Background
Ischemic heart disease (IHD) is characterized by myocardial ischemia caused by the narrowing of coronary vessels, which supply blood to the heart, and it represents one of the leading causes of deaths worldwide and causes a tremendous disease burden [
1]. During the past two decades, the disease burden of IHD has markedly increased in China, with the age-standardized mortality rate increased by 20.6% from 1990 to 2017, and IHD has already become the second leading cause of death in China and was surpassed only by stroke in 2017 [
2]. Epidemiological studies have demonstrated that temperature is an important environmental determinant among the many risk factors for IHD, with both extremely cold and heat associated with increased mortality risk from IHD [
3,
4].
Global climate change is one of the most serious environmental health problems currently faced by human populations [
5]. The Intergovernmental Panel on Climate Change Fifth Assessment Report (IPCC AR5) highlighted that the global mean surface temperature on earth is projected to rise in the future [
6]. Many countries worldwide have experienced tremendous burdens of heat-related and cold-related deaths under the current climate patterns [
7‐
9], and global climate change will likely exacerbate such risks [
10,
11]. Thus, creating projections of the temperature-related disease burden of IHD for future climate change scenarios is important for the development of environmental and public health policies to mitigate the health effects of extreme temperatures. However, evidence in this field is rare.
Aging has also emerged as a significant social and public health challenge. The elderly population will increase rapidly in the twenty-first century according to the United Nations (UN) projection [
12]. Considering the elderly are more vulnerable to temperature-related health risks [
13‐
15], this demographic change may lead to an increased disease burden from IHD in combination with extreme temperature exposure. Nevertheless, our understanding of the emerging temperature-related health risks of IHD in the elderly is limited because of the lack of estimates that integrate global warming and demographic change [
16].
In addition, based on the data from the National Bureau of Statistics of China, the population aged ≥65 years old in China has already reached 158 million in 2017, which accounted for 11.4% of the total population. This indicated China has already entered an “aging society” according to the definition by UN, and this trend is still accelerating [
17]. Thus, we selected the 65 and older group as the target population to evaluate the effects of global warming on the elderly population whose number will increase rapidly in the next decades.
Previous studies have used mortality as the main health outcome to assess the health impacts of temperature, and the influence of age at death was not taken into account [
3,
4,
18]. Years of life lost (YLL), which is an indicator of disease burden, accounts for both premature death and life expectancy at death [
19]. However, the availability of studies on the relationships between temperature and YLL from IHD in the elderly is rare, and projections have been insufficient up to now.
The effects of global warming on the disease burden of IHD in the elderly were estimated in a megacity of China in this study. Different climate change scenarios and demographic change in the elderly were considered in the projections. Considering the tremendous disease burden of IHD and its increasing trend in China, our study will provide important information for environmental and public health interventions aimed at reducing the temperature-related disease burden of IHD in an ageing population under global warming scenarios in the future.
Methods
Study area
The study area is Tianjin, the third largest municipality in China. It has a population of approximately 12.9 million, with the elderly (age ≥ 65 years) accounting for approximately 8.52% at baseline (Tianjin Statistical Information Site, 2006–2011,
http://stats.tj.gov.cn/Category_29/Index.aspx). Tianjin has a typical temperate climate characterized by hot, rainy summers and cold, dry winters.
Data collection
Daily mortality data on IHD in the elderly (age ≥ 65 years) from Jan 1st 2006 to Dec 31st 2011 in Tianjin were collected from Death Registration and Reporting System of the Chinese Centre for Disease Control and Prevention. The data were include all the residents of Tianjin, and the mortality data were representative of the study area. The causes of IHD were classified and coded according to the International Classification of Disease, 10th version (IHD: I20-I25). Data permission was obtained and the study was approved by the Ethics Committee/Institutional Review Board of Peking University Health Science Center.
Indicator of YLL was used in this study. The method to calculate YLL was used in previous studies as the follow equation [
19,
20].
$$ \mathrm{YLL}=\sum Yi\times Li $$
where
Yi is the death number for a specific age group
i, and
Li is the remaining life expectancy for a specific age group
i.
First, we matched each person’s age at death to the World Health Organization (WHO) standard life table (Additional file
1: Table S1). Then, the daily YLL of the elderly from IHD in our study area was estimated by summing the YLL of all individuals who aged ≥65 years and died from IHD on the same day.
