Table of Contents
New research links previously unknown genes to frailty, raising hope for early predicting and preventing age-related decline.
Study: Large-scale genome-wide analyses with proteomics integration reveal novel loci and biological insights into frailty. Image credit: Unai Huizi Photography/Shutterstock.com
A new study led by researchers from the Karolinska Institutet, Sweden, revealed genetic variants associated with the development of frailty in older people. The study, which is published in Nature Aging, provides new insights into the etiology of frailty.
Background
Aging is a complex process characterized by gradually deteriorating physiological functions necessary for survival and fertility. Frailty is a clinically relevant aging phenotype in which the body loses its resilience and becomes more vulnerable to falls, infections, and other stresses. Increased frailty may also contribute to the risk of hospitalization and death.
Currently, there is no gold standard to measure frailty. However, several assessment scales, including the Hospital Frailty Risk Score (HFRS), have been developed to identify at-risk populations. The HFRS is a relatively new scale to measure frailty, which overlaps with existing frailty definitions such as the frailty index (an in-depth overview of overall health) and the frailty phenotype (an in-depth overview of specific physical characteristics such as weakness, slowness, exhaustion, low physical activity, and weight loss).
Previous studies investigating causative factors related to frailty have focused on frailty index and phenotype models and identified specific genetic variants that increase the risk of frailty. The current study is the first to evaluate the genetics of frailty using the HFRS.
The study
The researchers conducted a genome-wide association study of the HFRS in FinnGen to identify frailty-associated genetic variants. FinnGen is a large national genetic resource containing genomes and health data from over 500,000 Finnish biobank donors. It is used to understand the genetic basis of diseases.
They replicated the significant variants in the UK Biobank, with a sample size of over 400,000 genomes, both at the individual variant level and through polygenic risk scores (PRSs). A meta-analysis of the FinnGen and UK Biobank results followed this. The HFRS-PRSs were calculated based on FinnGen’s genome-wide association study summary statistics. These results were then assessed for their association with mortality and hospitalizations in the UK Biobank.
They performed protein association and colocalization analyses to prioritize genes and identify causal variants.
Key findings
The study identified 53 significant genetic variants associated with frailty, 45 of which were novel and not previously reported for any trait. The variants mapped to 41 genes, 6 of which were novel. When examining replication, about 6% of the lead variants replicated at a strict genome-wide significant (P
The colocalization analysis identified several causal genes, including CHST9, C6orf106 (ILRUN), KHK, MET, APOE, CGREF1, and PPP6C. Among these genes, CHST9 encodes an enzyme essential for cell–cell interactions and signal transduction; C6orf106 (ILRUN) is a regulator of inflammation and lipid metabolism; CGREF1 is associated with cell cycle regulation and adhesion; APOE is significantly associated with Alzheimer’s disease; and PPP6C is involved in nuclear factor-κB pathway regulation.
Despite functional diversities, the C6orf106 (ILRUN), CHST9, CGREF1, and PPP6C genes collectively link immunoinflammatory modulation, cellular interactions, and cell adhesion to frailty.
The protein expression analysis revealed that elevated levels of CGREF1 and NECTIN2 and reduced levels of MET and APOC1 are associated with higher HFRS scores. Existing evidence has linked higher levels of NECTIN2 to Alzheimer’s disease and lower levels of APOC1 to cognitive decline and frailty. However, no studies have linked CGREF1 or MET to frailty, highlighting a novel association.
The cell-type enrichment analysis revealed higher expression of identified genes in various brain tissues, including the limbic system, cerebrum, visual cortex, cerebellar hemisphere, and cerebellum. These findings highlight the central nervous system’s involvement in frailty development.
The study found that the HFRS-PRSs effectively predict the risk of frailty, early-onset frailty, mortality, and hospitalizations. Since frailty develops at relatively older ages for most individuals, PRS-mediated risk assessment may help mitigate frailty at early stages through effective interventions. The study estimated the heritability of frailty as measured by HFRS to be about 6%, like previous estimates for other frailty measures.
Overall, the study reveals new genetic contributions to frailty and sheds light on its biological basis. This would help identify at-risk individuals as early as middle age, when there is still time to prevent frailty.
The study findings also support existing evidence on the involvement of immunoinflammatory and nervous system functions in the etiology of frailty. Future studies should explore the role of these functions in the development of cognitive frailty.
The study used clinical diagnoses in register data to define frailty. This approach has both advantages and disadvantages. One significant advantage is that public healthcare in Finland and the United Kingdom is primarily tax-funded, and each citizen has equal access. However, one disadvantage is that these registers may underreport in some conditions, or there may be a delay between the onset of symptoms and when a formal diagnosis is made for some conditions.
Moreover, the study found weaker genetic associations in the UK Biobank than FinnGen, which might be due to the differences in the enrollment process. In the UK Biobank, participation is voluntary, whereas FinnGen involves national cohorts and biobank samples of hospitalized individuals. The authors note that the generally lower prevalence of frailty in the UK Biobank may also have contributed to the lower replication rates in the dataset.
