Racial and ethnic differences in the impact of air pollution on the risk of Alzheimer's disease and related dementias in the Multiethnic Cohort Study.

Racial and ethnic differences in the impact of air pollution on the risk of Alzheimer’s disease and related dementias in the Multiethnic Cohort Study - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Search Log in Primary site navigation Search Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Environ Int . Author manuscript; available in PMC: 2026 May 31. Published in final edited form as: Environ Int. 2026 Feb 24;209:110169. doi: 10.1016/j.envint.2026.110169 Racial and ethnic differences in the impact of air pollution on the risk of Alzheimer’s disease and related dementias in the Multiethnic Cohort Study Anna H Wu a, * , Jun Wu b , Chiuchen Tseng a , Burcu F Darst c , Song-Yi Park d , Daniel O Stram a , Timothy Larson e , Scott Fruin a , Veronica W Setiawan a , Xian Yu a , Lynne R Wilkens d , Howard Hu a , Christopher Haiman a , Beate Ritz f , Eileen M Crimmins g , Unhee Lim d , Iona Cheng h , Loïc Le Marchand d Author information Article notes Copyright and License information a Department of Population and Public Health Sciences, Keck School of Medicine and Norris Comprehensive Cancer Center, University of Southern California, Los Angeles, CA 90033, United States b Department of Environmental and Occupational Health, Joe C. Wen School of Population and Public Health, University of California, Irvine CA 92697, United States c Public Health Sciences Division, Fred Hutchinson Cancer Center, Seattle, WA 98109, United States d Population Sciences in the Pacific Program, University of Hawaii Cancer Center, University of Hawaii at Manoa, Honolulu, HI 96813, United States e Department of Civil & Environmental Engineering and the Department of Environmental & Occupational Health Sciences, University of Washington, Seattle, WA 98109, United States f Division of Epidemiology, University of California, Los Angeles, CA 90025, United States g Leonard Davis School of Gerontology, Andrus Gerontology Center, University of Southern California, Los Angeles, CA 90089, United States h Department of Epidemiology and Biostatistics, University of California, San Francisco, CA 94143, United States * Corresponding author at: Department of Population and Public Health Sciences, Keck School of Medicine and Norris Comprehensive Cancer Center, University of Southern California, Los Angeles, CA 90033, United States. annawu@usc.edu (A.H. Wu). Issue date 2026 Mar. This is an open access article under the CC BY-NC license ( http://creativecommons.org/licenses/by-nc/4.0/ ). PMC Copyright notice PMCID: PMC13222114  NIHMSID: NIHMS2159068  PMID: 41793956 The publisher's version of this article is available at Environ Int Abstract Background and Objectives: Meta-analysis results, based largely among Whites, suggested that fine particulate matter (PM 2.5 ) exposure increases the risk of clinical dementia. This study investigated the association of air pollution and incidence of Alzheimer’s disease and related dementias (ADRD) by race and ethnicity. Methods: We investigated incidence of AD (n = 4,010) and other dementia (n = 4,971) among 44,954 California Multiethnic Cohort (MEC) participants (28% African American, 14% Japanese American, 44% Latino, 14% White adults) who were enrolled in the fee-for-service component of Medicare (2001–2016). We used Cox proportional hazards regression to examine associations between exposure to PM, airport-related ultrafine particles (aUFP) and gaseous pollutants and incidence of AD, other dementia, and ADRD in a minimally- and fully-adjusted model, considering 12 established ADRD risk factors. We conducted stratified analyses to examine associations by sex, and race/ethnicity. Results: ADRD incidence was associated with PM 2.5 (per 2 μg/m 3 ), airport-related UFP (aUFP, per 4400 particles/cm 3 ) and nitrogen dioxide (NO 2 , per 10 μg/m 3 ) with hazard ratios (HRs, 95%CI), respectively, of 1.04 (1.02–1.06), 1.03 (1.01–1.05) and 1.09 (1.06–1.12). The AD-associations with PM 2.5 and NO 2, were stronger than the corresponding associations with other dementia (P heterogeneity ≤ 0.003). Similar patterns of results were observed by sex and across race and ethnicity. Statistically significant findings for ADRD with PM 2.5, aUFP and NO 2 were observed among African American (respective HRs 1.03, 1.04, 1.09), and Latino and White participants for NO 2 (HR 1.10, 1.08). Results in all and African American participants remained statistically significant in fully-adjusted models. Although the effect of PM 2.5 was diluted in a co-pollutant with NO 2 , both PM 2.5 and aUFP were significantly associated with ADRD incidence in a co-pollutant model, and NO 2 and aUFP (but not PM 2.5 ) remained associated in a multipollutant model. We did not observe consistent modifying effects for any of the 12 established ADRD risk factors. Conclusions: In this multiethnic population, incidence of ADRD increased with exposures to PM 2.5 , aUFP, and NO 2 in all subjects and this pattern was most prominent among African American adults. These results emphasize that ADRD prevention should include not only individual-level factors but also population-wide policies and regulation to curb air pollution. 1. Introduction The number of people worldwide living with Alzheimer’s disease and related dementias (AD/ADRD) increased from ~36 million around 2012 to 57 million in 2019 and this figure is projected to double by 2050. (Collaborators GBDDF. Estimation of the global prevalence of dementia in, 2019) This rapid increase poses significant challenges to healthcare systems and economics globally. Sociodemographic, lifestyle, and cardiometabolic conditions are well-established risk factors for AD/ADRD. ( Livingston et al., 2020 ; Park et al., 2024 ) However, the role of ambient air pollution in the etiology of AD/ADRD remains equivocal. ( Tsai et al., Oct 2019 ; Fu and Yung, 2020 ) A 2023 meta -analysis of 16 studies found that exposure to fine particulate matter (PM 2.5 ) was associated with increased dementia risk (hazard ratio (HR) per 2 μg/m 3 =1.04, 95% CI 0.99–1.09), with substantially stronger associations in a subset of nine studies using active case ascertainment to diagnose dementia (HR=1.42, 95% CI 1.00–2.02). Risk increased in association with exposure to nitrogen dioxide (NO 2 ) and nitrogen oxide (NOx) but no association was observed with ozone (O 3 ) exposure. ( Wilker et al., 2023 ) An updated 2025 meta -analysis strengthened the overall evidence, showing a significant association between dementia incidence and exposure to PM 2.5 , NO 2 , and also PM 10 . ( Best Rogowski et al., Jul 2025 ) However, studies included in these meta -analyses were conducted among predominantly White populations in the US ( Ilango et al., Jun 2023 ; Shaffer et al., Aug 2021 ; Sullivan et al., Aug 2021 ; Shi et al., 2023 ; Yu et al., Mar 2023 ; Zhang et al., 2023; Wang et al., 2022 ) [two studies included Latino, ( Yu et al., Mar 2023 ; Zhang et al., 2023) three studies included African American ( Shi et al., 2023 ; Zhang et al., 2023; Wang et al., 2022 ), Canada, ( Chen et al., Nov 2017 ; Smargiassi et al., 2020 ; Yuchi et al., 2020 ; Ilango et al., 2020 ) United Kingdom, (Wood, 2024; Zhang et al., 2022 ; Carey et al., 2018 ) Europe, ( Mortamais et al., Mar 2021 ; Grande et al., 2020 ; Andersson et al., 2023 ; de Crom et al., 2023 ) and Australia, ( Bagheri et al., 2021 ; Trevenen et al., Dec 2022 ) with only two studies from Asia. ( Ran et al., 2021 ; Yan et al., 2022 ) Thus, while these meta -analyses investigated potential differences by geography, no information could be presented by race and ethnicity and had limited generalizability. The incidence of ADRD in the US is 1.5 to 2 times higher among African American and Latino populations than US White populations, which may be related, in part, to their higher prevalence of established risk factors including cardiometabolic conditions (e.g., heart disease, stroke, hypertension, diabetes), low education, low neighborhood socioeconomic status (nSES), not married, current smoking, low physical activity, obesity, poor diet quality and poor sleep. ( Park et al., 2024 ) In addition, minoritized population groups are more likely to live in neighborhoods of low SES, near sources of air pollutants due to decades of racial and ethnic segregation and may have greater susceptibility to the health effects of air pollution. ( Hajat et al., Dec 2015 ; Woo et al., Mar 2019 ) The geographical, social, and demographic imbalance of published studies highlight a knowledge gap of the role of air pollution on ADRD incidence across racially and ethnically diverse populations. ( Livingston et al., 2020 ) We address these unanswered questions by examining the role of ambient air pollution and incidence of ADRD using data from California participants in the Multiethnic Cohort (MEC) Study (over 70% African Americans and Latino participants) enrolled in Medicare Fee-for-Service (FFS). We report here the role of long-term (2001–2016) exposure to time-varying particulate matter (PM 2.5 , PM 10 , airport-related ultrafine particles (aUFP), <100 nm in diameter) and gaseous pollutants (NO 2 , a common traffic-related pollutant, NOx, carbon monoxide, ozone (O 3 )) in relation to incidence of ADRD with adjustment for sociodemographic, lifestyle, and cardiometabolic risk factors. These specific air pollutants were selected based on their prevalence in urban environments and their potential to cause neurological harm. ( Calderon-Garciduenas et al., Feb 2008 ; Oberdorster et al., Jun 2004 ; Hussain et al., Nov 2023 ) Both PM 2.5 and UFP can penetrate deeply into the respiratory system, cross the blood–brain barrier, and trigger adverse neurological effects. This 16-year exposure period aligns with the extended latency of ADRD development, allowing for a robust assessment of long-term air pollution effects. 