Food Insecurity and Perceived Financial Stress are Associated with Cost-related Medication Non-adherence in Stroke

Abstract

Objective.

To determine whether food insecurity and perceived financial stress contribute to cost-related medication non-adherence (CRN) in stroke.

Methods.

We conducted a retrospective study of adult stroke survivors in the National Health Interview Survey (2014–2018). Weighted prevalence of food insecurity, perceived financial stress, and CRN by age was calculated. Multiple logistic regression was conducted between food insecurity or perceived financial stress and CRN, adjusting for demographic and clinical variables.

Results.

Prevalence of food insecurity, perceived financial stress, and CRN respectively were 38%, 75%, and 26% (age 18–44), 38%, 76%, and 21% (age 45–64) and 17%, 43%, and 6% (age≥ 65). Food insecurity and perceived financial stress respectively were associated with CRN in stroke survivors aged 45–64 [odds ratio (95% CI) 1.35 (1.18–1.54) and 1.44 (1.29–1.61)] and age ≥ 65 [1.77 (1.52–2.06) and 1.51 (1.37–1.67)].

Conclusion.

Food insecurity and perceived financial stress are prevalent in stroke survivors and associated with CRN.

Stroke produces a heavy economic burden on the health care system and the stroke survivor. The estimated direct cost of caring for stroke patients totaled 30.8 billion dollars in the year 2016–2017.¹ Estimated out-of-pocket costs for Medicare beneficiaries with cardiovascular disease (heart disease and stroke) averaged $1,689 during the year 2014, 38% more than costs paid by those without cardiovascular disease.²

Financial hardship can have negative effects, affecting perceived financial stress, food security, and cost-related medication non-adherence (CRN).^3,4 Perceived financial stress, the psychological response to financial hardship, may adversely affect mental health outcomes.⁵ Food insecurity, decreased access to nutritionally adequate and safe foods due to insufficient financial resources, may exacerbate chronic illness and increase health care use.⁶ Non-adherence to medication due to cost is another consequence of financial hardship and is experienced by approximately 18% of stroke survivors.⁷ While a relationship between financial hardship and CRN has been demonstrated in stroke survivors,⁴ whether and to what magnitude perceived financial stress and food insecurity are associated with CRN in stroke survivors is not well known. Stroke survivors, with high out-of-pocket costs associated with potentially lifelong disability, may be particularly vulnerable to making trade-offs between buying food versus medication and the psychological impact of financial hardship. How and whether food insecurity and perceived financial stress relate to CRN may differ between middle-aged and older stroke survivors given the influence that age has on asset accumulation, social support networks,⁸ access to different forms of health insurance, and at times, access to social care programs.⁹ Defining the factors that influence CRN is important for developing policies to reduce the number of stroke survivors with financial hardship who experience CRN. Our objective was therefore to evaluate whether and to what magnitude perceived financial stress and food insecurity contribute to CRN in middle-aged and older stroke survivors.

Methods

Overview.

The study sample was a cohort of stroke survivors from the National Health Interview Survey (NHIS) during the years 2014–2018. The NHIS is a national annual survey conducted in person (and recently by telephone) to monitor the health of the United States non-institutionalized civilian population, selected by geographically clustered sampling techniques.¹⁰ Data from NHIS are publicly available at https://www.cdc.gov/nchs/nhis/data-questionnaires-documentation.html. Included were survey years after implementation of the Affordable Care Act (2014–2018) to mitigate effects of underinsurance on CRN. The study was deemed exempt by the Michigan Medicine Institutional Review Board as data were de-identified. Informed consent was not required for this study.

Study population.

Stroke survivors in NHIS during the years 2014–2018 who were at least 18 years of age were identified. Stroke survivors were those who answered, “Yes” to the question, “Have you ever been told by a doctor or other health professional that you had a stroke?” Adult stroke survivors were categorized by age as follows: young (between the ages of 18–44), middle-aged (between the ages of 45–64), older (65 years of age or older). All adult stroke survivors were eligible for inclusion. The non-stroke population was at least 18 years of age and answered, “No” to the NHIS question used to identify history of stroke. We report the prevalence of food insecurity and perceived financial stress in non-stroke survivors for comparison with stroke survivors. Self-reported history of stroke has been validated against clinical assessment.¹¹

Outcome.

The primary outcome was cost-related medication non-adherence, which was defined as being present (vs. absent) if the answer to any of the following NHIS questions was “yes”:

During the past 12 months, were any of the following true for you? … You skipped medication doses to save money; You took less medicine to save money; You delayed filling a prescription to save money.

