Reproducibility and Relative Validity of Food Group Intake in a Food Frequency Questionnaire Developed for the Tehran Lipid and Glucose Study

OBJECTIVE To examine the validity and reproducibility of food groups in the semi-quantitative food frequency questionnaire (FFQ) developed for the Tehran Lipid and Glucose Study (TLGS). METHODS To evaluate the reproducibility of food groups included in the FFQ, 132 subjects (61 men and 71 women) aged 20 years or older twice completed a 168-item FFQ (FFQ1, FFQ2), with a 14-month interval between FFQ1 and FFQ2. Over the 1-year interval, 12 dietary recalls (DRs) were collected (1 each month) to assess the validity of the FFQ. Seventeen food groups were derived from the FFQ based on methods described in previous studies. Age-adjusted and deattenuated Spearman correlation coefficients were used to assess the validity of the FFQ. RESULTS The mean (SD) age and body mass index of subjects were 35.5 (16.8) years and 25.5 (5.2) kg/m(2), respectively. Validity correlation coefficients ranged from 0.03 (liquid oil) to 0.77 (simple sugars) in men (median, 0.44), and from 0.12 (snacks) to 0.79 (simple sugars) in women (median, 0.37). The energy-adjusted intraclass correlation coefficient, which reflects the reproducibility of the FFQ, was 0.51 in men and was highest for tea and coffee (0.91); in women it was 0.59 and was highest for simple sugars (0.74). The highest percentage of complete agreement and disagreement was observed for snacks and desserts (60.6%) and potatoes and dairy products (12.8%), respectively, in men, and tea and coffee (62.9%) and legumes (15.7%) in women. CONCLUSIONS The FFQ that was designed for the TLGS was found to be reliable and valid for assessing the intake of several food groups.