PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of jgimedspringer.comThis journalToc AlertsSubmit OnlineOpen Choice
 
J Gen Intern Med. 2008 December; 23(12): 1992–1999.
Published online 2008 October 15. doi:  10.1007/s11606-008-0814-7
PMCID: PMC2596525

When Is Social Support Important? The Association of Family Support and Professional Support with Specific Diabetes Self-management Behaviors

Ann-Marie Rosland, MD, MS,corresponding author1 Edith Kieffer, PhD, MPH,2 Barbara Israel, DrPH, MPH,3 Marvis Cofield,5,6 Gloria Palmisano, MA,6 Brandy Sinco, MS,2 Michael Spencer, PhD, MSW,2 and Michele Heisler, MD, MPA1,4

Abstract

BACKGROUND

Social support is associated with better diabetes self-management behavior (SMB), yet interventions to increase family and friend support (FF support) have had inconsistent effects on SMB.

OBJECTIVE

To test whether FF support differentially affects specific SMBs and compare the influence of support from health professionals and psychological factors on specific SMBs to that of FF support.

DESIGN

Cross-sectional survey of people with diabetes recruited for a self-management intervention

PARTICIPANTS AND SETTING

One hundred sixty-four African-American and Latino adults with diabetes living in inner-city Detroit

MEASUREMENTS AND MAIN RESULTS

For every unit increase in FF support for glucose monitoring, the adjusted odds ratio (AOR) of completing testing as recommended was 1.77 (95% CI 1.21–2.58). FF support was not associated with four other SMBs (taking medicines, following a meal plan, physical activity, checking feet). Support from non-physician health professionals was associated with checking feet [AOR 1.72 (1.07–2.78)] and meal plan adherence [AOR = 1.61 (1.11–2.34)]. Diabetes self-efficacy was associated with testing sugar, meal plan adherence, and checking feet. Additional analyses suggested that self-efficacy was mediating the effect of FF support on diet and checking feet, but not the FF support effect on glucose monitoring.

CONCLUSIONS

The association between FF support and SMB performance was stronger for glucose monitoring than for other SMBs. Professional support and diabetes self-efficacy were each independently associated with performance of different SMBs. SMB interventions may need to differentially emphasize FF support, self-efficacy, or professional support depending on the SMB targeted for improvement.

KEY WORDS: diabetes, self-management, social support, community based participatory research, Hispanic Americans, African Americans

INTRODUCTION

Patients’ diabetes self-management is a key determinant of diabetes outcomes. Self-management tasks, such as glucose testing or insulin injection, often take place in social settings and can alter family and social routines. Patients’ family and friends can provide support in overcoming these social barriers and in executing complex self-management behaviors, but can also pose barriers to self-management behaviors. Practical and emotional assistance received in family and friend relationships, or social support, has a positive influence on global measures of diabetes self-management behavior (SMB).18 However, to date interventions seeking to improve SMB by increasing social support from family and friends (FF support) have had mixed results.9,10 In light of increasing calls to involve family in chronic disease care, it is critical to more closely examine the effects of FF support and possible reasons for intervention success or failure.

One hypothesis for inconsistent social support intervention effects is that FF support affects each SMB differently. For example, family support may be more important for SMBs that affect family routines, such as meal planning, or those that take place in social settings, such as glucose testing. Performance of some SMBs may depend on factors external to the family environment, such as neighborhood environment and community resources for healthful eating and physical activity.1115 Patients may receive more support for some SMBs from health professionals, making family support less important for those tasks.16 Moreover, according to social cognitive theory and social support theory, patient psychological factors may mediate the effect of social support on SMB, and it is possible that these mediators vary in their ability to affect SMB. For instance, social support may improve SMB by first leading to increased self-efficacy3,1719 or decreased depressive symptoms,2023 and self-management tasks that are affected more by these factors24 may be more affected by social support.