Exposures.

The two exposures of interest were perceived financial stress and food insecurity. Perceived financial stress was defined as being present (vs. absent) if the response to any of the following NHIS questions was “very worried or moderately worried”:

How worried are you right now about. … not being able to pay medical costs for normal health care? not being able to pay your rent/mortgage/housing costs? not being able to pay medical costs of a serious illness or accident? not having enough to pay your normal monthly bills? not having enough money for retirement? not being able to maintain the standard of living you enjoy?

A continuous variable for perceived financial stress was also created. Participants were allotted one point for each response that indicated perceived financial stress. Therefore, scores ranged from one to six. Higher scores indicated greater perceived financial stress.

Food insecurity was defined as being present (vs. absent) if the response to the NHIS question, “In the last 30 days, did you ever cut the size of your meals or skip meals because there wasn’t enough money for food?” was “Yes” or if the response to any of the following NHIS statements was “often true or sometimes true”:

I worried whether food would run out before I got money to buy more.

The food that I bought just didn’t last, and I didn’t have money to get more.

I couldn’t afford to eat balanced meals.

A continuous variable for food insecurity was also created. Participants were allotted one point for each response that indicated food insecurity. Therefore, scores ranged from one to four. Higher scores indicated greater food insecurity.

Covariates.

Covariates from the NHIS included demographic characteristics such as age (categorized as 18–44, 45–64, and 65 years of age and older); sex; marital status (categorized as married or not married); Hispanic ethnicity (yes/no); self-reported race (Black, non-Hispanic White, American Indian and Alaska Native, Asian, and Multiple Race); education (less than high school, high school, some college, associate’s degree, college graduate, graduate degree); income-to-needs ratio (below poverty, near poverty, not in poverty), which is a measure of total family income divided by the federal poverty level based on family size where an income-to-needs ratio less than one defines poverty, a ratio of one to two defines near poverty, and two or more defines not in poverty;¹² health insurance coverage (yes/no), and region (Northeast, Midwest, South, West). Clinical characteristics included number of comorbidities including hypertension, high cholesterol, coronary heart disease, angina, myocardial infarction, heart disease, emphysema, chronic obstructive pulmonary disease, diabetes, kidney disease, and arthritis.

Statistical analysis.

Analyses were computed in SAS software (v. 9.4). Copyright © 2013 SAS Institute Inc. SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc., Cary, NC, USA. SAS SURVEY and SUDAAN (v. 11.1) procedures were used to adjust for the NHIS complex cross-sectional sample design. The data were weighted using a scaled weight across all five years that summed to the United States population at the midpoint of the five-year period. Missing data were excluded because less than 10% of missing data were observed on variables in the analysis. Statistical significance was assigned based on a Type 1 error level of α = 0.05 or 95% confidence intervals for estimates of population statistics.

Comparisons were made of demographic and clinical characteristics between stroke patients by food insecurity status (food insecure/not food insecure) and by perceived financial stress (perceived financial stress/no perceived financial stress). Chi-squared test was used for comparison of categorical variables. T-tests were used for comparison of continuous variables. Statistical significance was set at a p-value of less than .05.

The weighted prevalence of food insecurity and perceived financial stress among all stroke survivors and in non-stroke survivors, for comparison, was described. The weighted prevalence of food insecurity, perceived financial stress, and cost-related nonadherence in stroke survivors were described by age category. Weighted prevalence of behaviors characterizing food insecurity and individual components of the perceived financial stress variable were also examined by age category.

Multiple logistic regression models were performed to evaluate the association between food insecurity as a continuous variable on the dichotomous outcome of CRN and perceived financial stress as a continuous variable on the dichotomous outcome of CRN in middle-aged and older adult stroke survivors separately. Young stroke survivors were not included in the multiple logistic regression due to our a priori hypothesis of low prevalence of food insecurity and perceived financial stress in this age group and the low prevalence of stroke in the young. Taylor Series variance estimation was used to capture the complex sample design. Clinical and demographic covariates that were significant in bivariate analyses with perceived financial stress or food insecurity at a p-value of <.05 were included as covariates in all multiple logistic regression models.

Results

There were 3,221 stroke survivors at least 18 years old during the years 2014–2018. Of these stroke survivors, there were 232 who were 18–44 years old; 976 who were 45–64 years old; and 2,013 who were 65 years of age or older. There were 81,891 adults without a history of stroke. All adult stroke survivors were included in the sample.

