abstract
- This study estimates the relative importance to child health (indicated by weight and height for age) of economic development level [gross domestic product (GDP) converted to international dollars using purchasing power parity (PPP) rates: GDP-PPP], household wealth and maternal education and examines the modifying influence of national contexts on these estimates. It uses information collected from mothers aged 15-49-years participating in Demographic Health Surveys (DHS) conducted in 42 developing countries. In multilevel regression models, the three study variables exhibited strong independent associations with child health: GDP-PPP accounted for the largest amount of unique variation, followed by maternal education and household wealth. There was also substantial overlap (shared variance) between maternal education and the other two study variables. The regressions of child health on household wealth and maternal education exhibited substantial cross-national variation in both strength and form of association. Although higher education levels were associated with disproportionately greater returns to child health, the pattern for household wealth was erratic: in many countries there were diminishing returns to child health at higher levels of household wealth. We conclude that there are inextricable links among different strategies for improving child health and that policy planners, associating benefits with these strategies, must take into account the strong moderating impact of national context.