PART II: RACE, ETHNICITY, AND HEALTH

Diabetes-Specific Emotional Distress among African Americans and Hispanics with Type 2 Diabetes Michael S. Spencer, PhD, MSW Edith C. Kieffer, PhD, MPH Brandy R. Sinco, MS Gloria Palmisano, MA J. Ricardo Guzman, MPH, MSW Sherman A. James, PhD Gwendolyn Graddy-Dansby, MD Jacqueline Two Feathers, PhD, MPH Michele Heisler, MD, MPA Abstract: This study examines baseline levels and correlates of diabetes-related emotional distress among inner-city African Americans and Hispanics with type 2 diabetes. The Problem Areas in Diabetes (PAID) scale, which measures diabetes-related emotional distress, was administered to 180 African American and Hispanic adults participating in the REACH Detroit Partnership. We examined bivariate and multivariate associations between emotional distress and biological, psychosocial, and quality of health care variables for African Americans and Hispanics. Scores were significantly higher among Hispanics than African Americans. Demographic factors were stronger predictors of emotional distress for Hispanics than for African Americans. Daily hassles, physician support, and perceived seriousness and understanding of diabetes were significant for African Americans. Understanding the personal, family and community context of living with diabetes and conducting interventions that provide support and coping strategies for self-management have important implications for reducing health disparities among disadvantaged racial and ethnic groups. Key words: Diabetes, health disparities, emotional distress, African Americans, Hispanics.

MICHAEL SPENCER and EDITH KIEFFER are Research Associate Professors at the University of Michigan School of Social Work, where BRANDY SINCO is a Research Associate/Statistician. JACQUELINE TWO FEATHERS is a graduate of the University of Michigan School of Public Health. GLORIA PALMISANO is the Project Manager of the REACH Detroit Partnership. J. RICARDO GUZMAN is the Chief Executive Officer of Community Health & Social Service Centers, Inc. in Detroit. SHERMAN JAMES is the Susan B. King Professor of Public Policy Studies at Duke University. GWENDOLYN GRADDY-DANSBY is the Director of the Center for Senior Independence at the Henry Ford Health System in Detroit. MICHELE HEISLER is a VA Research Scientist and Assistant Professor at the Veterans Affairs Center for Practice Management & Outcomes Research and the University of Michigan, respectively. Michael Spencer can be reached at [email protected]. Journal of Health Care for the Poor and Underserved  17 (2006): 88–105.

Spencer, Kieffer, Sinco, et al.

D

89

iabetes is a major clinical and public health challenge, especially in socially and economically disadvantaged populations.1–8 About 17 million Americans (6.2%) have type 2 diabetes mellitus (also known as adult-onset or non-insulin dependent diabetes),9 with over $100 billion in direct and indirect costs spent yearly on medical care for patients with type 2 diabetes.10 African Americans and Hispanics experience a 50–100% higher burden of illness and mortality due to type 2 diabetes than non-Hispanic White Americans.3–7 Nationwide, African American and Hispanic adults with diabetes have worse glycemic and blood pressure control than other groups, and there is growing evidence that there are racial and ethnic differences in diabetes processes of care, even among fully insured patients.11–14 In addition to the physical burden of diabetes, adults with type 2 diabetes are twice as likely as adults in the general population to experience serious psychological distress, including mental health problems.15 Moreover, adults with both serious psychological distress and diabetes are more likely than those with only diabetes to live in poverty, report poor health, and lack access to health care.15–16 Physical and environmental stressors, however, are not the only source of stress for adults with type 2 diabetes. Illness-related emotional distress can arise from the often burdensome and continual self-care demands related to diabetes, such as testing and monitoring blood sugar levels, taking medication, injecting insulin, monitoring food consumption, and engaging in regular physical activity.16 Diabetesrelated emotional distress in turn has been found to be a significant contributor both to poor adherence to diabetes self-care recommendations and to poor glycemic control.16–17 In addition to the often burdensome daily treatment regimens, adults with type 2 diabetes must also cope with the threat of serious diabetes-related complications, including increased incidence of kidney disease, amputation, blindness, and the potential for reduced life expectancy.18–19 The constant coping that is required of adults with diabetes can wear on an individual’s psyche and adversely affect interpersonal, social, and work functioning.16 The biopsychosocial model originally proposed by Engel offers additional insights into how chronic illnesses such as diabetes can affect daily life.20 The biopsychosocial model conceptualizes the mind and body as two important systems that are interlinked, treating the biological, psychological, and social issues as systems of the body. The model also makes a distinction between pathological processes that cause disease and patient illness, which includes the individual’s perception of his or her health and its subsequent impact on health. Applying this model to the present study, stress and coping with diabetes can directly affect the severity of disease through psychophysiologic processes and indirectly through patients’ own perception of illness by disrupting diabetes treatment adherence and daily functioning. Furthermore, social factors, such as the quality of health care services can further exacerbate perception of patient illness and illness-related emotional distress. While the difficulties many patients face adjusting to life with diabetes have been well documented,21–24 few studies have examined racial and ethnic differences in the association between type 2 diabetes and emotional distress.25 Existing studies suggest that racial and ethnic minority status is associated with higher levels of psychological distress and type 2 diabetes.24–31 Racial and ethnic minorities are

90

Diabetes-specific emotional distress

often disproportionately exposed to social risk factors for health and psychological problems. Cultural influences can also affect an individual’s experience with diabetes. For example, racial and ethnic minorities may differ in their perception of physical or emotional experiences as a problem, causal attribution about disease, understanding of self-care recommendations, and the daily implementation of these recommendations.32 Additional barriers to health include cultural inappropriateness of treatment; linguistic barriers; differential quality of interpersonal care or patient-provider communication; and access to and receipt of appropriate diagnostic, preventive, and therapeutic services and modalities.4,33–35 In the present study, we examine baseline levels and correlates of diabetes-related emotional distress among inner-city African American and Hispanic adults with type 2 diabetes who agreed to participate in a multi-faceted, community-based program to improve diabetes self-management and outcomes (REACH Detroit, described below).36 We use a biopsychosocial framework to identify potential correlates within each group.

