Weight-related changes in MRI-derived measures of body composition and liver steatosis: a large-scale analysis for obesity trial design
An erratum to this article was published (Int J Obes (Lond), May 2026 ↗)
Authors' conclusion
Using MRI, relative changes in body composition and liver steatosis resulting from weight loss can inform clinical trials, including placebo arm design and power estimations. Weight changes affect body composition and liver fat and can inform obesity trial design.
Study context
- Population
- Adults from the UK Biobank with repeat abdominal MRI scans 2.7 years apart; age 63 years, male 49%, 13% with obesity, 43% with overweight
- Participants
- 3,070 adults
- Exposure
- Weight change categories vs stable weight
- Outcome
- MRI-derived body composition and liver fat changes by weight change
- Follow-up
- 2.7 years
Methodological detailsUK Biobank, repeat abdominal MRI
- UK Biobank
- Repeat abdominal MRI
Main finding
As reported in the abstract
2.7 years
- Weight gain occurred
- 28%
- Weight loss occurred
- 28%
- Moderate or large weight gain increased liver fat content (LFC), VAT, subcutaneous adipose tissue (SAT), MFISM and SM
- all p < 0.001
- Moderate or large weight loss decreased LFC, VAT, SAT, SM, SMI and psoas cross-sectional area (CSA)
- all p < 0.001
Observational data: associations, not effects.
Further findings
Per 5% drop in weight, overweight or obesity
- Reduction in VAT
- 16%
- Reduction in SAT
- 11%
- Reduction in liver fat
- 24%
- Reduction in MFISM
- 2.3%
- Reduction in SM (or SMI)
- 1.5% (1.4%)
- Reduction in psoas CSA
- 2.1%
Modifiers of VAT and liver fat change
- Prevalent T2D exacerbated weight gain-related increases; blood pressure medication attenuated the impact of weight change
DXA lean mass change vs SM volume change
- Weak correlation (rho)
- 0.28-0.47
Original publication
Weight-related changes in MRI-derived measures of body composition and liver steatosis: a large-scale analysis for obesity trial design. Int J Obes (Lond). Published 3 March 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →