Introduction

In the United States, over 34 million people live with type 2 diabetes mellitus (T2DM), a debilitating and demanding chronic condition that influences several facets of an individual’s life [1]. T2DM, a leading cause of disability, morbidity, and mortality, is associated with higher rates of hospitalization [2] and medical expenditure burden of $46 billion per annum [3]. Additionally, T2DM and its associated complications such as neuropathy, nephropathy, retinopathy, cardiovascular disease, stroke, and peripheral vascular disease can negatively influence perceived health status and functional status, which have implications for quality of life [4–6], a construct significantly associated with glycemic control [7, 8].

The World Health Organization defines quality of life (QOL) as a person’s perception of his/her relative position within the context of his/her culture and values in comparison to his/her goals, expectations, and standards [9]. QOL has different domains including the physical and mental health components [9]. T2DM has been shown to have deleterious effects on both physical and mental aspects of QOL [10–13]. Studies have assessed factors associated with QOL in individuals with T2DM [11, 13–20]. These include demographic factors such as age [13]; self-care factors such as physical activity, glucose monitoring, diet [14], and medication adherence [11]; clinical factors such as fasting blood sugar [13], diabetes complications [14, 15], multimorbidity [19], insulin use [11, 16], and depression [17]; and psychosocial factors such as perceived control of diabetes [18], perceived discrimination [20], meaning of illness [21], and diabetes fatalism [22]. However, the distribution of these factors across racial groups in individuals with diabetes is not uniform. Though the evidence on racial differences in QOL in adults with diabetes is mixed [23–26], there is overwhelming evidence that ethnic and racial minorities with T2DM are worse in self-care and clinical measures associated with QOL including glycemic control [27] and diabetes-related complications [28].

Given the relationship between QOL and glycemic control, interventions have been developed and tested to improve QOL in adults with T2DM [29, 30]. However, the limited knowledge on racial and ethnic differences in factors associated with QOL may pose a barrier to the development of effective interventions that attend to the variations across racial and ethnic groups. To address this gap in the literature, this study aimed to examine racial differences in the unique contributions of demographic, social determinants, clinical, and self-care factors on QOL in a sample of adults with T2DM.

Methods

Study Population

The study recruited 601 White and Black adults. They were recruited from two adults primary care clinics in Southeastern United States. The Institutional Review Board of the Medical University of South Carolina and the Veterans Affairs (VA) Research and Development Committee approved all study procedures. Individuals aged 18 years and older, proficient in the English Language, and with a diagnosis of type 2 diabetes confirmed from the participants’ medical records were included in the study. Exclusion criteria included individuals with impaired cognition from significant dementia or active psychosis.

Before being consented, a detailed explanation of the study was provided to all eligible adults. The study questionnaire was administered to all. The validated questionnaire was informed, by the Brown et al. conceptual framework, and assessed demographic information, social determinants of health factors, psychosocial factors, self-care behaviors, and clinical factors [31].

Demographic Variables

The demographic variables assessed included age, sex, race, marital status, recruitment site, and insurance. Age was treated as a continuous variable. Race was dichotomized as White and Black. Insurance was categorized as none, private, Medicare, Medicaid, VA, and others. Marital status was categorized as never married, married, separated, and widowed. Recruitment occurred at two sites: The Medical University of South Carolina and the Ralph H. Johnson Veteran Affairs Medical Center.

Socioeconomic Factors

Years of education and income were assessed with a previously validated questionnaire from the 2002 National Health Interview Survey [32]. Education was treated as a continuous variable. Annual personal income was categorized as <$9999, $10,000–$14,999, $15,000–$19,999, $20,000–$24,999, $25,000–$34,999, $35,000–$49,000, $50,000–$74,999, and $75,000 and more. Hours of work per week was treated as a continuous variable.

Psychosocial Factors

Psychosocial factors assessed included diabetes fatalism, depression, diabetes distress, self-efficacy, social support, and perceived stress.

Depression

This was assessed with a PHQ-9 instrument. The PHQ-9 is a 9-item scale modeled after the DSM-IV criteria to diagnose depression. A higher score indicates more severe depression. A score of ≥10 from the PHQ-9 instrument has a specificity of 88% and a sensitivity of 88% for major depression [33].

