Relationship between Multiple Measures of Financial Hardship and Glycemic Control in Older Adults with Diabetes

Abstract

Aim:

Examine the relationship between multiple measures of financial hardship and glycemic control in older adults with diabetes.

Methods:

Using data from Health and Retirement Survey (HRS), we investigated 4 measures of financial hardship: difficulty paying bills, ongoing financial strain, decreasing food intake due to money, and taking less medication due to cost. Using linear regression models we investigated the relationship between each measure, and a cumulative score of hardships per person, on glycemic control (HbA1c).

Results:

After adjustment, a significant relationship existed with each increasing number of hardships associated increasing HbA1c (0.09, 95%CI 0.04, 0.14). Difficulty paying bills (0.25, 95%CI 0.14, 0.35) and decreased medication usage due to cost (0.17, 95%CI 0.03, 0.31) remained significantly associated with HbA1c.

Conclusion:

In older adults, difficulty paying bills and cost-related medication non-adherence is associated with glycemic control, and every additional financial hardship was associated with an increased HbA1c of nearly 0.1%.

Introduction

As the seventh leading cause of death and a driver of healthcare costs for those diagnosed, diabetes is a growing public health concern among adults in the United States (US), particularly in groups with higher risk of disease (Centers for Disease Control, 2017, American Diabetes Association, 2018). Adults with low socioeconomic status (SES) are disproportionately affected by diabetes and its complications (Brown et al., 2004; Chaturvedi, Jarrett, Shipley, & Fuller, 1998). For example, diabetes complications, including neuropathy, nephropathy, retinopathy, and cardiovascular disease, are associated with poor glycemic control and individuals with low SES tend to have worse glycemic control compared to those with higher SES (Houle et al., 2015; James et al., 2012; Jotkowitz et al., 2006). Lower SES has also been associated with increased risk factors and a higher mortality rate, particularly for individuals with diabetes (Dray-Spira, Gary-Webb, & Brancati, 2010; Laaksonen et al., 2008; O’Kane, McMenamin, Bunting, Moore, & Coates, 2010; Stringhini et al., 2010). The impact of diabetes in low SES populations reveal a need to better understand how financial hardship among those diagnosed with diabetes impact health outcomes.

Recent estimates indicate that older adults age 65 years and older are at higher risk for diabetes with approximately 25% of those over the age of 65 being diagnosed (Kirkman et al., 2012). Older adults with diabetes may experience increased financial burden and have lower economic resources compared to their middle-aged counter parts (DeNavas-Walt & Proctor., 2015). For example, it is estimated that nearly 15% of older adults in the US live below the federal poverty line (DeNavas-Walt & Proctor., 2015). These older adults must manage increasing housing and healthcare costs, lack of access to transportation, decreased savings, and job loss, often living at or below the poverty line (National Council on Aging, 2016). In addition, older adults with diabetes have greater risk of cognitive impairment, urinary incontinence, pain, and the need to manage multiple prescriptions. (American Diabetes Association, 2019) Individuals living with diabetes and high financial needs report the necessity of cost trade-offs such as decreased food intake and cutting down on medication to control their diabetes (Billimek & Sorkin, 2012; Silverman et al., 2015). These necessary decisions have health consequences; for instance, cost-related medication underuse is associated with poor diabetes control and increased ER/inpatient visits (Berkowitz et al., 2015; Kang, Lobo, Kim, & Sohn, 2018; Lee, Jiang, Dowdy, Hong, & Ory, 2018). Though many studies find younger adults have higher risk for cost-related non-adherence, often based on lack of insurance, a study of adults with Medicare Part D coverage found high rates of cost-related non-adherence particularly in cardiometabolic medications ((Ngo-Metzger, Sorkin, Billimek, Greenfield, & Kaplan, 2012; Williams, Steers, Ettner, Mangione, & Duru, 2013; Zhang, Lee, & Meltzer, 2014).

While definitions of SES vary across the literature, low SES is often characterized by lower education level and lower income. For instance, having lower education levels and low income is associated with worse glycemic control and increased risk of death in individuals with diabetes (Honda, Pun, Manjourides, & Suh, 2017; Saydah, Imperatore, & Beckles, 2013). More recently, additional factors related to low SES have been investigated, such as low-grade employment status or unemployment, and residential area characteristics, such as living in an urban area while impoverished and unemployed, or exposure to air pollution (Kumari, Head, & Marmot, 2004; Mirowsky et al., 2017). However, existing measures of SES, such as income and level of education, may underestimate the true impact of financial hardship on health, particularly among older adults. For example, food insecurity predicts chronic disease better than income (Gregory & Coleman-Jensen, 2017). In addition, while difficulty paying bills was associated with material hardship related to lower income, as well as financial pressure associated with accelerated aging in older adults, this relationship has not been investigated in adults with diabetes or without a focus on specific income levels (Levy, 2015; Simons et al., 2016).

