While the consequences of childbearing in adolescence have been well explored, little is known of the consequences of abortion which is, increasingly, a chosen outcome for unintended early conception. This study will collect background data on the approximately 600-700 young women of 17 and below who have not previously been pregnant and who present at the Comprehensive Care facility, in which they are enrolled, for pregnancy tests in a period of a year and one half; it will follow, prospectively, the 120 positives who elect induced abortion and, as controls, the 180 who carry to term and, among those whose tests are negative, a sample of 200 with age distributions similar to the study population. Among those who carry, spontaneous abortions will be monitored so that a small group of young women who miscarry can be identified as well, and included as a subgroup in the study population. Young women will be admitted to this 5 year study over a one and one half year period, and will be followed for 3 years with interviews at 1 year intervals, and telephone contacts each 6 months. Data collected at presentation and prospective observations will focus on variables in areas of (1) health; (2) physical development; (3) educational achievement; (4) living arrangements; (5) sexual behavior and fertility; (6) psychological measures of depression and mental health; and (7) economic well being and job history. The research will assess (1) the proportion of young women in the study and control populations who demonstrate changes over time in defined variables within each of these areas; (2) the significance of differences between the study and each control population in the proportion changing and/or departing from a normative expectation of change; and (3) the relationship between variables which are associated with change when it does take place within the study population; thus the study will attempt to establish which variables are associated with particular observed consequences of the pregnancy loss. Changes over time in these areas will be controlled by background information collected at presentation (for example, chronological and gynecological age at the index event) where such information is hypothesized as having an effect upon the expectation of change. Several regression techniques will be utilized in this analysis.