There were 868 stroke survivors with food insecurity (26% of stroke survivors) which represents close to two million stroke survivors with food insecurity. There were 12,699 non-stroke survivors with food insecurity (15% of non-stroke survivors) which represents 36.7 million non-stroke survivors with food insecurity. Compared with stroke survivors without food insecurity, stroke survivors with food insecurity were more likely to be middle-aged, female, unmarried, non-White race, Hispanic ethnicity, have less than a high school education, and have more medical comorbidities (Table 1).

Table 1.

DEMOGRAPHIC AND CLINICAL DIFFERENCES BY FOOD INSECURITY STATUS AND FINANCIAL STRESS AMONG STROKE SURVIVORS IN THE NATIONAL HEALTH INTERVIEW SURVEY (2014–2018)(N=3,221)

Factor	Food insecurity (n=868), 26% N(%)[95%CI]	No food insecurity (n=2353), 74% N(%)[95%CI]	Chi-squarea; p-Value	Financial stress (n=1732), 58% N(%)[95%CI]	No financial stress (n=1422), 42% N(%)[95%CI]	Chi-squarea; p-Value
Age (%)			p<.001			p<.001
≤44	90 (13.38) [10.06–16.70]	142 (7.85) [6.26–9.43]		170 (11.96) [9.72–14.21]	56 (5.27) [3.50–7.05]
45–64	419 (50.99) [46.47–55.51]	557 (28.99) [26.59–31.40]		710 (45.82) [42.83–48.80]	240 (19.43) [16.62–22.24]
≥65	359 (35.62) [31.45–39.78]	1654 (63.15) [60.72–65.57]		852 (42.2) [39.40–45.01]	1126 (75. 29) [72.29–78.28]
Sex (% Female)	511 (56.35) [52.03–60.67]	1227 (47.80) [45.30–50.31]	p<.001	979 (51.04) [47.94–54.14]	716 (47.82) [44.60–51.05]	p=.16
Married (% yes)	193 (32.45) [28.12–36.78]	964 (52.90) [50.39–55.40]	p<.001	594 (47.25) [44.25–50.26]	543 (48.76) [45.37–52.14]	p=.52
Hispanic (% yes)	110 (16.54) [12.41–20.66]	149 (9.28) [7.25–11.31]	p<.001	176 (14.78) [11.79–17.78]	77 (6.31) [4.28–8.34]	p<.001
Race (%)			p<.001			p<.002
White/Caucasian	547 (62.42) [57.99–66.85]	1949 (82.73) [80.44–85.01]		1296 (74.90) [72.08–77.72]	1149 (81.38) [78.59–84.17]
Black/African American	242 (27.99) [23.87–32.12]	277 (11.57) [9.68–13.45]		305 (17.11) [14.57–19.64]	200 (13.45) [10.93–15.98]
AIAN	16 (1.82) [.38–3.26]	17 (.83) [.31–1.34]		20 (1.20) [.47–1.94]	13 (.99) [.32–1.66]
Asian	24 (3.87) [1.93–5.82]	63 (3.15) [2.26–4.05]		46 (3.63) [2.36–4.90]	39 (2.94) [1.79–4.08]
Multiple Race	37 (3.86) [2.03–5.70]	46 (1.70) [1.09–2.31]		63 (3.13) [2.04–4.22]	20 (1.22) [.60–1.84]
Education (%)			p<.001			p<.01
Less than high school	283 (32.91) [28.55–37.27]	431 (17.86) [16.0–19.71]		406 (22.74) [20.14–25.34]	291 (19.65) [17.08–22.21]
High school	234 (26.23) [22.54–29.92]	673 (28.84) [26.53–31.15]		476 (27.63) [24.96–30.29]	412 (29.02) [26.00–32.04]
Some college	180 (20.07) [16.45–23.69]	428 (17.84) [15.77–19.92]		340 (19.21) [16.66–21.77]	257 (17.55) [15.25–19.85]
Associates degree	85 (11.20) [8.43–13.97]	267 (12.09) [10.32–13.86]		211 (13.13) [11.03–15.22]	134 (10.02) [8.09–11.96]
College graduate	58 (7.20) [4.49–9.90]	318 (14.36) [12.55–16.17]		187 (11.22) [9.26–13.18]	181 (14.52) [12.12–16.91]
Graduate degree	21 (2.37) [1.22–3.52]	215 (8.98) [7.58–10.38]		104 (6.04) [4.70–7.39]	129 (9.21) [7.41–11.02]
Income-to-needs ratiob (%)			p<.001			p<.001
Below Poverty	368 (41.03) [36.28–45.77]	276 (10.65) [9.13 -12.16]		436 (22.72) [20.00–25.45]	188 (12.43) [10.16–14.69]
Near Poverty	326 (36.67) [32.33–41.01]	530 (21.41) [19.22–23.59]		531 (29.36) [26.38–32.33]	313 (20.19) [17.54–22.85]
Not in Poverty	145 (22.29) [18.06–26.52]	1395 (67.93) [65.38–70.48]		695 (47.90) [44.57–51.24]	819 (67.36) [64.05–70.68]
Health insurance coverage (% yes)	807 (90.05) [86.62–93.48]	2289 (97.19) [96.31–98.07]	p<.001	1628 (92.82) [90.94–94.71]	1404 (98.71) [97.86–99.56]	p<.001
Region (%)			p<.001			p=.33
Northeast	108 (14.04) [10.64–17.45]	358 (16.03) [13.84–18.22]		235 (15.11) [12.47–17.75]	219 (15.86) [13.38–18.33]
Midwest	175 (21.07) [17.36–24.78]	536 (21.86) [19.46–24.27]		384 (22.50) [19.79–25.22]	318 (21.01) [18.06–23.97]
South	446 (49.88) [45.11–54.65]	926 (39.86) [36.77–42.95]		759 (43.28) [39.76–46.79]	581 (40.98) [37.31–44.65]
West	139 (14.99) [11.33–18.64]	533 (22.22) [19.58–24.87]		354 (19.09) [16.08–22.10]	304 (22.13) [19.06–25.20]
Number of comorbiditiesc (Mean)	4.15 [3.97–4.33]	1.07 [1.03–1.10]	p<.001	3.87 [3.75–3.99]	.90 [.88–.92]	p<.01