2. Materials and Methods 2.1. Study Participants The MEC was established in 1993–1996 with a population-based sample of more than 215,000 men and women, living in Hawaii or California (largely Los Angeles County) who completed a 26-page questionnaire at enrollment that assessed sociodemographic characteristics (e.g., race and ethnicity, education, marital status), lifestyle factors (e.g., body weight, height, smoking, alcohol, physical activity, diet), personal medical history and other factors. ( Kolonel et al., 2000 ) The cohort has been linked to the Centers for Medicare and Medicaid Services (CMS) claims data for 1999–2016 ( Setiawan et al., 2015 ) to identify chronic disease outcomes including ADRD based on the International Classification of Diseases (ICD) codes included in the claims, in addition to linkage to Hawaii and California death files and the National Death Index for vital status. However participants enrolled in managed care (i.e., Medicare Part C) were excluded because their medical services were not directly billed to Medicare. ( Setiawan et al., 2015 ) As in previous studies of late onset ADRD in the MEC, ( Park et al., 2024 ; Lim U, Wang S, Park SY, et al. Risk of Alzheimer’s disease and related dementia by sex and race/ethnicity: The Multiethnic Cohort Study. Alzheimers Dement. Dec 9, 2021 ) this analysis included Medicare FFS beneficiaries with available claims data (n=106,080) among the five racial and ethnic groups (African American, Japanese American, Latino, Native Hawaiian, and White). Because the availability of EPA air pollution monitor data are sparse in Hawaii, we have developed measures of criteria air pollutants only for the California component of the MEC and thus FFS Hawaii MEC participants (n=57,512) were excluded. Of the 48,568 FFS California MEC participants, 44,954 remained in the current analysis ( Fig. 1 ), after excluding 3,164 participants (598 had <1 month of follow-up after Medicare surveillance start in 1999 or at a later date, 2,064 exited MEC cohort before 2001, 876 had incomplete address information, and 78 self-identified as Native Hawaiian). Supplement Table 1 showed that the excluded participants were more likely to be African American and less likely to be Japanese American participants, and also had a higher prevalence of established ADRD risk factors than those remaining in the analyses. Fig. 1. Open in a new tab Flow chart of California Multiethnic Cohort Participants included in analysis on AD/ADRD and air pollution. 2.2. Standard Protocol Approvals, Registrations, and Patient Consents The IRBs at the UH (CHS 9575) and the USC (HS-17-00714) approved the MEC study protocol and agreed that implicit consent was granted by participants who returned a completed baseline questionnaire. We adhere to the Data Use Agreement with CMS on our Limited Data Set use for research. 2.3. ADRD Ascertainment As in previous publications on late-onset ADRD in the MEC, ( Park et al., 2024 ; Lim U, Wang S, Park SY, et al. Risk of Alzheimer’s disease and related dementia by sex and race/ethnicity: The Multiethnic Cohort Study. Alzheimers Dement. Dec 9, 2021 ; Park et al., 2022 ) cases were identified during the Medicare linkage period based on the ICD code using CMS chronic conditions algorithm. We defined ADRD as AD or related dementia based on the definition of Medicare ( Taylor et al., Sep 2002 ) and methods described by Goodman and colleagues, ( Goodman et al., Jan 2017 ) and obtained the following ADRD subtypes: AD-only, AD of mixed etiology, other dementia including degenerative dementia not otherwise specified (NOS), vascular dementia (VD)-only, Lewy body dementia (LBD)-only, and frontotemporal dementia (FTD). We present here results separately for AD (included AD-only and AD of mixed etiology), other dementia (included dementia NOS, VD, LBD, and FTD) and ADRD (AD and other dementia combined). 2.4. Address History, Geocoding, and Neighborhood Socioeconomic Status (nSES) The MEC maintains up-to-date addresses on all participants by periodic mailings of newsletters, follow-up questionnaires, and linkages to administrative data and registries. These geocoded addresses were used to assess nSES, using a composite measure of seven census-based indicators (education, median household income, percent living 200% below poverty level, percent blue-collar workers, percent ≥16 years in workforce without job, median rent, and median house value). ( Yost et al., Oct 2001 ; Yang et al., 2014 ) The nSES index was assigned to participants’ census block group in California for all addresses from the time of Medicare enrollment to the end of study follow-up (i.e., ADRD diagnosis, death, or December 31, 2016); low and high nSES were defined as quintiles 1–3 and quintiles 4–5, respectively. 2.5. Air Pollution Exposure Assessment Time-dependent cumulative average air pollution exposure estimates were developed for 44,954 FFS California MEC participants, using information on month and year living at each of the residential addresses to generate the cumulative average air pollution value based on each month at the residence. Proportional values are assigned for partial months, such as in the start year/month, end year/month and the year/months when the individual moved. If the start date for the address is before or on the 15 th day of the month, the calendar month for the date was used to compute year/month value, and if the start date is after the 15 th day of the month, the preceding calendar month was used to compute year/month value. Similarly, if the end date for the address is before or on the 15 th day of the month, the previous calendar month was used to compute year/month value, and if the date is after the 15 th day of the month, the calendar month was used. As in our previous studies, ( Wu et al., 2025 ; Bookstein et al., 2024 ; Wu et al., 2021 ; Cheng et al., 2022 ) estimates of satellite-based PM 2.5 concentrations (2001–2016) were obtained from a fine-resolution spatiotemporal model that provided validated and publicly available monthly PM 2.5 outputs at a 1-km resolution over North America by statistically fusing chemical transport model (GEOS-Chem) predictions and satellite observations of aerosol optical depth with ground-based observations. ( Meng et al., 2019 ) Monthly ultrafine particle concentrations (UFP) from Los Angeles International Airport (LAX) (referred to hereafter as airport-related UFP, or aUFP) emissions were estimated at a 1-km resolution using a recommended American Meteorological Society/Environmental Protection Agency (EPA) Regulatory Dispersion Model (AERMOD), which accounted for variations in airport flight activity and meteorology. ( Hudda et al., 2014 ) Concentrations of PM 10 and gaseous traffic-related pollutants (NO 2 , NOx, CO) and O 3 were assessed using monitoring data routinely collected by the U.S. EPA at participants’ addresses, applying empirical Bayesian kriging interpolation. California maintains one of the most extensive air monitoring networks in the world, with approximately 250 regulatory monitoring sites statewide (California Air Resources Board, 2023), including 35 permanent stations for criteria air pollutants in the Los Angeles Basin where the majority of MEC participants reside (South Coast AQMD, 2023). Among the 44,954 participants who remained in our analysis for PM and gaseous pollutants, 91% had complete data and 9% had some imputed data (5% had <10% imputed, 1% had 10-<20% imputed, and 3% had 20-<50% imputed); the exposure levels of the last known addresses were used for the subsequent period missing an address and requiring imputation. ( Cheng et al., 2020 ) HR estimates including the 9% participants with imputed air pollution measures were similar to the HR estimates that excluded them (data not shown) and all subsequent analyses included participants with imputed data. For aUFP only, we restricted our analysis to 68% of total subjects (n=30,703) with any residential addresses within the exposure grid zone surrounding LAX, given that UFP measurements were only available for this geographic area. Baseline demographic and lifestyle characteristics of participants without UFP data (n=14,251) were similar to the 44,954 participants except a higher percent of White and a lower percent of Japanese American participants were excluded (Column C, Table S1 ). Tables S2a and Table S2b present the correlation matrices of these air pollutants by race and ethnicity. Results for PM 2.5 , aUFP and NO 2 are shown in the main tables and results for PM 10 , NOx, CO and O 3 in the supplementary tables , in part, because of the lower penetration into and beyond the respiratory system of PM 10 compared with PM 2.5 , the high correlations between NO 2 and NOx (R=0.95) and between CO and NO 2 /NOx (R= 0.82 to 0.87), and because O 3 as modeled here is a secondary pollutant. 2.6. Statistical Analysis. We estimated incidence of AD, other dementia, and the combined ADRD in relation to air pollution using Cox proportional hazard models with monthly time-varying air pollutant exposures. The exposure period began 2 years after January 1, 1999 when the first Medicare surveillance data providing the ADRD diagnosis were available for participants who were Medicare beneficiaries as of January 1999 or 2 years after Medicare enrollment for those who enrolled after 1999 to avoid immortal person time. Age in calendar months was used as the time variable and defined a series of risks based on month/year at identification of each AD, other dementia (index case). Each risk set consisted of all MEC California participants, who remained alive and uncensored at the age of diagnosis of cases of AD and other dementia. For each member of each risk set (including the index case), the independent variable for exposure was the average air pollutant exposure for the period starting from the time of Medicare surveillance start (month/year) up to the age of AD and other dementia diagnosis of the index case in each risk set based on their residential history. Analyses were performed for the combined outcome of AD and other dementia, the outcome of AD where other dementia event was censored at diagnosis and the outcome of other dementia where AD was censored at diagnosis. We first conducted a minimally adjusted model which included as strata variables age at Medicare surveillance start (single-year age categories), and Medicare usage (in-patient, out-patient) and adjusted for race and ethnicity and sex (Model A). We then conducted a fully adjusted model (Model B), including 12 sociodemographic, lifestyle, and cardiometabolic risk factors assessed at cohort entry in 1993–1996 (education, marital status, history of hypertension, stroke, diabetes, heart disease, smoking status, physical activity, sleep duration, body mass index (BMI), and diet quality) and nSES at Medicare entry and at censoring. The associated population attributable fraction for ADRD in relation to these 12 factors combined was 24.0% in men and 22.8% in women in the MEC. ( Park et al., 2024 ) For comparison of our results to the recent meta-analysis results, ( Wilker et al., 2023 ) we estimated HRs and 95% CI for a fixed unit increase of PM (per 2 μg/m 3 for PM 10 and PM 2.5 ), NO 2 (10 μg/m 3 or 5.32 ppb), NOx (10 μg/m 3 or 5.23 ppb), CO (100 μg/m 3 or 86.96 ppb), and O 3 (5 μg/m 3 or 2.55 ppb) that are comparable to the interquartile range (IQR) of the average exposures over follow-up in the MEC (2.73 μg/m 3 for PM 2.5 , 3.98 μg/m 3 for PM 10 and 5.29, 14.25, 134.4, and 2.99 in ppb units, respectively for NO 2 , NOx, CO, and O 3 ). For aUFP, we used the IQR of 4400 particle counts/cm 3 as we have done in previous studies. ( Bookstein et al., 2024 ; Wu et al., 2021 ) Deviations from proportionality were checked using an analysis of Schoenfeld residuals and we found no violation of this assumption. Results obtained using Fine-Gray model to account for competing risks were similar to those using Lunn-McNeil augmentation approach; ( Lunn and McNeil, Jun 1995 ) we showed results obtained using the latter approach where AD or other dementia parameter estimates were fit by a cause-specific model in a separate stratum. Parameter estimates for AD and other dementia were compared using the Wald test. We also conducted stratified analyses by sex, race and ethnicity. As air pollution is a complex mixture of correlated pollutants, we conducted co-pollutant and multipollutant models to examine the concurrent contribution of PM 2.5, NO 2 and aUFP exposures to ADRD incidence in all subjects and by race/ethnicity. Sensitivity analyses were performed to check the robustness of the results from the main models. We investigated whether there are residual differences because participants entered Medicare at different ages/calendar years by further adjusting for Medicare entry year. In addition, we evaluated whether the air pollution-ADRD associations may be modified by the 12 ADRD risk factors. We also explored the role of genotype information (APOE) that was available in a subset of participants. To test for heterogeneity in HRs by race and ethnicity and other subgroups used in stratified analyses, we evaluated an interaction term in the model for each pollutant with subgroup membership using a Wald statistic. All statistical tests were based on two-tailed probability and a significance level was set at alpha <0.05. Participants were followed from January 2001 or 2 years after Medicare entry to the diagnosis of ADRD, or death, or end of follow-up (December 31, 2016). Death was identified through linkages to the National Death Index and state death certificate files. Mean follow-up time was 9.5±4.3 years. All analyses were performed using SAS 9.4 statistical software (SAS Institute, Cary NC). 3. Results 3.1. Study population and characteristics This analysis included 44,954 California MEC Medicare-FFS participants, of whom 28.2% self-reported as African American, 13.6% Japanese American, 43.6% Latino American, and 14.6% White American. The mean summary score of the 12 sociodemographic (education, marital status, nSES), lifestyle (smoking, physical activity, sleep, BMI, and diet quality) and cardiometabolic factors (hypertension, diabetes, heart disease, and stroke) was highest among African American (4.0 ±1.9), followed by Latino (3.8±1.7), White (2.6±1.7) and was lowest among Japanese American (2.4±1.5) participants ( Table 1 ). Air pollutant levels differed by race and ethnicity, especially for aUFP. Relative to other racial and ethnic groups, levels of aUFP were almost doubled in African American participants and levels of PM 2.5 and NO 2 were 4% to 9% higher among Latino participants (all P’s <0.001). These patterns by race and ethnicity were observed in male and female MEC participants ( Table S3 ). Table 1. Distribution of risk factors (in percents) at baseline (1993–1996) and mean (standard deviation) levels of air pollutants among California Multiethnic Cohort (MEC) participants included in the study on Alzheimer’s disease and related dementia (AD/ADRD), by race and ethnicity. All African American Japanese American Latino White N = 44,954 N = 12,674 N = 6,105 N = 19,617 N = 6,558 Mean (SD) age at cohort entry 60.1 (7.9) 60.8 (8.7) 61.0 (8.5) 59.1 (7.1) 60.8 (7.9) Mean (SD) age at AD/ADRD follow-up start in Medicare 69.8 (4.8) 70.8 (5.2) 70.4 (5.1) 68.9 (4.2) 70.3 (5.0) Percent with below characteristics % % % % % Less education (≤12th grade) 48.6 39.5 26.4 67.3 31.2 Low neighborhood socioeconomic status (nSES) a (Quintile 1–3) 65.1 74.2 38.3 75.9 40.1 Not married 38.6 54.8 24.5 33.7 34.9 Hypertension (yes) 39.0 53.1 36.6 33.2 31.4 Stroke (yes) 2.1 3.2 1.8 1.6 1.8 Diabetes (yes) 11.4 12.9 8.3 13.3 5.6 Heart disease (yes) 8.5 10.9 5.5 8.3 7.4 Current smoking 14.9 20.1 10.1 8.3 14.8 Physical inactivity (<30 min/day) 47.8 51.5 37.8 52.4 36.2 Short/long sleep (≤ 5 hr/ ≥ 9hr per day) 24.0 30.9 12.8 25.3 17.6 Obesity (BMI ≥ 30) 23.1 32.6 5.0 24.5 18.6 Low-quality diet (HEI-2015 < 60) 23.5 16.4 28.4 28.3 18.5 Mean (SD) # 12 risk factors b 3.5 (1.8) 4.0 (1.9) 2.4 (1.5) 3.8 (1.7) 2.6 (1.7) APOE (1 e4 allele) c 23.2 32.1 16.9 19.9 23.0 APOE (2 + e4 allele) 2.0 3.8 0.7 1.3 2.0 Mean (SD) air pollutants First 12 months PM 2.5 (μg/cm 3 ) d 19.1 (3.7) 18.3 (3.8) 18.9 (3.7) 19.7 (3.6) 18.8 (3.9) Average follow-up PM 2.5 15.6 (2.8) 15.6 (2.8) 15.7 (2.5) 16.3 (2.4) 16.0 (2.6) First 12 months NO 2 ((μg/cm 3 ) d 51.2 (11.4) 48.8 (11.2) 50.5 (11.5) 53.1 (10.8) 50.5 (12.2) Average follow-up NO 2 41.4 (8.6) 40.4 (8.7) 40.5 (8.3) 42.4 (8.2) 40.7 (9.6) First 12 months airport-related ultrafine particles (aUFP/cm 3 ) d 6,476 (6159) 9,881 (7186) 4,818 (3461) 5,323 (5356) 5,031 (6029) Average follow-up aUFP (particle #) 6,642 (6082) 10,165 (7451) 5,039 (3482) 5,401 (5127) 5,190 (5941) Participants with measures on PM 2.5 , NO 2 and aUFP N = 30703 N = 8718 N = 4688 N = 13381 N = 3916 First 12 months PM 2.5 (μg/cm 3 ) d 19.2 (3.2) 19.0 (3.1) 19.6 (3.3) 20.2 (3.2) 19.6 (3.3) Average follow-up PM 2.5 16.5 (2.1) 16.2 (2.1) 16.2 (2.0) 16.8 (2.0) 16.4 (2.3) First 12 months NO 2 ((μg/cm 3 ) d 53.8 (10.0) 51.6 (9.6) 52.6 (10.7) 55.5 (9.6) 54.2 (10.4) Average follow-up NO 2 43.4 (7.7) 42.5 (7.6) 42.1 (7.7) 44.4 (7.3) 43.8 (8.2) First 12 months airport-related ultrafine particles (aUFP/cm 3 ) d 6522 (6182) 9971 (7189) 4811 (3465) 5317 (5341) 5017 (5964) Average follow-up aUFP (particle #) 6692 (4783) 10,247 (6990) 5043 (3484) 5396 (5112) 5181 (5889) Open in a new tab a At time of Medicare entry. b Mean summary score of above 12 risk factors; a score of 1 was assigned to the category listed and a score of 0 was assigned otherwise. c APOE genotype information on was based on 15,640 participants. d Based on 44,954 with information on PM 2.5 ; 41,952 with information on NO 2, and 30,703 with information on airport-related UFP (aUFP). 3.2. Single pollutant results In minimally adjusted models, incidence of ADRD in men and women combined was positively associated with exposure to PM 2.5 (HR 1.04, 95% CI 1.02–1.06), aUFP (HR 1.03, 95% CI 1.01–1.05), and NO 2 (HR 1.09, 95% CI 1.06–1.12) ( Table 2 ). The respective HRs for AD were 1.07 (95% CI 1.04–1.09), 1.03 (95% CI 1.00–1.06), and 1.18 (95% CI 1.13–1.23) and the corresponding HRs for other dementia were 1.03 (95% CI 1.01–1.05), 1.03 (95% CI 1.01–1.06), and 1.06 (95% CI 1.02–1.10). In fully adjusted models for other dementia, all the 95% CI included the null, whereas the HRs for AD and ADRD associated with PM 2.5 and NO 2 remained statistically significant. HR estimates associated with these pollutants showed largely comparable associations by subtypes of dementia ( Table S4 ). Table 2. Association of AD (AD only or mixed etiology), other dementia, and ADRD incidence and exposure to fine particulate matter (PM 2.5 ), airport-related ultrafine particles (aUFP), and nitrogen dioxide (NO 2 ), in all subjects and by sex, California Multiethnic Cohort participants. Males and Females Males Females Pollutants in HR analysis a Event HR (95% CI) b Model A HR (95% CI) c Model B Event HR (95% CI) b Model A HR (95% CI) c Model B Event HR (95% CI) b Model A HR (95% CI) c Model B Phet e sex PM 2.5 per 2 μg/m 3 AD 4010 1.07 (1.04,1.09) 1.06 (1.03,1.08) 1389 1.07 (1.03,1.11) 1.06 (1.01,1.10) 2621 1.07(1.04,1.1) 1.06 (1.03,1.09) 0.89 Other dementia 4971 1.03 (1.01, 1.05) 1.01 (0.99,1.03) 1910 1.03 (1.00,1.06) 1.01 (0.97,1.04) 3061 1.03 (1.00,1.06) 1.01 (0.99,1.04) 0.94 Phet d 0.003 0.002 0.04 0.14 0.03 0.003 ADRD 8981 1.04 (1.02,1.06) 1.03 (1.01,1.04) 3299 1.04 (1.01,1.06) 1.02 (0.99,1.05) 5682 1.04 (1.02,1.06) 1.03 (1.01,1.05) 0.85 aUFP any grid Per 4400 particles/cm 3 AD 2793 1.03 (1.00,1.06) 1.03 (1.00,1.06) 979 1.05 (1.00,1.10) 1.04 (0.99,1.10) 1814 1.02 (0.99,1.06) 1.01 (0.98,1.05) 0.42 Other dementia 3402 1.03 (1.01,1.06) 1.02 (0.99,1.04) 1311 1.03 (0.99,1.08) 1.01 (0.96,1.06) 2091 1.03 (0.99,1.06) 1.02 (0.98,1.05) 0.78 Phet d 0.57 0.80 0.93 0.45 0.50 0.76 ADRD 6195 1.03 (1.01,1.05) 1.02 (1.00,1.04) 2290 1.04 (1.01,1.07) 1.02 (0.99,1.06) 3905 1.02 (1.00,1.05) 1.01 (0.99,1.04) 0.77 NO 2 per 10 μg/m 3 AD 3793 1.18 (1.13,1.23) 1.16 (1.11,1.22) 1333 1.13 (1.05,1.22) 1.12 (1.04,1.21) 2460 1.20 (1.14,1.27) e 1.19 (1.12,1.26) 0.19 Other dementia 4604 1.06 (1.02,1.10) 1.01 (0.97,1.06) 1808 1.06 (1.00,1.13) 1.01 (0.95,1.09) 2882 1.06 (1.01,1.11) 1.02 (0.97,1.07) 0.98 Phet d <0.0001 <0.0001 0.05 0.11 0.0001 0.0003 ADRD 8797 1.09 (1.06,1.12) 1.06 (1.03,1.09) 3141 1.08 (1.03,1.14) 1.05 (1.00,1.10) 5342 1.10 (1.06,1.14) 1.07 (1.03,1.11) 0.47 Open in a new tab a NO 2 10 μg/m 3 - =5.32 ppb; aUFP = airport-related ultrafine particles; ADRD = AD and other dementia. b MODEL A – Cox proportional hazards regression model using air pollution levels from 1999 to exit for age scale (from entry month to exit month) and stratum by age at start Medicare and Medicare usage (in = -patient, out-patient) and adjusted for race/ethnicity, sex (models in men and women combined). c MODEL B - Time-depend Cox