Introduction
As in many other countries, population ageing in China has emerged as an increasingly important social issue in the last few decades, driven especially by the combination of decreased birth rates and prolonged life expectancy [
1]. A key step to address the societal public health challenge of ageing is to quantify the health impact associated with population ageing accurately [
2].
Previous research has described changes in deaths, incidence, prevalence, and disability-adjusted life years (DALYs) associated with population ageing for specific diseases in particular countries and regions, including China [
3‐
17]. However, these studies suffer from methodological limitations. Studies implementing traditional decomposition methods fail to generate robust results because those results are sensitive to the selections of decomposing order of population size, age structure and age-specific rate, as well as the choice of the reference group for comparisons [
6,
7,
9,
10]. In addition, studies relying on projection models [
4,
5,
8] depend heavily on the validity of model assumptions, which can deviate from reality and prevent researchers from properly distinguishing the effects of population ageing from that of population growth.
Another limitation of existing research is in the use of data from different sources and/or different time periods for decomposition analyses [
15‐
17], making comparisons across studies difficult. In addition, many previous studies focus only on a single health outcome – either mortality [
3,
5,
12], prevalence [
9‐
11] or incidence [
4,
6‐
8,
12] – each capturing only a slice of the health effects of population ageing.
Amidst the broader global literature, several studies from China offer evidence for the need to study population ageing among the Chinese population. A study by Xu et al. [
11], for example, projected population ageing would cause significantly increased costs to treat dementia in China between 2020 and 2030. Three other studies quantified the impact of population ageing on cardiovascular incidences [
12,
13] and deaths [
14] in China. They reported an increase of coronary heart disease deaths related to population ageing in Beijing from 1990 to 2010 and predicted future cardiovascular incidence related to population ageing. One recent study quantified deaths related to population ageing in 195 countries/territories from 1990 to 2017, including China, and for 169 kinds of diseases and injuries [
18]. While offering valuable data for the field, this study excluded the effect of population ageing on prevalence cases (non-fatal health outcomes) from the data analyses and did not cover decomposition details across sex and type of disease. Thus, assessing the impact of population ageing on a broad spectrum of health outcomes in China remains highly needed.
Developed from many different data sources and sophisticated statistical models, the Global Burden of Diseases, Injuries, and Risk Factors Study 2017 (GBD 2017) offers estimates of key health outcomes across 359 diseases and injuries, 84 modifiable risk factors or clusters of risks, and sex-specific population sizes for 195 countries and territories from 1990 to 2017 [
19]. The dataset therefore provides comparable data across time periods and countries to permit accurate decomposition of the change in health outcomes associated population ageing.
Based on a recently-developed robust decomposition method [
20], we employed estimates from the GBD 2017 to extend the literature by examining two research questions:
(a)
Was the change in DALYs associated with population ageing in China from 1990 to 2017, and did the contribution of population ageing to the temporal change in DALYs vary by sex and by the type of disease?
(b)
Did the changes in population ageing and change in the age-specific DALY rates, including overall rate and the attributed rate defined by GBD 2017, counteract each other in their effects on the overall DALY change from 1990 to 2017 in China, and did the level of counteraction vary by sex and by type of disease?
Discussion
This study presents four major findings that are novel to the field. First, an increase of 92.8 million DALYs was associated with population ageing in China between 1990 and 2017, with 65.8% of that increase in YLLs. Second, in China, males suffered more health losses associated with population ageing compared to females between 1990 and 2017 (52.8 vs. 40.4 million); stroke, chronic obstructive pulmonary disease and ischemic heart disease were most affected by population ageing, contributing 54.9 million (59.2%) to total DALY loss. Third, the increase in DALYs related to population ageing between 1990 and 2017 was offset by the decrease in DALYs due to reduced DALY rates (including overall rate and -risk-attributable rates). Last, the counteracting effect between decreasing DALY rates and population ageing was roughly similar between males and females but varied greatly across the level-3 disease categories.
