Psychosocial Determinants of Mistimed and Unwanted Pregnancy: The Hamamatsu Birth Cohort (HBC) Study

Years of education mediate the association between adverse childhood experiences and unintended pregnancy: A population-based study in Japan

BACKGROUND

Adverse childhood experiences (ACEs) have been associated with unintended pregnancies, including mistimed pregnancies (MP) and unwanted pregnancies (UWP). However, it remains unknown which cluster of ACEs (i.e., child maltreatment/household dysfunction and deprivation/threat) are associated with MP/UWP and whether years of education mediate these associations.

OBJECTIVE

To investigate the association of the clusters of ACEs with MP and UWP, while also examining the mediating effect of education years.

PARTICIPANTS AND SETTING

A retrospective cohort study among 7652 postpartum women in Chiba, Japan.

METHODS

MP/UWP was defined by emotional responses to confirming pregnancy. Multinomial logistic regression analyses with multiple imputed datasets estimated the relative risk ratio (RRR) of MP/UWP by cumulative scores and each cluster of ACEs. Causal mediation analysis assessed the indirect effects of years of education.

RESULTS

Women with 4 or more ACEs were at a 2.4 times higher risk of MP (95 % confidence interval (CI): 1.6-3.8) and a 5.0 times higher risk of UWP (95 % CI: 3.1-8.2). Among ACE clusters, having 3 or more household dysfunction showed the strongest association with MP (RRR: 1.91, 95 % CI: 1.23-2.95), and having 3 or more deprivation showed the strongest association with UWP (RRR: 3.69, 95 % CI: 2.00-6.83). Education years mediated 16 % and 11 % of the association between total ACEs and MP/UWP, respectively, with a similar trend observed in each cluster.

CONCLUSIONS

Not only ACEs score but also each cluster of ACEs was associated with MP and UWP. The mediating effects of years of education were modest.

Teenage and young adult pregnancy and depression: findings from the Japan environment and children's study

Teenage pregnancy increases the threat of depression because of its many factors. Pregnancy during young adulthood may also have several risk factors for depression compared to older pregnancies. However, data on depression in young adult pregnancies are lacking. This study investigated the association between teenage and young adult pregnancy and depression. Data from the Japan Environment and Children's study was used as a nationwide multicenter prospective cohort study. A multivariate logistic regression was performed to investigate the association between age groups (14-19, 20-24, 25-29, 30-34, ≥ 35 years) and depression, adjusted for behavioral and sociodemographic characteristics. Depression was assessed using the Kessler Psychological Distress Scale. In total, 96,808 pregnant women responded to the questionnaire. Teenage (14-19 years) and young adult (20-24 years) pregnancy were associated with an increased risk of depression compared to older pregnancy (≥ 35 years) (teenage: OR 4.28, 95% confidence interval, CI [3.24-5.64]; young adult: OR 3.00, 95% CI [2.64-3.41]). After adjusting for covariates, the magnitude of the risk of depression was attenuated. However, teenage and young adult pregnancy remained at a significantly increased risk of depression compared to older pregnancy (teenage: OR 2.38, 95% CI [1.77-3.21]; young adult: OR 2.14, 95% CI [1.87-2.46]). Our findings indicate that teenage and young adults' pregnancy are at an increased risk of depression compared to older pregnancy. These findings suggest prioritizing teenage and young pregnant women for prevention and interventions related to depression.

Comparing pregnancy and pregnancy outcome rates between adolescents with and without pre-existing mental disorders

BACKGROUND

There are limited population-based data on the role of mental disorders in adolescent pregnancy, despite the presence of mental disorders that may affect adolescents' desires and decisions to become pregnant.

OBJECTIVE

This study aimed to examine the relationship between specific types of mental disorders and pregnancy rates and outcome types among adolescents aged 13-19 years, using single-year age groups.

METHODS

We conducted a retrospective cohort study using data from the Merative™ MarketScan Research Databases. The study population consisted of females aged 13-19 years with continuous insurance enrollment for three consecutive calendar years between 2005 and 2015. Pregnancy incidence rates were calculated both overall and within the different categories of mental disorders. The presence of mental disorders, identified through diagnosis codes, was classified into 15 categories. Pregnancy and pregnancy outcome types were determined using diagnosis and procedure codes indicating the pregnancy status or outcome. To address potential over- or underestimations of mental disorder-specific pregnancy rates resulting from variations in age distribution across different mental disorder types, we applied age standardization using 2010 U.S. Census data. Finally, multivariable logistic regression models were used to examine the relationships between 15 specific types of mental disorders and pregnancy incidence rates, stratified by age.