Methods Sample and procedures. The data for the present study come from the baseline survey of the REACH Detroit Family Intervention (N5180).36 Detroit is one of a number of cities in the United States to receive funding from the Centers for Disease Control and Prevention (CDC) to implement a Community Action Plan as part of the Racial and Ethnic Approaches to Community Health (REACH) 2010 Initiative.37 REACH 2010 is the CDC’s cornerstone initiative aimed at eliminating disparities in health status experienced by ethnic minority populations in key health areas, including diabetes, cardiovascular disease, infant mortality, breast and cervical cancer, HIV/AIDS, and child and adult immunizations. The purpose of REACH Detroit is to inform, educate, and empower families and communities to reduce risk factors for diabetes and its complications through family and health system interventions, social support group activities, and community-wide diabetes awareness and healthy lifestyle resource development activities among African American and Hispanic residents of the Eastside and Southwest Detroit communities. Eligible participants were at least 18 years of age, had physician-diagnosed type 2 diabetes, and were African American or Hispanic residents of Eastside or Southwest Detroit as defined by ZIP codes. People who already had significant ­diabetesrelated complications, such as blindness, amputated limbs, and kidney failure were excluded. Lists of eligible participants were identified and recruited through physicians’ offices at three Detroit health care systems, including a federally qualified comprehensive health center (FQHC) and two large urban hospitals. Because lab values of diabetes intermediate outcomes (e.g., HbA1cs) were necessary to assess intervention outcomes, the study participants had some form of health insurance or access to free medical care through one clinic site in Southwest Detroit. A total of 346 eligible adults were initially contacted to participate in the study and 180 (52%) of those contacted agreed to participate. Given our sample size, variance, and the minimum detectable difference in our dependent variable, we can detect mean

Spencer, Kieffer, Sinco, et al.

91

shifts of 1.0 at powers of .94 in the Hispanic data and 1.00 in the African American data. At powers of .80, we can detect a .80 shift among Hispanics and .50 among African Americans. Furthermore, we have significant statistical power to detect a mean shift of one-third of the standard deviation (2.4) for Hispanics and a mean shift of .28 of the standard deviation for African Americans (1.78). Measures. REACH Detroit baseline survey. A comprehensive, three-hour baseline survey was conducted in person, usually in the household of the participant, by trained staff, in the chosen language of the participant (Spanish or English) between February 2002 and October 2003. Bio-medical measurements were abstracted from medical records. The interview consisted of items from the Behavioral Risk Factor Surveillance System (BRFSS), a CDC-administered survey used to track health risks in the United States,38 and a battery of assessments about health, health care, behaviors and attitudes toward diabetes, quality of diabetes care, relations with health care providers, and dietary and physical activity practices. All instruments were translated into Spanish and pre-tested by bi-lingual interviewers as a means of testing the culturally and linguistically appropriate instruments. Socio-demographic variables included in all analyses as predetermined control variables were sex (female50, male51), household status (living with one or more persons50, living alone51), and age (under age 5550, age 55 or older51). Although information on income was obtained, the excessive amount of missing values led to its exclusion from further analyses. Thus, respondents’ education (less than a high school diploma50, high school graduate51) serves as our measure of socioeconomic status. Respondents’ primary language was included in analyses for Hispanics (English50, Spanish51). Of the Hispanics in our sample, 77.5% were immigrants and 76% of these individuals reported being in the United States for less than 5 years. Since number of years in the U.S. was highly correlated with respondents’ primary language being Spanish, only the primary language variable was used in our analyses. Since African Americans reside in both Eastside and Southwest Detroit, community of residence was a control variable in models for African Americans (Eastside Detroit50, Southwest Detroit51). All Hispanics in the study reside in Southwest Detroit; therefore community residence was not included as a control variable in models for Hispanics only. The two communities of interest in this study differ in very important ways. According to the 2000 U.S. Census, the Southwest was 44% Hispanic and 24% African American, compared with the Eastside, which is 89% African American and 3% Hispanic. In our sample, about 26% of the Southwest Detroit participants were African American. While the city of Detroit has seen a considerable decline in population and resources over the past 30 years, the Southwest has continued to attract new immigrants, has a strong commercial district, and is one of the most densely populated areas in Detroit. Although the Eastside is more representative of the larger city, it possesses unique resources and recently has been targeted for revitalization through new housing developments. Both areas maintain a strong sense of loyalty to community and boast a number of cultural and community organizations. Biological variables. Several biological factors hypothesized to be associated with

