Since the identification of general adaptation syndrome by Selye in 1936 [12], describing the impact of chronic stress on physiological systems, and the later establishment of allostatic theory by McEwen and colleagues [13,14,15], there has been a recognition that chronic stress influences multiple interlinked systems, producing pathophysiological changes that lead to chronic disease. Viewing these interlinked pathways as a network using a systems biology approach, the PINE model provides insights as to how behaviour of the entire network can explain transition from health to MDD and other NCDs [3]. This work was expanded to illustrate dynamic behaviour of the PINE network with chronic stress, including feedback loop transitions and a critical transition to disease [4,5]. The concept of a CIRN then proposed that the increased reciprocal risk between several major NCDs may reflect shared causative pathways within the PINE network; notably immune, autonomic and endocrine dysregulation [5].

Chronic stress acts on the autonomic nervous system (ANS) and HPA axis [16,17], resulting in sympathovagal imbalance, parasympathetic withdrawal [18,19] and sympathetic overdrive [19]. The long-term stimulation of the HPA axis results in cortisol resistance [16,19,20,21,22]. The negative regulatory feedback loops in the ANS and HPA axis that normally maintain homeostasis are therefore disrupted by chronic stress. Both vagal withdrawal and glucocorticoid resistance may ultimately push some negative feedback loops through a transition to positive feedback, driving a system-wide transition to disease [4]. Via neural and endocrine modulation of acute inflammation [23,24,25], immune function is intimately linked to this chronic stress response, increasing circulating inflammatory markers [3,4]. This systemic pro-inflammatory state promotes neuroinflammation, microglial proliferation and astrocyte loss [26,27,28], kynurenine pathway (KP) activation, tryptophan depletion and reduced CNS serotonin production [29]. Overactivity of the KP increases neurotoxic metabolites, inhibiting neurotrophic signalling and neuronal maintenance to further suppress astrocyte numbers [29,30,31]. Metabolite effects on hippocampal neurogenesis [32,33] in turn may influence learning [34] and memory [35]. These different CNS changes have all been linked to MDD: inflammation is strongly linked to MDD [36] and inflammatory mediators such as IL-1 drive sickness behaviours [37,38,39] that are identical to the neurovegetative symptoms and anhedonia of MDD [3,5,40]; reduced CNS serotonin and brain-derived neurotrophic factor (BDNF) are both linked to MDD [41], and shifts in neurogenesis and BDNF signalling may mediate cognitive changes in MDD [5,42].

Other systems constitute elements of the PINE network, notably the gut and gut microbiome. Dietary factors and inflammation affect gut motility [43] and permeability [44], facilitating the entry of bacteria and associated lipopolysaccharides (LPSs), inducing systemic pro-inflammatory cytokine release in a positive feedback loop [38,45,46]. Recent evidence indicates PM exposure also induces gut dysbiosis in animals and humans, promoting inflammation and metabolic disease [47,48,49]. The pro-inflammatory state arising with chronic stress increases leptin and reduces ghrelin secretion [50], leading to hypothalamic leptin resistance [51,52]. This central leptin resistance further increases circulating leptin levels, which in a positive feedback manner further enhances inflammatory cytokine release [53,54,55,56,57]. Through this array of changes, spanning multiple biological processes and pathways, long-term stress may give rise to a variety of chronic disorders [5]. We assert that PM may serve as an environmental trigger of PINE network disruption and then downstream, stress-related disease processes.

