Yang TY, Zhang YH, Sun QY, Wang YR, Wang KJ. Diurnal temperature range, daily mean temperature, and incident cardiovascular disease in older adults: A longitudinal cohort study. World J Cardiol 2026; 18(10): 125761 [DOI: 10.4330/wjc.125761]
Corresponding Author of This Article
Kai-Juan Wang, MD, PhD, Department of Epidemiology and Health Statistics, College of Public Health, Zhengzhou University, No. 100 Kexue Avenue, Zhengzhou 450001, Henan Province, China. kjwang@163.com
Research Domain of This Article
Medicine, Research & Experimental
Article-Type of This Article
research-article
Open-Access Policy of This Article
This article is an open-access article which was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/
Baishideng Publishing Group Inc, 7041 Koll Center Parkway, Suite 160, Pleasanton, CA 94566, USA
Share the Article
Yang TY, Zhang YH, Sun QY, Wang YR, Wang KJ. Diurnal temperature range, daily mean temperature, and incident cardiovascular disease in older adults: A longitudinal cohort study. World J Cardiol 2026; 18(10): 125761 [DOI: 10.4330/wjc.125761]
Tong-Yu Yang, Department of Immunization Program, Yuhong Center for Disease Control and Prevention, Shenyang 110000, Liaoning Province, China
Tong-Yu Yang, Yue-Hua Zhang, Qiu-Yu Sun, Kai-Juan Wang, Department of Epidemiology and Health Statistics, College of Public Health, Zhengzhou University, Zhengzhou 450001, Henan Province, China
Yi-Ran Wang, Henan Children’s Hospital Zhengzhou Children’s Hospital, Children’s Hospital Affiliated to Zhengzhou University, Zhengzhou 450018, Henan Province, China
Kai-Juan Wang, Henan Key Laboratory of Tumor Epidemiology, International Joint Laboratory of Tumor Biomarker and Molecular Imaging, Zhengzhou University, Zhengzhou 450052, Henan Province, China
Kai-Juan Wang, National Key Laboratory of Metabolism Disorder and Esophageal Cancer Prevention & Treatment, Zhengzhou University, Zhengzhou 450052, Henan Province, China
Author contributions: Yang TY and Wang KJ designed the study; Yang TY conducted the data analysis and wrote the first draft of the manuscript; Yang TY, Zhang YH, Sun QY, and Wang YR contributed to statistical analysis and data visualization; Wang KJ supervised the study and provided critical revisions and suggestions; and all authors read and approved the final manuscript.
AI contribution statement: The authors declare that no AI tools were used in the development or writing of this manuscript and take full responsibility for its integrity, accuracy, and originality.
Supported by the Henan Science and Technology Department, No. 251111313000.
Institutional review board statement: The CHARLS study was approved by the Ethics Review Committee of Peking University (No. IRB00001052-11015).
Informed consent statement: All participants provided written informed consent prior to enrollment.
Conflict-of-interest statement: All the authors report no relevant conflicts of interest for this article.
STROBE statement: The authors have read the STROBE Statement-checklist of items, and the manuscript was prepared and revised according to the STROBE Statement-checklist of items.
Data sharing statement: The datasets generated and/or analysed in the current study are available on the China Health and Retirement Longitudinal Study (CHARLS) website (https://charls.pku.edu.cn/) and the Chinese Research Data Services Platform (CNRDS, https://www.cnrds.com). The harmonized datasets used in this analysis are derived from the Harmonized CHARLS, available through the Gateway to Global Aging Data (https://g2aging.org).
Corresponding author: Kai-Juan Wang, MD, PhD, Department of Epidemiology and Health Statistics, College of Public Health, Zhengzhou University, No. 100 Kexue Avenue, Zhengzhou 450001, Henan Province, China. kjwang@163.com
Received: July 16, 2026 Revised: August 27, 2026 Accepted: October 8, 2026 Published online: October 26, 2026 Processing time: 99 Days and 20.7 Hours
Abstract
BACKGROUND
Cardiovascular disease (CVD) remains the leading cause of death globally. However, most studies have focused on acute temperature extremes—heatwaves and cold spells—with less attention to the potential cardiovascular effects of long-term exposure to ambient temperatures.