92

Diabetes-specific emotional distress

diabetes-related emotional distress were included in the analyses. Glycemic control, as measured by respondents’ hemoglobin A1C (HbA1c), was reported from the physicians’ office. Respondents’ self-reported diabetes-related complications (mean number of complications), and other chronic illnesses (mean number) were also included. Psychosocial variables. Two measures of stress were administered, including the four-item version of Cohen’s Perceived Stress Scale (PSS-4)39–40 and a 10-item scale adapted from the Delongis, Folkman, and Lazarus hassles and uplifts.41 The Perceived Stress Scale is a 4-item version of the longer PSS-14 scale that asks respondents in the last month: how often have you felt that you were unable to control the important things in your life; how often have you felt confident about your ability to handle your personal problems; how often have you felt that things were going your way; and how often have you felt difficulties were piling up so high that you could not overcome them? A reliability analysis from our sample resulted in a Cronbach’s alpha coefficients of .54 for the PSS-4. Cohen and Williamson report a reliability coefficient of .60 for the PSS-4 and demonstrate its construct validity in a national probability sample concluding that its use is appropriate for situations requiring a very brief measure of stress perceptions.40 The adapted 10 items from the daily hassles scale ask how often respondents were bothered in the past 12 months by: the way that your children are treated at school; problems with your children; hassles at work; trouble balancing work and family demands; problems in relationships with close friends, neighbors, relatives other than immediate family; family health problems; responsibilities for elder family members; problems with transportation or getting to places you need to go; problems with rats, mice, bugs in the place you are living now; and problems with city services, such as garbage pick up or snow removal. Reliability analysis from our sample resulted in a Cronbach’s alpha coefficient of .71. Additionally, two questions were used to measure the respondents’ perceived seriousness of their case of diabetes (How serious a case of diabetes do you think you have? 1 to 5, very mild to very serious) and their understanding of diabetes management (How well do you understand how to manage your diabetes? 1 to 5, not at all to very well). Quality of health care services. The Modified Health Care Climate Questionnaire (HCCQ) was used to assess respondents’ perceived support from their provider (sum of 5 items ranging 1 to 5, higher5strongly agree).42 Items for provider support include the following questions: You feel understood by your doctor, You feel a lot of trust in your doctor, Your doctor tries to understand how you see things before suggesting new ways to do things, You get all the support that you need from your doctor and Your doctor answers your questions fully and carefully. The modified HCCQ reports a Cronbach’s alpha of .80 and is correlated .91 with the original 15-item HCCQ. A factor analysis of the five items in various studies yielded a one-factor solution (eigenvalue53.0 with all factors loading above .74). In the present study, the Cronbach’s alpha for the five items was .95. Health care satisfaction was measured by a single item: How satisfied are you with the quality of health care you have received during the last 12 months? (mean, range 1 to 4, 45very dissatisfied).

Spencer, Kieffer, Sinco, et al.

93

Problem Areas in Diabetes Scale (PAID). The Problem Areas in Diabetes Scale (PAID) is the best-validated assessment tool used to measure diabetes-related emotional distress.16,23 The PAID is a self-report questionnaire that consists of 20 statements measuring emotional distress in managing and dealing with diabetes and its complications. Representative items include feeling scared when you think about living with diabetes; feelings of deprivation regarding food and meals; feeling depressed when you think about living with diabetes; feeling overwhelmed by your diabetes; feeling burned out by the constant effort to manage your diabetes. Each item can be rated on a 5-point Likert scale ranging from 1 (not a problem) to 6 (a serious problem). A previous study of the PAID reported a Cronbach’s alpha coefficient of .95.16 Also, concurrent validity has been found between the PAID and generalized distress, fear of hypoglycemia, disordered eating, and adherence to self care behaviors with significant correlations. A Cronbach’s alpha coefficient of .94 was found for our sample (.92 for Hispanics and .93 for African Americans). Analysis. SAS® 9.143 was used for all analyses. To facilitate interpretation, PAID scores were converted to a 0–100 scale, with higher scores indicating greater emotional distress.16 The PAID values were then log transformed to reduce skewness. Our analyses of PAID covariates were conducted in three stages: (1) examine differences in study variables including the PAID by race/ethnicity; (2) identify study variables that showed a significant (p..05) bivariate correlation with the PAID separately by race/ethnicity; and (3) examine within-group multivariable associations between the PAID and those study variables found to be significant in the bivariate analyses using ordinary least squares regression models. In the first and second stages, if the variable was binary, its effect on the PAID was assessed by comparing the PAID means at the two levels of the variable. The PAID means were then compared with the student t test for significance. If the variable was continuous or on a numerical scale with at least four levels, the Spearman correlation coefficient was used to measure correlation between the PAID and the variable. Spearman’s correlation coefficient was chosen because it measures the correlation between variables, even if it is non-linear.44 Finally, multiple hierarchical regression was used to test the association between the PAID and the study variables that were significantly correlated with the PAID in the bivariate analyses. The exception was the study’s predefined control variables noted above and glycemic control, which were included in all models. We use the biopsychosocial model to guide the order in which we entered our variables into the regression equations. We enter the control variables in the first step. We subsequently enter variables from proximal to distal (biological, psychological, and social) in following steps. Since glycemic control, our biological variable, is not statistically correlated with the PAID in any of our multivariate models, it is presented with our control variables in the first step. Separate regression models were computed for African Americans and Hispanics. Only variables that were significantly associated with the PAID at p,.05 in the separate bivariate analyses by race/ethnicity were included in the regression models. Study variables were entered into the hierarchical regression equation in four steps. In the first step, only the control variables and

94

Diabetes-specific emotional distress

glycemic control were entered into the model. The two traditional stress variables were entered in step 2, followed by other psychosocial variables in step 3, and the quality of health care variables in step 4.

