Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 May 1.
Published in final edited form as: Sci Total Environ. 2024 Mar 6;923:171535. doi: 10.1016/j.scitotenv.2024.171535

Joint Effects of Air Pollution and Neighborhood Socioeconomic Status on Cognitive Decline - Mediation by Depression, High Cholesterol Levels, and High Blood Pressure

Yiyang Mei 1, Grace M Christensen 1, Zhenjiang Li 2, Lance A Waller 2,3, Stefanie Ebelt 1,2, Michele Marcus 1,2, James J Lah 4, Aliza P Wingo 5,6, Thomas S Wingo 4,7, Anke Hüls 1,2
PMCID: PMC10965363  NIHMSID: NIHMS1974466  PMID: 38453069

Abstract

Air pollution and neighborhood socioeconomic status (N-SES) are associated with adverse cardiovascular health and neuropsychiatric functioning in older adults. This study examines the degree to which the joint effects of air pollution and N-SES on the cognitive decline are mediated by high cholesterol levels, high blood pressure (HBP), and depression. In the Emory Healthy Aging Study, 14,390 participants aged 50+ years from Metro Atlanta, GA, were assessed for subjective cognitive decline using the cognitive function instrument (CFI). Information on the prior diagnosis of high cholesterol, HBP, and depression was collected through the Health History Questionnaire. Participants’ census tracts were assigned 3-year average concentrations of 12 air pollutants and 16 N-SES characteristics. We used the unsupervised clustering algorithm Self-Organizing Maps (SOM) to create 6 exposure clusters based on the joint distribution of air pollution and N-SES in each census tract. Linear regression analysis was used to estimate the effects of the SOM cluster indicator on CFI, adjusting for age, race/ethnicity, education, and neighborhood residential stability. The proportion of the association mediated by high cholesterol levels, HBP, and depression was calculated by comparing the total and direct effects of SOM clusters on CFI. Depression mediated up to 87% of the association between SOM clusters and CFI. For example, participants living in the high N-SES and high air pollution cluster had CFI scores 0.05 (95%-CI:0.01,0.09) points higher on average compared to those from the high N-SES and low air pollution cluster; after adjusting for depression, this association was attenuated to 0.01 (95%-CI:−0.04,0.05). HBP mediated up to 8% of the association between SOM clusters and CFI and high cholesterol up to 5%. Air pollution and N-SES associated cognitive decline was partially mediated by depression. Only a small portion (<10%) of the association was mediated by HBP and high cholesterol.

Keywords: Air pollution, neighborhood deprivation, cognitive functioning, joint effects, environmental mixtures, epidemiology, mediation, depression, cholesterol, high blood pressure

Graphical Abstract

graphic file with name nihms-1974466-f0001.jpg

1. Introduction

Age is a significant risk factor for many diseases, and protecting the health of an aging population is a pressing public health issue. According to the World Health Organization, by 2030, 1 in 6 people in the world will be age 60 years or over (Rudnicka et al., 2020). As people age, several diseases may occur; many at the same time. Age-related diseases, such as cardiovascular disease and stroke, accelerate the rate of neuronal dysfunction and neuronal loss, which contribute to cognitive decline(Murman, 2015).

Numerous studies have investigated the association between exposure to air pollution and accelerated cognitive decline across multiple stages of life (Clifford et al., 2016; Kilian & Kitazawa, 2018; Power et al., 2016). Two recent critical reviews examining the association between exposure to ambient air pollutants and acceleration of cognitive decline suggested that their relationship is likely to be causal (Delgado-Saborit et al., 2021; Weuve et al., 2021). This result is further supported by animal model studies suggesting that particulate matter (PM) adversely impacts neuroanatomical and neuropathological changes in the brain, leading to the development of Alzheimer’s disease (AD)-like pathology via a multifactorial mechanism, including neurodegeneration, amyloid processing, and immune response (Kilian & Kitazawa, 2018). However, the causal pathway through which air pollution leads to cognitive decline, and particularly the role of common comorbidities such as cardiovascular and neuropsychiatric diseases in this association, is not yet clearly understood.

Cardiovascular and neuropsychiatric comorbidities have each been associated with both air pollution and cognitive decline. A number of epidemiologic studies have identified air pollution as a risk factor for cardiovascular morbidity and mortality (Manisalidis et al., 2020; Miller & Newby, 2020; Samet et al., 2000; Spiller et al., n.d.; Zong et al., 2022). Cardiovascular diseases (CVD) and cognitive decline are also strongly associated with each other (C. Qiu & Fratiglioni, 2015). In addition, neuropsychiatric comorbidity such as depression is also associated with air pollution and cognitive decline (Altuğ et al., 2020). Previous studies suggest that exposure to air pollution contributs to depression onset (Kioumourtzoglou et al., 2017; X. Qiu et al., 2022) as well as to the risk of acute hospital admissions for psychiatric disorders (X. Qiu et al., 2022). Furthermore, there is strong evidence that depression is associated with an increased risk of subsequent cognitive decline and dementia (Jorm, 2000; Perini et al., 2019).

Despite documented associations between air pollution and cardiovascular and neuropsychiatric comorbidities of cognitive decline, little is known about the potential mediating effects of these comorbidities on the association between air pollution and cognitive decline. Few studies have investigated the interrelationship between air pollution, depression, and cognitive decline. Among the limited number of papers that have done so, most explore of effect modification rather than mediation (Altuğ et al., 2020). To our knowledge, there are no studies investigating the mediation effects of depression on associations between outdoor air pollution and cognitive decline. Furthermore, only two studies explore mediation effects of cardiovascular disease on the association between outdoor air pollution and cognitive decline (Grande et al., 2020; Ilango et al., 2020). Both studies found some mediation of the association between air pollution and dementia by cardiovascular disease, but estimated proportions mediated ranged from 9% (Ilango et al., 2020) to 50% (Grande et al., 2020), emphasizing the need for additional studies on this topic.

Another understudied aspect concerning air pollution’s effect on cognitive decline is its joint effects with adverse neighborhood socioeconomic characteristics as a co-exposure. In several previous studies, low neighborhood socioeconomic status has had harmful effects on cognitive decline (Oi & Haas, 2019; Tang et al., 2022). Individuals living in disadvantaged neighborhoods are often exposed to a higher level of air pollution than individuals living in socioeconomically advantaged neighborhoods (Hajat et al., 2015). According to the theory of triple jeopardy, communities with low SES not only suffer from higher exposure to air pollutants and other environmental hazards but also have increased susceptibility to poor health because of psychosocial stressors (Ailshire & Clarke, 2015). These psychosocial stressors lower the brain’s threshold for neurotoxicity, making those living in disadvantaged neighborhoods more vulnerable to the harmful effects of air pollution (Christensen et al., 2022; Li et al., 2022; McEwen & Tucker, 2011). For those living in high SES communities, however, more advantageous SES can buffer the adverse effects of air pollution on cognitive decline, which can result in biased effect estimates when ignoring interconnections between air pollution and neighborhood SES (N-SES) (Christensen et al., 2022; Li et al., 2022).

Previous studies conducted in the Emory Health Aging Study (EHAS) have shown individual and joint effects of air pollution and N-SES on cognitive decline (Christensen et al., 2022; Li et al., 2022). This study aims to expand on that work by investigating the degree by which the joint effects of air pollution and N-SES on cognitive decline are mediated by high cholesterol levels, high blood pressure, and depression using data from 14,390 older adults of the EHAS in Metro Atlanta, GA.