Daily meteorological data, including the daily relative humidity and maximum temperature, were obtained from the Tianjin Meteorological Bureau, which were match with the daily mortality data of IHD during the study period from 2006 to 2011. The average daily concentrations of particulate matter with aerodynamic diameter ≤ 10 μm (PM10) were also collected from the Tianjin Environmental Monitoring Centre to allow for the evaluation of the confounding effects of air pollutants.
Statistical analyses
First, the baseline temperature-YLL relationships from 2006 to 2011 were established, then future assessments were made in combination with future temperature projections.
Distributed lag non-linear models (DLNMs) were used to measure the non-linear and delayed effects of temperature on YLL from IHD during the baseline period [
21]. Seasonal and long-term trend were adjusted using a natural cubic spline function with 7 degrees per year, and day of the week as a categorical variable. The daily PM
10 and relative humidity were adjusted using a natural cubic spline with 3 degrees of freedom. A natural cubic B-spline basis with 5 degrees of freedom for temperature and a maximum lag of 15 days between temperature and YLL with 5 degrees of freedom were chosen in the baseline analysis. Optimal temperature (OT) was used as a reference to calculate the temperature-related YLL related to high or low temperatures. The OT was determined according to the exposure-response curve for temperature in YLL for IHD during the baseline period.
Temperature projections were made according to Representative Concentration Pathways (RCPs) reported in the IPCC AR5. Three RCPs scenarios were selected in our analysis, including RCP2.6, RCP4.5 and RCP8.5, which represent the mild, the medium and the high emission scenarios. The future daily temperatures for periods of 2046 to 2065 centred on 2050s, and 2061 to 2080 centred on 2070s, were developed from 19 global-scale climate models (GCMs) from the World Climate Research Programme (WCRP) Coupled Model Intercomparison Project Phase 5 (CMIP 5) multi-model dataset (Additional file
1: Table S2).
As for the temperature calibration, we used method as follows: it started with the projected change in a weather variable (i.e. maximum temperature in June). This is computed as the (absolute or relative) difference between the output of the GCM run for the baseline years and for the target years (e.g. 2050s–2070s). These changes are then added to the observed baseline to create the projected temperature in a given year for a specific emission scenario (in this case the WorldClim Database,
http://www.worldclim.org/downscaling). Through this method, we could capture the trend of temperature change in future projection. And this approach was also used in previous study [
22].
The projection of temperature-related YLL from IHD in the elderly was calculated by integrating the temperature projections under different RCPs with baseline exposure-response relationships. The projections and changes of annual heat-related, cold-related and total temperature-related YLL were calculated.
Furthermore, adaptation of the population was taken into consideration because people could adapt to warmer climatic conditions through a number of measures [
23]. A 25% acclimatization factor was assumed according to the previous study of a U.S. city, which reported that excess mortality related with heat reduced by approximately 25%, indicating population adaptation to heat in recent decades through increased using of air conditioning, greater awareness of the risks by high temperature, and introduction of heat-warning systems, etc. [
24].
Considering the temperature-YLL relationship may not remain stable over time due to population adaptation, we modelled the adaptation by shifting the OT and shape of temperature-YLL curves [
25]. The model combining the absolute threshold shift of OT with the reduction in the slope of the heat exposure-response function was used in this study because it may balance the uncertainty among adaptation models, climate models and emissions [
26]. Previous studies have suggested that OT could continue to rise with increasing temperature due to adaptation [
26,
27]. Absolute threshold shift of OT is a popular method for modeling adaptation [
26], and applying a shift in absolute threshold of OT between 1.0 °C and 4.0 °C in future is recommended because this is broadly within the range of shifts in threshold temperature observed in previous epidemiological studies [
27‐
29]. Thus, we conservatively estimated that the OT will increase by 1.0 °C in 2050s, and by 1.2 °C in 2070s relative to the baseline [
30].
Demographic change was also taken into account. Based on the low, medium and high variant scenarios of population growth among the population aged 65 years and above in China employed by the UN, this population size will be 3.1 and 3.2 times greater than the baseline population in 2010 by the 2050s and 2070s, respectively [
12].
