The GBD study offers a tool for quantifying health loss from numerous diseases, injuries, and risk factors (14, 15). Asthma incidence and mortality data in China (without Taiwan province) for this study were retrieved from the Global Health Data Exchange (GHDx).¹ In GBD 2021, asthma is a chronic lung disease marked by spasms in the bronchi usually resulting from an allergic reaction or hypersensitivity and causing difficulty in breathing. The group defined asthma as a doctor’s diagnosis and wheezing in the past year. The relevant International Classification of Disease and Injuries (ICD-10) codes are J45 and J46. ICD-9 code is 493. Alternative case definitions include self-reported asthma in the past year, self-reported asthma ever, only a doctor’s diagnosis in the past year, and only wheezing in the past year. Vital registration and surveillance data from the cause of death (COD) database were used to estimate asthma mortality.

To assess asthma burden trends in China from 1990 to 2021, age-standardized incidence rate (ASIR) and age-standardized mortality rate (ASMR) were calculated. The estimated annual percentage change (EAPC) was used to evaluate age distribution differences across populations. To account for heterogeneity due to sampling error and non-sampling variance, an uncertainty interval (UI) analysis was conducted. The 95% UI was derived from 1,000 samples of the posterior distribution at each modeling step, with the 2.5th and 97.5th percentiles reported for each estimate.

Joinpoint regression model, conducted by Joinpoint software (version 5.2.0), was employed to analyze time trends systematically from 1990 to 2021 and assess the statistical significance of trends between join-points. To quantify the direction and magnitude of these trends, the annual percent change (APC) and average annual percentage change (AAPC) were calculated with confidence interval (CI). The APC evaluates changes within specific periods, while AAPC provides an overall measure of overall changes over time (16). Statistical significance was set at a two-tailed p-value <0.05.

The age-period-cohort model is a sophisticated research method that enhances traditional analyses in health and socio-economic studies, particularly for examining trends in chronic disease incidence and mortality. It captures disease trends by considering age, period, and cohort effects. The National Cancer Institute’s Age-Period-Cohort web tool was used for model fitting (17). In this framework, the age and period intervals must be consistent, thus, the analysis was conducted using data from 1992 to 2021, divided into six continuous 5-year periods (1992–1996 to 2017–2021). Age groups were also defined in 5-year intervals, ranging from under-5-year age to over-95-year age.

We predicted asthma-related new cases and deaths from 2022 to 2046 using the Bayesian age-period-cohort (BAPC) model with integrated nested Laplace approximations (INLA), based on asthma burden data from the GBD database from 1990 to 2021. The analysis was performed in R (version 4.4.1) using the “BAPC” (version 0.0.36) and “INLA” (version 24.06.27) packages. Data visualization was carried out with the “ggplot2” (version 3.5.1) package.

Between 1990 and 2021, China experienced a significant decline in both all-age asthma incidence and mortality (Table 1). The number of new asthma cases per 100,000 population decreased from 459.22 (95% UI, 394.46–637.92) to 276.57 (95% UI, 223.74–361.00), reflecting a 39.77% reduction. Similarly, deaths owing to asthma declined from 3.08 (95%UI, 2.43–4.42) to 1.84 (95%UI, 1.45–2.24), representing a 40.26% decrease. Corresponding trends were observed in the ASIR, which decreased from 524.81 (95% UI, 421.31–672.76) to 364.17 (95% UI, 283.22–494.10), and in the ASMR, which dropped from 5.82 (95% UI, 4.46–8.50) to 1.47 (95% UI, 1.15–1.79).

Both males and females exhibited similar trends in asthma incidence and mortality over the past decades, with higher rates observed in males. In males, new asthma cases decreased from 530.89 (95%UI, 421.17–690.47) in 1990 to 312.56 (95%UI, 251.83–411.59) in 2021, while the deaths decreased from 3.22 (95%UI, 2.30–5.24) to 2.03 (95% UI, 1.42–2.65). For females, new cases declined from 457.23 (95% UI, 367.07–578.61) to 238.84 (95% UI, 195.13–306.20), and mortality dropped from 2.94 (95% UI, 1.84–4.12) to 1.65 (95% UI, 1.13–2.31). Regarding age-standardized rate, both ASIR and ASMR were consistently higher in men than in women throughout the study period. Specifically, the ASIR for men decreased from 561.78 (95% UI, 448.43–725.78) to 404.54 (95% UI, 312.57–547.88), while for women, it decreased from 485.03 (95% UI, 392.61–617.31) to 319.56 (95% UI, 248.24–429.99). Both genders experienced a decline in ASMR, with a notable reduction in men from 7.11 (95% UI, 4.95–12.28) to 1.95 (95% UI, 1.38–2.50), and in women from 4.96 (95% UI, 3.10–7.21) to 1.14 (95% UI, 0.78–1.59).

From 1990 to 2021, the EAPC for asthma incidence in the entire population was −1.17 (95%UI, −1.56 to −0.78). Regard to gender, a greater annual reduction in incidence was observed among females. The EAPC for asthma mortality was −4.69 (95%UI, −4.93 to −4.45) for the overall population, and a more significant annual decrease in mortality was also observed in females than in males.

Asthma incidence and mortality varied across age groups. In both 1990 and 2021, the highest incidence occurred in the under-5 age group, with over 50% of cases in 1990 occurring in children under 10 years, and the majority in 2021 occurring in those under 15 years, highlighting the disease’s predominance in children and adolescents. Incidence also increased in individuals over 60 years of age (Figures 1A,C), as reflected in the crude incidence rates (Figures 2A,C). The age group with the highest number of deaths shifted from 75–79 years in 1990 to 80–84 years in 2021. In 1990, 72.28% of deaths were in individuals aged 60 or older, rising to 88.04% by 2021. Deaths in children under 5 decreased from 4.32% in 1990 to less than 0.1% in 2021 (Figures 1B,D). Mortality rates were significantly higher in older age groups, with a notable drop in death rates for males over 95 years, observed in both 1990 and 2021 (Figures 2B,D).