Diabetes Fatalism

This was assessed with the 12-item Diabetes Fatalism Scale (DFS-12). The items on the DFS-12 are scored using a 6-point Likert scale with scores ranging from 1 = strongly disagree to 6 = strongly agree. The higher score correlates to more diabetes fatalistic attitude [34]. DFS-12 has a Cronbach’s alpha of 0.80.

Diabetes Distress

This was assessed with the Diabetes Distress Scale (DDS). The 17-item instrument measures different dimensions of distress: access to care, support, emotional burden, and disease management [35].

Psychological Distress

This was assessed with the Kessler Psychological Distress Scale (K6). The instrument is a 6-item scale with higher scores correlating to a higher probability of severe mental illness. The psychometric properties of K6 have been shown to be consistent across multiple subpopulations [36].

Self-Efficacy

This was assessed with the Perceived Diabetes Self-Management Scale (PDSMS). A higher score on the 8-item instrument is indicative of a higher self-efficacy [37]. PDSMS has a Cronbach’s alpha of 0.83.

Social Support

This was assessed with the Medical Outcomes Study (MOS) Social Support Survey. The 19-item instrument measures positive social interaction, emotional, or informational support, affectionate support, and tangible support. The 1-year test-retest reliability for MOS-SS is 0.72 to 0.76 [38].

Perceived Stress

This was assessed with the Perceived Stress Scale (PSS). The 4-item scale measures the frequency of stressful events participants encountered over the previous month [39].

Built Environment Factors

Built environment was assessed using six scales and four indices based on a prior validation study [40]. A higher score is indicative of more perceived neighborhood problems.

The variables that define the built environment differ in the number of an item scale measures. The aesthetic environment consisted of 7 items; neighborhood walking/exercise environment consisted of 11 items; social cohesion consisted of 5 items; access to healthy foods consisted of 11 items. The scale’s response ranges from 1 (strongly agree) to 5 (strongly disagree). Recreation facilities index had 8 items which measure the presence of recreational facilities in the neighborhoods with a “yes” or “no” response. Neighborhood crime and violence have 4 items and 3 items, respectively. The response categories for neighborhood violence range from 1 (strongly agree) to 5 (strongly disagree).

Neighborhood Problems Index

The index was assessed with a 16-item scale. The scale response ranges from 1 to 3, with 1 indicating neighborhood attribute was not a problem; 2, it was somewhat of a problem; and 3, it was a big problem. The sum of the responses represented the neighborhood problems index score.

Neighborhood Participation Index

The neighborhood participation index was measured with a 12-item scale. The index measures an individual’s involvement in civic and political activities with their neighbors.

Clinical Factors

The Charlson comorbidity index was used to estimate comorbidity, and it was reported as a continuous score [41]. The duration of diabetes was collected in years and treated as a continuous variable. Body mass index (BMI) was estimated from the most recent height and weight obtained from the electronic medical record and used as a continuous variable. Participants were asked if they use insulin or had eye problems or kidney problems; these variables were dichotomized as yes or no. Perceived health status was used as a continuous variable.

Knowledge and Self-Care Factors

Diabetes Knowledge

This was assessed with the Diabetes Knowledge Questionnaire (DKQ). The final score from the 24-item scale is estimated from the percentage of correct scores [42].

Participants were asked about their blood sugar, foot care, exercise, and diet (general and specific). These variables were each treated as a continuous variable. Information on smoking status was collected and categorized as never smoker, current smoker, and former smoker.

Outcome Measure

Consisted of the physical health and mental health components, QOL was assessed with the SF-12. The 12-item scale provides a summary of mental health (MCS-12) and physical health (PCS-12) component outcome scores which were reported as continuous variables. The SF-12 is widely used because its outcome scores are similar to the longer SF-36 version in a variety of populations [43, 44]. Questions used to assess mental health include how often any emotional problems, such as feeling anxious or depressed, lead to respondents not doing activities as carefully as usual, or accomplishing less than one would like. Questions used to assess the physical health component include how often physical health limits activities such as climbing several flights of stairs or moving a table.

