NEW YORK — Through a genome-wide association study of 5.4 million people, a consortium of international researchers has uncovered thousands of SNPs associated with height.
Height is highly heritable, and genetic variants are thought to account for between 40 percent and 50 percent of variance in height. A previous analysis of about 250,000 people by the Genetic Investigation of Anthropometric Traits (GIANT) consortium in 2014 identified nearly 700 genetic variants associated with height, which was thought to explain about 36 percent of trait heritability, suggesting there was "missing heritability." A further meta-analysis in 2018 of about 700,000 individuals brought that to 712 genetic loci associated with height.
In a new analysis appearing Wednesday in Nature, GIANT consortium researchers conducted a GWAS of an even larger cohort. They homed in on 12,111 SNPs associated with height that account for 40 percent of variation in height within European-ancestry populations, suggesting that, for this population, nearly all the variants that influence height may have been found.
"We have accomplished a feat in studying the DNA of over 5 million people that was broadly considered impossible until recently," co-first author Eirini Marouli, senior lecturer in computational biology at Queen Mary University of London, said in a statement. She added that "[i]f we can get a clear picture of a trait such as height at a genomic level, we may then have the model to better diagnose and treat gene-influenced conditions like heart disease or schizophrenia, for example."
The researchers' genetic analysis included up to 5.38 million individuals from 281 studies from the GIANT consortium and 23andMe. While most of the participants — more than 4 million — were of European ancestry, the cohort also included 472,730 participants of East Asian ancestry, 455,180 participants of Hispanic ethnicity and admixed ancestry, 293,592 participants of African ancestry, and 77,890 participants of South Asian ancestry.
The researchers first tested in separate meta-analyses the association between standing height and more than 1.3 million SNPs in each of those five ancestral groups, then in a trans-ancestry meta-analysis identified 12,111 SNPs associated with height, which were found at 7,209 genomic loci.
For instance, 25 height-associated SNPs co-localized to within 100 kilobases of each other on chromosome 15 near the ACAN gene, in which mutations have been found in short stature and skeletal dysplasia syndromes.
On average, these height-linked loci are about 90 kilobases in length and cover 21 percent of the human genome.
Together, SNPs within these 7,209 genomic loci can explain 40 percent of variance in height — or about all that is thought to be attributable to genetic factors — among individuals of European ancestry.
"Hence, the missing heritability of height has been found. Further genetic studies in people of European ancestry will not contribute any more information — 'saturation' has been achieved for this trait in this ancestry group," Karoline Kuchenbaecker from University College London wrote in a related commentary appearing in Nature.
However, these SNPs only explain between 10 percent and 20 percent of phenotypic variance in height in non-European populations.
"While this is an achievement, the data is still skewed to people of European ancestry, a known problem in genetic studies," Loic Yengo, a statistical geneticist at the University of Queensland, said in a statement. "There is a growing number of worldwide initiatives to collect more diverse genetic data, because it is critical to widen the benefit of genetic studies to all populations."
Still, Kuchenbaecker added that these findings could spur scientists to seek such saturation for other traits and diseases, though she noted it would be a more difficult task for complex diseases and conditions for which outcomes are tough to measure, unlike height.