Despite discussion of these possible reasons for support to affect self-management tasks differently,4,9 previous studies have not directly compared family support, professional support, and psychological influences across SMB domains. Accordingly, we examined the relationship between perceived family and friend support for each of five specific SMBs (testing sugar, taking diabetes medicines, following a diabetes meal plan, physical activity, and checking feet) and the performance of that behavior, using data from a survey of African-American and Latino adults with diabetes. We then compared the effects of adjusting for professional support, self-efficacy, and depression in each FF support–SMB relationship.

RESEARCH DESIGN AND METHODS

Study Population

We analyzed data from surveys of 164 African-American and Latino adults with type 2 diabetes. The surveys were conducted by the REACH Detroit Partnership, a CDC-funded collaboration of community groups, academic partners, and local health-care providers working to improve diabetes outcomes among African Americans and Latinos in the Eastside and Southwest Detroit communities.25,26 The surveys were completed as part of a baseline assessment of participants enrolled in a community-health worker-led diabetes self-management intervention consisting of group self-management training sessions and individual support. Participants were age 18 or older, had physician-diagnosed type 2 diabetes, but no serious diabetes complications, such as blindness, amputation, or kidney failure, and received medical care through a local health system or a federally qualified community-health center. Participants were recruited from health-system diabetes registries. The average annual household income in the three zipcodes with the largest number of participants ranged from $24,256 to $25,546 (2000 US Census data).

Of 578 eligible participants, 183 agreed to participate, and 164 completed the full baseline assessment. Latinos were more likely to participate than African-Americans (49.7% vs 26.8%), and the mean age of participants (53.0) was significantly lower than the age for non-participants (58.9). Surveys were conducted in the participant’s home in their preferred language (English or Spanish) by a trained African-American or Latino interviewer.

Using a community-based participatory research approach,26,27 the REACH Detroit Partnership steering committee was actively involved in the development of the study questions, survey instrument, and methods for data collection, interpretation, and dissemination. The survey protocol received Institutional Review Board approval at the local participating health systems and the University of Michigan.

Study Measures

Sociodemographics were determined through self-report with survey items asking age, gender, education level, and race/ethnicity. Health Status was assessed with a question about activity limitations from the Behavioral Risk Factor Surveillance Survey (BRFSS)28 and self-reported diabetes medication regimen (see Appendix for question wording). Participants’ A1C levels were obtained from medical chart values closest to the time of the survey. Social support from family and friends (FF support) for each of five SMB tasks was assessed with items from the Diabetes Care Profile,29,30 and physician and other health-professional support for diabetes care was assessed with single questions (see Table 1 for question wording). Diabetes SMBs were assessed with items from the Survey of Diabetes Self-Care Activities (SDSCA).31,32 Physical activity was assessed using questions from the BRFSS.33 Diabetes care self-efficacy and depressive symptoms were measured with validated survey scales [the Perceived Competence for Diabetes Scale34 and the Patient Health Questionnaire (PHQ-9),35 respectively].

Table 1
Social Support Survey Questions

Statistical Analysis

Support for each SMB task was separately regressed onto the performance of the corresponding task in bivariate and multivariate models. Because SMB outcomes were collected in four categories of unequal intervals, they were treated as ordinal variables. Based on response distribution and clinical relevance of cutoff points to diabetes outcomes, taking medicines and checking feet were dichotomized to 7 days/week vs. less than 7 days/week, and testing sugar was dichotomized to 0–1 days/week vs. more than 1 day/week; these three SMBs were analyzed with logistic regression. The even response distribution in diabetes meal plan and physical activity categories allowed us to analyze these outcomes with ordered logistic regression.

