intervention completed
between
March 2012 and March
2014.
Food insecurity
determined using short
form of the US
Department of
Agriculture’s
Household
Food Security Survey
Module (HFSSM)
which measured
income-associated food
insecurity. Three
groups identified using
questionnaire: food
secure, low food
secure, very low food
secure
8 indicators of diabetes
self-management
examined:

No statistically significant
difference in mean
HbA1c (Table 1) or percentage of
participants with HbA1c
level above 8·5%
Tobacco use was more than twice
as frequent in the very
low-food-secure compared with the
food-secure group
dose–response relationship between
severity of food insecurity and
barriers to diabetes selfmanagement.

Demonstrates strong statistical
significance
OR indicates strong clinical
significance

83% response rate
Limitations
Categories used for Low food
security and very low food
security were decided by
convention
Cross sectional-cannot determine
causality
Control group significantly
smaller than cases
Recall bias with questionnaire
Potential selection bias:
 How FI is defined
 by selecting participants
accessing food pantry only
 food secure controls are
accessing food bank
 not enough “controls” or
food secure individuals
comparatively speaking

109
(1) HbA1c; (2) diabetes
self-efficacy; (3)
diabetes
distress; (4) medication
non-adherence; (5)
severe hypoglycaemia;
(6) depressive
symptoms; (7)
medication
affordability; and (8)
food–medicine
purchasing trade-offs.
HB A1C used to
measure diabetes
control
Self-efficacy measured
using 8-item
questionnaire; diabetes
distress measured using
2-item questionnaire;
Medication adherence
measures using a fouritem Medication
Adherence
Questionnaire;
Severe hypoglycemia
assessed by asking “In
the past 4 weeks, how
many times have
you had a severe low
blood sugar reaction,
such as passing
out or needing help to
treat the reaction?’
Depressive symptoms
assessed using patient
Health Questionnaire2;

Difficult to generalize some
results from United States:
 Medication and supply costs
in California unique creating
unique medicine-health
supply-food trade offs
 Medicine costs and
coverage based on insurance
affect affordability in a way
unique to California and the
individual
Cannot control for all potential
confounders, including medical
co-morbidities
Possible information bias: some
interviews done in person and
some over the phone
CASP rating:
High

110
Inability to afford
medicine was assessed
by asking:
‘In the last 12 months,
how often did you take
less medicine than you
were supposed to
because you could not
afford to buy more?’
Medicine-food trade
offs were assessed
using a 4-point
questionnaire.

Authors, Date
& Title

Aim

KollannoorSamuel,
Wagner,
Damio, SeguraPe´rez,
Chhabra, VegaLo´pez, &
Pe´rezEscamilla
(2011)

Identify risk
factors for
depression
amongst
individuals who
are food
insecure and
establish if
social support
mediates this
risk

Social Support
Modifies the
Association
Between
Household
Food

Study Design,
Sample &
Location
Cross sectional
211 low income
Latino
adults with T2D
Connecticut

Five covariates were
measured and
controlled for: age,
gender, race/ethnicity,
education and study
site
Methods & Analysis
Definition: “Limited
availability, access
and/or
intake of nutritionally
adequate foods”
Baseline data form
DIALBEST trial
(longitudinal RCT)
used.
Participants were
(1) 21 years old (2)
living in Hartford
County, CT, (3)
HbA1c >7%, and (4)
no medical conditions

Key Findings:

Strengths & Limitations

At each level of food insecurity,
greater social support led to lower
depression rates
support

Strengths
Validated tool for depression in
DM2

Results suggest
Social support buffers depression
risk in individuals experiencing
food insecurity
Increased depressive symptoms
associated with:
 lower household income,
 increased perceived
interference of DM2 on life

Covariates known to be
associated with depression were
included in multivariate analysis
Limitations
Cross sectional analysis
Cannot control for underlying
genetic predisposition for
depression or pre-existing traits

111
completely limiting
physical activity.