To investigate changes in asthma incidence and mortality in China from 1990 to 2021, a Joinpoint regression analysis was performed (Figure 3). Overall, both incidence and mortality exhibited a downward trend during this period.

Prior to 2005, the crude incidence rate decreased significantly, with the most notable decrease between 1996 and 2005. After 2005, the downward trend ceased, with a rise in incidence at an APC of 1.02%. However, the decline resumed after 2017. The AAPC for incidence over the entire period was −1.89%. Both male and female incidence rates followed similar trends, but a significance test revealed more join-points and greater fluctuations in females, particularly after 2005. Notably, after 2005, the crude incidence rate increased more in males than females (Figure 3A). Similarly, the ASIR for the total population also declined significantly until 2005, then began to rise. From 2010, the decline resumed, continuing through 2021. The AAPC for incidence over the entire period was −1.21%. Both male and female ASIRs followed the same general trend, with the gender gap in ASIR remaining relatively stable (Figure 3B).

The crude mortality rate of the total population exhibited a consistent downward trend, with an AAPC of −1.70%. A slight increase was observed between 1998 and 2004, though not statistically significant. The most substantial decline occurred between 2004 and 2007, with an APC of −7.90%. Analysis by gender revealed distinct trends. Female mortality declined continuously before 2004, while male mortality reached its lowest point in 1998, followed by a slight increase. Between 2004 and 2007, both male and female mortality rates declined sharply. However, after 2007, the rate of decline slowed, and female mortality showed a slight increase in 2013 through to 2021 (Figure 3C). The ASMR steadily decline from 1990 to 2021 for both the overall population and by gender. The AAPC for the entire population was −4.40%, indicating a significant reduction. While fluctuations occurred, the general trend was a decline, with a modest decrease of −1.44% between 1999 and 2004. After 2004, the decline accelerated, reaching an APC of more than −10%, before smoothing out in subsequent years. The most significant change in the ASMR for males occurred in 2004, following a slight decrease before that. For females, the steepest decline occurred in the decade after 2004. Unlike the ASIR, the gap between male and female ASMR narrowed, particularly after 2007 (Figure 3D).

An age-period-cohort model analysis was conducted to understand the independent effects of age, period, and birth cohort on asthma incidence and mortality. Local drift and net drift were calculated to illustrate the linear trends after adjusting for period and cohort effects. Net drift reflected the overall trend across all age groups, while local drift represented the trends within specific age groups. Between 1990 and 2021, the net drift for asthma incidence in China was −2.34% per year, and for mortality, it was −5.63% per year. The local drift for incidence initially decreased, then increased in older age groups, but remained negative throughout (Figure 4A). The greatest decline occurred in the under-5 age group, followed by a gradual decrease until the 20–25 age group, where a slight increase was observed. For those over 25 years, local drift continued to decline, reaching its lowest point in the 60–65 age group, before rising again in the 90–95 age group. In contrast, the local drift for mortality showed a different pattern but remained negative (Figure 4B). The lowest local drift for mortality was observed in the under-5 age group, increasing to the 35–40 age group, and then declining in the 60–65 age group. Mortality drift increased again for those aged 80 and above. A similar trend was observed in both male and female local drifts for incidence and mortality, with more significant changes seen in females, particularly for incidence (Supplementary Figures S1, S2).

After adjusting for period and cohort effects, we found that the age effect had a greater impact on asthma incidence in younger age groups (Figure 4C). While mortality was more strongly influenced by age in older groups, except for those under 5 years (Figure 4D). Gender-specific analysis showed a similar trend for incidence. However, a sharp decline in age attribution to mortality was observed in males over 95 years, with values halving compared to the previous age group. In females, age attribution to mortality was nearly six times higher in the under-5 group compared to older age groups, resulting in a notable shift in the longitudinal age curve for both the under-5 and over-95 age groups.

After controlling for age and cohort effects, the period effects on incidence and mortality were assessed using the ratio rates, with the median level for the period 2002–2006 as the baseline. The period effect of incidence declined from 1.47 in 1992–1996 to 0.82 in 2017–2021 (Figure 4E), and mortality risk dropped from 1.60 in 1992–1996 to 0.40 in 2017–2021 (Figure 4F). While the trends in period effects on incidence and mortality were consistent across gender, the decline in these ratios was more pronounced in females than in males.

After controlling for age and period effects, cohort effects on asthma incidence and mortality were assessed using the median level of the 1957–1961 cohort as the baseline. Cohort risk for incidence was significantly elevated in earlier birth cohorts and declined in more recent cohorts (Figure 4G), a pattern also observed for mortality (Figure 4H). Gender-specific analysis revealed higher incidence and mortality rate ratios in early birth cohorts for females compared to males within the same cohorts.

The BAPC model was employed to project trends in asthma incidence and mortality over the following 25 years. According to the model, asthma incidence is expected to decline continuously, reaching approximately 330 cases per 100,000 people by 2046 (Figure 5A; Supplementary Table S5). New cases are projected to hit a low point in 2032 before gradually increasing, with the total number of cases rising to around 4.5 million by 2046 (Figure 5C; Supplementary Table S6).

Similarly, the BAPC model predicts a decline in ASMR to about 0.69 per 100,000 by 2046 (Figure 5B; Supplementary Table S5). However, the absolute number of deaths is expected to rise sharply after 2040, culminating in over 80,000 deaths by 2046 (Figure 5D; Supplementary Table S6).