About

Joel D. Kaufman is a Professor in the Department of Epidemiology, Environmental and Occupational Health Sciences, and General Internal Medicine at the University of Washington. His educational background includes an MD from the University of Michigan obtained in 1986 and an MPH in Environmental Health from the University of Washington completed in 1990. His research focuses on environmental and occupational health, with particular interests in clean air, sustainable communities, environmental health, occupational medicine, pollution, and wildfires. Kaufman has contributed to understanding the health impacts of air pollution, environmental exposures, and community health, as evidenced by his recent publications on topics such as air pollution and brain structure, outdoor air pollution and cancer risk, and neighborhood environmental factors affecting health outcomes.

Research topics

• Medicine
• Internal medicine
• Environmental health
• Gerontology
• Psychiatry
• Environmental science
• Psychology
• Ecology
• Biology
• Demography
• Cardiology
• Geography
• Nursing
• Environmental planning
• Chromatography
• Chemistry
• Audiology
• Surgery
• Pathology
• Environmental chemistry

Selected publications

• The role of coronary artery calcification in metal-related cardiovascular disease

  Environment International · 2026-01-22

  articleOpen access

  • The association between urinary metals and cardiovascular disease (CVD) might be partially mediated by coronary artery calcification changes over time. • Cadmium showed a strong association with mortality for causes other than CVD, which lead to avoided cases of CVD due to premature death. • Other biological pathways unrelated to coronary artery calcification might mediate the association between copper and zinc and CVD. Metals are associated with cardiovascular disease (CVD), but the underlying pathways remain largely unclear. We evaluated the potential intermediate role of coronary artery calcification (CAC) trajectory on the association between urinary metals and incident CVD, accounting for competing risks by death from other causes. We used data from 6,459 participants of the Multi-Ethnic Study of Atherosclerosis (MESA). CAC was measured longitudinally using the spatially weighted calcium score in five exams, starting in 2000. Participants were followed for CVD events through 2019. Cadmium, cobalt, copper, uranium, tungsten, and zinc were measured in urine at the baseline visit (2000–2002). We used a causal inference algorithm with a path-specific effects approach for longitudinal mediation analysis to evaluate the intermediate role of CAC on the association between metals and incident CVD. The association with incident CVD mediated through the CAC trajectory was statistically significant for cadmium, cobalt, copper, tungsten, and zinc. The number of CVD cases (95% CI) per 100,000 person-years attributable to an interquartile range (IQR) increase in metal levels through the longitudinal trajectory of CAC was 44 (20, 72) for cadmium, 21 (6, 39) for cobalt, 19 (2, 36) for copper, 18 (2, 38) for tungsten, and 43 (26, 62) for zinc. This study supports that part of the association between urinary metals and CVD is attributable to changes in CAC over time. In particular, half of the association between urinary cadmium and CVD might be mediated by longitudinal changes in CAC. This study could inform strategies for early detection and prevention of CVD based on urinary metal levels.

  PublisherDOI

• Air pollution and cognitive performance in late life: the role of lipid profile – results from the Multi-Ethnic Study of Atherosclerosis (MESA)

  Environment International · 2025-08-18

  articleOpen access

  INTRODUCTION: Evidence on blood lipids in the air pollutant-cognition relationship is scarce. In this study, we identified the role of blood lipids in the air pollutant-cognition relationship in the Multi-Ethnic Study of Atherosclerosis (MESA) cohort. METHODS: between Exams 1 and 5 were assessed using validated spatiotemporal models. Total cholesterol (TC) and low-density lipoprotein cholesterol (LDL-C) concentrations in blood at Exam 5 were evaluated as mediators or modifiers using linear regression models. Models assessed gender and race/ethnicity related heterogeneity in air pollutant-cognition and blood lipid-cognition associations. RESULTS: ). Associations of cognitive performance with air pollutant exposure and blood lipids did not vary by gender or racial identity. DISCUSSION: TC and LDL-C were not mediators or clinically meaningful modifiers in the air pollutant-cognition relationship.