AIM
To investigate the associations of mean diurnal temperature range (mDTR) and mean daily mean temperature (mDMT) with incident CVD risk.
METHODS
Participants aged ≥ 60 years were recruited from the China Health and Retirement Longitudinal Study. Temperature exposure was derived from city-level meteorological data and defined as the average of mDTR and mDMT measured in 2011 and 2013. The incidence of CVD (physician-diagnosed heart disease or stroke) was ascertained at Wave 3 (2015) and Wave 4 (2018). The baseline CVD prevalence, mDTR, and mDMT were mapped at the provincial level. Cox proportional hazards models, restricted cubic splines, and Kaplan-Meier curves were used to assess the associations. Subgroup and sensitivity analyses were conducted to test the robustness of the findings.
RESULTS
During a median follow-up of 5.0 years, 982 incident CVD events occurred among 4333 participants (mean age 67.41 ± 6.35 years; 50.27% male). Each 1-SD increase in mDTR was associated with a 6% higher CVD risk, with Q4 vs Q1 associated with a 102% higher risk. Each 1-SD increase in mDMT was associated with a 6% lower risk, with Q4 vs Q1 associated with a 56% lower risk. Restricted cubic splines analyses indicated significant overall associations of both mDTR and mDMT with incident CVD, with significant evidence of nonlinearity for mDTR (P nonlinear < 0.001) but not for mDMT (P nonlinear = 0.808). The associations remained generally consistent across most sensitivity analyses and prespecified subgroups.
CONCLUSION
These findings suggest that temperature variability and average temperature may be important environmental factors to consider when assessing cardiovascular risk among older adults.
Core Tip: While acute temperature extremes are known to affect cardiovascular disease (CVD), the chronic effects of temperature variability and average levels in older adults remain unclear. Using data from the China Health and Retirement Longitudinal Study data, we found that higher average exposure to mean diurnal temperature range is associated with increased CVD risk, whereas higher average exposure to mean daily mean temperature is associated with decreased CVD risk. These associations remained generally consistent across prespecified subgroup and sensitivity analyses. These findings suggest that temperature variability and average temperature may be important environmental factors to consider when assessing cardiovascular risk among older adults.
Citation: Yang TY, Zhang YH, Sun QY, Wang YR, Wang KJ. Diurnal temperature range, daily mean temperature, and incident cardiovascular disease in older adults: A longitudinal cohort study. World J Cardiol 2026; 18(10): 125761
Cardiovascular disease (CVD) remains a leading cause of premature death and a major contributor to healthcare burden worldwide, accounting for an estimated 19.8 million deaths, representing 32% of all global deaths, in 2022[1,2]. In the context of global climate change, the relationship between ambient temperature exposure and cardiovascular health has received increasing attention. However, most previous studies have focused on short-term temperature extremes, such as heatwaves and cold spells, using time-series or case-crossover designs that capture acute cardiovascular responses over hours to days[3,4]. Although these studies have provided strong evidence for the immediate effects of extreme temperatures, the cardiovascular implications of various aspects of long-term temperature exposure, particularly among older adults, remain less well characterized[5]. Given that ambient temperature represents continuous environmental exposure, evaluating its long-term associations with cardiovascular risk is essential to better characterize the health impacts of climate change.
The diurnal temperature range (DTR) and daily mean temperature (DMT) represent two distinct dimensions of ambient temperature exposure[6]. The DTR reflects within-day temperature variability, whereas the DMT represents the overall thermal level of a day. These indicators may influence cardiovascular health through different pathways. Greater temperature variability may increase cardiovascular stress through repeated thermoregulatory responses, whereas lower ambient temperatures may activate sympathetic responses, promote vasoconstriction, increase blood pressure, and elevate cardiac workload[7-9]. Older adults may be particularly vulnerable to these effects because of age-related declines in vascular function and thermoregulatory capacity.
