---
title: "Air pollution and mortality in Acute Lymphoblastic Leukemia: Northern Thailand age-group analysis"
id: "plos-one-23-air-pollution-and-mortality-in-acute-lymphoblastic-leukemia-patients-across"
canonical_url: "https://medichelpline.com/clinical-feed/plos-one-23-air-pollution-and-mortality-in-acute-lymphoblastic-leukemia-patients-across"
content_type: "clinical_feed_article"
specialty: "Oncology"
source_name: "PLOS ONE (Medicine)"
source_url: "https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0355856"
published_at: "2026-08-13T14:00:00.000Z"
evidence_level: "Journal Feed"
license: "CC-BY-NC-4.0 / Informational Use"
---
# Air pollution and mortality in Acute Lymphoblastic Leukemia: Northern Thailand age-group analysis
## Provenance & Clinical Metadata
- **Canonical URL:** https://medichelpline.com/clinical-feed/plos-one-23-air-pollution-and-mortality-in-acute-lymphoblastic-leukemia-patients-across
- **Specialty:** [Oncology](https://medichelpline.com/clinical-feed/oncology.md)
- **Primary Source:** PLOS ONE (Medicine)
- **Source URL:** [Original Journal Publication](https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0355856)
- **Published At:** 2026-08-13T14:00:00.000Z
- **Evidence Rating:** Journal Feed
## Executive GIST (TL;DR)
- This retrospective cohort study evaluated 604 patients diagnosed with **acute lymphoblastic leukemia (ALL)** in eight provinces of upper northern Thailand from 1999 to 2020, followed through February 2024. - Patients were classified into three age groups: pediatric (<15 years; n=344, 57.0%), adolescent and young adult (AYA; 15–39 years; n=152, 25.1%), and adult (≥40 years; n=108, 17.9%). - Reported 5-year survival rates were 64.5% for pediatric, 19.6% for AYA, and 6.7% for adults, illustrating markedly worse survival with advancing age. - Annual average ambient pollutant concentrations (PM2.5, PM10, NO2, O3, CO) were estimated for 2003–2020 using the Copernicus Atmosphere Monitoring Service reanalysis and linked to patient district of residence at diagnosis. - Patients who died had significantly higher concentrations of **PM2.5**, **PM10**, **O3**, and **CO**, with the signal most pronounced among pediatric cases; AYA decedents also had higher PM2.5, PM10, and CO than survivors. - Adjusted hazard ratios for death were higher for AYA (aHR 3.76) and adult (aHR 8.10) groups relative to pediatric patients. Diagnosis before 2010 was associated with increased mortality (aHR 1.38). - A pediatric-specific association was reported for **PM2.5 ≥50 µg/m³** and increased mortality. - The study underscores a potential adverse effect of ambient **air pollution** on survival in ALL, particularly affecting children, in a region with seasonal high PM from biomass burning. - Data sources, exposure linkage method, and registry-derived clinical variables were described; the source text is truncated regarding assumptions about address stability, so details on residential mobility or exposure misclassification were not reported in the provided excerpt.