The principal independent variables (FF support, professional support, self-efficacy, and depressive symptoms) were measured with Likert scales and had near normal response distributions, so they were treated as continuous variables. Other independent variables were chosen based on theoretical and documented relationships with diabetes SMB and were limited in number by our sample size. They included age, gender, education, and race/ethnicity, medication regimen, and functional limitations. We tested for colinearity by examining bivariate associations of each of these additional independent variables with FF support and professional support using the Student’s t-test for continuous variables and Pearson’s chi-square test for categorical variables. Income was not used due to the high number of missing responses.

We tested for interactions between medication type (oral or insulin) and FF medication support, and between functional limitations and FF physical activity support, by evaluating the effect of adding the corresponding interaction terms to the medication and physical activity adherence models.

To illustrate the clinical relevance of differences in FF support, we calculated regression model-based predicted probabilities36 for performance of each self-management task significantly associated with FF support. Predictions were calculated from multivariate models controlled for sociodemographic and health variables for those with the lowest and highest level of FF support.

To compare the independent effect of professional support and psychological factors to that of FF support, these variables were added to the FF support–SMB multivariate models. Psychological factors and professional support were evaluated in separate models due to limitations of sample size. Because psychological factors were hypothesized to mediate the support–SMB relationship, we used the method described by Baron and Kenny 37 to assess for signs of mediation. According to this method, the independent variable (FF support) must be significantly associated with the proposed mediator variable (self-efficacy or depressive symptoms), which we tested with multivariate linear regression, adjusted for the same sociodemographic and health variables as SMB models. Then, when the mediator is included in the main analysis, the effect of the independent variable (FF support) on the dependent variable (SMB) should be lessened.

Regression diagnostic procedures yielded no evidence for multicollinearity or influential outliers in any model. Analyses were done with Stata 9.2.38

RESULTS

The mean age of the 164 participants was 52.6 years, the majority (71%) was female, and 59% had at least high school education (Table 2). Participants’ were diagnosed with diabetes an average of 8.5 years previously, and their average A1C was 8.5%. None of the sociodemographic or health status measures in Table 2 were significantly associated with FF support or professional support in bivariate tests.

Table 2
Characteristics of Study Participants (N = 164)

The mean level of FF support for the SMBs varied from 2.48 out of 5 for physical activity support to 2.79 for diabetes medication support (Table 3). For medically related SMB (testing blood sugar, taking diabetes medications, and checking feet), the majority (67%–84%) reported following doctor recommendations 7 days/week. For lifestyle-related SMB (following a diabetes meal plan, physical activity), full adherence was reported 33–36% of the time.

Table 3
Support for Diabetes Tasks, Self-Management Behavior, Psychological Factors

Associations between FF Support and SMB Task Performance

In model 1, each unit increase in FF support, adjusting for sociodemographic and health factors, increased the adjusted odds ratio (AOR) for testing sugar by 1.77 (95% CI 1.21–2.58), increased the AOR for being in a higher physical activity category by 1.31 (0.98–1.73), and increased the AOR for being in a higher diabetes meal plan adherence category by 1.29 (0.97–1.70) (Table 4, model 1). The predicted probability of testing sugar ranged from 37% for participants with the lowest FF support to 85% with the highest level of FF support. The interaction term was significantly positively associated with medication taking in all medication models, indicating that the association of FF support was greater with taking oral medications.

Table 4
Results of Multivariable Logistic Regression Models Evaluating the Association of Family/Friend Support (FF Support), Psychological Factors, and Professional Support with Diabetes Self-Management Behavior Adherence

The interaction term of FF physical activity support × functional limitations was not significant when added to physical activity models.

Additional Analyses

Multivariate linear regressions, adjusted for sociodemographic and health factors, showed that FF support had a positive association with self-efficacy for testing sugar (coefficient 0.36, p=0.03), following a diabetes meal plan (coefficient 0.37, p=0.04), and checking feet (coefficient 0.32, p=0.04) (data not shown in tables). There were no significant associations between any type of FF support and depressive symptoms. When psychological factors were added to multivariate models predicting SMB performance (model 2 in Table 4), self-efficacy was associated with testing sugar, following a diabetes meal plan, and checking feet, while greater depressive symptom severity was associated with lower adherence to a diabetes meal plan. The relationship between FF support for testing sugar and performance of testing sugar remained significant and did not decrease when psychological factors were added. However, in the models predicting following a diabetes meal plan and checking feet, the AOR of FF support decreased from model 1 to model 2 when psychological factors were added.

