Online

Eriksen, Harley

Nutritional and Food Sciences

2016

MS

To properly evaluate initiatives targeting children's fruit and vegetable (FV) consumption, we need feasible and valid dietary assessment methods that are time, cost, and resource effective. The objectives of this study were to test the feasibility and validity of two methods, digital imaging (DI) and aggregate plate waste (APW), for assessing children's FV consumption. UVM dietary assessment team graduate students prepared and distributed FV snacks in two pre-school classrooms over 30 consecutive school days. Feasibility of APW was tested by recording the frequency and weight of waste sorting errors by pre-school children and performing paired t-tests comparing uncorrected and corrected FV waste. Feasibility was tested for DI by determining the total number of individual FV snacks from which FV consumption could be estimated using the digital images. Validity was tested for DI using paired t-tests to compare FV consumption as assessed by DI against actual consumption as assessed by weighed plate waste (WPW). A total of 159 cluster APW weights were recorded during the 20 days of APW collection, with an overall mean difference of 0.57 grams (p=0.440) between uncorrected and corrected FV waste. Researchers were able to capture 100-percent usable digital images, effectively displaying 214 individual FV snack servings over the 10-day DI study period. Percent agreement between the two digital image coders was 99.1-percent. DI estimations for individual FV item and cumulative consumption were strongly correlated with WPW (all above r=0.97). Overall FV consumption as estimated by DI differed from WPW by less than one gram, and DI estimations for individual FV items differed from WPW by no more than two grams. Paired t-tests revealed no significant difference between DI estimations of clementine (p=0.954) and peapod (p=0.806) consumption and WPW measurements. However, paired t-tests indicated statistically significant differences between DI and WPW measurements for overall (p=0.001), grape (p=0.031), carrot (p=0.008), and pepper (p=0.027) consumption. Both methods were feasible for assessing mean FV consumption. DI estimations for individual FV items and cumulative consumption were strongly correlated with WPW, suggesting that despite statistical significance between DI and WPW measurements in some cases, the DI method is still precise for mean FV consumption evaluation. APW may be especially advantageous for rapid and efficient evaluation of behavior change in response to interventions targeting children's FV consumption. Due to ease of administration and instantaneous results, the APW method reduces the need for trained research staff to be present, drastically increasing accessibility to group-level dietary assessment.