China provides an important setting for evaluating the cardiovascular effects of long-term temperature exposure because of its extensive geographic area and substantial climatic heterogeneity. Northern regions generally experience colder conditions and larger diurnal temperature variations, whereas southern regions are characterized by warmer climates and smaller temperature fluctuations[10]. Such climatic differences may contribute to regional variations in vulnerability to temperature-related cardiovascular stress. Nevertheless, previous studies have mainly been conducted in specific cities or regions[11,12], limiting the generalizability of their findings to the national population level.
Therefore, using data from the China Health and Retirement Longitudinal Study (CHARLS), this prospective cohort study aimed to investigate the associations of mean DTR (mDTR) and mean DMT (mDMT), averaged across temperature estimates from 2011 and 2013, with incident CVD risk among Chinese adults aged 60 years and older. By examining both temperature variability and average temperature, this study sought to provide a more comprehensive assessment of the different dimensions of long-term ambient temperature exposure and their relationship to cardiovascular risk among older adults. These findings may provide further evidence regarding the cardiovascular effects of environmental temperature exposure and inform climate-related prevention strategies for aging populations.
MATERIALS AND METHODS
Study design and participants
This study utilized data from the CHARLS[13], a nationally representative longitudinal cohort of adults aged 45 years and older in China. The baseline survey was conducted in 2011 and 2012 (Wave 1). Follow-up waves were conducted biennially in 2013 (Wave 2), 2015 (Wave 3), 2018 (Wave 4), and 2020 (Wave 5). Ethical approval was obtained from the Biomedical Ethics Review Committee of Peking University (No. IRB00001052-11015), and all participants provided written informed consent.
Wave 1 was used to ascertain baseline characteristics and CVD status. Temperature exposure was assessed using temperature data from 2011 and 2013. Wave 2 was used as the landmark time point (time zero) for survival analysis. Participants were therefore required to be free of CVD at both Wave 1 and Wave 2 before entering follow-up. Incident CVD was subsequently assessed during Wave 3 and Wave 4. Individuals were excluded based on the following criteria: (1) Age < 60 years; (2) Missing CVD information at Wave 1 or Wave 2; (3) Prevalent CVD at Wave 1 or Wave 2; (4) Missing data on DTR or DMT; and (5) Missing or invalid follow-up information. A total of 4333 participants were included in the final analysis (Figure 1).
Figure 1 Flowchart of the study population.
CVD: Cardiovascular disease; mDTR: Mean diurnal temperature range; mDMT: Mean daily mean temperature.
Assessment of mDTR and mDMT
The primary exposures were mDTR and mDMT. Meteorological data were obtained from the China Research Data Services Platform (CNRDS, https://www.cnrds.com), which provided daily maximum and minimum temperatures from 124 cities across 28 provinces. DTR was calculated as the difference between daily maximum and minimum temperatures, and DMT as the average of the two [(maximum + minimum)/2]. mDTR and mDMT were calculated as the arithmetic means of the corresponding daily DTR and DMT values for each available year. City names in the meteorological and CHARLS datasets were harmonized and matched using city-level geographic identifiers. In this study, the temperature exposures from 2011 and 2013 were subsequently averaged to represent longer-term temperature exposure and reduce the influence of short-term variability.
Assessment of incident CVD
The primary outcome was incident CVD, defined as self-reported physician-diagnosed heart disease or stroke. Participants were asked: “Have you ever been diagnosed with a heart attack, coronary heart disease, angina, congestive heart failure, or other heart problems by a physician?” or “Have you ever been diagnosed with a stroke by a physician?”[14-16]. Incident CVD events were assessed at Wave 3 and Wave 4. The start of follow-up for the Cox proportional hazards analyses was defined as the Wave 2 interview date. For participants who first reported CVD at Wave 3, the event time was calculated as the elapsed time from the Wave 2 interview date to the Wave 3 interview date. For participants who remained free of CVD at Wave 3 but first reported CVD at Wave 4, the event time was calculated from the Wave 2 interview date to the Wave 4 interview date. Because the exact date of CVD onset was not available, the corresponding follow-up interview date at which CVD was first reported was used to define the event time.