Results Participant characteristics. Table 1 presents the descriptive statistics for the study variables. African Americans made up 60.5% of the total sample (n5109) and Hispanics (n571) made up the rest; women represented approximately 75% of the sample for both African Americans and Hispanics. Among Hispanics, 75% had HbA1c values greater than 7.0. About 65% of African Americans had HbA1c values greater than 7.0. In general, racial and ethnic differences were observed across participant characteristics. African Americans were significantly more likely than Hispanics to be high school graduates (71.6% compared with 19.7%), live alone (76.1% compared with 56.3%), have diabetes complications (95.4% compared with 75.7%), and report more understanding of diabetes management (3.54 compared with 3.13). The mean score on diabetes-related emotional distress as measured by the PAID was 23.9 for the entire sample (range 0–100). About 27% reported a serious problem on at least one of the PAID items, and 50% reported that at least one problem was somewhat serious. Hispanics reported mean PAID scores that were more than twice the mean reported for African Americans (36.75 compared to 15.59). In bivariate analyses between the study variables and the PAID (see Table 2), associations were found between emotional distress and many of the study variables for both African Americans and Hispanics. Among the socio-demographic variables for African Americans, only living with others was positively correlated with higher PAID scores. Among Hispanics, being female and being younger than age 55 were correlated with more emotional distress. Among the biological factors, an HbA1c value greater than 7 was associated with higher levels of emotional distress among Hispanics, but not African Americans. Experiencing more chronic illnesses was positively correlated with emotional distress for African Americans only. In general, psychosocial variables, including perceived stress, daily hassles and perceived seriousness of one’s diabetes were significantly correlated with emotional distress for both racial/ethnic groups, while a better understanding of diabetes self management was associated with less emotional distress. Less provider support and more health care dissatisfaction were significantly associated with higher PAID scores for African Americans, but not for Hispanics. Multivariate analyses. Regression models for African Americans and Hispanics are reported in Tables 3 and 4, respectively. Among African Americans, household status was associated with the PAID; living alone was associated with lower emotional distress scores compared with living with one or more people (Table 3). All four psychosocial variables examined were associated with the PAID, and both perceived stress and daily hassles were associated with increased emotional distress. Moreover, respondents who reported their case of diabetes as being more serious or having less of an understanding of how to manage their diabetes also reported increased

Spencer, Kieffer, Sinco, et al.

95

Table 1. Sociodemographic Characteristics and Descriptive Statistics, Hispanic and African American REACH-Detroit Participants Hispanic (n571)

African American (n5109)

Demographics Female (%) Age in years, mean (s.e.)  Age 55 years High school graduate (%) Living alone (%) Southwest community (%) Spanish primary language (%) PAID score,b, c mean (s.e.)

74.6 52.41 (1.74) 47.9 19.7 56.3 100.0 81.7 36.75 (2.40)

74.3 59.38 (1.22) 64.2 71.6 76.1 26.6 NA 15.59 (1.78)

.9597 .0009 .0301 ,.0001 .0052 NA NA ,.0001

Biological Hemoglobin A1c2, mean (s.e.) Hemoglobin A1c .7.0 (%) 1 Diabetes complications (%) 1 Other chronic illnesses (%)

8.64 (.28) 75.0 75.7 88.7

8.23 (.24) 64.9 95.4 95.3

.1882 .1694 ,.0001 .0978

1.93 (.07) 1.63 (.05)

1.96 (.05) 1.56 (.05)

.7803 .1924

3.56 (.10)

3.47 (.07)

.3108

3.13 (.11)

3.54 (.10)

.0067

4.03 (.10) 63.4

3.98 (.08) 59.6

.6549 .6142

Psycho-social Perceived stress, mean (s.e.) Daily hassles, mean (s.e.) Perceived seriousness of   diabetes case, mean (s.e.) Understanding diabetes   management, mean (s.e.) Quality of healthcare services Support from physician, mean (s.e.) Very satisfied with health care (%)

Note: Values in parentheses denote standard errors. a If categorical variable, Chi-Square test. If numerical variable, Student T-Test. b T-Test on log transform. c PAID5Problem Areas In Diabetes score.

p valuea

96

Diabetes-specific emotional distress

Table 2. Bivariate Associations Between Participant Characteristics and Diabetes Specific Emotional Distress,a,b Hispanic and African American REACH-Detroit Participants

Hispanic

African American

Demographics, mean (s.e.) Female Male # Age 54 years  Age 55 years , High school graduate High school graduate Lives with one or more people Lives alone English primary language Spanish primary language Eastside community Southwest community

40.03 (2.88)* 27.26 (3.46)* 43.10 (3.20)** 29.62 (3.22)** 37.69 (2.70) 32.98 (5.31) 37.63 (3.66) 36.08 (3.22) 34.50 (5.49) 37.26 (2.69) NA NA

17.97 (2.27) 8.70 (1.74) 21.04 (3.53) 12.55 (1.88) 13.74 (3.53) 16.32 (2.06) 27.85 (4.98)*** 11.75 (1.54)*** NA NA 16.10 (2.24) 14.18 (2.62)

Biological, mean (s.e.) Hemoglobin A1c #7.0 Hemoglobin A1c .7.0 No diabetes complications 1 or more diabetes complications No other chronic illnesses 1 or more other chronic illnesses

26.77 (4.56)* 40.43 (2.77)* 32.14 (3.17) 37.83 (3.02) 30.51 (4.82) 37.55 (2.63)

11.32 (2.48) 16.83 (2.23) 9.75 (4.53) 15.85 (1.87) 13.00 (4.79) 15.81 (1.89)

Psycho-social, correlation Perceived stress Daily hassles Perceived seriousness of diabetes case Understanding diabetes management

.2665* .3795** .2372* 2.1321

.2766** .4535*** .2407* 2.3378***

Quality of healthcare services Support from physician, correlation , Very satisfied with health care, mean (s.e.) Very satisfied with health care, mean (s.e.)

2.1874 43.30 (3.83) 33.11 (2.96)

2.2748** 21.22 (3.40)** 11.78 (1.78)**

Spearman correlation coefficients displayed for numerical variables and the PAID (Problem Areas In Diabetes) score; PAID means and standard errors displayed at each level of the categorical variables. b p values from Student T-Test on log transform of PAID. *p#.05, **p#.01, ***p#.001. a

Spencer, Kieffer, Sinco, et al.

97

emotional distress. Finally, respondents who perceived more support from their physicians had significantly less emotional distress. All of the variables included in the successive models remained significant throughout, with the exception of perceived stress, which was not significant in Model 2, but became significant when the psychosocial variables were entered in Model 3. The final model explained about 50% of the variance in emotional distress for African Americans. Among Hispanics, only demographic variables were associated with the PAID in the final model (Table 4). Being 55 years and older or having graduated from high school was associated with less emotional distress. Sex was significant in Models 1 through 3, but was not significant in Model 4 when health care quality was considered. While poor glycemic control was marginally associated with more emotional distress in Model 1, it was no longer significant once the stress variables were entered in Model 2. Similarly, daily hassles was marginally associated with emotional distress, but was non-significant when quality of health care variables were included in the final model. The final model explained about 39% of the variance in emotional distress for Hispanics.