Inhaled PM triggers pulmonary protein leakage and inflammation in a size-, dose- and solubility-dependent manner [58,59]. The persistence of inhaled particulates may prolong these effects. Exposure to PM influences multiple organ systems indirectly via the production of inflammatory factors and triggering of nervous reflexes, and directly via access to the circulation where they disrupt cellular, tissue and organ function to modulate autonomic nervous system (ANS) activity [60,61,62,63]. The impact of PM on pro-inflammatory responses are associated with the size of the particles. Due to their size, PM10 are restricted to the pulmonary system, where they locally increase reactive oxygen species (ROS) and inflammatory cytokines [64]. PM2.5 are associated with elevated inflammation and oxidative stress, together with changes to ANS activity [10]. Exposure to PM2.5 elevates inflammatory markers including CRP, TNF-α, PGE2 and Il-1α together with ET-1 [65,66]. PM0.1 are also heavily implicated in airway diseases, but have also been shown to be associated with inflammation in the cardiovascular and central nervous system [11] (Nelin et al., 2012).

Chronic inflammation is linked to proximity to roadways, where there are increased PM levels [67]. Inhalation of poorly soluble pollutants such as ozone, NO₂ and phosgene also induces pulmonary inflammation [68,69], with changes focused within the lower respiratory tract [70,71]. Conversely, water soluble sulphur dioxide, chlorine and ammonia may predominantly influence the nasopharynx [70,72,73,74]. Volatile organic chemicals in vehicle and other pollutant sources are also detrimental to cardiovascular health [75], and lead to immunosuppression [76] and Alzheimer’s disease [77]. For example, acute acrolein exposure induces dyslipidaemia [78] and vascular dysfunction [79,80], while chronic exposure promotes atherosclerosis and lesion rupture [80,81] and impairs protective anti-infarct signalling [82].

Microparticulate exposure additionally triggers adipose tissue production/release of cytokines, contributing to systemic inflammation and hepatic insulin-resistance [71]. There is also evidence of immune cell mobilisation from bone marrow in response to PM exposure [83,84,85]. Despite these clear pro-inflammatory changes, mixed effects of air pollution on circulating cytokines are still reported, ranging from increases in humans and animals [86,87,88] to no apparent change [89,90,91].

These different peripheral changes and inflammatory processes contribute to a neurogenic inflammation with air pollution exposure [92,93]. Activation of neuronal and epithelial transient receptor potential proteins (TRPs) by pollutant products triggers a tachykinin (e.g., substance P, neurokinin A) release [94,95,96], whilst capsaicin-sensitive TRPV1 channels evoke a neuronal tachykinin release [97], and ion channel receptors trigger a neuronal neurokinin release [98,99]. This induces neurokinin-receptor-dependent pulmonary inflammation [100,101], with pulmonary injury and cytokine production spilling over into the systemic circulation, contributing to metabolic changes and propagating vascular and CNS inflammation [102,103]. Neuroinflammation, in turn, contributes to mental health issues such as MDD and other NCDs via perturbation of the PINE network [3,5].

Inflammatory mediators produced with pollutant exposure may activate glial cells in the CNS [104]. Interestingly, cytokine-mediated NFκB signalling in the hypothalamus appears to enhance particulate-induced systemic inflammation and metabolic changes [71], supporting positive feedback augmentation of inflammation via the CNS. Central stress-axis regions such as the paraventricular nucleus (PVN) of the hypothalamus may also be activated via sensory irritant activation of the trigeminal and vagal nerves with pulmonary oxidative stress/irritants, stimulating sympathetic and HPA-axis activities and hormone release.

These central effects of air pollutants are influenced by, and in turn modulate, systemic metabolic changes. Experimental studies in rodents identify links between long-term particulate exposure and pollution-triggered inflammation with adipose and hepatic inflammation, hepatic ER stress, glucose intolerance and insulin-resistance [103,105]. These observations are consistent with links between chronic particulate exposure and insulin-resistance, hyperglycaemia and diabetes in humans [106]. Studies of acute exposure to gaseous pollutants reveal that lipid oxidation by-products are elevated, promoting skeletal muscle insulin-resistance [107].

