Accuracy of portion size estimation using the 2-dimensional food portion visual

Errors in self-reported food intake are, in part, attributable to portion size aids used in estimating food intake. We have previously reported differences between males and females in the aids used to estimate portion sizes. Females used common household measures more often than males, whereas males used the 2-D Food Portion visual (FPV) (Nutr. Consulting Enterprises, Framingham, MA) more often than females. The purpose of this study was to examine more closely the accuracy of the FPV used in estimating portion sizes. 24 h recalls were collected by telephone using the Minnesota Nutrition Data Systems from male (n=25) and female (n=18) subjects participating in the Dietary Effects on Upoproteins and Thrombogenic Activity (DELTA) Study (Protocol 1), a multicenter, well-controlled feeding study. The % differences between actual and reported intakes were calculated for food portions estimated using the FPV. The analyses included a comparison of Side A (volume or fluid ounce measures) and Side B (measures in dimensions). Using Side A of the FPV resulted in an overestimation of portion sizes by 21% for males and 45% for females. Using Side B, resulted in an underestimation of portion sizes by males (-1%) and females (-21%). For females, portion size estimations using Side B were more accurate than using Side A (p<0.0001). There were no statistically significant differences by side of FPV used for males. Overall, males tended to more accurately estimate portion sizes than females, although the differences were not statistically significant. These data suggest that accuracy in reporting portion sizes may be different for males and females when using the FPV. These differences should be further explored to improve the accuracy of self-reported food intakes.