Science & Technology (Commonwealth Union) – The role of advanced technology in data analysis and innovations continues to show promise in further progressing scientific development.
Scientists from the University of Cambridge, Tsinghua University and sleep technology company Sleep Cycle examined cough sounds recorded across 32 cities in 12 countries over a period of more than 500 days. Their findings revealed a strong association between higher daily air pollution and nighttime coughing, including in areas where pollution remained below the World Health Organization’s recommended daily exposure threshold.
The team analysed cough sounds passively captured through the microphones of mobile phones using the Sleep Cycle app. Coughs detected on individual devices were counted and then aggregated by city, creating an unusually large dataset for studying the relationship between air pollution and coughing. The researchers also controlled for other factors that could affect coughing frequency, including weather conditions and influenza prevalence.
The anonymised data showed that for every 10 micrograms per cubic metre increase in particulate matter pollution, coughing increased by 2.4%.
Los Angeles provided an additional point of evidence. Coughing rates in the city rose sharply around the time of the January 2025 wildfires, providing a quasi-natural experiment that supported the observed short-term link between daytime exposure to pollution and coughing during the night.
The study appeared in the journal Communications Health.
Air pollution is one of the major environmental threats to human health and is estimated to contribute to around seven million premature deaths each year. Of the pollutants present in the air, fine particulate matter measuring less than 2.5 micrometres (PM2.5) is especially harmful because it can travel deep into the lungs and enter the bloodstream. This exposure may trigger inflammation across the body and increase the risk of cardiovascular disease and respiratory illnesses, including asthma and chronic obstructive pulmonary disease (COPD).
The lead author Dr Tong Xia of Tsinghua University indicated that coughing from time to time is normal, but whether occasional or persistent, it can indicate irritation or inflammation in the lungs.
She pointed out that a cough may indicate an issue that has not yet developed into a chronic condition but could potentially do so in the future.
Sleep Cycle is a mobile application that monitors sleep quality by analysing sounds such as breathing, snoring and coughing. The audio is recorded on the user’s device and processed using an AI-based algorithm. Having assessed more than four billion sleep sessions, the company’s technology provides a potentially valuable source of data that can complement conventional public health monitoring and epidemiological studies.
Professor Cecilia Mascolo from the University of Cambridge, Department of Computer Science and Technology, who was the lead of the study indicated that the dataset from Sleep Cycle is highly beneficial as the app is widely utilized across the globe.
She pointed out that a majority of public health studies just have data from a single nation, so having a dataset of this magnitude provides it much greater power.
“Every night, Sleep Cycle’s technology turns millions of smartphones into precise sound analysis instruments,” explained co-author Dr Mikael Kågebäck from Sleep Cycle. “Coughs are detected and counted directly on each user’s device, so no audio ever leaves the phone. The combination of scale and privacy lets researchers see the health effects of air pollution city by city, night by night, at a scale no sleep lab could match.”
The dataset contained no personally identifiable information and included cough-sound data collected from tens of thousands of users over an 18-month period. Researchers focused on 32 cities around the world where the app had the highest levels of usage. These locations spanned 12 countries: Australia, Brazil, Chile, China, France, Germany, Italy, Japan, South Africa, Sweden, the United Kingdom and the United States.
The team compared the cough recordings with city-level PM2.5 pollution levels, weather conditions and influenza surveillance data collected over the same 18-month period. Even after taking into account the annual flu season, when coughing would typically increase, the analysis identified a strong association between daytime air pollution levels and the frequency of night coughing.