## Clinical Analysis & Structured Key Points
Air pollution and mortality in acute lymphoblastic leukemia patients across different age groups in Northern Thailand (1999–2020) | PLOS One Browse Subject Areas ? Click through the PLOS taxonomy to find articles in your field. For more information about PLOS Subject Areas, click here . Article Authors Metrics Comments Media Coverage Reader Comments Figures Figures Abstract This study investigated the impact of air pollution on 604 patients diagnosed with acute lymphoblastic leukemia (ALL) between 1999 and 2020 in upper northern Thailand. Patients were categorized into three age groups: pediatric (<15 years), adolescent and young adult (AYA) (15–39 years), and adult (≥40 years). The cohort included 344 (57.0%) pediatric, 152 (25.1%) AYA, and 108 (17.9%) adult patients, with 5-year survival rates of 64.5%, 19.6%, and 6.7%, respectively. Concentrations of PM 2.5 , PM 10 , O₃, and CO were significantly higher in patients who died, particularly among pediatric cases. Among AYA patients, those who died had higher levels of PM 2.5 , PM 10 , and CO than survivors. Compared to the pediatric group, the risk of death was higher in AYA (adjusted hazard ratio [aHR]: 3.76) and adult (aHR: 8.10) groups. Patients diagnosed before 2010 had an increased risk of death (aHR: 1.38). Notably, PM 2.5 levels ≥50 µg/m³ were associated with increased mortality in pediatric patients. These findings highlight the adverse effects of air pollution on survival outcomes in ALL patients, particularly children under 15 years of age. Citation: Traisathit P, Sathitsamitphong L, Rattanathammethee T, Srikummoon P, Kawilapat S, Thongsak N, et al. (2026) Air pollution and mortality in acute lymphoblastic leukemia patients across different age groups in Northern Thailand (1999–2020). PLoS One 21(8): e0355856. https://doi.org/10.1371/journal.pone.0355856 Editor: Phuping Sucharitakul, Tsinghua University, CHINA Received: February 27, 2026; Accepted: July 27, 2026; Published: August 13, 2026 Copyright: © 2026 Traisathit et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability: Data cannot be shared publicly because they contain potentially identifying information. Data are available from the Chiang Mai Cancer Registry Institutional Data Access/Ethics Committee (contact via +66 5393 6641 ext. 22 or researchmed@cmu.ac.th ) for researchers who meet the criteria for accessing confidential data. Funding: This research project was supported by Fundamental Fund 2023, Chiang Mai University, Thailand. The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist. Introduction Acute lymphoblastic leukemia (ALL), the most common cancer in children [ 1 ], is approximately four times higher in children than in adults [ 2 ], with the highest incidence occurring from 2 to 10 years of age [ 3 ]. Data from the Surveillance, Epidemiology, and End Results (SEER) program have shown significant variation in 5-year survival rates for ALL patients by age. Pediatric patients (<15 years) have the highest 5-year survival rate (85%), followed by those aged 15–19 years old (61%), 30–39 years old (40%), and 40 years and older (< 30%) [ 4 ]. Multicenter studies in Thailand have demonstrated similar trends of overall survival (OS) and event-free survival (EFS) among pediatric patients (< 15 years old), adolescents and young adults (AYA; 15–39 years old), and adults (≥ 40 years old). The 5-year OS and EFS have been reported as 75% and 81.7%, respectively, for ALL pediatric patients [ 5 ], while the 2-year OS and EFS are lower for AYA (54% and 45.5%, respectively) and the adult populations (44.1% and 26.2%, respectively) [ 6 ]. However, advancements in treatment modalities over time have led to a significantly improved ALL survival rate for all age groups since 2010, increasing from 51% before 1990 to 72% afterward [ 4 ]. Nevertheless, the prognosis remains poor, particularly for adults (≥ 40 years old) [ 1 ]. Air pollution is one of the most significant challenges of our time as it detrimentally affects health, thereby contributing to higher rates of illness and death [ 7 – 9 ]. In northern Thailand, air pollution, particularly from airborne particulate matter (PM) due to biomass burning, is especially severe