Date of Award




Document Type


Degree Name

Doctor of Philosophy (PhD)


Department of Epidemiology and Biostatistics



Content Description

1 online resource (iv, 82 pages) : illustrations (some color), maps (some color)

Dissertation/Thesis Chair

Akiko S Hosler

Committee Members

Xiaobo X Romeiko, Thomas J O’Grady


Chronic Disease, Food Access, Food Insecurity, Food Pantry, Fruit and Vegetable Intake, NYS Captial Region, Vegetables in human nutrition, Fruit in human nutrition, Diet therapy, Food habits, Food banks

Subject Categories

Biostatistics | Epidemiology | Nutrition


Background: Chronic diseases impose heavy burdens on individuals and the healthcare system in the US. Many factors were found to be associated with chronic diseases, including demographics, family history, social environmental factors, and individual behavioral factors such as diet and physical activity. Among those factors, fruit and vegetable intake can have substantial health impacts via a variety of causal pathways. Fruit and vegetable (F&V) consumption is generally lower among individuals living in households experiencing food insecurity and rely on food assistance programs. Decreased F&V intake among food pantry users may negatively impact health. However, conducting quantitative analysis on this population is difficult due to the lack of food pantry use data in most publicly available datasets.Methods: We built a predictive model of food pantry utilization using demographic variables from 3 health survey datasets collected from the NYS Capital Region, then applied this model to the SMART-BRFSS dataset to identify potential food pantry users. Next, we analyzed the SMART-BRFSS data to assess the association between fruit and vegetable intake and three major chronic diseases (diabetes, hypertension, and obesity), focusing on the effect modification of food pantry use. Finally, we estimated the amount of fresh produce donation needed to reduce chronic disease risks at the food pantry level, factoring in food wastes at the food pantry and household levels. Results: Among 5,454 SMART-BRFSS respondents, we identified 634 potential food pantry users using demographic variables including income, race/ethnicity, education, and geographic location. Consumption of vegetables was associated with decreased odds of hypertension and a lower BMI in both food pantry users and non-users. Consumption of fruits was associated with decreased odds of diabetes. Among food panty users, consumption of fruits was associated with a greater BMI, while among food pantry non-users, it was associated with a lower BMI. We estimated that if the current amount of fresh produce donation to food pantries is doubled, a 6% risk reduction in hypertension and 3% risk reduction in diabetes would be achieved, while a risk increase in obesity would be negligible (a 0.04 unit increase in BMI.) Discussion: The predictive model has good predictive ability to identify potential food pantry users using basic demographic variables. The potential protective effects of increased F&V consumption on chronic diseases suggest that increasing the availability of F&Vs in food pantries and encouraging their consumption not only reduces hunger but also leads to improved health. At food pantry-level, the current amount of produce distributed to Capital Region food pantry users has a positive health impact by lowering the risk of chronic diseases, despite its limited amount. More action is needed to achieve the goal of "doubling the donation."