Childhood pneumonia deaths fell sharply, but vaccine gaps keep the burden unequal

Lower respiratory infection remains one of the most consequential infectious causes of death and health loss in children younger than five years, and a new analysis published in Frontiers in Medicine quantifies exactly where the burden that survives three decades of progress now sits. Drawing on Global Burden of Disease data for 204 countries and territories from 1990 to 2021, together with an independent single-center cohort of pediatric pneumonia hospitalizations in Beijing, a team from China-Japan Friendship Hospital and the Chinese Academy of Medical Sciences reports that global under-five LRI deaths declined from 1,898,507 in 1990 to 503,409 in 2021, while disability-adjusted life-years, or DALYs, fell from 169,370,501 to 44,882,094 over the same period. The estimated annual percentage change was minus 4.3 percent for both deaths and DALYs and minus 3.3 percent for incidence, a sustained decline that nonetheless leaves hundreds of thousands of young children dying each year from an infection that vaccines, oxygen, and standardized care can largely address.

The technical architecture of the study is deliberately split into two parallel components that were never pooled or used to validate one another. The population component follows GBD modelling frameworks, which provide standardized, cross-nationally comparable estimates of incidence, deaths, DALYs, and pathogen-attributed burden, supplemented by WHO Global Health Estimates, UN-IGME child mortality estimates, WUENIC and WHO vaccine-introduction data, UNICEF indicators, and World Bank development indicators. The clinical component, by contrast, extracted 1,717 pediatric pneumonia hospitalization records from China-Japan Friendship Hospital between July 1, 2024 and December 31, 2025, capturing age, sex, recorded severity, respiratory support, length of stay, laboratory biomarkers, and pathogen testing. The authors followed GATHER principles for the population analyses and STROBE principles for the hospitalization-record component, and they prespecified children aged zero to four years as the main clinical subgroup because GBD focuses on children younger than five.

China’s recent national context comes from GBD 2023 estimates, which showed 2,895,598 incident LRI cases among children younger than five years, an incidence rate of 4,658.0 per 100,000, 4,859 deaths, and 436,398 DALYs. Within China’s pathogen-attributed burden, Streptococcus pneumoniae remained the largest source of deaths and DALYs among the pathogens considered, with an estimated 2,162 deaths and 193,740 DALYs, followed by the Haemophilus influenzae proxy at 145 deaths and 13,307 DALYs, respiratory syncytial virus at 121 deaths and 11,207 DALYs, and influenza virus at 81 deaths and 7,358 DALYs. The authors are explicit that these modelled national figures provide context for the hospitalization data but were not used to extrapolate incidence or mortality from the single-center cases, because the two components carry different denominators, sampling frames, and purposes of inference.

The most striking global finding concerns how the residual burden is distributed across pathogens and across the Socio-demographic Index, a composite measure of national development. In 2021, Streptococcus pneumoniae caused 166,379 under-five LRI deaths, representing 33.1 percent of all such deaths, and 14,794,643 DALYs, or 33.0 percent of the total. When the researchers compared age-standardized mortality rates between the lowest and highest SDI quintiles, pneumococcal disease alone accounted for 75.7 percent of the summed pathogen-specific excess mortality-rate difference, and together with H. influenzae-related burden the vaccine-relevant bacterial pathogens accounted for 81.0 percent of that pathogen-specific gap. Referenced to the overall LRI mortality-rate gap, the two bacterial categories together explained 36.0 percent, whereas RSV and influenza virus together accounted for only 8.5 percent, a distribution the authors interpret as evidence that SDI-related inequality is concentrated in bacterial pathogens closely tied to conjugate-vaccine policy.

Inequality metrics reveal a paradox at the heart of the decline. In 1990, the mean overall LRI age-standardized mortality rate was 538.4 per 100,000 in the lowest SDI quintile and 35.3 per 100,000 in the highest, a low-to-high ratio of 15.26. By 2021, the corresponding rates had fallen to 148.1 and 3.4 per 100,000, but the ratio had widened to 42.92. Pneumococcal mortality showed an even steeper gradient in 2021, at 49.4 per 100,000 in the lowest quintile versus 0.7 per 100,000 in the highest, a ratio of 70.56. The slope index of inequality and concentration index both remained negative throughout, indicating persistent concentration of burden in lower-SDI countries. In other words, a falling global average does not mean that gaps between countries have converged; the remaining mortality risk is more relatively concentrated than it was three decades ago.

Immunization data show that burden inequality and program inequality coexist in the same direction. In 2021, Hib3 coverage was 82.6 percent in the lowest SDI quintile and 94.4 percent in the highest, while PCV3 coverage was 79.5 percent and 91.2 percent, respectively. Vaccine-introduction timing differed even more: the median year of first Hib3 introduction was 14 years later in the lowest than in the highest SDI quintile, and PCV3 introduction was 6 years later. The researchers also ran country and year fixed-effects models with standard errors clustered by country, using 465 country-year observations from 62 countries for PCV3 and 1,111 country-year observations from 81 countries for Hib3, with a prespecified main lag of one year and Benjamini-Hochberg and Bonferroni corrections across 16 tests.

