Among sexually active adolescents interviewed at urban clinics in Connecticut,1 those who were pregnant were more likely than those who were not to report consistent use of hormonal contraceptives or dual methods six months later. By 12 months, the adolescents who were pregnant at baseline were still more likely to be consistent dual-method users, but the proportion reporting hormonal use had decreased substantially. The investigators infer that adolescent mothers may make a concerted effort to avoid a repeat pregnancy during their early postpartum months, but may have difficulty maintaining high levels of preventive behavior in the later postpartum period.
The participants, all aged 14-19 and nulliparous, were recruited in 1998-2000 at 10 clinics serving relatively poor communities in three Connecticut cities. Eligible teenagers were paid to participate in three structured, one-on-one interviews—at baseline, and six and 12 months later. At each interview, participants were tested for sexually transmitted diseases (STDs). Approximately half the 363 participants included in the analyses were pregnant, all in their third trimester, at baseline.
At baseline, researchers collected information on participants' social and demographic characteristics and sexual history. They also asked the women whether and on what proportion of occasions in the previous year they had used a condom or hormonal contraceptive (the pill, injection or levonorgestrel implant); at follow-up interviews, they asked about use in the past six months. At six and 12 months, participants reporting sexual activity in the previous six months were classified according to their consistency of method use—as consistent dual-method users, consistent users of hormonal methods only (used a hormonal, but not a condom, 100% of the time), consistent users of condoms only (used condoms always, but used hormonals less frequently or never) or inconsistent users (used neither method for at least one act of intercourse).
Investigators assessed any use of hormonals and condoms over time by using generalized estimating equations adjusted for social and demographic variables and sexual history. To assess consistency of method use for the two follow-up periods by baseline pregnancy status, they conducted multinomial logistic regression analysis adjusted for social and demographic variables, sexual history and baseline method use.
Among women pregnant at baseline and those not pregnant at baseline, the mean age was 17 years. In both groups, 39-44% of participants were Hispanic and 45-46% were black. Age at first sex was 15 years in each group; 29-30% of participants had had an STD in the past year. The proportion of women with a previous pregnancy was smaller among participants not pregnant at baseline than among those pregnant at baseline (23% vs. 51%). The nonpregnant group also had a smaller proportion of current students or high school graduates (11% vs. 23%). Among all participants, the median duration of sexual activity was 2.5 years, and the median number of partners per year of sexual activity was 1.2.
Of the women not pregnant at baseline, slightly more than half (52-59%) used a hormonal contraceptive in each study period. Pregnant women had predictably low use during the baseline period (18%) but subsequently reported a dramatic elevation in use (87%) for the early postpartum period. Although most of the women who were pregnant at baseline were still using a hormonal method in the later postpartum period (70%), the reduction in use from the first to second follow-up (17 percentage points) was statistically significant. For each follow-up period, the proportion of hormonal users among women pregnant at baseline differed significantly from that among women not pregnant at baseline. In contrast, the proportion of participants using condoms—throughout the study, in the range of 61-68% among women pregnant at baseline and 73-82% among the nonpregnant women—did not differ significantly at any time between the two groups and did not change substantially over time for either group.
In the first six months of follow-up, 23% of women who had been pregnant at baseline and 12% of those who had not been pregnant reported consistent dual-method use; 33% and 27%, respectively, were consistent users only of hormonals, 10% and 12% were consistent users only of condoms, 20% and 42% were inconsistent users, and 14% and 7% were not sexually active. During the second follow-up period, 16% and 10% used dual methods consistently, 30% and 26% used only hormonals consistently, 11% and 12% used only condoms consistently, 31% and 47% used both methods inconsistently, and 12% and 5% were not sexually active.
Pregnancy status at baseline was a strong predictor of contraceptive use throughout follow-up. At both six and 12 months, women pregnant at baseline were more likely than women not pregnant at baseline to be consistent dual-method users (relative risk ratios, 2.2-2.3) and not to be sexually active (4.6-6.0). In the first follow-up period, women pregnant at baseline also were more likely to be consistent users of hormonal methods (1.8).
Few other measures yielded similar results over time. Participants reporting recent hormonal use at baseline were more likely than others to be consistent users only of hormonal methods in each follow-up period (relative risks, 1.6-1.8). Women reporting no recent relationship at the six-month interview were more likely than women reporting a relationship lasting longer than six months to have had no sexual activity during either follow-up period (3.4-3.8).
In the first follow-up period, participants with fewer than 2.5 years of sexual activity or fewer than 1.2 sexual partners per year were more likely than others to use only condoms consistently (2.2-2.3). Participants aged 14-16 were more likely than older adolescents to be consistent users only of a hormonal method or to be consistent users of dual methods (1.7-1.8), and participants who had not been in a recent relationship were less likely than participants partnered for longer than six months to use dual methods consistently (0.3).
During the second follow-up period, young women sexually active for fewer than 2.5 years were more likely than other participants to be consistent users only of hormonal methods (1.5). Reporting consistent use only of condoms at 12 months was more likely if the participant was Hispanic rather than white (2.6), or if she had had a recent STD (1.9).
Although the authors concede the potential for bias in self-reported data, they comment that their study reveals noteworthy findings. In particular, they suggest that the results reveal "an active attempt to prevent repeat pregnancy," but not necessarily STDs, in adolescents' early postpartum months. They also observe, however, that "the effect of becoming a mother on subsequent contraceptive use weakens over time." Therefore, they conclude that "a 'booster' intervention may be needed in the later postpartum period to further reinforce contraceptive behavior and maintain consistent contraceptive use."
1. Kershaw TS et al., Short and long-term impact of adolescent pregnancy on postpartum contraceptive use: implications for prevention of repeat pregnancy, Journal of Adolescent Health, 2003, 33(5):359-368.