Validity of the estimated food diary: comparison of 2-day recorded and observed food and nutrient intakes

We assessed the validity of the estimated food diary by comparing recorded and observed food and nutrient intakes for 121 subjects (71 men and 50 women) aged 15 to 65 years. We obtained observed intake by unobtrusively recording the amounts of foods, preweighed or measured when served, at the four meals on each of the two study days. At the same time, subjects kept an estimated food diary. In general, the mean food intake recorded in the food diaries deviated from the observed intake by 15% at most. The deviation exceeded 20% for only a few food items. With the exception of calcium (-11%), starch (+10%) and thiamin (+ 8%), there was no difference or only a few percent difference in the means for energy and the 16 nutrients calculated. The validity of the food diary was equally high for men and women in the 15 to 34, 35 to 44, and 45 to 65 age groups. Product-moment correlation coefficients ranged from 0.68 to 0.97 for food intake and from 0.75 to 0.92 for nutrient intake. We conclude that validity is very satisfactory on the group level and is probably acceptable on the individual level