Taken together, evidence suggests that financial hardship independently influences diabetes outcomes (Walker, Smalls, Campbell, Strom Williams and Egede, 2014), however limited detail exists on the best measures to use to characterize SES within the context of health. Drawing from previous theories that helped identify factors contributing to the distribution of health in older age, the World Health Organization’s Commission on Social Determinants of Health developed a framework highlighting factors and pathways to healthy aging. (Sadana, Blas, Budhawani, Koller, and Paraje, 2016) Using this framework, we aimed to examine material factors (difficulty paying bills, medication non-adherence, food insecurity) categorized in the framework under strengths, exposures, and vulnerabilities, while controlling for socioeconomic factors the framework categorized under social position (education, wealth) in an effort to understand how best to intervene. (Sadana et al. 2016) Therefore, the aim of this analysis was to examine the relationship between multiple measures of financial hardship and glycemic control in older adults with diabetes.

METHODS

Data Source

The Health and Retirement Study (HRS) is a national survey of U.S. adults over age 50 and their spouses (Health and Retirement Survey, 2019). Its main goal is to provide longitudinal panel data that enable research and analysis in support of policies on retirement, health insurance, saving, and economic well-being. New cohorts of age 51 to 56 and their spouses are added to the survey every six years, maintaining its status as nationally representative of households with members over the age of 50. Biennial interviews were conducted through 2014. The enhanced face-to-face (EFTF) interview includes a set of physical performance measures, collection of biomarkers, and a Leave-Behind Questionnaire on psychosocial topics. A random one-half of households were pre-selected for the EFTF in 2006, with the other half of the sample selected for 2008. From that point on every household will repeat the EFTF portion every other wave (Health and Retirement Survey, 2019, Servais, M., 2010).

Study Population

This study population were adults, age 50 and older, who had self-reported diabetes, had an HbA1c measure, and responded to questions asked about financial hardship included in EFTF interviews during 2012 and 2014. Diabetes was defined as individuals answered ‘yes’ to question: “has a doctor ever told you that you have diabetes or high blood sugar?”. 14,419 participants who were ages 50 years and older participated in the blood-based biomarker data collection in the first EFTF interview. Among them, 3,645 participants self-reported with diagnosed diabetes. After excluding 78 participants without HbA1c values measured, and 885 participants without financial hardship information, in total, 2,682 participants were selected for analysis.

Primary Outcome

The primary outcome for this study was glycemic control (HbA1c). HbA1c was assayed from dried blood spot (DBS) samples collected during EFTF interview (Crimmins, Faul, Kim & Weir, 2015). Special informed consent was acquired for the blood acquisition process. Because the resulting biomarker values based on DBS vary across assays and laboratories and may be quite different from the more conventionally used whole blood assays, HRS constructed and released a NHANES equivalent assay value for each assay and recommended the NHANES equivalent assay values for analytic use (Crimmins et al.,2015). We used the NHANES equivalent HbA1c for analysis.

Primary Predictor

The primary predictor for this study was financial hardship. Measures of financial hardship were constructed using four components: (Smith, Ryan, Fisher, Sonnega & Weir, 2017)

1. Difficulty paying bills – dichotomous variable coded as 1 when participants answered somewhat, very, or completely difficult to the question “How difficult is it for you/your family to meet monthly payments on your/your family’s bills?”.

2. Food insecurity – dichotomous variable coded as 1 when participants answered “yes” to the whether they “eat less because there is not enough money”.

3. Medication cost non-adherence – dichotomous variable coded as 1 when participants answered “yes” to the whether they “ever take less medication because of cost”.

4. Ongoing financial strain – dichotomous variable coded as 1 when participants answered “yes” to whether they had “ongoing financial strain” over the past 12 months.

We then created two summary variables of financial hardship:

• 5.>1 Financial hardship – dichotomous variable coded as 1 if respondent answered yes to two or more of the 4 possible hardships.
• 6.Cumulative hardship score – count from 0 to 4 calculated by summing the number of financial hardships for each respondent. situations as a count from 0 to 4.