Insecurity and
Depression
Among Latinos
with
Uncontrolled
Type 2
Diabetes

Depressive symptoms
measured using the
Center for
Epidemiologic Studies
(CES-D) Scale
Food insecurity
measured using Short
form of the US
household food
security supplement
module [US-HFSSM]
Social support
measured using an 11
item-questionnaire
Multi Dimensional
Diabetes Questionnaire

Increased diabetes related
clinical symptoms

Limited generalizability: Puerto
Rican & predominantly female
participants

Our study suggests that depression
risk is associated
with the perception of how much
diabetes interferes with
work, income generation,
relationship with partner, daily
routine activities, and/or social
activities

Perceived social support limited
by recall bias



Increasing social capital for patients
with DM2 who are food insecure
may mediate depressive symptomsespecially amongst individuals in
low SES.

Post hoc- alpha Cronbach
demonstrates internal
consistency to support reliability
of surveys used
CASP rating
Low

Cultural, clinical,
psychosocial variables
measured using
validated tools.
Multivariate logistic
regression model used

Authors, Date
& Title

Aim

Shalowitz, Eng,
McKinney,

Determine if
food insecurity

Study Design,
Sample &
Location
longitudinal
observational study

Logistic regression
used to determine
social support-food
insecurity-depression
interaction
Methods & Analysis

Key Findings:

Strengths & Limitations

comprehensive care,
including complete

Insulin use and A1C increase with
increasing food insecurity.

Strengths
Longitudinal study

112
Krohn, Lapin,
Wang &
Nodine (2017)
Food security is
related to adult
type 2 diabetes
control over
time in a
United States
safety net
primary care
clinic
population

(FI) is
associated
longitudinally
with poor
glucose controleven when
barriers to
comprehensive
management
are minimized.

336 patients from a
multi-site federally
qualified health
centre (FQHC)
servicing
individuals with
DM2
Illinois

medication
management is
provided to
Individuals with DM2:
 Medication and
testing supply
assistance
 Diabetes selfmanagement
 Education
 One-on-one and
group support
 Retinal screenings
 Fitness
 programmes.
Data collected using
 baseline
assessment of
patients’ FI
using the short form
US Household Food
Security Module
follow up screening FI
annually and after
study completion
A1C levels recorded
throughout 24- month
study period (between
2-6 times depending on
participant) and once
during follow up.
longitudinal mixedeffects models using
 maximum
likelihood
estimating method

level of FI only slightly related to
BMI
low food security directly impacts
glucose control
Unlike individuals who are FS,
individuals with FI do not derive
benefit from
access to comprehensive
management (comprehensive care
included quarterly visits with A1C
testing,
medication and testing supply
assistance, diabetes selfmanagement education, one-on-one
and group support, retinal
screenings, and fitness
programmes).
Directly supports the significance
of the social determinants of
diabetes control
Near independent effect of FI on
diabetes control, confirmed with
multivariable model

Validated tool used for FI
Limitations
Generalizability limited:
 Results based on one
facility.
 Culturally limited to African
Americans, Hispanic, and
Caucasian
No power calculation
CASP rating
Moderate

113
unstructured
covariance
controlling
 covariate analysis
Methods & Analysis


Authors, Date
& Title

Aim

De Marchis;
Torres;
Fichtenberg;
& Gottlieb
(2019)

Examine the
validity and
acceptability of
food insecurity
(FI) screening
tools, and their
implementation
in health care
settings.

Identifying
Food Insecurity
in Health Care
Settings: A
Systematic
Scoping Review
of the Evidence

Study Design,
Sample &
Location
Systematic review
25 articles
United States

PubMed and Ovid
MEDLINE were
searched using
keywords from 3
domains: (1) FI; (2)
screening; and (3)
health care settings
Between June 30th,
2000July 27, 2017.
Articles excluded if
evidence for validity,
acceptability, or
feasibility of FI
screening not included
What was examined:
study design and
setting; FI screening
tools; barriers to
uptake; patient vs
clinician factors, FI
appropriateness,
acceptability, or
feasibility
Studies were assigned
quality ratings using
the Grading
Recommendations
Assessment
Development and

Key Findings:

Strengths & Limitations

Majority of articles conducted in
pediatric setting

Strengths
Covers several aspects regarding
the implementation of FI
screening

Brief FI screening instruments were
all derived from and validated
against the United States
Department of Agriculture-Food
Security Survey (USDA-FSS)
Brief (1- and 2-item) FI screening
tools are validated for use in US
healthcare settings, mostly among
caregivers in pediatric population
 1-item hunger screening
question
 SEEK (safe environment
for every kid)
 2-item Hunger Vital Sign
no difference in screening mode of
administration (based on one RCT)
66% -88% found it appropriate to
be asked about FI in healthcare
setting (based on 6 observational
studies of low quality of caregivers
for pediatric patients and young
adults)
between 80-89% of clinicians found
it acceptable to screen for FI in
healthcare visits (based on 2

Each article was assigned a
quality rating
Each finding listed the number
of articles and level of quality it
was based on.
Limitations
Validated against USDA Food
security scale
 Limited generalizability
 USDA-FSS is limited to
financially associated food
insecurity.
Limited mostly to pediatric,
young adult population
Some conclusions were derived
from only 3-4 studies.
19 of 25 studies examined were
rated low-very low quality.
CASP grade
Moderate

114
Evaluation approach
(GRADE)

qualitative articles of very low
quality)

Standardized mean
differences were
calculated using 2-by-2
frequency tables of
outcome frequencies

time required for FI screening was
varied (30 seconds to 10 minutes)
4 out of 10 studies reported time as
a barrier to FI screening in practice
while 6 out of 10 reported time
created minimal barriers to
implementation (all observational
studies)
6 articles found that educating on
the importance of FI screening and
impact of SDOH on patient health
improved FI screening competence
and uptake

Authors, Date
& Title

Aim

Jones, Ngure,
Pelto & Young
(2013)

1) examine
food security
(FS) as a
concept, the
different
methods used to
measure FS
along with their
purpose and
concepts of FS
used

What Are We
Assessing When
We Measure
Food
Security? A
Compendium
and Review
of Current
Metrics

2) Identify
challenges to
FS

Study Design,
Sample &
Location
Review
17 different FS
measures
Total sample
unknown, 17
articles pulled as
part of screening
for validation in its
use for FS
No restrictions on
location

Methods & Analysis

Not everyone who screens positive
for food insecurity wants help from
care providers
Key Findings:

PubMed, CABI,
Elsevier, Google
Scholar, and Web
of Science as well as
the Web sites of
several international
organizations

Concept of food security: having
consistent and stable, physical and
financial access and availability of
nutritionally adequate food and the
ability to utilize, acquire and
consume food in socially and
culturally acceptable ways

Included Screening
tools, metrics (proxy
measures) or any
method historically
correlated with FS.

Access needs to be further broken
down and understood as 3 different
aspects: access, acquisition, and
consumption
Food measures are not always clear
about what concepts of FS they are

Strengths & Limitations
Strengths
Assessed the individual quality
and strength/limitations of each
FS measure
Examine 17 different food
security measures
Limitations
CI and p-values provided for
validation measures only

115
measurement
and future
directions
needed

measuring, nor do they all have the
same purpose

Individual articles are not
assessed for their quality

Challenges to measuring food
insecurity
1) accurately measuring dietary
sufficiency
2) Determining which concepts of
FS the measurement aims to
address
3) discerning between various
components of food access
3) accurate cutoff for food
insecurity
4) controlling for
Response and recall bias
6) validating measures when they
are so diverse in what they measure

Unclear how information or
findings were grouped as based
on what level of evidence.

