The influence of health care utilization on the association between hormonal contraception initiation and subsequent depression diagnosis and antidepressant use

Objective: Assess the influence of healthcare utilization on previously reported associations between contraception initiation, diagnosis of depression, and dispensing of antidepressant medications. Study design: Retrospective cohort analysis of insurance records from 272,693 women ages 12–34 years old enrolled in the United States Military Healthcare System in September 2014 and followed for 12 months. We compared outcomes of women who initiated hormonal contraception with all women eligible for care and then with women who accessed care during the study month using Kaplan-Meier and Cox proportional hazards analyses. Results: Women age 12–34 who initiated hormonal contraception experienced a higher risk of depression diagnosis and antidepressant use when compared to all enrolled women but not when compared to women who accessed care. Among those who accessed care, some progestins (i.e., levonorgestrel, Hazard Ratio (HR) = 1.46, and norelgestromin, HR = 1.93) were associated with an increased rate of depression diagnosis but not antidepressant use; norethindrone (HR = 0.21) was associated with a decreased rate of depression diagnosis. Conclusion: When compared to women utilizing their health insurance, associations between initiating hormonal contraception and depression diagnosis and antidepressant use decreased or disappeared. This suggests that healthcare utilization may have confounded previous reports of an association between hormonal contraception use and depression and antidepressant use.