[Evaluation study on a new method of dietary assessment with instant photography applied in urban pregnant women in Nanjing city].

OBJECTIVE

To evaluate the usefulness and efficiency of a novel dietary method among urban pregnant women.

METHODS

Sixty one pregnant women were recruited from the ward and provided with a meal accurately weighed before cooking. The meal was photographed from three different angles before and after eating. The subjects were also interviewed for 24 h dietary recall by the investigators. Food weighting, image quantification and 24 h dietary recall were conducted by investigators from three different groups, and the messages were isolated from each other. Food consumption was analyzed on bases of classification and total summation. Nutrient intake from the meal was calculated for each subject. The data obtained from the dietary recall and the image quantification were compared with the actual values. Correlation and regression analyses were carried out on values between weight method and image quantification as well as dietary recall.

RESULTS

Total twenty three kinds of food including rice, vegetables, fish, meats and soy bean curd were included in the experimental meal for the study. Compared with data from 24 h dietary recall (r = 0.413, P < 0.05), food weight estimated by image quantification (r = 0.778, P < 0.05, n = 308) were more correlated with weighed data, and show more concentrated linear distribution. Absolute difference distribution between image quantification and weight method of all food was 77.23 ± 56.02 (P < 0.05, n = 61), which was much small than the difference (172.77 ± 115.18) between 24 h recall and weight method. Values of almost all nutrients, including energy, protein, fat, carbohydrate, vitamin A, vitamin C, calcium, iron and zine calculated based on food weight from image quantification were more close to those of weighed data compared with 24 h dietary recall (P < 0.01). The results found by the Bland Altman analysis showed that the majority of the measurements for nutrient intake, were scattered along the mean difference line and close to the equality line (difference = 0). The plots show fairly good agreement between estimated and actual food consumption. It indicate that the differences (including the outliers) were random and did not exhibit any systematic bias, being consistent over different levels of mean food amount. On the other hand, the questionnaire showed that fifty six pregnant women considered the image quantification was less time-consuming and burdened than 24 h recall. Fifty eight of them would like to use image quantification to know their dietary status.

CONCLUSION

The novel method which called instant photography (image quantification) for dietary assessment is more effective than conventional 24 h dietary recall and it also can obtain food intake values close to weighed data.