2. Methods

2.1. Study Population

Started in 2015, The Emory Healthy Aging Study (EHAS) is a large gerontology-based prospective study that focuses on the diseases of older adults. Anyone over the age of 18, living in the United States, and sufficiently fluent in English are encouraged to enroll. Individuals receiving health services at Emory Healthcare are the main focus of recruitment in the Metro-Atlanta area as well as their spouses, family members, and associated non-relatives. The recruitment is also open to the public. In general, EHAS aims to advance understanding of healthy aging and pathogenesis of age-related illnesses in well-characterized, community-based prospective cohorts and to identify biomarkers for the earliest manifestations of Alzheimer’s disease for the facilitation of preventative interventions (Goetz et al., 2019). EHAS participants included in this analysis were enrolled during the 2015–2020 period, were 50 years and older at baseline, and lived in the Metro Atlanta area. At enrollment, the participants were asked to complete a Health History Questionnaire which included demographic questions such as age, race and ethnicity, general and perceived health status, and memory decline. Prior to enrollment, all participants completed an online consent process. This study was approved by the Emory University Institutional Review Board.

2.2. Exposure Assessment

As described previously (Christensen et al., 2022), our exposure of interest was the joint exposure to air pollution and N-SES, which was estimated using Self-Organizing Maps (SOM), as described in the statistical analysis section. Air pollution and N-SES data, as described below, were used to create exposure profiles of each census tract in Metro-Atlanta. These exposure clusters represent the air pollution and N-SES exposure profile of a census tract. Using these exposure clusters we estimate the joint effect of the exposure profile on cognitive decline.

The average ambient air pollution concentrations from 2008–2010 were derived from the Community Multiscale Air Quality (CMAQ) chemical transport model at a grid resolution of 4-kilometers. We considered 12 pollutants: Nitrogen Oxides (NOx), Nitrogen Dioxide (NO2), Nitrate (NO3), Sulfur Dioxide (SO2), Ozone (O3), Carbon Monoxide (CO), Ammonium (NH4), Sulfate (SO4), Elemental Carbon (EC), Organic Carbon (OC), and Particulate Matter with diameters 10 microns or less (PM10) and 2.5 micrometers or less (PM2.5) (Senthilkumar et al., 2019). CMAQ predictive performance varies for for each pollutant, with spatial R2 ranging from 0.92 (SO4) to 0.44 (PM10), and is described in detail elsewhere (Senthilkumar et al., 2019). Each participant was assigned CMAQ pollutant concentrations based on their census tract of residence at baseline by geospatially matching the center of each CMAQ grid cell to its closest respective residential census tract, using center to center matching. The residential census tract rather than the street address was used here because the SOM requires all inputs, i.e., air pollution and N-SES, to be at the same spatial resolution.

N-SES data were obtained from the United States Census Bureau’s American Community Survey for 2013–2018 for each Metro Atlanta census tract. A total of 16 N-SES variables in the domains of income, racial composition, education, employment, occupation, and housing properties were selected to represent the mixture of exposure to N-SES (Messer et al., 2006). For each variable, 5-year average estimates were calculated through the R package tidycensus. To align interpretation with the other N-SES characteristics, the variable representing the median home value was multiplied by −1. Since residential census tract at baseline was used to connect both air pollution estimates and N-SES characteristics and we had no information on how long the participants lived at that address, there may be exposure misclassification by participants moving in the years prior to their baseline visit. To control for this, an indicator for residential stability representing the percentage of households in the participants’ residential census tract that moved into their current residence before 2010 was also added to the analysis (Christensen et al., 2022).

2.3. Outcome Assessment

The Cognitive Function Instrument (CFI) was employed to measure subjective cognitive decline. The CFI was self-administered online. It consisted of 14 questions related to the participants’ subjective cognitive concerns in daily life. Participants gave themselves scores based on their daily feelings. The total CFI score was calculated by scoring responses, with “yes” equals 1, “no” equals 0, and “maybe” equals 0.5. The higher the score, the more advanced perceived memory and cognitive decline (Amariglio et al., 2015). A higher CFI score is predictive of cognitive decline for the elderly because subjective experience of cognitive decline often marks the late stage of the preclinical, cognitively unimpaired phase of the Alzheimer’s disease continuum (Jack Jr. et al., 2018). Unlike other tests that assess cognitive impairment, the CFI can be completed remotely. It does not require an in-person interview or physician review. In our regression analyses, the CFI score was log-transformed to adjust for the right skew in the distribution of observed total score values.

2.4. Assessment of Potential Confounders

Potential confounders under consideration included participants’ individual age, race and ethnicity, education, and residential stability of the census tract. All information related to the confounders except residential stability was collected through an online Health History Questionnaire (HHQ), which is incorporated in a Qualtrics Survey (Qualtrics.com) and completed at baseline and annually thereafter (Goetz et al., 2019).

Race and ethnicity were considered to be separate variables. Race was divided into three categories: White, Black, and others. Ethnicity was classified as Hispanic yes or no. Education has 7 levels: less than high school, high school or General Education Development (GED), some college, associate degrees, bachelor’s degrees, masters’ degrees, and professional or doctorate degrees. Residential stability was measured through the percentage of residents living in the same household moving in before 2010. The median and interquartile range were calculated for continuous variables (age and residential stability). For categorical variables (education, race, and ethnicity), the percentages in each category were calculated.

2.5. Assessment of Potential Mediators

Cardiovascular diseases and mental health conditions were selected as potential mediators because they are associated with air pollution and N-SES, as well as risk factors for cognitive decline. High blood pressure and high cholesterol were selected as proxies for cardiovascular disease, and depression was selected as proxy for mental health in this analysis because a high proportion (>30%) of participants reported having these conditions. Information on participants’ cholesterol, blood pressure, and depression was collected through HHQ. Participants were asked whether they had ever been diagnosed by a doctor with high cholesterol levels, high blood pressure, or depression. All three variables were binary.

2.6. Statistical Analysis

As described previously (Christensen et al., 2022), we used the SOM algorithm (Pearce et al., 2014) to identify clusters of census tracts with similar air pollution and N-SES characteristics, which reflect the joint exposure of air pollution and N-SES. The number of clusters identified by the SOM algorithm was determined by identifying group structure using within-cluster sum of squares and between-cluster sum of square statistics, as well as visual inspection of cluster star plots. The method identifies clusters with exposure levels homogeneous within and heterogeneous between clusters. Next, census tract clusters were matched to EHAS participants using the census tract of the participants’ addresses. The SOM exposure clusters were then used as categorical exposure variables in the subsequent regression analyses to estimate the joint effects of air pollution and N-SES. The reference cluster designates the metro Atlanta area with the highest N-SES and lowest air pollution concentrations.

Following on our previous work that found an association between SOM clusters and CFI (Christensen et al., 2022), here we assessed the potential mediation effect of high cholesterol, high blood pressure, and depression on SOM clusters and CFI. Our analysis consisted of three steps. We estimated 1) the associations between SOM clusters and high cholesterol levels, high blood pressure, and depression (using logistic regression analysis); 2) the association between high cholesterol levels, high blood pressure, and depression and log-transformed CFI scores (lnCFI) (using linear regression models); and 3) the proportion of the association between SOM clusters and lnCFI mediated by high cholesterol levels, high blood pressure, and depression by comparing the total and direct effects from linear regression models (Robins & Greenland, 1992).

Mediation analyses rely on three assumptions. The first assumption is that there is no confounding between exposure (SOM cluster) and mediator (high cholesterol level, high blood pressure, and depression). The second assumption is that there is no confounding relationship between mediators and outcome (cognitive decline). The third assumption is that there is no confounding between exposure (SOM cluster) and outcome (cognitive decline). Confounders of each association were selected using directed acyclic graphs (DAGs) informed by the existing literature (Figure S1) and controlled in each analysis, as described below.