proportional hazards regression model using air pollution levels from 1999 to exit for age scale (from entry month to exit month) and stratum by age at start Medicare and Medicare usage (in = -patient, out-patient) and adjusted for continuous age at baseline, race/ethnicity, sex (models in men and women combined), education (<=high school graduate, some college, college graduate, graduate and professional school, missing), marital status (married, separated, divorced, widowed, never married, dk), hypertension (0,1), stroke (0,1), diabetes (0,1), heart attack (0,1), smoking status (no, ex, current, dk), moderate physical activity (<2.5 hr/week, 2.5-<5, 5-<7, 7-<14, 14+, dk), sleep (0,1), BMI at baseline (<25, 25-<30, 30+, dk), low diet quality (<60,>=60, DK), nSES at Medicare entry (1–5, dk) and current nSES (1–5, dk). d P heterogeneity (df = 1) comparing AD vs other ADRD. e P heterogeneity (df = 1) comparing HR results by sex; all P > 0.05 for Model A (not shown). Data shown are for Model B. Exposure to PM 10, NOx, and CO was significantly positively associated with increased AD and ADRD incidence but there was no association with O 3 exposure ( Table S5 ). In men and women combined, the magnitude of the HR estimates associated with PM 2.5 , PM 10 , NO 2 , NOx, and CO were significantly higher for AD than for other dementia in both minimally and fully adjusted models (P heterogeneity ranged from 0.01 for CO to <0.0001 for NO 2 ). HR estimates were comparable in females and males. All results we discuss below on ADRD are based on men and women combined. Table 3 shows the race/ethnic specific HR estimates for AD, other dementia, and ADRD. We found no evidence of significant heterogeneity in associations by race and ethnicity, but some suggestive differences in the magnitude of the effect estimates emerged. In minimally adjusted models, African American participants showed elevated HRs for ADRD in association with exposure to PM 2.5 (HR 1.03, 95% CI 1.01–1.06), aUFP (HR 1.04, 95% CI 1.01–1.07), and NO 2 (HR 1.09, 95% CI 1.04–1.14). Although these pollutants conferred elevated risks of ADRD among Japanese American, Latino and White participants, statistically significant HRs were observed only with NO 2 among Latino (HR 1.10, 95% CI 1.04–1.16) and White (HR 1.08, 95% CI 1.01–1.16) participants. However, the fully adjusted HRs for ADRD with aUFP and NO 2 remained statistically significant only in African American. HRs were higher for AD than for other dementia across race/ethnicity; the P heterogeneity were ≤0.05 among African American participants. HR estimates associated with exposures to PM 10, NOx, and CO were statistically significant for AD and ADRD among African American participants but not among Japanese American, Latino, and White participants ( Table S6 ) in both minimally and fully adjusted models (all P heterogeneity >0.05). There was no association with O 3 exposure. Table 3. Association of AD (AD only or mixed etiology), other dementia, and ADRD incidence and exposure to fine particulate matter (PM 2.5 ), airport-related ultrafine particles (aUFP), and nitrogen dioxide (NO 2 ), by race and ethnicity, California Multiethnic Cohort participants. Pollutants in HR analysis a African American Japanese American Latino White Phet (race) Event HR (95% CI) b Model A HR (95% CI) b Event HR (95% CI) b Model A HR (95% CI) b Event HR (95% CI) b Model A HR (95% CI) b Event HR (95% CI) b Model A HR (95% CI) b PM 2.5 per 2 μg/m 3 AD 1408 1.06 (1.03,1.10) 1.06 (1.02, 1.09) 534 1.09 (0.99,1.19) 1.09 (0.99, 1.20) 1473 1.05 (1.00,1.09) 1.04 (0.99,1.09) 595 1.08 (1.02,1.14) 1.07 (1.01,1.14) 0.69 Other dementia 1808 1.03 (1.00,1.06) 1.02 (0.99,1.05) 615 1.02 0.93,1.11) 1.01 (0.92, 1.10) 1781 1.03 (0.99,1.07) 1.01 (0.96,1.05) 767 1.02 (0.97,1.08) 1.00 (0.95,1.06) 0.74 Phet d 0.34 0.05 0.10 0.25 0.18 0.23 0.11 0.07 ADRD 3216 1.03 (1.01,1.06) 1.03 (1.00, 1.05) 1149 1.04 (0.98,1.10) 1.03 (0.97, 1.10) 3254 1.03 (1.00,1.06) 1.02 (0.99,1.05) 1362 1.04 (1.00,1.08) 1.03 (0.99,1.07) 0.93 aUFP per 4400 particles/cm 3 AD 873 1.05 (1.01,1.10) 1.05 (1.01, 1.10) 452 0.96 (0.83,1.11) 0.95 (0.82, 1.10) 1068 1.01 (0.96,1.07) 1.00 (0.95,1.06) 366 1.03 (0.96,1.10) 1.03 (0.96, 1.11) 0.56 Other dementia 1136 1.04 (1.00,1.08) 1.04 (1.00, 1.08) 500 1.02 (0.91,1.15) 1.04 (0.92, 1.17) 1294 1.05 (1.00,1.09) 1.02 (0.97, 1.07) 472 0.98 (0.91,1.06) 0.94 (0.87, 1.02) 0.36 Phet d 0.63 0.62 0.49 0.39 0.34 0.49 0.36 0.18 ADRD 2009 1.04 (1.01,1.07) 1.04 (1.01, 1.07) 952 1.00 (0.91,1.09) 1.00 (0.91, 1.10) 2362 1.03 (1.00,1.07) 1.02 (0.98, 1.05) 872 1.01 (0.96,1.06) 0.99 (0.94, 1.04) 0.61 NO 2 per 10 μg/m 3 AD 1288 1.21 (1.13,1.30) 1.21 (1.13,1.29) 529 1.11 (0.96,1.27) 1.12 (0.96,1.30) 1294 1.14 (1.04,1.24) 1.13 (1.03, 1.23) 562 1.16 (1.04,1.29) 1.15 (1.03,1.28) 0.36 Other dementia 1671 1.05 (0.99,1.11) 1.02 (0.96,1.09) 607 0.99 (0.86,1.13) 0.96 (0.84,1.11) 1709 1.08 (1.01,1.17) 1.01 (0.93, 1.09) 703 1.04 (0.95,1.15) 1.00 (0.91, 1.10) 0.44 Phet d 0.007 0.001 0.10 0.23 0.08 0.27 0.12 0.06 ADRD 2959 1.09 (1.04,1.14) 1.08 (1.03,1.13) 1136 1.02 (0.93,1.12) 1.01 (0.91,1.12) 3123 1.10 (1.04,1.16) 1.06 (1.00, 1.12) 1265 1.08 (1.01,1.16) 1.05 (0.98, 1.13) 0.38 Open in a new tab All the phets under Model A and B for African American vs Japanese American were also > 0.05. a NO 2 10 μg/m 3 - =5.32 ppb; ADRD = AD and other dementia. b MODEL A − Cox proportional hazards regression model using air pollution levels from 1999 to exit for age scale (from entry month to exit month) and stratum by age at start Medicare and Medicare usage (in -patient, out-patient) and adjusted for race/ethnicity, sex (models in men and women combined). c MODEL B- Time-dependent Cox proportional hazards regression model using air pollution levels from 1999 to exit for age scale (from entry month to exit month) and stratum by age at start Medicare and Medicare usage (in -patient, out-patient) and adjusted for continuous age at baseline, sex, education (<=high school graduate, some college, college graduate, graduate and professional school, missing), marital status (married, separated, divorced, widowed, never married, dk), hypertension (0,1), stroke (0,1), diabetes (0,1), heart attack (0,1), smoking status (no, ex, current, dk), moderate physical activity (<2.5 hr /week, 2.5-<5, 5-<7, 7-<14, 14+, dk), sleep (0,1), BMI at baseline (<25, 25-<30, 30+, dk), low diet quality (0,1), nSES at Medicare entry (1–5, dk) and current nSES (1–5, dk). d P heterogeneity (df = 1) comparing AD vs other dementia. e P heterogeneity (df = 3) comparing HR results by race/ethnicity, all P > 0.05 for Model A (not shown). Data shown are for Model B. 3.3. Co-pollutant and multi-pollutant results Data on PM 2.5 , NO 2 and aUFP were available on 30,703 MEC participants. This aUFP subgroup included slightly more African American (28.4%) and Japanese American (15.3%) and fewer White (12.8%) participants and also showed higher cumulative average levels of PM 2.5 (16.5 ± 2.1 ug/m 3 ) and NO 2 (43.4 ± 7.7 ug/m 3 ) ( Table 1 ) than the group of participants (n=14,251) without aUFP data (PM 2.5 13.3 ± 3.4 ug/m 3 , NO 2 35.6 ± 8.4 ug/m 3 ) ( Table S1 ). In this subgroup, the HRs for ADRD associated with PM 2.5 (1.11, 95% CI 1.06–1.16) and NO 2 (1.19, 95% CI 1.12–1.26) were higher than the respective HRs for all participants but the 95% CIs were overlapping ( Table 2 ). The ADRD-NO 2 association remained statistically significant when adjusting for PM 2.5 and aUFP in co- and multi-pollutant models. Similarly, the HR estimate for aUFP remained statistically significant when adjusting for PM 2.5 and NO 2 in co-pollutant models and multi-pollutant models. However, the HR associated with PM 2.5 weakened and the 95% CI included the null when adjusting for NO 2 but remained statistically significant when adjusting for aUFP in co-pollutant models. Both aUFP and exposures were statistically significantly associated with ADRD incidence in African American and Latino participants in multipollutant models including all three pollutants ( Table 4 ). These effect estimate patterns for co-pollutant and multipollutant models were largely observed for AD and other dementia ( Table S7 ). Table 4. Association of ADRD incidence and exposure to fine particulate matter (PM 2.5 ), airport-related ultrafine particles (aUFP), and nitrogen dioxide (NO 2 ) in co-pollutant and multipollutant models, among 30,703, California Multiethnic Cohort participants, in all subjects and by race and ethnicity. ADRD All African American Japanese American Latino White Phet by race & ethnicity N 6195 2009 952 2362 872 Single pollutant a PM 2.5 (HR 95% CI) 1.11(1.06,1.16) 1.10(0.99,1.21) 1.09(0.98,1.21) 1.07(1.01,1.15) 1.19(1.08,1.31) 0.51 NO 2 (HR 95% CI) 1.19(1.12,1.26) 1.23(1.06,1.41) 1.08(0.96,1.22) 1.19(1.08,1.31) 1.23(1.08,1.40) 0.40 aUFP (HR 95% CI) 1.03(1.01,1.05) 1.04(1.01,1.07) 1.00(0.91,1.09) 1.03(1.00,1.07) 1.01(0.96,1.06) 0.69 Co-pollutant model a PM 2.5 (HR 95% CI) a 1.02(0.96,1.09) 1.00 (0.88,1.14) 1.09(0.92,1.29) 0.96(0.88,1.06) 1.15(1.00,1.31) 0.23 NO 2 (HR 95% CI) 1.16(1.07,1.26) 1.23(1.02,1.48) 1.00(0.82,1.22) 1.24(1.08,1.42) 1.07(0.89,1.30) 0.22 PM 2.5 (HR 95% CI) 1.12(1.07,1.17) 1.12(1.01,1.24) 1.09(0.98,1.21) 1.08(1.01,1.16) 1.19(1.08,1.31) 0.54 aUFP (HR 95% CI) 1.04(1.02,1.06) 1.04(1.01,1.07) 1.01(0.92,1.10) 1.04(1.00,1.07) 1.01(0.96,1.07) 0.75 NO 2 (HR 95% CI) 1.23(1.16,1.30) 1.31(1.13,1.52) 1.10(0.96,1.25) 1.23(1.11,1.35) 1.26(1.10,1.44) 0.29 aUFP (HR 95% CI) 1.05(1.03,1.07) 1.05(1.02,1.09) 1.02(0.93,1.13) 1.05(1.01,1.09) 1.03(0.98,1.09) 0.89 Multi-pollutant model a PM 2.5 (HR 95% CI) 1.01(0.95,1.07) 0.98(0.86,1.12) 1.08(0.91,1.29) 0.95(0.87,1.05) 1.13(0.98,1.30) 0.22 NO 2 (HR 95% CI) 1.22(1.12,1.33) 1.33(1.10,1.62) 1.01(0.81,1.26) 1.29(1.12,1.48) 1.10(0.90,1.35) 0.15 aUFP (HR 95% CI) 1.05(1.03,1.07) 1.05(1.02,1.09) 1.01(0.91,1.12) 1.05(1.01,1.09) 1.02(0.97,1.08) 0.70 Open in a new tab a MODEL A − Cox proportional hazards regression model using air pollution levels from 1999 to exit for age scale (from entry month to exit month) and stratum by age at start Medicare and Medicare usage (in -patient, out-patient) and adjusted for race/ethnicity, sex (models in men and women combined). 