Stroke is a leading global cause of death and morbidity, ranking third among causes of years of life lost (GBD 2017 Causes of Death Collaborators, 2018) and first among causes of disability-adjusted life years (GBD 2016 Stroke Collaborators, 2019). Air pollution affects hypertension, oxidative stress, systemic inflammation, imbalance of the nervous system and atherosclerosis, which are all pathophysiological mechanisms associated with stroke (Lederer et al., 2021). Air pollutants with the strongest public health concern include particulate matter with diameter <2.5 μm (PM2.5) and nitrogen dioxide (NO2), and several other pollutants including elemental carbon (EC) and ultrafine particles (UFP, <0.1 μm diameter) indicate risk, but currently with insufficient data(WHO, 2021). A recent meta-analysis found that a 5nullμg/m3 increase in long-term exposure to air pollution with PM2.5 is associated with a 6.3% increased risk of stroke (Alexeeff et al., 2021). Studies on long-term exposure to air pollution with NO2 are less conclusive but generally indicate a positive association (Olaniyan et al., 2021). Only few studies have investigated the potentially more potent UFP, indicating an association with cardio- and cerebrovascular disease (Downward et al., 2018; Li et al., 2017; Poulsen et al., 2023a, Poulsen et al., 2023b). Studies on elemental carbon (EC) (or black carbon or PM2.5 absorbance) generally indicate no or a weak association with risk of stroke (Beelen et al., 2014; Downward et al., 2018; Ljungman et al., 2019; Poulsen et al., 2023a, Poulsen et al., 2023b; Stafoggia et al., 2014; Stockfelt et al., 2017; Wolf et al., 2021).

The association between air pollution and risk of stroke may be modified by other factors. In a Danish study, NO2 was associated with ischemic stroke only at residences with high noise levels (Sorensen et al., 2014). Some studies have linked access to green areas near the residence or living in rural areas with lower incidence and mortality from stroke or cardiovascular disease (Crouse et al., 2019; Hystad et al., 2020; Kim et al., 2019; Klompmaker et al., 2021; Seo et al., 2019; Stafoggia et al., 2014) possibly due to reduced psychological stress, increased physical activity, or by lower levels of air pollution (Nieuwenhuijsen et al., 2017).

Poor socioeconomic conditions are associated with risk of stroke, likely due to an unhealthy lifestyle, including smoking, alcohol consumption, stress, depression, poor diet, obesity, sedentary lifestyle and hypertension, which are all established risk factors for stroke (Boehme et al., 2017). Some studies have indicated stronger associations between air pollution and stroke among people with lower socioeconomic status (Rodins et al., 2020; Stafoggia et al., 2014; Yang et al., 2021), whereas this was not found in other studies (Hystad et al., 2020; Klompmaker et al., 2021).

Elderly people, children, women,and persons with pre-existing comorbidities including stress have been suggested to be more susceptible to the effects of air pollution (Clougherty, 2010; Schwartz et al., 2011). A recent meta-analysis of 11 cohort studies, found evidence of excess risk of ischemic heart disease in women compared to men in relation to long-term PM2.5 exposure, but no difference in risk between sexes in relation to stroke (Zhang et al., 2022). Hypertension and diabetes are established risk factors for stroke (Boehme et al., 2017), but studies investigating effect modification by comorbidity have produced conflicting results (Amini et al., 2020; Hart et al., 2015; Hystad et al., 2020; Olaniyan et al., 2021; Shin et al., 2019; Stafoggia et al., 2014). A study found that self-reported stress could modify the short-term association between PM2.5 and blood pressure (Hicken et al., 2014), indicating that stress conditions might also modify the association between air pollution and stroke.

At present, there is insufficient evidence to conclusively identify susceptible sub-populations for which the association between long-term exposure to air pollution and stroke is strongest, and where targeted interventions could be merited. When comparing associations between air pollution and risk of stroke in population groups, relative risk (i.e. how many times greater the hazard in one group compared to another, for example hazard ratios (HR)) and absolute risk (additional cases per 100,000 person-years, rate differences) might provide different results. The difference between these two measures arise when the basic disease rates between the groups of interest differ, for example when the incidence of disease is higher in persons of low versus high educational level, which is often the case. In such situations, the hazard ratio for a given exposure between these two groups may be similar, but the absolute rate difference will be higher in the group of low educational level compared to high – provided that the exposure is positively associated with the outcome. From a public health perspective, the knowledge on whether the absolute health effects of exposure to air pollution differ across subpopulations is important. We are only aware of one previous study of air pollution and risk of stroke, which has applied an additive model for estimation of absolute risk (Danesh Yazdi et al., 2021).

The aim of this nationwide study was to investigate if the association between long-term exposure to air pollution and risk of stroke differed by sociodemographic factors, financial stress, comorbid conditions, and road traffic noise, population density and green space at the residence.