Covariates
All covariates were assessed at Wave 1 and included: Demographic factors—age, sex (male/female), education (below upper secondary/upper secondary or above), residence (rural/urban), and marital status (partnered/not partnered); lifestyle factors—smoking status (ever/never) and drinking status (ever/never); socioeconomic status—household wealth (categorized into quartiles based on total asset value); anthropometric measure—body mass index (BMI); and environmental exposure—ambient fine particulate matter (PM2.5).
Handling of missing data
Missing data were handled according to a prespecified principle of no imputation for exposure or outcome variables, while imputation was permitted for covariates[17]. Completeness of the exposure data and availability of outcome information were prerequisites for inclusion in the main analysis. Missing covariates were handled using multiple imputation by chained equations, generating 20 imputed datasets with 20 iterations per dataset[18]. Each model was fitted separately across the multiple imputed datasets, and the results were pooled according to Rubin’s rules. The missing-data proportions for each variable and details of the multiple-imputation procedures are presented in Supplementary Table 1.
Statistical analysis
Baseline characteristics were compared according to incident CVD status. Continuous variables were summarized as mean ± SD and compared using Student’s t test or Wilcoxon rank-sum test, as appropriate. Categorical variables were presented as n (%) and compared using Pearson’s χ2 test.
Multivariable Cox proportional hazards models were used to estimate the association of mDTR and mDMT with incident CVD risk, reporting hazard ratios (HRs) and 95% confidence intervals (CIs). Four models were constructed: Model 1 was unadjusted; Model 2 adjusted for sex and age; Model 3 additionally adjusted for education, wealth, residence, marital status, smoking, drinking, and BMI; and Model 4 additionally adjusted for PM2.5. Each exposure was modeled in two ways: (1) As a continuous variable, per 1-SD increase; and (2) As a categorical variable, divided into quartiles, with the lowest quartile (Q1) as the reference, estimating HRs for Q2, Q3, and Q4, and calculating the P value for trend. The proportional hazards assumption was assessed using Schoenfeld residuals, and multicollinearity among covariates was assessed using variance inflation factors (VIFs). Spearman correlation analysis was performed to assess the correlation between mDTR and mDMT (Supplementary Figure 1). Restricted cubic splines (RCS) with 4 knots were fitted to examine the dose-response relationship, with tests for overall association and nonlinearity. Kaplan-Meier survival curves were used to visualize CVD-free survival probability over time across specific quartiles of mDTR and mDMT, and differences between groups were tested using the log-rank test.
Subgroup analyses were performed based on the fully adjusted model with continuous mDTR and mDMT. The models were repeated within strata defined by sex, age (< 65/≥ 65 years), education, wealth, residence, marital status, smoking, and drinking. Interaction terms between each exposure and the subgroup variable were used to test for heterogeneity of effects across strata.
Sensitivity analyses included: (1) Mutual adjustment of mDTR and mDMT; (2) Logistic regression treating incident CVD as a binary outcome; (3) Winsorization of mDTR/mDMT at the 1st and 99th percentiles; (4) Complete-case analysis; (5) Lag analysis, in which participants who developed CVD within the first 2 years of follow-up were excluded; and (6) Additional adjustment for province to account for broad geographic heterogeneity. The consistency of effect directions and approximate magnitudes across sensitivity analyses was used to assess the robustness of the findings.
All statistical analyses were performed using R software (version 4.5.2). A two-sided P value < 0.05 was considered statistically significant.
RESULTS
Baseline characteristics
A total of 4333 participants aged ≥ 60 years were included (mean age 67.41 ± 6.35 years; 50.27% male). Over a median follow-up of 5.0 years, 982 participants (22.66%) developed incident CVD; the provincial distribution of participants and incident CVD events is presented in Supplementary Table 2. Baseline characteristics stratified by incident CVD status are shown in Table 1. Compared with participants who remained free of CVD, those who developed CVD had higher BMI and PM2.5 levels, higher prevalence of hypertension, diabetes, and dyslipidemia, and a higher proportion of female participants. Regarding temperature exposures, participants who developed CVD had higher mDTR and lower mDMT values.