As of publication, no mediation R package could accommodate mediation with a multilevel categorical exposure variable (SOM clusters). Therefore we estimated the total effect, direct effect, and proportion mediatied as follows: The total effects were estimated using linear regression analysis for the association between SOM exposure clusters and lnCFI scores adjusting for individual age, race and ethnicity, education, and residential stability. The direct effects were derived by additionally adjusting these models for the potential mediators in individual linear regression models. The proportion mediated (PM) was calculated as (total effect – direct effect) / total effect.

To evaluate potential exposure-mediator interactions that could bias the results of our mediation analysis, we investigated effect modification of the association between SOM cluster and lnCFI score by depression, high cholesterol, and high blood pressure. Linear regression models with an interaction term between SOM cluster and potential mediators (depression, high cholesterol, and high blood pressure) were adjusted for confounding as described above.

3. Results

3.1. Population Characteristics

The final study sample contained 14,390 individuals, 50 years and older, living in the Metro Atlanta area. These participants had an average age of 61 years and were mostly white (80.3%). Participants of EHAS were highly educated − 74.3% of them had a bachelor’s degree or higher. The median CFI score was 1.5 (IQR: 2.5) (Table 1).

Table 1.

Study population characteristics of the Emory Healthy Aging Study (EHAS) stratified by Self-Organized Map (SOM) cluster.

SOM Cluster*
Total 1 2 3 4 5 6
A. Study Population Characteristics
N (%) 14390 3914(27.2) 707(4.9) 521(3.6) 5427(37.7) 2258(15.7) 1563(10.9)
Median CFI (IQR) 1.50 (2.50) 1.50 (2.50) 2.00 (2.50) 2.00 (3.00) 1.50 (2.50) 2.00 (2.50) 2.00 (2.50)
Median Age (IQR) 60.67 (12.81) 59.25 (14.67) 59.87 (13.11) 57.14 (13.82) 62.00 (11.68) 61.67 (11.07) 59.68 (12.76)
% Residents Moved in Before 2010 (IQR) 20.84 (18.94) 13.43 (12.21) 13.57 (13.02) 16.23 (14.26) 27.80 (21.12) 21.68 (15.86) 25.32 (19.70)
Race (%)
White 11441 (80.3) 3436 (88.6) 540 (77.8) 196 (37.8) 4849 (90.2) 1850 (82.9) 570 (36.8)
Black 2034 (14.3) 237 (6.1) 97 (14.0) 296 (57.1) 235 (4.4) 259 (11.6) 910 (58.8)
Other 767 (5.4) 204 (5.3) 57 (8.2) 26 (5.0) 290 (5.4) 123 (5.5) 67 (4.3)
Hispanic (%) 547 (3.8) 160 (4.1) 32 (4.6) 20 (3.8) 196 (3.6) 90 (4.0) 49 (3.1)
Education (%)
Less than High School 35 (0.2) 5 (0.1) 2 (0.3) 6 (1.2) 5 (0.1) 11 (0.5) 6 (0.4)
High School /GED 531 (3.7) 61 (1.6) 35 (5.0) 54 (10.4) 149 (2.7) 146 (6.5) 86 (5.5)
Some College, but no Degree 2133 (14.8) 355 (9.1) 128 (18.3) 109 (20.9) 704 (13.0) 516 (22.9) 321 (20.6)
Associates Degree 993 (6.9) 160 (4.1) 63 (9.0) 48 (9.2) 320 (5.9) 258 (11.4) 144 (9.2)
Bachelor’s Degree 4872 (33.9) 1376 (35.2) 233 (33.2) 147 (28.2) 1968 (36.3) 658 (29.2) 490 (31.5)
Master’s Degree 3837 (26.7) 1218 (31.2) 173 (24.7) 111 (21.3) 1510 (27.8) 443 (19.7) 382 (24.5)
Professional or Doctorate Degree 1963 (13.7) 730 (18.7) 67 (9.6) 46 (8.8) 769 (14.2) 222 (9.8) 129 (8.3)
B. Air Pollution Measurements – Median (IQR)
CO (ppm) 0.62 (0.21) 0.67 (0.13) 0.69 (0.13) 0.65 (0.15) 0.54 (0.23) 0.40 (0.24) 0.56 (0.16)
EC (μg/m3) 0.95 (0.33) 1.00 (0.26) 1.05 (0.25) 1.20 (0.45) 0.91 (0.22) 0.67 (0.38) 0.97 (0.29)
NH4(μg/m3) 1.04 (0.05) 1.05 (0.06) 1.08 (0.07) 1.04 (0.06) 1.04 (0.04) 1.02 (0.09) 1.02 (0.06)
NO2 (ppb) 23.63 (5.56) 24.79 (2.58) 25.12 (3.09) 25.10 (6.30) 22.59 (5.73) 17.20 (9.36) 22.54 (3.62)
NO3 (ppb) 0.62 (0.05) 0.61 (0.03) 0.64 (0.03) 0.62 (0.03) 0.63 (0.06) 0.62 (0.09) 0.61 (0.04)
NOX (ppm) 0.04 (0.02) 0.05 (0.02) 0.05 (0.00) 0.05 (0.02) 0.04 (0.02) 0.03 (0.02) 0.04 (0.01)
OC(μg/m3) 2.84 (0.20) 2.84 (0.13) 2.92 (0.37) 2.92 (0.26) 2.82 (0.20) 2.82 (0.27) 3.00 (0.17)
O3 (ppm) 0.04 (0.00) 0.04 (0.00) 0.04 (0.00) 0.04 (0.00) 0.04 (0.00) 0.04 (0.00) 0.04 (0.00)
PM10(μg/m3) 20.96 (0.20) 20.86 (0.13) 20.95 (0.19) 20.97 (0.16) 20.98 (0.19) 21.08 (0.16) 21.01 (0.23)
PM2.5(μg/m3) 12.56 (0.51) 12.52 (0.37) 12.83 (1.01) 12.85 (0.78) 12.52 (0.46) 12.54 (0.61) 12.85 (0.61)
SO2 (ppb) 8.37 (2.52) 8.92 (1.99) 8.98 (2.7) 8.19 (2.8) 8.33 (2.27) 6.48 (3.11) 7.18 (1.42)
SO4 (ppb) 2.95 (0.10) 2.97 (0.09) 3.01 (0.15) 2.95 (0.14) 2.95 (0.09) 2.87 (0.15) 2.89 (0.05)
C. N-SES Indicators – Median (IQR)
% Education Less than High School 5.50 (7.56) 3.60 (4.85) 21.39 (14.05) 16.25 (7.02) 3.20 (3.43) 11.04 (6.44) 9.75 (5.62)
Unemployment Rate 6.38 (4.03) 5.40 (3.76) 7.79 (3.85) 16.08 (5.35) 5.34 (2.89) 7.51 (2.71) 12.37 (4.89)
% Not in Labor Force 24.47 (8.33) 21.45 (8.86) 19.41 (13.30) 32.62 (12.42) 24.17 (7.13) 25.48 (6.84) 28.55 (6.73)
% Homes Vacant 8.12 (6.98) 10.11 (7.78) 11.79 (4.68) 16.71 (8.35) 5.13 (4.11) 7.53 (4.87) 11.33 (5.03)
% Homes Rented 28.76 (32.13) 48.72 (18.90) 64.55 (16.33) 63.97 (13.72) 12.39 (13.22) 23.66 (15.23) 30.04 (15.66)
% Homes Crowded 0.98 (1.85) 1.00 (1.58) 5.58 (3.04) 2.99 (2.61) 0.23 (0.93) 1.65 (1.91) 1.63 (1.93)
Median Home Value ($ in thousands) 247833.33 (183216.67) 316.33 (133.18) 158.85 (79.72) 99.05 (45.26) 329.20 (170.87) 166.48 (43.35) 128.93 (42.00)
% Male not in Management 48.12 (31.89) 37.64 (13.17) 76.28 (9.99) 79.71 (17.94) 39.61 (17.91) 68.17 (10.51) 70.44 (9.72)
% Female not in Management 46.17 (21.75) 35.83 (14.19) 65.03 (16.01) 69.66 (13.38) 39.28 (13.53) 59.62 (8.84) 58.45 (11.48)
% In Poverty 5.68 (7.82) 5.68 (6.44) 17.90 (8.49) 29.30 (8.77) 3.23 (2.63) 8.40 (5.78) 12.02 (8.27)
% Female Headed Households 5.09 (5.81) 4.08 (3.92) 9.24 (3.66) 14.52 (5.12) 3.52 (2.67) 6.73 (3.69) 11.28 (4.52)
% Income less than $35,000 21.03 (15.97) 24.79 (11.75) 39.33 (8.22) 54.24 (9.44) 12.69 (5.68) 23.80 (8.55) 31.09 (12.08)
% on Public assistance 0.87 (1.36) 0.80 (0.96) 1.67 (1.49) 3.18 (2.49) 0.52 (0.71) 1.50 (1.32) 1.85 (1.72)
% No Car 13.88 (12.16) 21.48 (5.66) 22.27 (3.91) 30.12 (7.49) 8.37 (5.05) 11.10 (4.73) 16.17 (7.10)
% Non-Hispanic Black 14.86 (21.24) 14.89 (13.58) 20.66 (18.52) 79.64 (26.51) 6.82 (9.06) 19.88 (18.32) 74.94 (30.69)
% Hispanic 5.38 (5.97) 5.59 (4.13) 36.41 (22.72) 3.62 (4.46) 4.55 (3.30) 9.93 (9.80) 4.17 (3.88)
D. Mediators
High Blood Pressure 5854 (41.6) 1343 (35.2) 307 (44.8) 252 (49.2) 2118 (39.8) 986 (44.9) 848 (55.4)
High Cholesterol 6778 (48.5) 1693 (44.5) 339 (49.3) 216 (42.5) 2638 (49.9) 1177 (54.0) 715 (47.2)
Depression 3864 (32.3) 1100 (33.8) 218 (36.9) 154 (34.4) 1366 (30.5) 659 (34.9) 367 (28.3)
Open in a new tab

