RESUMO
Though effective reproductive health interventions have been developed for adolescent girls and young women (AGYW) in sub-Saharan Africa, few have explored whether specific components of the interventions are responsible for observed changes in behaviour. Data for this longitudinal mediation analysis come from a quasi-experimental, sexual and reproductive health study conducted among AGYW (age 15-24) in Malawi. We assessed the extent to which the relationship between attending communication-specific small-group sessions and contraceptive use at one-year was mediated by contraceptive communication with partners at six months, using a bootstrapping procedure to estimate indirect effects. Of 358 participants, 44% attended communication-specific small-group sessions, 37% communicated with partners about contraception at six months, and 21% used non-barrier contraception at one-year. Participants who attended communication-specific small-group sessions had increased contraceptive communication with partners at six months (aOR = 1.48, 95% CI: 1.07-2.38) and increased non-barrier contraceptive use at one-year (aOR = 3.53, 95% CI: 1.86-6.69). The relationship between attending communication-specific small-group sessions and non-barrier contraceptive use at one-year was partially mediated by contraceptive communication with partners at six months (indirect effect = 0.04, 95% CI: 0.01-0.07). Our results suggest that contraceptive communication with partners is modifiable through interventions and important for AGYW non-barrier contraceptive uptake.