J Nutr Health.  2020 Jun;53(3):319-328. 10.4163/jnh.2020.53.3.319.

Validation of nutrient intake of smartphone application through comparison of photographs before and after meals

Affiliations
  • 1Department of Clinical Nutrition, Graduate School of Public Health, Dongduk Women's University, Seoul 02748, Korea
  • 2Department of Food and Nutrition, School of Natural Science, Dongduk Women's University, Seoul 02748, Korea
  • 3Samsung Genome Institute, Samsung Medical Center, Seoul 06351, Korea
  • 4Department of Molecular Cell Biology, Sungkyunkwan University School of Medicine, Suwon 16419, Korea
  • 5Machine Learning Lab, Samsung Advanced Institute of Technology, Suwon 16678, Korea

Abstract

Purpose
This study was conducted to evaluate the validity of the Gene-Health application in terms of estimating energy and macronutrients.
Methods
The subjects were 98 health adults participating in a weight-control intervention study. They recorded their diets in the Gene-Health application, took photographs before and after every meal on the same day, and uploaded them to the Gene-Health application. The amounts of foods and drinks consumed were estimated based on the photographs by trained experts, and the nutrient intakes were calculated using the CAN-Pro 5.0 program, which was named ‘Photo Estimation’. The energy and macronutrients estimated from the Gene-Health application were compared with those from a Photo Estimation. The mean differences in energy and macronutrient intakes between the two methods were compared using paired t-test.
Results
The mean energy intakes of Gene-Health and Photo Estimation were 1,937.0 kcal and 1,928.3 kcal, respectively. There were no significant differences in intakes of energy, carbohydrate, fat, and energy from fat (%) between two methods. The protein intake and energy from protein (%) of the Gene-Health were higher than those from the Photo Estimation. The energy from carbohydrate (%) for the Photo Estimation was higher than that of the Gene-Health. The Pearson correlation coefficients, weighted Kappa coefficients, and adjacent agreements for energy and macronutrient intakes between the two methods ranged from 0.382 to 0.607, 0.588 to 0.649, and 79.6% to 86.7%, respectively.
Conclusion
The Gene-Health application shows acceptable validity as a dietary intake assessment tool for energy and macronutrients. Further studies with female subjects and various age groups will be needed.

Keyword

smart phone; mobile application; validation study; nutrition assessment
Full Text Links
  • JNH
Actions
Cited
CITED
export Copy
Close
Share
  • Twitter
  • Facebook
Similar articles
    DB Error: unknown error