Discussion The results of the analyses highlight important differences in correlates of emotional distress among African Americans and Hispanics. Significant associations between diabetes-related emotional distress and psychosocial variables, including daily hassles, perceptions of seriousness of one’s case of diabetes, and understanding of self management were found for African Americans. Clearly, the literature acknowledges the role of life stressors and their relationship to psychological distress, and acknowledges it specifically in the lives of individuals living with chronic illnesses such as type 2 diabetes.21–24 These psychosocial variables had a greater impact on emotional distress than biological variables; this is evident in the analyses when all variables were held constant. Among African Americans, quality of health care services also was significantly associated with emotional distress. The negative effect of psychosocial stress and perceived poor quality health care on diabetes-related emotional distress among African Americans should not be surprising and has important implications for service use. Disparities of service use for both chronic diseases and mental health problems among African Americans is well documented in the recent health care literature.2,4,45 In the case of routine, noninvasive processes of care, such as those for diabetes, the interactions between health care providers and patients may significantly contribute to racial and ethnic variations in health care.4,46 Physicians often provide less information to minority patients, and Black patients rate their physician visits as less participatory than do non-Hispanic white patients.47–50 Many of the same factors that pose challenges to type 2 diabetes preventive services are similar to those challenges related to the use of psychological services, particularly among vulnerable populations. African Americans and Hispanics use fewer mental health services and receive poorer mental health care than non-Hispanic whites, even when health care access, income, and symptom severity are controlled for.33–35,51–52 Similarly, the same barriers posed by

98

Diabetes-specific emotional distress

Table 3. Multivariate Regression Coefficients for Association Between Participant Characteristics and Diabetes Specific Emotional Distress, REACH-Detroit Hispanic Participantsa

Model 1b

Model 2b

Model 3b

Model 4b

.3072

.3679

.3738

.3917

2.53** .16 2.56*** 2.45* 2.02

2.44* .24 2.48** 2.49* .04

2.41* .24 2.46** 2.50* .05

2.39 .26 2.43* 2.53* .02

.30

.23

.23

.24

Psycho-social Perceived stress .17 Daily hassles .32 Perceived seriousness of diabetes case

.16 .34 .07

.16 .32 .05

R-Squaredc Demographics Maled Live alone Age 55 years or older High school graduate Spanish primary languaged Biological Hemoglobin A1c .7.0d

Quality of healthcare services Very satisfied with health cared

2.21

Regression on log transform of the PAID, Problem Areas In Diabetes score. Model 1: Demographics and Hemoglobin A1c. Model 2: Demographics, Hemoglobin A1c, Stress. Model 3: Demographics, Hemoglobin A1c, Stress, Understanding of Diabetes. Model 4: Demographics, Hemoglobin A1c, Stress, Understanding of Diabetes, Healthcare. c Least Squares Regression Coefficient Squared. d Reference Groups: Male gender referenced to female gender. Spanish primary language referenced to English. Hemoglobin A1c .7.0 referenced to Hemoglobin A1c #7.0. Very satisfied with healthcare, referenced to less than very satisfied. *p#.05, **p#.01, ***p#.001. a

b

low socioeconomic status to healthy eating and physical activity also may contribute to both type 2 diabetes and emotional distress.53 Among Hispanics, demographic factors (e.g., low educational level and young age) were associated with diabetes-related emotional distress, suggesting that structural and personal factors may outweigh process variables observed in this study. Structural barriers, such as low access to education that affects subsequent literacy and English proficiency, can lead to distress. Health literacy is an important con-

Spencer, Kieffer, Sinco, et al.

99

Table 4. Multivariate Regression Coefficients for Association Between Participant Characteristics and Diabetes Specific Emotional Distress, REACH-Detroit African American Participantsa R-Squaredc

Model 1b

Model 2b

Model 3b

Model 4b

.1219

.2388

.4410

.5042

Demographics Maled Live alone Age 55 years or older High school graduate Southwest communityd

.37 21.01** 2.10 2.01 2.33

Biological Hemoglobin A1c .7.0d

.42

2.31 2.61 .06 .01 2.23 .37

Psycho-social Perceived stress .43 .57* Daily hassles 1.10*** .92** Perceived seriousness of diabetes case Understanding of diabetes management

.08 2.71* .15 .03 2.08

.29 2.78* .20 .09 .19

.27

.17

.56* .33** 2.63***

0.93** 0.34** 2.61***

Quality of healthcare services Support from physician Very satisfied with health cared

2.43* .13

Regression on log transform of the PAID, Problem Areas In Diabetes score. Model 1: Demographics and Hemoglobin A1c. Model 2: Demographics, Hemoglobin A1c, Stress. Model 3: Demographics, Hemoglobin A1c, Stress, Understanding of Diabetes. Model 4: Demographics, Hemoglobin A1c, Stress, Understanding of Diabetes, Healthcare. c Least Squares Regression Coefficient Squared. d Reference Groups: Male gender referenced to female gender. Southwest Detroit, referenced to Eastside Detroit. Hemoglobin A1c .7.0 referenced to Hemoglobin A1c #7.0. Very satisfied with healthcare, referenced to less than very satisfied. *p#.05, **p#.01, ***p#.001. a

b

sideration, even when individuals are English proficient. For example, Heisler and colleagues found that formal education was associated with increases in patients’ HbA1c knowledge, which in turn was associated with accurately assessing diabetes control and better understanding of diabetes care.54 It is not surprising that individuals who develop diabetes at younger ages have higher levels of emotional distress

