Dev48
Language
  • About
  • Services
  • Industries
  • Technologies
  • Articles
  • Contacts
Book a call
    Home/Articles/Sleep app tracks link between air pollution and a bad nights sleep
Dev48
info@dev48.org+375 29 320-40-96WhatsApp@isky_by

© 2026 · All rights reserved.

Sleep app tracks link between air pollution and a bad night’s sleep

Source: Cambridge LTL•September 26, 2026 at 10:27 AM

Sleep app tracks link between air pollution and a bad night’s sleep

Analysing data collected via a sleep app, researchers have found that daytime pollution is strongly linked to higher levels of coughing at night, even where pollution is below the World Health Organization's recommended daily exposure limit.

Daytime pollution is strongly linked to higher levels of coughing at night, even where local pollution levels are below the World Health Organization's (WHO) recommended daily exposure limit.

That's the concerning finding made by a team of researchers from this Department, Tsinghua University, and the sleep technology company Sleep Cycle.

The researchers used coughing sounds that had been passively recorded by the Sleep Cycle mobile app through the microphone in app users' phones. The sounds were recorded in 32 cities in 12 countries over 500 days. Cough sounds were counted on individual devices and pooled by city to build an unprecedented dataset of this kind. The researchers accounted for external factors that could lead to an increase in coughing levels, such as weather conditions or influenza levels.

Analysis of the data, which was anonymised, also found that when pollution levels rose by 10 micrograms of particulate matter per metre cubed, there was a 2.4% rise in coughing levels.

One of the cities included in the dataset was Los Angeles, and coughing levels there spiked at the same time as the January 2025 wildfires. This acted as a quasi-natural experiment that further supported the short-term association between daytime pollution and nighttime coughing. The results are reported in the journal Communications Health.

Four billion sleep sessions

Sleep Cycle is a mobile app that uses sounds – such as breathing, snoring or coughing – to track sleep quality. These sounds are recorded locally and analysed by an AI algorithm. The company's technology has analysed more than four billion sleep sessions, making it a useful complement to traditional public health surveillance and epidemiological research.

"The dataset from Sleep Cycle is really valuable, because the app is widely used worldwide," says Cecilia Mascolo, Professor of Mobile Systems here, who led the research. "Most public health studies only have data from one country, so having a dataset of this size gives it much more power."

The dataset – which included no personal identifiers – contained information on cough sounds from tens of thousands of users over 18 months. The researchers linked the cough data with city-level PM2.5 concentrations, meteorological variables and influenza surveillance data over a period of 18 months. Analysis of the data, even when accounting for the annual flu season when cough rates would normally be higher, showed a strong link between daytime pollution levels and nighttime cough frequency.

'No safe threshold for exposure to air pollution'

"This suggests that there may be no such thing as a safe threshold for exposure to air pollution, although our research is purely observational," said Mascolo. "However, people may wish to consider how they might mitigate their own exposure to air pollution, by using masks or air filters, for example."

Read the full story on the University of Cambridge website.

  • Tong Xia et al. '.' Communications Health (2026). DOI: 10.1038/s44528-026-00030-5

Tong Xia et al. '.' Communications Health (2026). DOI: 10.1038/s44528-026-00030-5

← All articles

What this article says

More in AI & Machine Learning

All →
Anthropic launches roadshow for proposed initial public offering
Claude

Anthropic launches roadshow for proposed initial public offering

Introducing AlphaProof Code: a breakthrough in verifying complex software at depth
DeepMind

Introducing AlphaProof Code: a breakthrough in verifying complex software at depth

Introducing the Waymo Accessibility Network UK: Co-Designing the Future of Autonomous Mobility
Waymo

Introducing the Waymo Accessibility Network UK: Co-Designing the Future of Autonomous Mobility

Closing the AI guidance gap in education
OpenAI

Closing the AI guidance gap in education

Form field detection: fewer missed fields, cleaner labels
Datalab

Form field detection: fewer missed fields, cleaner labels

Transitioning Height Inspection From Sample Inspection to 100% Inspection: Designing From Reference Plane, Tact Time, and Threshold
Keyence

Transitioning Height Inspection From Sample Inspection to 100% Inspection: Designing From Reference Plane, Tact Time, and Threshold

More from Cambridge LTL

Researcher joins new £10m national initiative to tackle respiratory infections
Cambridge LTL

Researcher joins new £10m national initiative to tackle respiratory infections

Heatwaves are threatening Europe's power supplies, researchers warn
Cambridge LTL

Heatwaves are threatening Europe's power supplies, researchers warn

Researchers receive funding to explore ambitious new ways of improving AI security
Cambridge LTL

Researchers receive funding to explore ambitious new ways of improving AI security

Undergraduates’ work is presented at major conference
Cambridge LTL

Undergraduates’ work is presented at major conference

National Cyber Award for our Professor of Emergent Harms
Cambridge LTL

National Cyber Award for our Professor of Emergent Harms