Browsing by Author "Göransson, Annette"

ABSTRACT Background and aim of the study: Recently the reliability and accuracy of traditional research methods of food intake have been questioned. Traditional methods have a tendency to underestimate the energy intake of the subject. Individuals’ ability to estimate the portion size, memory, and self-report affects the reliability. New technology offers new tools for nutritional research. Individuals’ food and energy intake can in theory be estimated from photographs showing the food portions and left over food. In this master’s thesis the aim is to investigate if food photographs of preschool children can be used in assessment of food consumption. The reference method in this research was weighted portion sizes. Active food photography method has not in fact been investigated on small children before. Design and methods: This master’s thesis is part of the DAGIS research project. The subjects (n=22) were children aged 3–6 years in two day care centers in Helsinki. The food intake of the subjects was photographed during lunch hour and snack time in the daycare setting with two cameras. All food items and left over food was also weighed. Research material includes also the children’s food diaries filled in by the daycare personnel. Portion size estimates from food photographs and food diaries were compared against the weighted portion sizes. Arithmetic means and standard deviations were calculated for different food groups. The difference between the weighed portion size and the estimated portion size was calculated with 95 % confidence interval. The geometric means with 95 % limits of agreement were also calculated for the ratio of weighted and estimated portion size. Bland-Altman figures were drawn to assess the agreement between methods. Results: This master’s thesis found that on average food photographs overestimated the portion sizes compared to the weighted portions (geometric mean 1,06). The estimations made from the photographs were in line with the food records in if they over or under estimated the portions in different food groups. When comparing the energy intake from the study meals this study found that there was no difference between the estimated energy intake from photographs and the energy intake from weighted portions (geometric mean 1,01). The differences made in the portion size estimation (over or underestimation) were seen only in part of the food groups in both estimation methods. Conclusion: This study found no difference between the energy intake estimated from the food photographs and weighed portions. This was an exceptionally good result compared to earlier studies where food photographs have underestimated subjects’ energy intake. The explanation for the difference in results may possibly be explained by the study design. In this study the food photographs were taken at the day care centers by researchers and the day care personnel filled in the food records. Third party report is different from self-report. Food photographs have hardly been studied in food consumption studies with small children before. The heterogeneity of the previous studies and the lack of validation of the different food photograph methods make the comparison of the study results difficult. This study suggests that food photographs could someday even replace the traditional methods. Before that can happen, we need more studies where the method has been validated and the study settings is well planned forehand. Food photographs would be a useful tool for nutritional studies especially in a busy day care environment. KEYWORDS: Dietary assessment: food intake, energy intake, digital photography, food photography, food records, portion size estimation, children