Supplementary Online Tables

Table e-1. Effect of TC, HDL-C and TC/HDL-c ratio on 10 year Stroke Incidence adjusted for interim MI events

IS / ABI
Cases/N / 269/6276 / 190/6276
HR [95% CI] / HR [95% CI]
TC per10 / 1.02[0.99-1.05] / 1.04[1.00-1.07]
TC≥240 / 1.00[0.77-1.31] / 1.10[0.81-1.51]
HDL^ / 0.68[0.60-0.78]*** / 0.71[0.61-0.82]***
HDL≤40 / 2.02[1.56-2.62]*** / 1.98[1.46-2.70]***
TC/HDL^ / 1.39[1.24-1.56]*** / 1.41[1.23-1.61]***
TC/HDL ≥5 / 1.71[1.34-2.19]*** / 1.81[1.35-2.43]***

* 0.01≤p<0.05; ** 0.001≤p<0.01; *** p<0.001

^log-transformed, increase per approximately 1SD

Table e-2. Effect of TC, HDL-C and TC/HDL-C ratio on 10Year Ischemic stroke based on stroke subtypes

IS / CE / ABI / Lacunar / Non-Lacunar
(LAA +Cryptogenic)
Case/N / 301/6276 / 90/6276 / 211/6276 / 57/6276 / 154/6276
HR [95% CI] / HR [95% CI] / HR [95% CI] / HR [95% CI] / HR [95% CI]
TC per 10 / Model 1 / 1.02 [1.00-1.05] / 0.99 [0.94-1.04] / 1.04 [1.01-1.08]* / 1.02 [0.95-1.09] / 1.05 [1.01-1.09]*
TC≥240 / 1.04 [0.81-1.34] / 0.82 [0.51-1.32] / 1.15 [0.86-1.55] / 1.19 [0.67-2.13] / 1.14 [0.81-1.61]
HDL^ / 0.67 [0.59-0.76]*** / 0.63 [0.50-0.78]*** / 0.69 [0.60-0.80]*** / 0.71 [0.53-0.94]* / 0.69 [0.58-0.81]***
HDL≤40 / 2.05 [1.61-2.61]*** / 2.16 [1.38-3.37]*** / 2.01 [1.50-2.68]*** / 1.59 [0.92-2.77] / 2.19 [1.56-3.07]***
TC/HDL^ / 1.42 [1.28-1.58]*** / 1.38 [1.14-1.68]** / 1.44 [1.27-1.63]*** / 1.35 [1.06-1.73]* / 1.47 [1.27-1.70]***
TC/HDL ≥5 / 1.76 [1.40-2.23]*** / 1.61 [1.05-2.46]* / 1.84 [1.39-2.43]*** / 1.98 [1.15-3.42]* / 1.78 [1.29-2.46]***
IS / CE / ABI / Lacunar / Non-Lacunar
(LAA +Cryptogenic)
Case/N / 291/6101 / 86/6101 / 205/6101 / 56/6101 / 149/6101
HR [95% CI] / HR [95% CI] / HR [95% CI] / HR [95% CI] / HR [95% CI]
TC per10 / Model 2 / 1.02 [0.99-1.05] / 1.02 [0.97-1.08] / 1.03 [0.99-1.06] / 1.02 [0.95-1.08] / 1.03 [0.99-1.07]
TC≥240 / 1.01 [0.78-1.30] / 0.95 [0.59-1.55] / 1.04 [0.77-1.41] / 1.20 [0.67-2.16] / 0.99 [0.69-1.41]
HDL^ / 0.77 [0.68-0.87]*** / 0.68 [0.53-0.86]** / 0.81 [0.70-0.95]** / 0.86 [0.64-1.16] / 0.80 [0.67-0.95]*
HDL≤40 / 1.59 [1.23-2.05]*** / 1.84 [1.15-2.96]* / 1.50 [1.11-2.03]** / 1.13 [0.64-2.00] / 1.68 [1.18-2.39]**
TC/HDL^ / 1.29 [1.15-1.44]*** / 1.41 [1.14-1.74]** / 1.24 [1.09-1.42]** / 1.17 [0.91-1.52] / 1.27 [1.08-1.48]**
TC/HDL ≥5 / 1.47 [1.15-1.87]** / 1.53 [0.98-2.39] / 1.44 [1.08-1.92]* / 1.52 [0.87-2.66] / 1.41 [1.00-1.97]*

*0.01<p<0.05, **0.001<p<0.01, ***p<0.001

Table e- 3. Effect of HDL-C on 10 Year Incident IS,ABI and MI and net reclassification HR.

IS / ABI / MI
Case/N / 284/6022 / 201/6022 / 372/5642
HR [95% CI] / HR [95% CI] / HR [95% CI]
Age/5 years / 1.51*** / 1.57*** / 1.33***
Male sex / 0.92 / 1.15 / 2.14***
HTN / 1.62*** / 1.65** / 1.34*
Smoking / 1.70*** / 1.68 / 1.69***
Prevalent CVD / 1.36* / 1.31 / 1.72***
Prevalent AF / 2.12** / 0.36 / 1.08
Diabetes / 2.06*** / 2.59*** / 1.89***
BMI / 1.03* / 1.05*** / 1.02
HDL≤40 / 1.59*** / 1.50** / 1.53***
Addition of HDL≤40to model:
Continuous NRI / 0.307 [0.185,0.417] / 0.319 [0.188,0.456] / 0.401 [0.292,0.507]

*0.01<p<0.05, **0.001<p<0.01, ***p<0.001

Supplementary Appendixe-1.

