A Machine-Learning Assisted Genetic Risk Score Identifies Improved Weight Loss After Endoscopic Sleeve Gastroplasty

Area of Science:

• Genetics and Genomics
• Obesity Medicine
• Medical Technology

Background:

• Obesity is a widespread epidemic with serious health implications.
• Endoscopic sleeve gastroplasty (ESG) promotes substantial weight loss, but outcomes vary.
• Previous research developed a machine-learning (ML) genetic risk score (GRS) for calorie intake and satiety, predicting response to anti-obesity drugs.

Purpose of the Study:

• To assess novel GRSs for emotional hunger and calories to satiation (CTS).
• To determine the predictive capability of these GRSs for weight loss following ESG.
• To evaluate the performance of high or low CTS GRS in predicting ESG outcomes.

Main Methods:

• Forty participants undergoing ESG completed genetic testing using the MyPhenome test.
• ML-assisted GRSs for high/low CTS and emotional hunger were utilized.
• Total body weight loss (TBWL) at 12 and 24 months was the primary endpoint, analyzed with LOCF, ANOVA, and Tukey's HSD.

Main Results:

• The low CTS GRS group exhibited greater TBWL compared to other groups across all time points (3-24 months).
• At 12 months, the low CTS GRS group showed significant TBWL (21.4%) compared to emotional hunger (13.7%) and high CTS GRS (14.9%) groups.
• This significant weight loss difference persisted through 24 months.

Conclusions:

• ML-assisted GRSs are linked to enhanced weight loss after ESG.
• Identifying patients with a higher likelihood of superior weight loss response can optimize ESG candidate selection.
• GRS may serve as a valuable tool for personalizing obesity treatment strategies.