Air pollution, stroke, and anxietyBMJ 2015; 350 doi: https://doi.org/10.1136/bmj.h1510 (Published 24 March 2015) Cite this as: BMJ 2015;350:h1510
- Michael Brauer, professor
The effects of air pollution on the lungs and heart are now widely appreciated, with expanding evidence for an important role in cardiac disease.1 The Global Burden of Disease Study identified fine particulate matter (PM2.5) in outdoor air and household air pollution from use of solid fuels as the ninth and fourth leading risk factors, respectively, for disease worldwide,2 and the World Health Organization attributes one in every eight deaths to air pollution.3 The effects of air pollution are not limited to cardiopulmonary diseases. Recent evidence suggests a role in diverse outcomes, including diabetes,4 low birth weight, and preterm birth.5 This research stems from improved understanding of the role of air pollution in initiating systemic inflammation, a response that may affect multiple organ systems. Two linked studies (doi:10.1136/bmj.h1295, doi:10.1136/bmj.h1111) add to growing evidence that air pollution is an important risk factor for an increasing number of common diseases.6 7
In the first of the two papers, Shah and colleagues6 systematically reviewed and meta-analysed 103 studies conducted in 28 countries and including 6.2 million events to assess the role of short term fluctuations in air pollution as a trigger for stroke. Although evidence from several cohort studies of long term exposure to particulate matter indicates associations with stroke mortality, such findings are not universal.8
The role of air pollution as a possible trigger for stroke has important implications for disease burden, especially in China where air pollution and the incidence of (especially haemorrhagic) stroke are high. In their analysis, Shah and colleagues found that increases in each of the common gaseous and particulate air pollutants were significantly associated with admission to hospital for stroke or stroke related mortality, with associations strongest for strokes on the same day as exposure; increased ozone was only weakly associated with cerebrovascular events.
Air pollution remained significantly associated with stroke in sensitivity analyses that adjusted for potential biases related to quality of outcome ascertainment, assessment of exposure, and adjustment for confounders. This analysis supports a role for air pollution as a modifiable risk factor for stroke, although associations with air pollution were less precise for haemorrhagic stroke than for ischaemic stroke. The impact of chronic exposure to air pollution on development of carotid atherosclerosis (a precursor for stroke) remains unclear. Although this is not covered in the analysis, evidence of an association is growing.9
Since air pollution causes systemic inflammation, it is reasonable that researchers have now turned to the arena of mental health, a leading priority for research given the relative absence of known modifiable risk factors and a high and growing disease burden.10 In the second linked paper, Power and colleagues exploit rich data in the Nurse’s Health Study cohort to assess the role of particulate pollution on prevalent anxiety symptoms.7 They found an exposure dependent association between higher levels of PM2.5 and increased symptoms of anxiety, and indications that associations were stronger for exposures in the month immediately preceding the scoring of anxiety.
These observations were supported by several sensitivity analyses, which indicated that associations were robust to broad geographical region, health status (to control for the possibility of anxiety as a sequela of cardiopulmonary effects of air pollution), and demographic characteristics, although the study was limited to older women. Power and colleagues’ findings add to a growing literature on the mental health effects of air pollution, including a small but intriguing body of research linking short term variability in air pollution to suicide.11
Power and colleagues used spatiotemporal exposure estimates and reported stronger effects for more recent exposures, reducing confounding by spatially varying factors correlated with air pollution. Since effects were observed over all time periods, spatial variation seems to have had an important influence on effect estimates. Furthermore, although effects were observed in all geographical regions, the investigators did not examine other potentially adverse (for example, noise, barometric pressure, solar intensity) or healthy (for example, natural spaces) environmental exposures that may operate at different scales. Indeed, evidence is accumulating that natural spaces may have beneficial effects on stress and social cohesion, both of which deserve further study in relation to mental health.12
As with any observational study, questions remain, as the authors acknowledge, and the findings should be replicated in other populations and with other study designs. Moreover, although these observations are biologically plausible, given links between inflammation and anxiety there is a need for greater mechanistic supporting evidence, of the type that now exists for associations between particulate matter and pulmonary, cardiac, and circulatory disease.
The findings of these two studies support a sharper focus on air pollution as a leading global health concern. They also suggest opportunities for reducing the prevalence of two debilitating and common diseases. One of the unique features of air pollution as a risk factor for disease is that exposure to air pollution is almost universal. While this is a primary reason for the large disease burden attributable to outdoor air pollution, it also follows that even modest reductions in pollution could have widespread benefits throughout populations. The two linked papers in this issue confirm the urgent need to manage air pollution globally as a cause of ill health and offer the promise that reducing pollution could be a cost effective way to reduce the large burden of disease from both stroke and poor mental health.
Cite this as: BMJ 2015;350:h1510
Competing interests: I have read and understood the BMJ policy on declaration of interests and declare the following: none.
Provenance and peer review: Commissioned; not externally peer reviewed.