Very interesting paper. I couldn't find the definition of what groups are being compared for effect size/hazard ratios in Figure 4. Is it also for an 1-SD increment as in Figure 2E?

Also supplementary material are not available on medRxiv.