Childhood obesity is a global public health concern, but the long-term causal effects of early-life adiposity on adult regional fat distribution and the biological pathways involved are incompletely understood. The authors aimed to test whether childhood BMI causally influences specific adult adipose depots and to evaluate whether circulating plasma proteins mediate any observed depot-specific relationships.
The study applied a two-sample Mendelian randomisation (MR) framework. Genetic instruments for childhood BMI were derived from published supplementary data (Helgeland et al., 2022) and were available across 12 childhood timepoints spanning ages 3 to 18 years. The authors used these instruments to obtain genetically predicted childhood BMI at successive ages and then tested causal effects on adult fat phenotypes.
Two-step MR methods were implemented to assess proteomic mediation: first to identify proteins influenced by childhood BMI during a developmentally sensitive window, and second to test whether those proteins causally affected adult fat depots in a direction consistent with mediation.
Adult adipose phenotypes were MRI-derived measures available in UK Biobank. Five regional fat depots were examined:
Two-sample MR analyses used public GWAS summary statistics for liver and pancreatic fat (NHGRI-EBI GWAS Catalog accession numbers reported) and ASAT, GFAT, and VAT GWAS summary statistics available from the Cardiovascular Disease Knowledge Portal.
The authors integrated pQTL summary statistics from the UK Biobank Pharma Proteomics Project (UKB-PPP; Olink platform) to examine 2,940 circulating plasma proteins. The two-step MR approach first identified proteins whose circulating levels were causally influenced by genetically predicted childhood BMI during a specified developmental window, then evaluated whether those proteins causally affected adult regional fat depots in a manner consistent with mediation.
Temporal, statistical, and directional evidence across childhood ages were combined to prioritise proteins plausibly operating in a developmentally sensitive period.
Genetically predicted higher childhood BMI from approximately age 7 years onward was associated with higher adult ASAT and GFAT. These associations persisted into adolescence based on the temporal pattern of genetic instrument effects across childhood.
In contrast, the study reported little evidence that genetically predicted childhood BMI had causal effects on visceral fat, liver fat, or pancreatic fat in adulthood.
Overall, the results indicate that childhood adiposity exerts depot-specific effects on adult body composition, with a particular influence on subcutaneous adipose tissue depots.
Using the two-step MR mediation framework, the authors identified 140 plasma proteins whose circulating levels were affected by childhood BMI during a developmentally sensitive window. Of these, seven proteins showed directionally consistent evidence supporting mediation of the relationship between childhood BMI and adult subcutaneous fat depots.
Proteins implicated for abdominal subcutaneous fat included ACAN, CCL7, and CLIC5. Proteins implicated for gluteofemoral fat included CLIC5, BMP10, CLEC10A, KIT, and IGSF3. The overlap of CLIC5 across both subcutaneous depots suggests some shared biology, while other proteins indicate partially distinct pathways between depots.
The authors state that all summary statistics and data sources used are publicly available. Specific sources cited include GWAS summary statistics for liver and pancreas fat from the NHGRI-EBI GWAS Catalog, ASAT/GFAT/VAT GWAS from the Cardiovascular Disease Knowledge Portal, childhood BMI genetic instruments from Helgeland et al. (2022), and proteomic GWAS (pQTL) data from the UKB-PPP via Synapse.
Ethical declarations indicate that data were openly available prior to the study and that relevant approvals and participant consents were obtained. The manuscript is posted as a preprint and has not been peer-reviewed; the authors caution it should not guide clinical practice.
A competing interest disclosure notes that one author (Dimitri J. Pournaras) has received funding, consultancy fees, and payments from industry sources as detailed in the source text.
The analyses support a causal link between higher childhood BMI (from ~7 years) and increased adult subcutaneous fat mass (ASAT and GFAT), but do not support causal effects on visceral, liver, or pancreatic fat. The proteomic mediation results identify candidate circulating proteins that may lie on pathways connecting early-life adiposity to adult subcutaneous adipose expansion, and suggest partially distinct biological mechanisms across depots.
These findings highlight childhood as a potentially critical period for determining adult adipose tissue distribution and point to molecular targets for further mechanistic study or intervention research. As reported by the authors, this work is a preprint and has not undergone peer review; additional validation, detailed methodological appraisal, and replication in independent datasets are warranted before clinical translation.