The effect of education programs on nutrition outcomes is difficult to assess because programs often have poor baseline data or nutrition outcomes are not evaluated (174, 182). Studies that used longitudinal analyses and “natural” experiments (e.g., before and after a national education policy) found that education was associated with reduced fertility (183, 184), and delayed early marriages and pregnancies (184–187). The impact was more significant for higher levels of education (185). However, 1 study in Malawi identified negative associations between education and timing of first birth, although these findings were largely not statistically significant (188). Secondary education for adolescents and women of reproductive age also showed no impact on women's empowerment (184), although it did show an impact on improved literacy and leadership (174). Educational interventions that provided conditional cash transfers (CCTs) and school feeding, as well as other forms of social protection to families of enrolled girls, were associated with greater school enrollment and attendance (189–191), improved test scores (189, 190), reduced gender gaps (192), and reduced hunger (190, 191).
  Community centers  ↑/NC knowledge about hygiene and sanitation, ↑ water quality, ↓ diarrheal morbidity  ↑/NC knowledge about hygiene and sanitation, ↑ water quality, ↓ diarrheal morbidity  ↑/NC knowledge about hygiene and sanitation, ↑/NC hand-washing, ↑ water quality, NC waste disposal, ↓ diarrheal morbidity  ↑/NC knowledge about hygiene and sanitation, ↑ water quality, ↓ diarrheal morbidity 
Our findings identified gaps and limitations in the evaluation, scope, targeting, and delivery platforms of nutrition interventions in low- and middle-income countries. First, the monitoring and evaluation of nutrition programs that reported on women's nutrition outcomes was generally inadequate. Many of the studies we identified included small-scale efficacy trials. Although there were many large-scale programs that targeted women and adolescent girls with nutrition-specific and nutrition-sensitive approaches, they lacked rigorous evaluation. Whether the evidence about women's outcomes was limited because they are not systematically measured or because they are not well reported is not clear. Negative results are often not published, and many evaluations of nutrition interventions that are conducted by the same groups responsible for implementing them are typically presented positively. This may have also skewed our findings. More intentional research-quality program evaluation, including of large-scale programs, would provide a stronger evidence base. Of the studies identified in this review, many reported on short-term findings such as changes in knowledge, dietary behaviors, and program coverage. They were limited in their ability to report clinical and anthropometric outcomes for women, the duration of those outcomes, and the feasibility of scaling up programs. There is also a need for systematic, long-term evaluations of interventions whose effects on nutrition outcomes are more distal (e.g., nutrition education compared with micronutrient supplementation). The effects of multisectoral interventions are even more complex to measure. However, frameworks exist to evaluate complex interventions (102) and could be utilized to evaluate the impact of interventions across the life course.