Obesity

Polygenic prediction of body mass index and obesity through the life course and across ancestries

Nat MedPublished Genetic association and prediction study

Authors' conclusion

Overall, these data show that PGSs have the potential to improve obesity prediction, particularly when implemented early in life.

Study context

Population
Up to 5.1 million people (4.6% African, 14.4% American, 8.4% East Asian, 71.1% European, 1.5% South Asian ancestry) from GIANT consortium and 23andMe; validation in UK Biobank, ALSPAC and lifestyle intervention trialsup to
Participants
5.1 million participants
Outcome
BMI variation explained by multi-ancestry PGS
Methodological detailsmulti-ancestry, ancestry-specific and multi-ancestry PGSs
  • Multi-ancestry
  • Ancestry-specific and multi-ancestry PGSs

Main finding

As reported in the abstract

UK Biobank participants of European ancestry

17.6%

East Asian-Americans

16%

Rural Ugandans

2.2%

Further findings

Childhood body mass index (BMI) trajectory (ALSPAC) · age 2.5 years to adolescence

Higher PGS: accelerated BMI gain and earlier adiposity rebound

Explained variance for BMI, predictors at birth ± PGS · age 8

Adding PGS nearly doubled explained variance from age 5 onward; at age 8
from 11% to 21%

Prediction of BMI at age 18, early-life BMI ± PGS

Adding PGS improved prediction up to age 5; at age 5
from 22% to 35%

Adult weight gain

Higher PGS associated with greater adult weight gain

Weight loss in intensive lifestyle intervention trials · first year

Higher PGS: modestly more weight lost, per s.d.
0.55 kg
Higher PGS: more likely to regain weight

Original publication

Polygenic prediction of body mass index and obesity through the life course and across ancestries. Nat Med. Published 21 July 2025. doi:10.1038/s41591-025-03827-z

Composed by AI from the published abstract and checked against it. How Briefs are made →

Get them every week

New studies on this topic in My Week and in your Monday e-mail.

We e-mail you a sign-in link and a code. No password, no spam.