3.2. Exposure Characteristics

Based on the air pollution and N-SES characteristics of the census tracts, the SOM identified 6 clusters (Figure 1A). Most participants lived in cluster 4 (n=5427; 37.71%), characterized by high N-SES and relatively low air pollution, while relatively few lived in cluster 2 (n = 707, 4.91%) and 3 (n=521, 3.62%), both clusters characterized by low N-SES and high air pollution. In the map of Metro Atlanta (Figure 1B), census tracts were color-coded based on their SOM cluster assignment. Cluster 4, seen on the map in dark blue, primarily appears in the northern half of the Metro-Atlanta outside the City of Atlanta. Clusters 1, 2, and 3, all characterized by high air pollution, appear primarily in the City of Atlanta around highways (black lines on the map). These three clusters have the highest proportions of rented homes and percentage of dwellers not owning a car. Historically, because of Atlanta’s segregation policies, the highest proportion of non-Hispanic Black participants reside within clusters 3 and 6, which are located in southern Atlanta. Cluster 2 has the highest percentage of Hispanic residents. Participants in all clusters had similar median ages, with the youngest residing in cluster 3 (median 57 years; IQR: 13.82) and the oldest in cluster 4 (median 62 years; IQR: 11.68). Cluster 4 was used as the reference group for regression and mediation analyses because of its low concentration of air pollution and high N-SES. Correlations between air pollution and N-SES characteristics of the census tracts are shown in Figure S2.

Figure 1. Self-Organized Map (SOM) clusters.

Figure 1.

Open in a new tab

A. SOM cluster star plot, slices represent median values of a mixture component, each circle is a SOM cluster. Blue slices correspond with N-SES indicators, while pink slices correspond with air pollutants. B. Map of census tracts in Metro-Atlanta by cluster. Black lines represent major highways.

3.3. Association between Exposure and Mediators

Living in a neighborhood assigned to SOM cluster 1 exposure profile had lower odds of having high cholesterol (OR: 0.86; 95% CI: 0.78, 0.94), compared to living in a neighborhood with a cluster 4 exposure profile (Figure 2A). Participants assigned to cluster 4 had higher O3 and PM10 exposure concentrations, lower exposure concentrations of the other air pollutants, and a similarly high N-SES as those in cluster 1. In contrast, living in a neighborhood with a cluster 5 exposure profile had higher odds of high cholesterol (1.18; 1.06, 1.31), compared to a cluster 4 exposure profile (Figure 2A, Table S2A). Participants assigned to cluster 5 had similar air pollution concentrations as participants assigned to cluster 4 (reference category) but a lower N-SES.

Figure 2. Association between SOM clusters (joint effects of air pollution and N-SES) and common cardiovascular and neuropsychiatric comorbidities.

Figure 2.

Open in a new tab

Association between the joint effects of air pollution and N-SES and A. high blood pressure, B. high cholesterol levels, and C. depression. Associations were adjusted for age, race/ethnicity, education, and neighborhood residential stability.

We observed slightly stronger associations with narrower confidence intervals when comparing high blood pressure status among SOM clusters, with significant associations observed for the comparison of clusters 1, 2, 5, and 6 to the reference cluster 4. Similar to the associations with high cholesterol, cluster 1 was protective of high blood pressure (OR: 0.89; 95% CI: 0.81, 0.98) and we observed higher odds for cluster 5 (OR: 1.11; 95% CI: 1.00, 1.24). In addition, we observed adverse effects for clusters 2 (OR: 1.22; 95% CI: 1.02, 1.45) and 6 (OR: 1.27; 95% CI: 1.11, 1.45), which are both characterized by high air pollution concentrations and low N-SES.

For depression, we observed non-significant higher odds of depression for people in clusters 2, 3, 5 and 6 compared to reference cluster 4 (Figure 2C, Table S2A).

3.4. Association between Mediator and Outcome

High cholesterol level, high blood pressure, and depression were associated with higher lnCFI scores (Figure 3). Participants with these health conditions had lnCFI scores 0.39 (95% CI: 0.36, 0.42; depression), 0.12 (95% CI: 0.09, 0.15; high blood pressure), and 0.13 (95% CI: 0.10, 0.15; high cholesterol levels) points higher than those without these conditions (Table S2B).

Figure 3. Association between a prior diagnosis of common cardiovascular and neuropsychiatric comorbidities and cognitive functioning.

Figure 3.

Open in a new tab

Association between high blood pressure (HBP), high cholesterol level (CHOL), depression and cognitive decline. The association was adjusted for individual age, race/ethnicity, education, and neighborhood residential stability.

3.5. Mediation by high cholesterol level, high blood pressure, and depression

A prior study in this population found a significant total effect of SOM exposure cluster on lnCFI score (Christensen et al., 2022). The mediation analysis focused on the mediating effects of high cholesterol levels, high blood pressure, and depression on the association between SOM exposure clusters and lnCFI scores (Figure 4). The largest mediated proportion was found for depression with 19% (for cluster 3) to 87% (for cluster 1) of the association between SOM cluster and lnCFI being mediated by depression (Figure 4, table S1). For example, prior to adjusting for depression, on average, participants living in cluster 1 had CFI scores 0.05 (CI: 0.01, 0.09) (total effect) higher than those from cluster 4. After adjusting for depression, the association was attenuated (0.01; CI: −0.04, 0.05; direct effect).