When we added professional support for diabetes care to multivariate models of SMB performance (model 3 in Table 4), professional support from non-physicians was associated with increased diabetes meal plan and checking feet adherence. In contrast, increased physician support was associated with decreased diabetes meal plan adherence. Of note, the relationship between FF support and adherence to a diabetes meal plan became statistically significant after adjusting for professional support.

DISCUSSION

In adjusted analyses in our sample of urban African-American and Latino adults with diabetes, social support from family and friends was positively associated with testing sugar and with following a diabetes meal plan when adjusted for professional support, but was not significantly associated with other SMBs tested. FF support had a significantly more positive effect on taking oral medications than on taking insulin. Self-efficacy was significantly associated with testing sugar, following a diabetes meal plan, and checking feet, while depressive symptoms were significantly associated with diabetes meal plan adherence only. Additional analyses suggested that self-efficacy could be mediating the effect of FF support on diabetes meal plan adherence and checking feet. Support from non-physician health professionals was positively associated with following a diabetes meal plan and checking feet, while support from physicians was negatively associated with diabetes meal plan adherence. This study is the most detailed and comprehensive evaluation of the disease-specific family support–SMB relationship that we are aware of, as we examined the relationship of support to a wide variety of SMBs, compared the effects of family support with professional support on specific SMBs, and compared the effects of psychological factors and family support on SMB.

Our study builds on the findings of two prior studies on the association of disease support with specific SMBs. Bailey and colleagues39 found unadjusted correlations between overall diabetes social support and diet and exercise, but not with medications or testing sugar in 104 predominantly white participants. Shaw and colleagues’15 study of 208 participants, 57% of whom were white, found that family and friend support was associated with diet and foot care, but not physical activity or testing sugar, after adjusting for age, gender, and education. Our study assessed a wider range of SMBs than prior studies, analyzed differences in support effects on oral medications vs. insulin, and included additional potential confounders such as participants’ medication regimen and functional status. In this study three SMBs (taking medications, following a diabetes meal plan, and physical activity) had associations with FF support that were close to statistically significant. Given these results, it is possible that FF support had an effect on these SMBs that we were underpowered to detect at 95% confidence levels, but it is also possible that the FF support–SMB relationship for these SMBs did not vary in our participants.

Our findings add valuable knowledge about FF support in African-American and Latino adults with diabetes. Adherence to SMB recommendations, such as testing sugar, diet plans, and checking feet, have been found to be lower in African-Americans and Latinos,4043 and factors such as family characteristics and cultural food practices have been shown to affect SMB in African-Americans and Latinos more than non-Hispanic whites.4446 When comparing the effect of support on SMB in white and African-American patients,39,47 Bailey and colleagues found that support had stronger associations with SMB in African-Americans, while Fitzgerald and colleagues found that support mattered more to diet adherence in whites than African-Americans with diabetes. Studies with African-American4850 or Latino5153 participants have had conflicting results on whether social support is related to global measures of diabetes SMB. Our finding that social support is related to some SMBs and not others in these groups could help to explain these disparate findings and adds evidence that SMB interventions with African Americans and Latinos should consider the varying effects of social support on specific SMBs. We are not aware of any studies that directly compare the effects of social support in African-Americans and Latinos.

