Data Sheet 1_Performance of cohort-adapted dietary and lifestyle inflammation scores among Hispanic adults.docx

BackgroundDietary and lifestyle inflammation scores (DIS/LIS) were created to assess their contributions to systemic inflammation; however, there is little understanding of their validity in Hispanic adults.
ObjectiveThis study aims to utilize DIS and LIS in the Boston Puerto Rican Health Study (BPRHS), as previously published, and create cohort-adapted scores. Adapted scores were validated in The Puerto Rico Observational Study of Psychosocial, Environmental, and Chronic disease Trends (PROSPECT).
MethodsThis cross-sectional analysis assessed diet from a food-frequency questionnaire, self-reported lifestyle information, and inflammation from fasted blood. Published food and lifestyle groups were used to create DIS-1 and LIS-1. Cohort-adapted food and lifestyle groups were used to create DIS-2 and LIS-2 in the BPRHS (n = 854). Factor analysis was used to create DIS-3. The associations among DIS-1, LIS-1, DIS-3, and a biomarker score (continuous and dichotomized) were tested in BPRHS using multivariable linear and logistic regression. The associations among DIS-2, LIS-2, DIS-3, and inflammation (continuous concentration and higher versus lower hsCRP) were tested in PROSPECT (n = 835) using multivariable linear and logistic regression.
ResultsIn BPRHS, DIS-1 showed a 1.07 (95% CI: 1.01, 1.15) times greater odds of a high inflammation BMS with each 1-unit increase in the DIS-1 and a 2.86 (95% CI: 2.14, 3.56) times greater odds of a high BMS with each 1-unit increase in the LIS-1. In PROSPECT, LIS-1 showed a 2.91 (95% CI: 2.33, 3.67) times greater odds of high hsCRP with each 1-unit increase in LIS-1; results were similar in linear analyses. DIS-3 was characterized by three factors. In BPRHS, high (quintile 5) vs. low (quintile 1) adherence to factor 1 (healthy diet) showed a 0.57 (95% CI: 0.35, 0.92) times lower odds of high BMS. DIS-3 was not associated with high hsCRP in PROSPECT.
ConclusionPredetermined lifestyle inflammation scores were associated with inflammation in this population, but dietary inflammation scores were inconsistent in their association with inflammation. Cohort-specific adaptation did not improve the scores’ association with inflammatory status. Further work is needed to understand the role of diet in the development of inflammation in these populations of Hispanic adults.