A statistical approach for estimating the distribution of usual dietary intake to assess nutritionally at-risk populations based on the new Japanese Dietary Reference Intakes (DRIs).

The health of individuals is not markedly affected by the nutrients ingested in a single day; rather it is influenced by dietary habits over a long period of time. However, it is impossible to survey usual intake directly because recording diet over a long period of time is burdensome for the subjects, so usual dietary intake is rarely measured directly. Instead, we estimated the distribution of selected nutrients in subjects' usual dietary intake using a statistical method (Best-Power method) described previously. And we assessed the proportion of nutritionally at-risk subjects in individual groups based on the new Japanese Dietary Reference Intakes (DRIs), the Estimated Average Requirement (EAR), or the tentative dietary goal for preventing life-style related diseases (DG) as cut-points. We collected the survey data from 208 men and 251 women, aged 50 to 69 y in 2004 and 2005. The survey was carried out on three non-consecutive days four times each year: in spring, summer, autumn, and winter. The distribution of nutrients in the usual intake was estimated from a dietary survey of 3 d using one-way analyses of variance. We found that the proportion of the population at risk for nutrient deficiency was overestimated in the 1-d intake distribution. On the other hand, the fraction that was nutritionally at-risk in terms of salt intake, according to DG as the cut-point, was underestimated in the 1-d intake distribution: 74.0% of men and 82.5% of women in the 1-d intake, and 90.5% and 93.2%, respectively, for the estimated usual intake adjusted for seasonal variation.