AI Atlas Reveals How Body Fat Distribution Predicts Deadly Health Risks

A new AI-powered study has created a detailed atlas of human body composition, revealing clear links between fat distribution and serious health risks. Researchers analysed data from 66,608 participants, uncovering how muscle and fat levels influence disease and mortality. The study used deep learning algorithms to measure fat and muscle from whole-body MRI scans. It tracked subcutaneous fat, visceral fat, skeletal muscle volume, and muscle quality across different ages, sexes, and heights. The findings challenge the reliability of BMI as a health indicator.

Reference curves were generated to show how body composition changes with age. These curves, separated by sex and height, provide a clearer picture than traditional measurements. The AI framework behind this work is open-source and automated, meaning it can be applied to routine CT or MRI scans in clinical settings. Key results highlighted major health risks tied to fat levels. High intramuscular fat raised the likelihood of cardiovascular events by 1.54 times. Visceral fat was linked to a 2.26-fold increase in diabetes risk. Meanwhile, low skeletal muscle mass independently predicted a 1.44-fold higher chance of death from any cause.

The atlas offers precise, AI-driven insights into how body fat and muscle affect long-term health. Clinicians can now use this automated tool on standard imaging scans to better assess patient risks. The findings may lead to more personalised approaches in preventing diabetes, heart disease, and early mortality.