Air Pollution and Parkinson’s Disease

Abstract

Parkinson’s disease (PD) is the second most common neurodegenerative disease of unclear etiology that is thought to be caused by a combination of genetic and environmental factors. Air pollution, the largest environmental health risk globally, has been suggested to be associated with PD risk, while not all results are uniform. In this chapter, we summarize the recent advances in the epidemiology of six criteria air pollutants-fine particulate matter (PM2.5), inhalable particles (PM10), nitrogen dioxide (NO2), sulfur dioxide (SO2), carbon monoxide(CO), and ozone exposure with PD risk, and provided an overview of the potential mechanisms of air pollution on PD. Based on the current evidence from the human’s studies and animal models, this chapter provides a novel insight for the understanding of how environmental exposure influences the pathogenesis of neurodegeneration and prevents the occurrence or development of PD.

1. Introduction

Air pollution, a major cause of premature death and disease, is the largest environmental health risk globally [1, 2]. According to the Global Burden of Disease (GBD) study in 2019, air pollution has been ranked as the fourth death cause in the world [3], which poses a heavy threat to both individuals and society. With rapid urbanization and industrialization, air pollution levels have continuously increased over the past decades. Parkinson’s disease (PD), the second most common neurodegenerative disease, is characterized by the pathological accumulation of proteins, inflammation, and neuron loss [4]. It is reported that the global prevalence of PD doubled over the next 30 years [4]. Recent epidemiological studies have found that air pollutants exposure is significantly associated with the increased risk of PD, while not all results are uniform [5, 6, 7, 8]. The variability among these studies is likely attributed to the measurement of air pollutants, exposure assessment and duration, and correction for other confounding. In this chapter, we summarize the recent advances in the epidemiology of air pollution exposure and PD, including the evidence of the effects of six criteria air pollutants-fine particulate matter (PM_2.5), inhalable particles (PM₁₀), nitrogen dioxide (NO₂), sulfur dioxide (SO₂), carbon monoxide (CO), and ozone with PD, and provided a narrative review of the potential mechanisms of air pollution on PD risk. This chapter serves to provide a novel insight for the understanding of how environmental exposure influences the pathogenesis of neurodegeneration and reduces the risk of PD.

2. Air pollution: sources, composition, and monitoring

Air pollution is a heterogeneous mixture composed of particulate matter (PM) and gaseous components that are released directly into the atmosphere (primary pollutants) but also generated by reacting with other components (secondary pollutants) that vary both temporally and spatially [9]. Expansion of industry and vehicle traffic had a major impact on the overall air quality in urban areas over the past decades. In addition, indoor solid fuels burning, agricultural production, and waste incineration also contribute to negative effects on air quality.

Urban airborne PM typically can be described by four modes (Figure 1), which primarily composed of PM₁₀ (particles <10 μm in diameter), PM_2.5 (particles <2.5 μm in diameter), and ultrafine particles (UFP, PM_0.1, particles <0.1 μm in diameter). However, only PM_2.5 and PM₁₀ are monitored by regulatory agencies in different countries (i.e., the Environmental Protection Agency in the USA) due to the “mass-based” estimation approach [10]. The sources of PM are diverse and mainly include anthropogenic causes (e.g., traffic gas emissions and fossil fuel burning) and natural sources such as dust storms, wildfires, and volcanic eruptions, which are related to climate change.

Figure 1.

The size distribution of airborne particulate matter (Adapted from USA. EPA 2004).

Nitrogen oxides (NOx), SO₂, CO, and ozone are the critical gaseous components in the atmosphere. NOx is a mixture of gases that consists of nitrogen and oxygen, such as NO₂ and nitric oxide (NO) that can be produced by traffic or indoor cooking stoves or the burning of coal, oil, or natural gas. The sources of SO₂ are primarily from burning fossil fuels, non-ferrous metal smelting, steel, and industry production process. CO is produced by motor vehicle emissions, industrial boilers, and waste incineration. Ozone, a secondary pollutant, is formed when NOx reacts with volatile organic carbons (VOCs) and oxygen in the presence of heat and light. Its main sources include traffic gas, factories, and electric utilities. To accelerate the control of air pollution, a series of air quality standards for the primary air pollutants have been established in the USA, EU, and China (Table 1).

