Date of Award

Summer 7-7-2016

Degree Type


Degree Name

Master of Science (MS)



First Advisor

Dr. Megan McCrory

Second Advisor

Barbara Hopkins

Third Advisor

Dr. Lara Dugas


Objective: To conduct a secondary analysis of dietary variety consumed by individuals of African-origin in two countries with differing stages of economic development. Our overall aim is to determine the relationships of two different dietary variety scores developed previously in our laboratory with reported energy intake (rEI), ER (which will be a more accurate reflection of true EI) and BMI in the total sample and the plausibly reporting subsample.

Methods: Data for this analysis were collected as part of METS between January 2010 to September 2011, whose purpose was to elucidate the associations of physical activity and diet with body weight, diabetes, and risk of cardiovascular disease. Five communities of African-origin and in different countries were selected based on their different levels of economic development, as measured using the UN Human Development Index. A subsample of 141 (Ghana, n=70 and U.S., n=71) men and women with an average age of 35.1±0.5 years and an average BMI of 27.5±0.6 kg/m2 were randomly selected to have their total energy expenditure (TEE) measured by the doubly labeled water (DLW) method. Participants were interviewed using the multiple-pass method designed by the Medical Research Council of South Africa to estimate their dietary intake the day after consumption. Data was transferred to Nutrient Data System for Research (NDSR) ver. 2011 and dietary variety scores (DVS) were calculated for combination and ingredient varieties. Combination variety was defined as the total number of unique foods and beverages consumed in a day. Ingredient variety was the total number of unique ingredients consumed in a day. Implausibility of rEI was controlled for by calculating rEI as a percentage of TEE. Associations of dietary variety scores with total energy intake and BMI were assessed for both the total sample and plausible subsample using SPSS version 22 through univariate analyses of variance and correlations.

Results: Both combination and ingredient variety were positively associated with rEI in both countries when implausible reporting was not controlled, but no significant association was observed in both countries when implausible reporting was controlled. Ingredient variety was negatively associated with TEE when implausible reporting was both controlled and uncontrolled in the U.S. (p= 0.029), but no association was observed in Ghana. Ingredient and combination variety were also negatively associated with log BMI, percent body fat, and weight in U.S. when implausible reporting was not controlled but not in Ghana’s. However, in Ghana, combination variety was positively associated with percent body fat (p=0.041) and log BMI (p= 0.027) when plausible reporting was controlled but was not significant when implausible reporting was uncontrolled.

Conclusion: Dietary variety was positively associated with rEI in both countries when implausible reporting was not controlled and with obesity markers in Ghana when plausible reporting was controlled.

Available for download on Saturday, July 15, 2017