Education, Health Knowledge & Child Health in Ghana

Catherine N. Stiff, Brown University

This paper examines the relationships between education, “health knowledge” (beliefs about disease causation), and child health behaviors and outcomes in Ghana, West Africa. Data are from a representative sample of 2500 Ghanaian men and women interviewed in mid-2002. Preliminary fieldwork in coastal Ghana in 2001 highlighted an apparent disconnect between knowledge of disease etiology on the one hand, and realized health behaviors on the other. Respondents expressed high levels of health knowledge (e.g., reported hygiene- and contagion-related causes of child illness), but practiced relatively low levels of household hygiene. Although child survival has generally increased in sub-Saharan Africa over the past several decades, the region continues to suffer from comparatively high levels of infant and child mortality. Moreover, the experience of child illness and death is not uniformly distributed throughout the sub-Saharan region; large differentials in infant and child morbidity and mortality occur by demographic, social, economic, and geographic characteristics. For example, infant and child mortality is generally higher in rural areas than in urban, among the uneducated than among the educated, and among the poor than among the wealthy. This research is aimed at exploring some of the determinants of child health in Ghana. More specifically, the work examines – at the household level – the link between knowledge and subsequent health-related practices. Theories of mortality change and the determinants of child health inform this research design. Preston and Haines (1991), in their analysis of child mortality in the U.S. in the late nineteenth century, identified health “know how” – acceptance of the germ theory of disease causality and associated hygienic practices – as an important determinant of mortality decline. Moreover, Preston (1985) theorized that health knowledge was a product of formal education. Similarly, Caldwell (1979), in research in Nigeria, hypothesized that maternal education has an independent and direct effect on child health, and identified health knowledge as one causal mechanism through which education operates. Yet despite theorists’ emphasis on the importance of health beliefs in morbidity and mortality research, empirical research on this topic in Ghana is limited. Recent qualitative research in Ghana (Wyllie 1994) and elsewhere (Pebley et al. 1999, McLennan 1998, Azevedo et al. 1991) reiterates the importance of health beliefs in child survival and suggests the continued existence of traditional beliefs in non-biomedical causes of illness. Toward exploring health knowledge as an intermediate variable in the education-child survival relationship and investigating health knowledge in a representative sample, this study includes both a population-based survey of 1300 households in coastal Ghana and qualitative research (to be conducted in January 2003) on beliefs about child illnesses. This region manifests the usual range of major childhood diseases and has also been undergoing significant economic and social change. The results of the household-based survey are emphasized here. While large-scale surveys such as the Demographic and Health Survey (DHS) are good data sources for quantitative analysis of some socioeconomic determinants of infant and child morbidity and mortality, these sources lack measures of health knowledge. Our 2002 Ghana Population & Environment Survey was intended to fill this empirical gap. In addition to information on child morbidity and mortality and health-related behaviors such as personal and household hygiene, we also collected information on attitudes and behaviors with respect to three serious child illnesses in Ghana – malaria, diarrhoeal disease, and acute respiratory infection (ARI) – including symptoms, causes, prevention, and treatment.