Figure 4. Mediation by common cardiovascular and neuropsychiatric comorbidities (A. Mediation by high cholesterol levels, B. Mediation by high blood pressure, C. Mediation by depression).

Figure 4.

Open in a new tab

Associations between SOM clusters (joint effects of air pollution and N-SES) and cognitive functioning before (total effects) and after adjusting for potential mediators (direct effects). Proportion mediated (PM) was calculated as (total effect – direct effect) / total effect. Associations were adjusted for individual age, race/ethnicity, education, and neighborhood residential stability.

The maximum proportion mediated by high cholesterol and high blood pressure for the association between SOM clusters and cognitive decline did not exceed 10%. Up to 8% of the association between SOM clusters and cognitive decline was mediated by high blood pressure (for cluster 3), with only 5% at maximum for high cholesterol levels (for cluster 6).

Next, we evaluated potential interaction between SOM clusters and mediators (high blood pressure, high cholesterol, and depression) that could bias the results of our mediation analysis. Participants in cluster 3 who were diagnosed with depression had significantly higher lnCFI scores than those without depression (p-value: 0.01; Figure S3A). There were no significant exposure-mediator interactions for the other clusters, e.g. for cluster 1 (interaction p-value: 0.56), for which we found the largest mediation effects. There was no effect modification of the association between SOM cluster and lnCFI score by high cholesterol or high blood pressure (Figure S3B+C).

4. Discussion

In a prevous study of 14,390 individuals 50 years and older from Metro Atlanta, we found a significant total effect of air pollution and N-SES on cognitive decline (Christensen et al., 2022). In the current study on the same study population, we showed that up to 87% of the association between air pollution and N-SES on cognitive decline was mediated by depression, whereas only a small portion (<10%) was mediated by high blood pressure and high cholesterol levels. High blood pressure, high cholesterol levels, and depression were all associated with cognitive decline. We further identified a significant joint effect of air pollution and N-SES on high blood pressure and cholesterol, whereas the joint effect of these exposures and depression was weaker and not significant. We observed adverse joint effects of exposure to air pollution and low N-SES on depression, but the associations were not statistically significant, potentially due to a lack of statistical power. While research on the joint effects of air pollution and N-SES on depression is scarce, the direction of association is in line with several studies that have reported an association between air pollution and increased odds of depression or the use of antidepressants (Ali & Khoja, 2019; Fan et al., 2020; Kioumourtzoglou et al., 2017; Lim et al., 2012; Vert et al., 2017; Yang et al., 2023) as well as between low N-SES and depression (Jakobsen et al., 2022; Mair et al., 2008; Richardson et al., 2015).

Depression was associated with cognitive decline in our study and the strongest predictor of subjective cognitive decline in comparison to other comorbidities (high blood pressure, high cholesterol), consistent with the existing literature. Substantial evidence suggests that cognitive deficits could persist after remission of a major depressive episode (Formánek et al., 2020; Perini et al., 2019; Semkovska et al., 2019). In a review of biological mechanisms relating the putative mechanisms of cognitive decline in depression, Dobielska and colleagues reported reduced cognitive function in two-thirds of the depressed patients and at least one-third of the remitted patients studied (Dobielska et al., 2022). Their hypothesis is that inflammation drives a decrease in neuroplasticity and damages the brain structure, mainly the hippocampus. The atrophy of the hippocampus has proven to be a strong predictor of cognitive decline; therefore, depression-associated inflammation may contribute to cognitive decline. Of note, inflammation is also one of the biological pathways hypothesized to link air pollution exposure with cognitive decline (Dobielska et al., 2022).

In our study, depression mediated 19 to 87% of the association between SOM cluster and cognitive decline. As discussed above, SOM cluster was not significantly associated with depression in exposure-mediator analysises, which limits the precision of the estimated proportion mediated. So far, very few studies investigate the interconnections between air pollution and N-SES, depression, and cognitive decline. One study investigated the role of cognitive impairment as a potential effect modifier of the association between air pollution and depression (Altuğ et al., 2020). Another study investigated the association between natural outdoor environments (parks, forests, and recreation areas) and cognitive functions, and used loneliness and mental health as potential mediators. However, they did not find indications for mediation by loneliness and mental health (Zijlema et al., 2017). Therefore, our findings contribute to the current literature by providing the first suggestion of a mediating role of depression for the association between the joint exposure to air pollution and low N-SES and cognitive decline. Further studies with more statistical power should investigate the role of depresson in mediating this association.

We found significant associations between SOM cluster and cardiovascular comorbidities, particularly with high blood pressure. This finding is consistent with the current literature, which shows a robust association between air pollution and high blood pressure (Giorgini et al., 2016) as well as some evidence of an association with high cholesterol levels (McGuinn et al., 2019). Low N-SES has also been consistently reported to be associated with higher blood pressure and cholesterol levels (Espírito Santo et al., 2019; Grotto et al., 2008; Jenkins & Ofstedal, 2014; Leng et al., 2015).

High blood pressure and high cholesterol were associated with cognitive decline, though the effect estimates were smaller than for depression. This result is consistent with a systematic review conducted in 2019 summarizing results from 50 studies (Forte et al., 2019). High blood pressure affects cerebral perfusion which can cause alterations in the physiological processes of cerebral blood flow regulation, making hypertensive patients more vulnerable to the development of cerebrovascular damage and vascular dementia (Jennings et al., 2005; Moretti et al., 2008). As for the association between high cholesterol levels and dementia, our findings contribute to the currently mixed results by adding to the pool of evidence that increased cholesterol level is a risk factor for dementia (Anstey et al., 2008; Peters et al., 2021).

High blood pressure and high cholesterol level mediated a small percentage of the association between SOM clusters and cognitive decline. This study is one of very few exploring cardiovascular comorbidities as potential mediators of the association between air pollution and cognitive decline. One study evaluated the role of CVD as a mediator of the association between indoor unclean fuel and adult cognitive function and indoor unclean fuel (Cong et al., 2021). They found that hypertension explained more than 50% of such a fuel-related decline in the verbal memory (Cong et al., 2021). So far, only two studies have investigated CVD as a potential mediator of the association between outdoor air pollution and cognitive decline (Grande et al., 2020; Ilango et al., 2020). The proportions of the association mediated by CVD varied widely and ranged from 9% (Ilango et al., 2020) to 50% (Grande et al., 2020). In our study, up to 8% of the association between air pollution and N-SES on subjective cognitive decline was mediated by high blood pressure and high cholesterol levels, which is more in line with the estimates from Ilango et al. (2020). However, important differences between their and our study are that we looked at the joint effects of air pollution and N-SES, whereas they only focused on air pollution, and differences in the definition of CVD. Both previous studies focused on severe cardiovascular events (e.g., ischemic heart disease, heart failure, atrial fibrillation, and stroke) (Grande et al., 2020; Ilango et al., 2020), whereas we captured with high blood pressure and high cholesterol levels more common and less severe CVD, which are often well-controlled with medications. This difference in mediator assessment could explain the smaller proportions mediated observed for CVD in our study.

