Methods: The study selection followed Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA). Extant literature published between 2001 to 2018 were searched using eight search engines: PsycINFO, CINAHL, SocIndex, Social Work Abstracts, AgeLine, MedLine, Cochrane Library, and Academic Search Complete. Additional searches were conducted in gray literature. The inclusion criteria for participants were community-dwelling middle-aged and older adults aged 45 or above diagnosed of diabetes. Assessment of methodological quality and risk biases were conducted using the Item Bank for Assessing Risk of Bias and Confounding for Observational Studies of Interventions or Exposures by Agency for Healthcare Research and Quality.
Results: In total 12 studies met the inclusion criteria. The inter-rater reliability between the first and second coder, represented by the Cohen’s Kappa coefficient, was .57, indicating moderate agreement. The sample sizes ranged from 50 to 1834, and the mean ages ranged from 55 to 70. Of the included 12 studies, 10 studies found significant associations between self-efficacy and overall diabetes self-management or at least one diabetes self-management regimen, identifying self-efficacy as a significant predictor for diet and blood glucose monitoring. However, only five studies found significant relationships between self-efficacy and foot care, and two studies found that self-efficacy was not significant predictors for exercise and medication adherence respectively. In terms of self-efficacy and diabetes self-management by race, higher self-efficacy in Hispanics predicted better self-management behaviors, whereas mixed relations between self-efficacy and diabetes self-management were found in samples including black and white diabetes patients. Native Americans and Asian Americans were constantly underrepresented in study samples.
Conclusions and Implications: Strong evidence show that higher self-efficacy is highly predictive of improved behaviors in diet and blood glucose monitoring. However, mixed findings on exercise suggest that consistent measures of self-efficacy should be used to assess patients’ confidence in managing diabetes. In addition, it is worthwhile to consider focusing on improving the self-efficacy in middle-aged and older Mexican Americans in the development of diabetes education. Mixed findings in white and black population indicated that factors other than self-efficacy might be more salient in predicting patients’ self-management behaviors. More studies on Asian Americans are needed to better study the association between self-efficacy and diabetes self-management. In conclusion, health social workers working with middle-aged and older adults with diabetes could focus efforts on increasing self-efficacy, especially for Mexican Americans. Still, specific needs for particular racial groups should be considered as well.