Covariates

All data came from the biennial core interview data, except total wealth and household income which came from RAND data. Demographic factors included gender, age (grouped into 50-59, 60-69, 70-79, 80+), race/ethnicity (categorized as non-Hispanic White, non-Hispanic Black, Hispanic, and other minority), marital status (yes and no to married or living with a partner), education (grouped as no degree, high school diploma/GED, and higher education); and household income and assets (grouped into quarters). Job status was grouped into working now, retired, and other.

Two variables were included to capture aspects of disease severity: comorbidities and medication usage. Comorbidity included the following diseases: high blood pressure, cancer, lung disease, heart condition, stroke, emotional/psychiatric problems, depression, Alzheimer’s, and arthritis, we grouped comorbidity count 0-2 as low comorbidity, 3-4 as moderate comorbidity, 5 and above as high comorbidity. Medication usage was based on self-report and categorized as No Medication, Oral Medication only, Insulin only, and Both Oral and Insulin.

Statistical Analysis

To analyze whether financial hardship was associated with glycemic control, we conducted univariate analyses and then adjusted analyses on all 6 measures of financial hardship. First, we investigated the characteristics of study participants and the different financial hardship variables. Secondly, we compared participant characteristics by financial hardship response using chi-square tests and investigated univariate relationship of HbA1c on participant characteristics using ANOVA. Thirdly, univariate generalized linear models (GLM) were developed to test the associations between each financial hardship measure as the primary independent variable and the continuous measure of HbA1c as the outcome. We tested interactions for each of the financial hardship measures against ages 50-64 vs. 65+ to ensure there were no differential relationships by age. All interaction terms were not significant, so we did not stratify final models. Finally, we used multivariable GLM models adjusted for demographics and comorbidity covariates to investigate the independent relationship of financial hardship on glycemic control. For each model one of the six financial hardship measures was entered as the primary independent variable, and covariates included age, gender, race/ethnicity, education, marital status, employment, household income/assets, and comorbidity count. All p-values were 2-sided and p<0.05 was considered statistically significant. Statistical analysis was performed with SAS version 9.4 (SAS Institute).

Results

The sample included 2,682 adults with diabetes with a mean age of 69 years. The majority of the sample was non-Hispanic white (59.7%), married or living with a partner (62.9%) and had a high school diploma or higher education (54.8% with high school diploma/GED and 22.7% with higher education). Additional demographic information can be found in Table 1.

Table 1:

Sample demographics of older adults with diabetes (n=2,682)

	Mean or %
Age Group
50-59 years	20.3%
60-69 years	31.1%
70-79 years	33.7%
80+ years	14.9%
Gender
Male	45.2%
Female	54.8%
Race/Ethnicity
NH White	59.7%
NH Black	21.5%
Hispanic	15.6%
Other	3.2%
Education level
No degree	22.5%
High school diploma/GED	54.8%
Higher education	22.7%
Married or living with a partner
Yes	62.9%
No	37.1%
Living with partner/spouse or children
Living with partner/spouse only	45.8%
Living with partner/spouse and children	17.1%
Living with children	10.6%
Living alone	26.5%
Employment status
Working now	22.6%
Retired	52.2%
Other	25.2%
Household income and assets
Mean of 1st Quartile	$7,500
Mean of 2nd Quartile	$73,670
Mean of 3rd Quartile	$211.298
Mean of 4th Quartile	$955,261
Comorbidity count
Low comorbidity (0-1)	45.7%
Moderate comorbidity (2-3)	38.6%
High comorbidity (4+)	15.7%
Medication Use
No medication	17.4%
Oral medication	57.6%
Insulin	9.1%
Both oral and insulin	13.8%

Table 2 provides summary information on glycemic control and responses to financial hardship measures. The mean HbA1c was 6.9% with approximately one third having uncontrolled diabetes (20.1% with HbA1c of 7-8% and 15.7% with HbA1c over 8%). Half of the sample reported ongoing financial strain (50.5%), 38.1% indicated difficulty paying bills, 14.7% indicated medication cost non-adherence, and 7.9% indicated food insecurity.