  PublisherDOI

• Neighborhood environment and incident diabetes, a neighborhood environment-wide association study (‘NE-WAS’): Results from the Hispanic Community Health Study/Study of Latinos (HCHS/SOL)

  PLoS ONE · 2025-07-29

  articleOpen accessCorresponding

  The prevalence of type 2 diabetes is increasing among the Hispanic/Latino population. Type 2 diabetes incidence rates vary between neighborhoods, but no single aspect of the neighborhood environment is known to cause type 2 diabetes. Using data from the Hispanic Community Health Study/Study of Latinos cohort of 16,415 Hispanic/Latino adults in four major US cities, we conducted a neighborhood environment-wide association study to identify neighborhood measures or clusters of measures associated with diabetes incidence. Two-hundred and four neighborhood measures were calculated at the census tract level or within a 1-km buffer of participants' residential addresses. Independent covariate-adjusted and survey-weighted Poisson regressions were run for each neighborhood measure and incident diabetes. Principal component analysis of neighborhood measures was conducted to reduce dimensionality. No coherent pattern of neighborhood measures or principal component scores were associated with diabetes incidence within the cohort, though established individual-level risk factors such as age and family history were strongly associated with diabetes incidence. Results from our analysis did not indicate specific neighborhood measures, clusters, or patterns. Individual, rather than neighborhood, factors distinguish incident diabetes cases from non-cases.

  PublisherDOI

• Associations between Ambient Air Pollutants and Clonal Hematopoiesis of Indeterminate Potential.

  UNC Libraries · 2025-07-26

  articleOpen access

  BACKGROUND: Clonal hematopoiesis of indeterminate potential (CHIP) is an age-related somatic mutation associated with incident hematologic cancer. Environmental stressors which, like air pollution, generate oxidative stress at the cellular level, may induce somatic mutations and some mutations may provide a selection advantage for persistence and expansion of specific clones. METHODS: We used data from the Multi-Ethnic Study of Atherosclerosis (MESA) N = 4,379 and the Women's Health Initiative (WHI) N = 7,701 to estimate cross-sectional associations between annual average air pollution concentrations at participant address the year before blood draw using validated spatiotemporal models. We used covariate-adjusted logistic regression to estimate risk of CHIP per interquartile range increases in particulate matter (PM2.5; 4 μg/m3) and nitrogen dioxide (NO2; 10 ppb) as ORs (95% confidence intervals). RESULTS: Prevalence of CHIP at blood draw (variant allele fraction > 2%) was 4.4% and 8.7% in MESA and WHI, respectively. The most common CHIP driver mutation was in DNMT3A. Neither pollutant was associated with CHIP: ORMESA PM2.5 = 1.00 (0.68-1.45), ORMESA NO2 = 1.05 (0.69-1.61), ORWHI PM2.5 = 0.97 (0.86-1.09), ORWHI NO2 = 0.98 (0.88-1.10); or with DNMT3A-driven CHIP. CONCLUSIONS: We did not find evidence that air pollution contributes to CHIP prevalence in two large observational cohorts. IMPACT: This is the first study to estimate associations between air pollution and CHIP.

  PublisherDOI

• Occupational Exposure to Ambient Air Pollution: At-Risk Worker Groups, Regulatory and Research Needs An Official American Thoracic Society Workshop Report