We analyzed two factors that could affect the FF support-SMB relationship. First, we examined psychological mediators of the FF support-SMB relationship and found that self-efficacy and depressive symptoms did not show signs of mediating the association between testing sugar and FF support. However, we found evidence that self-efficacy could mediate the association of FF support with diabetes meal plan adherence and checking feet. We noted a pattern of independent relationships between self-efficacy and SMB similar to that found by others,24 but we know of no other study looking simultaneously at FF support and self-efficacy with specific SMBs. It is possible that social support may affect some SMBs by increasing self-efficacy, while self-efficacy and support may act independently on other SMBs; this could be further tested in a prospective analysis.

Second, we tested the effect of professional support on the FF support-SMB relationship, finding that non-physician professional support was most strongly associated with tasks that had no significant relationship with FF support: checking feet and diabetes meal plan adherence. Past research has shown that certain provider communication styles are more influential on some SMB than others.54 Our findings could reflect the ability of professionals to give more effective support for tasks that require specialized or complex knowledge. Alternatively, when support from one source is ineffectual at supporting a task, patients may look to alternate support sources for that task. Interestingly, physician support was negatively associated with meal plan adherence, while non-physician support was positively associated. While it is possible that physician support is detrimental to those trying to follow a meal plan or that non-physician professionals are more effective at supporting healthy eating, it is also possible that people who have more severe diabetes or who are not doing well with their meal plan require more attention from their physician providers.

There are other factors that could explain differences in FF support effects on SMBs. The quality of family support for various tasks needs to be evaluated. For example, families may have more accurate knowledge of — and confidence in — providing support for a simpler task like testing sugar than a more complex task like following a diabetes meal plan. In addition, the influence of factors external to the family on the ability for support to affect SMB needs to be evaluated. For example, the availability of medical supplies, healthy food, and exercise facilities may influence the ability of patients to ‘act on’ support from others,1114 and these barriers are perceived as common in Detroit and other urban communities.27,55

This study has several limitations. First, our small sample size may have been underpowered to detect some statistically significant support-SMB associations. Our sample size also limited our ability to include more correlates of SMB in our models, such as comorbidities and cognitive function, although other measures of health status and functional status were included. Second, our participants were urban-dwelling, minority race or ethnicity, and were predominantly female, so these results may not be generalizable to other patient populations. The low response rate among African-Americans may have biased our sample and further limited generalizability. Third, it has been well documented that self-report of SMB leads to over-reporting of adherence,5658 although we relied on categories of adherence whose distribution probably still reflects the relative level of adherence between groups. Fourth, this study examines only perceived disease-specific support and does not distinguish between recognized types of support (such as instrumental, emotional, and informational) or between support from family versus friends. Since it appears that different types of support are important for different SMBs, future studies of social support and SMB should measure these specific types of support. Finally, cross-sectional data make it impossible to assess causality, as it is possible that FF support could be increased in response to more intensive SMB needs. Longitudinal analyses in progress could clarify the cause and effect relationship between support and SMB in this sample.

Despite these limitations, our findings have important implications for future diabetes self-management research and interventions. Future studies of social support and SMB should examine effects on each SMB separately. Studies need to move beyond assessing whether support affects SMB and focus on how support affects SMB, by examining moderators and mediators of the support-SMB relationship. SMB interventions might be improved by focusing on FF support or professional support for those SMBs that are most influenced by such support. Alternatively, interventions could attempt to increase the quality of social support for those behaviors currently not influenced by social support. Our results further suggest that interventions could affect some SMB by directly increasing self-efficacy, but in other cases intervening on both FF support and self-efficacy could be necessary.

In conclusion, in this group of urban African-American and Latino people with diabetes, diabetes-specific social support from family and friends was associated with different self-management behaviors than support from health professionals, and self-efficacy may have mediated the effect of support on some SMBs more than others. Prospective studies that assess the effects of social support with its mediators and moderators on diabetes self-management behaviors are a key next step in understanding the importance of social support to SMB. Communities and researchers working in partnership to improve diabetes SMB need to consider the effects of — and how best to mobilize — social support for specific SMBs in the design, implementation, and evaluation of their interventions.