Table 1 Baseline characteristics of participants according to incident cardiovascular disease status, n (%).
At the provincial level, CVD prevalence varied substantially across regions, ranging from 8.51% in Guizhou to 46.43% in Heilongjiang, with higher prevalence generally observed in northern provinces and lower prevalence in southern provinces (Supplementary Table 3, Figure 2A). mDTR ranged from 8.67 °C (Shanghai) to 26.6 °C (Xinjiang), with higher values in northern and western regions (Supplementary Table 4, Figure 2B). mDMT ranged from 1.93 °C (Inner Mongolia) to 22.6 °C (Guangdong), showing a clear north-south gradient consistent with the spatial distribution of CVD prevalence (Supplementary Table 5, Figure 2C).
Figure 2 Spatial distribution of cardiovascular disease prevalence, mean diurnal temperature range, and mean daily mean temperature across Chinese provinces.
A: Baseline cardiovascular disease prevalence; B: Average mean diurnal temperature range (2011 and 2013); C: Average mean daily mean temperature (2011 and 2013). Meteorological data were available for 28 provinces covered by the China Health and Retirement Longitudinal Study survey; white areas indicate missing data across all panels. CVD: Cardiovascular disease; mDTR: Mean diurnal temperature range; mDMT: Mean daily mean temperature.
Associations of mDTR and mDMT with incident CVD risk
Table 2 presents the associations of mDTR and mDMT with incident CVD risk. In the fully adjusted model (Model 4), each 1-SD increase in mDTR was associated with a 6% higher risk of incident CVD (HR = 1.06, 95%CI: 1.05-1.08, P < 0.001). When analyzed by quartiles, participants in Q4 had a 102% higher risk than those in Q1 (HR = 2.02, 95%CI: 1.67-2.46, P < 0.001). In contrast, each 1-SD increase in mDMT was associated with a 6% lower risk of incident CVD (HR = 0.94, 95%CI: 0.93-0.95, P < 0.001), and participants in Q4 had a 56% lower risk compared with Q1 (HR = 0.44, 95%CI: 0.36-0.53, P < 0.001). The distribution of participants, person-years, and incident CVD events across quartiles of mDTR and mDMT is presented in Supplementary Table 6. The proportional hazards assumption was satisfied for both mDTR and mDMT, as assessed using Schoenfeld residuals (Supplementary Tables 7 and 8). No substantial multicollinearity was observed among the covariates in Model 4, with all VIFs below 1.5 (Supplementary Tables 9 and 10).
Table 2 Cox regression models for the association of mean diurnal temperature range and mean daily mean temperature with incident cardiovascular disease risk.
Kaplan-Meier survival curves were constructed to compare CVD-free survival probability across quartiles of mDMT and mDTR (Figure 3A and B). For mDTR, participants in the higher quartiles consistently showed lower CVD-free survival probabilities throughout the follow-up period (log-rank P < 0.001). Conversely, for mDMT, participants in the higher quartiles consistently showed higher CVD-free survival probabilities (log-rank P < 0.001). These findings were consistent with the results from the Cox regression models.
Figure 3 Kaplan-Meier survival curves for incident cardiovascular disease across quartiles of A mean daily mean temperature and B mean diurnal temperature range.
Cardiovascular disease (CVD)-free survival probability was compared using the log-rank test, and the corresponding P values are shown in each panel. Dose-response relationship between C mean daily mean temperature and D mean diurnal temperature range and incident CVD risk using restricted cubic spline models. Models were adjusted for age, sex, education, wealth, residence, marital status, smoking, drinking, body mass index, and fine particulate matter with an aerodynamic diameter ≤ 2.5 μm. CVD: Cardiovascular disease; mDTR: Mean diurnal temperature range; mDMT: Mean daily mean temperature.
