Depression and unintended pregnancy in the National Longitudinal Survey of Youth: a cohort studyBMJ 2002; 324 doi: https://doi.org/10.1136/bmj.324.7330.151 (Published 19 January 2002) Cite this as: BMJ 2002;324:151
- David C Reardon, director of research (, )
- Jesse R Cougle, researcher
- Correspondence to: D C Reardon
- Accepted 29 August 2001
Psychological maladjustments after abortion are significantly associated with a history of depression.1 It has been suggested that prior psychological state is equally predictive of subsequent depression among women with unintended pregnancies regardless of whether they abort or carry to term.1 To examine this hypothesis we examined the National Longitudinal Study of Youth begun in 1979 with a nationwide cohort of 12 686 American youths aged 14-21.
Methods and results
The outcome variable of interest, depression, was assessed in 1992 for a subset of 4463 women using the Center for Epidemiological Studies depression (CES-D) scale scored by professional interviewers. This 20 item scale has good test-retest reliability among diverse population subgroups. In 1992 women were also asked whether their first delivered pregnancy had been the result of an intended pregnancy. Women who responded “yes” or “didn't matter” were excluded from our sample.
To control for prior psychiatric state we used the four item version of the Rotter internal-external locus of control scale, which was administered in 1979 (n=6215; mean 8.95, SD 2.1). The Rotter scale is intended to measure the extent to which people feel in control of their own destiny as opposed to having their fate decided by environment or chance. This abbreviated scale correlates well with self esteem, social class, and education. Higher external scores on the Rotter scale have been found to correlate with higher depression scores.2
Among all women surveyed depression scores were found to correlate with total family income in 1992 (r(3762)=−0.104, P<0.0001), highest educational grade completed in 1992 (r(4459)= −0.202, P<0.0001), age at first pregnancy event (r(3363)= −0.164, P<0.0001), and 1979 Rotter scores (r(4423)=0.135, P<0.0001). Depression scores were also significantly different between white (mean 9.48, SD 9.5) and non-white people (mean 11.69, SD 10.0; t=7.47, P<0.0001). These variables were used as controls.
The final sample used in these analyses includes only women for whom all control variables were available and who had their first abortion or first unintended delivery between 1980 and 1992 (n=421). On average, aborting women had had their first pregnancy in 1984 at age 22, and delivering women gave birth in 1986 at age 24. The mean Rotter score for aborting women and delivering women was 8.88 (SD 2.1) and 9.09 (2.2) respectively.
The table shows the results of a logistic regression analyses, stratified by marital status in 1992. Among married women, those who aborted were significantly more likely to be at “high risk” of clinical depression compared with those who delivered unintended pregnancies. The difference was not significant among unmarried women.
Our results do not support the hypothesis that depression after unintended pregnancies is independent of outcome. Among married women those who previously carried an unintended first pregnancy to term were at less risk of subsequent depression than women who aborted. Rates of high risk depression scores were comparable among unmarried women, however. This may be due to the stress unmarried women may experience in raising a child without support or it may be related to this group's higher rate of concealing past abortions. Compared with national averages, unmarried women in the National Longitudinal Study of Youth report only 30% of the expected abortions compared with married women, who report 74% of the expected abortions.3 Since shame, secrecy, and thought suppression regarding an abortion are all associated with greater post-abortion depression, anxiety, and hostility,4 the high rate of concealing past abortions in the National Longitudinal Study of Youth sample, 60% overall,3 may significantly dilute the observed effect.
Contributors: DR and JC jointly conceived and designed the study and contributed to the analysis and interpretation. JC did the statistical calculations, and DR wrote the paper with substantive participation by JC.
Funding Elliot Institute. The National Longitudinal Survey of Youth survey is conducted by the Center for Human Resource Research at Ohio State University and funded by the United States Department of Labor.
Conflicts of interest None.