Background & Aims
Low back pain (LBP) is a complex and debilitating condition that affects a significant proportion of the population globally, leading to significant individual and societal burdens[1]. Understanding the genetic factors contributing to chronic LBP susceptibility is crucial to elucidating its underlying mechanisms and developing personalized interventions.
By exploring various phenotyping approaches, we aimed to leverage genome-wide association (GWA) data from large-scale biobanks, the United Kingdom Biobank (UKB) and the Canadian Longitudinal Study on Aging (CLSA) to develop and validate a polygenic risk score (PRS) for chronic LBP.
Methods
We performed GWA scans using Regenie v3.1.2 for each phenotype based on 3 overlapping samples from the UKB comprising participants with 1. chronic back pain at baseline & first follow-up (FUP1) (cases: n=64,689;controls: n=145,124); 2. any chronic pain site at baseline & FUP1 (cases: n=161,117;controls n=145,241); 3. >=2 chronic pain sites known as multisite chronic pain (MCP), at baseline & FUP1 (cases: n=70,803;controls: n=145,315). Controls for each group were those with no pain site at baseline & FUP1. Post-GWAS filtering included Hardy-Weinberg equilibrium (HWE P<1.0e-12) and minor allele count (MAC<100). We computed SNP heritabilities using LDSC v1.0.1. We then derived PRSs and validated using clumping + thresholding. We defined chronic back pain in the CLSA as those who had back pain >3 months, & controls were those usually free of pain or discomfort & had no history of back pain lasting >=1 month(cases: n=5,109;controls: n=13,577). At both stages, we adjusted for covariates.
Results
Heritability estimates for the 3 phenotypes were as follows: 0.0904 (0.0044) for chronic back pain; 0.0699 (0.003) for any chronic pain; and 0.1053 (0.0044) for MCP. The highest pseudo-R2 values for the full models were 3.03%, 2.88%, 3.05% respectively, at individual marker association test P-value thresholds of 0.4 for chronic back pain, 0.5 for any chronic pain, and 0.3 for MCP PRS. Their AUC values (95% CI) were 0.595 (0.583-0.606), 0.592 (0.581-0.603), 0.595 (0.584-0.606) respectively.
Conclusions
Although MCP showed slightly better AUC, all three PRSs had comparable performance in the CLSA.
References
1.Global, regional, and national incidence, prevalence, and years lived with disability for 310 diseases and injuries, 1990-2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet, 2016. 388(10053): p. 1545-1602.
Presenting Author
Carolina Beraldo Meloto
Poster Authors
Topics
- Genetics