The longitudinal association between depressive symptoms and initiation of insulin therapy in people with type 2 diabetes in primary care.

Abstract

OBJECTIVE: To examine whether depressive symptoms are associated with time to insulin initiation in insulin-naïve people with type 2 diabetes in primary care. METHODS: 1,389 participants completed the Edinburgh Depression Scale (EDS) in 2005 and were followed until: 1) insulin therapy was started, 2) death, 3) an oral antihyperglycemic drug (OAD) prescription gap >1 year, 4) last OAD prescription in 2010 or 5) the end of the study (December 31, 2010). Cox regression analyses were used to determine whether there was a difference in time to insulin initiation between people with a low versus a high depression score at baseline, adjusting for potential demographic and clinical confounders, including HbA1c levels. RESULTS: The prevalence of depression (EDS ≥ 12) was 12% (n=168). After a mean follow-up of 1,597 ± 537 days, 253 (18%) participants had started insulin therapy. The rate of insulin initiation did not differ between depressed and non-depressed participants. People with depression were not more likely to start insulin therapy earlier or later than their non-depressed counterparts (HR=0.98, 95% CI 0.66-1.45), also after adjustment for sex and age (HR=0.95, 0.64-1.42). The association remained non-significant when individual candidate confounders were added to the age- and sex-adjusted base model. CONCLUSIONS: In the present study, depression was not associated with time to insulin initiation. The hypothesis that depression is associated with delayed initiation of insulin therapy merits more thorough testing, preferably in studies where more information is available about patient-, provider- and health care system factors that may influence the decision to initiate insulin.