Measuring the double burden of air pollution

In the latest Global Burden of Disease study, ambient and household air pollution* were together ranked the 4th leading risk factors of disease burden (after dietary risks, tobacco smoke and high blood pressure), accounting for 6.5 million premature deaths in 2015. Up to 90% of these deaths occurred in low-and middle-income countries (LMICs), many of which are undergoing industrialisation and suffering from ever-worsening ambient air pollution (AAP), as well as unresolved household air pollution (HAP) from domestic solid fuel use – this is the  ‘double burden of air pollution’.

Available evidence

While AAP has been rising on the global public health agenda in recent years, early epidemiological investigations begun after the infamous 1952 London smog event. Since then, there has been an ongoing effort to evaluate the health impacts of APP, including several landmark cohort studies conducted in the United States and Europe, which have contributed significantly to current estimations of the global disease burden associated with AAP. Despite years of research, direct prospective evidence from LMICs remains extremely rare.

Unlike AAP data, scientific evidence on HAP is predominantly from LMICs where 2.7 billion individuals are exposed to HAP from domestic solid fuel use. However, most of these studies are small, cross-sectional, and often not carefully conducted. Similar to AAP, there is very limited prospective evidence on the health effects of HAP.

With a lack of reliable LMIC-specific evidence, the importance of tackling air pollution can easily be overshadowed by other social or political priorities –leaving a major public health challenge unattended.

Data collection barriers

The absence of large prospective cohort with long-term follow-up of individuals’ risk of disease development, and, more-importantly, the difficulties inherent in accurately quantifying personal exposure – a result of complex interactions between the source of air pollution, behaviour, climate and more – have resulted in huge knowledge gaps.

For AAP, conventional exposure assessment methods rely heavily on data from air quality monitoring stations, which are expensive (~USD 200,000 each) and therefore unaffordable in many LIMCs. Even for rapidly developing countries like China, such stations are scattered (only 880 national stations in 2015), and millions of individuals may share only one monitoring station. Furthermore, conventional methods are based on predictive modelling which assign exposure levels at a community level (for example, individual’s ZIP codes were used in a recent study published in the New England Journal of Medicine), which may not precisely reflect personal exposure. For HAP, direct measurement of exposure is impossible without deploying air quality monitors at a household- or individual-level. This imposes enormous financial and logistical challenges in accurately estimating personal exposure to HAP at a large enough scale, let alone the availability of practical and reliable monitors.

In the past 10 years, several large prospective cohorts of up to 0.5 million participants have been established in China, India and Mexico, with more under development. These studies offer new opportunities for reliable assessment of the long-term health effects of air pollution in LMICs. So, the current pressing question is, “how can we obtain accurate personal exposure data?”

High precision air quality monitors are available from specialised companies. However, these products are still relatively expensive, large and heavy, have short battery lives and are infeasible to deploy on a large-enough scale. Meanwhile, to overcome the impracticalities of these devices, there is an emerging effort to develop low-cost, light-weight, portable static and personal monitors through two primary development models.

One typical model of development is that academic researchers initiate the design and testing of a device, and eventually commercialise the product. Particle and Temperature Sensor (PATS+), a portable device that records real-time fine particulate matter (PM2.5) concentration, initially developed at the Berkeley School of Public Health, is one example of this. The academic model of development takes advantage of robust testing and evaluation processes and, later, the self-sustainability of a commercial company for further development. However, an academic-led approach often incurs prolonged iterative cycles of grant application, design, testing and evaluation before it can be applied in epidemiological studies or commercialised.

The other model is commercial-based. Increasing public concern about air quality has created a thriving market of personalised air quality monitoring products, thus attracting tech-giants such as Xiaomi and Huaiwei, as well as small start-ups. Driven by growing consumer demand and potential profit, this model has dominated recent air pollution monitor developments in terms of quantity. Nonetheless, most of these consumer products have not been carefully tested in accordance with high academic standards. This casts doubt on data quality and applicability in epidemiological studies.

By simultaneously taking advantage of these two models, there is great potential for an integrative partnership approach. With academics’ research expertise and business enterprises’ financial and technological capacity, the development, evaluation and application of modern air quality monitors could be significantly streamlined while having quality assured. This could be revolutionary. However, akin to modern drug development, ultimately the device would be a profit-making product, which may render it unaffordable to populations in greatest need. To mitigate this inequality, an agreement between academics, NGOs and commercial partners has to be reached to ensure non-commercial use in resource-deprived settings at an affordable cost. Eventually, the profit from the commercial market could create a sustainable feedback loop of funding and development of more advanced personal monitoring technology, which will benefit all parties.

Constructive collaboration between different stakeholders is urgently needed to bring about a new era of personal exposure monitoring, advance our knowledge in the health effects of air pollution and inform policy actions to tackle the double burden of air pollution.

(* AAP is mainly from fossil fuel combustion for power generation, heavy industry or mobile vehicles, and HAP is primarily from domestic solid fuel use for cooking and heating)

Peter Ka Hung Chan is a PhD student at the Nuffield Department of Population Health, University of Oxford. Peter has studied at the Chinese University of Hong Kong, the University of Copenhagen and the Johns Hopkins Bloomberg School of Public Health, and completed a MSc in Global Health Science at Oxford. His main research interest is in environmental health, particularly air pollution and climate change impacts on global health. For his PhD research, Peter is investigating the relationship between household air pollution and smoking and cardiovascular diseases in the China Kadoorie Biobank, a 0.5 million participants prospective cohort. 

Blog post originally published on Plos Blog.