Include the health association, but discount it because the evidence is observational
Married people had about 12% lower mortality than unmarried people, while people who had never married had about 24% higher mortality than married people. This does not establish an effect of marriage. Healthier, wealthier, more socially connected people are both more likely to marry and more likely to live longer. The mostly Western figures offer only a rough scale for China.
What it takes
No money or time. Remember that married people having lower mortality is an association, not proof that marriage extends life.
What you may gain
A meta-analysis of older adults pooled 53 comparisons involving over 250,000 people. Married versus unmarried all-cause mortality had RR 0.88, 95% CI 0.85–0.91: about 12% lower risk during follow-up. RR is a risk ratio and CI an uncertainty interval. In the same study, widowed, divorced or separated, and never-married groups had ratios of 1.11, 1.16 and 1.11 relative to married people: about 11%, 16% and 11% higher. There was no sex difference. The most conservative estimate was RR 0.94, only about 6% lower. A second meta-analysis compared never-married with married people across 95 publications and over 500 million people: HR 1.24 (1.19–1.30), approximately 24% higher mortality during follow-up, with a smaller gap at older ages. A sex-specific meta-analysis of 21 studies and over 7.89 million people associated being unmarried with all-cause, cancer and cardiovascular mortality in both sexes. Men had larger cardiovascular and all-cause gaps. The paper describes sex differences corresponding to 31% for stroke mortality and 9% for all-cause mortality, based on women-to-men ratios of the marital-status risk ratios. These compare the strength of the unmarried-versus-married association between sexes, not the direct mortality rates of unmarried men and unmarried women.
Context & considerations
Disputed because selection into marriage matters: healthier, higher-income and more socially connected people are more likely both to marry and to live longer. All three meta-analyses used observational population data, not randomized comparisons. Their authors reported publication bias or methodological limitations. Roelfs also found smaller differences in more rigorous studies. The mainly Western populations offer only an approximate scale for China.
Research & references
Manzoli L, et al. (2007). Marital status and mortality in the elderly: a systematic review and meta-analysis. Soc Sci Med 64:77–94. https://doi.org/10.1016/j.socscimed.2006.08.031; Roelfs DJ, et al. (2011). The rising relative risk of mortality for singles: meta-analysis and meta-regression. Am J Epidemiol 174(4):379–389. https://doi.org/10.1093/aje/kwr111; Wang Y, et al. (2020). Sex differences in the association between marital status and the risk of cardiovascular, cancer, and all-cause mortality: a systematic review and meta-analysis of 7,881,040 individuals. Glob Health Res Policy 5:4. https://doi.org/10.1186/s41256-020-00133-8