Acknowledgements

Contributors We thank the CHASS/REACH Detroit Partnership staff, the REACH Detroit Partnership Steering Committee (www.reachdetroit.org), and the REACH Detroit Partnership intervention participants for their involvement in this study. The REACH Detroit Partnership is affiliated with the Detroit Community-Academic Urban Research Center (www.sph.umich.edu/URC/).

Funding Support for this study was provided by the Centers for Disease Control and Prevention Cooperative Agreement no. U50/CCU417409, the Michigan Diabetes Research and Training Center (NIDDK P60DK-20572). Michele Heisler is a VA HSR&D Career Development awardee. Ann-Marie Rosland is a Robert Wood Johnson Clinical Scholar.

Presentations Results from this paper were presented at the Society of General Internal Medicine Midwest Region Annual Meeting on 29 September 2007.

Conflict of Interest None disclosed.

Appendix

Table thumbnail

References

1. Lloyd CE, Wing RR, et al. Psychosocial correlates of glycemic control: the Pittsburgh Epidemiology of Diabetes Complications (EDC) study. Diabetes Res Clin Pract. 1993;21(2–3):187–195. [PubMed]
2. Sherbourne CD, Hays RD, et al. Antecedents of adherence to medical recommendations: results from the Medical Outcomes Study. J Behav Med. 1992;15(5):447–68. [PubMed]
3. Tillotson LM, Smith MS. Locus of control, social support, and adherence to the diabetes regimen. Diabetes Educ. 1996;22(2):133–9. [PubMed]
4. Gallant MP. The influence of social support on chronic illness self-management: a review and directions for research. Health Educ Behav. 2003;30(2):170–95. [PubMed]
5. DiMatteo MR. Social support and patient adherence to medical treatment: a meta-analysis. Health Psychol. 2004;23(2):207–18. [PubMed]
6. Shenkel RJ, Rogers JP, et al. Importance of "significant others" in predicting cooperation with diabetic regimen. Int J Psychiatry Med. 1985;15(2):149–55. [PubMed]
7. Levy RL. Social support and compliance: a selective review and critique of treatment integrity and outcome measurement. Soc Sci Med. 1983;17(18):1329–38. [PubMed]
8. Schafer LC, McCaul KD, et al. Supportive and nonsupportive family behaviors: relationships to adherence and metabolic control in persons with type I diabetes. Diabetes Care. 1986;9(2):179–85. [PubMed]
9. Hogan BE, Linden W, et al. Social support interventions: do they work? Clin Psychol Rev. 2002;22(3):381–440. [PubMed]
10. van Dam HA, van der Horst FG, et al. Social support in diabetes: a systematic review of controlled intervention studies. Patient Educ Couns. 2005;59(1):1–12. [PubMed]
11. Cohen DA, McKenzie TL, et al. Contribution of public parks to physical activity. Am J Public Health. 2007;97(3):509–14. [PubMed]
12. Diez Roux AV, Evenson KR, et al. Availability of recreational resources and physical activity in adults. Am J Public Health. 2007;97(3):493–9. [PubMed]
13. Laraia BA, Siega-Riz AM, et al. Proximity of supermarkets is positively associated with diet quality index for pregnancy. Prev Med. 2004;39(5):869–75. [PubMed]
14. Horowitz CR, Colson KA, et al. Barriers to buying healthy foods for people with diabetes: evidence of environmental disparities. Am J Public Health. 2004;94(9):1549–54. [PubMed]
15. Shaw BA, Gallant MP, et al. Assessing sources of support for diabetes self-care in urban and rural underserved communities. J Commun Health. 2006;31(5):393–412. [PubMed]
16. Greene J, Yedidia MJ, et al. Provider behaviors contributing to patient self-management of chronic illness among underserved populations. Journal of Health Care for the Poor & Underserved. 2005;16(4):808–24. [PubMed]
