Objective: To examine the relationship between diet quality and source of food among U.S. adults.
Design: Cross-sectional sample of adults who participated in the 1994-1996 Continuing Survey of Food Intakes by Individuals divided into clusters based on their food source.
Setting: In-person interviews conducted nationwide.
Participants: A national sample of 9407 adults over 18 years of age.
Method: Data gathered from the interviews consisted of two 24-hour dietary recalls; self-reported demographic characteristics; body mass index (BMI); and health behaviors such as physical activity, smoking, and alcohol consumption.
Main outcome measures: The United States Department of Agriculture (USDA) Healthy Eating Index (HEI) score and the percentage of total grams of food consumed from each source.
Analysis: Multivariate linear regression evaluated the association between food source pattern determined by cluster analysis, and HEI scores, while controlling for demographic characteristics, BMI, and health behaviors.
Results: Ten primary food source patterns were identified from the clusters. The food source clusters were statistically significant in predicting HEI score and were stronger predictors than most demographic and health behaviors included.
Implications for research and practice: The success of cluster analysis in identifying food source patterns and associating them with diet quality suggests that identifying an individual's food source pattern may be more predictive of diet quality than demographic characteristics, and other health behaviors.