ObjectivesTo test the relationship of diet quality and health behavior indexes derived from a validated food liking survey with cardiometabolic health in a convenience sample of non-diabetic patients with diagnosed depression from a psychiatric facility and age-gender matched students from a University setting. MethodsOne hundred six patients and 106 controls (62% female, mean age = 21) completed a 100-item liking survey comprised of foods, beverages, and physical and sedentary activities. Nutritional items were conceptually grouped, weighted and averaged into a Diet Quality Index (DQI). A Healthy Behavior Index (HBI) was the average of the weighted nutritional and physical activity groups. Higher indexes reflected healthier behaviors. Multiple linear regression was used to relate DQI and HBI with blood pressure and fasting insulin, glucose and serum lipids. BMI, biologic sex and patient status were included as covariates. Some serum markers required transformation to approach normal distribution. ResultsFrom BMI, 4% were underweight, 57% normal, 25% overweight, and 14% obese. The liking survey took minutes for participants to complete and little processing to generate the indexes. The DQI and HBI were internally reliable (α = 0.68–0.69), reflected greater than one dimension (principal component analysis), and were normally distributed. Higher DQI scores were significantly related to higher HDL (standardized β = 0.16, P = 0.019) and lower glucose (β = –0.19, P = 0.005), insulin (β = –0.18, P = 0.005) and diastolic blood pressure (β = –0.32, P < 0.001). Adding physical activity, HBI also was related to glucose (β = –0.19, P = 0.008), triglycerides (β =–0.14, P = 0.04) and insulin (β = –0.18, P = 0.006). Patient status was not a significant independent contributor in the models. ConclusionsA simple liking survey can produce reliable and valid indexes of diet quality and health behaviors that significantly associated with multiple cardiometabolic risk factors. Funding SourcesUSDA NIFA, Hatch project.