Premium
Does antidepressant medication use affect persistence with diabetes medicines?
Author(s) -
Caughey Gillian E.,
Preiss Adrian K.,
Vitry Agnes I.,
Gilbert Andrew L.,
Ryan Philip,
Shakib Sepehr,
Esterman Adrian,
McDermott Robyn A.,
Roughead Elizabeth E.
Publication year - 2013
Publication title -
pharmacoepidemiology and drug safety
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.023
H-Index - 96
eISSN - 1099-1557
pISSN - 1053-8569
DOI - 10.1002/pds.3424
Subject(s) - medicine , discontinuation , metformin , sulfonylurea , depression (economics) , diabetes mellitus , antidepressant , population , veterans affairs , pharmacoepidemiology , psychiatry , medical prescription , pharmacology , anxiety , environmental health , endocrinology , economics , macroeconomics
Purpose This study aimed to examine the effect of antidepressant use on persistence with newly initiated oral antidiabetic medicines in older people. Methods A retrospective study of administrative claims data from the Australian Government Department of Veterans' Affairs, from 1 July 2000 to 30 June 2008 of new users of oral antidiabetic medicines (metformin or sulfonylurea). Antidepressant medicine use was determined in the 6 months preceding the index date of the first dispensing of an oral antidiabetic medicine. The outcome was time to discontinuation of diabetes therapy in those with antidepressant use compared with those without. Competing risks regression analyses were conducted with adjustment for covariates. Results A total of 29 710 new users of metformin or sulfonylurea were identified, with 7171 (24.2%) dispensed an antidepressant. Median duration of oral antidiabetic medicines was 1.81 years (95% CI 1.72–1.94) for those who received an antidepressant at the time of diabetes medicine initiation, by comparison to 3.23 years (95% CI 3.10–3.40) for those who did not receive an antidepressant. Competing risk analyses showed a 42% increased likelihood of discontinuation of diabetes medications in persons who received an antidepressant (subdistribution hazard ratio 1.42, 95% CI 1.37–1.47, p < 0.001). Conclusions The results of this large population‐based study demonstrate that depression may be contributing to non‐compliance with medicines for diabetes and highlight the need to provide additional services to support appropriate medicine use in those initiating diabetes medicines with co‐morbid depression. Copyright © 2013 John Wiley & Sons, Ltd.