Affective and Substance Abuse Disorders Following Abortion by Pregnancy Intention in the United States: A Longitudinal Cohort Study

Abstract

Background and Objectives: Psychological outcomes following termination of wanted pregnancies have not previously been studied. Does excluding such abortions affect estimates of psychological distress following abortion? To address this question this study examines long-term psychological outcomes by pregnancy intention (wanted or unwanted) following induced abortion relative to childbirth in the United States. Materials and Methods: Panel data on a nationally-representative cohort of 3935 ever-pregnant women assessed at mean age of 15, 22, and 28 years were examined from the National Longitudinal Survey of Adolescent to Adult Health (Add Health). Relative risk (RR) and incident rate ratios (IRR) for time-dynamic mental health outcomes, conditioned by pregnancy intention and abortion exposure, were estimated from population-averaged longitudinal logistic and Poisson regression models, with extensive adjustment for sociodemographic differences, pregnancy and mental health history, and other confounding factors. Outcomes were assessed using the Diagnostic and Statistical Manual, Version 4, American Psychiatric Association (DSM-IV) diagnostic criteria or another validated index for suicidal ideation, depression, and anxiety (affective problems); drug abuse, opioid abuse, alcohol abuse, and cannabis abuse (substance abuse problems); and summary total disorders. Results: Women who terminated one or more wanted pregnancies experienced a 43% higher risk of affective problems (RR 1.69, 95% CI 1.3–2.2) relative to childbirth, compared to women terminating only unwanted pregnancies (RR 1.18, 95% CI 1.0–1.4). Risks of depression (RR 2.22, 95% CI 1.3–3.8) and suicidality (RR 3.44 95% CI 1.5–7.7) were especially elevated with wanted pregnancy abortion. Relative risk of substance abuse disorders with any abortion was high, at about 2.0, but unaffected by pregnancy intention. Excluding wanted pregnancies artifactually reduced estimates of affective disorders by 72% from unity, substance abuse disorders by 11% from unity, and total disorders by 21% from unity. Conclusions: Excluding wanted pregnancies moderately understates overall risk and strongly understates affective risk of mental health difficulties for women following abortion. Compared to corresponding births, abortions of wanted pregnancies are associated with a greater risk of negative psychological affect, particularly depression and suicide ideation, but not greater risk of substance abuse, than are abortions of unwanted pregnancies. Clinical, research, and policy implications are discussed briefly.