100

Diabetes-specific emotional distress

than their older counterparts as they consider their mortality and management of a chronic illness in mid-life. There are significantly more Hispanics who are under 55 in our study than African Americans, which may explain why this variable was significant for Hispanics only. Although health care variables were not significantly associated with emotional distress for Hispanics, the study’s context provides a reasonable explanation. Hispanics in REACH Detroit were nearly all recruited from one highly regarded health center in Southwest Detroit that provides comprehensive, bilingual services to people regardless of income or citizenship. Hispanics in the study overall reported greater satisfaction with health care. Finally, African Americans in our sample reported notably low levels of emotional distress (mean score of 15.59), particularly when compared with Hispanics (mean score of 36.75). Other studies of the PAID have found mean scores ranging from 33.4 to 49.3 for African American women with type 2 diabetes.55–56 Low PAID scores are evident despite high levels of poor glycemic control. Unlike Hispanics, the PAID is not correlated with poor glycemic control at the bivariate level for African Americans. Low PAID scores among African Americans compared with Hispanics can be partially explained by their higher educational levels, higher percentage of individuals over age 55, their higher level of understanding of diabetes management, and their slightly lower HbA1c scores. Additionally, the low levels of diabetes-specific emotional distress could suggest that these individuals are experiencing multiple other stressors and priorities that are more distressing than their diabetes. However, at the same time, these same environmental stressors also contribute to poor diabetes management and poor glycemic control. Further investigation and replication of these results are necessary before firm conclusions can be drawn from these findings. There are several limitations to this study. First, a small sample size may have provided inadequate statistical power to detect some significant differences. Second, the study uses a convenience sample of individuals who have a regular source of care, and therefore the findings are not generalizable to all African Americans and Hispanics with type 2 diabetes in the targeted ZIP codes. More research with large, representative samples of diverse populations is needed. Research with diverse Hispanic populations is especially needed, as there may be other sociocultural variables influencing emotional distress that were not measured in this study. Specifically, we note that our sample of Hispanics were largely new immigrants. Our sample and its measures precludes our ability to conduct refined analyses on generational effects, nativity, or acculturative effects on diabetes-related emotional distress, although we acknowledge the possibility that cultural adaptation may play a significant role in health literacy and health care barriers. Third, many of our measures are self-reported. However, we used well-validated self-report instruments that have been found to be valid and reliable measures in a number of health surveys. The study is further strengthened by the use of objective lab values obtained from physician’s offices. Another limitation related to instrumentation is the use of measures that were not originally developed in Spanish. Although we took great steps to assure the

Spencer, Kieffer, Sinco, et al.

101

cultural and linguistic appropriateness of our instruments, it should be noted that we did not address issues of cultural equivalence, which requires an investigative process of testing participants to determine whether the questions were understood as intended.57–58 Disregarding the influence of culture can lead to systematic bias in measuring distress among different groups.59 While the issues of cultural equivalence is important in all transcultural research, and therefore is important for African Americans as well, it is particularly important for Spanish-speaking Hispanics given the wide variations in vocabulary across regions and countries that can lead to subtle differences in meaning. Unfortunately, evidence of cultural equivalence for the measures that we use in this study is not available in the current literature, as such studies are still quite rare. Future research must address the cultural equivalence of measures, such as the PAID and other stress measures, if we are to conduct sound transcultural research. Last, while the study includes traditional measures of stress, it does not include a measure of mental illness such as depression or anxiety. It seems likely that individuals with mental illness would report higher rates of diabetes-related emotional distress, since many of the items in the PAID scale reflect a sense of depression or anxiety around diabetes and its management.

Conclusion In conclusion, researchers, providers, and policymakers must become aware of the adverse psychological impact that diabetes carries and actively seek to intervene on the factors that may exacerbate diabetes-related emotional distress among African Americans and Hispanics. At the individual level, factors beyond the biomedical realm should be considered, such as negative life events and the stressors and hassles of everyday life. Knowledge of how to best manage one’s diabetes is also important. This information should be available in languages other than English in order to reach immigrant and non-English speaking racial/ethnic minorities. Finally, quality health care is imperative. This study builds on the health disparities literature by demonstrating the significant impact of health care providers’ communication and support among African Americans on the psychological well being associated with diabetes. Interventions should recognize the personal, family, and community contexts in which African Americans and Hispanics live with type 2 diabetes. The influence of these factors on serious emotional distress may, in turn, have an impact on managing and coping with this, and other, chronic diseases. Greater attention to treating illness by addressing its psychological side through support and coping strategies, and not just the disease, has important implications for reducing health disparities among disadvantaged racial and ethnic groups.

Acknowledgments The authors acknowledge the generous support of the REACH Detroit Partnership Steering Committee and the Detroit-Community Academic Urban Research Center.

102

Diabetes-specific emotional distress

The authors also wish to thank Dean Paula Allen-Meares (University of Michigan School of Social Work), the University of Michigan Office of the Vice President for Research and Office of the Provost, Ms. Sara McNary and Ms. Nkenge Jack (CDC), Michael Anderson (REACH Detroit FHA Supervisor), the REACH Detroit Family Health Advocates, our REACH Detroit partners, the UM REACH Detroit staff, and the families that participated in the REACH Detroit Family Intervention. This work was supported by a grant from the Centers for Disease Control and Prevention (CDC), Racial and Ethnic Approaches to Community Health (REACH) 2010 grant #U50CCU522189.