The fixed-effects results were directionally consistent but statistically fragile. Each 10-percentage-point increase in PCV3 coverage was associated with a 0.9 percent relative decrease in pneumococcal age-standardized mortality and DALY rates, with a 95 percent confidence interval of minus 1.9 to 0.2 that crossed zero. Each 10-percentage-point increase in Hib3 coverage was associated with a 1.8 percent relative decrease in the corresponding rates for the H. influenzae proxy, with a confidence interval of minus 3.4 to minus 0.2. However, across the prespecified 16 tests, neither main result remained significant after false-discovery-rate correction at 0.05, and no result passed Bonferroni correction. The authors interpret these coefficients as country-level ecological associations rather than causal vaccine-effect estimates, noting that fixed effects cannot fully remove time-varying confounding from nutrition, care-seeking behavior, oxygen availability, antibiotic supply, and health-system improvement, and that staggered vaccine introduction can bias results when pretrends are not parallel.

The independent clinical cohort describes what pneumonia requires once children reach hospital care, information that population estimates cannot provide. Among the 1,717 pediatric pneumonia hospitalization records, 889, or 51.8 percent, were among children aged zero to four years. In this subgroup, 170 records, or 19.1 percent, were classified as severe, 184, or 20.7 percent, involved at least one form of respiratory support, and 92, or 10.3 percent, involved advanced respiratory support defined as high-flow oxygen therapy, non-invasive ventilation, or invasive mechanical ventilation. The median length of stay was 6.0 days. Exploratory adjusted models found that higher log-transformed C-reactive protein was associated with a greater likelihood of being recorded as severe, with an odds ratio of 1.22 and a 95 percent confidence interval of 1.07 to 1.39, while text-derived Mycoplasma pneumoniae records were associated with an odds ratio of 2.07 and a confidence interval of 1.23 to 3.49.

Hospital pathogen testing showed a clinical picture shaped by tested populations and test panels rather than by population infection rates. Among tested hospitalization records for children younger than five years, laboratory positivity was 106 of 449, or 23.6 percent, for Mycoplasma pneumoniae; 100 of 446, or 22.4 percent, for RSV; 102 of 444, or 23.0 percent, for human rhinovirus; and 18 of 446, or 4.0 percent, for influenza virus. Streptococcus pneumoniae and H. influenzae were infrequently tested or recorded in the hospital cohort, so the authors state that hospital data are not suitable for evaluating GBD attribution estimates for these pathogens. RSV and influenza virus provide a natural connection between the two components because they appear in both GBD pathogen attribution and hospitalization testing, while M. pneumoniae and rhinovirus mainly reflect clinical testing features of recent pediatric pneumonia admissions in China, consistent with previous Chinese studies showing that pathogens vary by age, region, and season.

The study carries clear limitations that the authors state directly. GBD results are modelled population-level estimates whose pathogen attribution depends on input data quality and modelling assumptions, and the vaccine-coverage analyses reflect country-year ecological associations rather than individual vaccine effectiveness. The hospitalization data came from a single center and included only hospitalized children, so they cannot estimate national incidence, mortality, DALYs, or severe-pneumonia incidence, and hospital pathogen positivity was affected by testing indications, test panels, timing, and documentation completeness. Age-stratified analyses added a further nuance: infants aged one to five months accounted for a relatively large share of under-five LRI deaths in 2021, whereas coverage-burden associations were more apparent in children aged two to four years, suggesting that the group with the highest mortality risk is not necessarily the group in which a national annual coverage signal is easiest to detect. The authors conclude that childhood LRI control should address vaccine equity alongside early recognition, oxygen capacity, referral, and inpatient pneumonia management, treating prevention and clinical care as complementary parts of the same pathway.

Subject of Research: Global and Chinese burden of lower respiratory infection in children under 5 years, vaccine coverage inequality, and pediatric pneumonia hospitalization care needs

Article Title: Global and China lower respiratory infection burden in children under 5 years: implications for vaccine equity and pediatric clinical care

Article References: Su, X., Liu, D., Liu, J., Lv, D., Guo, J., Hui, Q., Ye, F., Chen, Y., Wang, C., Tang, L., Xia, M., & Zhang, Q. (2026). Global and China lower respiratory infection burden in children under 5 years: implications for vaccine equity and pediatric clinical care. Frontiers in Medicine, 13, Article 1971991. https://doi.org/10.3389/fmed.2026.1971991

Image Credits: AI Generated

DOI: 10.3389/fmed.2026.1971991

Keywords: lower respiratory infection, children under 5, Global Burden of Disease, Streptococcus pneumoniae, pneumococcal conjugate vaccine, Haemophilus influenzae type b, vaccine equity, pediatric pneumonia, respiratory support, Socio-demographic Index, China, DALYs