RCS analysis
RCS analysis with four knots based on the fully adjusted model was used to examine the dose-response relationships of mDMT and mDTR with incident CVD risk (Figure 3C and D). For mDMT, the overall association was statistically significant, with no evidence of nonlinearity (P overall < 0.001; P nonlinear = 0.808). The HR decreased progressively with increasing mDMT and crossed HR = 1 at approximately 16.0 °C, with higher mDMT associated with a lower risk of incident CVD across the exposure range. For mDTR, both the overall association and nonlinear component were statistically significant (P overall < 0.001; P nonlinear < 0.001). The HR increased with increasing mDTR, crossing HR = 1 at approximately 12.7 °C, followed by a less steep increase at higher mDTR levels.
Subgroup analysis
Subgroup analyses were conducted according to sex, age group, education, wealth, smoking, drinking, marital status, and residence (Figure 4 and Supplementary Figure 2). For both mDTR and mDMT, the associations with incident CVD were generally consistent across the examined subgroup strata, with mDTR showing positive associations and mDMT showing inverse associations. However, neither association was statistically significant among participants with upper secondary education or above. No statistically significant interaction was observed for mDMT across any of the examined subgroup factors. For mDTR, a statistically significant interaction was observed with drinking status (P for interaction = 0.013), with a stronger positive association among ever drinkers than among non-drinkers. However, given the multiple subgroup and interaction tests performed, this finding was considered exploratory and should be interpreted cautiously.
Figure 4 Forest plot of subgroup analyses for the association between mean diurnal temperature range and incident cardiovascular disease risk.
Hazard ratios and 95% confidence intervals are shown for each subgroup. P values for interaction are indicated on the right. Models were adjusted for age, sex, education, wealth, residence, marital status, smoking, drinking, body mass index, and fine particulate matter with an aerodynamic diameter ≤ 2.5 μm. CI: Confidence interval; CVD: Cardiovascular disease; mDTR: Mean diurnal temperature range; HR: Hazard ratio.
Sensitivity analyses
We conducted a series of sensitivity analyses to assess the robustness of the findings (Supplementary Table 11). After mutual adjustment, the association for mDTR was attenuated but remained directionally consistent (HR = 1.02, 95%CI: 1.00-1.04), while the inverse association between mDMT and incident CVD remained statistically significant (HR = 0.95, 95%CI: 0.94-0.97). Multicollinearity assessment showed no substantial multicollinearity in the mutual-adjustment model (Supplementary Table 12). Logistic regression analyses yielded results consistent with the primary analysis, with ORs of 1.06 (95%CI: 1.04-1.08) for mDTR and 0.94 (95%CI: 0.93-0.95) for mDMT. Complete-case analysis, 2-year lag analysis, and winsorization of exposure values at the 1st and 99th percentiles yielded broadly similar results. In the province-adjusted analysis, the associations remained evident for both mDTR (HR = 1.06, 95%CI: 1.03-1.09) and mDMT (HR = 0.92, 95%CI: 0.88-0.96). Overall, the direction and magnitude of the associations were generally consistent across sensitivity analyses.
DISCUSSION
Using data from the nationally representative CHARLS cohort, we systematically evaluated the associations of ambient temperature exposure and incident CVD among Chinese older adults. A higher mDTR was associated with an increased risk of incident CVD, whereas a higher mDMT was associated with a lower risk of incident CVD. RCS analyses revealed significant overall associations for both temperature indicators, with significant evidence of nonlinearity for mDTR but no evidence of nonlinearity for mDMT. The associations remained generally consistent across the prespecified subgroup and sensitivity analyses. The findings suggest that temperature variability and average temperature may both be relevant to CVD risk assessment in older populations and may provide complementary information for climate-related health assessments.
