A study paradigm integrating prospective epidemiologic cohorts and electronic health records to identify disease biomarkers.

Defining the full spectrum of human disease associated with a biomarker is necessary to advance the biomarker into clinical practice. We hypothesize that associating biomarker measurements with electronic health record (EHR) populations based on shared genetic architectures would establish the clinical epidemiology of the biomarker. We use Bayesian sparse linear mixed modeling to calculate SNP weightings for 53 biomarkers from the Atherosclerosis Risk in Communities study. We use the SNP weightings to computed predicted biomarker values in an EHR population and test associations with 1139 diagnoses. Here we report 116 associations meeting a Bonferroni level of significance. A false discovery rate (FDR)-based significance threshold reveals more known and undescribed associations across a broad range of biomarkers, including biometric measures, plasma proteins and metabolites, functional assays, and behaviors. We confirm an inverse association between LDL-cholesterol level and septicemia risk in an independent epidemiological cohort. This approach efficiently discovers biomarker-disease associations.

Pharmacogenetic predictors of statin-mediated low-density lipoprotein cholesterol reduction and dose response.

BACKGROUND: There is interindividual variation in low-density lipoprotein cholesterol (LDLc) lowering by statins and limited study into the genetic associations of the dose dependant LDLc lowering by statins. METHODS AND RESULTS: Five hundred nine patients with hyperlipidemia were randomly assigned atorvastatin 10 mg, simvastatin 20 mg, or pravastatin 10 mg (low-dose phase) followed by 80 mg, 80 mg, and 40 mg (high-dose phase), respectively. Thirty-one genes in statin, cholesterol, and lipoprotein metabolism were sequenced and 489 single nucleotide polymorphisms with minor allele frequencies >2% were tested for associations with percentage LDLc lowering at low doses using multivariable adjusted general linear regression. Significant associations from the analysis at low dose were then repeated at high-dose statins. At low doses, only 1 single nucleotide polymorphism met our experiment-wide significance level, ABCA1 rs12003906. Twenty-six subjects carried the minor allele of rs12003906, which was associated with an attenuated LDLc reduction (LDLc reduction in carriers versus noncarriers -24.1+/-2.6% versus -32.2+/-1.5%; P=0.0001). In addition, we replicated the association with the APOE epsilon3 allele and a reduced LDLc reduction. At high doses, carriers of the minor allele of ABCA1 rs12003906 and the APOE epsilon3 allele improved their LDLc reduction but continued to have a diminished LDLc reduction compared with noncarriers (-30.5+/-4.0% versus -42.0+/-2.4%; P=0.005) and (-38.5+/-1.9% versus -45.3+/-2.8%; P=0.009), respectively. CONCLUSIONS: An intronic single nucleotide polymorphism in ABCA1 and the APOE epsilon3 allele are associated with reduced LDLc lowering by statins and identify individuals who may be resistant to maximal LDLc lowering by statins.