Sensitivity analyses were performed to test whether the results were robust to changes in the parameters in the model, including using 4 degrees of freedom of relative humidity, and a natural cubic B-spline with 5 degrees of freedom for temperature and a maximum lag of 15 days between temperature and YLL with 6 degrees of freedom.
Discussion
This study estimated future temperature-related YLL from IHD in the elderly in Tianjin, China, in the 2050s and 2070s. To the best of our knowledge, this is the first study that has estimated the effects of global warming on the disease burden of IHD in the elderly using the 19GCMs as well as the RCP scenarios and considering ageing population growth over the twenty-first century.
We found substantial increases in heat-related IHD YLL but relatively slight decreases in cold-related IHD YLL under future global warming scenarios. The increments in heat-related IHD YLL will not be offset by the reduction in cold-related IHD YLL. The finding indicated an increasing disease burden from IHD in the elderly caused by rising temperatures in the future. And the results were consistent with previous studies which also found the increase in the heat-related deaths is unlikely to be offset by the decrease in cold-related deaths [
5,
31].
Exposure to heat may induce profound physiological changes, including increased blood viscosity and cardiac output, which can lead to dehydration, hypotension, increased surface blood circulation and even endothelial cell damage [
32]. These changes can cause haemoconcentration and overload the function of the heart.
A study conducted in Beijing, China pointed out that the heat-related cardiovascular deaths in the whole population are projected to increase by 135% under RCP8.5 in the 2080s compared with the baseline [
33]. While in our study, more than 150% increase was found under RCP8.5 in the 2070s compared with the baseline in the elderly, which indicated more pronounced effect was observed in the elderly.
Considering that aging reduces the ability to thermoregulate, disrupts homoeostasis and the elderly were more likely to living alone, the elderly were more susceptible to the adverse effects of extreme temperatures [
13]. Thus more attention should be paid to the temperature-related IHD disease burden in the elderly under global warming scenarios.
Monthly analyses showed that large percent increases occurred mainly in the warm season months from May to September. Considering global climate change is likely to produce more frequent, more intense and longer-lasting heatwaves, the elderly will experience additional stress in the warmer months. These findings will provide important information for environmental and public health resource distribution for IHD according to the monthly change patterns in future.
Because people may adapt to warming climates by increasing their use of air conditioning, improving building design and city planning, implementing early warning systems and changing behaviours [
31], human adaptation to a warming climate was also considered in this study. However, the increment in heat-related YLL will not completely offset under the medium emission scenario in the 2070s or the high emission scenario in the 2050s and 2070s, even with an assumed 25% adaptation.
Demographic change is another important factor that should be considered when exploring the health impacts of climate change. Nevertheless, most previous studies have focused only on climate change and ignored possible demographic change [
18,
34]. According to UN projections, the world population is ageing rapidly [
12]. Therefore, if population growth in the elderly is not properly considered, the temperature-related health burden will be greatly underestimated. Compared with the baseline, the total temperature-related YLL in the elderly will increase rapidly in the 2050s and 2070s by 158.4 to 196.6% in Tianjin, China.
There are several strengths of this study. First, to our knowledge, it is the first time to explore the temperature-related disease burden of IHD in the elderly under future climate change scenarios. These data will provide scientific evidence for policy making on the important issue of global warming on the elderly. Second, monthly analyses were conducted to explore changes in temperature-related YLL from IHD by different months, which can be crucial for resources allocation for IHD according to different monthly patterns. Third, the disease burden was estimated using the indicator of YLL, which is an informative measurement for quantifying premature mortality [
35]. In addition, projections were made that accounted for both the adaptation of the population to the warming climate and demographic change in the elderly.
This study also had certain limitations. First, the data in our study were limited to one megacity with typical temperate climate, thus generalizing the results to other geographic areas should be performed with caution. Second, temperature data from the fixed sites rather than the individual data were used, but measurement errors may bias the results towards the null hypothesis [
36]. Third, climate change is likely to produce more frequent and more intense heatwaves. Because additional effects of heat waves may be observed [
37], our results may have underestimated the future IHD YLL in the elderly due to climate change.
Acknowledgements
We thank the Chinese Centres for Disease Control and Prevention for providing the daily mortality data on IHD, the Tianjin Meteorological Bureau for providing the daily meteorological data, and the Tianjin Environmental Monitoring Centre for providing the average daily PM10 data.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.