Though fewer studies implicate stress hormones and their receptors in environmental disease susceptibilities [108], there is substantial evidence of stress-axis modulation in response to air pollutants [58]. Ozone has been extensively studied as a specific airborne toxin; however, any air pollutant that interacts with biological components of the airway and triggers neural responses can stimulate a HPA-mediated stress response [58]. Virgolini and colleagues [109,110] demonstrated permanent alteration in HPA-axis function with lead exposure, and heavy metals can form a significant component of air pollution. Heavy metal concentration can also be found in bush fires, with fires in Lithuania being linked to increases in heavy metals such as copper, lead and zinc contaminating river water [111], while ash from Californian wildfires in 2007 was shown to contain substantial levels of arsenic, cadmium, copper and lead [112].

It has been argued that the environmental stress of air pollution and psychological stress may act synergistically in disrupting health. Clougherty and Kubzansky [113] provide a compelling case for air pollution and stress potentiating respiratory disease onset and severity, with a lucid discussion of the central role of the HPA-axis, while Olvera Alvarez et al. [114] propose that early life stress and air pollution act synergistically to increase the risk of chronic diseases, including MDD, CHD, T2DM and lung and brain cancer [115]. These authors suggest early life stress results in long-term modification of HPA-axis function (e.g., via NR3C1 glucocorticoid receptor gene methylation), rendering the HPA axis susceptible to dysfunction with environmental stressors such as air pollution [114]. Similarly, [3,5] suggest that early developmental diathesis contributes to later disease risk via the PINE network, a ramification of which is that people experiencing early life stress and subsequently exposed to events such as bushfires or elevated levels of PM may be at greater risk of multiple chronic illnesses, including MDD, CHD or T2DM.

Air pollution can be linked to these chronic disorders via inflammatory, chemical and ischemic influences on the ANS [116]. To measure cardiac autonomic control, heart rate variability (HRV) is a well-recognised non-invasive and quantitative marker reflecting rhythmic activity of the sinus node, which is analysed in time, frequency or non-linear domains [117]. While HRV measures generally reflect sympathovagal balance, observed HRV may be primarily driven by vagal activity [118,119,120], with reduced HRV linked to vagal withdrawal [18,19]). Reduced HRV is associated with poor health outcomes and is linked to conditions such as MDD and CHD [121] and pathophysiological changes in the PINE network (discussed above).

Air pollution, especially PM2.5, is associated with decreased HRV in different populations, including healthy young adults (particularly when simultaneously exposed to amplified noises) [122], healthy adult boilermakers [123], healthy senior adults [124,125,126], and those with cardiovascular diseases [127,128,129] also identified a decrease in HRV with PM exposure in elderly people, who were non-smokers and had no serious medical conditions. Additional smaller studies compared personal exposure to PM with HRV in healthy and diseased individuals [130,131,132]. For individuals, particularly the elderly, with certain underlying respiratory conditions, PM exposure can have up to a 4-fold increase in autonomic instability compared healthy young individuals [132]. Alongside reducing HRV, there is also limited evidence that PM increases coagulation [131]. Although limited by small sample sizes, these studies add to a body of evidence linking PM exposure, autonomic dysregulation and inflammation and indicating that pre-existing diseases can increase vulnerability to autonomic imbalance.

There is noted heterogeneity within study findings. [130] (2003) report that for (a limited cohort) patients with stable, severe CHD, CO but not PM exposure may briefly modify autonomic control. Furthermore, studies of older people presenting with chronic obstructive pulmonary disease (COPD), found no associations between HRV and PM2.5 [133,134], potentially reflecting differences in systemic inflammation compared to other subgroups [135], including a baseline lower HRV in individuals with COPD [136,137]. Taken together, these results suggest that the acute cardiovascular toxicity of PM alters the autonomic control of the heart depending on the underlying health status of the individual.