Statistical Analysis

The sociodemographic characteristics of the sample were assessed using chi square for categorical variables and t-test for continuous variables and are presented by race. Linear regression models were then used to model the relationship between the demographic, social determinants, clinical, knowledge, and self-care factors and the two components of QOL for the entire sample. This was followed by linear regression models to assess the relationship between the demographic, social determinants, clinical, knowledge, and self-care factors and QOL for Whites and Blacks. The analyses were conducted using STATA SE 15.1, and statistical significance was assessed using a two-sided p-value of 0.05.

Results

The baseline characteristics by race for the 601 adults with T2DM participating in this study are presented in Table 1. Two-thirds of the participants were Black, and one-third was White. The baseline demographics differed by race with the exception of mean duration of diabetes, mean comorbidity, site, and perceived health status. White participants were more likely to be older (65 years versus 60 years), married (63% versus 42%), have more years of education (14 years versus 13 years), and have higher income (55% versus 22% with income $35K or more) compared with Blacks. Black participants were more likely to be female (47% versus 23%), separated/divorced (32% versus 22%), and earn less (68% versus 46% with income $35K or less). White and Black adults reported similar physical component score (PCS) and mental component score (MCS) of quality of life.

Table 1 Sample demographics by race
Full size table

Table 2 shows the results of the adjusted linear regression model examining the relationship between demographic, social determinants, clinical, knowledge, and self-care factors, PCS, and MCS for the full sample. For PCS, significant associations were observed with psychological distress (β = 0.02, p < 0.01), neighborhood aesthetics (β = 0.05, p < 0.01), neighborhood safety (β = 0.01, p < 0.05), neighborhood crime (β = −0.15, p < 0.05), and neighborhood comparison (β = 0.13, p < 0.05). For MCS, significant associations were observed with depression (β = −0.06, p < 0.05), psychological distress (β = −0.09, p < 0.001), perceived stress (β = −0.12, p < 0.01), and perceived health status (β = −0.33, p < 0.01).

Table 2 Adjusted linear regression model of the relationship between demographic, social determinants, clinical, knowledge, and self-care factors, PCS, and MCS
Full size table

Table 3 provides the results of adjusted linear regression model examining the relationship between demographic, social determinants, clinical, knowledge, and self-care factors and PCS by race. Significant relationships existed between PCS and psychological distress (β = 0.02, p < 0.01), neighborhood aesthetics (β = 0.05, p < 0.01), neighborhood walking environment (β = −0.02, p < 0.05), access to healthy food (β = 0.01, p < 0.05), neighborhood crime (β = −0.15, p < 0.05), and neighborhood comparison (β = 0.13, p < 0.05). After stratification by race, there were significant relationships between PCS and Medicaid (β = −0.86, p < 0.05), neighborhood participation index (β = −0.05, p < 0.05), and perceived health status (β = −0.19, p < 0.05) in Whites. In Blacks, there were significant relationships between PCS and hours worked per week (β = 0.01, p < 0.05), psychological distress (β = 0.03, p < 0.05), neighborhood aesthetics (β = 0.06, p < 0.05), neighborhood walking environment (β = −0.03, p < 0.05), and neighborhood crime (β = −0.24, p < 0.05).

Table 3 Adjusted linear regression model of the relationship between demographic, social determinants, clinical, knowledge, and self-care factors and PCS stratified by race
Full size table

Table 4 provides the results of adjusted linear regression models examining the relationship between demographic, social determinants, clinical, knowledge, and self-care factors and MCS by race. Significant relationships existed between MCS and depression (β = −0.06, p < 0.05), psychological distress (β = −0.09, p < 0.001), perceived stress (β = −0.12, p < 0.01), and perceived health status (β = −0.33, p < 0.01). After stratification by race, there were significant relationships between MCS and Medicaid (β = −2.42, p < 0.01), other insurance (β = −1.90, p < 0.05), depression (β = −0.09, p < 0.05), and neighborhood problem index (β = 0.15, p < 0.05) in Whites, while significant relationships existed between MCS and Medicaid (β = 1.33, p < 0.05), psychological distress (β = −0.11, p < 0.001), perceived stress (β = −0.13, p < 0.01), perceived health status (β = −0.35, p < 0.05), and diabetes knowledge (β = −0.08, p < 0.05) for Blacks.