RESULTS

The age-standardized pregnancy rate among adolescents diagnosed with at least one mental disorder was 15.4 per 1,000 person-years, compared to 8.5 per 1,000 person-years among adolescents without a mental disorder diagnosis. Compared to pregnant adolescents without a mental disorder diagnosis, those with a mental disorder diagnosis had a slightly but significantly higher abortion rate (26.7% vs 23.8%, P-value < 0.001). Multivariable logistic regression models showed that substance use-related disorders had the highest odds ratios (ORs) for pregnancy incidence, ranging from 2.4 [95% confidence interval (CI): 2.1-2.7] to 4.5 [95% CI:2.1-9.5] across different age groups. Overall, bipolar disorders (OR range: 1.6 [95% CI:1.4-1.9]- 1.8 [95% CI: 1.7-2.0]), depressive disorders (OR range: 1.4 [95% CI: 1.3-1.5]- 2.7 [95% CI: 2.3-3.1]), alcohol-related disorders (OR range: 1.2 [95% CI: 1.1-1.4]- 14.5 [95% CI: 1.2-178.6]), and attention-deficit/conduct/disruptive behavior disorders (OR range: 1.1 [95% CI: 1.0-1.1]- 1.8 [95% CI: 1.1-3.0]) were also significantly associated with adolescent pregnancy, compared to adolescents without diagnosed mental disorders of the same age.

CONCLUSION

This study emphasizes the elevated rates of pregnancy and pregnancy ending in abortion among adolescents diagnosed with mental disorders, and identifies the particular mental disorders associated with higher pregnancy rates.

Pregnancy intention and preconception contraceptive behaviors and substandard prenatal care in France

OBJECTIVE

Evaluate the association between a combined measure of time-based pregnancy intention and preconception contraceptive behavior and suboptimal prenatal care.

POPULATION AND METHOD

Women delivering a live birth in all maternity units during one week in March 2016 were interviewed in the postpartum ward (N = 13,132). Multinomial logistic regression models were used to assess the association between the indicator of pregnancy intention and substandard prenatal care (late initiation of care and less than the recommended number of prenatal visits (<60% recommended)).

RESULTS

83.6% of women had timed pregnancies, 4.7% had mistimed pregnancies but discontinued contraception to conceive, 8.0% had mistimed pregnancies without discontinuing contraception to conceive and 3.7% had unwanted pregnancies. Women with timed pregnancies or mistimed pregnancies despite discontinuing contraception to conceive were more socially advantaged than those who had an unwanted pregnancy or a mistimed pregnancy without discontinuing contraception to conceive. 3.3% of women had a substandard number of prenatal visits and 2.5% had delayed prenatal care initiation. The adjusted odds ratios (aOR) of substandard prenatal visits were high among women with unwanted pregnancies (aOR=2.78; 95% confidence interval [1.91-4.05]) and women with mistimed pregnancies who had not discontinued contraception to conceive (aOR=1.69; [1.21-2.35]) compared to women with timed pregnancies. No difference was observed for women with mistimed pregnancies who discontinued contraception to conceive (aOR=1.22; [0.70-2.12]).

CONCLUSION

Using routinely collected information on preconception contraception allows a more nuanced assessment of pregnancy intentions that can help caregivers identify women at greater risk of substandard prenatal care.

Psychiatric vulnerability and the risk for unintended pregnancies, a systematic review and meta-analysis

Background

Unintended pregnancies (UPs) are a global health problem as they contribute to adverse maternal and offspring outcomes, which underscores the need for prevention. As psychiatric vulnerability has previously been linked to sexual risk behavior, planning capacities and compliance with contraception methods, we aim to explore whether it is a risk factor for UPs.

Methods

Electronic databases were searched in November 2020. All articles in English language with data on women with age ≥ 18 with a psychiatric diagnosis at time of conception and reported pregnancy intention were included, irrespective of obstetric outcome (fetal loss, livebirth, or abortion). Studies on women with intellectual disabilities were excluded. We used the National Institutes of Health tool for assessment of bias in individual studies and the Grading of Recommendations Assessment, Development and Evaluation method for assessment of quality of the primary outcome.