Extending previous research on decomposing mortality changes in China [
18], this study offers an overall picture of the health impact of population ageing in China using DALY as a metric. DALY offers a summary measure of morbidity, disability, and mortality, and therefore reflects overall health impact of diseases and injuries in a country compared to single fatal or non-fatal health indicators. The decomposition results obtained via applying a robust method [
20] to GBD 2017 data is a valuable addition to the literature [
12‐
14], providing rigorous estimates for China that can readily be compared across different time periods and with other countries for a variety of health outcomes [
19]. We recommend the same decomposition method for analyzing GBD 2017 estimates for other countries to maximize comparability.
The small fluctuation in DALYs associated with population ageing we discovered before 2002 might be primarily due to the joint effects of temporal variations of both age structure and DALY rates across age groups in the population. For example, the proportion of people aged 30 to 65 with a fairly low DALY rate increased from 35.6 to 45.9% between 1990 and 2002, and the proportion of the under-five age group with an extremely high DALY rate decreased from 13.4 to 6.6% in the same time period.
The steady increase of DALYs associated with population ageing since 2002 was primarily caused by extremely high DALY rates among older people and the rise of the number of older people in the population rather than changes in population size or changes in age-specific rates. Ageing typically is related to progressive loss of physiological integrity, leading to impaired functions and increased vulnerability to morbidity and mortality among people in the oldest age groups [
26,
27]. The increase in numbers of older people in China corresponds with prolonged life expectancy, which is attributed to substantial social development, especially improvement in health care services [
28]. Between 1990 and 2017, the life expectancy of Chinese people rose from 68.7 to 70.1 years [
29].
Interestingly, DALYs associated with population ageing in China were significantly larger among males than among females even though females had a comparatively higher life expectancy, a higher proportion of old people, and a larger number of old people. This finding reflects the comparatively higher DALY rate among males (29,019 per 100,000 population in 2017) compared to females (23,464 per 100,000 population) [
21]. Similarly, variations in health losses associated with population ageing across different disease categories reflect varying changes in DALY rates.
Encouragingly, the health losses associated with population ageing in China were wholly counteracted by the reduction in DALY rates over time. Notably, the reduction of risk-attributable rates adequately offset the increase in DALYs associated with population ageing for COPD, hypertensive heart disease, chronic kidney disease and tuberculosis. This reflects considerable progress in prevention and control efforts for certain diseases such as COPD and stomach cancer. Progress has been inconsistent across disease categories, however, diseases like ischemic heart disease and diabetes should be targeted for successful ageing because the earn from reductions in DALY rates was insufficient to offset the health loss from population aging, and continued work is needed to control preventable health outcomes through evidenced-based strategies like prohibiting smoking in public places, reducing harmful use of alcohol, and promoting physical activity that have potential to continue such progress [
30‐
32].
For diseases with substantial burden that is difficult to control, like dementia and age-related hearing loss, DALY rate reductions between 1990 to 2017 were minimal, and insufficient to offset the burden associated with population ageing. Continued research to understand and ultimately prevent such diseases is recommended [
33,
34].
Our findings have three major implications. First, they demonstrate the enormity of health loss associated with population ageing in China. As life expectancy continues to rise, we might expect increased health losses in the future. This calls for urgent action to address the challenges of an ageing society. Second, our results illustrate how the health burden associated with population ageing can be offset by prevention and treatment efforts. Holistic efforts should be made to reduce population exposures to known modifiable risk factors for diseases. These efforts will require government investments to disseminate effective strategies that are available but not fully implemented nationwide; examples include creating age-friendly products and a built environment and promoting a healthy lifestyle among older adults [
35]. Last, our findings underscore the value of increasing government investment in scientific research. Research is needed to generate effective prevention and treatment strategies, including for currently unpreventable diseases, and to explain observed changes in DALY rates that cannot be attributed to the known modifiable risk factors.
The primary limitation of this study is in its use of the GBD 2017 dataset. As previously noted [
36‐
38], GBD 2017 estimates are affected by the paucity and low quality of raw data although advanced and complex models have been adopted to impute missing values and optimize the estimation for each country or territory, including China. Furthermore, because full posterior samples of cause-specific rates stratified by age sex, location, and year are not freely accessible for GBD 2017 [
19], we were unable to provide 95% uncertainty intervals for the estimates.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.