Notes   1. Vinicor F. Is diabetes a public-health disorder? Diabetes Care. 1994 Jun;17 Suppl 1:22–7.   2. U.S. Department of Health and Human Services (DHHS). Healthy People 2010: Understanding and Improving Health. 2nd Ed. Washington, DC: U.S. Government Printing Office, 2000 Nov.   3. Flegal KM, Ezzati TM, Harris MI, et al. Prevalence of diabetes in Mexican Americans, Cubans and Puerto Ricans from the Hispanic Health and Nutrition Examination Survey, 1982–1984. Diabetes Care. 1991 Jul;14(7):628–38.   4. Smedley BD, Stith AY, Nelson AR, et al., eds. Institute of Medicine, Committee on Understanding and Eliminating Racial and Ethnic Disparities in Health Care. Unequal treatment: confronting racial and ethnic disparities in healthcare. Washington, DC: The National Academies Press, 2003.   5. Harris MI, Eastman RC, Cowie CC, et al. Racial and ethnic differences in glycemic control of adults with type 2 diabetes. Diabetes Care. 1999 Mar;22(3):403–8.   6. Karter AJ, Ferrara A, Liu JY, et al. Ethnic disparities in diabetic complications in an insured population. JAMA. 2002 May;287(19):2519–27.   7. Cowie CC, Port FK, Wolfe RA, et al. Disparities in incidence of diabetic end-stage renal disease according to race and type of diabetes. N Engl J Med. 1989;321:1074–9.   8. Pugh JA, Stern MP, Haffner SM, et al. Excess incidence of treatment of end-stage renal disease in Mexican Americans. Am J Epidemiol. 1988 Jan;127(1):135–44.   9. Centers for Disease Control and Prevention (CDC). National diabetes fact sheet: general information and national estimates on diabetes in the United States, 2003. Atlanta, GA: CDC, 2004. 10. Rubin RJ, Altman WM, Mendelson DN. Health care expenditures for people with diabetes mellitus, 1992. J Clin Endocrinol Metab. 1994 Apr;78(4):809A–809F. 11. Saaddine JB, Engelgau MM, Beckles GL, et al. A diabetes report card for the United States: quality of care in the 1990s. Ann Intern Med. 2002 Apr;136(8):565–74. 12. Cowie CC, Harris MI. Ambulatory medical care for non-Hispanic whites, AfricanAmericans, and Mexican-Americans with NIDDM in the U.S. Diabetes Care. 1997 Feb;20(2):142–7. 13. Chin MH, Zhang JX, Merrell K. Diabetes in the African-American Medicare population. Morbidity, quality of care, and resource utilization. Diabetes Care. 1998 Jul; 21(7):1090–5. 14. Heisler M, Smith DM, Hayward RA, et al. Racial disparities in diabetes care processes, outcomes, and treatment intensity. Med Care. 2003 Nov;41(11):1221–32.

Spencer, Kieffer, Sinco, et al. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.

25. 26.

27.

28. 29.

30.

31. 32.

33.

34.

103

Anderson RJ, Freedland KE, Clouse RE, et al. The prevalence of comorbid depression in adults with diabetes: a meta-analysis. Diabetes Care. 2001 Jun; 24(6):1069–78. Polonsky WH, Anderson BJ, Lohrer PA, et al. Assessment of diabetes-related distress. Diabetes Care. 1995 Jun;18:(6)754–60. Gary TL, Crum RM, Cooper-Patrick L, et al. Depressive symptoms and metabolic control in African Americans with type 2 diabetes. Diabetes Care. 2000 Jan;23(1):23–9. American Diabetes Association. Diabetes 1996 vital statistics. Alexandria, VA: American Diabetes Association, 1996. Roy A, Roy M. Depressive symptoms in African-Americans type 1 diabetics. Depress Anxiety. 2001;13(1):28–31. Engel GL. The need for a new medical model: a challenge for biomedicine. Science. 1977 Apr 8;196(4286):129–36. Jacobson, AM. The psychological care of patients with insulin-dependent diabetes mellitus. N Engl J Med. 1996 May 9;334(19):1249–53. Rubin RR, Peyrot M. Psychosocial problems and interventions in diabetes. A review of the literature. Diabetes Care. 1992 Nov;15(11):1640–57. Polonsky WH, Welch G. Listening to our patients’ concerns: understanding and addressing diabetes-specific emotional distress. Diabetes Spectrum. 1996;9:8–11. Black SA, Ray LA, Markides KS. The prevalence and health burden of self-reported diabetes in older Mexican Americans: findings from the Hispanic established populations for epidemiologic studies of the elderly. Am J Public Health. 1999 Apr; 89(4):546–52. de Groot M, Lustman PJ. Depression among African-Americans with diabetes: a dearth of studies. Diabetes Care. 2001 Feb;24:(2)407–8. Gary TL, Crum RM, Cooper-Patrick L, et al. Depressive symptoms and metabolic control in African-Americans with type 2 diabetes. Diabetes Care. 2000 Jan; 23(1):23–9. de Groot M, Auslander W, Williams JH, et al. Depression and poverty among African American women at risk for type 2 diabetes. Ann Behav Med. 2003 Summer; 25(3):172–81. Fisher L, Chesla CA, Mullan JT, et al. Contributors to depression in Latino and EuropeanAmerican patients with type 2 diabetes. Diabetes Care. 2001 Oct;24(10):1751–57. Bazargan M, Bazargan-Hejazi S, Baker RS. Treatment of self-reported depression among Hispanics and African Americans. J Health Care Poor Underserved. 2005 May;16(2):328–44. Alvidrez J, Havassy BE. Racial distribution of dual-diagnosis clients in public sector mental health and drug treatment settings. J Health Care Poor Underserved. 2005 Feb;16(1):53–62. Sambamoorthi U, Olfson M, Wei W, et al. Diabetes and depression care among Medicaid beneficiaries. J Health Care Poor Underserved. 2006 Feb;17(1):141–61. de Groot M, Welch G, Buckland GT 3rd, et al. Cultural orientation and diabetes self-care in low-income African Americans with type 2 diabetes. Ethn Dis. 2003 Winter;13(1):6–14. Alegria M, Canino G, Rios R, et al. Inequalities in use of specialty mental health services among Latinos, African Americans, and non-Latino whites. Psychiatr Serv. 2002 Dec;53:1547–55. Melfi CM, Croghan TW, Hanna MP. Access to treatment for depression in a Medicaid population. J Health Care Poor Underserved. 1999 May;10(2):201–15.