Findings

Eleven studies reporting on psychiatric vulnerability and UPs were included. The participants of these studies were diagnosed with mood, anxiety, psychotic, substance use, conduct and eating disorders. The studies that have been conducted show that women with a psychiatric vulnerability (n = 2650) have an overall higher risk of UPs compared to women without a psychiatric vulnerability (n = 16,031) (OR 1.34, CI 1.08–1.67) and an overall weighed prevalence of UPs of 65% (CI 0.43–0.82) (n = 3881).

Interpretation

Studies conducted on psychiatric vulnerability and UPs are sparse and many (common) psychiatric vulnerabilities have not yet been studied in relation to UPs. The quality of the included studies was rated fair to poor due to difficulties with measuring the outcome pregnancy intention (use of various methods of assessment and use of retrospective study designs with risk of bias) and absence of a control group in most of the studies. The findings suggest an increased risk of UPs in women with psychiatric vulnerability. As UPs have important consequences for mother and child, discussing family planning in women with psychiatric vulnerabilities is of utmost importance.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12884-022-04452-1.

Association Between First Depressive Episode in the Same Year as Sexual Debut and Teenage Pregnancy

PURPOSE

This study aimed to examine whether the timing of depression onset relative to age at sexual debut is associated with teenage pregnancy.

METHODS

Using data from 1,025 adolescent girls who reported having had sex in the National Comorbidity Survey-Adolescent Supplement, we applied cox proportional hazards models to test whether depression onset before first sex, at the same age as first sex, or after first sex compared with no depression onset was associated with experiencing a first teenage pregnancy. We examined the unadjusted risk by depression status as well as risk adjusted for adolescents' race/ethnicity, marital status, poverty level, whether the adolescent lived in a metropolitan area, living status, age at first sex, parental education, and age of mother when the adolescent was born.

RESULTS

In both unadjusted and adjusted models, we found that adolescents with depression onset at the same age as having initiated sex were at an increased risk of experiencing a teenage pregnancy (unadjusted hazard ratio [HR] = 2.5, 95% confidence interval [CI]: 1.08-5.96; adjusted HR = 2.7, 95% CI: 1.15-6.34) compared with those with no depression onset. Moreover, compared with those with no depression onset, the risk of pregnancy for girls experiencing depression onset before first sex also increased but was not significant (adjusted HR = 1.5, 95% CI: .82-2.76).

CONCLUSIONS

Timing of first depressive episode relative to age at first sexual intercourse plays a critical role in determining the risk of teenage pregnancy. Timely diagnosis and treatment of depression may not only help adolescents' mental well-being but may also help them prevent teenage pregnancy.

Is Preconception Substance Use Associated With Unplanned or Poorly Timed Pregnancy?

OBJECTIVE

Unplanned and poorly timed pregnancies are associated with adverse maternal and neonatal outcomes. Further understanding of preconception substance use with unplanned and poorly timed pregnancy is warranted.

METHODS

Data were analyzed from a prospective study enrolling women early in pregnancy. Preconception tobacco, alcohol, marijuana, opioid, and cocaine use was ascertained. Participants reported whether their current pregnancy was planned and whether it was a good time to be pregnant. Multivariable logistic regression modeling generated risk estimates for preconception substance use, and pregnancy planning and timing, adjusting for confounders.

RESULTS

Overall, 37.2% reported unplanned pregnancy, 13.0% poorly timed pregnancy, and 39.0% reported either unplanned and/or poorly timed pregnancy. Within 6 months preconception, one-fifth (20.2%) reported nicotine cigarette use. In the month before conception, 71.8% reported alcohol use, 6.5% marijuana, and approximately 1% opioid or cocaine use. Multivariable analysis demonstrated preconception opioid use was associated with increased odds of poorly timed pregnancy (odds ratio [OR] 2.87, 95% confidence interval [CI] 1.03-7.99). Binge drinking the month before conception was associated with increased odds of poorly timed pregnancy and unplanned pregnancy (OR 1.75, 95% CI 1.01-3.05; and OR 1.68, 95% CI 1.01-2.79, respectively). Marijuana use 2 to 3 times in the month preconception was associated with increased risk of unplanned pregnancy, and unplanned and/or poorly timed pregnancy compared with nonuse (OR 1.78, 95% CI 1.03-3.08; and OR 1.79, 95% CI 1.01-3.17, respectively). Preconception tobacco or cocaine use was not associated with unplanned or poorly timed pregnancy following adjustment.

