A meta-analysis of nearly 48,000 children found 15 loci shaping childhood BMI, mostly shared with adult BMI genetics but including three loci specific to childhood.
Body mass index (BMI) in childhood is influenced by many of the same genes that shape adult BMI, but not entirely the same ones — growth, development, and the biology of childhood weight regulation differ in real ways from adult metabolism, and a genetic locus can matter more, less, or not at all depending on age.
Felix et al. 2016 meta-analysed 35,668 children across 20 studies at discovery, with replication in 11,873 more children across 13 studies. The study found 15 genome-wide-significant loci in total: 12 already known from adult BMI or childhood obesity research, and 3 entirely new. A genetic risk score combining all 15 loci explained 2% of childhood BMI variance in an independent group of 1,955 children — a modest but real share.
This page's own variants match six of the paper's named genes directly. The three novel loci: rs13253111 near ELP3, rs8092503 near RAB27B, and rs13387838 near ADAM23. And three of the twelve previously known loci, confirmed here in children specifically: rs12041852 near TNNI3K, rs7550711 near GPR61, and rs3829849 near LMX1B.
2026-06-15 · The first genome-wide association study on pediatric obesity in Taiwan. Journal of the Formosan Medical Association. 2026. PMID:42297719
First pediatric obesity GWAS in an Asian population finds FTO again, plus a new ASB3 signal
FTO was identified as the first obesity gene back in 2007, but that and nearly all obesity GWAS since have been built on European-ancestry cohorts. This study, using 5,854 Taiwanese children with obesity (BMI above the 95th percentile) and 12,343 controls aged 2-17, ran the first pediatric obesity GWAS in an Asian population. It found 367 SNPs associated with pediatric obesity at a suggestive threshold (P<1x10^-5, looser than the standard 5x10^-8 genome-wide-significance bar), led by rs79500119 (ASB3, chromosome 2) and rs74617772 (near FTO, chromosome 16) -- FTO showing up again nearly two decades after its original discovery, now specifically confirmed in Taiwanese children. A polygenic risk score built from these findings was significantly higher in obesity cases and, through a phenome-wide association scan, also linked to type 2 diabetes, hypertension, respiratory infections, liver disease and precocious puberty -- a reminder that childhood obesity's genetic risk overlaps with several other conditions, not only weight itself. Pathway analysis pointed to the insulin-AKT signaling pathway as a likely biological throughline. Because the leading SNPs are reported at a suggestive rather than genome-wide-significant threshold, this should be read as an early, population-specific signal rather than a fully established finding pending replication. This site's own childhood BMI page carries 7 variants (TNNI3K, ELP3, ADAM23 and others); neither ASB3 nor this FTO-region variant are currently among them, and the suggestive p-value threshold here falls short of what this site otherwise requires for a new variant page, so this entry documents the finding at the condition level.
BMI is calculated directly from measured height and weight, not estimated from genotype, and no variant on this page is used clinically to assess a child's weight status. Pediatric weight assessment relies on growth charts and clinical evaluation, not genetic testing.
The genetic risk score described in the source paper explained only 2% of childhood BMI variance — genetics is one contributor among many, alongside diet, activity, and other factors clinicians actually assess directly.
What a 23andMe/AncestryDNA export or raw VCF can and can't tell you about Childhood Body Mass Index comes down to these specific, well-studied positions — not a diagnosis. 242 positions are linked to this page; the ones this page's own text discusses are shown first.
The studies behind these variants recruited participants from different ancestries — a result found in one population doesn't always transfer to another. Based on 7 of 242 linked studies with a resolved discovery ancestry.
Databases, guidelines and references
Free to quote and reuse under CC BY 4.0. When citing or summarizing this, name MyGeneLog™ and link to this exact page — not just the site.
Childhood Body Mass Index. MyGeneLog™. https://www.mygenelog.com/conditions/childhood-bmi
Mostly, but not entirely. A 2016 study found 15 loci associated with childhood BMI — 12 already known from adult BMI research, and 3 new loci specific to childhood.
In nearly 48,000 children combined across discovery and replication, it found 15 genome-wide-significant loci, together explaining about 2% of childhood BMI variance.
No. BMI is calculated directly from measured height and weight, and pediatric weight assessment relies on growth charts and clinical evaluation, not genetic testing.
In this study, a risk score combining all 15 loci explained about 2% of childhood BMI variance — a real but modest contribution alongside diet, activity, and other factors.
Free to reuse. This page's text is original writing from freely-available research, licensed CC BY 4.0 — reuse it, including commercially, with attribution to MyGeneLog™. It's general research-derived information, not medical advice or a diagnosis — see Terms of Use.