ABSTRACT
The food frequency questionnaire (FFQ) is the preferred method to evaluate long-term usual dietary intake in population-based epidemiological studies because it is simple, easy to administer and requires minimal effort from the subjects. Therefore, we validated a food frequency interview schedule (FFIS) to estimate the dietary intakes of the urban population of Hyderabad city. Methods: A cross-sectional study was conducted among five socio-economic sections of Hyderabad. Areas for the survey were selected by cluster random sampling and households in each area were selected by simple random sampling. The FFIS was developed and validated against a 6-day 24-hour dietary recall (HDR) method. The instruments were administered to the participants six months apart to check for reproducibility. Statistical analyses for validation and reproducibility included correlation, regression analyses and paired t-test. Results: Means of intakes of foods measured by 24-HDR were significantly lower than those measured by FFIS for some foods at alpha levels of 0.05. Pearson's correlation (r) for the intakes by the two methods ranged from 0.12 to 0.85. Regression coefficients were significant for 12 food groups. Correlation coefficients for the two FFISs were between 0.31 (spices) and 0.81 (carbonated beverages) and showed good reproducibility. Intakes of conventional foods like cereals, pulses, vegetables etc. by FFIS correlated better with 24-HDR than the processed foods such as breakfast cereals and bakery items. Conclusion: The data suggests that the FFIS is a well-validated, reproducible tool for assessment of long term dietary habits of a specific population. However, its use for populations of other regions requires specific modifications.