Comparison of micronutrient-intake of lactating mothers from the Hlabisa district in KwaZulu-Natal using two different dietary intake methods
INTRODUCTION: The objective of this research study was to analyze previously collected dietary intake data using multiple 24-hour dietary recalls and semiquantitative food frequency questionnaires (FFQ’s) in a group of HIV-positive and HIV-negative breastfeeding women from a rural region in KwaZulu-Natal in order to compare the intake of selective micronutrients obtained with the two instruments. Identifying the pattern of food intake and the contribution of different foods to the micronutrient intake in this population group will contribute to possible recommendations aimed at dietary changes to improve dietary micronutrient intake. This study was designed as a sub-study of a longitudinal prospective cohort study and subjects (N=108) were lactating mothers enrolled in a cohort which investigated the combined effect of HIV-infection and breastfeeding on women’s nutritional status. METHOD: A locally constructed FFQ and 24h-recall were used to collect dietary intake data from 108 subjects on three occasions, (~6 weeks, 14- and 24-weeks post partum). Analysis was done using the Food Finder Program™2. Micronutrients under investigation were iron, zinc, copper, selenium, vitamin A, B6, C, D and E, thiamin, riboflavin and folic acid and were selected on their relevance in HIV (AIDS). Descriptive statistics was used to determine the consumption of food items as percentage of all food items consumed and to calculate mean, mode, median and range of serving sizes for the ten food items most frequently consumed (measured with the 24h and FFQ respectively). Data was not normally distributed (indicated by the paired t-test and confirmed with a RM ANOVA nonparametric test). The F-value was determined (using Wilcoxon matched pairs test) and the significance of the difference between the micronutrient intakes measured with the two instruments (p<0.05) calculated. To investigate the strength of the correlation between the two dietary intake measures, Spearman’s correlation coefficients were determined for the nutrients under investigation. The significance level for these measurements was 95% (p<0.05). RESULTS: Both methods identified maize meal and mahewu, bread, chicken, dried beans, cabbage, onion, bananas, oranges and green leaves as the foods most often consumed. Bread, dried beans, maas, pilchards, mango and green wild leaves were the foods that contributed the most to the micronutrients under investigation. Although maize meal (in the form of phutu or mahewu) was the food item most frequently consumed in large portions, it was not in the top ten food items for any micronutrient contribution, except for selenium. Correlation coefficients (unadjusted for energy) in this study were very poor, ranging from 0.038 for vitamin B12 up to 0.48 for iron. All correlations (except vitamin B12) were poor but significant (p<0.05). CONCLUSION: There was some agreement found in the type of foods most frequently consumed and their contribution to the micronutrient intake of this population group, when using three 24h-recalls and FFQ’s and therefore in describing the habitual food intake of the population group. There was however no agreement between the micronutrient intake measured with three 24h-recalls and three FFQ’s (p<0.05). Further analysis of the data and comparisons with the biochemical results reported in another study, is recommended.