Table 2:

Glycemic Control and Measures of Financial Hardship in Sample

	Mean (SD) or %
Baseline Blood Hemoglobin (HbA1C)
Mean (SD)	6.93 (1.33)
<6.5%	44.1%
6.5-<7.0%	20.1%
7.0-<8.0%	20.1%
8.0%+	15.7%
Ongoing financial strain
Yes	50.5%
Difficulty paying bills
Yes	38.1%
Medication cost non-adherence
Yes	14.7%
Food insecurity
Yes	7.9%
>1 Financial Hardship Measure
Yes	37.5%
Financial Hardship Count
0	41.5%
1	21.0%
2	25.2%
3	9.3%
4	3.0%

Table 3 provides unadjusted and Table 4 provides adjusted results of the linear models investigating the relationship between measures of financial hardship and glycemic control (HbA1c). In the unadjusted model for the cumulative measure of financial hardship, HbA1c increased by 0.16 for every additional financial hardship measure reported (0.16, 95%CI 0.11,0.20). After adjustment for demographics, socioeconomic status, comorbidities, and medication usage, this relationship remained significant with HbA1c increasing by 0.09 for every additional financial hardship measure reported (0.09, 95%CI 0.04, 0.14). Additional significant relationships after adjustment included the dichotomous measure of >1 financial hardship (0.17, 95%CI 0.07, 0.28), difficulty paying bills (0.25, 95%CI 0.14, 0.35), and medication cost non-adherence (0.17, 95%CI 0.03, 0.31). While socioeconomic status measures, including employment, and income were significantly associated with glycemic control when unadjusted, these relationships lost significance after adjustment for financial hardship measures.

Table 3:

Unadjusted Relationship between Measures of Financial Hardship, Socioeconomic Status and Glycemic Control

	β (95% CI)
Ongoing financial strain	0.19 (0.09, 0.29)
Difficulty paying bills	0.38 (0.27, 0.48)
Medication cost non-adherence	0.36 (0.22, 0.50)
Food insecurity	0.31 (0.12, 0.50)
>1 Financial Hardship Measure	0.32 (0.21, 0.42)
Financial Hardship Count	0.16 (0.11, 0.20)

Bold indicates significant at p<0.05 level

Table 4:

Six Models Investigating the Adjusted Relationship between Financial Hardship and Glycemic Control