Pollutants	Averaging period	USA	E.U.	China
Sulfur dioxide	1 hour	75 ppb	350 μg/m^3	150 μg/m^3
Carbon monoxide	8 hour	10 mg/m^3	10 μg/m^3	10 mg/m^3a
Nitrogen dioxide	1 hour	100 ppb	200 μg/m^3	200 μg/m^3
Ozone	8 hour	0.07 ppm	120 μg/m^3	100 μg/m^3
PM2.5	1 year	15 μg/m^3	25 μg/m^3	15 μg/m^3
PM10	1 day	150 μg/m^3	50 μg/m^3	50 μg/m^3
Lead	1 year	0.15 μg/m^3	0.5 μg/m^3	0.5 μg/m^3

Table 1.

Air quality standards in the USA, EU, and in China.

^a

The averaging period for carbon monoxide is 1 hour.

In general, ambient air pollutants are monitored by traditional monitoring methods with fixed air stations, while they are expensive, sparsely distributed, and require high maintenance. To overcome the limitations in the traditional methods, a growing body of relatively new approaches including land-use regression (LUR) models, satellite remote sensing, and disperse models have been developed to estimate the air pollution levels with high spatial resolution. In the LUR modeling approach, levels of vehicle exhaust markers, such as NOx and PM, are measured simultaneously at many locations throughout an urban area using relatively inexpensive passive monitors. Various geographic information system (GIS) parameters (such as traffic and roadway density, land use, and population density) are used to predict the measured concentrations. For satellite-based models, Shi et al. used the V4.CH.02 product of the Dalhousie University Atmospheric Composition Analysis Group to simulate the annual average PM_2.5 and its chemical constituents at approximately 0.01° × 0.01° resolution [11]. This dataset combines satellite retrievals and simulation of aerosol optical depth (AOD) from multiple sources (MISR, MODIS Dark Target, MODIS, and SeaWiFS Deep Blue, MODIS MAIAC, and GEOS-Chem), chemical transported models, and near-surface PM_2.5 concentrations [11, 12]. Additionally, to measure an individual’s exposure of air pollution, low-cost sensors with real-time monitoring are becoming popular in the field of environmental epidemiology.

3. Epidemiological findings of air pollution with PD risk

3.1 Particulate matter and PD risk

PM has been suggested to be related to the risk of PD, while the evidence from different study designs is inconsistent. Among others, PM_2.5 is the most consistent and robust indicator of air pollution [11]. A prospective cohort study of 2.19 million participants in Ontario showed that long-term exposure to PM_2.5 was associated with a 4% increase in incident PD (95%CI:1.01, 1.08) [7]. Another nationwide cohort study among 63.03 million individuals aged ≥65 years in the USA found that annual PM_2.5 exposure was positively associated with an increased risk of first hospital admission for PD (hazard ratio, HR = 1.13, 95%CI:1.12, 1.14) [13]. Regarding the short-term effects of PM_2.5, Zanobetti et al. found that each 10 μg/m³ increase in lag 2-day average PM_2.5 was positively associated with the hospitalization risk of PD (3.23%, 95%CI:1.08, 5.43) [14]. Moreover, a study performed in New York State to explore the chemical constituents of PM_2.5 with PD hospitalization by using a satellite-based model at 1 km × 1 km spatial resolution and the results showed that organic matter (OM) and nitrate exposure significantly increased the risk of PD aggravation [15]. However, the associations of PM exposure are not uniform in some studies [6, 16]. A large cohort from the Health Professionals Follow-up Study (HPFS) showed that exposure to ambient PM₁₀ or PM_coarse was not significantly associated with PD risk among US men [16]. A recent meta-analysis conducted in 2020 involving 10,077,029 participants presented no significant relationship between long-term PM₁₀ exposure and PD incidence. With every 10 μg/m³ increment, the relative risks (RRs) and 95% CIs were 1.01 (0.97, 1.05) for PM₁₀ exposure [17].