from January to April each year. In particular, this issue is exacerbated by smoke from burning activities in neighboring countries [ 10 , 11 ]. The World Health Organization (WHO) estimates that approximately 7 million deaths, predominantly from noncommunicable diseases, are linked to the combined impact of ambient and household air pollution [ 12 ]. Regarding the WHO’s 2021 guidelines, the 24-hour average concentration of PM 2.5 should not exceed 15 µg/m 3 [ 12 ]. However, data from northern Thailand in 2020 revealed that PM 2.5 concentrations ranged from 16 to 195 µg/m 3 [ 11 ], greatly exceeding the levels recommended by the WHO. Therefore, air pollution poses a serious health concern, and effective pollution management is needed to protect public health. Over the past few years, numerous studies have highlighted the negative health impacts associated with both short-term and long-term exposure to air pollutants [ 8 ]. In particular, exposure to various air pollutants, including PM, nitrogen dioxide (NO 2 ), ozone (O 3 ), and carbon monoxide (CO), has been linked with increased mortality and constitutes a significant health risk [ 7 , 8 , 13 ]. Epidemiological evidence on the impact of outdoor air pollution on cancers other than lung cancer is relatively sparse, highlighting the need to understand its effects on cancer incidence and survival across other cancer types [ 14 ]. An increase in PM 2.5 concentration by 5 µg/m 3 has been associated with mortality in young children with lymphoid leukemia [5-year hazard ratio (HR 5-year ): 1.32; 95% confidence interval (CI): 1.02–1.71]. Nevertheless, no significant association between an increase in PM 2.5 concentration and leukemia mortality in AYA has been demonstrated [ 15 ]. However, evidence on the impact of ambient air pollution on the survival of patients with ALL, particularly across age groups in regions with extreme exposure, remains limited. The aim of the present study is to address this knowledge gap by investigating whether PM 2.5 , PM 10 , NO 2 , O 3 , and CO exposure is associated with the mortality rates of pediatric, AYA, and adult patients diagnosed with ALL in eight provinces in upper northern Thailand. Materials and methods Study design and population This was a retrospective study involving a cohort of 604 pediatric patients (< 15 years old), AYA (15-39 years old), and adults (≥ 40 years old) in eight provinces in upper northern region of Thailand: Chiang Mai, Chiang Rai, Nan, Phrae, Phayao, Lampang, Lamphun, and Mae Hong Son diagnosed with ALL between January 1, 1999 and December 31, 2020. The patients were tracked from their registration date until the end of February 2024 to assess their survival rates. We access this clinical data on May 1, 2024. Data collection and measurements The cancer characteristics and stage at diagnosis, all demographic data available in the Chiang Mai Cancer Registry database (such as sex, age, body mass index (BMI), smoking history, hill tribe ethnicity, and a family history of cancer) were included in the analysis to identify any potential risk factors. For the pediatric patients, additional demographic data (such as maternal smoking history, age at delivery and parity, and the family’s educational background) were also included. Cause of death was obtained from the Chiang Mai Cancer Registry based on linked hospital records and national death certificate data. Data on the hourly concentrations of PM 2.5 , PM 10 , NO 2 , O 3 , and CO were obtained from the Copernicus Atmosphere Monitoring Service at the European Centre for Medium-Range Weather Forecasts (ECMWF) [ 16 , 17 ]. This dataset comprising the most recent global reanalysis of atmospheric composition includes consistent 3D fields of various aerosols and chemical species with a horizontal resolution of approximately 80 km (often interpolated to a 0.75° * 0.75° grid). Using this data, we calculated the annual average concentrations of PM 2.5 (µg/m³), PM 10 (µg/m³), NO 2 (ppb), O 3 (ppb), and CO (ppb) for 2003–2020 for each of the districts in the eight provinces in upper northern Thailand. These averages were then linked to the corresponding districts for the addresses of 604 patients provided in the Chiang Mai Cancer Registry