07/17/2025
By Kaylea Flanagan

The Zuckerberg College of Health Sciences, Department of Public Health, invites you to attend a doctoral dissertation defense by Kaylea Flangan, MPH, RD on “Ultra-Processed Food Consumption, Cognitive Health, and Diabetes Control among Puerto Rican Adults.”

Candidate Name: Kaylea Flanagan, MPH, RD
Degree: Doctoral 
Defense Date: Thursday, July 24, 2025
Time: 3:30-5 p.m.
Location: Virtual via Zoom 
Thesis/Dissertation Title: “Ultra-Processed Food Consumption, Cognitive Health, and Diabetes Control among Puerto Rican Adults”

Committee: 

  • Advisor: Sabrina E. Noel Ph.D. R.D., Department of Public Health and Biomedical and Nutritional Sciences, University of Massachusetts Lowell
  • Katherine L. Tucker, Ph.D., Department of Biomedical and Nutritional Sciences, University of Massachusetts Lowell
  • Tammy Scott, Ph.D., Friedman School of Nutrition Science and Policy, Tufts University
  • Karyn Heavner, Ph.D., Department of Public Health, University of Massachusetts Lowell

Brief Abstract:
Ultra-processed foods (UPF) consumption has been on the rise and have been linked to several health outcomes, such as cognitive decline and type II diabetes (T2D). Many research studies on UPF have been conducted in non-Hispanic white populations, and among the studies conducted on Hispanic individuals, studies are limited on Puerto Rican adults, a population with high prevalence of cognitive impairment and T2D. Better understanding risk factors, such as diet, and measurable biological measures associated with UPF intake, are imperative among this population. This project aims to examine associations between UPF intake, associated metabolic pathways, cognition, and T2D control among Puerto Rican adults living in MA by (1) analyzing associations between UPF intake and cognitive measures (executive function, memory, and global cognitive score (GCS)), and brain MRI measures (brain age, total brain volume, hippocampal volume, and white matter hyperintensities), (2) assessing the relationship between UPF intake and insulin resistance (using the homeostatic model for insulin resistance (HOMA-IR)), glycemic control (hemoglobin A1c (HbA1c)), and incident T2D, and (3) determining plasma metabolites positively associated with UPF and how they are related to cognitive health and T2D.

Both cross-sectional and longitudinal analyses were conducted using data from the Boston Puerto Rican Health Study (BPRHS). Nutritional data was obtained via food frequency questionnaire (FFQ), and the NOVA classification system was utilized to identify UPF. Three factors of UPF identified using principal component analysis (PCA) were also used as primary predictors in some analyses. Multivariable linear and logistic regression analyses were utilized to assess the association between UPF intake and cognitive measures (memory, executive function, GCS, binary decline in these scores) and MRI outcomes (brain age, total brain volume, hippocampal volume, and white mattery hyperintensities) (Aim1). Multivariable linear and logistic regression analyses were also utilized to assess the association between UPF intake and continuous T2D control measures (HOMA-IR, HbA1c), HOMA-IR ≥3.8, HbA1c ≥6.5%, and incident T2D (Aim 2). Lastly, multivariable linear regression analyses were used to assess relationships between plasma metabolites and UPF. Metabolites that were positively associated with UPF factors were then utilized as primary predictors in regression models with cognitive measures and T2D (Aim 3).

Mean intake of UPF was lower in the BPRHS (30-31% kcal/d). Generally, those with T2D at baseline tended to have poorer cognition. Longitudinal models indicated that those with T2D at baseline significant decline in executive function and borderline significant GCS decline at wave four. No significant associations were observed between UPF and MRI outcomes among those with T2D at baseline. UPF was associated with minimal percentage of change in continuous HOMA-IR and HbA1c. Those without T2D who consumed more UPF and had a BMI ≥30 kg/m2 had higher odds of insulin resistance at wave two. Those without T2D and who were at least 60 y old at baseline also had higher odds of poor glycemic control at wave two. Adherence to UPF patterns and associations with T2D control-related measures varied. Among those without T2D, adherence to the desserts and snacks pattern was associated with lower odds of poor glycemic control at wave three. Those with T2D and the highest intakes of desserts and snacks at baseline also had lower odds of poor glycemic control at wave three, yet higher intakes of mixed dishes/fast and fried foods had higher odds of poor glycemic control at wave three among this group. Higher intake of mixed dishes/fast and fried foods was also associated with increased odds of incident T2D. Significant associations between 19 metabolites and UPF factors (8 with desserts and snacks, 7 with mixed dishes/fast and fried foods, and 4 with ready-to-eat diet and meal replacement products) we observed. No significant findings were observed between metabolites positively associated with UPF and baseline cognition, however, of the 12 metabolites that were positively associated with UPF patterns, 6 were associated with odds of T2D.

Although UPF intake was lower in the BPRHS compared to other study populations, intakes and adherence to certain UPF patterns were associated with various cognitive and T2D-related health outcomes. Metabolomic analyses indicated biological pathways that may be associated with UPF. Findings from this work may provide inference for future dietary recommendations and/or biological measures associated with UPF, however future studies are needed to confirm these associations.