Skip to content
Heavy Metal Index

Arsenic exposure and respiratory outcomes

Evidence synthesis on arsenic exposure and respiratory outcomes, including lung function impairment, airway obstruction, lung cancer, and pulmonary fibrosis.

Arsenic is an established respiratory carcinogen (IARC Group 1) and is associated with non-malignant respiratory outcomes including obstructive lung disease, reduced lung function, and pulmonary fibrosis. The respiratory evidence base covers three distinct exposure pathways: (1) chronic drinking water arsenic ingestion producing systemic arsenic body burden; (2) inhalation of arsenic-contaminated particulates in occupational settings, ambient air pollution, and disaster contexts; and (3) early-life or prenatal exposure, where developmental effects on the lung persist into adulthood. Dietary arsenic contributes to systemic arsenic burden in low-water-arsenic populations, and the Spanish birth cohort evidence reviewed below suggests that even low dietary inorganic arsenic may be detectable in childhood respiratory symptom counts [1]. The majority of the epidemiological evidence for lung function impairment comes from populations with high drinking water arsenic.

Lung function impairment in human cohort and cross-sectional studies

The most direct human epidemiological evidence for arsenic-associated lung function impairment comes from studies in Bangladesh and from post-disaster cohorts.

A cross-sectional study in Dhaka, Bangladesh enrolled 269 adults with longstanding drinking water arsenic exposure (median toenail arsenic 4.7 µg/g, interquartile range 2.2-10.4 µg/g) [2]. Spirometry was completed by 166 participants. Airflow obstruction (forced expiratory volume in 1 second to forced vital capacity ratio below 0.70) was present in 19.9% (33 of 166 participants). In unadjusted logistic regression, each interquartile range increment in toenail arsenic was associated with an odds ratio of 1.97 (95% CI 1.06-3.67, p = 0.03) for airflow obstruction. The study also identified impaired cystic fibrosis transmembrane conductance regulator (CFTR) function as a candidate mechanism: each 1 µg/g increase in toenail arsenic was associated with 0.59 mmol/L higher sweat chloride concentration (p less than 0.001), consistent with arsenic-induced CFTR degradation promoting mucus plugging and obstructive physiology. Genetic screening confirmed that the elevated sweat chloride was not attributable to pathogenic CFTR variants [2].

A prospective cohort study of 1,400 adults in Maui, Hawaii enrolled 6 to 18 months after the August 2023 Lahaina wildfires found that internal urinary arsenic concentrations were approximately twice the US population reference (median 18.10 µg/L vs. 8.09 µg/L in NHANES, p less than 0.001) [3]. In weighted quantile sum mixture analyses, each one-quantile increment in the metal mixture index was associated with odds ratios of 1.84 (95% CI 1.37-2.45) for below-lower-limit-of-normal forced vital capacity, 1.53 (95% CI 1.07-2.17) for FEV1, 1.83 (95% CI 1.37-2.45) for the FEV1/FVC ratio, and 1.84 (95% CI 1.33-2.56) for the FEF25-75 (small airways measure). Arsenic was identified as the dominant positive contributor to the mixture, with a single-metal arsenic association of OR 1.26 (95% CI 1.06-1.51) for below-LLN FVC [3]. Because wildfire smoke exposure mobilizes legacy contaminants from built infrastructure rather than dietary intake, this study characterizes an inhalation and environmental contamination pathway.

A narrative review synthesizing COPD epidemiology across multiple arsenic-exposed populations reported that arsenic exposure was associated with approximately 30-36% increased risk of COPD, with one cohort study reporting 36% and 30% increased risk at 5-year and 10-year follow-up respectively, and population data from Pabna, Bangladesh showing negative correlation of arsenic exposure with FEV1/FVC ratio [4]. The same review cited negative correlations of urinary cadmium with FEV1/FVC (r = -0.199, p = 0.013), FEV1 (r = -0.221, p = 0.007), and FVC (r = -0.201, p = 0.011) [4].

