Heat-Wave and Cold-Spell Vulnerability Analysis by Physical and Cognitive Functions: A Prospective Longitudinal Cohort Study in China

BACKGROUND AND AIM: There is limited evidence regarding the interaction between individual behavioural characteristics and temperature-related premature death. We aimed to investigate the short-term effects of heat-waves and cold-spells on elderly mortality by physical and cognitive functions. METHOD: We conducted a longitudinal cohort study utilizing the Chinese Longitudinal Healthy Longevity Survey (CLHLS, 2008-2018). Our study population consisted of age groups (65, 80, and 100 years). Heat-wave events were defined as three or more consecutive days with daily minimum temperature exceeding the 95th percentile (May to September), while cold-spell events were defined as three or more consecutive days with a daily maximum temperature below the 5th percentile (November to March). Demographic and socio-economic data, as well as physical (Activities of Daily Living, ADL; Instrumental Activities of Daily Living, IADL) and cognitive function (Mini-Mental State Examination, MMSE), were also collected. We used time-dependent Cox proportional hazard models, adjusted for possible confounders including age, sex, administrative region, urban or rural residence, co-residence, air pollutants (PM2.5 and ozone), and green and blue space around the residential locations. RESULTS: We followed 11,903 participants for 64,372.42 person-year (median baseline age: 86.87, 57.43% female). Both the occurrence of heat-wave day (HR: 1.33, 95% CI: 1.09, 1.60) and cold-spell day (HR: 1.28, 95% CI: 1.10, 1.49) were associated with increased premature mortality compared to normal days. High age was a risk factor of heat-wave (HR: 1.05, 95%CI:1.04, 1.06) and cold-spell (HR:1.34, 95%CI:1.13, 1.57). Individuals with severe ADL impairments were the most vulnerable to the short-term effects of heat-waves (HR: 2.13, 95% CI: 1.71, 3.86). Those needing partial help getting dressed were the most vulnerable subgroup to heat waves (HR:5.41, 95% CI:1.58, 18.56). CONCLUSIONS: It is crucial to understand the individual behavioural characteristics contributing to increased or decreased risk for mortality associated with extreme temperature; we found specific disabilities for public health interventions.