Statistical approach to identify food categories that determine daily intake levels of total and inorganic arsenic, lead and aluminium of Japanese diet.

Food categories that significantly contribute to daily intake of total arsenic (TAs), inorganic arsenic (iAs), lead (Pb) and aluminium (Al) were statistically sought for the 949 duplicate diet samples collected from 319 households in Japan. Daily intakes of TAs, iAs, Pb and Al were calculated based on measured element concentration in duplicate diet samples and weight of the samples. Amounts of consumption of 12 food categories of each duplicate diet sample were self-reported (beverage), interviewed (fat and oil, spices, and sugar) and actually weighted (other 9 categories) at the time of diet sampling. Multiple regression analyses were conducted by using daily element intake as dependent variable and amounts of daily consumption of the 12 food categories as independent variables. The most significant predictors were amount of daily consumption of 'Fish & Shellfish' for daily TAs intake and 'Rice & Rice products' for iAs intake. Other food categories contributed to daily intakes of the four elements with smaller coefficient. Some of the significant predictors identified in this study were consistent with the findings of the previous market basket surveys while others were not. The 12 food category consumption data moderately explained daily intake of TAs (Coefficient of determination adjusted for degree of freedom (adjusted R2) = 0.447) and iAs (0.307) while they only poorly explained daily intakes of Pb (0.229) and Al (0.117). The present results suggest that a Total Diet Study based on food consumption survey data and element contents of food from data base/literature will result in large errors in the estimation of daily element intake.