Table 4 Adjusted linear model of the relationship between demographic, social determinants, clinical, knowledge, and self-care factors and MCS stratified by race
Full size table

Discussion

This study assessed racial differences in factors independently associated with PCS and MCS in adults with T2DM. In the total sample, psychological distress, neighborhood aesthetics, neighborhood walking environment, access to healthy food, neighborhood crime, and neighborhood comparison were significantly associated with the physical component of QOL, and depression, psychological distress, perceived stress, and perceived health status were significantly associated with the mental health component of quality of life. In the stratified models, there were racial differences in the factors independently associated with both PCS and MCS. For Whites, PCS was significantly associated with Medicaid, neighborhood participation index, and perceived health status, while MCS was significantly associated with Medicaid, other insurance, depression, and neighborhood participation index. In Blacks, PCS was significantly associated with hours worked per week, psychological distress, neighborhood aesthetics, neighborhood walking environment, neighborhood participation index, and perceived health status, while MCS was significantly associated with Medicaid, psychological distress, social support, perceived stress, perceived health status, and diabetes knowledge. These findings suggest that there are racial differences in the factors associated with QOL, and therefore, interventions to improve quality of life in adults with T2DM should be tailored accordingly.

Prior studies have examined the construct of quality of life in individuals with T2DM and demonstrate an association between quality of life and various self-care and clinical constructs [11, 13, 14, 16, 17]. Though limited evidence exists on racial differences in QOL across racial and ethnic groups with diabetes, a study examining health-related quality of life (HRQL) in a sample of 186 Black adults with T2DM found that HRQL was associated with socioeconomic and familial factors (money, housing, crime, family, caretaking responsibilities), obesity, insulin, and comorbidity [45]. However, the current study is novel in that it provides unique insight into racial differences in the demographic, socioeconomic, clinical, and self-care correlates of quality of life in individuals with T2DM.

Given the persistent health disparities in the USA, there is a need for innovative interventions that improve health outcomes and quality of life in adults with T2DM. Several studies have tested the efficacy and effectiveness of interventions in improving quality of life in ethnically and racially heterogenous samples of adults with diabetes and have shown promising results [29, 30]. Additionally, studies have tested interventions focused on psychological distress, social support, and knowledge in Blacks demonstrating that incorporating these constructs in interventions are effective in improving targeted outcomes in Blacks with T2DM [46, 47]. However, effective measures to ameliorate disparities in diabetes-related outcomes, particularly quality of life, in adults with T2DM remain elusive. Hence, it is important that clinicians caring for patients with T2DM should be informed about in factors associated with quality of life across various racial and ethnic groups, especially since a standardized, uniform approach for all patients may not be as effective. While it is important that interventions are culturally appropriate, it also vital that interventions should address factors that have been shown to be scientifically relevant for the target populations.

The current study has some limitations that need to be acknowledged. First, this was a cross-sectional study, which precludes any assumptions of causal associations between the variables being studied and QOL. Second, the data used in this study were consisted of self-reported measures, which could be a potential source of bias. Third, the participants in the study were recruited from two primary clinics in Southeastern United States, which limits our ability to generalize our findings to other regions of the USA. Lastly, subgroup analyses were not performed in the current study; hence, it is challenging to ascertain if the differences in MCS and PCS can be explained by the differences in demographic characteristics among Black and White participants. In spite of these weaknesses, the use of validated instruments and our large sample size are notable strengths of the study.

Conclusion

The findings from this study illustrate racial differences in factors associated with quality of life in a sample of adults with T2DM. Our analysis revealed variations in factors associated with quality of life among Whites and Blacks with T2DM, which could prove useful in advancing efforts to develop tailored health interventions that consider the unique needs of specific racial groups. Future studies should seek to understand racial differences in the effectiveness of multilevel interventions that address relevant social determinants, clinical, knowledge, and self-care factors aimed at improving quality of life in adults with T2DM. Future studies should also investigate disparities in QOL outcomes using larger sample size and across diverse geographical areas.