63. Association of Hormonal Contraception Initiation With Subsequent Depression Diagnosis And Antidepressant Use In United States Military Health System Beneficiaries: A Cohort Study

Abstract

Examine the association of hormonal contraception initiation with subsequent depression diagnoses or Selective Serotonin Reuptake Inhibitor (SSRI) use among women enrolled in the United States Military Healthcare System (MHS) before and after accounting for the effect of healthcare utilization. Secondary analysis of insurance records from women ages 12-34 years old with 12 months of previous enrollment in the MHS during September 2014. We excluded women who were pregnant, using contraception, diagnosed with a mood disorder, or prescribed SSRIs in the prior 12 months. We used Kaplan-Meier and Cox Proportional Hazards regression analysis to assess the relationship of contraceptive initiation during September 2014 with antidepressant use or diagnosis with a depressive disorder during the subsequent 12 months. Women were censored from further analysis if they discontinued the contraceptive method initiated, started a new contraceptive method, became pregnant, or terminated enrollment in TRICARE Prime. We used 2 different control groups for our analyses. Control Group 1 (n=269,078) included all eligible women who did not start hormonal contraception. Control Group 2 (n=94,700) included only enrollees that accessed any inpatient, outpatient, or pharmacy services, excluding contraception, during September 2014. Contraception was initiated by 3,615 women (Pill/Patch/Ring: 84.9%, Intrauterine: 5.4%, Implant: 5.1%, Injectable: 4.5%). The most common progestin types used were Norgestimate (27.6%) and Levonorgestrel (16.4%). The depression diagnosis rate among women who started an Implant (12.9%), Intrauterine Contraception (11.2%), or a Pill/Patch/Ring (8.7%) was significantly higher than the rate in Control Group 1 (6.2%) but not Control Group 2 (9.0%). The SSRI rate was significantly higher among women who initiated intrauterine contraception (10.3%) or a Pill/Patch/Ring (5.6%) when compared to Control Group 1 (4.6%) but not Control Group 2 (6.8%). In multivariable analyses, adjusting for the effect of demographic factors and progestin type used, a higher hazard of depression diagnoses and SSRI use was seen among women with a history of military service and/or a junior enlisted insurance sponsor. Women aged 12- 19 years old were more likely to be diagnosed with depression (HR:1.12 (95%CI:1.05-1.19), p<0.001) and less likely to be started on SSRIs (0.63 (0.58-0.68), p<0.001). Compared to Control Group 1, higher depression rates were seen among women using Norgestimate, Levonorgestrel, Etonogestrel, or Norelgestromin. Compared to Control Group 2, Levonorgestrel and Norelgestromin use was associated with higher rates of depression (1.42 (1.05-1.92), p=0.024 and 1.93 (1.04-3.60), p=0.037), while use of Norethindrone containing contraception was protective (0.21 (0.05-0.85), p=0.028). Compared to Control Group 1, use of Norgestimate, Levonorgestrel, or Nogestrel was associated with a higher hazard of SSRI use. However, these associations did not persist when analyses were restricted to Control Group 2. Similar to other studies, our initial analysis demonstrated an association of initiating hormonal contraception with subsequent depression diagnoses or SSRI use. However, this association decreased or disappeared when we restricted our analysis to women who accessed care. This suggests that healthcare utilization rates may influence the association between contraception use and depression seen in previous studies. Further analysis is needed to confirm these findings.