Genotyping methods

Genotyping was performed on the Affymetrix GeneChip® Human Mapping 500K Array Set and 50K Human Gene Focused Panel® through the SHARe project (SNP-Health Association Resource), and these data were used to impute to the 2.5 million non-monomorphic, autosomal single nucleotide polymorphisms (SNPs) described in the HapMap (CEU population). Genotyping, quality control, and imputation methods have been described earlier.

Supplementary Appendix e-2.

Power calculation for Mendelian Randomization

Denote as the coefficient of genetic score on IS, . Under the null hypothesis,  = 0. Since we are interested in testing if the genetic score association is in the same direction as the epidemiology association of HDL-C with IS, the alternative hypothesis is one-sided, e.g.  0 . Therefore, the power for detecting  =1 at alpha=0.05 is

where 1.96 is the critical value of a one sided Z test at alpha=0.05.

1 corresponding to 80% power can be obtained by solving above equation at power=80%

i.e. where 1.28 is the standard normal quantile corresponding to 80% left tail probability. Thus is the expected effect size that can be detected with 80% power. For a dichotomous (survival) outcome, , is the log odds ratio (log hazard ratio) and same power calculation as continuous trait can be used to obtain . Odds ratio (hazard ratio) corresponding to 80% power can be obtained by transforming.

Given the SE of genetic score association with ischemic stroke is 0.3, the study has 80% power to detect beta =-(1.96+1.28)*0.3=-1.0449 or less.

Table e-3a. Effect of TG and LDL-C on 10 Year Stroke and MI Incidence stratified by Sex and Age

Model 2 / IS Events
Case/N / All / Women / Men / <65 / ≥65
69/2801 / 25/1417 / 44/1384 / 43/2368 / 25/408
HR [95% CI] / HR [95% CI] / HR [95% CI] / HR [95% CI] / HR [95% CI]
TG^ / 1.17[0.93-1.48] / 1.30[0.89-1.91] / 1.12[0.84-1.49] / 1.22[0.90-1.64] / 1.11[0.74-1.65]
TC≥240 / 1.23[0.71-2.15] / 1.58[0.60-4.16] / 1.10[0.56-2.17] / 1.42[0.73-2.76] / 1.03[0.36-2.95]
LDL-C^ / 1.08[0.85-1.36] / 1.44[1.03-2.01]* / 0.86[0.63-1.19] / 0.99[0.73-1.34] / 1.06[0.69-1.63]
LDL-C≥160 / 1.02[0.58-1.80] / 2.48[1.09-5.67]* / 0.47[0.19-1.21] / 1.12[0.56-2.26] / 0.65[0.22-1.93]
ABI Events
Case/N / All / Women / Men / <65 / ≥65
54/2801 / 18/1417 / 36/1384 / 30/2368 / 23/408
HR [95% CI] / HR [95% CI] / HR [95% CI] / HR [95% CI] / HR [95% CI]
TG^ / 1.17[0.91-1.52] / 1.19[0.77-1.84] / 1.19[0.87-1.63] / 1.18[0.83-1.67] / 1.23[0.83-1.82]
TC≥240 / 1.18[0.63-2.21] / 1.14[0.33-3.90] / 1.26[0.61-2.59] / 1.32[0.60-2.92] / 1.21[0.43-3.47]
LDL-C^ / 1.15[0.88-1.50] / 1.67[1.15-2.43]** / 0.87[0.61-1.23] / 1.02[0.71-1.46] / 1.14[0.73-1.77]
LDL-C≥160 / 1.06[0.56-1.99] / 2.93[1.12-7.65]* / 0.45[0.16-1.29] / 1.18[0.52-2.70] / 0.71[0.24-2.11]
MI Events
Case/N / All / Women / Men / <65 / ≥65
100/2669 / 25/1395 / 75/1274 / 78/2273 / 20/374
HR [95% CI] / HR [95% CI] / HR [95% CI] / HR [95% CI] / HR [95% CI]
TG^ / 1.36[1.14-1.63]*** / 1.38[0.94-2.01] / 1.39[1.14-1.71]** / 1.34[1.09-1.64]** / 1.24[0.84-1.84]
TC≥240 / 1.31[0.82-2.08] / 2.51[0.98-6.43] / 1.13[0.67-1.92] / 1.21[0.71-2.04] / 1.40[0.46-4.26]
LDL-C^ / 1.16[0.96-1.41] / 1.26[0.90-1.77] / 1.12[0.89-1.41] / 1.17[0.94-1.45] / 1.12[0.68-1.87]
LDL-C≥160 / 1.61[1.04-2.49]* / 1.80[0.77-4.19] / 1.57[0.94-2.62] / 1.67[1.03-2.70]* / 1.29[0.40-4.12]

* 0.01≤p<0.05; ** 0.001≤p<0.01; *** p<0.001

^log-transformed, increase per approximately 1SD

Model 2- adjusted for age, sex, HTN, smoking, prevalent CVD, AF, DM, BMI

1