Date of Award


Document Type

Open Access Dissertation


Epidemiology and Biostatistics

First Advisor

James R. Hebert


Background: Diet and its components are known to play an important factor in the process of inflammation and in turn on the health effects that are related to inflammation like cancer, and cardiovascular diseases. Previous research so far has mainly looked at the effect of specific food or nutrients on inflammation and health outcomes. In this regard a new literature derived and population based dietary index called Dietary Inflammatory Index (DII) was developed after carefully screening around 6000 articles that looked at the association between 45 food parameters and 6 commonly studied inflammatory markers. The list of food parameters includes various nutrients like vitamins, minerals, macronutrients; food items like garlic, onion, ginger; and bioactive compounds flavonoids. The purpose of this research is to explore the following associations: 1. Association between DII and C - reactive protein (CRP) in NHANES 2005-2010; 2. Association between DII and mortality in NHANES III study and 3. Association between DII and inflammatory and metabolic biomarkers in CAN DO intervention study.

Methods: First data from NHANES 2005-2010 was used to examine the association between DII and CRP among the United States population by race/ethnicity and diabetes status.CRP was analyzed as both continuous and as categorical variable (based on the cut-off of 3mg/l). As CRP was not normally distributed it was log transformed and analyzed. In the same dataset HEI-2010 was calculated and used to predict CRP and results were compared with those of DII. Multivariate linear and logistic regression was used for the analyses. Next, we examined the ability of a newly developed dietary inflammatory index (DII) to predict mortality in the National Health and Nutrition Examination Survey (NHANES) III cohort study. The DII was computed based on baseline dietary intake assessed 24-h dietary recalls (1988-94). Mortality was determined from the National Death Index records through 2006. Cox proportional hazards regression was used to estimate hazard ratios. During the follow-up period through the end of 2006, 2795 deaths were identified, including 615 cancer, 158 digestive cancer and 1233 cardiovascular (CVD) deaths. Following this for the third aim data was used from the CAN DO study, a dietary intervention study with a sample size of 234 individuals with metabolic syndrome. The two interventions were 1) a high fiber diet and 2) the American Heart Association (AHA) diet. DII was calculated using 24-h dietary recalls at baseline, 6, and 12 months and was tested against metabolic markers (insulin, blood glucose and homeostasis model assessment (HOMA-IR)) and inflammatory markers, namely C-reactive protein (CRP), IL-6 and TNF-α using linear mixed models adjusted for covariates. All the biomarkers were log transformed and the results are back-transformed and expressed as the ratio of the geometric means.

Results: Multivariate analysis revealed CRP to be positively associated with DIIQuartile4vs1 (β =0.19, C.I. 0.13, 0.24), and HEI-2010 Quartile1vs4 (β =0.15, C.I. 0.10, 0.20). Similar associations were observed when CRP was categorized (>3 mg/l), DIIQuartile4vs1 (OR= 1.37, C.I. 1.27, 1.71), and with HEI-2010Quartile1vs4 (OR= 1.31, C.I. 1.12, 1.56). Multivariable analysis, adjusting for , race, diabetes status, hypertension, physical activity, BMI, poverty index and smoking, revealed positive associations between higher DII and overall mortality (HR for DII Tertile3vs1=1.34; 95%CI 1.19- 1.51, p-trend-<0.0001), cancer related mortality (HR for DII Tertile3vs1=1.46; 95%CI 1.10- 1.96, p-trend-0.01), digestive cancer mortality (HR for DII Tertile3vs1=2.10; 95%CI 1.15- 3.84, p-trend-0.03) and CVD mortality (HR for DII Tertile3vs1=1.46; 95%CI 1.18- 1.81, p-trend-0.0006). Across time points DII was lower in the ‘high fiber’ group compared to the ‘AHA’ group. For metabolic factors after multivariate analysis compared to tertile 1 participants in tertile 3 had higher insulin level (1.41; 95% CI 1.02, 1.91), glucose level (1.07; 95% CI 1.02, 1.12) and HOMA-IR (1.51; 95% CI 1.07, 2.09) and for inflammatory biomarkers, compared to tertile 1 participants in tertile 3 had higher IL-6 values (1.35; 95% CI 1.05, 1.78). No significant associations were observed with CRP and TNF-α.

Conclusion: The results from all the three aims reinforce the fact that diet as a whole plays an important role in modifying inflammation and health outcomes related to inflammation. Results from aim 1show the DII can predict inflammation in general population with different ethnicities and from different regions of US and also comparatively DII has a slightly better predictive ability (6%) compared to HEI-2010. Results from second aim showed that a pro-inflammatory diet, as indicated by higher DII scores, was associated with overall, cancer and CVD mortality and finally results from the intervention study provide further evidence that fiber has a major effect in reducing inflammation and insulin resistance and also reinforce the opinion that DII can be used as a tool to detect the levels of metabolic and inflammatory biomarkers. Finally all the above mentioned results suggest that diet has a major role in controlling inflammation and thereby plays an important role in the development or prevention of various chronic diseases, hence public health steps should be taken to modify individual’s whole diet rather than the intake of specific nutrients.

Included in

Public Health Commons