Related Experiment Video
Updated: Aug 28, 2025

Establishing a Competing Risk Regression Nomogram Model for Survival Data
Published on: October 23, 2020
How do cohabitation and marital status affect mortality risk? Results from a cohort study in Thailand
Jiaying Zhao1, Chi Kin Law2, Matthew Kelly3
1School of Demography, Australian National University, Canberra, Australian Capital Territory, Australia.
Objective:
To examine the relationship between baseline union status (ie, including marriage and cohabitation) and mortality, paying attention to gender differentials, through an 11-year follow-up of a large cohort in Thailand.
Design:
Cohort data from Thai Cohort Study (TCS) were linked official death records over an 11-year follow-up period.
Setting:
Community-based adults in Thailand.
Participants:
87 151 Thai adults participated in TCS cohort.
Method:
Cox regression models measured longitudinal associations between union status and 11-year mortality.
Results:
From 2005 (baseline) to 2016, persons who cohabited and lived with a partner, married persons but not living with a partner and separated/divorced/widowed people were more likely to die compared with those married and living together with a partner. Those who did not have good family support had a higher death risk than those having good family support.Single or cohabiting women had higher risks of mortality than women who were married and living together with a partner throughout follow-up, while separated/divorced/widowed men had higher risks of mortality than counterpart males.
Conclusions:
Our study reveals the protective effect of marriage and living together on mortality in Thailand, an understudied setting where institutionalisation of cohabitation is low leading to a limited mortality protection. Public policies for moderating mortality should thus be gender nuanced, culturally and institutionally specific. Also, we demonstrate that in settings such as Thailand, where marital status is not always defined in the same way as in western cultures, the need to measure cohabitation in locally relevant terms is important.
Related Concept Videos
Applications of Life Tables
Relationship Formation
Assumptions of Survival Analysis
Relative Risk
Bias in Epidemiological Studies
Life Tables