A strength of this study is its large sample size. EHAS collected information from a diverse city. The racial, ethnic, and socioeconomic landscape in Atlanta allowed for the definition of diverse profiles of the exposure mixture by SOM clustering algorithm. The second strength is the use of CFI to determine subjective cognitive decline. This measurement strengthens this analysis because subjective cognitive decline is one of the first signs of progression to dementia (Amariglio et al., 2015; Jessen et al., 2014)

There are several limitations in this study. The first is the cross-sectional nature of the study data with subjective cognitive decline over the last year (outcome) being assessed at the same time as the mediators (ever diagnosed with high blood pressure, high cholesterol, or depression). However, while the mediator and outcome assessment were done on the same day, the doctor’s diagnosis of comorbidities (high blood pressure, high cholesterol levels, and depression) must have occurred prior to their baseline visit to be reported, which limits the risk of reverse causality. On the other hand, the outcome (subjective cognitive decline) describes one of the first signs of cognitive decline, which is assumed to be a very recent event. Additionally, the air pollution measurements are from 2008–2010, which is prior to the enrollment window and measurements of the N-SES characteristics, mediators, and outcome. This may introduce possible exposure misclassification, based on participants moving prior to their enrollment in the EHAS cohort. However, any misclassification is likely to be non-differential by CFI score. The second limitation is the lack of representativeness for the information being collected. EHAS is not a population-based sample. The majority of the participants are white and live in areas with high N-SES, which is not representative of all residents in Atlanta. This could have limited the statistical power of our association analysis, particularly for the association between SOM clusters and depression. The grid size of the CMAQ chemical transport model is also a limitation. The 4 km resolution grids may not capture granular enough variation within air pollution levels in Atlanta. However, another study using 250m resolution did report similar harmful effects of air pollutants on cognitive decline (Li et al., 2022). The fourth limitation focuses on the self-reporting status of high cholesterol levels, high blood pressure, depression, and subjective cognitive decline. Participants were asked to report doctors’ diagnoses. Recall bias may be present in data collection leading to misclassification of the mediator. There is also possible misclassification of the exposure based on participants moving prior to their enrollment in EHAS, although this is likely to be non-differential based on their CFI scores. There may also be residual or unmeasured confounding from mediator-outcome confounders such as diet that were unmeasured in this study. The final limitation is that, due to the categorical nature of the exposure with a total of six categories, we could not use one of the recent R packages for causal mediation analyses (e.g., R packages mediation or CMAverse) that would have allowed us to incorporate exposure-mediator interactions. However, our sensitivity analysis analysis did not suggest any strong exposure-mediator interactions that could have biased our mediation results.

5. Conclusions

In our study of 14,390 participants from metro Atlanta, GA, depression explained more of the association between the joint effects of air pollution and N-SES on cognitive decline than high blood pressure and high cholesterol levels. Future studies are needed to replicate our findings and to investigate whether these results can help to understand the type of dementia that is most commonly caused by air pollution exposure (e.g., vascular dementia or Alzheimer’s disease).

Supplementary Material

1

Highlights.

  • Air pollution and neighborhood socioeconomic status (nSES) jointly impact cognition

  • Cognition was associated with diagnosis of depression and cardiovascular outcomes

  • Joint effects of air pollution and nSES on cognition mediated by depression

  • A smaller portion of the association was mediated by cardiovascular comorbidities

Sources of support:

This work was based on information from the Emory Healthy Aging study, supported by the National Institute on Aging (NIA) [R01AG070937 (Lah) and R01AG079170 (Hüls/Wingo)] and a HERCULES Pilot Project via National Institute of Environmental Health Sciences (NIEHS) P30ES019776 (Hüls). Grace M. Christensen was supported by the National Institute of Environmental Health Sciences under Award Number 5T32ES12870.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Declaration of interests

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.