	Financial Hardship Variable Entered as Primary Independent Variable
	Ongoing financial strain	Difficulty paying bills	Medication cost non-adherence	Food insecurity	>1 Financial Hardship Measure	Financial Hardship Count
Financial Hardship Measure	0.03 (−0.07, 0.13)	0.25 (0.14, 0.35)	0.17 (0.03, 0.31)	0.16 (−0.02, 0.34)	0.17 (0.07, 0.28)	0.09 (0.04, 0.14)
Age (Ref = 50-59 years)
60-69 years	−0.26 (−0.40, −0.12)	−0.24 (−0.38, −0.10)	−0.25 (−0.39, −0.11)	−0.26 (−0.40, −0.12)	−0.24 (−0.38, −0.10)	−0.23 (−0.38, −0.09)
70-70 years	−0.37 (−0.53, −0.21)	−0.34 (−0.50, −0.19)	−0.36 (−0.51, −0.20)	−0.37 (−0.52, −0.21)	−0.34 (−0.50, −0.19)	−0.33 (−0.49, −0.18)
80+ years	−0.40 (−0.60, −0.21)	−0.36 (−0.55, −0.16)	−0.38 (−0.58, −0.19)	−0.40 (−0.59, −0.21)	−0.36 (−0.56, −0.17)	−0.35 (−0.54, −0.15)
Gender (Ref = male)	0.001 (−0.10, 0.10)	−0.004 (−0.10, 0.09)	−0.002 (−0.10, 0.10)	0.002 (−0.10, 0.10)	−0.003 (−0.10, 0.10)	−0.003 (−0.10, 0.10)
Race/Ethnicity (Ref = NHW)
NH Black	0.25 (0.12, 0.37)	0.24 (0.11, 0.36)	0.24 (0.11, 0.37)	0.24 (0.12, 0.37)	0.24 (0.11, 0.36)	0.23 (0.11, 0.36)
Hispanic	0.21 (0.06, 0.36)	0.21 (0.06, 0.36)	0.20 (0.06, 0.35)	0.20 (0.05, 0.35)	0.21 (0.06, 0.36)	0.21 (0.06, 0.36)
Other	0.20 (−0.07, 0.47)	0.19 (−0.08, 0.46)	0.19 (−0.08, 0.46)	0.19 (−0.08, 0.46)	0.19 (−0.08, 0.46)	0.18 (−0.09, 0.45)
Education (Ref = no degree)
High school diploma/GED	−0.02 (−0.15, 0.10)	−0.02 (−0.15, 0.10)	−0.03 (−0.15, 0.10)	−0.02 (−0.15, 0.10)	−0.03 (−0.16, 0.09)	−0.03 (−0.16, 0.09)
Higher education	−0.07 (−0.23, 0.09)	−0.07 (−0.22, 0.09)	−0.08 (−0.23, 0.08)	−0.07 (−0.23, 0.09)	−0.08 (−0.23, 0.08)	−0.08 (−0.24, 0.08)
Marital status (Ref = married)	−0.07 (−0.18, 0.03)	−0.06 (−0.17, 0.05)	−0.07 (−0.18, 0.04)	−0.08 (−0.19, 0.03)	−0.07 (−0.18, 0.04)	−0.07 (−0.18, 0.04)
Employment (Ref = working now)
Retired	−0.10 (−0.24, 0.04)	−0.08 (−0.22, 0.06)	−0.10 (−0.24, 0.04)	−0.10 (−0.24, 0.04)	−0.09 (−0.23, 0.05)	−0.08 (−0.22, 0.06)
Other	−0.18 (−0.33, −0.04)	−0.19 (−0.34, −0.05)	−0.18 (−0.33, −0.04)	−0.19 (−0.34, −0.05)	−0.19 (−0.33, −0.04)	−0.18 (−0.33, −0.04)
Household income and assets (Ref = 1st quartile)
2nd quartile	−0.005 (−0.14, 0.15)	0.04 (−0.10, 0.18)	0.01 (−0.13, 0.15)	0.02 (−0.12, 0.16)	0.02 (−0.12, 0.17)	0.03 (−0.11, 0.18)
3rd quartile	−0.14 (−0.29, 0.02)	−0.08 (−0.23, 0.08)	−0.12 (−0.28, 0.03)	−0.11 (−0.27, 0.04)	−0.10 (−0.25, 0.06)	−0.08 (−0.24, 0.07)
4th quartile	−0.06 (−0.22, 0.11)	0.04 (−0.13, 0.21)	−0.05 (−0.21, 0.12)	−0.04 (−0.21, 0.12)	0.005 (−0.16, 0.17)	0.03 (−0.14, 0.20)
Comorbidity count (Ref = 0-2 comorbidities)
3-4 comorbidities	−0.10 (−0.21, 0.004)	−0.11 (−0.22, −0.01)	−0.10 (−0.21, 0.000)	−0.10 (−0.21, 0.003)	−0.11 (−0.21, −0.005)	−0.11 (−0.22, −0.01)
5+ comorbidities	−0.19 (−0.34, −0.05)	−0.22 (−0.36, −0.07)	−0.21 (−0.36, −0.07)	−0.20 (−0.34, −0.05)	−0.22 (−0.36, −0.07)	−0.23 (−0.38, −0.09)
Medication Use (Ref = no medication)
Oral medication	0.54 (0.42, 0.67)	0.54 (0.41, 0.67)	0.55 (0.42, 0.67)	0.55 (0.42, 0.68)	0.54 (0.42, 0.67)	0.55 (0.42, 0.67)
Insulin	1.44 (1.25, 1.63)	1.43 (1.24, 1.61)	1.44 (1.25, 1.63)	1.45 (1.26, 1.63)	1.43 (1.25, 1.62)	1.44 (1.25, 1.62)
Both oral and insulin	1.57 (1.40, 1.74)	1.57 (1.40, 1.74)	1.57 (1.40, 1.73)	1.57 (1.41, 1.74)	1.57 (1.40, 1.74)	1.57 (1.40, 1.74)

Bold indicates significance at p<0.05 level

In each model the outcome was glycemic control and the primary independent variable was one of six financial hardship variables as indicated in first row.

Discussion

This study found among a sample of older adults with diabetes that financial hardship had a cumulative influence on glycemic control, with each additional hardship associated with approximately a 0.1% increase in HbA1c. In addition, difficulty paying bills and cost related medication non-adherence were associated with worse glycemic control (higher HbA1c) after adjusting for demographics, socioeconomic status, and comorbidities. This relationship was significant in unadjusted models for ongoing financial strain and food insecurity, however, it lost significance after adjustment. Standard socioeconomic status measures, such as education, employment, and income were not independently associated with glycemic control after adjusting for financial hardship factors. These results suggest that in older adults, financial hardship has an influence on glycemic control and specific measures of hardship may be better to investigate the influence of socioeconomic status on clinical outcomes.

