Effect of Hygiene and Menstrual Hygiene Interventions on Learning and Psychosocial Wellbeing in Madagascar

Disentangling the Effects of Infrastructure, Sensitization, and Sanitary Pad Distribution on Health, Learning, and Psychosocial Wellbeing Among Adolescent Girls in Rural Madagascar: A Cluster-Randomized Trial

A previous study in schools in rural Madagascar found that a bundle of hygiene-focused interventions substantially improved girls' learning outcomes. This study disentangles the effects of physical infrastructure (latrines, handwashing basins) from other intervention components (peer sensitization and free sanitary pad distribution) in a cluster-randomized trial across 140 secondary schools in the Amoron'i Mania region. The trial originally measured learning and psychosocial wellbeing outcomes and was prospectively registered with the AEA RCT Registry (AEARCTR-0014423). The study team is now also collecting secondary health biomarker outcomes (e.g., salivary cortisol and blood-based C-reactive protein) and is registering those health biomarker endpoints with ClinicalTrials.gov.

140 secondary schools were randomly assigned to one of four arms: (1) control, (2) infrastructure only, (3) sensitization + free sanitary pads, or (4) the full combined package. Outcomes are assessed via girls' surveys, teacher/director surveys, and school-level surveys at baseline (Oct 2024), midline, and endline.

This registration draws on two overlapping analytic samples. The first is the full 140-school evaluation sample (35 schools/arm; approximately 3,080 girls, ~22/school), used for the cognitive/academic composite, self-reported psychosocial well-being (GAD-7, PHQ-9), and health outcomes that do not require biospecimen collection or laboratory processing - self-reported infectious disease symptoms, mid-upper-arm circumference (MUAC), and self-reported urinary tract infection. These are collected during the social-science endline survey (May-June 2026), concurrent with the parent trial's learning and psychosocial data collection. The second is a 80-school biomarker subsample (approximately 20 schools/arm; ~1,360 girls, ~17/school), used specifically for salivary cortisol and blood-based C-reactive protein (CRP) drawn from a venipuncture. These outcomes require biospecimen collection, cold-chain transport, and laboratory assay, and are therefore restricted to the subset of schools logistically accessible for that purpose. They are collected at a separate endline health/biomarker visit at a Basic Health Center (July-August 2026, ~19 months post-baseline).

This registration covers the biomarker outcomes component of the trial, with data collection and analysis completing by approximately December 2026; the broader parent Kilonga 2 trial (registered with the AEA RCT Registry, AEARCTR-0014423) has a longer overall completion date of December 2027, covering the full set of learning and psychosocial endpoints. The biospecimen collection protocol includes additional biomarkers, cortisol and C-reactive protein, as registered secondary outcomes.

This trial was registered with ClinicalTrials.gov retrospectively, after participant enrollment had already begun. At inception (October 2024), the study was designed and IRB-approved (Paris School of Economics IRB #2024-046, approved September 17, 2024) as a social science evaluation focused on learning outcomes, and was prospectively registered with the AEA RCT Registry (AEARCTR-0014423) on September 23, 2024. Registration with ClinicalTrials.gov was initiated upon expanding the study to include health-related biomarker secondary endpoints, in order to comply with ICMJE requirements for biomedical journal publication. The Pre-Analysis Plan covering these health endpoints will be timestamped on the Open Science Framework prior to the start of health data collection in July 2026.

Inclusion Criteria:

• Girl identified in the Kilonga 2 evaluation sample (randomly selected during the December 2024-January 2025 baseline from class registers across the 140 participating schools); participants were 11-20 years old at baseline enrollment and are approximately 12-21 years old at the time of endline/health-outcomes data collection (May-August 2026)
• Full evaluation sample (n~3,080; ~22 girls/school): all girls in the Kilonga 2 sample across the 140 schools are eligible for the cognitive, self-reported psychosocial, and non-biospecimen health outcomes, collected at the social-science endline.
• Biomarker subsample (n~1,360; ~17 girls/school): girls attending one of the 80 schools selected for the health biomarker subsample (logistically accessible for biospecimen collection and cold-chain maintenance) are additionally eligible for salivary cortisol and blood-based C-reactive protein outcomes.
• Informed consent obtained from parent/legal guardian, and assent obtained from the participant herself

Exclusion Criteria:

• Adolescents not meeting the inclusion criteria above
• Adolescents (or their parent/legal guardian) who refuse to participate or who withdraw consent/assent at any time
• Adolescents for whom sample collection is judged impossible or contraindicated at the time of collection (e.g., health status precluding collection)