Notes

^at-test was used for the comparison of number of comorbidities with food insecurity or perceived financial stress.

^btotal family income divided by the federal poverty level based on family size where an income-to-needs ratio < 1 defines poverty, a ratio of 1–2 defines near poverty, and ≥2 defines not in poverty.

^ccomorbidities include hypertension, high cholesterol, coronary heart disease, angina, myocardial infarction, heart disease, emphysema, chronic obstructive pulmonary disease, diabetes, kidney disease, and arthritis. Values in the table represent unit (mean or %) and [95% CI].

AIAN=American Indian or Alaska Native.

There were 1,732 stroke survivors with perceived financial stress (58% of stroke survivors) which represents 4.3 million stroke survivors with perceived financial stress. There were 47,238 non-stroke survivors with perceived financial stress (61% of non-stroke survivors) which represents 142 million non-stroke survivors with perceived financial stress. Compared with stroke survivors without perceived financial stress, those who had perceived financial stress were more likely to be middle-aged, Hispanic ethnicity, and have a higher number of medical comorbidities (Table 1).

Weighted prevalence of food insecurity, perceived financial stress, and CRN by age.

The prevalence of food insecurity was numerically the highest in stroke survivors aged 18–44 and stroke survivors aged 45–64, among whom 38% were food-insecure compared with 17% of stroke survivors aged 65 and older. The prevalence of perceived financial stress was numerically similar and high in stroke survivors aged 18–44 and 45–64 years old, with a prevalence of 75% and 76% respectively compared with 43% in stroke survivors aged 65 and older (Figure 1). Young and middle-aged stroke survivors had similar prevalence of cost-related medication non-adherence (26% [95% confidence interval (CI): 17–34%] and 21% [95% CI: 17–24%], respectively) compared with 6% (95% CI: 5–8%) in stroke survivors aged 65 and older.

Figure 1.

Prevalence of food insecurity and perceived financial stress by age.^a

Note:

^aThe bar graph shows the prevalence of food insecurity and perceived financial stress categorized by age and for the total sample of stroke survivors in the National Health Interview Survey (2014–2018). There were 3,221 stroke survivors (18–44 years old: N=232; 45–64 years old: N=976; 65 years and older: N=2013). The error bars represent the 95% confidence intervals. The number above each error bar indicates the prevalence of food insecurity or perceived financial stress for that age category (i.e., the value for prevalence on the y-axis).

Prevalence of components of food insecurity and perceived financial stress by age category.