3.4. Sensitivity analysis HR results associated with PM 2.5 , NO 2 and aUFP for AD, other dementia and ADRD were largely unchanged with the addition of Medicare entry year as a linear term, suggesting little residual differences related to Medicare entry at different ages and calendar years ( Table S8 ). We found little evidence for modification of PM 2.5 , NO 2 and aUFP and ADRD incidence effect measures by each of the sociodemographic, lifestyle and cardiometabolic factors ( Table 5 ). Although there was a suggestion of higher HR for the PM 2-- ADRD association among participants who did not have the cardiometabolic conditions, this was not observed for NO 2 or aUFP. In the about one-third of study participants with genotype information, there was no modifying effects by APOE e4 carrier status ( Table S9 ). The evidence for mediation by cardiometabolic condition or lifestyle factors combined was also weak ( Tables S10 ). Table 5. Association of ADRD incidence and exposure to PM 2..5 , NO 2 and aUFP by each of the 12 established sociodemographic, lifestyle and cardiometabolic risk factors, California Multiethnic Cohort participants. PM 2.5 NO 2 aUFP ADRD risk factors b ADRD HR (95% CI) a ADRD HR (95% CI) a ADRD HR (95% CI) a Married Not 4086 1.04(1.02,1.07) 3841 1.09(1.05,1.14) 2815 1.02(0.99,1.05) yes 4771 1.03(1.01,1.05) 4527 1.08(1.04,1.13) 3292 1.04(1.01,1.07) P heterogeneity (df = 1) 0.48 0.25 0.21 Education Low 4877 1.03(1.01,1.05) 4629 1.08(1.04,1.12) 3402 1.03(1.00,1.05) High 3946 1.05(1.03,1.07) 3706 1.10(1.06,1.15) 2688 1.04(1.01,1.07) P heterogeneity (df = 1) 0.08 0.33 0.77 nSES (medicare entry) Low 6028 1.03(1.01,1.05) 5730 1.08(1.04,1.12) 4358 1.03(1.01,1.06) High 2580 1.04(1.01,1.07) 2458 1.07(1.02,1.13) 1663 1.00(0.96,1.04) P heterogeneity (df = 1) 0.29 0.91 0.15 Body mass index Obese 2102 1.03(1.00,1.06) 1969 1.07(1.01,1.14) 1428 1.01(0.98,1.05) Not Obese (< 30) 6879 1.04(1.02,1.06) 6514 1.10(1.06,1.13) 4767 1.03(1.01,1.06) P heterogeneity (df = 1) 0.50 0.51 0.64 Current smoking yes 1172 1.05(1.01,1.10) 1101 1.12(1.03,1.21) 825 1.03(0.97,1.09) No 7554 1.04(1.02,1.05) 7139 1.09(1.05,1.12) 5188 1.03(1.01,1.05) P heterogeneity (df = 1) 0.99 0.87 0.35 Sleep Short/long 2547 1.05(1.02,1.08) 2392 1.10(1.04,1.16) 1731 1.02(0.98,1.05) Normal 6434 1.04(1.02,1.06) 6091 1.09(1.05,1.13) 4464 1.04(1.01,1.06) P heterogeneity (df = 1) 0.58 0.67 0.38 Physical activity Low 4628 1.02(1.00,1.05) 4376 1.06(1.01,1.10) 3225 1.04(1.01,1.06) High 4353 1.05(1.03,1.08) 4107 1.12(1.08,1.17) 2970 1.02(0.99,1.05) P heterogeneity (df = 1) 0.10 0.06 0.48 Diet quality Low (< 60) 1750 1.04(1.00,1.08) 1668 1.07(1.00,1.14) 1229 1.03(0.98,1.07) High 6833 1.04(1.02,1.06) 6436 1.10(1.06,1.13) 4682 1.03(1.01,1.05) P heterogeneity (df = 1) 0.69 0.26 0.80 Heart attack Yes 1106 1.01(0.96,1.05) 1040 1.01(0.93,1.09) 716 1.04(0.99,1.10) No 7875 1.04(1.03,1.06) 7443 1.10(1.07,1.14) 5479 1.03(1.01,1.05) P heterogeneity (df = 1) 0.09 0.02 0.39 Hypertension Yes 4369 1.03(1.04,1.05) 4121 1.08(1.04,1.13) 2957 1.04(1.01,1.07) No 4612 1.05(1.03,1.07) 4362 1.10(1.06,1.15) 3238 1.02(1.00,1.05) P heterogeneity (df = 1) 0.16 0.73 0.47 Diabetes Yes 1508 1.03(0.99,1.07) 1425 1.11(1.04,1.19) 1046 1.00(0.95,1.04) No 7473 1.04(1.02,1.06) 7058 1.09(1.05,1.12) 5149 1.04(1.01,1.06) P heterogeneity (df = 1) 0.80 0.26 0.09 Stroke Yes 335 1.08(1.00,1.17) 317 1.16(1.00,1.34) 213 1.00(0.90,1.11) No 8646 1.04(1.02,1.05) 8166 1.09(1.06,1.12) 5982 1.03(1.01,1.05) P heterogeneity (df = 1) 0.25 0.37 0.49 Cardiometabolic risk factors (heart, hypertension, diabetes, stroke) Yes 5301 1.02(1.00,1.04) 5001 1.07(1.03,1.11) 3590 1.03(1.01,1.06) No 3680 1.06(1.04,1.09) 3482 1.12(1.07,1.18) 2605 1.02(0.99,1.05) P heterogeneity (df = 1) 0.03 0.23 0.62 12 risk factors combined score 0, 1, 2 factors 2335 1.04(1.01,1.07) 2204 1.09(1.03,1.15) 1574 1.02(0.98,1.07) 3, 4 factors 3421 1.03(1.01,1.06) 3228 1.09(1.04,1.14) 2377 1.02(0.99,1.05) 5 + factors 3225 1.01(0.98,1.04) 3051 1.04(0.99,1.09) 2244 1.03(0.99,1.06) P heterogeneity (df = 2) 0.31 0.36 0.88 Open in a new tab a MODEL A – Cox proportional hazards regression model using air pollution levels from 1999 to exit for age scale (from entry month to exit month) and stratum by age at start Medicare and Medicare usage (in = -patient, out-patient) and adjusted for continuous age at baseline, race/ethnicity, sex (models in men and women combined). b Except for nSES at Medicare entry, all the other variables are from baseline questionnaire and defined as follows: marital status (married vs not, included separated, divorced, widowed, never married), education (low, <=high school graduate vs high, some college, college graduate, graduate and professional), nSES (neighborhood SES) at Medicare entry (low, Q1-Q3 vs high, Q4-Q5), BMI (obese, 30 + kg/m2 vs not obese, <30 kg/m2); current smoking (current smoker vs former + never smoker), sleep (short or long < 6 or > 8 hrs/day, vs normal 6–8 hrs/day); diet quality (HEI 2005 index < 60 (low) vs high, ≥60); cardiometabolic conditions (no to all 4 conditions, diabetes, hypertension, heart attack or stroke vs yes to any of the conditions). For the combined index of 12 risk factors, we assigned a score of 1 to each ADRD high-risk category of these 12 factors. 4. Discussion In this large prospective study of California MEC participants (73% African American and Latino adults) enrolled in the FFS component of Medicare, incidence of ADRD was found to increase with traffic- and airport-related air pollutants, including particulate matter (PM 2.5 , PM 10, aUFP) and gaseous pollutants (NO 2 , NOx, and CO) in single pollutant models. The correlation between PM 2.5 and NO 2 was high (0.89), and the effect estimate for PM 2.5 moved towards the null when both were included in a co-pollutant model. ADRD incidence increased statistically significantly with PM 2.5 and aUFP in co-pollutant models, and both NO 2 and aUFP were associated with ADRD incidence in a multi-pollutant model adjusting for PM 2.5 . Risk patterns were similar across race and ethnicity, but African American participants showed more consistently elevated HRs for ADRD with all pollutants (except ozone). These results add to the literature, representing one of the largest prospective analyses of long-term air pollution and incidence of ADRD in African American and Latino populations using individual-level geocoded address-based exposure assessment rather than zip-code level data. Dementia incidence increased significantly 8% and 3% with exposures to PM 2.5 and NO 2 in a 2025 meta-analysis and HRs were comparable for AD and vascular dementia although the latter result, based on fewer studies, was not statistically significant. ( Best Rogowski et al., Jul 2025 ) As this and other recent meta-analyses concurred that the PM 2.5 -dementia association was stronger in studies that used active ascertainment of dementia diagnosis, ( Wilker et al., 2023 ; Best Rogowski et al., Jul 2025 ; Huang et al., 2025 ) the smaller HR estimate associated with PM 2.5 among MEC participants is likely an underestimate. Two recent studies not included in the 2025 meta-analysis are unlikely to change the overall result. ( Best Rogowski et al., Jul 2025 ) In the Atherosclerosis Risk in Communities (ARIC) study, PM 2.5 and NO 2 exposures were not associated with incident dementia (719 African American, 1,792 White) after adjustment for education and other factors. ( Song et al., 2024 ) In contrast, in a large Danish nationwide administrative cohort, dementia incidence increased strongly by 14% and 25%, respectively, in relation to PM 2.5 or NO 2 exposures. ( Andersen et al., Aug 2025 ) The consistent associations between air pollution and ADRD incidence among African American participants in the MEC add to the sparse literature in this racial group within a cohort of a robust sample size and carefully accounted for established risk factors. These associations may reflect that African American participants in the MEC had the highest prevalence of 8 of the 12 ADRD risk factors (low nSES, not married, history of hypertension, heart attack and stroke, current smoking, short/long sleep and obese) and almost 2-times higher levels of exposure to aUFP compared to other race/ethnic groups ( Table 1 ). The higher prevalence of established risk factors and exposure to other unmeasured toxicants, may explain their longstanding higher incidence of dementia. ( Eiser, Oct 2024 ; Backhouse et al., 2024 ) Published results on African American participants are mixed. In the Women’s Health Initiative Memory Study (WHIMS), PM 2.5 exposure was associated with a HR of 2.6 for AD for African American compared to HR of 1.14 in White women but results were based on 19 and 111 AD cases, respectively. ( Younan et al., 2021 ) A large Medicare data study also found stronger associations of PM 2.5 and NO 2 and ADRD incidence in African American than White adults, ( Shi et al., 2021 ) but this study used zip code-level exposure assignment, and lacked geocoded address locations and individual-level SES and other covariates. However, the Environmental Predictors of Cognitive Health and Aging (EPOCH) cohort reported no significant racial/ethnic differences in associations between incident dementia and total PM 2.5 mass. (Zhang et al., 2023) Our finding on air pollution and ADRD incidence among Latino adults, based on nearly 3000 events, adds substantially to the existing data in Latinos. ( Yu et al., Mar 2023 ; Paul et al., 2020 ) Our findings linking airport-related UFP and incidence of ADRD risk are supportive of a recent nationwide Medicare-based study which investigated the role of UFP using zip-code level data from a single year. ( Zhu et al., 2025 ) However, our results differed from a Seattle-based study reporting no association between dementia incidence and exposure to UFP and NO 2 ( Zhu et al., 2025 ) but a positive association with PM 2.5 exposure. ( Shaffer et al., Aug 2021 ) The mean UFP (10,537 particle/cm 3 ) levels in this Seattle study were higher than UFP levels among African American subjects in the MEC (9,881 particle/cm 3 ), but their levels of NO 2 (9.9 ppb) and PM 2.5 (10.1 μg/cm 3 ) were lower than the respective levels (15.2 ppb and 15.6 μg/cm 3 ) in Los Angeles County. Another Seattle study found that air pollution levels differed by income, race and ethnicity; this was particularly marked for UFP levels which were also highest among African American