CONCLUSIONS

We demonstrate increased odds of unplanned or poorly timed pregnancy among women with preconception binge drinking, marijuana use, and opioid use; however, no association is observed with other substances after multivariable adjustment, including tobacco. Further research to evaluate high-level preconception substance use and substance disorders with pregnancy planning and timing is warranted. Focused efforts optimizing preconception health behaviors and reducing risk of unplanned or poorly timed pregnancy are needed.

Adverse life experiences and risk of unintended pregnancy in adolescence and early adulthood: Implications for toxic stress and reproductive health

Objective

We examined the effects of adverse life experiences (ALEs) on rates of unintended first pregnancy, including differential effects by race/ethnicity and socioeconomic status, among women in a national longitudinal cohort study.

Methods

We drew upon 15-years of data from 8810 adolescent and young adult females in the National Longitudinal Study of Adolescent to Adult Health. Using 40 different ALEs reported across childhood and adolescence, we created an additive ALE index, whereby higher scores indicated greater ALE exposure. We employed Cox proportional hazard models, including models stratified by racial/ethnic and socioeconomic groups, to estimate the effects of ALEs on time to first unintended pregnancy, controlling for time-varying sociodemographic, health and reproductive covariates.

Results

Among all women, a 1-standard deviation increase in ALE scores was associated with an increased rate of unintended first pregnancy (adjusted Hazard Ratio 1.11, 95% Confidence Interval=1.04-1.17). In stratified models, associations between ALE scores and risk of unintended pregnancy varied across racial/ethnic, socioeconomic, and age groups and according to various elevated ALE thresholds. For example, the 1-standard deviation increase in ALE score indicator increased the unintended pregnancy risk for African-American (aHR=1.12, CI=1.01-1.25), Asian (aHR 1.69, CI=1.26-2.26), and White women (aHR=1.12, CI=1.03-1.22), women in the lowest ($0-$19,999; aHR=1.21, CI = 1.03-1.23) and highest (>$75,000; aHR=1.36, CI=1.12-1.66) income categories, and women aged 20-24 (aHR=1.13, CI=1.04-1.24) and >24 years (aHR 1.25, CI=1.06-1.47), but not among the other sociodemographic groups.

Conclusion

ALEs increased the risk of unintended first pregnancy overall, and different levels of exposure impacting the risk of pregnancy differently for different sub-groups of women. Our ongoing research is further investigating the role of stress-associated adversity in shaping reproductive health outcomes and disparities in the United States.

Risk factors for antenatal anxiety: A systematic review of the literature

BACKGROUND

Given the prevalence of antenatal anxiety and its consistent associations with adverse pregnancy and child outcomes, early detection and management of anxiety are essential.

OBJECTIVE

The aim was to identify risk factors for anxiety among pregnant women by systematically reviewing original research.

METHODS

Cross-sectional, case-control and cohort studies that examined associations between antenatal anxiety and at least one potential risk factor prospectively or retrospectively and measured anxiety independent from other mental health conditions were included. Studies rated strong/moderate in methodological quality appraisal were used to synthesise the evidence.

RESULTS

Factors associated with greater risk of anxiety included previous pregnancy loss, medical complications, childhood abuse, intimate partner violence, denial/acceptance coping styles, personality traits, inadequate social support, history of mental health problems, high perceived stress and adverse life events.

CONCLUSIONS

Several risk factors identified in this review are detectable in routine prenatal care visits (e.g. previous pregnancy loss, pregnancy complications), potentially modifiable (e.g. coping styles, social support, partner factors) and can be identified prior to pregnancy (e.g. psychosocial factors), underlining the significance of pre-conception mental health screening.

Social Disparities in the Relationship Between Depression and Unintended Pregnancy During Adolescence and Young Adulthood

PURPOSE

We investigated the influence of depression on subsequent risk of unintended pregnancy and social disparities within this relationship, during adolescence and young adulthood.