Among food-insecure young, middle-aged, and older stroke survivors, the prevalence of cutting size or skipping meals because of not enough money was numerically higher (58% for young, 51% for middle-aged, and 38% for older stroke survivors) than the prevalence of performing other food-insecure behaviors, which were quantitatively similar (range of prevalence of 25–35% for young, range of prevalence of 22–33% for middle-aged, and range of prevalence of 11–18% for older stroke survivors) (Figure 2A). Young, middle-aged, and older stroke survivors reported perceived financial stress across a range of financial responsibilities, with the lowest prevalence reported by middle-aged and older stroke survivors for paying for children’s college (prevalence of 17% for middle-aged and 1% for older stroke survivors) (Figure 2B).

Figure 2.

Prevalence of components of food insecurity and perceived financial stress by age.^a

Notes:

^a(A) The bar graph shows the prevalence of the components of food insecurity categorized by age and for the total sample for stroke survivors in the National Health Interview Survey (2014–2018) (There were 3,221 stroke survivors (18–44 years old: N=232; 45–64 years old: N=976; 65 years and older: N=2013). The error bars represent the 95% confidence intervals. The number above each error bar indicates the prevalence of the component of food insecurity for that age category (i.e. the value for prevalence on the y-axis). (B) The bar graph shows the prevalence of the components of perceived financial stress categorized by age and for the total sample for stroke survivors in the National Health Interview Survey (2014–2018). The error bars represent the 95% confidence intervals. The number above each error bar indicates the prevalence of the component of perceived financial stress for that age category (i.e. the value for prevalence on the y-axis).

The association of food insecurity and perceived financial stress with CRN in stroke survivors.

Greater food insecurity was associated with higher odds of CRN in both middle-aged and older stroke survivors (odds ratio (OR) 1.35 [95% CI 1.18–1.54] and OR 1.77 [95% CI 1.52–2.06], respectively) adjusting for demographic and clinical covariates (Table 2). Greater perceived financial stress was associated with higher odds of CRN in both middle-aged and older stroke survivors (OR 1.44 [95% CI 1.29–1.61] and OR 1.51 [95% CI 1.37–1.67], respectively) adjusting for demographic and clinical covariates (Table 2).

Table 2.

MULTIPLE VARIABLE REGRESSION MODELS EXAMINING THE ASSOCIATION BETWEEN FOOD INSECURITY AND COST-RELATED NON-ADHERENCE BEHAVIORS (CRN) AND PERCEIVED FINANCIAL STRESS AND COST-RELATED NON-ADHERENCE AMONG THE POPULATION WITH STROKE IN THE NATIONAL HEALTH INTERVIEW SURVEY (2014–2018) IN MIDDLE-AGED AND OLDER ADULTS

Factor	Cost-related non-adherence behaviors OR [95% CI]; p-value
	Ages 45–64	Ages 65+
Modela
Food insecurity	1.35 [1.18–1.54]*	1.77 [1.52–2.06]*
Perceived financial stress	1.44 [1.29–1.61]*	1.51 [1.37–1.67]*

Notes

^aData shown are for separate models of food insecurity and perceived financial stress by age (4 separate models). Model is adjusted for sociodemographics (sex, race, ethnicity, education, marital status), region, income-to-needs ratio, health insurance coverage, and medical comorbidities.

^*p<.001

OR= Odds Ratio

CI= Confidence Interval

Discussion

In this nationally representative sample of stroke survivors, food insecurity and perceived financial stress were both associated with CRN in middle-aged and older stroke survivors even with adjustment for demographic and clinical covariates, and health insurance status. The prevalence of food insecurity and perceived financial stress were higher in middle-aged than older stroke survivors. Both middle-aged and older stroke survivors reported performing a range of food-insecure behaviors and reported perceived financial stress across a variety of financial responsibilities.

Social determinants of health such as food insecurity and perceived financial stress influence CRN. While health insurance can help mitigate the cost of medical treatment, concerns about finances or food availability remain an important determinant of whether stroke survivors adhere to their prescribed treatment. In a study evaluating the contribution of food insecurity to CRN in patients with chronic conditions, the odds of CRN was four times higher in chronically-ill adult patients with food insecurity than in those who were food-secure.¹³ Also similar to what was reported here, food-insecure patients were more likely to be of non-White race and Hispanic ethnicity than patients who were not food-insecure.¹³ In a study examining the relationship between social determinants of health and adherence to anti-hypertensive medications in Medicare recipients age 65 or older, poverty/food insecurity had the strongest association with anti-hypertensive medication non-adherence.¹⁴ The current study adds that food insecurity is also associated with CRN in stroke survivors, and the prevalence of food insecurity is numerically higher in middle-age stroke survivors than their older counterparts. Perhaps middle-aged adults are more likely to allocate limited financial resources to financial demands other than food or medications given potentially greater financial responsibilities in middle age compared with old age. Furthermore, the existence of food assistance programs for older adults might result in lower prevalence of food insecurity compared with middle-aged stroke survivors.

