Improving Menstrual Health Through BCC Among Bangladeshi University Students

Overview

This study evaluated the effectiveness of a Behavior Change Communication (BCC) intervention in improving menstrual health among female university students in Bangladesh. The intervention focused on three outcomes: dysmenorrhea, premenstrual syndrome (PMS), and menstrual irregularity (MI), promoting a non-pharmacological, sustainable approach through behavioral and lifestyle changes. Three structured educational sessions on menstrual health, healthy diet, and physical activity were delivered by trained female educators between May 20 and June 30, 2023, at Patuakhali Science and Technology University, Barishal University, and Khulna University. Participation was voluntary. Reinforcement follow-ups were conducted every two months for six months. A quasi-experimental design was employed. Post-intervention data were collected from February to March 2024. Based on sample size calculations assuming a 30% reduction in disorder prevalence, 498 students (249 per group) were initially enrolled. After exclusions, 234 intervention and 238 control participants were analyzed for dysmenorrhea; 228 per group for PMS and MI. Propensity score matching (1:1, caliper 0.01, no replacement) yielded 98 matched participants per group for final analysis. Eligible participants were female students aged ≥19 years, residing in university dormitories. Data collection used validated Bengali-language, interviewer-administered questionnaires, based on relevant literature and menstrual health IEC materials. Trained female enumerators ensured data quality and participant comfort. Primary outcomes were: (1) Dysmenorrhea, measured by the Andersch and Milsom score (dichotomized Yes/No); (2) PMS, assessed using the Premenstrual Symptoms Screening Tool, with no/mild PMS and moderate to severe PMS/PMDD as the outcome; (3) MI, defined as self-reported menstrual cycles \<21 or \>35 days. The treatment variable was BCC exposure (Yes/No). Covariates included physical activity, BMI, dietary diversity (DDS ≥5), food cravings, breakfast skipping, sleep duration, caffeine use, bedtime, family history of menstrual disorders, age at menarche, marital status, residence, and parental education and occupation. Baseline differences were assessed using descriptive statistics, chi-square, and t-tests. Conditional logistic regression estimated intervention effects on matched data. Sensitivity analyses (ATE, ATT, balance diagnostics, Bayesian logistic regression) supported the robustness of results.