METHODS

Drawing upon 15-year, nationally representative data from 8,810 young U.S. women in the National Longitudinal Study of Adolescent to Adult Health, we estimated associations between depression and time to first pregnancies reported as unintended, overall and stratified by race/ethnicity, socioeconomic status, and age with Cox proportional hazard models.

RESULTS

Moderate/severe depression symptoms were associated with an increased risk of unintended first pregnancy (hazard ratio [HR], 1.21; confidence interval [CI], 1.02-1.44). In stratified models, depression increased the pregnancy risk for all minority groups (HRs, 1.36-3.25) but not white women. Depression increased the pregnancy risk for women with $0-$19,999 (HR, 1.48; CI, 1.11-1.98) and $20,000-$49,999 (HR, 1.33; CI, 1.05-1.68) income levels but not those at higher levels. Depression increased the pregnancy risk for adolescents <20 years (HR, 1.35; CI, 1.07-1.71) but decreased the risk for women >24 years (HR, .47; CI, .25-.86).

CONCLUSIONS

Findings may inform more equitable, holistic public health strategies that target depression as a modifiable risk factor for adverse reproductive outcomes during adolescence and young adulthood.

The prevalence of unplanned pregnancy ending in birth, associated factors, and health outcomes

STUDY QUESTION

What are associated factors of unplanned pregnancies ending in birth?

SUMMARY ANSWER

Pregnancies that were less planned were associated with women of lower socio-economic status (SES), an unhealthier lifestyle before and during the pregnancy, more stress, and less social support.

WHAT IS KNOWN ALREADY

In Europe, the prevalence of unplanned pregnancy leading to birth varies. Unplanned pregnancy is more common among socially disadvantaged women, and associated with adverse pregnancy outcomes.

STUDY DESIGN, SIZE, DURATION

In a cross-sectional study, 517 women were recruited from May through September 2015.

PARTICIPANTS/MATERIALS, SETTING, METHODS

Women were recruited from six hospitals in Flanders, Belgium. Data from self-report and medical records were collected during the first 5 days postpartum. The validated London Measure of Unplanned Pregnancy was used to collect data regarding pregnancy planning. Data were analysed with Mann-Whitney U tests, Kruskal-Wallis tests, and multiple linear regression analysis.

MAIN RESULTS AND THE ROLE OF CHANCE

The majority of the pregnancies (83%) ending in birth were planned, 15% were ambivalent, and 2% unplanned. Women who are multigravida (95% CI -0.30 to -0.02), less well educated (95% CI 0.07-0.85), single or having a non-cohabiting relationship (95% CI 0.01-2.53), having history of drug abuse (95% CI -2.07 to -0.35), and experiencing intimate partner violence (95% CI -3.82 to -1.59) tended to have a significantly higher risk of a less planned pregnancy. Less planned pregnancies were significantly associated with initially unwanted pregnancies (P < 0.001), no folic acid or vitamin use before pregnancy (P < 0.001), lower number of prenatal visits (P = 0.03), smoking during pregnancy (P < 0.001), more stress (P = 0.002), lower relationship satisfaction (P = 0.001), and less social support (P < 0.001). Less planned pregnancies were also significantly associated with hyperemesis (P < 0.001) and shorter duration of delivery (P = 0.03). No differences were found in neonatal outcomes.

LIMITATIONS, REASONS FOR CAUTION

The prevalence of unplanned pregnancies is probably underestimated due to overrepresentation of women with higher SES in this study. Women's emotions may have influenced the answer to certain questions. Owing to the cross-sectional design, no causal relationships could be established.

WIDER IMPLICATIONS OF THE FINDINGS

This study emphasizes the importance of targeting socially disadvantaged women in the prevention of unplanned pregnancies.

STUDY FUNDING/COMPETING INTERESTS

This study was funded by the Research Foundation - Flanders (FWO). The authors have no conflict of interests.

TRIAL REGISTRATION NUMBER

Not applicable.

Are pregnancy planning and timing associated with preterm or small for gestational age births?

OBJECTIVE

To investigate whether unplanned or poorly timed pregnancies (self-reported at enrollment) are associated with preterm or small for gestational age births.

DESIGN

Prospective cohort study.

SETTING

Not applicable.

PATIENT(S)

Two thousand six hundred fifty-four pregnant women <18 weeks estimated gestational age with a singleton pregnancy.

