Impacts of multisectoral cash plus programs after four years in an urban informal settlement: Adolescent Girls Initiative–Kenya (AGI–K) randomized trial

Document Type

Article (peer-reviewed)

Publication Date



Background: The vast majority of adolescent births occur in low- and middle-income countries and are associated with negative outcomes for both the mother and her child. A multitude of risk factors may explain why few programs have been successful in delaying childbearing and suggest that multisectoral interventions may be necessary. This study examines the longer-term impact of a two-year (2015–17) multisectoral program on early sexual debut and fertility in an urban informal settlement in Kenya. Methods: The study used a randomized trial design, longitudinally following 2,075 girls 11–14 years old in 2015 until 2019. The interventions included community dialogues on unequal gender norms and their consequences (violence prevention), a conditional cash transfer (education), health and life skills training (health), and financial literacy training and savings activities (wealth). Girls were randomized to one of four study arms: 1) violence prevention only (V-only); 2) V-only and education (VE); 3) VE and health (VEH); or 4) all four interventions (VEHW). We used ANCOVA to estimate intent-to-treat (ITT) impacts of each study arm and of pooled study arms VE, VEH, and VEHW relative to the V-only arm, on primary outcomes of fertility and herpes simplex virus-2 (HSV-2) infection, and secondary outcomes of education, health knowledge, and wealth creation. Post-hoc analysis was carried out on older girls who were 13–14-years-old at baseline. In 2018, in the VEHW arm, in-depth qualitative evaluation were carried out with adolescent girls, their parents, school staff, mentors, community conversation facilitators, and community gatekeepers. The trial is registered at ISRCTN: ISRCTN77455458. Results: At endline in the V-only study arm, 21.0 percent of girls reported having had sex, 7.7 percent having ever been pregnant and 6.6 percent having ever given birth, with higher rates for the older subsample at 32.5 percent, 11.8 percent, and 10.1 percent, respectively. In the full sample, ever having given birth was reduced by 2.3 percentage points (pp) in the VE and VEHW study arms, significant at 10 percent. For the older subsample there were larger and significant reductions in the percent ever having had sex (8.2 pp), HSV-2 prevalence (7.5 pp) and HSV-2 incidence (5.6 pp) in the VE arm. Two years after the end of the interventions, girls continued to have increased schooling, sexual and reproductive health knowledge, and improved financial savings behaviors. Qualitatively, respondents reported that girls were likely to have sex as a result of child sexual exploitation, peer pressure or influence from the media, as well as for sexual adventure and as a mark of maturity. Conclusion: This study demonstrates that multisectoral cash plus interventions targeting the community and household level, combined with interventions in the education, health, and wealth-creation sectors that directly target individual girls in early adolescence, generate protective factors against early pregnancy during adolescence. Such interventions, therefore, potentially have beneficial impacts on the longer-term health and economic outcomes of girls residing in impoverished settings.