3.2 Nitrogen oxides with PD risk

Nitrogen oxides (NOx) are recognized as traffic-related air pollutants, which are released during any high-temperature combustion and then rapidly converted to NO₂. The associations found between NO₂ exposure and PD risk are mixed. In a Denmark study, Ritz et al. used a dispersion model to estimate the NO₂ exposure and the results showed positive relationships with PD risk, with a 9% higher risk (95% CI:3, 16.0%) [18]. Similar associations were also observed in a nationally representative cohort of 78,830 individuals [19], the findings indicated that long-term exposure to ambient NO₂ was related to an increased risk of PD in Korea (HR for highest vs. lowest quartile, 1.41; 95% CI: 1.02, 1.95; P_trend = 0.045). However, a matched case-control study in the Netherlands reported no significant association between 16 years of residential exposure to ambient NO₂ and the development of PD (aOR = 0.87, 95% CI: 0.54, 1.41) [6]. Moreover, a nested case–control study in Taiwan investigated multiple chemical compounds exposure with the incidence of PD, and no significant associations for NO₂, NO, and NO_X exposure were observed [8]. A recent meta-analysis summarized the studies of ambient air pollution with PD risk, the pooled odds ratio (OR) for the effect of NO₂ (per 1 μg/m³) on PD was 1.01 (95%CI,1.00, 1.02, I² = 69%) [20].

3.3 Sulfur dioxide with PD risk

Studies of the association between SO₂ exposure and PD risk are relatively fewer. Most studies reported no significant effects of ambient SO₂ exposure on the risk of PD. A nested case-control study performed in Taiwan using the National Health Insurance Research Dataset (NHIRD) explored the multiple air pollutants exposure with incident PD, and the findings showed no significant relationship for SO₂ exposure. Jo et al. [19] used the data from the Korea National Health Insurance Service and estimated the pollutants levels based on the nearest air monitoring stations to examine such association. The results found no statistically significant relationship between long-term exposure to ambient SO₂ and incident PD (HR = 1.02, 95%CI: 0.74, 1.41). Another population-based cohort study was also conducted in Korea, which examined the short-term effects of air pollutants exposure on PD aggravation, with a positive association. Each unit increase in the 8-day moving average of SO₂level was significantly related to PD aggravation (OR = 1.54, 95%CI:1.11, 2.14 per 1 ppb) [21]. As reported in a recent meta-analysis in 2022, the pooled association was not statistically significant for SO₂ exposure [20].

3.4 Carbon monoxide with PD risk

Overall, the research on CO has demonstrated an effect on the risk of PD, but there are inconsistencies across studies. A population-based case-control study in Taiwan suggested that traffic-related pollutants exposure such as NOx and CO increased PD risk in Taiwanese population. The multi-pollutant models showed that the OR was 1.17 (1.07, 1.27) for ambient CO above the 75th percentile exposure compared with the lowest percentile [22]. A case-crossover study in Seoul found that short-term exposure to ambient CO significantly related to the risk of PD aggravation. The OR was 1.46 (95%CI, 1.05, 2.04) for each 0.1 ppm increment in the 8-day moving average of CO concentrations [21]. Conversely, some studies found no significant effects of ambient CO exposure on PD risk [8, 19]. Given the inconsistencies of studies, a systematic review and meta-analysis were conducted in 2019 to summarize the results from 10 studies. The pooled association indicated that the risk of PD was 1.65 (1.10, 2.48) for each 1 ppm increment of ambient CO exposure [23].

3.5 Ozone with PD risk

Studies of ambient ozone exposure with the risk of PD have been reported but remain inconclusive. Several studies found significant associations between exposure to ozone and PD risk [7, 24]. Zhao et al. [24] estimated the ambient average level of ozone by a combination of chemical transport models and ground measurement. The association analysis indicated that long-term exposure to ozone was significantly related to the increased risk of mortality due to PD in Canada (HR = 1.09, 95%CI:1.04, 1.14). A systematic review and meta-analysis included 21 studies with 222,051 individuals who found that ozone exposure might contribute to a higher risk of PD. The pooled results presented that with each 10 μg/m³ increase in the concentration of ozone, the adjusted RR was 1.01 (95%CI: 1.00, 1.02) [25]. Furthermore, a most up-to-date meta-analysis by Dhiman et al. further confirms the positive association of ozone exposure [20].

