A longitudinal assessment of the diet and growth of malnourished children participating in nutrition rehabilitation centres in Accra, Ghana Esi K Colecraft 1 , Grace S Marquis 1, *, Alfred A Bartolucci 2 , LeaVonne Pulley 3 , W Bruce Owusu 4 and H Michael Maetz 5 1 Department of Food Science and Human Nutrition, 1127 HNSB, Iowa State University, Ames, IA 50014, USA: 2 Department of Biostatistics, School of Public Health, University of Alabama at Birmingham, USA: 3 Department of Health Behaviour, School of Public Health, University of Alabama at Birmingham, USA: 4 Department of Nutrition and Food Science, University of Ghana, Legon, Ghana: 5 Department of Epidemiology, School of Public Health, University of Alabama at Birmingham, USA Submitted 19 August 2003: Accepted 13 October 2003 Abstract Objective: To examine the adoption of feeding recommendations among caregivers of children recuperating from malnutrition and assess the determinants of growth of children attending a nutrition rehabilitation centre (NRC) in Accra, Ghana. Design: Longitudinal study in which attendance and maternal programme participation were recorded daily and children’s anthropometry and dietary intake were measured at four time points (admission, interim, exit, post-exit) at the NRC and participants’ homes. Setting: NRCs at four polyclinics and participants’ homes in Accra, Ghana. Subjects: One hundred and eight caregivers and their 116 children referred to an NRC between November 1999 and July 2000. Results: Most caregivers attended the NRC sporadically (effective length of stay was 1.4 ^ 0.1 months). Use of NRC-promoted foods in the home after discharge was low due to inaccessibility of the food items, lack of preparation knowledge or money, child preferences and the common practice of purchasing ready-to-eat foods. Although there were significant increases in children’s weight-for-age (P ¼ 0.048) and weight-for-height (P ¼ 0.002) Z-scores between enrolment and discharge, most children discontinued programme participation before adequate recuperation. Conclusions: The NRC education did not address the use of street foods for child feeding and was unsuccessful in changing in-home feeding behaviour. The prominence of street foods in children’s diets warrants re-evaluation of the NRC’s educational approaches to enhance their responsiveness to caregivers’ needs and effectiveness for the continued recuperation of malnourished children at home. NRC feeding strategies need improvement to ensure adequate provision of energy and nutrients to support catch-up growth in children. Keywords Protein – energy malnutrition Feeding programmes Child anthropometry Dietary intake Child growth Complementary feeding Koko Ghana Protein – energy malnutrition (PEM) is a problem of public health significance in most developing countries and is associated with increased childhood morbidity and mortality 1,2 . In Ghana, about 40% of all child deaths beyond infancy are due to PEM, making it the greatest single cause of child mortality 3 . With approximately one- quarter of Ghanaian children either underweight or stunted 4 , Ghana is classified in the medium to high range of severity of malnutrition in a population 5 . The debilitating effects of malnutrition have prompted initiatives to improve children’s nutritional well-being. In the mid-1950s, Bengoa proposed the nutrition rehabilita- tion centre (NRC) as a strategy for combating childhood PEM 6 . This system was piloted in Ghana in the 1970s 7 and later integrated into primary healthcare facilities (poly- clinics) throughout the country. Presently, the centres are operated as day-care facilities where malnourished children attend daily with their caregivers. The pro- gramme objectives are to recuperate malnourished children through supplemental feeding and to educate caregivers on optimal feeding practices through partici- pation in food preparation activities. The NRC programme in Ghana has not been system- atically evaluated to determine its effectiveness. Evalu- ations of NRC programmes in other settings have reported mixed results 8–11 . Many of these studies are now dated q The Authors 2003 *Corresponding author: Email gmarquis@iastate.edu Public Health Nutrition: 7(4), 487–494 DOI: 10.1079/PHN2003553