Substantial changes in PM2.5 pollution and corresponding premature deaths across China during 2015–2019: A model prospective

Kamal Jyoti Maji

Research output: Contribution to journalArticlepeer-review

56 Citations (Scopus)


Long-term exposure to the ambient fine particulate matter (PM2.5) is the major public health risk factor in China. Several past studies have assessed premature mortalities associated with PM2.5 in China at varying levels of temporal and spatial scales using different methodological approaches. However, recently developed global exposure mortality model [GEMM NCD + LRI and GEMM 5-COD] provides a much more sophisticated methodology in capturing mortality due to PM2.5-exposure than the commonly accepted integrated exposure-response (IER) model, which this study applied to China. This study provides a comparative assessment of the excess long-term PM2.5-attributed nonaccidental deaths as well as cause-specific deaths for 349 cities in mainland China during five years (from 2015 to 2019) and compares the results with the spatial resolution scale of 0.1° × 0.1° across overall China. The results demonstrate that the national annual average PM2.5 concentration declined from 51.9 ± 18.2 μg/m3 in 2015 to 39.0 ± 13.2 μg/m3 in 2019, and the overall annual negative trend was around −3.1 ± 2.2 μg/m3/year [−5.6 ± 3.4%/year] across China. Consequently, the number of PM2.5-related deaths decreased by 383 thousand [95% CI: 331–429] to 1755 thousand [95% Confidence Interval: 1470–2025; GEMM NCD + LRI]; 315 thousand [95% CI: 227–370] to 1380 thousand [95% CI: 948–1740; GEMM 5-COD] and 125 thousand [95% CI: 64–140] to 876 thousand [95% CI: 394–1262; IER] in 2019, derived from the pre-established models (GEMM and IER). The estimate PM2.5-attributed death with a spatial resolution of 0.1° × 0.1° was 2419 thousand [95% CI: 2041–2771; GEMM NCD + LRI], 1918 thousand [95% CI: 1333–2377; GEMM 5-COD] and 1162 thousand [95% CI: 534–1611; IER] in 2015, which is about 11–16% higher value than the city-level health risk assessment study. The estimated deaths by GEMM NCD + LRI and GEMM 5-COD were 104% and 61% higher than the estimated by IER, highlighting that total premature mortalities associated with PM2.5 were substantially left behind based on the pre-existing model. The “other noncommunicable diseases” mortality, which IER method doesn't account for, was 375 thousand in 2019, 68 thousand less than in 2015. Such significant mortality was previously overlooked in estimation methods, which should now be considered for the air pollution-related policy development in China. The high number of premature deaths in central and northern parts of China, calls for the need for the Government to quickly impose even more stringent and effective pollution control measures.
Original languageEnglish
Article number138838
Number of pages11
JournalScience of the Total Environment
Early online date23 Apr 2020
Publication statusPublished - 10 Aug 2020
Externally publishedYes


Dive into the research topics of 'Substantial changes in PM2.5 pollution and corresponding premature deaths across China during 2015–2019: A model prospective'. Together they form a unique fingerprint.

Cite this