References

  1. Ailshire JA, & Clarke P (2015). Fine Particulate Matter Air Pollution and Cognitive Function Among U.S. Older Adults. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 70(2), 322–328. 10.1093/geronb/gbu064 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Ali NA, & Khoja A (2019). Growing Evidence for the Impact of Air Pollution on Depression. The Ochsner Journal, 19(1), 4. 10.31486/toj.19.0011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Altuğ H, Fuks KB, Hüls A, Mayer A-K, Tham R, Krutmann J, & Schikowski T (2020). Air pollution is associated with depressive symptoms in elderly women with cognitive impairment. Environment International, 136, 105448. 10.1016/j.envint.2019.105448 [DOI] [PubMed] [Google Scholar]
  4. Amariglio RE, Donohue MC, Marshall GA, Rentz DM, Salmon DP, Ferris SH, Karantzoulis S, Aisen PS, Sperling RA, & Alzheimer’s Disease Cooperative Study. (2015). Tracking early decline in cognitive function in older individuals at risk for Alzheimer disease dementia: The Alzheimer’s Disease Cooperative Study Cognitive Function Instrument. JAMA Neurology, 72(4), 446–454. 10.1001/jamaneurol.2014.3375 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Anstey KJ, Lipnicki DM, & Low L-F (2008). Cholesterol as a risk factor for dementia and cognitive decline: A systematic review of prospective studies with meta-analysis. The American Journal of Geriatric Psychiatry: Official Journal of the American Association for Geriatric Psychiatry, 16(5), 343–354. 10.1097/JGP.0b013e31816b72d4 [DOI] [PubMed] [Google Scholar]
  6. Christensen GM, Li Z, Pearce J, Marcus M, Lah JJ, Waller LA, Ebelt S, & Hüls A (2022). The complex relationship of air pollution and neighborhood socioeconomic status and their association with cognitive decline. Environment International, 167, 107416. 10.1016/j.envint.2022.107416 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Clifford A, Lang L, Chen R, Anstey KJ, & Seaton A (2016). Exposure to air pollution and cognitive functioning across the life course—A systematic literature review. Environmental Research, 147, 383–398. 10.1016/j.envres.2016.01.018 [DOI] [PubMed] [Google Scholar]
  8. Cong X, Zhang J, Sun R, & Pu Y (2021). Indoor unclean fuel cessation linked with adult cognitive performance in China. Science of The Total Environment, 775, 145518. 10.1016/j.scitotenv.2021.145518 [DOI] [PubMed] [Google Scholar]
  9. Delgado-Saborit JM, Guercio V, Gowers AM, Shaddick G, Fox NC, & Love S (2021). A critical review of the epidemiological evidence of effects of air pollution on dementia, cognitive function and cognitive decline in adult population. Science of The Total Environment, 757, 143734. 10.1016/j.scitotenv.2020.143734 [DOI] [PubMed] [Google Scholar]
  10. Dobielska M, Bartosik NK, Zyzik KA, Kowalczyk E, & Karbownik MS (2022). Mechanisms of Cognitive Impairment in Depression. May Probiotics Help? Frontiers in Psychiatry, 13, 904426. 10.3389/fpsyt.2022.904426 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Espírito Santo LR, Faria TO, Silva CSO, Xavier LA, Reis VC, Mota GA, Silveira MF, Mill JG, & Baldo MP (2019). Socioeconomic status and education level are associated with dyslipidemia in adults not taking lipid-lowering medication: A population-based study. International Health, ihz089. 10.1093/inthealth/ihz089 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Fan S-J, Heinrich J, Bloom MS, Zhao T-Y, Shi T-X, Feng W-R, Sun Y, Shen J-C, Yang Z-C, Yang B-Y, & Dong G-H (2020). Ambient air pollution and depression: A systematic review with meta-analysis up to 2019. Science of The Total Environment, 701, 134721. 10.1016/j.scitotenv.2019.134721 [DOI] [PubMed] [Google Scholar]
  13. Formánek T, Csajbók Z, Wolfová K, Kučera M, Tom S, Aarsland D, & Cermakova P (2020). Trajectories of depressive symptoms and associated patterns of cognitive decline. Scientific Reports, 10(1), 20888. 10.1038/s41598-020-77866-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Forte G, De Pascalis V, Favieri F, & Casagrande M (2019). Effects of Blood Pressure on Cognitive Performance: A Systematic Review. Journal of Clinical Medicine, 9(1), 34. 10.3390/jcm9010034 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Giorgini P, Di Giosia P, Grassi D, Rubenfire M, Brook RD, & Ferri C (2016). Air Pollution Exposure and Blood Pressure: An Updated Review of the Literature. Current Pharmaceutical Design, 22(1), 28–51. 10.2174/1381612822666151109111712 [DOI] [PubMed] [Google Scholar]
  16. Goetz ME, Hanfelt JJ, John SE, Bergquist SH, Loring DW, Quyyumi A, Clifford GD, Vaccarino V, Goldstein F, Johnson TM 2nd, Kuerston R, Marcus M, Levey AI, & Lah JJ (2019). Rationale and Design of the Emory Healthy Aging and Emory Healthy Brain Studies. Neuroepidemiology, 53(3–4), 187–200. 10.1159/000501856 [DOI] [PubMed] [Google Scholar]
  17. 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 Neurology, 77(7), 801–809. 10.1001/jamaneurol.2019.4914 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Grotto I, Huerta M, & Sharabi Y (2008). Hypertension and socioeconomic status. Current Opinion in Cardiology, 23(4), 335–339. 10.1097/HCO.0b013e3283021c70 [DOI] [PubMed] [Google Scholar]
  19. Hajat A, Hsia C, & O’Neill MS (2015). Socioeconomic Disparities and Air Pollution Exposure: A Global Review. Current Environmental Health Reports, 2(4), 440–450. 10.1007/s40572-015-0069-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Ilango SD, Chen H, Hystad P, van Donkelaar A, Kwong JC, Tu K, Martin RV, & Benmarhnia T (2020). The role of cardiovascular disease in the relationship between air pollution and incident dementia: A population-based cohort study. International Journal of Epidemiology, 49(1), 36–44. 10.1093/ije/dyz154 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Jack CR Jr., Bennett DA, Blennow K, Carrillo MC, Dunn B, Haeberlein SB, Holtzman DM, Jagust W, Jessen F, Karlawish J, Liu E, Molinuevo JL, Montine T, Phelps C, Rankin KP, Rowe CC, Scheltens P, Siemers E, Snyder HM, … Silverberg N (2018). NIA-AA Research Framework: Toward a biological definition of Alzheimer’s disease. Alzheimer’s & Dementia, 14(4), 535–562. 10.1016/j.jalz.2018.02.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Jakobsen AL, Jørgensen A, Tølbøll L, & Johnsen SB (2022). Opening the black box of the relationship between neighborhood socioeconomic status and mental health: Neighborhood social-interactive characteristics as contextual mechanisms. Health & Place, 77, 102905. 10.1016/j.healthplace.2022.102905 [DOI] [PubMed] [Google Scholar]
  23. Jenkins KR, & Ofstedal MB (2014). The association between socioeconomic status and cardiovascular risk factors among middle-aged and older men and women. Women & Health, 54(1), 15–34. 10.1080/03630242.2013.858098 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Jennings JR, Muldoon MF, Ryan C, Price JC, Greer P, Sutton-Tyrrell K, van der Veen FM, & Meltzer CC (2005). Reduced cerebral blood flow response and compensation among patients with untreated hypertension. Neurology, 64(8), 1358–1365. 10.1212/01.WNL.0000158283.28251.3C [DOI] [PubMed] [Google Scholar]
  25. Jessen F, Amariglio RE, van Boxtel M, Breteler M, Ceccaldi M, Chételat G, Dubois B, Dufouil C, Ellis KA, van der Flier WM, Glodzik L, van Harten AC, de Leon MJ, McHugh P, Mielke MM, Molinuevo JL, Mosconi L, Osorio RS, Perrotin A, … Subjective Cognitive Decline Initiative (SCD-I) Working Group. (2014). A conceptual framework for research on subjective cognitive decline in preclinical Alzheimer’s disease. Alzheimer’s & Dementia: The Journal of the Alzheimer’s Association, 10(6), 844–852. 10.1016/j.jalz.2014.01.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Jorm AF (2000). Is Depression a Risk Factor for Dementia or Cognitive Decline? Gerontology, 46(4), 219–227. 10.1159/000022163 [DOI] [PubMed] [Google Scholar]
  27. Kilian J, & Kitazawa M (2018). The emerging risk of exposure to air pollution on cognitive decline and Alzheimer’s disease—Evidence from epidemiological and animal studies. Biomedical Journal, 41(3), 141–162. 10.1016/j.bj.2018.06.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Kioumourtzoglou M-A, Power MC, Hart JE, Okereke OI, Coull BA, Laden F, & Weisskopf MG (2017). The Association Between Air Pollution and Onset of Depression Among Middle-Aged and Older Women. American Journal of Epidemiology, 185(9), 801–809. 10.1093/aje/kww163 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Leng B, Jin Y, Li G, Chen L, & Jin N (2015). Socioeconomic status and hypertension: A meta-analysis. Journal of Hypertension, 33(2), 221–229. 10.1097/HJH.0000000000000428 [DOI] [PubMed] [Google Scholar]