17. Anderson ES, Wojcik JR, et al. Social-cognitive determinants of physical activity: the influence of social support, self-efficacy, outcome expectations, and self-regulation among participants in a church-based health promotion study. Health Psychol. 2006;25(4):510–20. [PubMed]
18. Cohen S. Psychosocial models of the role of social support in the etiology of physical disease. Health Psychol. 1988;7(3):269–97. [PubMed]
19. DiIorio C, Shafer PO, et al. Project EASE: a study to test a psychosocial model of epilepsy medication managment. Epilepsy Behav. 2004;5(6):926–36. [PubMed]
20. Ciechanowski PS, Katon WJ, et al. Depression and diabetes: impact of depressive symptoms on adherence, function, and costs. Arch Intern Med. 2000;160(21):3278–85. [PubMed]
21. Park H, Hong Y, et al. Individuals with type 2 diabetes and depressive symptoms exhibited lower adherence with self-care. J Clin Epidemiol. 2004;57(9):978–84. [PubMed]
22. Connell CM, Davis WK, et al. Impact of social support, social cognitive variables, and perceived threat on depression among adults with diabetes. Health Psychol. 1994;13(3):263–73. [PubMed]
23. Warren BJ. Depression, stressful life events, social support, and self-esteem in middle class African American women. Arch Psychiatr Nurs. 1997;11(3):107–17. [PubMed]
24. Sarkar U, Fisher L, et al. Is self-efficacy associated with diabetes self-management across race/ethnicity and health literacy? Diabetes Care. 2006;29(4):823–9. [PubMed]
25. Two Feathers J, Kieffer EC, 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;95(9):1552–60. [PubMed]
26. Feathers JT, Kieffer EC, et al. The development, implementation, and process evaluation of the REACH Detroit Partnership’s Diabetes Lifestyle Intervention. Diabetes Educ. 2007;33(3):509–20. [PubMed]
27. Kieffer EC, Willis SK, 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;14(3 Suppl 1):S27–37. [PubMed]
28. Behavioral Risk Factor Surveillance System Survey Questionnaire. In. U.S. Department of Health and Human Services: Centers for Disease Control and Prevention (CDC); 2004.
29. Fitzgerald J, Davis W, et al. Development and validation of the diabetes care profile. Evaluation and the Health Professions. 1996;19(2):209–231. [PubMed]
30. Fitzgerald J, Anderson R, et al. The reliability of the diabetes care profile for African Americans. Evaluation and the Health Professions. 1998;21(1):52–65. [PubMed]
31. Toobert DJ, Hampson SE, et al. The summary of diabetes self-care activities measure. Diabetes Care. 2000;23(7):943–50. [PubMed]
32. Toobert DJ, Glasgow RE. Assessing diabetes self-management: the summary of diabetes self-care activities questionnaire. In: Clare B, ed. Handbook of psychology and diabetes. Amsterdam: Harwood Academic Publishers; 1994:351–375.
33. Behavioral Risk Factor Surveillance System Survey Questionnaire. In. U.S. Department of Health and Human Services: Centers for Disease Control and Prevention (CDC); 2003.
34. Williams GC, Freedman ZR, et al. Supporting Autonomy to Motivate Patients with Diabetes for Glucose Control. Diabetes Care. 1998;21(10):1644–1651. [PubMed]
35. Kroenke K, Spitzer R, et al. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. 2001;16(9):606–13. [PMC free article] [PubMed]
36. Long J.Regression models for categorical and limited dependent variables. Sage Publications; 1997.
37. Baron R, Kenny D. The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. J Pers Soc Psychol. 1986;51(6):1173–82. [PubMed]
38. StataCorp.Stata Statistical Software: Release 9. In. College Station, TX: StataCorp LP; 2005.
