
NEW YORK – An international team led by investigators at the University of Gothenburg has focused in on blood plasma protein patterns associated with the risk of future hip fractures.
In the process, the investigators developed a proteomic risk score that appears to complement established clinical risk fracture prediction strategies such as the "fracture risk assessment tool" (FRAX) algorithm. Their findings appeared in Nature Aging on Monday.
"No omics-derived factor[s] have previously been shown to improve hip fracture beyond the clinical[ly] used fracture prediction algorithms," senior and corresponding author Claes Ohlsson, an internal medicine and clinical nutrition researcher at the University of Gothenburg and Sahlgrenska University Hospital, said in an email, noting that the "developed protein-based risk score improves hip fracture prediction beyond" these algorithms.
As part of the Cardiovascular Health Study, the researchers relied on the SomaLogic SomaScan 5K aptamer platform to profile blood plasma proteins from 3,171 individuals with an average age of 74 years, including 456 individuals who experienced incident hip fractures.
The team reasoned that proteins circulating in the blood can provide a window into risk of such fractures and related health conditions such as osteoporosis, since the proteome is influenced by not only an individual's genetic features but also environmental exposures.
"Protein profiles are dynamic and may integrate information on genetic variations and environmental factors," the authors explained, noting that these proteomic patterns "also reflect ongoing biological processes and may, thereby, reflect current health status and disease risk."
Using the Cardiovascular Health Study data, the team came up with three hip fracture-associated proteomic risk scores, narrowing in on a weighted score that appeared to outperform the machine learning-based alternative scores.
The researchers went on to validate their weighted proteomics score using SomaScan 5K or 7K data for thousands more individuals from the Trøndelag Health Study (HUNT), including 342 individuals with hip fractures and more than 4,900 unaffected participants, and in 50,876 UK Biobank participants with or without hip fracture who had plasma profiling with an Olink double antibody proximity extension assay approach.
"We developed a proteomic risk score that improved hip fracture prediction and discrimination in three independent validation cohorts, analyzed by two substantially different proteomic platforms," the authors wrote, noting that the proteomic score "predicted hip fractures similarly in both young and old participants, and in both men and women."
Along with follow-up Mendelian randomization analyses aimed at untangling causal contributors to the hip fracture risk process, the team went on to assess the hip fracture prediction capabilities of the proteomic signature in relation to existing clinical methods such as the FRAX algorithm, in contexts where estimated bone mineral density (eBMD) was or was not available.
While the investigators noted that further research is needed to assess the proteomic risk score in a clinical setting, including cost-effectiveness analyses, they noted that it appeared to provide additional fracture risk clues compared to prediction capabilities associated with previously reported polygenic risk scores (PRSs) for hip fracture, femoral neck bone mineral density, or bone mass.
"When added to FRAX with or without information on eBMD, the proteomic risk score, but not available PRSs, improved fracture discrimination," the authors reported. "We propose that the developed proteomic risk score is a biomarker candidate to be included as a new risk marker in future updates of the fracture prediction tool FRAX."
Even so, they noted that hip fracture prediction improvements associated with the plasma proteomic risk score appeared relatively modest so far, and suggested that "its clinical utility beyond FRAX with information on femoral neck bone mineral density remains to be determined."