In recent years, amid global climate change, the relationship between ambient temperature and cardiovascular health has become a focal point in public health research. However, most previous studies have concentrated on short-term temperature fluctuations, including heatwaves, cold spells, and abrupt temperature changes, and have focused primarily on examining their acute effects on cardiovascular events[3,4,19-22]. Although short-term exposure may induce immediate physiological responses, the potential health implications of multi-year average temperature exposure remain less understood. Existing evidence on temperature exposure and CVD has largely focused on mortality outcomes[11,23], with limited longitudinal evidence examining new cases of CVD. Mortality is influenced not only by disease occurrence but also by subsequent survival and healthcare access. In contrast, studying incident CVD may provide a clearer understanding of the association between environmental exposure and disease development. In addition, China’s vast geographic area and substantial climatic heterogeneity offer a valuable context for examining temperature-related cardiovascular risks. Previous studies on temperature and cardiovascular health have largely been conducted in specific regions, limiting the generalizability of findings across populations with diverse climatic conditions[11,24]. Given that adults aged 60 years and older are particularly susceptible to CVD[25], we used nationally representative CHARLS data to evaluate the association between multi-year average ambient temperature exposure and incident CVD risk among older Chinese adults.
We observed a positive association between mDTR and incident CVD risk: Each 1-SD increase in mDTR was associated with approximately a 6% higher risk, and participants in Q4 had a 102% higher risk than those in Q1. These findings are consistent with emerging evidence linking long-term temperature variability to increased cardiometabolic burden[5]. Several biological mechanisms may underlie this association. Rapid intraday temperature fluctuations may challenge the adaptive capacity of the thermoregulatory, autonomic nervous, and cardiovascular systems[26]. Sustained exposure to pronounced diurnal temperature variation may disrupt homeostasis of the autonomic nervous system. Repeated cycles of peripheral vasoconstriction and vasodilation induced by daily temperature fluctuations may contribute to sustained sympathetic nervous system activation and abnormal circadian blood pressure rhythms. For older adults with reduced vascular elasticity, this repeated regulatory demand may impose persistent physiological stress and potentially contribute to the development of hypertension and the progression of atherosclerosis, thereby potentially increasing the risk of incident CVD when mDTR is elevated[27-29]. In addition, substantial diurnal temperature variability may influence outdoor activity patterns and other health-related behaviors, which may further contribute to adverse cardiovascular outcomes[30]. In the present study, the HR increased with increasing mDTR, crossing HR = 1 at approximately 12.7 °C, followed by a less steep increase at higher mDTR levels. These findings may have implications for regions characterized by greater diurnal temperature variability. Climate adaptation measures for older populations, including community-based early warning systems for substantial temperature fluctuations and efforts to improve indoor thermal environments, may warrant further consideration in future research[31,32].
In contrast to mDTR, mDMT was inversely associated with incident CVD risk, with a consistent dose-response pattern across quartiles. This inverse association aligns with previous cross-sectional evidence from China[33]. Populations living in colder environments may experience a greater cardiovascular burden, and a substantial proportion of temperature-related morbidity and mortality has been attributed to cold exposure[34]. Several mechanisms may explain this association. Prolonged exposure to low temperatures may sustain sympathetic nervous system activation to maintain core body temperature, resulting in peripheral vasoconstriction, elevated blood pressure, and increased cardiac workload[7,8]. In addition, chronic cold exposure may stimulate thermogenic metabolism and alter lipid metabolism, potentially contributing to the accumulation of atherogenic lipoprotein remnants and promoting atherosclerotic progression[35,36]. Cold exposure may also increase coagulation activity, as indicated by elevated fibrinogen levels, thereby increasing blood viscosity and thrombotic risk[37,38]. These findings highlight the potential importance of cardiovascular risk management during cold seasons, particularly among vulnerable older populations. Potential strategies that could be considered include improving heating infrastructure, establishing community-based warming programs, and strengthening health monitoring among vulnerable populations during cold seasons[39]. In addition, further adjustment for province did not materially alter the associations of mDTR or mDMT with incident CVD, suggesting that the findings were not substantially explained by broad geographic heterogeneity. Nevertheless, residual confounding at finer geographic or local environmental levels cannot be completely excluded.