  30. Li Z, Christensen GM, Lah JJ, Marcus M, Russell AG, Ebelt S, Waller LA, & Hüls A (2022). Neighborhood characteristics as confounders and effect modifiers for the association between air pollution exposure and subjective cognitive functioning. Environmental Research, 212(Pt A), 113221. 10.1016/j.envres.2022.113221 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Lim Y-H, Kim H, Kim JH, Bae S, Park HY, & Hong Y-C (2012). Air Pollution and Symptoms of Depression in Elderly Adults. Environmental Health Perspectives, 120(7), 1023–1028. 10.1289/ehp.1104100 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Mair C, Roux AVD, & Galea S (2008). Are neighbourhood characteristics associated with depressive symptoms? A review of evidence. Journal of Epidemiology & Community Health, 62(11), 940–946. 10.1136/jech.2007.066605 [DOI] [PubMed] [Google Scholar]
  33. Manisalidis I, Stavropoulou E, Stavropoulos A, & Bezirtzoglou E (2020). Environmental and Health Impacts of Air Pollution: A Review. Frontiers in Public Health, 8, 14. 10.3389/fpubh.2020.00014 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. McEwen BS, & Tucker P (2011). Critical Biological Pathways for Chronic Psychosocial Stress and Research Opportunities to Advance the Consideration of Stress in Chemical Risk Assessment. American Journal of Public Health, 101(Suppl 1), S131–S139. 10.2105/AJPH.2011.300270 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. McGuinn LA, Schneider A, McGarrah RW, Ward-Caviness C, Neas LM, Di Q, Schwartz J, Hauser ER, Kraus WE, Cascio WE, Diaz-Sanchez D, & Devlin RB (2019). Association of long-term PM2.5 exposure with traditional and novel lipid measures related to cardiovascular disease risk. Environment International, 122, 193–200. 10.1016/j.envint.2018.11.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Messer LC, Laraia BA, Kaufman JS, Eyster J, Holzman C, Culhane J, Elo I, Burke JG, & O’Campo P (2006). The Development of a Standardized Neighborhood Deprivation Index. Journal of Urban Health, 83(6), 1041–1062. 10.1007/s11524-006-9094-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Miller MR, & Newby DE (2020). Air pollution and cardiovascular disease: Car sick. Cardiovascular Research, 116(2), 279–294. 10.1093/cvr/cvz228 [DOI] [PubMed] [Google Scholar]
  38. Moretti R, Torre P, Antonello RM, Manganaro D, Vilotti C, & Pizzolato G (2008). Risk factors for vascular dementia: Hypotension as a key point. Vascular Health and Risk Management, 4(2), 395–402. 10.2147/vhrm.s2434 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Murman DL (2015). The Impact of Age on Cognition. Seminars in Hearing, 36(3), 111–121. 10.1055/s-0035-1555115 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Oi K, & Haas S (2019). Cardiometabolic Risk and Cognitive Decline: The Role of Socioeconomic Status in Childhood and Adulthood. Journal of Health and Social Behavior, 60(3), 326–343. 10.1177/0022146519867924 [DOI] [PubMed] [Google Scholar]
  41. Pearce JL, Waller LA, Chang HH, Klein M, Mulholland JA, Sarnat JA, Sarnat SE, Strickland MJ, & Tolbert PE (2014). Using self-organizing maps to develop ambient air quality classifications: A time series example. Environmental Health, 13(1), 56. 10.1186/1476-069X-13-56 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Perini G, Cotta Ramusino M, Sinforiani E, Bernini S, Petrachi R, & Costa A (2019). Cognitive impairment in depression: Recent advances and novel treatments. Neuropsychiatric Disease and Treatment, 15, 1249–1258. 10.2147/NDT.S199746 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Peters R, Xu Y, Antikainen R, Beckett N, Gussekloo J, Jagger C, Jukema JW, Keinanen-Kiukaanniemi S, Rydén L, Skoog I, Staessen JA, Thijs L, Trompet S, Tully PJ, Tzourio C, & Anstey KJ (2021). Evaluation of High Cholesterol and Risk of Dementia and Cognitive Decline in Older Adults Using Individual Patient Meta-Analysis. Dementia and Geriatric Cognitive Disorders, 50(4), 318–325. 10.1159/000519452 [DOI] [PubMed] [Google Scholar]
  44. Power MC, Adar SD, Yanosky JD, & Weuve J (2016). Exposure to air pollution as a potential contributor to cognitive function, cognitive decline, brain imaging, and dementia: A systematic review of epidemiologic research. Neurotoxicology, 56, 235–253. 10.1016/j.neuro.2016.06.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Qiu C, & Fratiglioni L (2015). A major role for cardiovascular burden in age-related cognitive decline. Nature Reviews Cardiology, 12(5), Article 5. 10.1038/nrcardio.2014.223 [DOI] [PubMed] [Google Scholar]
  46. Qiu X, Danesh-Yazdi M, Wei Y, Di Q, Just A, Zanobetti A, Weisskopf M, Dominici F, & Schwartz J (2022). Associations of short-term exposure to air pollution and increased ambient temperature with psychiatric hospital admissions in older adults in the USA: A case–crossover study. The Lancet Planetary Health, 6(4), e331–e341. 10.1016/S2542-5196(22)00017-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Richardson R, Westley T, Gariépy G, Austin N, & Nandi A (2015). Neighborhood socioeconomic conditions and depression: A systematic review and meta-analysis. Social Psychiatry and Psychiatric Epidemiology, 50(11), 1641–1656. 10.1007/s00127-015-1092-4 [DOI] [PubMed] [Google Scholar]
  48. Robins JM, & Greenland S (1992). Identifiability and Exchangeability for Direct and Indirect Effects. Epidemiology, 3(2), 143–155. [DOI] [PubMed] [Google Scholar]
  49. Rudnicka E, Napierała P, Podfigurna A, Męczekalski B, Smolarczyk R, & Grymowicz M (2020). The World Health Organization (WHO) approach to healthy ageing. Maturitas, 139, 6–11. 10.1016/j.maturitas.2020.05.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Samet JM, Zeger SL, Dominici F, Curriero F, Coursac I, Dockery DW, Schwartz J, & Zanobetti A (2000). The National Morbidity, Mortality, and Air Pollution Study. Part II: Morbidity and mortality from air pollution in the United States. Research Report (Health Effects Institute), 94(Pt 2), 5–70; discussion 71–79. [PubMed] [Google Scholar]
  51. Semkovska M, Quinlivan L, O’Grady T, Johnson R, Collins A, O’Connor J, Knittle H, Ahern E, & Gload T (2019). Cognitive function following a major depressive episode: A systematic review and meta-analysis. The Lancet. Psychiatry, 6(10), 851–861. 10.1016/S2215-0366(19)30291-3 [DOI] [PubMed] [Google Scholar]
  52. Senthilkumar N, Gilfether M, Metcalf F, Russell AG, Mulholland JA, & Chang HH (2019). Application of a Fusion Method for Gas and Particle Air Pollutants between Observational Data and Chemical Transport Model Simulations Over the Contiguous United States for 2005–2014. International Journal of Environmental Research and Public Health, 16(18), Article 18. 10.3390/ijerph16183314 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Spiller E, Proville J, Roy A, & Muller NZ (n.d.). Mortality Risk from PM2.5: A Comparison of Modeling Approaches to Identify Disparities across Racial/Ethnic Groups in Policy Outcomes. Environmental Health Perspectives, 129(12), 127004. 10.1289/EHP9001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Tang F, Li K, Rosso AL, Jiang Y, & Li M (2022). Neighborhood segregation, socioeconomic status, and cognitive function among older Chinese immigrants. Journal of the American Geriatrics Society. 10.1111/jgs.18167 [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Vert C, Sánchez-Benavides G, Martínez D, Gotsens X, Gramunt N, Cirach M, Molinuevo JL, Sunyer J, Nieuwenhuijsen MJ, Crous-Bou M, & Gascon M (2017). Effect of long-term exposure to air pollution on anxiety and depression in adults: A cross-sectional study. International Journal of Hygiene and Environmental Health, 220(6), 1074–1080. 10.1016/j.ijheh.2017.06.009 [DOI] [PubMed] [Google Scholar]
  56. Weuve J, Bennett EE, Ranker L, Gianattasio KZ, Pedde M, Adar SD, Yanosky JD, & Power MC (2021). Exposure to Air Pollution in Relation to Risk of Dementia and Related Outcomes: An Updated Systematic Review of the Epidemiological Literature. Environmental Health Perspectives, 129(9), 096001. 10.1289/EHP8716 [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Yang T, Wang J, Huang J, Kelly FJ, & Li G (2023). Long-term Exposure to Multiple Ambient Air Pollutants and Association With Incident Depression and Anxiety. JAMA Psychiatry, 80(4), 305–313. 10.1001/jamapsychiatry.2022.4812 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Zijlema WL, Triguero-Mas M, Smith G, Cirach M, Martinez D, Dadvand P, Gascon M, Jones M, Gidlow C, Hurst G, Masterson D, Ellis N, van den Berg M, Maas J, van Kamp I, van den Hazel P, Kruize H, Nieuwenhuijsen MJ, & Julvez J (2017). The relationship between natural outdoor environments and cognitive functioning and its mediators. Environmental Research, 155, 268–275. 10.1016/j.envres.2017.02.017 [DOI] [PubMed] [Google Scholar]
  59. Zong Z, Zhang M, Xu K, Zhang Y, & Hu C (2022). Association between Short-Term Exposure to Ozone and Heart Rate Variability: A Systematic Review and Meta-Analysis. International Journal of Environmental Research and Public Health, 19(18), 11186. 10.3390/ijerph191811186 [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
NIHMS1974466-supplement-1.docx (816.1KB, docx)

RESOURCES