dataset at the year of diagnosis. We assumed that the patients’ recorded addresses did not change during the study period. The pollution data were updated annually until the patient’s death, loss to follow-up, or data censoring. Since residential mobility during follow-up could not be assessed, exposure misclassification is possible if patients relocated after diagnosis, thereby potentially attenuating the observed associations. ALL cases had been histologically confirmed (ICD 10: C91.0) and recorded in the Chiang Mai Cancer Registry. Patients were excluded if they had secondary or therapy-related leukemia, missing or unverifiable residential district information at the time of diagnosis (precluding exposure assignment), resided outside of the eight provinces included in the study, had an unknown vital status, or had incomplete follow-up data. Patients diagnosed before 2003 were retained. For the included cases, air pollution exposure was assigned beginning in 2003 (when CAMS data became available). We have also clarified that missing covariate data were handled via a complete case analysis of each model. Ethical considerations The study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of the Faculty of Medicine, Chiang Mai University (No. 200/2021). Patient consent was waived due to anonymous data recorded in the present study. Statistical analysis The baseline characteristics of the study participants are described using medians and interquartile ranges (IQRs) for continuous variables and frequencies and percentages for categorical variables. The follow-up period was measured from the date of diagnosis to either the date of death from any cause or the last follow-up date, whichever occurred first. The annually averaged concentrations of PM 2.5 , PM 10 , NO 2 , O 3 , and CO were segmented into two categories using quartiles, to which dichotomization was applied when deemed suitable. The Shapiro–Wilk test was utilized to assess the normality of continuous variables including demographics and pollutant concentrations. For data exhibiting normal distributions, parametric tests were employed, such as the independent t-test or one-way analysis of variance for comparing mean concentrations between groups. Conversely, non-normally distributed data were analyzed using non-parametric tests, including the Mann-Whitney U test or Kruskal-Wallis test for comparing median concentrations between groups. The overall mortality rate and the rates associated with each variable were determined by dividing the number of deaths by the total person-years of follow-up. The date used for survival analysis was the date of diagnosis, as recorded in the cancer registry database. Kaplan-Meier curves were used to estimate survival rates, and the significance of the differences in survival probabilities between groups for each variable was assessed using log-rank tests. Cox proportional hazard models were used to investigate the associations between the risk of death in ALL patients and various risk factors, including gender, age, hill tribe ethnicity, year of diagnosis, a family history of cancer, smoking history, and time-updated concentrations of PM 2.5 , PM 10 , NO 2 , O 3 , and CO. Variables with p < 0.25 in the univariable analysis were included in the multivariable model [ 18 ]. All statistical analyses were performed using Stata (version 18). Results Characteristics of the ALL patients Of the 604 patients with ALL registered between 1999 and 2020, 344 (57.0%) were pediatric, 152 (25.1%) were AYA, and 108 (17.9%) were adults. The median age at diagnosis was 11.7 years old (IQR: 4.6–30.7), with 55.8% of the patients being male. The median ages of the pediatric, AYA, and adult patient groups were 5.1 years old (IQR: 3.0–8.5), 22.3 years old (IQR: 17.6–31.3), and 55.6 years old (IQR: 47.7–63.9), respectively ( Table 1 ). Download: PNG larger image TIFF original image Table 1. Characteristics of the study population by age group. https://doi.org/10.1371/journal.pone.0355856.t001 Survival rates The highest proportion of deaths occurred among the adult group (89.8%), followed by the AYA group (71.1%) and pediatric group (36.6%). For the pediatric group, the percentage of deaths decreased significantly from 44.4% in