This study aimed to assess the effectiveness of a structured Behavior Change Communication (BCC) intervention in improving menstrual health among female university students in Bangladesh. Specifically, it targeted three key menstrual health outcomes: dysmenorrhea, premenstrual syndrome (PMS), and menstrual irregularity (MI). The overarching goal was to promote a sustainable, non-pharmacological strategy for managing common menstrual disorders by fostering positive lifestyle and behavioral changes-such as healthier eating habits, increased physical activity, and greater awareness about menstrual health. The study employed a quasi-experimental design involved two groups: an intervention group that received BCC module and a control group that did not. Participants were female undergraduate students at three public universities in Bangladesh i.e. Patuakhali Science and Technology University (PSTU), Barishal University (BU), and Khulna University (KU). Ethical approval was obtained from the Shahjalal University of Science and Technology Research Ethics Board (Ref. No. AST/002/258), and all participants provided written informed consent. The BCC intervention comprised three interactive educational sessions delivered between May 20 and June 30, 2023, by trained female facilitators. The first session addressed menstrual disorders and associated risk factors, the second focused on healthy dietary practices, and the third emphasized physical activity (including yoga) and other relevant lifestyle modifications. To enhance retention and adherence, follow-up visits were conducted bi-monthly for six months after the intervention to reinforce key messages and address participant questions or barriers to behavior change. Post-intervention data collection was carried out between February and March 2024. Participants were eligible if they were female students aged 19 years or older, residing full-time in university dormitories, and not currently taking hormonal treatments or diagnosed with chronic illnesses affecting menstruation. Recruitment occurred through record book of student dormitories at each university and student networks. Informed written consent was obtained prior to participation. Data were collected through interviewer-administered questionnaires, adapted and validated in Bengali for cultural relevance and comprehension. Instruments were based on existing literature and government-endorsed information, education, and communication (IEC) materials on menstrual health. All interviews were conducted by trained female enumerators to ensure a comfortable and respectful environment for participants. The sample size was calculated to detect a 30% reduction in the prevalence of menstrual disorders, with 80% statistical power and a 95% confidence interval, accounting for an anticipated dropout rate. Initially, 498 participants were recruited-249 each in the intervention and control groups. After excluding individuals with missing follow-up data or incomplete attendance, 234 participants in the intervention group and 238 in the control group were included in the analysis for dysmenorrhea. For PMS and MI outcomes, 228 participants per group had valid data.To control for potential confounding, 1:1 propensity score matching (PSM) was conducted using nearest-neighbor matching with a caliper of 0.01 and without replacement. Matching was based on baseline characteristics to ensure statistical equivalence between groups. After matching, a balanced analytical sample of 196 participants (98 in each group) was retained for all outcome analyses. Among all the primary outcomes, dysmenorrhea was assessed using the Andersch and Milsom four-point scale and dichotomized as no pain (Grade 0) vs. any pain (Grades 1 - 3). PMS was evaluated by the premenstrual symptoms screening tool (PSST) and recoded into a binary variable: No/Mild PMS (code = 0) and Moderate to Severe PMS or premenstrual dysphoric disorder (PMDD) (code = 1). Additionally, Menstrual irregularity was determined based on self-reported cycle length and coded as irregular (1) if \<21 or \>35 days, and not irregular (0) if within the 21-35 day range. The primary exposure variable was participation in the BCC intervention, coded as yes (intervention group) or no (control group). The covariates included physical activity, obesity, dietary habits, lifestyle factors, and socio-demographic characteristics. Physical activity was classified according to WHO guidelines as: (i) sedentary: light-intensity activities; (ii) active: moderate-intensity activities for 150-300 minutes per week; and (iii) athlete: vigorous-intensity activities for 75-150 minutes per week. Obesity was assessed using body mass index (BMI, kg/m²) and categorized according to WHO Asian-specific cut-offs: (i) underweight (\<18.5), (ii) normal weight (18.5-22.9), (iii) overweight (23.0-27.5), and (iv) obese (\>27.5). Dietary information was evaluated through 24-hour dietary recall using FAO's Minimum Dietary Diversity for Women (MDD-W) scoring system, with scores ≥5 indicating adequate diversity was collected using a five-day 24-hour recall method, and dietary diversity score (DDS) was calculated following FAO guidelines. Lifestyle-related covariates included food cravings for high-fat or sweet foods, skipping breakfast, sleep duration, bedtime and caffeine intake. Socio-demographic variables included family history of menstrual disorders, age at menarche, marital status, and participants' residence as well as parental educational and occupational status. Data analysis was performed using Stata 17. The following steps were undertaken: 1. Descriptive statistics summarized baseline demographic and behavioral characteristics. Chi-square (χ²) tests were used for categorical variables, and independent-sample t-tests for continuous variables. 2. Propensity score matching (PSM) was conducted using logistic regression to estimate the probability of receiving the intervention, based on the identified covariates. Nearest-neighbor matching (1:1) with a caliper width of 0.01 (no replacement) was applied. Balance between groups post-matching was assessed using standardized mean differences (SMD), with values \<10% indicating adequate balance. 3. Primary analyses of intervention effects were conducted using conditional logistic regression on the matched sample to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for each outcome. 4. To confirm the robustness of the findings, bivariate and multivariate logistic regression analyses were also performed on the unmatched sample, adjusting for covariates. 5. Sensitivity analyses were carried out to assess the stability of the results, including: covariate balance diagnostics (before and after matching); estimation of Average Treatment Effect (ATE) and Average Treatment Effect on the Treated (ATT); Bayesian logistic regression, including log Bayes factor interpretation, to evaluate model fit and strength of evidence for intervention effects. However, All statistical tests were two-sided, with a significance level set at p \< 0.05.