Smartphone data may help detect rare lung disease earlier, study suggests
Findings suggest new way to track disease and monitor progression
Written by |
Data from smartphones and wearable devices may help with the early detection of idiopathic pulmonary arterial hypertension (IPAH), a rare lung disease with no known cause, a study suggests.
Using up to eight years of data obtained via the My Heart Counts iOS app, device-derived data helped distinguish people with IPAH from healthy individuals and from those with conditions that can cause similar symptoms, and tracked disease progression before and after diagnosis.
“Our study highlights the potential of leveraging real-world activity and questionnaire data from smartphones and wearable devices to improve early detection and longitudinal monitoring of PAH,” the study authors wrote.
The proof-of-concept study, “Assessing the feasibility of using smartphone data to identify risk of idiopathic pulmonary arterial hypertension,” was published in npj Cardiovascular Health.
IPAH is rare and often diagnosed late due to vague symptoms
PAH is a rare, progressive disease marked by abnormally high blood pressure in the arteries supplying the lungs. People with the condition may experience symptoms such as severe shortness of breath, fatigue, weakness, chest pain, dizziness, and fainting.
Because many of these symptoms are non-specific and resemble those of other diseases, PAH diagnosis can be delayed. This delay means patients may be diagnosed at a more advanced stage of disease, often with additional health conditions and a higher risk of death.
“Reducing the time to diagnosis is therefore essential to improve patient outcomes, minimise the number of investigations, and allow treatment initiation at an earlier disease stage, when therapies may be more effective,” wrote a team led by researchers at Imperial College London, in the U.K.
The team explored whether data collected by smartphones and smartwatches could help identify people at risk of IPAH earlier.
Study used smartphone app and wearable data to track activity
The researchers used a free iPhone app called My Heart Counts, which collects health data through Apple’s HealthKit platform and includes questionnaires about lifestyle, sleep, exercise habits, and overall health and well-being.
In the U.K., 109 participants were enrolled. Of these, 33 had confirmed IPAH and 61 were healthy controls. A group of 14 comparison participants was also included, 12 of whom had been hospitalized with severe COVID-19 and two who had been evaluated for PAH but had normal pulmonary artery pressures.
Participants were asked to download the app and were provided with an Apple Watch Series 4. Data were collected between September 2014 and August 2024, with some participants contributing data for up to 8.4 years. Pre-diagnosis data were available for 21 IPAH patients, covering a median of 10 months before diagnosis.
In HealthKit activity data, IPAH patients had significantly fewer steps, a lower average walking pace, and a lower maximum walking pace than healthy controls. They also climbed fewer flights of stairs, at a lower average and maximum pace. These differences were present even before diagnosis. After diagnosis, step count and stair-climbing pace increased significantly, while both resting and walking heart rates decreased.
Wearable data revealed differences in heart rate and activity levels
From watch-derived HealthKit activity starting at study recruitment, IPAH patients had significantly higher average and walking heart rates and lower heart rate variability than healthy controls. They also had significantly lower VO2max, the body’s maximum capacity to use oxygen during exercise, and burned fewer active calories than both control groups.
For sleep, data were limited, with detailed assessments available for 11 participants. Still, IPAH patients spent more time awake at night than the control groups (an average of 56 vs. 15 minutes).
To validate the approach, the researchers compared the app-derived metrics to two standard clinical tests for physical function: the six-minute walk test (6MWT) and the incremental shuttle walk test (ISWT). Results showed correlations between the 6MWT and measures such as flights of stairs climbed, heart rate, and energy burned. The ISWT was associated with step count, heart rate reserve, heart rate variability, and both resting and walking heart rate.
Pre-diagnosis phone data alone classified participants as IPAH with an AUC of 0.81 (an AUC of 1 indicates perfect classification). Watch data performed better, reaching an AUC of 0.87. When post-diagnosis data were used instead, the model’s ability to distinguish between groups dropped substantially, to 0.48 for phone data and 0.64 for watch data.
Combining pre-diagnosis phone activity metrics with questionnaire responses on lifestyle factors yielded an AUC of 0.91. For watch data, combining activity metrics with life-satisfaction survey responses raised the AUC to 0.94. Across all models, the most important classification metrics were those related to walking pace and stair climbing.
Model performance improved after retraining with U.S. data
For comparison, a separate U.S. group of 73 participants was used to assess the model’s performance outside the U.K. When the U.K.-based model was applied directly to this group, it achieved an AUC of less than 0.5.
The researchers identified differences in physical activity levels between the two populations. For example, the maximum walking pace in the U.K. group was a median of 112 steps per minute, compared with 86 steps per minute in the U.S. group.
To address this, they retrained the model by adding 20% of U.S. data to the training set alongside the U.K. data. This approach achieved an AUC of 0.74 on the remaining U.S. data using phone data combined with lifestyle survey responses.
“This study was designed as proof-of-concept to evaluate whether data collected via a smartphone app could support the identification of individuals at increased risk of IPAH,” the researchers wrote. “These non-invasive measures hold promise as a scalable, patient-centred solution for empowering patients and doctors to track disease progression more closely.”
