Polygenic risk for triglyceride levels in the presence of a high impact rare variant

• Published in: BMC medical genomics 2023-11, Vol.16 (1), p.1-281, Article 281
• Main Authors: Ying, Shengjie, Heung, Tracy, Thiruvahindrapuram, Bhooma, Engchuan, Worrawat, Yin, Yue, Blagojevic, Christina, Zhang, Zhaolei, Hegele, Robert A, Yuen, Ryan K. C, Bassett, Anne S
• Format: Article
• Language: English
• Subjects: 22q11.2 microdeletion; Cardiovascular diseases; Cholesterol; DNA sequencing; Genes; Genetic aspects; Genetic diversity; Genetics; Genome sequencing; Genomes; Genomics; Hypertriglyceridemia; Lipid; Lipids; Metabolism; Nucleotide sequencing; Obesity; Polygenic risk score; Psychosis; Regression analysis; Risk factors; Schizophrenia; Sex; Triglyceride; Triglycerides; Type 2 diabetes; Variables
• ISSN: 1755-8794; 1755-8794
• DOI: 10.1186/s12920-023-01717-2

Elevated triglyceride (TG) levels are a heritable and modifiable risk factor for cardiovascular disease and have well-established associations with common genetic variation captured in a polygenic risk score (PRS). In young adulthood, the 22q11.2 microdeletion conveys a 2-fold increased risk for mild-moderate hypertriglyceridemia. This study aimed to assess the role of the TG-PRS in individuals with this elevated baseline risk for mild-moderate hypertriglyceridemia. We studied a deeply phenotyped cohort of adults (n = 157, median age 34 years) with a 22q11.2 microdeletion and available genome sequencing, lipid level, and other clinical data. The association between a previously developed TG-PRS and TG levels was assessed using a multivariable regression model adjusting for effects of sex, BMI, and other covariates. We also constructed receiver operating characteristic (ROC) curves using logistic regression models to assess the ability of TG-PRS and significant clinical variables to predict mild-moderate hypertriglyceridemia status. The TG-PRS was a significant predictor of TG-levels (p = 1.52E-04), along with male sex and BMI, in a multivariable model (p.sub.model = 7.26E-05). The effect of TG-PRS appeared to be slightly stronger in individuals with obesity (BMI [greater than or equal to] 30) (beta = 0.4617) than without (beta = 0.1778), in a model unadjusted for other covariates (p-interaction = 0.045). Among ROC curves constructed, the inclusion of TG-PRS, sex, and BMI as predictor variables produced the greatest area under the curve (0.749) for classifying those with mild-moderate hypertriglyceridemia, achieving an optimal sensitivity and specificity of 0.746 and 0.707, respectively. These results demonstrate that in addition to significant effects of sex and BMI, genome-wide common variation captured in a PRS also contributes to the variable expression of the 22q11.2 microdeletion with respect to elevated TG levels.