BMI Deviations from Genetic Prediction Are Associated with Distinct Molecular Signatures in Finnish Population Cohorts
- 1Institute for Molecular Medicine Finland (FIMM), Helsinki Institute of Life Science (HiLIFE), University of Helsinki, Helsinki, Finland; Institute of Biotechnol
Abstract
ObjectivePolygenic risk scores (PRS) estimate genetically predicted BMI, but lifestyle and environmental factors may cause deviations between observed and predicted BMI. We investigated whether such deviations are associated with distinct molecular profiles and whether these differ by direction of deviation.
MethodsWe analyzed 2,508 adults from the Finnish Twin Cohort and 1,715 from the Young Finns Study with clinical, metabolomic, DNA methylation, and proteomic data. PRS-informed residual BMI (rBMI) quantified deviations between observed and genetically predicted BMI. Associations between rBMI and multi-omic profiles were assessed using linear mixed-effects models with a hinge term to compare effects above vs. below genetic prediction. Twin analyses further accounted for shared genetic and environmental factors.
ResultsrBMI is associated with distinct cardiometabolic and molecular profiles depending on whether BMI is above or below genetic prediction. When BMI exceeds genetic prediction, rBMI shows stronger associations with inflammatory and cardiovascular markers. In contrast, several metabolites and plasma proteins show stronger associations when BMI is below genetic prediction. Epigenetic aging increases with higher rBMI regardless of direction. Similar patterns are observed in BMI-discordant twin pairs.
ConclusionDeviations from genetically predicted BMI are associated with distinct clinical and molecular signatures beyond absolute BMI alone.
Related
- Transcriptomic Profiling of Oxidative Stress-Associated Immune and Apoptotic Signatures and In Silico Prediction of Therapeutic Candidates for Sjögren’s Syndrome Dry Eye
- Nicotinamide Mononucleotide Is Associated with Attenuation of OGD/R-Induced Pro-Inflammatory Programs in Microglia: Transcriptomic Evidence Converging on the NLRP3 Inflammasome and Lipid-Droplet-Related Molecular Signatures
- Proteomic machine-learning signatures associated with carotid plaque: A population-based hypothesis-generating study
- Omics-derived signatures in metabolic dysfunction-associated steatotic liver disease: from molecular characterization to clinical translation.
- Candidate molecular endotypes and dynamic risk prediction in anti-MDA5-positive dermatomyositis-associated interstitial lung disease: a systematic scoping review
BibTeX
@misc{zhu2026deviations,
title = {{BMI Deviations from Genetic Prediction Are Associated with Distinct Molecular Signatures in Finnish Population Cohorts}},
author = {T. Zhu and G. Drouard and A. Heikkinen and M. Hukkanen and T. Palviainen and B. J. Berntzen and L.-P. Lyytikainen and E. Raitoharju and K. H. Pietilainen and T. Lehtimaki and O. Raitakari and M. Ala-Korpela and J. Kaprio and R. Cristofari and M. Ollikainen},
howpublished = {medRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.21.26366088},
url = {https://www.medrxiv.org/content/10.64898/2026.10.21.26366088v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - BMI Deviations from Genetic Prediction Are Associated with Distinct Molecular Signatures in Finnish Population Cohorts
AU - T. Zhu
AU - G. Drouard
AU - A. Heikkinen
AU - M. Hukkanen
AU - T. Palviainen
AU - B. J. Berntzen
AU - L.-P. Lyytikainen
AU - E. Raitoharju
AU - K. H. Pietilainen
AU - T. Lehtimaki
AU - O. Raitakari
AU - M. Ala-Korpela
AU - J. Kaprio
AU - R. Cristofari
AU - M. Ollikainen
PY - 2026
DA - 2026/10/21
PB - medRxiv
DO - 10.64898/2026.10.21.26366088
UR - https://www.medrxiv.org/content/10.64898/2026.10.21.26366088v1
AB - ObjectivePolygenic risk scores (PRS) estimate genetically predicted BMI, but lifestyle and environmental factors may cause deviations between observed and predicted BMI. We investigated whether such deviations are associated with distinct molecular profiles and whether these differ by direction of deviation. MethodsWe analyzed 2,508 adults from the Finnish Twin Cohort and 1,715 from the Young Finns Study with clinical, metabolomic, DNA methylation, and proteomic data. PRS-informed residual BMI (rBMI) quantified deviations between observed and genetically predicted BMI. Associations between rBMI and multi-omic profiles were assessed using linear mixed-effects models with a hinge term to compare effects above vs. below genetic prediction. Twin analyses further accounted for shared genetic and environmental factors. ResultsrBMI is associated with distinct cardiometabolic and molecular profiles depending on whether BMI is above or below genetic prediction. When BMI exceeds genetic prediction, rBMI shows stronger associations with inflammatory and cardiovascular markers. In contrast, several metabolites and plasma proteins show stronger associations when BMI is below genetic prediction. Epigenetic aging increases with higher rBMI regardless of direction. Similar patterns are observed in BMI-discordant twin pairs. ConclusionDeviations from genetically predicted BMI are associated with distinct clinical and molecular signatures beyond absolute BMI alone.
ER -