Adaptation of locally available portion sizes for food frequency questionnaires in nutritional epidemiological studies: how much difference does it make?

Abstract

Introduction: There is need for employing strategies to minimize measurement errors while administering semi-quantitative FFQ. The current study was planned to adapt and standardize locally available portion sizes for semi-quantitative FFQ to improve its validity and document the process of standardization of portion sizes. Methodology: The study was conducted in 9 villages of the INCLEN-SOMAARTH DDESS (Demographic, Development and Environmental Surveillance Site), Palwal district, Haryana, India. The subjects in these nine villages are part of a cohort study to assess the interaction between societal and household factors with food intake and physical activity of children. Systematic utensil survey was undertaken in 3 randomly chosen households per village i.e. 27 households and the portion sizes were derived from a total of 74 serving utensils. The derived sizes were classified as small (240 ml), medium (320 ml) and large (600 ml). The semi-quantitative FFQ with locally derived portion sizes was then administered to 63 children in 6-12 year age group. Results: The volume of food measured by the reference portion sizes generally being employed in the national surveys, would have been underestimated the food consumed by the child by 55-60% as compared to what was being consumed by the children in the study area. The correlation between food intake assessed by 24-hr recall method and FFQ using derived (local) portion sizes was better as compared to that obtained with the semi-quantitative FFQ done with reference portions. Conclusion: In conclusion, local portion size adaptation of FFQ for semi-quantification is useful to mitigate measurement errors associated with this technique.