The acute or direct health effects of elevated background PM exposure and following significant events, such as bush fires, has been relatively well documented [138]. For instance, exposure to smoke is associated with increased hospitalisations [139], respiratory morbidity such as asthma and COPD [140], stroke [141], cardiovascular disease [142], poorer birth outcomes [143] and premature deaths [144]. Analysis of acute smoke exposure and premature deaths in Sydney from 1994 to 2007 reveals a 5% increase in mortality during bushfires [145]. In terms of the acute effects of recent fires in eastern and southern Australia, [146] estimated ~420 excess deaths from this smoke exposure, together with 3151 additional cardio-respiratory related hospitalisations and 1305 emergency department attendances for asthma.

In addition to immediate or early effects, air pollution has been linked to later development of chronic diseases and emerging evidence supports epigenetic promotion of such disease risk across generations. The WHO estimates that >80% of those individuals in urban areas are exposed to air pollution levels exceeding guideline limits [147], with pollution disproportionately impacting those in less economically developed populations [148], mirroring the demographics for major NCDs. Inhaled pollutants impact organ systems beyond the lungs (Kurt et al., 2016), including the heart and blood vessels [149,150,151], liver [152], kidneys [153,154] and as detailed above, the CNS [155]. Air pollution also affects developmental programming [156].

Beyond predictable associations with respiratory disorders [157], air pollution is linked to NCDs including steatohepatitis [158], diabetes [106], neurodegenerative diseases [159] and cancers [115,160]. Cardiovascular disorders, such as hypertension [161] and CHD [162,163], are particularly strongly associated with air pollution. Even brief exposure is linked to AMI, stroke, arrhythmias, worsening of heart failure and hypertension [164,165,166,167], while chronic exposure accelerates atherosclerosis, impacts blood pressure control, thrombosis, endothelial function, insulin sensitivity [164,166] and increases the risk of hypertension, Long-term effects of ambient PM2. 5 on hypertension and blood pressure and attributable risk among older Chinese adults [167,168]. The WHO estimates that air pollution contributes to ~7 million premature deaths globally per annum, with cardiovascular impacts, for example, rivalling the most widely studied and powerful drivers of smoking, hypertension and inactivity [165]. More recent analysis ascribes 9 million premature deaths to pollution in 2015 (16% of global deaths), and welfare costs approaching USD 5 trillion annually [169]. There appear to be no safe levels of air pollution in terms of increased mortality [170].

Exposure to PM2.5, predominantly generated from fossil fuel combustion and bushfires, is linked to premature death due to cancer, respiratory, metabolic and cardiovascular diseases [138]. As with air pollution more broadly, those with existing conditions are particularly sensitive to PM exposure [166], exacerbating acute cardiovascular events and promoting chronic CVD. While the relatively large size of PM10 particles (predominantly pollen, dust and construction or agriculture by-products) limits penetration to the upper respiratory tract, PM2.5 and smaller reach the alveoli which may underlie greater impacts of PM2.5 on multiple organ systems.

One body of evidence linking airborne PM to NCDs comes from studies of air pollution exposure in socioeconomically disadvantaged people. The markedly increased NCD risks in minority and low-socioeconomic-standing (SES) groups may involve disproportionate exposures to pollutants. For example, diabetes is linked to increased pollutant exposure in high-risk groups in the USA [171,172,173]. Proximity to major roadways, strongly linked to low SES, is associated with increased carotid-intima thickness [174], abdominal adiposity [175], hypertension [77], the risk of AMI [176] sudden cardiac death [177], mortality due to acute heart failure [178] and stroke [179]. Concentration of chemical plants and waste sites in low SES communities also contributes to increased exposure and thus disease risk [180]. The health impacts of pollution are also influenced by SES, dietary and other factors. For example, a lack of association between PM exposure and CVD in a study of male health professionals has been attributed to a higher SES and healthier lifestyle of this sub-population [181]. In summary, PM is linked to NCDs in both short-term and longer-term timeframes. We have detailed how PM impacts PINE subsystems and have shown that PM is linked to not only respiratory disease, but NCDs that have previously been linked to PINE network pathophysiology [5].