Comparability of nutrient estimation by three food frequency questionnaires for use in epidemiological studies

Replication of results is an important issue in studies of diet and disease , possibly dependent on data collection method. We compared assessments fro m the Health Habits and History Questionnaire (HHHQ), the Harvard Semiquant itative Food Frequency Questionnaire (HFFQ), and the New York State Cohort Food Frequency Questionnaire (CFFQ) for estimates of daily intakes of energ y, protein, carbohydrates, total fat, dietary fiber, cholesterol, vitamins A, C, and E, and carotenoids. Fifty-nine men and 50 women aged 35-73 years completed the HHHQ and HFFQ as interviews and the 44-food CFFQ as a self-ad ministered mailed questionnaire. Comparability was assessed with Spearman c orrelation coefficients. Quantitation of nutrient intake differed by nutrie nt, questionnaire, and nutrient calculation method. Ranking on energy and m acronutrient intake for the HHHQ and HFFQ ranged from 0.62 to 0.80; ranking for micronutrient intake ranged from 0.56 to 080. Far the CFFQ with the HH HQ or HFFQ, correlations ranged between 0.29 and 062. The CFFQ performs com parably to the HHHQ and HFFQ far some, but not all, nutrients; our results suggest that the HHHQ and HFFQ can be used interchangeably with reasonable confidence in studies of diet and disease.