One is not necessarily superior:
rather it is important to choose
A screening tool that is informed
by the concept that one intends to
measure.
Data across measures collected at
differing levels (national, regional,
household, and/or individual
Levels)
Acquiring foods in socially
acceptable/unacceptable ways is not
measured
Safety of food acquired is not
measured
Nutritionally adequate diet/diet
quality not measured

CASP rating
Low

116
Difficult to identify chronic, vs
transitory vs seasonal food
insecurity
Important trade-off to consider with
selecting food security screening
tools: exchanging
comprehensibility and contextual
detail for simplicity
and comparability
No singular way to validate a
screening tool-the process is varied
depending on what the tool is trying
to measure.
One metric is unlikely to be able to
address all domains of FS.
Consider selecting FS measures that
identify aspects that can be
addressed or is suited to a particular
program’s resources
Household consumption and
expenditure surveys or dietary
diversity surveys may be best in
identifying household level
behaviors, however there is a lot of
room for error.
Authors, Date
& Title

Aim

Ashby, Kleve,
McKechnie, &
Palermo (2016)

Identify and
review all
multi-item tools
measuring food
insecurity (FI)
available in

Measurement
of the

Study Design,
Sample &
Location
Systematic Review
13 articles
All developed
countries

Methods & Analysis

Key Findings:

Strengths & Limitations

CENTRAL, CINAHL
plus, EMBASE,
MEDLINE, TRIP were
searched

8 tools were identified in total

Strengths
quality of the study
and risk of bias for every study is
assessed

All tools assessed food access as a
dimension and it is addressed
adequately in each tool.

117
dimensions of
food insecurity
in developed
countries: a
systematic
literature
review

developed
countries and
explore the
dimensions of
FI they address.

English studies since
1999.
USDA Food Security
Survey Module
excluded due to the
known
reliability/validity
established in earlier
(Marques et al. (2014)
systematic reviews.
Data were summarized
against the dimensions
of food insecurity
quality of the study
and risk of bias as
assessed using the
American Dietetic
Association Evidence
Analysis Manual

Authors, Date
& Title

Aim

Marques,
Reichenheim,
de Moraes,
Antunes, &
Salles-Costa
(2014)

Identify &
describe
household food
security scales
available in
peer reviewed
literature and
examine, plus

Study Design,
Sample &
Location
Systematic review
24 different tools
from 184 articles
written either in
English, Spanish or
Portuguese

Most tools only assessed food
access as a dimension
 physical and economic
resources to access food
Tools do not explore other
components of FI.
Radimer/Cornell tool enquired
about anxiety
around eating a good meal due to
assistance required
with preparing food, and food
utilization

Discussed each measurement’s
reliability and validity
Limitations
None
CASP rating
High

Kuyper tool attempted
to measure stability over time
Only one tool measures both
household food insecurity, and
individual food insecurity
All studies used Cronbach’s alpha
(coefficient of
internal consistency) to measure
reliability and scored moderate to
high.

Methods & Analysis

All tools are self-report
Key Findings:

Databases used:
MEDLINE, LILACS
and SciELO databases,

Instruments use different terms to
describe same concept of household
food insecurity

Terms used: (‘food
insecurity’ OR ‘food
security’) AND

All instruments or tools were
developed using the same
underlying concept of food
insecurity (FI): poor access to food

Strengths & Limitations
Strengths
Minimized publication bias by
examining
the reference sections of all the
articles to ensure articles weren’t
missed using the algorithmbased search method

118
Household food
insecurity: a
systematic
review of the
measuring
instruments
used in
epidemiological
studies

synthesize their
psychometric
properties.

Tools developed in
United States
Canada,
Venezuela,
Colombia, Costa
Rica,
Burkina Faso,
Kenya, Iran,
Bangladesh and
Indonesia

(‘questionnaires’ OR
‘scales’
OR ‘validity’ OR
‘reliability’)
No time period limit
applied
Use household food
insecurity (HFI), not
individual food
insecurity
Article review carried
out by independent
reviewers
Reliability was
classified by using
intra-observer
reliability, interobserver reliability and
internal
consistency
Validity: examined for
(i) face or content
validity
(ii) structural
(dimensional)
Validity
(iii) criterion or
concurrent validity
(iv) construct
Validity
Tools were only
included if they
contained 3 or items.