INTERVENTION(S)

None.

MAIN OUTCOME MEASURE(S)

Preterm and small for gestational age births.

RESULT(S)

In adjusted analyses, pregnancy planning was not statistically significantly associated with preterm (odds ratio [OR] 1.18; 95% confidence interval [CI], 0.85-1.65) or small for gestational age birth (OR 1.17; 95% CI, 0.69-1.97). Similarly, poorly timed pregnancies were not statistically significantly associated with preterm (OR 0.85; 95% CI, 0.53-1.38) or small for gestational age birth (OR 0.92; 95% CI, 0.65-1.29). Combining pregnancy planning (yes/no) and timing (yes/no) into a 4-level category showed no statistically significant association with preterm birth or small for gestational age.

CONCLUSION(S)

In a large cohort with antenatally assessed pregnancy planning and timing, outcome data collected from medical record abstraction, and robust analysis adjusting for multiple confounding factors including maternal demographics, medical conditions, and other risk factors, neither pregnancy planning nor pregnancy timing showed a statistically significant association with preterm or small for gestational age infants. This study improves upon previous analyses that lacked adjustment for confounding and used retrospective self-reporting to assess pregnancy planning and timing, and preterm and small for gestational age births. Findings may differ in higher risk populations with higher prevalence of preterm or small for gestational age births.

FACTORS INFLUENCING MISTIMED AND UNWANTED PREGNANCIES AMONG NEPALI WOMEN

This paper assesses the factors influencing mistimed and unwanted pregnancies in Nepal separately using data from the 2011 Nepal Demographic and Health Survey. Women who had given birth within the five years before the survey were interviewed about the intendedness of their last pregnancy. The data were analysed with a chi-squared test, followed by multiple logistic regression analysis. Among the total 5391 participants, 11.29% and 13.13% reported their last pregnancy as mistimed and unwanted respectively. Logistic regression analysis showed that women from the hill region were more likely to report mistimed pregnancy, while women from the Western and Far-Western development regions were less likely to report mistimed pregnancy. Education status was positively correlated with the reporting of mistimed pregnancy. Women involved in agriculture, with full autonomy on household decision, with some exposure to mass media, belonging to higher age group and having third or higher parity were less likely to report mistimed pregnancy. Similarly, women from the Western development region had relatively higher odds of reporting unwanted pregnancy. Women with husbands involved in a paid job had lower odds of unwanted pregnancy. Women's autonomy was also positively correlated with unwanted pregnancy. Women with the intention to use contraceptive had lower odds of unwanted pregnancy. Interventions targeting the factors identified by this study could be useful in reduction of mistimed and unwanted pregnancies among Nepali women.

Risk factors of transient and persistent anxiety during pregnancy

PURPOSE

chronic poor mental health over the course of pregnancy contributes to greater adverse maternal and child outcomes. Identifying women with chronic depressive or anxiety symptoms can provide opportunities to reduce distress and improve pregnancy outcomes. The objective of this study was to determine risk factors of chronic antenatal depressive and anxiety symptoms using a longitudinal pregnancy cohort in Alberta, Canada.

METHODS

women with singleton pregnancies were included (N=3021). Anxiety and depressive symptoms were measured in the second and third trimesters using the Spielberger State-Trait Anxiety Inventory and the Edinburgh Postnatal Depression Scale, respectively. On the basis of the timing and persistence of symptoms, the following three mutually exclusive subgroups for each anxiety and depressive symptoms were created: never symptomatic, symptomatic only in the second trimester, and symptomatic at both time points. Separate logistic regression models were used to derive risk factors for each subgroup.

FINDINGS

women with chronic anxiety or depressive symptoms were distinguished from those with transient symptoms or no symptoms by their optimism scores, in which less optimistic pregnant women had a four-fold increased risk for developing chronic depressive or anxiety symptoms compared with more optimistic women (AOR varied from 4.30 to 4.93). Additionally, high perceived stress, low social support, history of mental health issues were common predictors of chronic anxiety and depressive symptoms in pregnancy. Partner tension was the exclusive predictor of anxiety symptoms (AOR varied from 1.94 to 2.31) and poor physical health (AOR 2.54; 95% CI 1.32-4.89), unplanned pregnancy (AOR 3.05; 95% CI 1.61-5.79), and infertility treatments (AOR 4.98; 95% CI 1.85-13.39) were unique predictors of chronic depressive symptoms.