The following table summarizes the principal human epidemiological studies on arsenic and lung function:

Study / locationDesignExposure metric and matrixPrincipal outcomeEffect estimatenLimitation
Lee et al. 2025 [2]; BangladeshCross-sectionalToenail arsenic (ICP-MS); median 4.7 µg/gAirflow obstruction (FEV1/FVC below 0.70)OR 1.97 (95% CI 1.06-3.67) per IQR166 (spirometry)High-exposure context; cross-sectional
Maunakea et al. 2026 [3]; Maui, USAProspective post-disaster cohortUrinary metals mixture (WQS); arsenic median 18.10 µg/LFVC below LLNMixture OR 1.84; arsenic-only OR 1.26 (1.06-1.51)1,400No pre-fire baseline; mixed pathways
Zhang et al. 2026 [4]; reviewNarrative reviewDrinking water or urinary arsenic; multiple studiesCOPD risk30-36% increased risk (various studies cited)n/aNarrative; primary studies variable design
Signes-Pastor et al. 2022 [1]; SpainProspective birth cohortUrinary sum arsenic; median 4.92 µg/L at age 4Cumulative respiratory symptoms (GEE-Poisson)Expected count 0.63 at average vs. 1.33 at 99th percentile400Single exposure timepoint; exploratory

Early-life and prenatal arsenic effects on lung development

Arsenic exposure during fetal development and early childhood appears to program lasting pulmonary deficits that persist into adulthood, independent of subsequent exposure.

A prospective cross-sectional study in Antofagasta, Chile recruited 150 adults (75 males, 75 females) born during 1958-1970 when municipal drinking water arsenic reached approximately 870 µg/L, and compared their pulmonary function and exhaled breath condensate metabolomics to adults born in the same city during a low-exposure era [5]. Participants were assessed at ages 47-58 years, approximately 47 years after the high-exposure period ended. Forced vital capacity and FEV1 were significantly reduced in the high-exposure group in both sexes. The high-exposure group also showed significantly elevated 8-iso-prostaglandin F2alpha (p = 0.019), a marker of oxidative lipid stress, in exhaled breath condensate. Untargeted metabolomics identified 201 peaks uniquely associated with high early-life arsenic exposure, with a partial least squares discriminant analysis area under the curve of 0.97 for distinguishing exposure groups [5]. Current urinary arsenic levels in both groups were similar, confirming that the metabolic and pulmonary differences reflect early-life rather than current exposure.

Review evidence from ambient air metal pollution in vulnerable populations has described arsenic-induced impairment of mouse lung branching morphogenesis in prenatal models, with downregulation of fibroblast growth factor signaling cascades [6]. The same review documented ambient air arsenic concentrations of 0.800-15.7 ng/m3 at various monitoring stations and noted that the US Occupational Safety and Health Administration permissible exposure limit for inorganic arsenic is 0.01 mg/m3 (10 µg/m3) with the NIOSH recommended exposure limit set at 0.002 mg/m3 (2 µg/m3) [6].

A Spanish birth cohort study (INMA project; n = 400 children) using urinary arsenic speciation at age 4 found increasing expected counts of cumulative respiratory symptoms (wheeze, asthma, eczema, sneezing) with increasing urinary summed arsenic (inorganic arsenic plus monomethylarsonic acid plus dimethylarsinic acid), with adjusted expected counts of 0.63 at the average exposure (ln-summed arsenic 1.57) and 1.33 at the 99th percentile (ln-summed arsenic 4.00) [1]. In this low-water-arsenic Spanish population, food (primarily rice and rice-based products) was the dominant arsenic exposure pathway.

Arsenic carcinogenesis in the lung

Arsenic is classified by IARC as a Group 1 human carcinogen for lung cancer via both inhalation and ingestion routes [7, 8, 9, 10]. The carcinogenic mechanisms are established across multiple lines of evidence.

