Long-term incidence and recurrence of common mental disorders after abortion. A Dutch prospective cohort study

https://doi.org/10.1016/j.jpsychires.2018.04.001Get rights and content

Abstract

In a previous study (Van Ditzhuijzen et al., 2017) we investigated the incidence and recurrence of mental disorders 2.5 to 3 years post-abortion. The aim of the current study was to extend these findings with longer term follow up data, up until 5–6 years post-abortion. We compared data of women who had had an abortion of the Dutch Abortion and Mental Health Study (DAMHS) to women who did not have an abortion from the Netherlands Mental Health Survey and Incidence Study-2 (NEMESIS-2) (Ntotal = 2227). We used 1-to-1 matching on background confounding variables and measured post-abortion incidence and recurrence of common DSM-IV mental disorders (mood, anxiety, and substance use disorders) using the Composite International Diagnostic Interview (CIDI) version 3.0. After matching on confounding variables, abortion did not increase the likelihood that women had incident or recurrent mental disorders in the 5–6 years post-abortion (any incident mental disorder: OR = 3.66, p = .16; any recurrent mental disorder: OR = 0.22, p = .47). We found no evidence that experiencing an abortion increases the risk on new or recurrent mental disorders on the longer term.

Introduction

Credible review studies have stated that there is no evidence that abortion causes adverse mental health outcomes (APA, 2008; Charles et al., 2008; NCCMH, 2011). However, the field is characterized by methodological limitations such as insufficient confounder control, and conclusions have mostly been based on cross-sectional or short-term follow-up studies. It could therefore be argued that mental health problems that might arise only years after an abortion, are being missed in research; but the few high-quality studies that have focused on long-term postabortion mental health have found no evidence for this (e.g., Biggs et al., 2015, 2017; Munk-Olsen et al., 2011). These studies measured either the prevalence of (sub)clinical symptoms of a limited selection of mental disorder categories with brief inventories (Biggs et al., 2015, 2017), or psychiatric contacts from population-based registry data, which might underestimate actual psychiatric morbidity (Munk-Olsen et al., 2011). So far, no studies have used structured diagnostic interviews which assess long-term incidence and recurrence of a wide array of common clinical mental disorders. The current study aims to fill this gap.

In an earlier case-controlled cohort study, we found that abortion was not associated with higher incidence of disorders in the 2.5–3 years post-abortion, but women who had an abortion seemed at slightly increased risk for recurrence of any mental disorder (Van Ditzhuijzen et al., 2017). In the current study, we investigated whether these results hold in the long term. The aim was to answer the question whether the life event of terminating an unwanted pregnancy increases the risk on the development of clinical DSM-IV mental disorders 5–6 years postabortion.

Section snippets

Study design

In this prospective three-wave cohort study we compared women who had an abortion with women from the general population who never had an abortion. In research like this, it is impossible to use random assignment to treatment (abortion) or control condition, which implies that selectivity and confounding are inherent to this type of research. We used Coarsened Exact Matching (CEM) to deal with this general problem (Stuart, 2010; Rosenbaum and Rubin, 1983; Dehejia and Wahba, 2002; Cook et al.,

Participants

At T0, we included 325 participants in DAMHS. The initial response rate was 36.5%, mainly because we were unable to schedule interviews with women who had agreed to participate. Participant flow and response analysis results at T0 are described extensively in an earlier study (Van Ditzhuijzen et al., 2013). Of the initial 325 participants in DAMHS, 264 (81.2%) were re-interviewed at T1, 40 women could not be traced, 13 refused participation, and 8 women did not show up at the interview. At T2,

Discussion

In the unmatched (unadjusted) data, we found differences between cohorts for long-term incidence and recurrence of substance use disorders, and in incidence of any mental disorder, but not for other categories of disorders. However, these differences were strongly reduced after 1-to-1 matching on confounding variables. The initial differences seem not attributable to the abortion itself, but rather related to confounding variables.

Comparing the current findings to the findings after 2.5–3 years

Contributors

Authors Vollebergh and Van Nijnatten designed the study, in cooperation with authors De Graaf and Ten Have. Author Van Ditzhuijzen managed the literature searches, data collection and analysis, supervised by author Ten Have. Author Van Ditzhuijzen wrote the first draft of the manuscript. All authors contributed to and have approved the final manuscript.

Conflict of interest

None.

Financial support

This work was supported by the Dutch Ministry of Public health, Welfare and Sports, via the Netherlands Organization for Health Research and Development (ZonMw) [grant number 127000003].

Acknowledgements

We thank Ellen Wittkampf, MSc, who assisted in the coordination of the T2 data collection; Peter Lugtig, PhD, for statistical advice; and all the interviewers and participants, without whom this study could not have been conducted.

References (27)

  • R. De Graaf et al.

    The Netherlands mental health Survey and incidence Study-2 (NEMESIS-2): design and methods

    Int. J. Meth. Psychiat. Res.

    (2010)
  • R.H. Dehejia et al.

    Propensity score-matching methods for nonexperimental causal studies

    Rev. Econ. Stat.

    (2002)
  • D.G. Foster et al.

    A comparison of depression and anxiety symptom trajectories between women who had an abortion and women denied one

    Psychol. Med.

    (2015)
    • View full text
    © 2018 Elsevier Ltd. All rights reserved.