104

35. 36.

37.

38.

39. 40.

41. 42. 43. 44. 45.

46.

47.

48. 49.

50.

51.

52.

Diabetes-specific emotional distress Sue S. Psychotherapeutic services for ethnic minorities. Two decades of research findings. Am Psychol. 1988 Apr;43(4):301–8. Two Feathers J, Kieffer EC, Palmisano G, et al. Racial and Ethnic Approaches to Community Health (REACH) Detroit partnership: improving diabetes-related outcomes among African American and Latino adults. Am J Public Health. 2005 Sep;95(9):1552–60. Kieffer EC, Willis SK, Odoms-Young AM, et al. Reducing disparities in diabetes among African-American and Latino residents of Detroit: the essential role of community planning focus groups. Ethn Dis. 2004 Summer;14(3 Suppl 1):S27–37. Centers for Disease Control and Prevention (CDC). Behavioral Risk Factor Surveillance System Survey Data. Prevalence Data: Nationwide—2000 Women’s Health, Have you ever had a mammogram? Atlanta, GA: CDC, 2005. Available at http://apps.nccd .cdc.gov/brfss/race.asp?cat5WH&yr52000&qkey5311&state5US. Cohen S, Kamarck T, Mermelstein R. A global measure of perceived stress. J Health Soc Behav. 1983 Dec;24(4):385–96. Cohen S, Williamson GM. Perceived stress in a probability sample of the United States. In: Spacapan S, Oskamp S, eds. The social psychology of health. Newbury Park, CA: Sage, 1988. Delongis A, Coyne JC, Dakof G, et al. Relationships of daily hassles, uplifts, and major life events to health status. Health Psychol. 1982 Spring;1(2):119–36. Williams GC, Freedman ZR, Deci EL. Supporting autonomy to motivate patients with diabetes for glucose control. Diabetes Care. 1998 Oct;21(10):1644–51. SAS Institute. SAS®: Statistical Analysis Software, Release 9.1. Cary, NC: SAS Institute, 2002. Conover WJ. Practical nonparametric statistics. 2nd Ed. New York: John Wiley & Sons, 1980; 252–7. U.S. Public Health Service (PHS), Office of the Surgeon General. Mental health: a report of the Surgeon General: executive summary. Rockville, MD: DHHS, PHS, 1999. Johnson RL, Roter D, Powe NR, et al. Patient race/ethnicity and quality of patientphysician communication during medical visits. Am J Public Health. 2004 Dec; 94(12):2084–90. DiMatteo MR, Sherbourne CD, Hays RD, et al. Physicians’ characteristics influence patients’ adherence to medical treatment: results from the Medical Outcomes Study. Health Psychol. 1993 Mar;12(2):93–102. Cooper-Patrick L, Gallo JJ, Gonzales JJ, et al. Race, gender, and partnership in the patient-physician relationship. JAMA. 1999 Aug 11;282(6):583–9. Kaplan SH, Gandek B, Greenfield S, et al. Patient and visit characteristics related to physicians’ participatory decision-making style. Results from the Medical Outcomes Study. Med Care. 1995 Dec;33(12):1176–87. Cooper LA, Roter DL, Johnson RL, et al. Patient-centered communication, ratings of care, and concordance of patient and physician race. Ann Intern Med. 2003 Dec; 139(11):907–15. Padgett DK, Patrick C, Burns BJ, et al. Ethnicity and the use of outpatient mental health services in a national insured population. Am J Public Health. 1994 Feb;84(2): 222–6. Hu TW, Snowden LR, Jewell JM, et al. Ethnic populations in public mental health: service choice and level of use. Am J Public Health. 1991 Nov;81(11):429–34.

Spencer, Kieffer, Sinco, et al. 53.

54.

55.

56. 57.

58.

59.

105

Everson SA, Maty SC, Lynch JW, et al. Epidemiologic evidence for the relation between socioeconomic status and depression, obesity and diabetes. J Psychosom Res. 2002 Oct;53(4):891–5. Heisler M, Piette JD, Spencer M, et al. The relationship between knowledge of recent HbA1c values and diabetes care understanding and self-management. Diabetes Care. 2005 Apr;28(4):816–22. D’Eramo-Melkus G, Spollett G, Jefferson V, et al. A culturally competent intervention of education and care for black women with type 2 diabetes. Appl Nurs Res. 2004 Feb;17(1):10–20. Newlin K, Melkus G, Chyun D, et al. The relationship of spirituality and health outcomes in black women with type 2 diabetes. Ethn Dis. 2003 Winter;13(1):61–8. Johnson TP, O’Rourke D, Chavez N, et al. Social cognition and responses to health survey questions among minority populations: preliminary evidence. In: Proceedings of the 1993 Public Health Conference on Records and Statistics. Toward the Year 2000. Refining the Measures, Washington (DC), July 19–21, 1993:357–62. (DHHS Pub. No. PHS 94-124.) Hyattsville, MD: National Center for Health Statistics, 1994. Rogler LH, Cortes DE, Malgady RG. The mental health relevance of idioms of distress. Anger and perceptions of injustice among New York Puerto Ricans. J Nerv Ment Dis. 1994 Jun;182(6):327–30. Williams D, Harris-Reid M. Race and mental health: emerging patterns and promising approaches. In: Horwitz AV, Scheid TL, eds. A handbook for the study of mental health: social contexts, theories, and systems. New York: Cambridge University Press,1999; 295–314.