Random errors, from any source, will attenuate epidemiological risk estimates. Before we launched the Shanghai Men's Health Study (SMHS), a large population-based cohort study investigating the diet–cancer association among Chinese men, a dietary calibration study was conducted among 96 men aged 40–75 years (mean age 56.5 years), with biweekly 24-hour dietary recalls (24HDRs) implemented over a 1-year period. Data from this study were analysed to evaluate the nature and magnitude of variances for intake of 26 nutrients among SMHS participants, to compare variance ratios of 26 nutrients among Chinese men and women and individuals in other studies, and to estimate the number of 24HDRs required for future dietary calibration studies in similar populations.

Ninety-six healthy, free-living men in Shanghai were administered biweekly 24HDR interviews 24 times over a 1-year period. To assess between-individual and within-individual contributions to variance, a mixed effects model was fitted and ratios of within-individual to between-individual dietary intake variances were computed.

Shanghai, China.

In agreement with reports from studies conducted in the USA and many other countries, we found that within-individual variances were usually larger than between-individual variances in dietary intake for all nutrients. The sum of all other variation (e.g. weekday and weekend, seasonal, interviewer) accounted for less than 5% of total variation. Ratios of within- to between-individual variances (for log-transformed data) ranged from 1.25 for carbohydrate intake to near 8 for δ-tocopherol intake.

The results of this study suggest that among middle-aged and elderly Chinese men in Shanghai, within- and between-individual variation account for more than 95% of the total variation for 26 nutrients. Further dietary validation studies in the same population could be adequately carried out with only 12 days of dietary recalls, if 100 participants were enrolled.