In a longitudinal study based on a year's worth of weekly survey data from 18–20-year-old women at risk of pregnancy, baseline levels of depression and stress were associated with contraceptive use.[1] Women who had symptoms of moderate or severe depression or stress were less likely to use long-acting methods, and more likely to use relatively ineffective ones, than they were to use the pill. In addition, those with elevated stress levels had higher odds than others of reporting that they had used no method during a week in which they had sexual activity.
The study was conducted in Michigan, among a population-based cohort of women recruited in 2008–2009. Participants completed an in-person baseline interview that covered their background characteristics, reproductive attitudes and intentions, contraceptive history and mental health. They then completed weekly surveys (online or by phone), in which they reported on their sexual activity and contraceptive use. To assess relationships between symptoms of stress and depression and contraceptive use, researchers analyzed data for weeks during the year following study enrollment in which women were sexually active and not pregnant.
A total of 689 women were included in the analyses, most (92%) of whom were 18 or 19 years old. The majority of participants were white (58%) or black (32%) and were in school (67%). Half were employed, and three in 10 were receiving public assistance. One in five attended religious services at least weekly, and about the same proportion never did. Eighty-five percent were in a romantic relationship, and 21% were living with a partner. Seven in 10 women had begun having vaginal sex at age 15 or later, and seven in 10 had had two or more partners; one-quarter had been pregnant at least once. Although 88% of participants had used contraceptives, 51% had had unprotected intercourse. Sixty percent were using a method at baseline, predominantly the pill. As assessed with standard screening tools, 27% of women had symptoms of moderate or severe depression; 25%, moderate or severe stress.
Survey responses provided information on 8,877 weeks of sexual activity. Contraceptive use was reported in 90% of weeks—noncoital methods (IUDs and hormonal methods other than emergency contraception) in 48% and coital ones (barrier methods, withdrawal, emergency contraception, rhythm and unspecified other methods) in 42%. During weeks in which contraceptives were used, the most common methods were the pill (44%), condoms (27%) and withdrawal (17%). In most weeks (63%), women reported using only one method; the main dual-method strategies were use of the pill with either condoms (18%) or withdrawal (14%).
Results of chi-square tests suggested that contraceptive use varied by women's baseline characteristics, including their scores on the depression and stress screening tools. These relationships were further examined in random effects multivariate logistic regression analyses. Overall, the multivariate model showed that women with moderate or severe stress symptoms had higher odds than others of reporting a week in which they had sex but did not use a contraceptive (odds ratio, 2.2). Level of depression was not associated with nonuse.
Among women using contraceptives, those with moderate or severe depression were less likely to use a long-acting method (IUD, implant or injectable), and more likely to use the ring or patch, than they were to use the pill (relative risks, 0.5 and 2.8, respectively). They also were more likely to use both the pill and withdrawal than they were to use a single method (1.2). Contraceptive users with moderate or severe stress had a lower risk of relying on long-acting methods than of being pill users (0.8), but an elevated risk of using the ring or patch (3.0), condoms (1.2) or withdrawal (1.3). Their risk of using various combinations of methods for dual protection was lower than their risk of using a single method (0.7–0.8).
The researchers acknowledge "several notable limitations" of their study, including the reliance on self-reported data, the use of psychological measures assessed only at baseline, the lack of information about aspects of women's health histories that may influence contraceptive choices and the limited generalizability of the sample. Additionally, they note that because the study did not explore reasons for women's contraceptive choices, their results "do not illuminate why women with psychological symptoms may resort to less effective methods or forgo contraception altogether." However, they conclude that routine screening can help clinicians to identify women with mental health symptoms, "who may be at risk for contraceptive nonuse and misuse."—D. Hollander
Reference
1. Hall KS et al., Role of young women's depression and stress symptoms in their weekly use and nonuse of contraceptive methods, Journal of Adolescent Health, 2013,