NEW YORK (GenomeWeb News) – A large international research team has tracked down a slew of genomic sites with ties to osteoporosis, including more than a dozen loci with additional links to bone fracture risk. The study was reported online yesterday in Nature Genetics.
With access to data on nearly 33,000 European, North American, East Asian, and Australian individuals collected through 17 past genome-wide association studies, the researchers did a discovery meta-analysis looking for variants linked to bone mineral density, a key indicator of osteoporosis that's used to predict the risk of bone fractures.
After replication studies involving tens of thousands more individuals, the team was left with 32 new and 24 known risk loci for bone mineral density. In the mix: 14 loci that appeared to contribute to fracture risk in follow-up analyses involving more than 130,000 affected individuals and unaffected controls.
"We're learning that the genetic architecture of disease is very complex," co-senior author John Ioannidis, director of the Stanford Prevention Research Center, said in a statement, adding that this presents a challenge for those interested in applying the genetic information for clinical or risk prediction purposes.
"[T]he next step of incorporating this information into basic patient care is not clear," he said. "Each variant conveys a small quantum of risk or benefit. We can't predict exactly who will or won't get a fracture."
Although around two-dozen sites in the genome have been tied to bone mineral density in previous GWAS, the researchers explained, none of the loci have been conclusively linked to bone fracture, despite nominal associations found for a few of the bone mineral density-associated loci.
To explore the genetics of osteoporosis and fracture risk in more detail, collaborators involved in the current effort started by bringing together data from 17 GWAS done on five continents.
Through a meta-analysis of data at roughly 2.5 million genotyped or imputed SNPs for 32,961 individuals assessed for femoral neck bone mineral density and 31,800 assessed for lumbar spine bone density, investigators found 82 autosomal SNPs with apparent ties to one or both of the bone mineral density measures.
After looking at X chromosome markers and doing additional conditional analyses, they added another 14 suspicious SNPs to take forward to replication testing.
They then genotyped all 96 of these SNPs in another 50,933 cases and controls and did further meta-analyses with data from both the discovery and replication groups. Sixty-four of the SNPs could be replicated in these steps, with 56 showing significant associations with bone mineral density, including two-dozen found before and 32 new variants.
When the team used the same 96 SNPs taken forward from the discovery analysis to look at fracture risk, using data from 31,016 individuals who had experienced fractures and more than 102,000 unaffected controls, it found 14 significantly associated sites in the genome. Six of these sites — in and around the FAM210A, SLC25A13, LRP5, MEPE, SPTBN1, and DKK1 genes — were strongly associated with the risk of bone breakage.
They also developed a genetic model based on some of the SNPs falling out of the meta-analyses that could correctly classify post-menopausal women from a Danish study into higher or lower osteoporosis risk groups.
While the model had only limited ability to predict an individual's osteoporosis or fracture risk, researchers explained, it supported the notion that the loci identified are relevant to understanding disease biology.
"[T]he predictive power of the study for individuals is relatively low: those with multiple risk-increasing variants are only about three to four times more likely than those with the fewest variants to have lower bone mineral density and experience fractures," Ioannidis said.
In their subsequent functional analyses of the bone mineral density and fracture risk associated SNPs, combined with expression quantitative trait locus analyses, the investigators found clues that some of the loci affect genes in bone-related pathways, including genes from ossification and Wnt signaling pathways.
Still, study authors cautioned that the overall biology of osteoporosis appears to be complex and likely includes some genetic contributors that are specific to men and women and to the affected site in the body.