Reproducibility and validity of a quantitative food frequency questionnaire in an urban and rural area of northern India

Background. Food frequency questionnaires (FFQs) have been used in epidemiological studies across the world to capture the usual food intake of individuals. As food habits vary in different population groups, FFQs should be validated before use. Hence, we determined the reproducibility and validity of FFQs designed for urban and rural populations of northern India. Methods. Separate FFQs, designed for urban and rural populations using standard methods, were administered to a sample of 200 subjects (100 urban and 100 rural) in the age group of 35-70 years in the beginning (baseline FFQ) of the study and after an interval of 1 year (1-year FFQ) to assess their reproducibility. Six 24-hour dietary-recalls, taken at an interval of 2 months over a period of 1 year, were used as a reference method to test the validity. Crude and energy-adjusted nutrient intakes estimated from FFQs and 24-hour dietary-recalls were compared using Pearson correlation coefficients. Bland and Altman plots were also used to test the agreement between the two methods. Results. Nutrient intakes were found to be similar at the baseline and 1-year FFQs in urban and rural areas. The unadjusted Pearson correlation between 24-hour dietary-recalls and 1-year FFQ ranged from 0.22 for vitamin C to 0.63 for iron in the urban area. It ranged from 0.06 for vitamin C to 0.74 for energy in the rural area. The correlations lowered after adjusting for energy and there was a minimal increase after de-attenuation. Conclusion. The FFQs were reproducible and valid for assessing nutrient intakes except for some micronutrients.