Higher levels of fine-particle air pollution were associated with an increased risk of invasive pneumococcal disease in South Africa, according to a large study that found the strongest link among adults aged 65 and older.
The peer-reviewed research, published in Nature Microbiology on September 14, 2026, examined reported disease cases, bacterial genomic data and air-pollution exposure over nearly two decades. Its findings indicate that the size and timing of pollution-associated risk varied according to patients’ ages, disease presentation and the genetic characteristics of the pneumococcal bacteria circulating in the population.
The study, titled “Pneumococcal population structure influences the effects of air pollution on invasive disease risk in South Africa,” included 59,017 identified cases reported by 531 hospitals across 52 districts between 2005 and 2023. The full dataset comprised 32,639 cases of bacteraemia, 21,167 cases of meningitis and 5,211 other invasive pneumococcal disease cases.
For their principal models, the researchers restricted the analysis to 52,437 cases recorded from 2005 through 2019. That period overlapped with the available genomic data and avoided the substantial decline in reported cases during the COVID-19 pandemic.
Risk accumulated over several weeks
At a weekly average concentration of 50 micrograms of PM2.5 per cubic metre of air, the model estimated a 3.4% increase in cumulative invasive pneumococcal disease risk over eight weeks. The corresponding cumulative relative risk was 1.03, with a 95% confidence interval of 1.02 to 1.05.
PM2.5 refers to fine airborne particles measured by their diameter. The study assessed how weekly exposure levels were associated with disease risk over subsequent weeks rather than treating exposure and illness as events occurring only at the same time.
The association was notably stronger in the oldest age group. Among adults aged 65 and older, a weekly average PM2.5 concentration of 50 micrograms per cubic metre was associated with an eight-week cumulative relative risk of 1.16. The 95% confidence interval ranged from 1.10 to 1.22.
The results did not suggest a uniform response across all pneumococcal infections. Associations differed by age, clinical presentation, bacterial serotype and genomic lineage. Serotypes 4, 8 and 23F showed an immediate rise in risk following high PM2.5 exposure, while other bacterial groups had different lag patterns.
First author Dr Sophie Belman told IOL/The Star that the bacterial strain could affect both when disease occurs and which people may face greater risk. “The strain of bacteria impacts the timing of disease and who might be more at risk depending on their respiratory microbiome,” she said.
Genomic data added to disease surveillance
The analysis incorporated 4,350 genome-sequenced pneumococcal isolates collected through 2019. Researchers used Bayesian spatiotemporal models to evaluate patterns across locations and time while accounting for factors including seasonality, population density and periods of pneumococcal-vaccine implementation.
By combining genomic information with surveillance records, the researchers were able to assess whether pollution associations changed with the composition of the pneumococcal population. That distinction matters because broad case totals can obscure differences among serotypes and lineages that may respond to environmental conditions on different timelines.
The study establishes population-level associations, however, and does not show that PM2.5 exposure caused any individual infection. The authors also identified limited ground-based air-quality monitoring in parts of South Africa as a constraint. Sparse monitoring can make it more difficult to estimate local exposure accurately, particularly across a geographically broad national analysis.
The findings nevertheless point to a possible role for environmental data in infectious-disease preparedness. The researchers concluded that combining information about air quality with data on the pneumococcal strains circulating in communities could improve models used to consider vaccination strategies and anticipate hospital-capacity needs.
Such planning would need to account for the study’s central finding that pollution-associated risk was not evenly distributed. Older adults experienced the highest estimated risk, and bacterial groups differed in both the strength and timing of their associations with elevated PM2.5.