secondary to financial constraints.
Other underlying aspects of FI
unexplored

Limitations
Tools with less than 3 items
were not included

For most tools there is very limited
appraisal

Language bias

Core Food Security
Measurement/Household Food
Security Survey Module
(CFSM/HFSSM) has been
significantly analyzed in literature,
while SPHFSS and HFSSM-6SF
have been moderately analyzed in
the literature. Half of the tools had
only been tested or studied in the
literature 1-2 times.
CFSM/HFSSM and HFSSM-6SF
studied/validated in
many countries worldwide
Cronbach’s α coefficients
demonstrates good internal
consistency for CFSM/HFSSMand
HFSSM-6SF
& SPHFSS with α
varying from 0·82 to 0·94.
Few studies measured other
Dimensions of reliability e.g. test–
retest reliability
All 3 had studies that measured
construct validity, structural
validity, convergent validity
Most tools examined only
reliability and validity Future
studies should aim to examine other

Selection bias: FI tools not
previously part of a peer
reviewed study not included
Casp Rating
High

119
psychometric properties not yet
explored
HFSSM-6SF has demonstrated
acceptable screening capability in 2
validity studies
Choosing an appropriate instrument
for use requires significant in-depth
analysis, but to date has not been
done in the majority of the tools
measuring FI
Authors, Date
& Title

Aim

Makelarski,
Abramsohn,
Benjamin, Du,
& Lindau
(2017)

To test the
diagnostic
accuracy of the
American
Academy of
Pediatrics
(AAP) and
Hunger Vital
Sign (HVS)
against gold
standard
Household
Food
Security Screen
short form
(HFSS-sf)

Diagnostic
Accuracy of
Two Food
Insecurity
Screeners
Recommended
for Use in
Health Care
Settings

Study Design,
Sample &
Location
Case control
N= 342
patients,
parents, caregivers,
family
members, and
friends of patients
>18 years of age
Emergencies
Chicago Illinois
between July-Nov
2016

Methods & Analysis

Key Findings:

Strengths & Limitations

Definition used:
“limited
or uncertain
availability of
nutritionally adequate
and safe foods”

Compared to gold standard (HFSS),
the AAP lacks sensitivity: Nearly
25% of individuals who positively
identified with FI through HFSS,
were negative with the AAP.

Strengths
No priming effect noted based
on the order screening tools were
employed.

A self-administered
survey was given
which included all 3
tools. Approximately
half the participants
completed based on a
30-day recall and the
other half with a 12month recall.
Sensitivity specificity,
positive likelihood
ratios, negative
likelihood ratios and
95% confidence
intervals (CIs) were
calculated

AAP tool slightly more sensitive
amongst households with children
compared to without children

High response level
Valid results

HVS was highly sensitive in
detecting FI compared to HFSS-sf,
but has lower specificity than AAP.

Limitations
Statistically significant results
had wide confidence intervals,
lack precision.

HVS viable alternative to HFSS-sf,
however AAP is lacking sensitivity

Small sample size. Lacks power
to identify small differences

Yes or no i.e. dichotomous surveys
for FI might be prone to more
measurement error than surveys
with more response options.

Selection bias-people who were
unwell were excluded

Stigmatization may play a role in
answers and survey accuracy-

Recall bias with self
questionnaires

120
answering “yes” is more
stigmatizing than “sometimes true”

Measurement bias-some surveys
had missing data
Generalizability- 76% African
American respondents from one
hospital in Chicago
Using HFSS-sf as gold standard.
It is not absolute in detecting all
forms of FI.
Possible social desirability biaseasier to say “sometimes true”
than to answer “yes” to FI.
Some mention of, but no
adjustment for confounding
factors

Authors, Date
& Title

Aim

Thomas,
Fitzpatrick,
Sidani, &
Gucciardi
(2018)

Examine
acceptability &
feasibility of a
food insecurity
(FI) screening
initiative that
will help health
providers
modify diabetes
management
strategies

Developing and
Implementing a
Food Insecurity
Screening
Initiative for
Adult Patients
Living With
Type 2
Diabetes

Study Design,
Sample &
Location
Systematic
literature reviews,
qualitative
interviews &

Methods & Analysis

Key Findings:

Definition: inadequate
or insecure access to
food because of
financial constraints

39% found to be food insecure

4 diabetes care
sites in Toronto
were used

Part 1: Choosing and
piloting FI
questionnaires with
population of interest,
including qualitative
interviews with 10
patients and 15 care
providers
Single- and 2-item
versions of HFSS were

71% patients were willing and
comfortable to discuss their
experiences with food insecurity
with their healthcare provider
Was found both relevant and nonintrusive to discuss in healthcare
setting. No embarrassment or
negative reactions reported
Perceived time barriers of the
provider and discomfort with
discussing financial hardship may

CASP rating
Low
Strengths & Limitations
Strengths
Care algorithm based on
systematic review of evidence
Recruitment strategy well
described and validated by
participants
Limitations
Screen only for incomeassociated FI as per HFSS
Assumes poverty is root cause of
FI

121
used to assess food
consumption over last
3 months
Part 2:
A systematic literature
review to develop a FI
algorithm and
guidelines for
managing individuals
with DM 2 who are
food insecure.
Part 3:
registered nurses&
registered dietitians
working in diabetes
consented to the FI
screening initiative to
determining
acceptability and
feasibility of Part 2 by
(a) employing initiative
with 33 patients & 5
care providers (b)
Interviewing
7 patients and 5 care
providers to discuss
acceptability and
feasibility.
Recruitment occurred
by handing out flyers
during diabetes classes
or appointments and by
using posters
Participants were > 18
years and were

be barriers in patients choosing to
disclose their FI status
Strong rapport with a provider,
perceiving the questions as
important to their diabetes
management and overall literacy
increased the likelihood of
answering the FI questions
Screening elicited how patients
cope with FI and how it affects
diabetes self-management.
Information provided was helpful to
providers in guiding care and
support for patients.
Systematic review that informed FI
algorithm encouraged the following
assessment & management plan for
individuals identified with FI:
Assess
 Medical history
 Diabetes knowledge
 Diabetes self-management
skills
 Mental health
 Literacy level
 Physical limitations
 Housing
 Financial support
 Social support
 Access to food
 Cultural influences
 Smoking habits
Assess risk of hypoglycemia:

Results limited to the experience
of dieticians one RN screening,
not PCPs
Unclear how comfortable PCP
feel discussing FI
Unclear how often screening
should occur, especially if
temporarily food insecure
Unsure if saturation of data
reached vs screening was
completed during a
given/specific period of time
CASP rating
Moderate

122
diagnosed with
diabetes
Inductive qualitative
approach used
Interviewed audiorecorded and coded in
NVivo
Thematic analysis used
to recognize
developing themes
Consensus among
researchers was used to
recap final set of
themes







Review how to treat
hypoglycemia
Replace glucose lowering
medications that cause
hypoglycemia
Schedule medications with
meals
Consider a more flexible
insulin regimen
Consider adjusting glycemic
target upwards

Management
 Develop collaborative &
realistic treatment plan
 Encourage inclusion of family
members in education
 Consider client’s medication
coverage
Consider referrals:
 Dietitian
 Nurse
 Social worker
 Respiratory Therapist for
smoking cessation
 Local community food
resources (food banks,
community kitchens, low cost
grocery stores)
 Food skills classes
 Assess if eligible for Special
Diet Allowance funding
Dietary management
 Individualize nutrition
recommendations based on the
client’s budget, food skills, and

123




available food storage and
cooking equipment
Provide information on low
cost community foods
recourses and meals
Provide low cost ideas for
treating hypoglycemia

Assessing for FI can assist healthcare providers in supporting
diabetes self-management.
Further evaluation needed to
establish how FI screening may
improve a diabetes selfmanagement and outcomes.
Important to acknowledge when
screening may not be necessary or
important.
May bring meaning and
understanding to providers why
patients might not follow diabetes
management recommendations.
FI can be best addressed through
income-based interventions.