CONCLUSIONS

knowledge of the risk factors of chronic poor mental health during pregnancy might inform the development of effective strategies within the limited resources of health-care systems to target populations with greater needs for interventions.

Stress symptoms and frequency of sexual intercourse among young women

INTRODUCTION

We have previously documented the relationships between stress and depression symptoms and adolescent women's nonuse and misuse of condoms and other contraceptive methods and on their unintended pregnancy rates.

AIM

Here, we examine relationships between mental health symptoms and another understudied adolescent reproductive health behavior-frequency of sexual intercourse.

MAIN OUTCOME MEASURE

Our outcome was weekly sexual intercourse activity.

METHODS

We used panel data from a longitudinal, population-based cohort study of 992 women ages 18-20. Weekly journals measured sociodemographic, relationship, reproductive, and mental health characteristics, sexual and contraceptive behaviors, and pregnancy history. We examined 27,130 surveys from 952 women during the first study year. Predictors of weekly sexual intercourse were moderate to severe stress (Perceived Stress Scale-4) and depression (Center for Epidemiologic Studies Depression Scale-5) symptoms measured at baseline. Multilevel, mixed-effects logistic regression models estimated the relationships between stress and depression symptoms and the weekly odds of sexual intercourse while adjusting for covariate fixed effects and random woman effects.

RESULTS

Nearly a quarter of the sample had moderate to severe stress (23%) and depression (24%) symptoms at baseline. Women reported sexual intercourse in 36% of weeks. Proportions of sexually active weeks were higher among women with stress (43%) and depression (40%) compared with those without symptoms (35% and 35%, respectively; P values<0.001). Controlling for covariates, women with baseline stress symptoms had 1.6 times higher weekly odds of sexual intercourse compared with women without stress (adjusted odds ratio 1.6, confidence interval [1.1, 2.5]; P=0.04). Depression symptoms were not associated with sexual intercourse frequency in adjusted models.

CONCLUSIONS

Stress symptoms were positively associated with sexual intercourse frequency among these young women. Research and practice efforts are needed to identify effective sexual health promotion and risk-reduction strategies, including contraceptive education and counseling, in the context of mental health symptoms and unintended pregnancy.

The risk of unintended pregnancy among young women with mental health symptoms

Depression and stress have been linked with poor contraceptive behavior, but whether existing mental health symptoms influence women's subsequent risk of unintended pregnancy is unclear. We prospectively examined the effect of depression and stress symptoms on young women's pregnancy risk over one year. We used panel data from a longitudinal study of 992 U.S. women ages 18-20 years who reported a strong desire to avoid pregnancy. Weekly journal surveys measured relationship, contraceptive use and pregnancy outcomes. We examined 27,572 journal surveys from 940 women over the first study year. Our outcome was self-reported pregnancy. At baseline, we assessed moderate/severe depression (CESD-5) and stress (PSS-4) symptoms. We estimated the effect of baseline mental health symptoms on pregnancy risk with discrete-time, mixed-effects, proportional hazard models using logistic regression. At baseline, 24% and 23% of women reported moderate/severe depression and stress symptoms, respectively. Ten percent of young women not intending pregnancy became pregnant during the study. Rates of pregnancy were higher among women with baseline depression (14% versus 9%, p = 0.04) and stress (15% versus 9%, p = 0.03) compared to women without symptoms. In multivariable models, the risk of pregnancy was 1.6 times higher among women with stress symptoms compared to those without stress (aRR 1.6, CI 1.1,2.7). Women with co-occurring stress and depression symptoms had over twice the risk of pregnancy (aRR 2.1, CI 1.1,3.8) compared to those without symptoms. Among women without a prior pregnancy, having co-occurring stress and depression symptoms was the strongest predictor of subsequent pregnancy (aRR 2.3, CI 1.2,4.3), while stress alone was the strongest predictor among women with a prior pregnancy (aRR 3.0, CI 1.1,8.8). Depression symptoms were not independently associated with young women's pregnancy risk. In conclusion, stress, and especially co-occurring stress and depression symptoms, consistently and adversely influenced these young women's risk of unintended pregnancy over one year.