This study is unique in that it employed four measures of financial hardship directly related to diabetes outcomes: difficulty paying bills, food insecurity, medication cost non-adherence, and ongoing financial strain. While prior research found a relationship between cost related medication non-adherence and outcomes in diabetes (Berkowitz et al., 2015; Berkowitz, Seligman, & Choudhry, 2014; Kang et al., 2018; Lee et al., 2018; Ngo-Metzger et al., 2012; Williams et al., 2013; Zhang et al., 2014) this is one of the first studies to highlight the importance, from a health standpoint, of understanding if older adults with diabetes are struggling with paying bills (Levy, 2015; Simons et al., 2016). In addition, by investigating a count of financial hardship measures, we found a cumulative relationship with glycemic control. In a prior study of different kinds of material needs, Berkowitz et al. (2015) found that increased number of material needs, such as food, medication, housing, and energy was associated with worse diabetes control and increased health care use. Approximately 40% of respondents in the clinic-based population from Berkowitz et al.’s (2015) paper and the older adult population from this study reported >1 hardship. There was lower reporting of cost related medication non-adherence and food insecurity in our study, which may be due to the singular questions asked, as opposed validated scales used by Berkowitz that capture multiple domains around food and medication shortage (Berkowitz et al., 2015). However, we found 38% of the sample of older adults reported difficulty paying bills and 50% reported ongoing financial strain, which is quite high and highlights the financial hardship experienced by many older adults with diabetes.

Based on our results and the existing literature, relying on traditional measures of SES (such as income and education) without accounting for measures of financial hardship may mask the relationship between financial hardship and glycemic control (Assari, Moghani Lankarani, Piette, & Aikens, 2017). For example, a recent study examining the role of SES on glycemic control in adults with diabetes found that SES was marginally associated with glycemic control in subpopulations of the sample, however SES overall was not associated with glycemic control across the study population. Houle et al. (2016) also examined the role of SES on glycemic control measuring the direct relationship of income, education, and employment with glycemic control and found that each SES measure was mediated by psychosocial and health behavior factors. Taken together, reliance on traditional measures of SES may not be sufficient to understanding how SES influences glycemic control in older adults with diabetes (Houle et al., 2016). However, the results of the current study suggest considering SES as an important social determinant of health that can be captured by more specific measures of financial hardship. A more comprehensive assessment of financial hardship may need to be considered when designing surveys intending to capture the role of SES in health. Interventions directly addressing social determinants, such as financial hardship, in patients with diabetes is lacking. In addition, these results suggest the importance of considering new lines of questions for providers to assess financial hardship when treating older adult patients. Rather than relying on traditional measures, such as income, or only one type of financial hardship, it may be important to understand the broader financial context and how this influences a patients’ life. (Hill, Nielsen, & Fox, 2013) Questions regarding possible mediators of the relationship between financial hardship and health, such as stress, should also be considered. In addition, these results highlight the importance of interdisciplinary care teams that include clinicians and social workers trained in addressing financial challenges. (Ambros-Miller & Ashcroft, 2016; Del Valle & McDonnell, 2018; McGill, Blonde, Chan, et al., 2017)

While this study focused on a diverse and nationally representative older adult sample, there are limitations that are worth mentioning. First, data used in this analysis is cross-sectional and therefore, cannot be used to infer causality. Second, there are a number of variables with an important role in glycemic control, such as diet and physical exercise that were not captured in the analysis. Third, as financial hardship can change over time, this analysis can only comment on the relationship at the time questions were asked. Further studies should seek to understand the longitudinal relationship between financial hardship and diabetes outcomes, and the possible influence of trajectories of SES rather than one-time measures. Finally, this analysis does not investigate the mechanism for the relationship between financial hardship and glycemic control. Further studies using longitudinal data or path analysis should investigate possible pathways, such as change in employment, lack of support family, changes in self-care behaviors, and stress.

Conclusion

In conclusion, in a sample of older adults, financial hardship is associated with glycemic control, specifically, difficulty paying bills and cost-related medication non-adherence. In addition, every additional financial hardship was associated with an increased HbA1c of 0.1%. Given the increasing costs of care, and limited financial resources of many older adult populations, strategies are needed to address the ongoing financial hardship in older adults with diabetes to achieve optimal diabetes control.