Dissertations and Theses

Date of Degree


Document Type


Degree Name

Doctor of Philosophy (Ph.D.)


Epidemiology and Biostatistics


Heidi Jones

Committee Members

Elizabeth Kelvin

Mary Schooling

Eva Goldfarb

Subject Categories

Epidemiology | Public Health | Public Health Education and Promotion


School-based sex education, sexual and reproductive health, adolescent health, epidemiology


Background: There are significant sexual and reproductive health disparities in the United States (US). A significant proportion of sexual health disparities among adolescents is likely due in-part to inadequate school-based sex education. Sex education encourages sustainable and informed sexual behavior and has the potential reach adolescents prior to sexual debut and throughout the years at highest risk for adverse sexual health outcomes. However, the way in which sex education is defined and operationalized, as described in the literature, varies substantially, which may lead to wide variation in implementation in schools and this might explain the disparities seen. State-level sex education policy, which has the potential to standardize sex education, but impact studies at the state-level have focused solely on abstinence-only policies. In response to these challenges, this dissertation aims to 1) summarize the content and structure of evaluations of school-based sex education interventions and develop a new method to classify comprehensiveness of sex education interventions, 2) statistically evaluate whether school-based sex education programs evaluated in the published literature that are more comprehensive are more effective at improving health outcomes compared to those that are less comprehensive, and 3) evaluate the immediate impact of the first state-wide comprehensive sex education policy that was enacted in 2016 in California, titled the California Healthy Youth Act (CHYA).

Methods: For Aim 1, a search was completed using ERIC and PsycINFO on the Ebsco platform, as well as MEDLINE on Web of Science, to identify studies that meet the following criteria: (1) US-based quantitative-only studies, (2) published from 2000-2020 in a peer-reviewed journal, (3) setting exclusively K-12 school-based, (4) an evaluation study, and (5) included a control group. Three comprehensiveness scales were created based on common approaches to scale development (additive, additive grouping, meaningful approach) to summarize differences in intervention by topics covered, based on the 2012 National Sexuality Education Standards. In Aim 2 we included studies identified in Aim 1 that evaluated the following as primary outcomes: (1) sexual debut, (2) condom use at last sex, or (3) interpersonal violence victimization. Heterogeneity was assessed with forest plots. Random effects models were run, and pooled odds ratios (OR) with 95% confidence intervals (CIs) were estimated using summary statistics from the included studies. Publication bias was evaluated with a funnel plot and Egger’s test. To evaluate differences by comprehensiveness, studies were grouped and compared by the meaningful approach scale category and separate random effect models were run by category to result in pooled estimates. For Aim 3, a segmented interrupted time series was used to evaluate the effect of CHYA on state-level reportable outcomes among teens: chlamydia, gonorrhea, syphilis, birth, and high school graduation rates. As a sensitivity analysis, all outcomes were lagged at one year to allow for behaviors to change.

Results: In Aim 1, we found that very few interventions were evaluated among elementary school students, were most often implemented by regular classroom teachers, and consisted of fewer than 30 sessions in total. Identity and personal safety were the two topics least covered, while pregnancy and reproduction and STIs/HIV were most frequently covered. All studies included in the systematic review reported at least one positive outcome or change due to the intervention compared to the study defined control group. When stratified by the comprehensiveness scales, the additive or additive grouping scale approaches (which count the number of topics covered) showed no clear pattern while the meaningful grouping scale approach suggested that highly comprehensive studies tend to report, on average, a greater number of measured outcomes with a positive improvement due to the intervention compared to the control group, compared to those classified as minimally comprehensive. In Aim 2, among the studies that evaluated sexual debut (n=14), the pooled estimate showed that intervention groups were statistically significantly successful at delaying sexual debut among target students compared to the control (OR: 0.79, 95% CI: 0.70, 0.89), and this association held among those considered highly comprehensive. There was no statistical significance among those considered moderately or minimally comprehensive, and the differences between the three categories were not statistically significant, which could be due to low sample size. Among the studies examining condom use (n=7), the pooled estimate showed that the intervention groups were more likely to report using a condom at last sex, although not statistically significant, which remained across all categories of comprehensiveness. Among the studies that examined IPV (n=3), results were null, with too few studies available to compare across comprehensiveness categories. In Aim 3, there was a significant reduction in teen birth rate over time; however, when effects were lagged a year to allow time for the policy to ‘”reach” adolescents, that effect was only borderline significant. There were two unexpected findings: a significant increase in chlamydia among teens and reduction in high school graduation rates. None of the other outcomes were significantly associated with CHYA.

Discussion: Aim 1 is one of the largest systematic reviews of school-based sex education interventions conducted to date, focusing on quantifying components and characteristics of interventions to better understand the heterogeneity regarding comprehensive sex education interventions. When comparing by comprehensiveness in Aim 2, results suggested that highly comprehensive interventions do show a protective effect for sexual debut that is not present among moderately or minimally comprehensive interventions, though due to minimal sample size in condom use at last sex and IPV, it is unclear how consistent or accurate comprehensiveness is capturing the studies included. In Aim 3, preliminary ecologic analyses did not find any large, state-wide significant improvements in reportable teen STI rates, teen births, or high school graduation rates associated with implementation of the CHYA. It is possible that despite the state-wide mandate, there was regional variation in implementation, and other policies may have impacted secular trends. Future analysis might explore the association at smaller geographic units, such as the district-level. This dissertation project confirms the heterogeneity present in evaluated school-based sex education interventions. While there may not have been enough time for CHYA to be effective, the policy requirements and implementation might also not be sufficient or consistent enough to make an impact at the state-level. Considering how few interactions with students are required under the CHYA policy as compared to the interventions that were evaluated from Aims 1 and 2, there seems to be a lack of consistent approaches seen across school-based sex education interventions. This research should inform dedicated comparative effectiveness research to aid in state-level sex education policies.



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.