Epidemiological review: a comprehensive environmental pollution and lung cancer review documented a case-control study finding an odds ratio of 4.41 (95% CI 2.01-9.67) for lung cancer comparing the top versus bottom quartile of occupational cadmium exposure, as well as mechanistic evidence for arsenic-induced inhibition of autophagy (ACSS2/Atg5 pathway) driving lung adenocarcinoma invasion and metastasis [7]. IARC’s classification rests on inhalation and ingestion routes; this review also noted 25% significant increase in lung cancer incidence per standard increment of cadmium exposure in a prospective study [7].

Epigenetic mechanisms: a review of metal carcinogens and cancer epigenetics identified arsenic as producing dose-dependent DNA hypomethylation, with persistence after the exposure period ends, and inhibition of tumor suppressor genes MAP2K3 and DKK1 [8]. Arsenic was listed among the five most dangerous occupational lung carcinogens [8].

Toxicogenomic analysis: a Comparative Toxicogenomics Database in silico analysis of six metals in lung cancer found that arsenic interacts with 10,260 genes in the CTD, of which 199 are associated with lung cancer, 113 with non-small-cell lung cancer, and 27 with small-cell lung cancer [10]. TP53 was the only gene common to all six metals across all three lung cancer categories, and PI3K-AKT-mTOR pathway enrichment was identified for both NSCLC and SCLC [10]. These findings characterize shared mechanistic space among arsenic, cadmium, lead, mercury, nickel, and chromium(VI) in lung carcinogenesis.

Mechanistic evidence: cellular and molecular pathways

The mechanistic literature establishes three converging axes by which arsenic injures respiratory epithelium: oxidative stress, pro-fibrotic signaling, and direct genotoxicity.

Oxidative stress: an in vitro study using the BEAS-2B bronchial epithelial cell line showed that arsenic trioxide at 1.0-2.0 µM reduced cell viability and elevated reactive oxygen species (ROS); treatment with the ROS scavenger N-acetyl cysteine reversed both outcomes, confirming the ROS-IRE1alpha-apoptosis and ROS-IRE1alpha-autophagy axes [11].

Pro-fibrotic signaling: a mechanistic review identified the TGF-beta/Smad pathway as the central fibrosis axis for arsenic and co-occurring metals, with persistent TLR4/NF-kappaB/NLRP3 inflammatory activation, autophagy/lysosomal dysfunction, and fibroblast-to-myofibroblast transition contributing to alveolar epithelial injury and extracellular matrix deposition [4].

Mine dust inhalation model: an air-liquid interface bronchial epithelial cell study exposed cells to aerosolized arsenic-containing mine dust from tailings samples, documenting epithelial barrier disruption (reduced transepithelial electrical resistance), elevated inflammatory cytokines, and reduced cell viability [12]. This model is more physiologically relevant than submerged culture for inhalation exposure conditions [12].

Pleural and airway pathology: a review of airborne metals in pleural effusion reported that arsenic causes a spectrum from nasal and airway irritation through impaired mucociliary clearance to lung cancer via IARC-classified mechanisms [9]. The same review measured significantly elevated zinc in malignant pleural effusion versus benign effusion, with elevated superoxide dismutase (SOD) activity [9].

Hexavalent chromium and inhalation carcinogenesis: contextual evidence

Hexavalent chromium is included in this synthesis because it represents the best-mechanistically-characterized inhalation lung carcinogen in the same IARC Group 1 category, and several reviewed papers analyzed arsenic and chromium(VI) in direct mechanistic comparison [10]. A guinea pig model study confirmed that zinc chromate induces DNA double-strand breaks (measured by gamma-H2AX foci) specifically in bronchiolar epithelial cells at the lowest tested dose (0.2 mg/kg) after acute exposure, and inhibits homologous recombination DNA repair (measured by RAD51 foci) after 90-day subchronic exposure [13]. Males had overall higher DNA double-strand break levels than females. These findings are specific to chromium(VI) and not directly transferable to arsenic, but they establish the bronchiole as the site of IARC-classified inhalation carcinogenesis across metal classes, consistent with the arsenic lung adenocarcinoma mechanistic literature [13].