  Annals of the American Thoracic Society · 2025-10-31 · 1 citations

  articleOpen access

  Although health effects of ambient air pollution are well established in the general population, the impact of exposure in working populations remains poorly understood. Outdoor workers are disproportionately exposed to ambient air pollution, particularly with increasing wildfire smoke events and global climate change. An international interdisciplinary group of experts including worker representation assembled to review the current state of knowledge regarding the impact of occupational air pollution exposure on worker health and develop recommendations for research and actions to evaluate, mitigate, and regulate occupational air pollution exposure. The group identified health risks likely resulting from air pollution based on studies of the general population, noting that additional risks may be encountered from coexposures, as well as exertion increasing the work of breathing. High-risk groups were identified, including agricultural workers, construction workers, and wildland firefighters; others working in warehouses and indoor spaces are likely at risk via ambient air pollutant infiltration. It was estimated that at least 20 million outdoor U.S. workers are exposed to air pollution at work, which limits productivity and increases absenteeism. Participants recommended using air quality to guide work modifications and adoption of the hierarchy of exposure controls as a model to reduce exposures, as used by some states and proposed by the National Institute for Occupational Safety and Health for agricultural and other outdoor workers. Existing research supports the urgent need for policies to protect workers from exposure. Research gaps remain, including medical surveillance strategies, improved technology to protect workers, and studies specifically evaluating the impacts of occupational air pollution exposure.

  PublisherOA PDFDOI

• Ambient Air Pollution Exposure and Mortality in the Pulmonary Hypertension Association Registry

  Annals of the American Thoracic Society · 2025-06-16 · 1 citations

  articleOpen access

  Abstract Rationale The effects of long-term ambient air pollution exposure on survival in pulmonary arterial hypertension (PAH) and chronic thromboembolic pulmonary hypertension (CTEPH) remain unclear. Objectives To evaluate the association between exposure to particulate matter with aerodynamic diameter ≤ 2.5 μm (PM2.5), NO2, and O3 and mortality or lung transplant in the Pulmonary Hypertension Association Registry (PHAR). Methods A total of 2,196 adult patients enrolled in the PHAR provided data between 2015 and 2024. Annual average concentrations of air pollutants, including PM2.5, NO2, and O3, estimated from validated spatiotemporal models in 2015 were linked to each participant’s residential address. Cox proportional hazards models evaluated the associations between air pollutant exposures and risk of death or lung transplant, adjusting for baseline demographics, individual and neighborhood socioeconomic factors, disease severity, and spatial confounders. Additional analyses were adjusted for and stratified by nine U.S. census divisions. Results Study participants were broadly distributed across U.S. regions, with 72.0% female and a mean age of 55.7 years. A proportion of 35.6% had idiopathic PAH, 26.5% had connective tissue disease–associated PAH, and 14.5% had CTEPH. In models adjusted for demographics and individual and neighborhood socioeconomic factors, each interquartile range increase of PM2.5 was associated with a mortality or lung transplant hazard ratio of 1.16 (95% confidence interval, 1.01–1.33). This association was marginally attenuated and not statistically significant after adjusting for spatial covariates, with a hazard ratio of 1.12 (95% confidence interval, 0.95–1.31) per interquartile range increase in PM2.5. We noted regional variation in the observed associations. No significant associations were found with NO2 or O3. Conclusions Long-term ambient air pollution exposure was not significantly associated with survival in PHAR patients with PAH or CTEPH. Future research should investigate potential modifying effects of regional social determinants and healthcare-related factors on the relationship between air pollution exposure and mortality in these conditions.

  PublisherOA PDFDOI

• Pre- and post-move exposure to air pollution and neighborhood socioeconomic status

