Includes bibliographical references (leaves 72-73). / Christel House South Africa (CHSA) is an independent primary and secondary school that offers a comprehensive educational programme that includes two meals and snacks to children from disadvantaged areas known as the Cape Flats. Anthropometric measurements such as weight and height are widely recognised as a reliable way to assess growth and nutritional status of children. Underweight and wasting are indicators of recent and severe malnutrition while stunting is associated with poor socio-economic conditions. Academic performance is associated with many factors including genetics, socio-economic circumstances, parental education and ill-health. However, there is no convincing evidence supporting a direct link between improved academic performance and improvement in nutritional status. This study aims to evaluate the comprehensive programme that CHSA provides by assessing anthropometric measurements and academic performance as well as absenteeism among the learners over a two year period. In particular, to assess whether those learners who showed improved anthropometry also demonstrated improved academic performance and reduced absenteeism between 2002 and 2004. The study is a cross-sectional survey with a retrospective cohort analysis. Existing data, part of the school's routine data collection, were used. The data consisted of weight and height measurements, end-of-year grades for English and Mathematics, and absenteeism records, all for 2002 and 2004. After obtaining informed consent from parents and guardians, a sample of 175 learners was established and used for cross-sectional analyses. This represented 65% of all eligible learners in the sample population. The median age was 8.5 years. Sex and age specific anthropometric measurements, expressed as weight-for-age, height-for-age and weight-for-height Z-scores, were created using the Epi Nut software in EPilnf06. Due to age and height limits in data that can be entered in this software, a cohort sub-sample of 81 learners, median age 7.2 years, was created and used for cohort analyses. As a result of measurement error with regard to height measurements in 2002, height-for-age and weight-for-height outcomes for 2002 were rendered invalid and excluded from statistical analyses of height-for-age and weight-for-height Z-score change between 2002 and 2004. However, by creating a 'gain' variable that took account of the measurement error, selected comparisons could still be done. The study found a prevalence of underweight of 8% in 2002 and 1.7% in 2004 (n=175). The reduction in underweight was statistically significant (P<0.05). There was a significant improvement in weight-for-age, English performance and Mathematics performance between 2002 and 2004, both in the complete sample (n=175) and the cohort (n=81). These improvements were statistically significant for both sexes, except for weight-for-age among girls in the cohort. There was no significant reduction in absenteeism during this period, either in the complete sample or the cohort. For 2004, bi-variate analyses showed a modest but significant positive correlation between weight-for-age Z-score and English (r=0.21, P<0.05) and Math (r=0.25, P<0.05) performance respectively, and a significant modest negative correlation with absenteeism (r=-0.18, P<0.05) (n=175). These correlations were not apparent in 2002. Multivariate analyses using logistic regression showed no associations between weight-for-age Z-score gain or height-for-age Z-score gain and improvement in English, Math and absenteeism, respectively (n=81). A negative association (OR=0.3, 95% CI 0.11-0.76) was found between weight-for-height Z- score gain and improvement in absenteeism. All analyses were adjusted for sex, age and having failed a year in school between 2002 and 2004. No significant associations were found when assessing improvement in English, Math and absenteeism in relation to quartiles of change in weight-for-age Z-score. A modest positive correlation was found between reduction in absenteeism and improvement in English (r=0.3, P<0.05) (n=81). Limitations of the study include measurement error in height in 2002, leading to the exclusion of height-for-age and weight-for-height in most of the analyses, and eliminating stunting and wasting as indicators of malnutrition from the overall assessment. Secondly, as a result of the limitations for age and height in the EpiNut software, a cohort was created. Analysis indicated that those included in the cohort were significantly different 'from those not included in the cohort in terms of age and sex. In particular, selection bias caused the exclusion of older and taller girls from the cohort. As such, findings for the cohort may not be representative of the study population. In spite of its limitations, this study identified a number of significant findings in respect of academic performance and anthropometric outcomes, suggesting that the programmes as provided by Christel House South Africa benefit their learners. In view of future monitoring and evaluation it is recommended that a protocol for weight and height measurement taking is agreed on, that the same instruments are used consistently, and that these instruments are regularly calibrated. It is also recommended that any future research related to child growth and academic performance at CHSA include data on socio-economic circumstances and parental education. Lastly, it is recommended that CHSA makes an assessment of the quality and nutritional value of the food provided, to ensure that the meals and snacks are healthy, varied and nutritionally balanced.
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:uct/oai:localhost:11427/9332 |
| Date | January 2007 |
| Creators | Kamminga, Froukje |
| Contributors | London, Leslie |
| Publisher | University of Cape Town, Faculty of Health Sciences, Department of Public Health and Family Medicine |
| Source Sets | South African National ETD Portal |
| Language | English |
| Detected Language | English |
| Type | Master Thesis, Masters, MPH |
| Format | application/pdf |
Page generated in 0.003 seconds