Date of Award

January 2021

Document Type


Degree Name

Master of Public Health (MPH)


School of Public Health

First Advisor

Yasmmyn D. Salinas


Though obesity is typically associated with metabolic dysfunction and cardiometabolic diseases, there are both obese individuals who have favorable metabolic profiles (metabolically healthy obese; MHO) and individuals who present obesity alterations despite their normal weight (metabolically obese normal weight; MONW). Genetic studies suggest that there is a genetic component underlying these phenotypes. However, previous research has been restricted by the use of the candidate gene approach, limited choice of metabolic indicators, and low statistical power. A genome-wide association (GWA) study that examines both MHO and MONW in a single large-scale population cohort is necessary to uncover the genetic determinants of these phenotypes. This will deepen our understanding of the mechanistic underpinnings for these traits and may help uncover viable targets for the treatment of these specific phenotypes.We conducted GWA analyses of MHO, MONW, and metabolically unhealthy obesity (MUO) among 194,885 white British subjects in the UK Biobank using linear mixed models. To define the phenotypes, we considered seven metabolic indicators: lipoprotein (a), glucose, systolic blood pressure, diastolic blood pressure, triglycerides, high-density lipoprotein, and low-density lipoprotein. We categorized subjects with BMI ≥ 30 into MHO (having zero alterations for the metabolic indicators) and MUO (at least one alteration), and categorized subjects of 18.5 ≤ BMI < 30.0 into MONW (at least one alteration) and common reference group (zero alteration). Genome-wide significant signals (p < 5.0 ×10−8) were investigated among 28,335 non-British white subjects in the replication study. We then dissected the MHO and MONW-associated signals to determined which of the seven indicators were contributing to them. We identified 15 variants (reside on SLC22A2, SLC22A3, LPA genes and the TOMM40/APOE/APOC1 region) that were associated with a paradoxical relationship between adiposity and metabolic risks. Univariate analyses confirmed the paradoxical relationships and suggested that Lipoprotein (a) and HDL may play important roles in the manifestation of these genetic components.


This thesis is restricted to Yale network users only. It will be made publicly available on 06/01/2023