4. Major study design in air pollution epidemiology

4.1 Studies on acute health effects

In the field of air pollution epidemiology, time-series study, case-crossover study, and panel study are often used to assess the acute health effects of pollutants. A time-series design by controlling for confounders that do not vary temporally but can only address short-term acute effects [26]. In recent years, time-series study has been widely applied to investigate the short-term exposure to air pollution on various health endpoints. Similar findings have been documented in different areas with different air pollution backgrounds, as well as in different study populations around the world [27, 28, 29]. This method is particularly advantageous where the catchment area is unclear, for hospital-based studies in densely populated areas where not all hospitals can be included, counts of admissions or outpatients might be comparable for high-polluted versus low-polluted days. Therefore, time-based comparisons within a population are useful for assessing acute health effects from community-wide exposures and may provide more valid estimates than comparisons between communities. The statistic methods of the generalized additive model (GAM) or Poisson regression models are often applied to analyze in these studies, whereas temperature and relative humidity are controlled for potential meteorological effects [30].

Case cross-over study is another typical design used for examining the acute effects of air pollution. The concept of “case-crossover study” is firstly named by Dr. Maclure at the Harvard University in 1991. The key feature of this study design is that each case serves its own control. The method is analogous to a crossover experiment viewed retrospectively, except that researcher does not control when a patient starts and stops being exposed to the possible trigger. Furthermore, the exposure frequency is measured in only a sample of the total period when the patient was at risk of the onset of disease [31]. Confounding from individual time-invariant characteristics is completely controlled, as the individual supplies his/her own referent periods. Compared with the time-series study, the strength of case-crossover study not only can control many potential confounding by its novel design rather than statistical models but also can help avoid many ethical issues.

Panel study, sample a set of fixed individuals on whom observations are made at regular time intervals, is usually completed in time. In recent decades, a growing number of epidemiological studies have applied the panel design to investigate the association of individual-level air pollutants exposure with health outcomes. Panel studies are usually performed over a short period, require intensive observation within this period, and have a relatively narrow focus. Thus, it may be difficult for recruitment and retention as they are demanding on participants, leading to issues with completeness and sample size [32]. Linear mixed models, mixed-effect models, or generalized estimating equations (GEEs) are usually used in this study design to examine the health effects of air pollutants.

4.2 Studies on long-term health effects

Compared with acute health effects, studies on long-term exposure of air pollutants are more common in the field of environmental epidemiology. Cohort studies, case-control studies, or cross-sectional studies are often used to analyze the relationship. The ecological study, a type of cross-sectional design, can be used to determine the regional characteristics of air pollution. Ecologic studies utilize group-level data on outcomes (i.e., rates of disease, prevalence proportions, or mean measurements) in relation to group-level data on exposures. With the emergence of some new technologies such as LUR model or satellite-based models, the estimation of the temporal and spatial changes of ambient concentrations of air pollutants has been improved [11, 24]. As reported in a recent model study, satellite-retrieved AOD provides a unique opportunity to characterize the long-term trends of ground-level PM_2.5 at high spatial resolution [33]. Given the ecological study is easily subject to ecological fallacy and confounding, results from ecological studies should be viewed critically.

Cohort study is widely recognized as an ideal approach to investigating the long-term effects of air pollutants exposure on health outcomes. However, due to the high expense and time-consuming nature, it brings many challenges to the practices of scientific research.

5. Potential mechanisms of air pollution on Parkinson’s disease

The primary pathology in PD involves dopaminergic neuron loss, particularly in the substantia nigra (SN), and systematic inflammation. In the context of air pollution, studies have indicated that diesel exhaust can increase α-synuclein (α-syn) levels and lead to neuroinflammation, which is associated with the development of PD. It is suggested that several pathways by which air pollutants can affect the central nervous system (CNS) and contribute to the pathogenesis of PD. The direct neurotoxicity and neuroinflammation, air pollutants-lung-brain connection, and changes of gut and microbiome play important roles in the occurrence and development of PD. The details of the possible mechanisms are summarized in Figure 2.

