Added: Kristofer Salyer - Date: 05.11.2021 02:36 - Views: 45912 - Clicks: 3327
Try out PMC Labs and tell us what you think. Learn More. Past studies on the influence of sexual activity on contraceptive behaviours are inconclusive, relying heavily on cross-sectional data sources. We used a population-based longitudinal sample of young women in a Michigan county to evaluate weekly associations between sexual activity and contraceptive use.
We used multinomial logistic regression ing for correlations within partnerships and women, assessing effects of woman-level, partnership-level, and week-level sexual activity on contraceptive use. Relative to use of least effective methods, weekly sexual activity was associated with condom aRRR For non-barrier methods, partnership-level and woman-level effects were similar to week-level effects.
For condoms, there was no ificant woman-level effect.
Evidence of immediate effects of sexual activity on contraceptive use within a single week highlights the importance of longitudinal data. These dynamics may be diluted or missed altogether when relying on cross-sectional data approaches that compare groups of individuals.
Young adults in the United States are the demographic that experiences the highest unintended pregnancy rates and the greatest share of unintended births. One in three unintended births in the United States is to a woman aged 20—24 Mosher et al. Almost nine out of ten women ages 20—24 who are at risk of an unintended pregnancy are current contraceptive users Mosher et al. Understanding when and why young women use specific contraceptive methods can help inform strategies to improve reproductive autonomy in this age group and prevent unintended pregnancies.
Sporadic patterns of sexual activity, not uncommon in this stage of the life course, may be an important factor that influences contraceptive use. This dynamic, however, has not been thoroughly examined in the research literature. Infrequent sexual activity is one of the main reasons that women who are at risk of unintended pregnancy cite for not using contraception Frost et al. While the association between more frequent sexual activity and increased use of any contraceptive method has been substantiated in national cross-sectional surveys Frost et al.
Several studies that have assessed use of specific types of contraceptive methods have also found that more frequent sexual activity is associated with use of more effective methods Frost and Darroch ; Kusunoki and Upchurch ; Wildsmith et al. Interpretation and comparison of these studies can be difficult, as the measures vary widely, particularly with respect to the amount of time over which measures of sexual activity and contraceptive use are summarised.
Measures used in prior research often summarise behaviours over the course of many months rather than assessing short-term associations over days or weeks, a timeframe in which women likely make short-term and medium-term contraceptive decisions. Inconsistent findings in the literature also reflect the limitations of using cross-sectional data sources to study associations between patterns of sexual activity and contraceptive use.Megan Fox \u0026 Machine Gun Kelly Ask Each Other 40 Questions - The Couples Quiz - GQ
For behaviours that change over time, analyses of cross-sectional data rely on the assumption that comparisons at a single point in time between individuals with different behaviours will approximate, on average, the associations for individuals as their behaviours vary over time.
This assumption may not be accurate for sexual activity and contraceptive behaviours. A longitudinal study of sexual health risk behaviours found that across multiple sexual health outcomes, including sex with a high risk partner, partner risk communication, alcohol use before sex, and condom use, there was substantially more variance in these outcomes within individuals over time than between individuals Cooper This preliminary evidence suggests that the variation in sexual activity and contraceptive behaviours within individuals over time may be substantial and not adequately captured in cross-sectional data sources.
The first objective of this study is to examine weekly associations between sexual activity and those contraceptive behaviours. According to multiple theories of health behaviour, perceived risk or perceived susceptibility to a health outcome is important in shaping health behaviours Brewer and Rimer A woman may therefore adjust her contraceptive behaviour according to her level of sexual activity because of a change in perceived risk of unintended pregnancy, an association that has been theorized and substantiated in prior research DuRant, Jay et al. Thus, our first hypothesis is that increased weekly sexual activity will be associated with use of more effective and longer acting methods.
The second objective of this study is to compare the associations between contraceptive use and sexual activity across different levels of measurement. Specifically, we will compare the associations between i week-level sexual activity and immediate contraceptive use; ii partnership-level sexual activity and contraceptive use across all weeks in a partnership; and iii women-level sexual activity and contraceptive use across all partnerships. The independent effect at each of these three levels will be assessed to identify the extent to which the approach taken in cross-sectional research woman-level associations reflects the shorter-term associations that occur over time for individual women week-level associationswhile also taking into variation in behaviours between partnerships partner-level associations.
Our second hypothesis is that as the data are summarised across longer timeframes effect sizes will be attenuated, with the greatest amount of information in week-level associations and the least information in woman-level associations.
Data for these secondary analyses come from the Relationship Dynamics and Social Life RDSL study, a longitudinal study with the broad goal of assessing the dynamics of unintended pregnancy through frequent measurement of pregnancy intentions, contraceptive use, and relationship characteristics Barber et al. The population-based sample of women ages 18—19 at baseline was randomly selected from a racially and socioeconomically diverse county in Michigan.
The baseline survey lasted approximately one hour and assessed participant sociodemographic characteristics, relationship experiences, and sexual and reproductive history. Participants were asked weekly about sexual activity, pregnancy intentions, contraceptive use, and pregnancy, with other questions asked only when applicable. Additional detail about the implementation of the study has been published elsewhere Barber et al. We conducted analyses of contraceptive method use for women ever sexually active in weeks where they were with a partner, were not pregnant, and indicated clearly not wanting to become pregnant.
The analytic sample included a total of women and 24, weeks. There were no missing data for potential confounders except for partner fertility desires, which were missing in 0.