those diagnosed before 2010 to 26.0% in those diagnosed since 2010, representing a reduction of 18.4% (95% CI: 8.5–28.3; p = 0.001). A similar trend was seen in the AYA group (from 78.6% in those diagnosed before 2010 to 61.2% in those diagnosed since 2010), signifying a reduction of 11.6% (95% CI: 2.2–31.3; p = 0.02). However, the decrease was only slight in the adult group (from 90.9% in those diagnosed before 2010 to 82.7% in those diagnosed since 2010); this reduction of 4.2% (95% CI: −2.0 to 10.3; p = 0.07) was not statistically significant ( Fig 1 ). Download: PNG larger image TIFF original image Fig 1. Bar charts showing percentages of deceased acute lymphoblastic leukemia patients by age group and time period of diagnosis. https://doi.org/10.1371/journal.pone.0355856.g001 The survival rate for the adult group was poorer than those for the pediatric and AYA groups ( p < 0.001) ( Fig 2(a) ), which highlights the impact of age on survival time. Moreover, the survival probability of the ALL patients improved when diagnosed since 2010 compared to those diagnosed beforehand ( p = 0.003) ( Fig 2(b) ). Fig 2(c) illustrates that pediatric patients diagnosed since 2010 had a significantly better survival rate compared to those diagnosed beforehand ( p = 0.002). Similarly, AYA patients diagnosed since 2010 showed an improved survival probability compared to those diagnosed beforehand, albeit that the statistical significance is borderline ( p = 0.05) ( Fig 2(d) ). For adults diagnosed since 2010, the survival rate was slightly better than those diagnosed beforehand, which was also statistically significant ( p = 0.04) ( Fig 2(e) ). Download: PNG larger image TIFF original image Fig 2. The survival rates of participants diagnosed with acute lymphoblastic leukemia categorized by (a) age group, (pediatric (age < 15 years old), AYA (adolescent and young adult; age 15–39 years old), and adult (age ≥ 40 years old)) and stratified by year of diagnosis (1999–2009 or 2010–2020): (c) the pediatric group, (d) the AYA group, and (e) the adult group. https://doi.org/10.1371/journal.pone.0355856.g002 The results in Table 2 indicate that the overall median follow-up time for the cohort was 1.9 years (IQR: 0.5–6.4) and, specifically, 4.3 years (IQR: 1.2–9.5), 1.0 years (IQR: 0.4–2.4), and 0.4 years (IQR: 0.1–1.1) for the pediatric, AYA, and adult groups, respectively. The overall 5-year survival rate was 43.4%, with a range from 39.2% to 47.6%, reflecting a median survival time of 3.1 years. The 5-year survival rates were 64.5% for the pediatric group, 19.6% for the AYA group, and 6.7% for the adult group. Survival rates stratified by the time of diagnosis indicate that the 5-year survival rate increased from 38.8% for those diagnosed before 2010 to 49.4% for those diagnosed since 2010, and the median survival time was extended by 4.8 years. The 5-year survival rate for the adult group was the least favorable at 6.7% (95% CI: 2.8%−12.9%) with a median survival time of 0.5 years. When stratified by the time of diagnosis, adults diagnosed before 2010 had a survival rate of 4.4% (95% CI: 0.8%−13.1%), improving to 9.0% (95% CI: 3.0%−19.1%) for those diagnosed since 2010. Download: PNG larger image TIFF original image Table 2. The 5-year overall survival rate by age group and year of diagnosis. https://doi.org/10.1371/journal.pone.0355856.t002 Pollutant concentrations The median annual concentrations of the pollutants, as depicted in S1 Fig , suggest a temporal decrease in levels for PM 2.5 , PM 10 , and CO. Given the non-normal distribution of pollutant concentrations ( S1 Table ), median comparisons between patients who survived and died, as well as across years of diagnosis were presented in S2 Table . The median concentrations of PM 2.5 , PM 10 , NO 2 , O 3 , and CO were 35.8 µg/m³, 49.8 µg/m³, 7.2 ppb, 36.4 ppb, and 378.9 ppb, respectively. A detailed comparison of median concentrations for each pollutant across age groups and years of diagnosis within those age group is presented in S2 Fig . This comparison reveals no significant differences among the concentrations of most of the pollutants (PM 2.5 , PM 10 , O 3 , and CO) across the age groups, suggesting consistent exposure levels irrespective