RESUMO
PURPOSE: The purpose of this meta-analysis was to investigate whether locus of control (LOC) for internal LOC, powerful others, or chance is correlated with how well people control diabetes (measured by H1C). METHODS: The literature search included Medline, CINAHL, PsychINFO, PsychArticles, Health Source, Academic Search Elite, EMBASE, Current Contents, and BIOsis databases, which yielded 296 articles. Selection criteria for inclusion were adult participants with diabetes mellitus, use of glycosylated hemoglobin (H1C) as a measure of glycemic control, measurement of a locus-of control scale, and a Pearson correlation (r) value. Seventeen articles met all inclusion criteria. The R package "metacor," was used to perform a random effects meta-analysis and estimate correlation coefficients between the 3 LOC measurements and H1C. RESULTS: The strength of subjects ascribing an internal LOC (across 13 studies) was uncorrelated with H1C (r = -0.0099; 95% confidence interval [CI], -0.1092, 0.0893). A LOC ascribed to powerful others LOC (across 9 studies) was similarly uncorrelated with H1C (r = 0.0928; 95% CI, -0.0136, 0.1993). Ascribing LOC to chance (across 9 studies) was also uncorrelated with H1C (r = 0.0926; 95% CI, -0.0398, 0.2250). CONCLUSIONS: This meta-analysis of 17 studies found no correlation between the control of diabetes and LOC. At best, there may be a weak correlation between powerful others and chance LOC with the metabolic control of diabetes. The use of LOC as a means of designing diabetes care can be discarded as this random effects meta-analysis found no significant correlation with effectiveness of diabetes control in adult patients.