Missing data were not imputed due to the very low frequency of missingness. Each week, women were asked whether they had vaginal sexual intercourse in the prior week. The primary exposure for this analysis is a dichotomous measure of weekly sexual activity. We also calculated a summary measure of partnership-level sexual activity, defined as the proportion of weeks sexually active among all weeks in a specific partnership. Finally, we calculated a measure of woman-level sexual activity, the proportion of weeks sexually active among all weeks in which women had a partner during the study.
Including these summary measures of sexual activity in models permits assessment of the effect of weekly sexual activity on contraceptive use independently of the effect of the overall level of sexual activity within a partnership and for an individual woman. This analytic approach is described in greater detail as part of our statistical methods.
This approach has been detailed in the statistical literature Begg and Paridesand has been applied to other longitudinal analyses of other sexual health behaviours Weir and Latkin ; Harvey et al. Contraceptive use was ascertained every week by asking women if they used anything to prevent pregnancy.
Those who answered affirmatively were then asked sequentially about specific non-coital methods prior to asking about sexual activity and coital methods. In weeks when women reported sexual activity, they were asked whether they had used a condom that week. Methods were grouped according to effectiveness Trusselland weeks were classified into four according to the most effective method used that week: i least effective methods, such as withdrawal, or no method referred to as least effective methods ; ii condoms; iii pill, patch, and ring referred to as pill ; and iv highly effective methods, including the injectable, IUD, implant, and male and female sterilisation.
For the purpose of descriptive statistics, we also summarized contraceptive use at the partnership and woman levels, calculating the percentage of analytic study weeks during which women reported using each type of method overall and within each partnership. We considered a series of potential confounders that we included in adjusted models as indicated.
Potential sexual and reproductive health history confounders were also assessed at baseline and included first sex by age 15, multiple sexual partners prior to enrolment, and any pregnancies prior to enrolment. Partnership qualities and partner sociodemographic characteristics were also assessed for potential confounding. Relationships were typified using a time-varying variable according to seriousness and time intensity into six : casual, nonexclusive dating, long distance, exclusive dating, cohabiting, and married or engaged Kusunoki We used a time-varying measure of whether the relationship had a prior breakup and reconciliation over the course of the study to assess relationship stability.Do girls also want sex as boys want
We considered a time-varying measure of total relationship duration with a specific partner. For partnerships that were ongoing at the start of the study, women were asked when they began this relationship in order to calculate duration.
We also included a time-varying measure of whether a participant ever had a pregnancy with her current partner. Characteristics of partners included partner age difference where a partner was three or more years older than a participant, education differential in which a partner was reported to have more education than the participant, and whether a partner had children from a relationship.
Finally, we constructed a weekly measure of whether perceived partner fertility desires were strongly anti-conception or not. We calculated descriptive statistics at the woman level for individual-level characteristics and at the week level for partnership-level characteristics. The outcome and exposure variables were summarised at the woman level, partnership level, and week level.
For contraceptive outcomes at the woman and partnership levels we calculated the mean percentage of weeks that each contraceptive method was used across women and partnerships respectively, with standard deviations expressed in percentage points. To decide whether to include candidate confounders in adjusted models, we first examined the association between each confounder with weekly measures of sexual activity and contraceptive use. As needed, we then considered associations between confounders and partnership-level and woman-level measures of sexual activity.
Woman-level candidate confounders that were associated with woman-level measures of sexual activity were included in adjusted models, as were candidate confounders associated with partnership-level or week-level sexual activity. Relationship duration was modelled as a linear spline with knot at one year in all models to allow different associations during and after the first year of the relationship. Multicollinearity of the adjusted model was assessed by estimating variance inflation factors VIFswhich averaged 1.
To estimate associations with specific contraceptive method use, we fit multinomial logistic regression models with robust standard errors to for the clustering of weekly data. To for clustering of partnerships within women and of weeks within partnerships, we estimated parameter confidence intervals using bootstrapped models with clustered resampling at the level of women in the sample. We preserved all partnerships and weeks for individual women during resampling and estimated each model with 1, bootstrap replications. Bias corrected and accelerated BCa confidence intervals were estimated for each model.
This approach of bootstrapping and estimating the BCa confidence intervals provides more accurate estimates of the standard errors of the regression coefficients and corrects the variance for the correlation among weeks within partnerships and partnerships within women Carpenter and Bithell These centered variables are calculated by subtracting, for example, the overall proportion of weeks sexually active in a specific partnership from the dichotomous measure of weekly sexual activity, or by subtracting the woman-level proportion of weeks sexually active from the partnership-level proportion of weeks sexually active:.
In this multinomial logistic regression equation, the log of the relative risk ratio for use of method k e. For the model reported in tables, partnership-level and woman-level sexual activity measures were scaled by a factor of ten, so that estimates reflect the effect of a ten percentage point increase in the proportion of weeks sexually active at each level. Models were also estimated with unscaled measures of partnership-level and woman-level sexual activity. Using from these unscaled models we compared woman-level effects and partnership-level effects to the week-level effects using postestimation Wald tests.
All analyses were conducted using StataMP Women reported data on a total of 2, partners with an average of 2.
The sample was socially and demographically diverse Table 1. Mean high school grade point average GPA was 3. One third of weeks were spent in exclusive dating relationships.Michigan girls wanting sex
email: [email protected] - phone:(177) 898-3489 x 7338
Horny local girls in MICHIGAN