Extreme Heat and Cold Each Raised Chronic Lung Disease Risk in a 9,458-Person Cohort
A long-running Chinese cohort study tied both cold spells and heat waves to a higher incidence of chronic lung disease in middle-aged and older adults, with risk rising across 50 different temperature-duration combinations. Over a median 7.0 years of follow-up, 1,562 participants developed chronic lung disease, an incidence of 16.51% in the analytic sample of 9,458 people drawn from five waves of CHARLS between 2011 and 2020.
The result is notable not just because both temperature extremes tracked with disease, but because moderate, repeated exposures carried a heavier burden than rarer, more intense events in several comparisons. In practical terms, that pattern fits a cumulative injury model: airways and pulmonary vasculature may tolerate a single short burst of environmental stress, yet repeated thermal stress can push susceptible lungs toward chronic inflammation, impaired mucociliary clearance, bronchoconstriction, and accelerated decline in respiratory reserve.
Fifty Exposure Definitions, One Consistent Direction of Effect
Investigators constructed 50 exposure metrics by combining percentile thresholds with duration windows. Cold spells ranged from the P2.5 to P10 thresholds, while heat waves ranged from P85 to P97.5, each paired with durations from at least 1 day to at least 5 days. Across that grid, every metric was associated with increased chronic lung disease risk after Benjamini-Hochberg false discovery rate correction, and all adjusted P values were below 0.01.
At the longest duration examined, the pattern ran against the assumption that only the most severe thermal events matter most. For cold spells lasting at least 5 days, the P10 threshold carried an odds ratio of 1.34 (95% CI 1.27–1.42), exceeding the effect size for the more extreme P2.5 threshold, which had an odds ratio of 1.13 (95% CI 1.07–1.20). Heat waves showed the same broad shape: the P85 threshold at ≥ 5 days reached an odds ratio of 1.32 (95% CI 1.25–1.40), while the P97.5 threshold reached 1.20 (95% CI 1.13–1.27).
A blunt clinical implication follows from those numbers. Repeated moderate exposure may be doing more population-level damage than dramatic but less frequent weather episodes.
Cold Spell Associations Were Nonlinear; Heat Wave Effects Stayed Linear
The exposure-response curves separated into two biologically plausible patterns. Cold spells displayed nonlinear relationships, with a P5 model showing a significant non-linearity signal (P-non-linearity < 0.001). Heat waves, by contrast, behaved in a linear fashion across the P90 to P97.5 range, with all corresponding P-non-linearity values above 0.50.
Those shapes matter because they suggest different physiologic pathways. Cold air can trigger abrupt airway cooling, bronchospasm, altered epithelial barrier function, and changes in airway blood flow, effects that may intensify once a certain threshold of duration is crossed. Heat exposure may act more steadily through dehydration, oxidative stress, higher cardiopulmonary strain, and worsened pollutant interactions, producing a more graded rise in risk rather than a threshold-driven jump.
The study also reported one effect modifier in the FDR-adjusted interaction analysis for cold spells: residence. The adjusted P-interaction was 0.042, and the cold-spell association was stronger among rural residents. Rural populations may face greater exposure to heating insecurity, housing insulation deficits, and occupational outdoor work, all of which can magnify the respiratory toll of sustained cold.
Multiple Sensitivity Analyses Pointed in the Same Direction
The primary associations held up when examined with alternative modeling strategies. Cox models yielded a hazard ratio of 1.28 (95% CI 1.22–1.35), and Fine-Gray competing risk models produced a subdistribution hazard ratio of 1.26 (95% CI 1.20–1.33). Fixed exposure window validation and other sensitivity analyses supported the same direction of effect.
That consistency strengthens the inference that the findings were not an artifact of one statistical approach. It also suggests the signal was robust to competing events and to the way exposure windows were specified. In a cohort of older adults, those checks matter because mortality and attrition can distort respiratory incidence estimates if they are not handled carefully.
The study design cannot establish a direct causal mechanism, but the epidemiology fits known respiratory biology. Chronic lung disease develops when repeated insults converge on airway inflammation, remodeling, and loss of pulmonary resilience. Temperature extremes can plausibly accelerate that process by worsening bronchial reactivity, impairing host defense, and increasing susceptibility to respiratory infections and pollutant-related injury. In older adults, whose physiologic reserve is already limited, even modest environmental stress can shift the balance toward symptomatic and clinically recognized disease.
What the Cohort Means for Climate-Health Planning
This was a prospective cohort study using data from CHARLS, a large national platform with longitudinal follow-up and written informed consent under approved ethics procedures. The study’s public-health message is straightforward: climate adaptation planning cannot focus only on catastrophic weather events. Sustained moderate cold and heat may produce a broader burden of chronic respiratory disease than the most extreme episodes, especially in older adults and in rural communities.
For clinicians, the findings reinforce a familiar but often underemphasized risk conversation. Patients with chronic cough, dyspnea, prior smoking exposure, or limited access to climate control may need anticipatory counseling before prolonged cold or heat periods. For health systems, the study argues for respiratory risk mapping that includes housing quality, local weather patterns, and the cumulative duration of exposure, not just peak temperature anomalies.
The analysis stops short of identifying individual-level prevention thresholds or proving which adaptation measure would blunt the risk most effectively. It also does not separate specific chronic lung diagnoses within the composite endpoint. Even so, the message from this cohort is hard to ignore: temperature extremes, repeated over days and seasons, were consistently associated with incident chronic lung disease, and rural residents appeared to bear a heavier cold-related burden.