Figure 2.

The potential mechanisms of air pollutants exposure on risk of PD.

5.1 Direct neurotoxicity and neuroinflammation

It is believed that many components of air pollution can reach the brain and thus contribute to the pathogenesis of PD by direct neurotoxicity and/or neuroinflammation. The two fractions of PM-PM_2.5 and UFP are predominantly implicated in CNS effects. Both of them are acutely toxic to cardiovascular and lung tissue. Given the small particle size and high activity, these particles can cross the blood-air barrier of the lungs, gaining access to peripheral circulation and the brain. The nasal olfactory pathway is suggested to be a critical portal of entry, where inhaled UFP reaches trigeminal nerves, brainstem, and hippocampus [34, 35]. Furthermore, a growing number of studies have indicated these PMs can enter the brain and may be related to neurodegenerative pathology in vivo [36, 37, 38]. Additionally, evidence has shown that the concentrations of some polycyclic aromatic hydrocarbons (PAHs) in human brains are very high. This can help support the concept that specific components of air pollution can bioaccumulate in the nervous system, poses significant risks through direct neurotoxicity [39].

Oxidative stress and inflammation have been associated with the neurodegenerative process including PD. In vitro studies have reported that PM exposure can induce inflammation in airway epithelial cells mediated by oxidative stress [40]. It is believed that oxidative stress plays a role in PD pathogenesis and can cause α-syn aggregation [41]. Pathological α-syn aggregates appear to spread throughout the CNS in a predictable manner, which determines the clinical symptom of PD. Mitochondrial damage has also been implicated in the development of PD and can lead to oxidative stress and neuronal loss. Dysfunction of mitochondria promotes the formation of reactive oxygen species (ROS), which can induce α-syn aggregation [42]. In addition, animal models showed that systematic inflammation could result in neuroinflammation and loss of dopaminergic neurons, especially in combination with increased α-syn levels, which posed an elevated risk for incident PD.

5.2 Lung-brain connection

Exposure to air pollution also leads to peripheral or systematic inflammation, which, in turn, can contribute to CNS inflammation. Epithelial cells lining the airway can physically block larger size PM and secrete cytokines including tumor necrosis factor-α (TNF-α) and interleukin-1 beta (IL-1β), which promote the synthesis of other cytokines and lead to immune cell activation. Just as an example, chronic low-level inflammation is linked to multiple systemic injections of low lipopolysaccharide (LPS), a cell wall component of Gram-negative bacteria that is a potent pro-inflammatory stimulus, rendering animals more vulnerable to further pro-inflammatory insult [43]. The blood-brain barrier (BBB) is weakened by systemic inflammation-derived proinflammatory cytokines and chemokines, which can allow proinflammatory cytokines and inflammatory cells to enter the brain. Once passing to the brain, these factors from the periphery, along with brain-derived cytokines, chemokines, α-syn, and amyloid precursor proteins, can activate CNS immune cells and induce downstream effects [44, 45]. In brief, air pollutants exposure induces a systemic inflammatory response, which may lead to neuroinflammation and an elevated risk of PD.

5.3 Gut-brain connection

Studies from both animal experiments and humans have supported the hypothesis that pathological α-syn can accumulate in the gut, spread to the brainstem by the vagus nerve, and eventually induce neuronal loss in the SN. To our knowledge, there is little direct evidence of air pollutants can induce α-syn aggregates in the gut, which spreads to the CNS, while there is an increasing body of publications demonstrating that air pollutants can change the gut mucosa, which is thought to promote α-syn pathology. In animal studies, α-syn preformed fibrils injected into the duodenum induces α-syn spread into brainstem nuclei and then to the SN [46]. Moreover, it is reported that air pollution exposure can induce inflammation and leakiness in the gut. It may be a trigger for α-syn aggregates and alter the risk of inflammatory bowel disease (IBD) [44]. To date, it remains unclear how air pollutants exert these changes in the gut, while they may act or partly affect by altering the microbiome.