residents. ( Zhu et al., 2025 ; Bramble et al., Jul 2023 ) UFPs may impact brain health through unique mechanisms. Their small sizes, high concentrations in ambient environments, large surface area per mass relative to their size, and long retention in the lung enable them to penetrate extra-pulmonary organs and cross the blood brain barrier. There is increasing evidence that UFPs can impact the brain directly via the olfactory bulb, impairing mitochondrial functions and eliciting their toxic effects. ( Mussalo et al., Sep 2024 ) Hussian et al hypothesized that inhaled PM, particularly UFP, may impair glymphatic clearance, and that glymphatic transport may facilitate the spread of UFPs and their adverse effects to different brain regions. ( Hussain et al., Nov 2023 ) Airport-related UFP has been associated with risk of malignant brain tumors in a Canadian study ( Weichenthal et al., Mar 2020 ) and among MEC participants in Los Angeles County; these latter results were also stronger among African American subjects. ( Wu et al., 2021 ) In a cross-sectional study, long-term exposure to UFP was associated with a 3.6% increase in high sensitivity C-reactive protein, which increased to 6.3% when PM 2.5 was adjusted for in two-pollutant models. ( Pilz et al., Apr 2018 ) Thus, high exposures to UFP may increase risk of ADRD via both brain specific effects such as reducing the efficiency of glymphatic clearance of waste metabolites and other systemic inflammatory pathways. Air pollution is a heterogenous mixture with highly correlated components and common sources, making it challenging to identify the effects of individual pollutants or components on ADRD incidence. We investigated the effects of PM 2.5 , NO 2 , and aUFP in co-pollutant and multipollutant models among 30,703 MEC participants with data on all three pollutants ( Table 4 ). Our findings of statistically significant associations for both PM 2.5 and aUFP in co-pollutant models are supportive of findings in the nationwide Medicare cohort, ( Zhu et al., 2025 ) and add new information on mutually adjusted effects of NO 2 and aUFP. The weakened PM 2.5 effect estimate when adjusting for NO 2 among MEC participants differed from what has been reported in previous studies. In Southern California, both PM 2.5 and NO 2 represent traffic-related air pollutants, and the weakened effect estimate for PM 2.5 may reflect that our measure of PM 2.5 -derived from a chemical transport model captures contributions from traffic less well than the NO 2 measure. Nevertheless, we presume that the traffic exhaust components of PM 2.5 and NO 2 are important for the development of dementia. The relative importance of different particle sources contributing to the PM 2.5 mixture and different pollutant constituents in different geographic areas may have contributed to the divergent co-pollutant results for PM 2.5 and NO 2 in published studies. ( Yan et al., 2022 ; Shi et al., 2023 ; Andersen et al., Aug 2025 ) We did not observe consistent evidence of a modifying effect of the air pollution-ADRD association by any of the 12 established sociodemographic, lifestyle, and cardiometabolic risk factors. Although we observed stronger associations s of PM 2.5 for ADRD incidence among those without a history of cardiometabolic risk factors than those with a history, this was not observed in relation to NO 2 and aUFP exposures. Our information on risk factors was based on self-reports at cohort entry, about 10 years before Medicare entry. It is very likely that the number of heart attacks increase with increasing age and there is misclassificaton for this covariate.. Published results on potential modifying effect of comorbidities on air pollution-ADRD associations are mixed and difficult to interpret because the specific chronic conditions that were investigated differed. In some studies, modifying effects were found with one pollutant but not another pollutant, or the results related to ADRD, or only AD, or only other dementia. Since information on medications or treatment of these comorbidities were generally not available, differences in findings on modifying effects of comorbidities may be related to severity of these conditions and other unmeasured factors. ( Chen et al., Nov 2017 ; Ran et al., 2021 ; Younan et al., 2021 ; Zhang et al., 2023; Zhu et al., 2025 ) We did not find any differences in risk association by APOE e4 carrier status, consistent with other reports. ( Wang et al., 2022 ; Mortamais et al., Mar 2021 ; Semmens et al., Feb 2023 ; Parra et al., Jun 2022 ) Air pollution exposure has been estimated to be responsible for 16% of mortality worldwide, adversely affecting brain health via oxidative stress, neuroinflammation, microglial activation and vascular damage. In animal models, exposure to fine particulate matter has been associated with brain structural atrophy in a manner hypothesized to increase the risk of ADRD. ( Haghani et al., 2020 ; Woodward et al., May 2017 ) Using individual-level estimates of ambient concentrations of PM 2.5 and NO 2 to assess improvement in air quality over the span of 10 years in the WHIMS cohort, Wang and colleagues reported that reductions in PM 2.5 and NO 2 levels were associated with significantly 20% reduced risk of dementia. ( Wang et al., 2022 ) These results highlight the importance of air pollution reduction to lower ADRD incidence on a population level. Strengths of this study include the prospective design, large numbers of male and female participants, a long-term follow-up, racial, ethnic, and socioeconomic diversity, and consideration of 12 ADRD risk factors as well as APOE genotype, as adjustment variables and in stratified analyses exploring effect modification. Our study also benefited from having updated geocoded residential history since cohort entry and thus time-dependent, high-resolution air pollution measures with satisfactory model performance ( Table S11 ). Yet, there are some study limitations. We did not have information on air pollution during travel, work, or time indoors and thus exposure misclassification cannot be ruled out. High aUFP exposure is likely to coincide with elevated aircraft noise, and may introduce confounding bias although this is not likely to be a serious concern based on another study we conducted in Southern California. ( Wing et al., Apr 2020 ) African American participants are the subgroup with the highest aUFP exposures across follow-up and the highest concentration of residents living within the modeled UFP exposure grid, ( Bookstein et al., 2024 ; Wu et al., 2021 ) which may have contributed to the stronger associations. The ADRD case definition was based solely on Medicare claims data, which may not have fully captured incident cases. Our analysis was restricted to Medicare FFS beneficiaries, representing approximately ~60% of the original MEC participants, and while our results may not be generalizable to the entire MEC cohort, our study subgroup was large and had ample racial/ethnic and socioeconomic diversity. Further, the under-detection of cases might occur more often for those with mild symptoms than for severe dementia. The stronger association between air pollution and dementia incidence, for studies that used active case ascertainment instead of electronic health records suggests a higher rate of misclassification of dementia status with the latter. ( Wilker et al., 2023 ) However, this would mean that our HR estimates are likely underestimations of the true effect sizes for these air pollutants. In addition, our results only apply to those with FFS Medicare claims and we lacked information on individuals with earlier onset ADRD or those without Medicare coverage. Finally, despite the large sample size, statistical power might still be limited for some of the subgroups, and information on APOE genotype was available on only a subset of MEC participants. In summary, long-term air pollution exposure is associated with ADRD incidence, specifically AD and possibly dementia in this multiethnic cohort of older adults. As air pollution is a complex mixture of highly correlated particulate matter, gaseous pollutants and air toxics from numerous sources, it is challenging to isolate the effects for each individual pollutant and we interpret our findings to reflect an impact of traffic- and airport-related air pollution. These findings were largely consistent across race and ethnicity, but subgroup analyses suggest that ambient air pollutants may have stronger estimated effects among African American and Latino adults. Our results suggest further investigations into the role of UFP and other pollutants, including refined measures of ADRD diagnoses, and of younger adults are still needed. Supplementary Material 1 NIHMS2159068-supplement-1.docx (118.7KB, docx) Study Funding This work was supported by grants from the National Cancer Institute, a supplement (CA 164973 08S1) from the National Institute on Aging at the NIH to the Multiethnic Cohort (U01 CA 164973) and a pilot grant from Center for Advancing Sociodemographic and Economic Study of Alzheimer’s Disease and Related Dementias (CeASES ADRD) funded by the National Institute on Aging (P30AG066589) Glossary AD Alzheimer’s disease ADRD Alzheimer’s disease and related dementia APOE apolipoprotein E aUFP airport related ultrafine particles BMI body mass index CO carbon monoxide CMS Centers for Medicare & Medicaid Services EPA Environmental Protection Agency FFS fee for service HR hazard ratio HRS Health Retirement Study IRB Institutional Review Board IQR interquartile range LAX Los Angeles International Airport MEC Multiethnic Cohort Study nSES neighborhood socioeconomic status NO 2 nitrogen dioxide NO nitrogen oxide O 3 ozone PM 2.5 fine particulate matter PPB parts per billion UK United Kingdom UH University of Hawaii USC University of Southern California WHIMS Women’s Health Initiative Memory Study Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.envint.2026.110169 . Footnotes CRediT authorship contribution statement Anna H. Wu: Writing – original draft, Supervision, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Conceptualization. Jun Wu: Writing – review & editing, Resources, Funding acquisition, Formal analysis. Chiuchen Tseng: Writing – review & editing, Investigation, Formal analysis. Burcu F. Darst: Writing – review & editing, Methodology, Investigation. Song-Yi Park: Writing – review & editing, Methodology, Investigation. Daniel O. Stram: Writing – review & editing, Methodology, Funding acquisition, Formal analysis. Timothy Larson: Writing – review & editing, Methodology, Data curation. Scott Fruin: Writing – review & editing, Methodology, Data curation. Veronica W. Setiawan: Writing – review & editing, Resources, Data curation. Xian Yu: Writing – review & editing, Resources, Data curation. Lynne R. Wilkens: Writing – review & editing, Resources, Project administration, Methodology, Investigation. Howard Hu: Writing – review & editing, Methodology, Investigation. Christopher Haiman: Writing – review & editing, Resources, Project administration, Methodology, Investigation. Beate Ritz: Writing – review & editing, Methodology, Investigation. Eileen M. Crimmins: Writing – review & editing, Methodology, Investigation. Unhee Lim: Writing – review & editing, Validation, Resources, Methodology, Investigation, Data curation. Iona Cheng: Writing – review & editing, Resources, Investigation, Funding acquisition, Data curation. Loïc Le Marchand: Writing – review & editing, Resources, Project administration, Methodology, Investigation. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Data availability The ADRD outcome data were obtained under a data use agreement with CMS that explicitly prohibits data sharing. The MEC baseline data are shared in dbGaP (accession number: phs002183.v1.p1). References Andersen ZJ, Lim YH, Zhang J, et al. , Aug 2025. Long-term exposure to air pollution and risk of dementia among older individuals of a Danish nationwide administrative cohort. Environ Int. 202, 109607. 10.1016/j.envint.2025.109607. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Andersson J, Sundstrom A, Nordin M, et al. , 2023. PM2.5 and Dementia in a Low Exposure Setting: The Influence of Odor Identification Ability and APOE. J Alzheimers Dis. 92 (2), 679–689. 10.3233/JAD-220469. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Backhouse EV, Bauermeister S, Wardlaw JM, 2024. Lifetime influences on imaging markers of adverse brain health and vascular disease. Cereb Circ Cogn Behav. 6, 100194. 10.1016/j.cccb.2023.100194. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Bagheri N, Mavoa S, Tabatabaei-Jafari H, et al. , 2021. The Impact of Built and Social Environmental Characteristics on Diagnosed and Estimated Future Risk of Dementia. J Alzheimers Dis. 84 (2), 621–632. 10.3233/JAD-210208. [ DOI ] [ PubMed ] [ Google Scholar ] Best Rogowski CB, Bredell C, Shi Y, et al. , Jul 2025. Long-term air pollution exposure and incident dementia: a systematic review and meta-analysis. Lancet Planet Health. 9 (7), 101266. 10.1016/S2542-5196(25)00118-4. [ DOI ] [ PubMed ] [ Google Scholar ] Bookstein A, Po J, Tseng C, et al. , 2024. Association between Airport Ultrafine Particles and Lung Cancer Risk: The Multiethnic Cohort Study. Cancer Epidemiol Biomarkers Prev. 33 (5), 703–711. 10.1158/1055-9965.EPI-23-0924. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Bramble K, Blanco MN, Doubleday A, et al. , Jul 2023. Exposure Disparities by Income, Race and Ethnicity, and Historic Redlining Grade in the Greater Seattle Area for Ultrafine Particles and Other Air Pollutants. Environ Health Perspect. 131 (7), 77004. 10.1289/EHP11662. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Calderon-Garciduenas L, Solt AC, Henriquez-Roldan C, et al. , Feb 2008. Long-term air pollution exposure is associated with neuroinflammation, an altered innate immune response, disruption of the blood-brain barrier, ultrafine particulate deposition, and accumulation of amyloid beta-42 and alpha-synuclein in children and young adults. Toxicol Pathol. 36 (2), 289–310. 10.1177/0192623307313011. [ DOI ] [ PubMed ] [ Google Scholar ] Carey IM, Anderson HR, Atkinson RW, et al. Are noise and air pollution related to the incidence of dementia? A cohort study in London, England. BMJ Open. Sep 11 2018;8 (9):e022404. doi: 10.1136/bmjopen-2018-022404. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Chen H, Kwong JC, Copes R, et al. , Nov 2017. Exposure to ambient air pollution and the incidence of dementia: A population-based cohort study. Environ Int. 108, 271–277. 10.1016/j.envint.2017.08.020. [ DOI ] [ PubMed ] [ Google Scholar ] Cheng I, Tseng C, Wu J, et al. , 2020. Association between ambient air pollution and breast cancer risk: The multiethnic cohort study. Int J Cancer. 146 (3), 699–711. 10.1002/ijc.32308. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Cheng I, Yang J, Tseng C, et al. , 2022. Traffic-Related Air Pollution and Lung Cancer Incidence: The California Multiethnic Cohort Study. Am J Respir Crit Care Med. Jun 1. 10.1164/rccm.202107-1770OC. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Collaborators GBDDF. Estimation of the global prevalence of dementia in 2019 and forecasted prevalence in 2050: an analysis for the Global Burden of Disease Study 2019. Lancet Public Health. Feb 2022;7(2):e105–e125. doi: 10.1016/S2468-2667(21)00249-8. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] de Crom TOE, Ginos BNR, Oudin A, Ikram MK, Voortman T, Ikram MA, 2023. Air Pollution and the Risk of Dementia: The Rotterdam Study. J Alzheimers Dis. 91 (2), 603–613. 10.3233/JAD-220804. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Eiser AR, Oct 2024. Environmental, Metabolic, and Nutritional Factors Concerning Dementia in African American and Hispanic American Populations. Am J Med. 137 (10), 939–942. 10.1016/j.amjmed.2024.06.023. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Fu P, Yung KKL, 2020. Air Pollution and Alzheimer’s Disease: A Systematic Review and Meta-Analysis. J Alzheimers Dis. 77 (2), 701–714. 10.3233/JAD-200483. [ DOI ] [ PubMed ] [ Google Scholar ] Goodman RA, Lochner KA, Thambisetty M, Wingo TS, Posner SF, Ling SM, Jan 2017. Prevalence of dementia subtypes in United States Medicare fee-for-service beneficiaries, 2011–2013. Alzheimers Dement. 13 (1), 28–37. 10.1016/j.jalz.2016.04.002. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Grande G, Ljungman PLS, Eneroth K, Bellander T, Rizzuto D, 2020. Association Between Cardiovascular Disease and Long-term Exposure to Air Pollution With the Risk of Dementia. JAMA Neurol. 77 (7), 801–809. 10.1001/jamaneurol.2019.4914. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Haghani A, Morgan TE, Forman HJ, Finch CE, 2020. Air Pollution Neurotoxicity in the Adult Brain: Emerging Concepts from Experimental Findings. J Alzheimers Dis. 76 (3), 773–797. 10.3233/JAD-200377. [ DOI ] [ PubMed ] [ Google Scholar ] Hajat A, Hsia C, O’Neill MS, Dec 2015. Socioeconomic Disparities and Air Pollution Exposure: a Global Review. Curr Environ Health Rep. 2 (4), 440–450. 10.1007/s40572-015-0069-5. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Huang X, Steinmetz J, Marsh EK, et al. A systematic review with a Burden of Proof meta-analysis of health effects of long-term ambient fine particulate matter (PM(2.5)) exposure on dementia. Nat Aging. May 2025;5(5):897–908. doi: 10.1038/s43587-025-00844-y. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Hudda N, Gould T, Hartin K, Larson TV, Fruin SA, 2014. Emissions from an international airport increase particle number concentrations 4-fold at 10 km downwind. Environ Sci Technol. 48 (12), 6628–6635. 10.1021/es5001566. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Hussain R, Graham U, Elder A, Nedergaard M, Nov 2023. Air pollution, glymphatic impairment, and Alzheimer’s disease. Trends Neurosci. 46 (11), 901–911. 10.1016/j.tins.2023.08.010. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Ilango SD, Chen H, Hystad P, et al. , 2020. The role of cardiovascular disease in the relationship between air pollution and incident dementia: a population-based cohort study. Int J Epidemiol. 49 (1), 36–44. 10.1093/ije/dyz154. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Ilango SD, Leary CS, Ritchie E, et al. , Jun 2023. An Examination of the Joint Effect of the Social Environment and Air Pollution on Dementia Among US Older Adults. Environ Epidemiol. 7 (3), e250. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Kolonel LN, Henderson BE, Hankin JH, et al. , 2000. A multiethnic cohort in Hawaii and Los Angeles: baseline characteristics. Am J Epidemiol. 151 (4), 346–357. 10.1093/oxfordjournals.aje.a010213. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Lim U, Wang S, Park SY, et al. Risk of Alzheimer’s disease and related dementia by sex and race/ethnicity: The Multiethnic Cohort Study. Alzheimers Dement. Dec 9 2021; doi: 10.1002/alz.12528. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Livingston G, Huntley J, Sommerlad A, et al. , 2020. Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Lancet. 396 (10248), 413–446. 10.1016/S0140-6736(20)30367-6. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Lunn M, McNeil D, Jun 1995. Applying Cox regression to competing risks. Biometrics. 51 (2), 524–532. [ PubMed ] [ Google Scholar ] Meng J, Li C, Martin RV, van Donkelaar A, Hystad P, Brauer M, 2019. Estimated Long-Term (1981–2016) Concentrations of Ambient Fine Particulate Matter across North America from Chemical Transport Modeling, Satellite Remote Sensing, and Ground-Based Measurements. Environ Sci Technol. 53 (9), 5071–5079. 10.1021/acs.est.8b06875. [ DOI ] [ PubMed ] [ Google Scholar ] Mortamais M, Gutierrez LA, de Hoogh K, et al. , Mar 2021. Long-term exposure to ambient air pollution and risk of dementia: Results of the prospective Three-City Study. Environ Int. 148, 106376. 10.1016/j.envint.2020.106376. [ DOI ] [ PubMed ] [ Google Scholar ] Mussalo L, Lampinen R, Avesani S, et al. , Sep 2024. Traffic-related ultrafine particles impair mitochondrial functions in human olfactory mucosa cells - Implications for Alzheimer’s disease. Redox Biol. 75, 103272. 10.1016/j.redox.2024.103272. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Oberdorster G, Sharp Z, Atudorei V, et al. , Jun 2004. Translocation of inhaled ultrafine particles to the brain. Inhal Toxicol. 16 (6–7), 437–445. 10.1080/08958370490439597. [ DOI ] [ PubMed ] [ Google Scholar ] Park SY, Setiawan VW, White LR, et al. Modifying effects of race and ethnicity and APOE on the association of physical activity with risk of Alzheimer’s disease and related dementias. Alzheimers Dement. Apr 27 2022;doi: 10.1002/alz.12677. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Park SY, Setiawan VW, Crimmins EM, et al. Racial and Ethnic Differences in the Population-Attributable Fractions of Alzheimer Disease and Related Dementias. Neurology. Feb 13 2024;102(3):e208116. doi: 10.1212/WNL.0000000000208116. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Parra KL, Alexander GE, Raichlen DA, Klimentidis YC, Furlong MA, Jun 2022. Exposure to air pollution and risk of incident dementia in the UK Biobank. Environ Res. 209, 112895. 10.1016/j.envres.2022.112895. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Paul KC, Haan M, Yu Y, et al. , 2020. Traffic-Related Air Pollution and Incident Dementia: Direct and Indirect Pathways Through Metabolic Dysfunction. J Alzheimers Dis. 76 (4), 1477–1491. 10.3233/JAD-200320. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Pilz V, Wolf K, Breitner S, et al. , Apr 2018. C-reactive protein (CRP) and long-term air pollution with a focus on ultrafine particles. Int J Hyg Environ Health. 221 (3), 510–518. 10.1016/j.ijheh.2018.01.016. [ DOI ] [ PubMed ] [ Google Scholar ] Ran J, Schooling CM, Han L, et al. , 2021. Long-term exposure to fine particulate matter and dementia incidence: A cohort study in Hong Kong. Environ Pollut. Feb 15 271, 116303. 10.1016/j.envpol.2020.116303. [ DOI ] [ PubMed ] [ Google Scholar ] Semmens EO, Leary CS, Fitzpatrick AL, et al. , Feb 2023. Air pollution and dementia in older adults in the Ginkgo Evaluation of Memory Study. Alzheimers Dement. 19 (2), 549–559. 10.1002/alz.12654. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Setiawan VW, Virnig BA, Porcel J, et al. , 2015. Linking data from the Multiethnic Cohort Study to Medicare data: linkage results and application to chronic disease research. Am J Epidemiol. 181 (11), 917–919. 10.1093/aje/kwv055. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Shaffer RM, Blanco MN, Li G, et al. , Aug 2021. Fine Particulate Matter and Dementia Incidence in the Adult Changes in Thought Study. Environ Health Perspect. 129 (8), 87001. 10.1289/EHP9018. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Shi L, Steenland K, Li H, et al. A national cohort study (2000–2018) of long-term air pollution exposure and incident dementia in older adults in the United States. Nat Commun. Nov 19 2021;12(1):6754. doi: 10.1038/s41467-021-27049-2. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Shi L, Zhu Q, Wang Y, et al. , 2023. Incident dementia and long-term exposure to constituents of fine particle air pollution: A national cohort study in the United States. Proc Natl Acad Sci U S A. Jan 3 120 (1), e2211282119. 10.1073/pnas.2211282119. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Smargiassi A, Sidi EAL, Robert LE, et al. Exposure to ambient air pollutants and the onset of dementia in Quebec, Canada. Environ Res. Nov 2020;190:109870. doi: 10.1016/j.envres.2020.109870. [ DOI ] [ PubMed ] [ Google Scholar ] Song Z, Lynch K, Parker-Allotey NA, et al. , 2024. Association of midlife air pollution exposures and residential road proximity with incident dementia: The Atherosclerosis Risk in Communities (ARIC) study. Environ Res. Oct 1 258, 119425. 10.1016/j.envres.2024.119425. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Sullivan KJ, Ran X, Wu F, et al. , Aug 2021. Ambient fine particulate matter exposure and incident mild cognitive impairment and dementia. J Am Geriatr Soc. 69 (8), 2185–2194. 10.1111/jgs.17188. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Taylor DH Jr., Fillenbaum GG, Ezell ME, Sep 2002. The accuracy of medicare claims data in identifying Alzheimer’s disease. J Clin Epidemiol. 55 (9), 929–937. 10.1016/s0895-4356(02)00452-3. [ DOI ] [ PubMed ] [ Google Scholar ] Trevenen ML, Heyworth J, Almeida OP, et al. , Dec 2022. Ambient air pollution and risk of incident dementia in older men living in a region with relatively low concentrations of pollutants: The Health in Men Study. Environ Res. 215 (Pt 2), 114349. 10.1016/j.envres.2022.114349. [ DOI ] [ PubMed ] [ Google Scholar ] Tsai TL, Lin YT, Hwang BF, et al. , Oct 2019. Fine particulate matter is a potential determinant of Alzheimer’s disease: A systemic review and meta-analysis. Environ Res. 177, 108638. 10.1016/j.envres.2019.108638. [ DOI ] [ PubMed ] [ Google Scholar ] Wang X, Younan D, Millstein J, et al. , 2022. Association of improved air quality with lower dementia risk in older women. Proc Natl Acad Sci U S A. Jan 11 119(2)doi: 10.1073/pnas.2107833119. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Weichenthal S, Olaniyan T, Christidis T, et al. , Mar 2020. Within-city Spatial Variations in Ambient Ultrafine Particle Concentrations and Incident Brain Tumors in Adults. Epidemiology. 31 (2), 177–183. 10.1097/EDE.0000000000001137. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Wilker EH, Osman M, Weisskopf MG, 2023. Ambient air pollution and clinical dementia: systematic review and meta-analysis. BMJ. Apr 5 381, e071620. 10.1136/bmj-2022-071620. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Wing SE, Larson TV, Hudda N, Boonyarattaphan S, Fruin S, Ritz B, Apr 2020. Preterm Birth among Infants Exposed to in Utero Ultrafine Particles from Aircraft Emissions. Environ Health Perspect. 128 (4), 47002. 10.1289/EHP5732. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Woo B, Kravitz-Wirtz N, Sass V, Crowder K, Teixeira S, Takeuchi DT, Mar 2019. Residential Segregation and Racial/Ethnic Disparities in Ambient Air Pollution. Race Soc Probl. 11 (1), 60–67. 10.1007/s12552-018-9254-0. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Wood D, Evangelopoulos D, Beevers S, Kitwiroon N, Demakakos P, Katsouyanni K. Exposure to ambient air pollution and cognitive function: an analysis of the English Longitudinal Study of Ageing cohort. Environ Health. Apr 5 2024;23(1):35. doi: 10.1186/s12940-024-01075-1. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Woodward NC, Pakbin P, Saffari A, et al. , May 2017. Traffic-related air pollution impact on mouse brain accelerates myelin and neuritic aging changes with specificity for CA1 neurons. Neurobiol Aging. 53, 48–58. 10.1016/j.neurobiolaging.2017.01.007. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Wu AH, Fruin S, Larson TV, et al. , 2021. Association between Airport-Related Ultrafine Particles and Risk of Malignant Brain Cancer: A Multiethnic Cohort Study. Cancer Res. 81 (16), 4360–4369. 10.1158/0008-5472.CAN-21-1138. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Wu AH, Wu J, Tseng C, et al. , 2025. Air Pollution and Breast Cancer Incidence in the Multiethnic Cohort Study. J Clin Oncol. 43 (3), 273–284. 10.1200/JCO.24.00418. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Yan YH, Chen TB, Yang CP, et al. Long-term exposure to particulate matter was associated with increased dementia risk using both traditional approaches and novel machine learning methods. Sci Rep. Oct 12 2022;12(1):17130. doi: 10.1038/s41598-022-22100-8. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Yang J, Schupp CW, Harrati A, Ca C, Keegan THM, Gomez SL, 2014. Developing an area-based socioeconomic measure from American Community Survey data. Cancer Prevention Institute of California, Fremont, Calfornia. [ Google Scholar ] Yost K, Perkins C, Cohen R, Morris C, Wright W, Oct 2001. Socioeconomic status and breast cancer incidence in California for different race/ethnic groups. Cancer Causes Control. 12 (8), 703–711. 10.1023/a:1011240019516. [ DOI ] [ PubMed ] [ Google Scholar ] Younan D, Wang X, Gruenewald T, et al. Racial/Ethnic Disparities in Alzheimer’s Disease Risk: Role of Exposure to Ambient Fine Particles. J Gerontol A Biol Sci Med Sci. Aug 12 2021;doi: 10.1093/gerona/glab231. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Yu Y, Su J, Jerrett M, et al. , Mar 2023. Air pollution and traffic noise interact to affect cognitive health in older Mexican Americans. Environ Int. 173, 107810. 10.1016/j.envint.2023.107810. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Yuchi W, Sbihi H, Davies H, Tamburic L, Brauer M, 2020. Road proximity, air pollution, noise, green space and neurologic disease incidence: a population-based cohort study. Environ Health. Jan 21 19(1):8. 10.1186/s12940-020-0565-4. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Zhang B, Langa KM, Weuve J, et al. Hypertension and Stroke as Mediators of Air Pollution Exposure and Incident Dementia. JAMA Netw Open. Sep 5 2023;6(9): e2333470. doi: 10.1001/jamanetworkopen.2023.33470. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Zhang B, Weuve J, Langa KM, et al. , 2023. Comparison of Particulate Air Pollution From Different Emission Sources and Incident Dementia in the US. JAMA Intern Med. 183 (10), 1080–1089. 10.1001/jamainternmed.2023.3300. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Zhang Z, Chen L, Wang X, et al. Associations of Air Pollution and Genetic Risk with Incident Dementia: A Prospective Cohort Study. Am J Epidemiol. Oct 21 2022;doi: 10.1093/aje/kwac188. [ DOI ] [ PubMed ] [ Google Scholar ] Zhu Q, Deng YL, Liu Y, Steenland K, 2025. Associations between Ultrafine Particles and Incident Dementia in Older Adults. Environ Sci Technol. 59 (11), 5443–5451. 10.1021/acs.est.4c10574. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials 1 NIHMS2159068-supplement-1.docx (118.7KB, docx) Data Availability Statement The ADRD outcome data were obtained under a data use agreement with CMS that explicitly prohibits data sharing. The MEC baseline data are shared in dbGaP (accession number: phs002183.v1.p1). ACTIONS View on publisher site PDF (466.0 KB) Cite Collections Permalink RESOURCES Similar articles Cited by other articles Links to NCBI Databases Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top

Related documents

• Environmental and Health Impacts of Air Pollution: A Review#7469
• Delayed associations between air pollution and population health across the life course#429594
• Prevalence and Risk Factors of Dementia Among Older People of Gelao Ethnicity in China: A National Cross-Sectional Study.#174431
• Hyper-localized measures of air pollution and risk of preterm birth in Oakland and San Jose, California#629275
• Air pollution and health#7475