Perceived financial stress, independent of poverty status, was associated with CRN in both middle-aged and older stroke survivors. A similar finding has been reported in a previous study of medication adherence and persistence in stroke survivors, which showed that stroke survivors who self-reported adequate income to meet their household needs had higher odds of medication adherence at 12 months than those who did not.¹⁵ An association between perceived financial stress and CRN has also been shown among patients with other chronic conditions.¹⁶ In the general adult population, a measure of chronic stress that includes ongoing financial stress was associated with medication non-adherence independent of income.¹⁷ Taken together, these findings suggest that the perception of one’s financial condition plays a key role in medication adherence, in general, and CRN in particular. Given the high prevalence of perceived financial stress observed in young and middle-aged stroke survivors (75% and 76%, respectively) further research on the impact of strategies to reduce perceived financial stress on CRN should include young and middle-aged stroke survivors. Although this study was not designed to identify the factors contributing to perceived financial stress in stroke survivors, prior research has shown that actual financial hardship after stroke is lessened by social contact.¹⁸ Perhaps, the reduction in social network size through loss of employment after stroke contributes to increased perceived and actual financial stress of young and middle-aged stroke survivors, which is an important area of future research.

The prevalence of food insecurity, perceived financial stress, and cost-related medication nonadherence were similar among young and middle-aged stroke survivors, contrary to what we predicted. Our findings suggest that young adults with chronic conditions are as vulnerable as middle-aged adults to the effects of financial hardship. Indeed, almost one in three young adults with type 2 diabetes reported food insecurity during the years 2016–2019,¹⁹ which is like the prevalence of food insecurity among young adult stroke survivors in our sample. Clinicians should screen all patients for financial hardship, particularly as the prevalence of financial hardship is likely to have increased with the COVID-19 pandemic.

If these findings are replicated in other samples of stroke survivors, they have broader implications for policy. Strategies for reduction of food insecurity, such as broadening eligibility for meal assistance programs, increasing awareness of existing programs, and strategies to reduce perceived financial stress may have a positive impact on adherence and cardiovascular outcomes. More research is needed on the association between CRN and outcomes such as stroke mortality and recurrence.²⁰ It is also vitally important to evaluate whether addressing social needs such as food insecurity and perceived financial stress reduces health care costs and utilization.²¹

Strengths of this study include the use of a nationally representative sample and inclusion of stroke survivors across the age spectrum. The limitations include the fact that the study period was restricted to 2014–2018. Although recent years were not included, subsequent years of economic downturn and the COVID-19 pandemic might increase the number of stroke survivors with food insecurity or perceived financial stress but should not affect the strength of association with CRN. As this study was a cross-sectional analysis of NHIS data, it cannot be said that food insecurity and perceived financial stress cause CRN. It is also uncertain whether those reporting CRN are cutting back on secondary stroke prevention medications or medications for other health conditions. We report the prevalence of the components of food insecurity and perceived financial stress by age category without testing the statistical significance of the comparisons, as these data were not central to our research objective. We did not investigate sex-related, racial, or ethnic differences in the contribution of food insecurity and perceived financial stress to CRN, which is a potential area of future research. We did not investigate the relationship between CRN and stroke mortality. This was a secondary analysis of survey data. Items used to generate the burden-related constructs in this study and CRN behaviors are based upon available data, and there may be other items or ways to frame questions that would strengthen reliability. There were no available data to verify the CRN behaviors reported in this study against pharmacy-claims or medical records, which would strengthen confidence in self-report data, however the rates of CRN reported here are in line with those reported in other work.^7,22

Conclusion.

In this nationally representative sample of stroke survivors, food insecurity and perceived financial stress were associated with CRN. Both food insecurity and perceived financial stress are prevalent in stroke survivors of all ages, but the higher prevalence of food insecurity and perceived financial stress in middle-aged compared with older stroke survivors suggests that stroke survivors in middle age may be at particular risk. Future research should examine ways to reduce food insecurity and perceived financial stress and the impact on CRN among stroke survivors.