“…In the primary meta-analysis combining adjusted ORs (of early mortality) and HRs (of overall mortality) for married versus unmarried (unmarried (see Supplementary References 4 and 6), 7,8–10 never married (see Supplementary Reference 5), 2,4,5 and single) 1,6 patients demonstrated that married patients were associated with significantly lower early (OR 0.55, 95% CI 0.36–0.86; I 2 = 87%, P = 0.008; Figure 1, upper panel) and overall mortality (HR 0.75, 95% CI 0.65–0.87; I 2 = 55%, P = 0.0001; Figure 1, lower panel) compared with unmarried patients. In several studies (see Supplementary References 4 and 5), 1,2,4,5–7 multiple ORs/HRs (in addition to the OR/HR for married vs. unmarried patients) were available (e.g.…”