VIENNA – Long-term use of antidepressant medication does not appear to cause diabetes, according to findings from a longitudinal study.
The team, from the Institut National de la Santé et de la Recherche Médicale (INSERM) in Villejuif, France, found no association between fasting plasma glucose (FPG) or hemoglobin A1c levels measured over a 9-year study period and the use of antidepressants.
“Several studies have shown an association between antidepressant use and the risk of type 2 diabetes,” reported Marine Azevedo Da Silva at the annual meeting of the European Association for the Study of Diabetes.
The reason for this remains unclear, so the hypothesis for the study was that “if antidepressant use is causally associated with type 2 diabetes, then it should also be associated with glucose dysregulation,” said Ms. Azevedo Da Silva, a doctoral student at the Versailles (France) Saint-Quentin-en-Yvelines University.
While other research teams have examined this hypothesis, the results have been inconsistent, Ms. Azevedo Da Silva added. To solve some of the methodological issues of the prior studies, she and her colleagues used data from an epidemiological study on the insulin resistance syndrome (DESIR). This is a longitudinal cohort of 5,212 men and women without diabetes who were aged 30-60 years at recruitment.
Health assessments, including antidepressant use, FPG, and hemoglobin A1c, were performed at four time points: at recruitment in 1994 to 1996, and then at 3-year intervals between 1997-1999, 2000-2002, and 2003-2005. Data on 4,869 participants who did not have diabetes at baseline were available for analysis. Diabetes was defined as an FPG of less than 7 mmol/L (126 mg/dL) and no use of glucose-lowering medication. The majority (96%) of participants were not using antidepressants at baseline.
Baseline mean FPG and hemoglobin A1c levels were similar among antidepressant and non–antidepressant users, at 5.22 mmol/L (96 mg/dL) and 5.22%, versus 5.32 mmol/L (96 mg/dL) and 5.35%, respectively. Among the 4% of subjects who used antidepressants, there was no difference in these blood glucose measures according to the type of antidepressant used.
There were similar increases in both FPG and hemoglobin A1c in the two groups over time, at 0.02 mmol/L (0.36 mg/dL) per year for FPG in both antidepressant users and nonusers, and 0.01% per year for hemoglobin A1c.
Results had been adjusted for multiple confounding factors, including gender, age, marital status, education, and employment status; and several other health factors, such as alcohol use, smoking, body mass index, family diabetes history, hypertension, and other medication use.
Antidepressant use was self-reported, noted Ms. Azevedo Da Silva, which was one of the main limitations of the study. There was also a lack of information on the actual antidepressants used and the doses taken. In addition, there is the possibility of “confounding depression,” she acknowledged.
Nevertheless, this is a large observational, longitudinal cohort with an extended period of follow-up. FPG and hemoglobin A1c were robust clinical measures, and the team was able to study a large number of participants.
“Our results suggest that the association highlighted in recent studies between antidepressant use and type 2 diabetes may not be causal,” Ms. Azevedo Da Silva concluded.
This study is the fourth to find no association between antidepressant use and glucose dysregulation (Biol. Psych. 2011;70:978-84; Ann. Med. 2012;44:279-88; Psychopharmacology 2013;227:467-77), with only two other studies suggesting that there might be a link (Int. J. Clin. Pharm. 2011;33:484-92; PloS One 2011;6:e21551).
There is now a need to look at other reasons that might be responsible for the purported association, such as insulin resistance, she said.
The Ministry of French Research supported the study. Ms. Azevedo Da Silva had no conflicts of interest.