  The Science of The Total Environment · 2025-10-28

  articleOpen access

  Nearly 10 % of the U.S. population moves in a given year. We aimed to examine differences in neighborhood socioeconomic status (nSES) and ambient air pollution concentrations before and after a residential move, whether change varies by distance of the move, and whether moving is a potential source of bias in estimating relationships between contextual exposures and outcomes. We used data from participants in the Atherosclerosis Risk in Communities (ARIC) Study with geocoded addresses at Visit 2 (Visit 2; 1990–1992) and Visit 3 (Visit 3; 1993–1995). We quantified nSES using the 1990 census and estimated concentrations of fine particulate matter (PM 2.5 ), ozone (O 3 ) and nitrogen dioxide (NO 2 ). We defined short-distance and long-distance movers as having moved less and more than the median move distance of 4.3 miles, respectively. We quantified Visits 2 to 3 change in nSES and air pollution for short and long-distance movers, examined change in air pollution after accounting for temporal trends, and conducted a quantitative bias analysis to estimate the bias introduced by not accounting for moving in example studies of PM 2.5 and health. On average, compared to non-movers, short-distance movers relocated to neighborhoods with better nSES and similar air pollution concentrations, while long-distance movers relocated to neighborhoods with worse nSES but lower exposure to air pollution. The quantitative bias analysis suggested there was little to no bias in the PM 2.5 -health relationship from not accounting for relocation, likely due to relatively little overall difference in PM 2.5 exposures observed with relocation. Change in nSES and air pollution with a residential move differs by move distance. However, not accounting for relocation may not lead to significant bias in the relationship with the health outcome. Understanding the extent to which contextual exposures change after a move can clarify whether accounting for bias due to relocation is necessary. • Moving changes exposure to neighborhood SES (nSES) and air pollution • Short-distance moves led to better nSES, with little change in pollution levels. • Long-distance moves led to worse nSES but lower air pollution exposures. • PM 2.5 -health relationship showed minimal bias from not accounting for moving.

  PublisherDOI

• Table S1 from Associations between Ambient Air Pollutants and Clonal Hematopoiesis of Indeterminate Potential

  2025-11-26

  articleOpen accessSenior author

  <p>Supplementary table1 shows additional results for correction for sampling in the WHI cohort</p>

  PublisherDOI

• Associations Between Urinary Metal Levels and Incident Heart Failure

  JACC Heart Failure · 2025-06-17 · 5 citations

  article
  PublisherDOI

• Marginal structural models for quantifying the causal effects of exposure to ambient air pollution on progression of CT emphysema in the MESA lung and MESA air studies

  American Journal of Epidemiology · 2025-11-14

  articleOpen access

  Associations between exposure to ambient air pollution and progression of emphysema have been identified in longitudinal observational studies. However, previous work has not used statistical causal inference methods tailored to address bias from time-varying confounding. The objective of this study is to propose an analytical approach for estimating longitudinal health effects of air pollution while accounting for time-varying confounding using marginal structural models and to re-analyze data on air pollution and emphysema progression from the Multi-Ethnic Study of Atherosclerosis using this analytical approach. We estimate weights for continuous exposure levels using two techniques: quantile binning of the exposure and a semiparametric model for the requisite conditional densities. The latter approach incorporates flexible machine learning methods. We find evidence for the harmful effects of ambient ozone pollution during study follow-up on the progression of emphysema, consistent with previously reported results. We find no evidence of effects of NOx during study follow-up. This investigation demonstrates that analyses based on marginal structural models are feasible in studies of the health effects of air pollution and may address possible sources of bias that traditional regression-based methods fail to address. Further investigation is warranted to understand differences between our findings and previously published results.

  PublisherOA PDFDOI

Recent grants

• Integrative Environmental Health Sciences Facility Core

  NIH · $34.7M · 1997–2027

• NIH Grant R01OH003445

  NIH · $839k · 2002

• NIH Grant R01ES023500

  NIH · $3.2M · 2020

• NIH Grant R56ES026528

  NIH · $601k · 2020

• NIH Grant M01RR000037

  NIH · $22.1M · 2009

Frequent coauthors

• Lianne Sheppard

  University of Washington

  247 shared

• R. Graham Barr

  Columbia University Irving Medical Center

  217 shared

• Adam A. Szpiro

  University of Washington

  204 shared

• Timothy V. Larson

  University of Washington

  179 shared

• Kiang Liu

  Northwestern University

  175 shared

• James H. Stein

  172 shared

• Richard A. Kronmal

  University of Washington

  155 shared

• Patrick L. Kinney

  Boston University

  142 shared

Labs

• Department of EpidemiologyPI

Education

• Ph.D., Epidemiology

  University of Washington

  1990

• M.S., Epidemiology

  University of Washington

  1985

• B.S., Public Health

  University of California, Berkeley

  1983