124

Appendix C
Food Insecurity Tools
Screening Tool

Purpose/ Intent

Hager two-item
screen

Identifies income
associated food
insecurity
Identifies income
associated food
insecurity
Measures access and
Anxiety or uncertainty
surrounding food
supply
Access to quality and
variety of food
Insufficient food
intake and its physical
consequences
Identifies income
associated food
insecurity
Identifies income
associated food
insecurity
Identifies income
associated food
insecurity
Detect incomeassociated food
insecurity

Girard four-point
tool
Household Food
Insecurity Access
Scale (HFIAS)

Townsend Food
Behaviour
Checklist
Radimer/Cornell
Modified
Radimer/Cornell
One-Item Hunger
Screening
Question

6
7

Dimension
measured
Economic
Access7

Food Insecurity
Level
Household

No. of
Questions
2

Sensitivity/
Specificity6
97% / 83%

Validated in
Healthcare Setting?
No

Comments

Economic
Access

Household
Individual

4

n/a

No

Economic
Access

Individual

9

n/a

No

Economic
Access

Individual

2

n/a

No

Economic
Access
Utilization
Economic
Access
Utilization
Economic
Access

Individual

8

n/a

No

Well analyzed in peerreviewed literature

Individual

10

n/a

No

Well analyzed in peerreviewed literature

Household

1

83% / 80%

Yes

Diagnostic accuracy in capturing the specific dimension of food insecurity listed in each study. Note: studies have near exclusively determined this value by validating the tool against the HFSSM
“Economic access” to food refers to access to food income-associated food insecurity

125
Two-Item
Hunger 8Vital
Sign (HVS)

Household
Dietary Diversity
Score (HDDS)

Household Food
Insecurity Access
Scale
(HFIAS)
Household
Hunger Scale
(HHS)

Core Food
Security
Measurement
/Household Food
Security Survey
Module
(CFSM/HFSSM)

8

Detect incomeassociated food
insecurity and hunger
for the purpose of
linking screened
individuals to
resources
Measures dietary
diversity as a proxy
for food security.
Measures 12 different
food groups
Can be used to
measure food security
changes over time

Economic
Access

Household

2

Food quality

National
Regional
Household

12

No

Used to measure FS at
regional or household
levels
Used to monitor
impact of
interventions
Measures hunger as a
proxy for food
security
Can be repeated to
measure the impact of
interventions on
hunger
Measures 4 areas of
IAFI:
1) anxiety about
household food
supplies
2) perceptions of
accessibility to food
quality or quantity

Anxiety
Food
Preferences
Economic
access
Food quantity
Economic
Access
Food quantity

Regional
Household

9

No

Regional
Household

3

No

Economic
Access
Anxiety

Household

18

Yes

Note: Two-question short-form from HFSS tool

89-97% /
74-84%

Yes

Well analyzed in peer
reviewed literature
Gold standard

126

Household Food
Security Survey
Module SF
(HFSSM-sf)
American
Academy of
Pediatrics (AAP)
Tool9

3-item Food
Security tool11

Food Insecurity
by Elders (FIE)

3) any reductions in
food intake
4) children’s reduced
food intake
Validated short form
of HFSSM to capture
IAFI

Economic
Access

Household

6

98%/ 92%

Yes

Well analyzed in peer
reviewed literature

Detect incomeassociated food
insecurity and hunger
for the purpose of
linking screened
individuals to
resources
Adapted from the
HFSSM to identify
food insecurity
amongst a diabetic
population

Economic
Access

Household

2

71%-82% /
96%10

Yes

Comparative studies
demonstrate AAP
missed a quarter of
food insecure
individuals compared
to HVS

Economic
Access

Household

3

n/a

No

Elderly augmented
HFSSM

Economic
Access
Physical
Access
Utilization

Individual-

1412

89%/69%

No

Not validated. Used
for one pilot study
Validated using 3 food
secure and 3 food
insecure individuals
with DM 2
Extra questions
created based on one
study’s findings

Validated against HVS, and does not specifically measure pediatric population
Values reflect families without children
11
Adapted using 3 questions from validated HFSSM. Not validated to identify food insecurity and no sensitivity/specificity testing.
12
In addition to the HFSSM
9

10