Reproducibility of a semi-quantitative food frequency questionnaire in Chaoshan area, China.

OBJECTIVE To examine reproducibility of assessed intake of foods and nutrients according to a semi-quantitative food frequency questionnaire (SQFFQ) in adult doctors and nurses residing in Chaoshan area of China. SUBJECTS The SQFFQ was administered first in October to December of 2004 to 120 adult doctors and nurses living in Chaoshan area of China and was then re-administered to 102 three months later between January and March of 2005 (SQFFQ 1 and SQFFQ 2). METHODS Reproducibility was evaluated in terms of consumption of 10 food groups and energy and 34 macro- and micro-nutrients based on the SQFFQ from the 102 doctors and nurses. RESULTS For intake of foods, Pearsom's correlation coefficients (CCs) with log-transformation and energy adjustment (minimum - median - maximum) range from 0.43 (eggs) - 0.84 - 0.90 (teas). Spearman's rank CCs with energy adjustment ranged from 0.77 (cereals) - 0.84 - 0.94 (milks). Kappa statistics with energy adjustment ranged from 0.53 (vegetables) - 0.63 - 0.82 (teas). For consumption of nutrients, Pearson's correlation coefficients (CCs) with log-transformation and energy adjustment (minimum - median - maximum) range from 0.83 (docosahexaenoic acid and oryzanin) - 0.88 - 0.90 (linolenic acid, vitamin A, folic acid, vitamin E, calcium, sodium, selenium and magnesium). Spearman's rank CCs with energy adjustment ranged from 0.81 (oryzanin and vitamin C) - 0.86 - 0.90 (sodium). Kappa statistics with energy adjustment ranged from 0.49 (protein) - 0.60 - 0.77 (sodium). CONCLUSION Substantially high reproducibility was observed; it is possible to use the tailored, relatively simple, but comprehensive, self-administered SQFFQ to facilitate assessment of the association between lifestyle and health/disease in large-scale epidemiological studies.