rs114898510 (CDKAL1): Blood Biomarker Variant

Key takeaways

• rs114898510 sits in the CDKAL1 gene region, found in a UK Biobank genome-wide scan of 35 blood and urine biomarkers across 363,228 participants
• The study mapped 1,857 genetic loci to at least one biomarker, producing 3,374 fine-mapped associations (a process of narrowing down which specific variant in a region is most likely to be causal)
• Polygenic risk score models (weighted sums of many risk variants) built from these biomarker loci improved type 2 diabetes risk prediction in a separate Finnish cohort of 135,500 people
• Mendelian Randomization (a technique that uses genetic variants as natural experiments to test causal relationships) identified 51 causal links between biomarker levels and disease outcomes

What the research says A 2021 Nature Genetics study analyzed 35 blood and urine biomarkers in 363,228 UK Biobank participants (meta-analysis n=355,891), identifying 1,857 loci and 3,374 fine-mapped associations across protein-altering, non-coding, HLA, and copy-number variants, including variants at the CDKAL1 locus. Common-variant heritability across biomarkers ranged from 0.6% (lipoprotein A) to 23.9% (IGF-1) by LD Score regression (a method for estimating what fraction of trait variance is explained by common genetic variants), and from 3.2% (microalbumin in urine) to 57% (total bilirubin) by HESS (Heritability Estimator from Summary Statistics, an alternative approach applied at local genomic regions). Multi-biomarker polygenic risk score models built from these findings improved genetic risk stratification for type 2 diabetes and chronic kidney disease when validated in FinnGen (n=135,500), outperforming single-disease polygenic score approaches.

Reported associations

• Blood and urine biomarker levels: rs114898510 in CDKAL1 is among the variants mapped in this UK Biobank biomarker GWAS (n=363,228), which covered glycemic traits, lipids, kidney function tests, liver function tests, and other commonly measured laboratory values; specific effect sizes for this locus are reported in the study supplementary tables
• Type 2 diabetes genetic risk stratification: Multi-biomarker polygenic risk score models from this study improved type 2 diabetes risk classification in FinnGen (n=135,500) compared to single-disease polygenic score approaches, demonstrating the utility of biomarker-associated loci for disease prediction

Evidence quality Evidence comes from Sinnott-Armstrong et al. (Nature Genetics, 2021), one of the largest biomarker genome-wide association studies to date (meta-analysis n=355,891, independent replication n=135,500). Strict significance thresholds were applied: Bonferroni-corrected p < 5 x 10-9 for imputed variants, Bonferroni-corrected p < 1 x 10-6 for copy-number variants, and Benjamini-Yekutieli-adjusted p < 0.05 for HLA alleles. LD Score intercepts from 0.999 to 1.137 across all 35 phenotypes indicate well-controlled population stratification. Effect size estimates showed strong agreement with 42 previously published cohorts for 25 of the biomarkers. The study covered five ancestry groups (White British n=318,953, non-British White n=23,582, African n=6,019, South Asian n=7,338, East Asian n=1,082). Specific association statistics for rs114898510, such as effect sizes and confidence intervals, are contained in the study supplementary tables rather than the main text excerpt available here.

Lifestyle considerations No lifestyle considerations on file for this variant.