of age. However, an exception was noted for NO 2 where the adult group exhibited the highest median concentration at 6.8 ppb, followed by AYA at 7.9 ppb and pediatric at 8.2 ppb (p < 0.001) ( S2 Fig . (e)). According to S2 Table , concentrations of PM 2.5 ( p < 0.001), PM 10 ( p < 0.001), O 3 ( p = 0.007), and CO ( p < 0.001) were significantly higher in the patients who had died. Among pediatrics, PM 2.5 ( p < 0.001), PM 10 ( p < 0.001), O 3 ( p < 0.001), and CO ( p < 0.001) concentrations were also significantly higher in those who had died. For the AYA group, PM 2.5 ( p = 0.001), PM 10 ( p = 0.001), and CO ( p < 0.001) concentrations were significantly higher in the patients who had died compared to those who survived. Conversely, in the adult group, there were no significant differences in the median of PM 2.5 , PM 10 , NO 2 , and O 3 concentrations between those who had died and those who survived, except for CO ( p = 0.036). Further analysis, presented in S2 Table and S2 Fig , compares the median annual pollutant concentrations in the residential districts of the patients stratified by years of diagnosis (1999–2009 vs. 2010–2020) for each age group. It was found that the concentrations of all pollutants were significantly lower in the later period across all age groups, with the exception of O 3 concentration in the adult group ( p = 0.309). Death risk factors The results of the univariable and multivariable analyses for determining the risk factors for death in ALL patients are provided in Table 3 . The results of the univariable analysis indicate that age, specifically the AYA and adult groups, posed the highest risk factors for death ( p < 0.001). The results of the multivariable analysis further demonstrate that being in the AYA group (aHR: 3.76, 95% CI: 2.77–5.10) or the adult group (aHR: 8.10, 95% CI: 5.85–11.23), or being diagnosed with ALL before 2010 (aHR: 1.38, 95% CI: 1.07–1.78), were independently associated with an increased risk of death. Subgroup analysis by age group reveals that the pediatric patients exposed to higher concentrations of PM 2.5 (≥ 50 µg/m 3 ) had an increased risk of death (HR: 2.02; 95% CI: 1.10–3.70) in the univariable analysis ( S3 Table ). However, this association was not statistically significant in the multivariable analysis (aHR: 1.92; 95% CI: 0.96–3.83). Despite this, ALL diagnosis before 2010 still posed a higher risk of death in the pediatric group (aHR: 1.72; 95% CI: 1.06–2.80) ( S3 Table ). For the AYA and adult groups, no significant links were found
## Related Clinical Research

- [Burden and 15-Year Prediction of Early-Onset Cancer Attributable to Overweight and Obesity in Chin](https://medichelpline.com/clinical-feed/pubmed-42660835.md) (DOI: 10.3881/j.issn.1000-503X.17015)
- [Fingolimod triggers apoptosis-like death and cytotoxic stress in AML including P-glycoprotein–over](https://medichelpline.com/clinical-feed/pubmed-42269275.md) (DOI: 10.1016/j.bbrc.2026.154118)
- [MYC–MIZ1 Enhancer Complexes Regulate Neuroendocrine Identity in Small Cell Lung Cancer](https://medichelpline.com/clinical-feed/biorxiv-7-myc-miz1-complexes-at-enhancers-tune-neuroendocrine-identity-of-small-cell-lung.md)
- [PDLIM2 in Lung Adenocarcinoma: Prognostic Biomarker and Therapeutic Candidate](https://medichelpline.com/clinical-feed/british-journal-of-cancer-0-exploring-pdlim2-as-a-prognostic-biomarker-and-therapeutic-target-in-lung.md)
- [Translational modelling questions receptor‑occupancy‑based dosing for pembrolizumab (PD‑1 inhibito](https://medichelpline.com/clinical-feed/british-journal-of-cancer-0-translational-modelling-challenges-receptor-occupancy-based-dosing-of-pd-1.md)

## Navigation
- [← Back to Oncology Feed](https://medichelpline.com/clinical-feed/oncology.md)
- [← All Clinical Specialties](https://medichelpline.com/clinical-feed.md)
## Medical & Regulatory Disclaimer

> [!CAUTION]
> MedicHelpline content is structured for research, educational, and professional discovery purposes. It does not constitute individual medical advice, clinical diagnosis, or treatment recommendations.
> Always verify dosing, contraindications, and regulatory alerts against official product labeling and primary regulatory sources before clinical decision-making.