Table 1.
Examples of studies evaluating the relationship between genetic risk scores and cardiovascular outcomes
| Reference (first author, year)a | Outcomeb | Samplec | Source of SNPs for GRSd | GRS (no. of SNPs, weights)e | Effect estimatef | Clinical utilityg |
|---|---|---|---|---|---|---|
| Paynter, 2010 [39] | I-CVD | 19,313 EA (women) | CVD outcomes (multiple sources from [57]) | 12, U | HRh = 1.04 (1.00–1.08) per risk allele | ΔC index = 0.001 (p = 0.12) NRI = 0.05 (p = 0.59) |
| Cox et al., 2014 [41] | CVD, CVD mortality | 1,175 EA (T2D cases) | CHD/MI (7 sources, see [40]) | 13, W | CVDi: OR = 1.51 (1.22–1.86) per risk allele CVD mortalityi: HR = 1.35 (1.01–1.81) per risk allele |
CVD: NRI = 0.03 (p = 0.13) CVD mortality: NRI = 0.06 (p = 0.06) |
| Tikkanen, 2013 [24] | I-CHD, I-CVD | 24,124 N. Eur |
MI, CHD (4 sources, see [24]) | 28, W | I-CHDj: HR = 1.27 (1.20–1.35) per GRS SD I-CVDj: HR = 1.18 (1.12–1.24) per GRS SD |
For I-CHD, ΔC index ≈ 0.004 (p ≤ 0.001) NRI for CHD = 0.05 (p = 0.01) Clinical NRI for CHD = 0.27 (p = 1.1 × 10 –8) |
| Vaarhorst, 2012 [36] | I-CHD | 2,559 EA |
CHD (multiple sources from [57]) | 29, W | HRk = 1.12 (1.04–1.21) per risk allele | ΔC index = 0.002 (p = 0.13) NRI = 0.03 (p = 0.03) |
| Havulinna, 2013 [28] | I-CHD, I-stroke, I-CVD | 32,669 N. Eur |
SBP, DPB (4 sources, see [28]) | 32, W | For GRS quantiles 5 vs. 1, I-CHDl, GRSSBP: HR = 1.25 (1.07–1.46) I-CHDl, GRSDBP: HR = 1.23 (1.05–1.43) I-Strokel, GRSSBP: HR = 1.24 (1.01–1.53) I-Strokel, GRSDBP: HR = 1.35 (1.09–1.66) I-CVDl, GRSSBP: HR = 1.23 (1.08–1.40) I-CVDl, GRSDBP: HR = 1.26 (1.11–1.44) |
For 10-year I-CVD using GRSSBP, NRI = −0.0008 (p = 0.92) Clinical NRI = 0.04 (p = 5 × 10 −5) |
| Fava, 2014 [31] | IS | 6,092 N. Eur |
SBP, DPB (4 sources, see [31]) | 29, W | ORm = 1.09 (1.03–1.15) per GRS SD | ΔAUC = 0.003 (p > 0.05) NRI = 0.07 (p = 0.013) |
| Ibrahim-Verbaas, 2014 [35•] | IS, all stroke | 22,720 EA (age > 55 years) | Stroke, 9 risk factors (>30 sources, see [35•]) | 324, W | – | ISn: ΔAUC = 0.02 (p = 3.7 × 10
−7) All stroken: ΔAUC = 0.02 (p = 2.3 × 10 −6) |
| Ehret, 2011 [26] | HTN | 23,294 EA (women) | SBP, DBP [26] | 29, W | HTNo: OR = 1.23 (1.19–1.28) per GRS SD | – |
p values < 0.05 are in bold
I incident, EA European ancestry, N. Eur. Northern European ancestry, W weighted, U unweighted, SD standard deviation, Δ change, SNP single nucleotide polymorphism, GRS genetic risk score, GWAS genome-wide association study, CI confidence interval, CVD cardiovascular disease, CHD coronary heart disease, IS ischemic stroke, HTN hypertension, MI myocardial infarction, SBP systolic blood pressure, DBP diastolic blood pressure, HDL-C high-density lipoprotein cholesterol, T2D type 2 diabetes, BMI body mass index, HR hazard ratio, OR odds ratio, NRI net reclassification improvement index, AUC area under the receiver-operator curve
aOnly selected GRS analyses from each study are reported in this table
bCardiovascular outcome tested for association with GRS
cRefers only to the sample used for evaluating the association between the outcome and GRS, corresponding to the effect estimates listed in this table
dAll studies used GWAS/meta-analyses as the SNP source for the GRS. The outcome/trait used in the GWAS is reported
eNumber of SNPs in the GRS
fEffect estimate (95 % CI) for association between outcome and GRS. Adjustment variables are listed below
gEstimates for clinical utility compare a prediction model including the GRS and traditional CVD risk factors (i.e., the adjustment variables described in the effect estimate column) to a model with only traditional CVD risk factors, followed by p value
hAdjustment variables: Age, SBP, hypertensive medication use, smoking, diabetes, total cholesterol, HDL-C
iAdjustment variables: Age, sex, T2D, BMI, current smoking, HTN, dyslipidemia
jAdjustment variables: Sex, total cholesterol, HDL-C, BMI, SBP, antihypertensive treatment, smoking, T2D
kAdjustment variables: Sex, current smoking, SBP, total cholesterol, HDL-C, self-reported diabetes, BMI, parental history of MI
lAdjustment variables: Age, age squared, sex
mAdjustment variables: Age, sex, T2D, smoking habits, HTN
nAdjustment variables: Age, sex, Framingham Stroke Risk Score (SBP, T2D, smoking, prior CVD, atrial fibrillation, left ventricular hypertrophy, antihypertensive medication)
oAdjustment variables: Age, age-squared, BMI