Date of Award

Spring 2021

Document Type

Open Access Dissertation

Department

Epidemiology and Biostatistics

First Advisor

James Herbert

Abstract

Background: Diet is now known to play an important role in the process of inflammation and subsequent chronic health events, such as cancer, metabolic syndrome (MetS), anxiety, depression, and cardiovascular disease (CVD). The Dietary Inflammatory Index (DII®) and energy-adjusted DII (E-DIITM) are tools that can estimate the inflammatory potential of diet of individuals. Previous research has focused predominantly on diet-associated inflammation and adverse health effects in populations that were not composed of cancer survivors. As of the time of this dissertation, none have focused on childhood cancer survivors. Childhood cancer survivors are at an elevated risk of suffering from an adverse chronic health event at an earlier age compared to the general population. The St. Jude Lifetime Cohort Study (SJLIFE) provides the unique opportunity to investigate potential associations between the inflammatory potential of diet and adverse chronic health problems among this group of survivors and controls. The purpose of this research is to explore the following associations: 1. Association between diet-associated inflammation and C-reactive Protein (CRP); 2. Association between diet-associated inflammation and Metabolic Health (hemoglobin A1C, HOMA-IR, fasting glucose, fasting insulin, waist circumference, high-density lipoproteins, triglycerides, and metabolic syndrome); 3. Association between E-DII and Mental Health (anxiety and depression); and 4. Association between diet-associated inflammation and Cardiovascular Health (hypertension, hyperlipidemia, cardiomyopathy, coronary artery disease (CAD), cardiac dysrhythmia, and cerebrovascular accidents (CVA).

Methods: Data from SJLIFE among cancer survivors diagnosed between 1962 and 2012 we used to examine associations between diet-associated inflammation and adverse health outcomes among a cohort of childhood cancer survivors and controls. The E-DII who used in all analyses based on better model fit (lower AIC values). E-DII scores were computed using the 2005 Block Food Frequency Questionnaire (FFQ). The E-DII was analyzed as both as a continuous variable and categorized into quartiles where quartile 1 represents the most anti-inflammatory diet. First, we investigated the association between the E-DII and CRP among SJLIFE cohort members. CRP was analyzed as a continuous, binary (based on the cut-off 3mg/L), and categorical outcome (based on two cut-offs: <1 mg/L, 1-3 mg/L, and >3 mg/L). Second, we examined the associations between the E-DII and hemoglobin A1C, fasting glucose, fasting insulin, Homeostatic Model of Assessment of Insulin Resistance (HOMA-IR), waist circumference, high-density lipoproteins (HDL), triglycerides, and metabolic syndrome (MetS). Multivariable linear and logistic regression models were used to assess these associations. Third, the Brief Symptom Inventory-18 (BSI-18) was used to investigate acute symptoms of anxiety and depression. Multivariable linear and logistic regression models were used to evaluate the associations between the inflammatory potential of diet and symptoms of anxiety and depression. Fourth, we examined the associations between the E-DII and hypertension, hyperlipidemia, cardiomyopathy, coronary artery disease (CAD), cardiac dysrhythmia, and cerebrovascular accidents (CVA). Multivariable logistic models were used to investigate the associations between the E-DII and CVD outcomes. Results: Multivariable analysis revealed that the inflammatory potential of diet is associated with insulin resistance (HOMA-IR) (ORQuartile 4vs1 = 1.28, 95% CI: [1.13, 1.44]) and HDL cholesterol (ORQuartile 4vs1, = 1.23, 95 CI: [1.09, 1.38]). Multivariable analysis also revealed that diet-associated inflammation was not significantly associated with hs-CRP in this population. Multivariable analysis revealed that the E-DII was not associated with symptoms of anxiety. In contrast, multivariable analysis revealed that depressive symptoms were significantly associated with the E-DII as a continuous and categorical variable (OR =1.07, 95% CI: [1.02, 1.13]; ORQuartile 4vs1 = 1.63, 95% CI: [1.19, 2.23], respectively). Age-adjusted logistic models revealed that the E-DII as a continuous variable was associated with hypertension (OR=1.06, 95% CI: [1.03, 1.10], and CAD (OR=1.14, 95% CI: [1.05, 1.24]. Age-adjusted logistic models revealed that a pro-inflammatory diet was associated with hypertension (ORQuartile 4vs1 = 1.46, 95% CI: [1.17, 1.81]), and CAD (ORQuartile 4vs1 = 1.97, 95% CI: [1.16, 3.35]). Multivariable logistic models revealed that the E-DII, as a continuous variable, was associated with a significantly lower odds of hyperlipidemia (OR=0.91, 95% CI: [0.86, 0.97]). Multivariable analysis showed that a pro-inflammatory diet was not significantly associated with hypertension, hyperlipidemia, cardiomyopathy, CAD, cardiac dysrhythmia, or CVA. Conclusion: Results for the first aim is consistent with our hypothesis that diet plays an important role in regulating insulin resistance and HDL cholesterol levels. Results from the second aim indicates that diet-associated inflammation plays an important role in regulating depressive symptoms. Finally, results for the third aim reveals inconsistencies in the associations between the inflammatory potential of diet and CVD outcomes. Previous literature has shown that insulin resistance, low HDL levels, and depressive symptoms have the potential to increase the risk of CVD outcomes among childhood cancer survivors and controls. Using intervention strategies that promote an anti-inflammatory diet to decrease the risk insulin resistance, HDL, and depression has the potential to decrease incidence and mortality from CVD observed among childhood cancer survivors. Analyses of longitudinal data collected over long time periods are needed to better understand these associations. These studies and, perhaps, others need to investigate biological mechanisms behind these associations.

Included in

Epidemiology Commons

Share

COinS