Several limitations warrant consideration. First, CVD outcomes were based on self-reported physician diagnoses, which may introduce reporting or recall bias; however, the CHARLS questionnaire has been rigorously validated, and self-reported disease information has been widely used in previous studies. Second, temperature exposure was assigned using city-level meteorological data rather than individual-level microenvironmental monitoring. This approach may not fully capture individual-level exposure or within-city spatial variability, and the absence of information on individual-level time-activity patterns may have resulted in exposure misclassification. Third, despite adjustment for a wide range of covariates, residual confounding from unmeasured or incompletely measured factors cannot be completely excluded. In particular, spatial confounding and clustering of temperature exposure at the city level may have contributed to residual geographic confounding. Regional healthcare accessibility, dietary patterns, built environment characteristics, and unmeasured socioeconomic factors were not fully captured in the available data. Information on indoor heating/cooling facilities and residential building characteristics was unavailable or incomplete in Wave 1 and therefore could not be incorporated into the multivariable models. Finally, the study population was restricted to Chinese adults aged 60 years and older. Given the notable differences in climatic conditions, population characteristics, living environments, and healthcare systems across countries and regions, the generalizability of our findings to other settings or age groups should therefore be interpreted with caution. Future comparative studies across countries and climatic zones are warranted to further elucidate the association between temperature exposure and cardiovascular health in diverse populations and to explore the underlying mechanisms.
CONCLUSION
Based on the CHARLS cohort, we found that among adults aged 60 years and older, higher mDTR was associated with an increased risk of incident CVD, whereas higher mDMT was associated with a reduced risk. Our findings suggest that both diurnal temperature variability and average temperature may be relevant to cardiovascular health among older populations. Further studies are needed to clarify the causal nature of these associations and to determine whether and how specific climate adaptation and cardiovascular prevention strategies may help reduce temperature-related cardiovascular risks.
ACKNOWLEDGEMENTS
The authors express their sincere gratitude to the CHARLS research team for their time and effort they have devoted to their respective projects, and to all individuals who participated in the study. This analysis uses data or information from the Harmonized CHARLS dataset and Codebook, Version D as of June 2021, developed by the Gateway to Global Aging Data.
Wang L, Liu J, Yin P, Gao Y, Jiang Y, Kan H, Zhou M, Ao H, Chen R. Mortality risk and burden of sudden cardiac arrest associated with hot nights, heatwaves, cold spells, and non-optimum temperatures in 0.88 million patients: An individual-level case-crossover study.Sci Total Environ. 2024;949:175208.
[RCA] [PubMed] [DOI] [Full Text][Cited by in RCA: 9][Reference Citation Analysis (0)]
Lechner K, Zhang S, Krüger N, Krefting J, Wolf K, Dallavalle M, Dreischulte T, Offenborn F, Starnecker F, von Scheidt M, Chen K, Peters A, Breitner S, Schneider A, Schunkert H. Upward Air Temperature Shifts and Acute Cardiovascular Events in Individuals with Atherosclerotic Cardiovascular Disease: A Time-Stratified Case-Crossover Study.Lancet Reg Health Eur. 2026;67:101730.
[RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)][Cited by in RCA: 1][Reference Citation Analysis (0)]
Austin PC, Giardiello D, van Buuren S. Impute-then-exclude versus exclude-then-impute: Lessons when imputing a variable used both in cohort creation and as an independent variable in the analysis model.Stat Med. 2023;42:1525-1541.
[RCA] [PubMed] [DOI] [Full Text][Cited by in RCA: 6][Reference Citation Analysis (0)]
Gasparrini A, Guo Y, Hashizume M, Lavigne E, Zanobetti A, Schwartz J, Tobias A, Tong S, Rocklöv J, Forsberg B, Leone M, De Sario M, Bell ML, Guo YL, Wu CF, Kan H, Yi SM, de Sousa Zanotti Stagliorio Coelho M, Saldiva PH, Honda Y, Kim H, Armstrong B. Mortality risk attributable to high and low ambient temperature: a multicountry observational study.Lancet. 2015;386:369-375.
[RCA] [PubMed] [DOI] [Full Text][Cited by in Crossref: 2738][Cited by in RCA: 1725][Article Influence: 156.8][Reference Citation Analysis (0)]