5.4 Microbiome-brain connection

Microbiome has recently encountered the interest of neuroscience, which may be associated with the risk of PD. Air pollutants exposure has been linked to alterations in the microbiome, while primarily in animals. The imbalances in the microbiome can lead to disruption of the epithelial barrier of the gut and allow various bacterial metabolites and virulence factors to pass through the intestinal lining, enter the bloodstream and subsequently across the BBB. Metabolites of gut microbes, such as short-chain fatty acids (SCFA, i.e., acetate, propionate, and butyrate), have also been suggested to activate microglia and increase neuroinflammation [47]. In addition, some molecules including synaptogenic proteins, SCFA, levodopa (L-dopa), γ-aminobutyric acid, and serotonin can be produced, suppressed, and overused by strains of microbiota. These changes can have a direct effect on neurological function. As reported in a mice model, PD-derived gut microbiota may exacerbate α-syn-mediated motor impairments and neuronal disease, whereas germ-free mice displayed milder α-syn pathology [48]. An experimental study in mice found that ambient PM_2.5 exposure exhibited significant changes in gut microbial diversity [49]. A significant increment was observed in Bacteroidales, which probably involved degradation of the mucous layer and elevated gut permeability [44]. Though the mice microbiome is different from that in humans, these studies may help provide implications for air pollution affecting the PD risk by changing the microbiome. Future studies are warranted to validate these findings.

6. Summary and future perspectives

The etiology of PD is complex but undoubtedly involves a combination of gene and environmental factors. To our knowledge, epidemiological investigations of air pollution exposure and PD risk have covered a wide variety of sizes and types of pollutants, populations, geographical locations, and related factors. Air pollution has been suggested to contribute to a significant percentage of PD cases worldwide. Several potential mechanisms include direct neural toxicity, CNS inflammation, and alternations in the microbiome may promote neurodegeneration and increase the risk of PD. However, this field is still in its early stage, some important questions remain unanswered, and some challenges are listed below.

1. PM_2.5 constituents with PD risk

   Since PM_2.5 is a complex mixture of more than 50 chemical constituents (such as black carbon, OM, water-soluble ions, and metals), different chemical constituents of PM_2.5 on PD risk as well as the possible mechanisms may be varied. It is meaningful for identifying the roles of various constituents of PM_2.5 played in the development of PD.

2. Critical exposure windows

   Given the development of PD is a chronic condition with a latency of about a decade [50], the critical windows of air pollutants exposure on PD risk needed to be identified in the future. With the rapid development of wearable personal monitoring sensors, we may foresee future possibilities for more precise and accurate personal exposure to pollutants, which can greatly benefit studies of air pollution and PD.

3. Interaction of gut microbiome and air pollutants

   The involvement of human gut microbe with PD risk is an exciting emergent field of research. Besides the inhalation pathway, the PMs can also be swallowed and end up in the gut. The particles can then induce systematic inflammation, or interact with the gut microbiome in other ways, thus potentially impacting PD risk. Future investigations are needed to explore the interaction between air pollutants and gut-microbiome concerning the risk of PD.

4. Gene modifications and vulnerable population

   As reported that the PD risk increased threefold when joint exposure to the AA genotype of the IL-1β gene and ambient NO₂ [22]. It would be interesting to explore the interactions of the gene (such as APOEε4) with air pollutants on PD risk. Additionally, it is significant to identify the vulnerable characteristics of the population in future studies.

We believe that the understanding of the relationship between air pollution and PD will improve in the coming years as more investigations are conducted and more reproducible findings are reported.

Conflict of interests

None declared.

Abbreviations

PD	Parkinson’s disease
PM2.5	fine particulate matter
PM10	inhalable particles
NO2	nitrogen dioxide
SO2	sulfur dioxide
CO	carbon monoxide
UFP	ultrafine particles
RR	relative risk
OR	odds ratio
HR	hazard risk
95%CI	95% confidence interval
α-syn	α-synuclein
CNS	central nervous system
BBB	blood-brain barrier
SN	substantia nigra
OM	organic matter
AOD	aerosol optical depth