Inhalation and environmental contamination contexts

Three reviewed papers document arsenic in environmental contamination settings with respiratory relevance, without directly measuring dietary arsenic.

Coke oven emissions: a systematic review and meta-analysis of 38 studies (1972-2024; heterogeneity I-squared = 98.9%) reported that coke oven worker populations showed 37% elevated incidence of lung cancer and 18% elevated incidence of COPD or chronic bronchitis, with a significant association with malignant respiratory mortality [14]. Arsenic is a constituent of coke oven emissions alongside polycyclic aromatic hydrocarbons; these findings are specific to occupational inhalation.

Mining regions: a cross-sectional study in Chumbivilcas province, Cusco, Peru (n = 370 adults; mean urinary arsenic 39.30 µg/g creatinine, 68.2% above the 20 µg/g creatinine threshold for high exposure) found that occupational exposure at work was associated with higher urinary arsenic prevalence (adjusted prevalence ratio 1.29, 95% CI 1.07-1.56), and spring water consumption was associated with higher prevalence (aPR 1.19, 95% CI 1.01-1.40) [15]. This study characterizes high endemic-area arsenic exposure in a mining region, without measuring respiratory outcomes directly [15].

Disaster contamination: a retrospective ecological study of 203 municipalities in the Doce River Basin following the 2015 Fundao tailings dam collapse (60 million tons of mining waste released) identified arsenic as a primary metal exceeding Brazilian maximum residue limits in drinking water, with approximately 300 cancer cases attributable to drinking water contamination and elevated cancer mortality for ovarian, rectal, oropharyngeal, and central nervous system cancers [16]. A formal difference-in-differences test was non-significant, requiring caution on causal attribution [16]. A scoping review of 49 tailings dam disasters worldwide 2015-2025 (670 deaths, 112 injuries) found arsenic and lead to be the most frequently reported potentially toxic elements in tailings, while only 2 of 49 disasters had been investigated in population-based health studies [17].

A review of heavy metal air pollution and health documented ambient air arsenic concentrations of 2.87 +/- 2.08 ng/m3 in PM2.5 at measured sites, arsenic IARC Group 1 classification, and OSHA/NIOSH occupational exposure limits of 10 and 2 µg/m3 respectively, with copper smelter worker exposures approaching or exceeding the OSHA permissible exposure limit [18].

What remains uncertain

The relationship between dietary inorganic arsenic and non-malignant respiratory outcomes is not established with quantitative dose-response estimates in adult populations. The Spanish birth cohort data [1] represent the strongest available evidence linking low dietary arsenic exposure to childhood respiratory symptom burden, but the analysis was exploratory and lacked statistical power to identify specific respiratory diagnoses.

The CFTR dysfunction pathway [2] requires replication. The mediation analysis in the Bangladesh study did not reach statistical significance, so the degree to which CFTR-mediated chloride transport impairment drives arsenic-associated obstructive disease remains unresolved.

For early-life exposure, the Antofagasta cohort [5] provides compelling evidence of lifelong pulmonary function deficits, but the study population experienced unusually high arsenic concentrations (drinking water up to approximately 870 µg/L) compared to contemporary dietary exposures in high-income countries. Whether the persistent EBC metabolomic signature scales to lower exposures is unknown.

The mechanistic literature is primarily in vitro or animal-based. The arsenic concentrations used in cellular models (micromolar range) generally exceed those associated with drinking water or dietary exposures; concentration-response concordance between in vitro and human epidemiological data has not been formally demonstrated for respiratory endpoints.

Respiratory outcomes are understudied relative to bladder, skin, and lung cancer endpoints in the arsenic health effects literature. Most epidemiological studies in this corpus report on COPD or airflow obstruction as incidental outcomes rather than primary endpoints, limiting precision.

Update history

No substantive edit history is available in this build. The full commit record is available in git.