Background Primary school children in low-income countries are at risk of many diseases and poor health affects attendance, cognition and ability to learn. Developing school health and nutrition strategies has been extensively highlighted as a global priority, with a particular focus on complex programme design. However, such programmes are relatively untested in low-income settings.

Method We implemented a complex school health and nutrition programme in two schools in Western Kenya over 3 years. There were numerous elements covering health policy, skills-based health education, infrastructure and disease prevention. A local non-governmental organisation, with involvement from local government and the community, performed programme implementation. Height-for-age, weight-for-age,height-for-weight, anaemia prevalence, academic performance and school attendance were the primary outcome measures.

Results The programme improved nutrition, academic performance and anaemia prevalence. The number of underweight children fell from 20% to 11% (OR 0.51 95% CI 0.39 to 0.68 p=<0.01) and stunting prevalence fell from 29.9% to 20% (OR 0.59 95% CI 0.50 to 0.68 p=<0.01). Academic performance improved with a 74% reduction in odds of failing assessments (OR 0.26 95% CI 0.22 to 0.29 p=<0.01). Anaemia prevalence fell from 17.2% to 11%. The programme showed an increase in low body mass index prevalence and no effect on school attendance, the reasons for which are unclear.

Discussion These results are encouraging and demonstrate that complex schools health programmes can lead to positive gains in health, nutrition and importantly academic performance. There is a need for further evaluation of comprehensive school health interventions in poor communities.