摘要：CONTEXT: Several studies have found a relationship between abortion and prior substance use, suggesting that a reduction
in substance use might help decrease abortion rates. However, such a conclusion requires a greater understanding
of the processes linking abortion and prior substance use.
METHOD: Path analysis of longitudinal data from 1,224 women was used to simultaneously test two pathways from
adolescent substance use to abortion by age 29, one mediated by higher rates of unplanned pregnancy and the other
independent of unplanned pregnancy rates. The model was then expanded to examine how these pathways change
when unconventional attitudes and behaviors (such as rebelliousness and low religiosity) are taken into consideration.
RESULTS: In the analysis that did not control for unconventionality, women who reported smoking cigarettes or using
marijuana or hard drugs at age 18 had an increased likelihood of subsequent unplanned pregnancy and, as a result,
higher rates of abortion. In addition, women who had used marijuana had an increased likelihood of abortion independent
of unplanned pregnancy rates. In the final model, unconventionality strongly predicted both abortion and
unplanned pregnancy. Moreover, it explained the associations between the use of hard drugs or marijuana and abortion
that were due to higher unplanned pregnancy rates.
CONCLUSIONS: Unconventionality mediates certain associations between substance use and abortion, perhaps because
unconventional women are more likely both to use substances and to engage in behaviors that increase their
risk of unplanned pregnancy. Hence, it seems unlikely that reducing substance use will result in substantially fewer
abortions.