39. Bailey BJ, Lherisson-Cedeno D. Diabetes outcomes and practices: comparison of African Americans and Caucasians. J Natl Black Nurses Assoc. 1997;9(1):66–75. [PubMed]
40. Karter AJ, Ferrara A, et al. Self-monitoring of blood glucose: language and financial barriers in a managed care population with diabetes. Diabetes Care. 2000;23(4):477–83. [PubMed]
41. Kirk JK, Graves DE, et al. Racial and ethnic disparities in self-monitoring of blood glucose among US adults: a qualitative review. Ethn Dis. 2007;17(1):135–42. [PubMed]
42. Oster NV, Welch V, et al. Differences in self-management behaviors and use of preventive services among diabetes management enrollees by race and ethnicity. Dis Manage. 2006;9(3):167–75. [PubMed]
43. Thackeray R, Merrill RM, et al. Disparities in diabetes management practice between racial and ethnic groups in the United States. Diabetes Educ. 2004;30(4):665–75. [PubMed]
44. de Groot M, Welch G, et al. Cultural orientation and diabetes self-care in low-income African Americans with type 2 diabetes mellitus. Ethn Dis. 2003;13(1):6–14. [PubMed]
45. Fisher L, Chesla CA, et al. The family and disease management in Hispanic and European-American patients with type 2 diabetes. Diabetes Care. 2000;23(3):267–72. [PubMed]
46. Chesla CA, Fisher L, et al. Family predictors of disease management over one year in Latino and European American patients with type 2 diabetes. Fam Proc. 2003;42(3):375–90. [PubMed]
47. Fitzgerald JT, Anderson RM, et al. Differences in the impact of dietary restrictions on African Americans and Caucasians with NIDDM. Diabetes Educ. 1997;23(1):41–7. [PubMed]
48. Chlebowy DO, Garvin BJ. Social support, self-efficacy, and outcome expectations: impact on self-care behaviors and glycemic control in Caucasian and African American adults with type 2 diabetes. Diabetes Educ. 2006;32(5):777–86. [PubMed]
49. Ford ME, Tilley BC, et al. Social support among African-American adults with diabetes, Part 2: A review. J Natl Med Assoc. 1998;90(7):425–32. [PMC free article] [PubMed]
50. McDonald PE, Wykle ML, et al. Predictors of social support, acceptance, health-promoting behaviors, and glycemic control in African-Americans with type 2 diabetes. J Natl Black Nurses Assoc. 2002;13(1):23–30. [PubMed]
51. Carranza SN, LeBaron S. Adherence among Mexican Americans with type 2 diabetes: behavioral attribution, social support, and poverty. Fam Med. 2004;36(8):539–40. [PubMed]
52. Gleeson-Kreig J, Bernal H, et al. The role of social support in the self-management of diabetes mellitus among a Hispanic population. Public Health Nurs. 2002;19(3):215–22. [PubMed]
53. Wen LK, Shepherd MD, et al. Family support, diet, and exercise among older Mexican Americans with type 2 diabetes. Diabetes Educ. 2004;30(6):980–93. [PubMed]
54. Heisler M, Vijan S, et al. When do patients and their physicians agree on diabetes treatment goals and strategies, and what difference does it make? J Gen Intern Med. 2003;18(11):893–902. [PMC free article] [PubMed]
55. Zenk SN, Schulz AJ, et al. Fruit and vegetable intake in African Americans income and store characteristics. Am J Prev Med. 2005;29(1):1–9. [PubMed]
56. Johnson SB. Methodological issues in diabetes research. Measuring adherence. Diabetes Care. 1992;15(11):1658–67. [PubMed]
57. Vermeire E, Hearnshaw H, et al. Patient adherence to treatment: three decades of research. A comprehensive review. J Clin Pharm Ther. 2001;26(5):331–42. [PubMed]
58. Stewart M. The validity of an interview to assess a patient’s drug taking. Am J Prev Med. 1987;3(2):95–100. [PubMed]

Articles from Journal of General Internal Medicine are provided here courtesy of Society of General Internal Medicine