Daily smartphone voice recordings revealed subtle changes from the very first day of an asthma or COPD flare, highlighting a promising new approach to remote monitoring that could one day enable early intervention and improve patient care.
Study: Voice as a biomarker for the early detection of exacerbations in asthma and COPD: the TACTICAS study. Image credit: Pixel-Shot/Shutterstock.com
Asthma and chronic obstructive pulmonary disease (COPD) are chronic and painful, often disabling conditions, with frequent exacerbations. Early detection of exacerbations can allow early treatment, potentially improving outcomes. A recent study published in the journal ERJ Open Research found that voice changes can occur as early as the first day of an exacerbation, supporting the future use of voice analysis for remote monitoring and early treatment.
Voice analysis offers a new monitoring approach for asthma and COPD
Asthma and COPD exacerbations are episodes in which symptoms suddenly worsen, often leading to poor quality of life, increased medication use, hospitalization, and a greater risk of death. Because early treatment is associated with better outcomes, finding reliable ways to detect these flares as they begin has become an important goal in the management of both conditions.
Current approaches to remote monitoring rely on measures such as pulmonary function tests, blood samples, or vital signs, but these can be invasive, labor intensive, or require specialized equipment. Therefore, researchers have turned their attention to digital voice biomarkers, since changes in the airways during an exacerbation can change the way a person speaks. Although voice analysis has shown promise for detecting a range of other health conditions, its potential for monitoring asthma and COPD exacerbations remains largely unexplored.
Patients recorded speech during the three-month follow-up period
The current study aimed to understand how voice changes might reflect exacerbations of these two conditions, using specific voice characteristics as biomarkers of exacerbation onset and different types of exacerbation symptoms. They also investigated how well participants would follow the voice-based app for daily tracking.
The researchers conducted a prospective cohort study (the TACTICAS (Telemonitoring for Asthma and COPD Through VoICE AnalysiS) study), including 73 participants recruited in respiratory clinics. Of these, 35 had asthma and 38 had COPD, with a mean age of 62 years.
Participants recorded their own voices three times a day for three months using their own mobile phones. The app elicited speech recordings in the morning, after light exercise in the afternoon, and again in the evening along with symptom questionnaires. They used the TACTICAS app, co-developed with asthma and COPD patients and designed for home use.
The voice capture consisted of a fixed vowel sound and either a reading of a paragraph or an answer to a question. Participants also used the validated EXACT questionnaire to detect exacerbations, either by increasing their symptom score by more than nine points from baseline over three days, or by scoring more than 12 points for two consecutive days.
In addition, researchers checked in with participants every four weeks, and data about medication or symptom escalation was stored in the electronic case record form (eCRF).
Exacerbations were clinically verified using electronic health records (EHR) in conjunction with medication use and patient-reported symptoms, and only validated events were included in the analysis. Events not reported but identified by eCRF or EHR analysis were also included.
The researchers obtained a total of 23,799 recordings, of which 2,737 occurred during exacerbations. Of these, 227 were on the start, recovery or peak day. Within the study period, they recorded 38 exacerbations, mostly mild and moderate. Most were recorded by the questionnaire.
Voice changes occurred from the onset of exacerbations
The analysis showed that measurable changes in speech were already present on the first day of an outbreak. Multilevel analysis identified 13 auditory speech characteristics that changed significantly at baseline, including lower average loudness, shorter sustained vowels, and greater gloss, an indication of irregular vocal cord vibrations. The noise to harmonic ratio also increased, indicating an overall drop in voice quality.
Rather than occurring only at the beginning of an episode, these changes evolved throughout the outbreak. By collecting daily voice recordings, the researchers were able to track how speech changed from onset to recovery, providing a detailed picture of the progression of an exacerbation.
The study also found that different voice characteristics appeared to reflect different symptom patterns. Shimmer and noise-to-harmonic ratio were more strongly associated with chest symptoms, possibly because increased airway obstruction reduces airflow during speech, while minimal loudness showed a stronger association with dyspnea. The authors caution that these findings are exploratory and require confirmation in larger studies.
Interestingly, changes in voice were similar in participants with asthma and those with COPD. The researchers suggest that this may indicate that underlying physiological changes that occur during an exacerbation, rather than the specific diagnosis, drive the changes in speech. These changes may result from a combination of increased airway obstruction, sputum impaction, inflammation of the vocal cords, and changes affecting the larynx. However, they also stress that the study was not large enough to determine this for sure.
The authors suggest that these changes result from worsening airway obstruction during an exacerbation, which places mechanical stress on both the respiratory system and structures involved in speech production. As airway resistance increases, the respiratory muscles work harder, lung expansion is limited, and the pressure below the vocal cords falls.
This can make it more difficult to maintain a steady voice, shorten speech duration due to reduced airflow, and create a more breathy voice with increased sparkle. As the participants recovered, peak voice volume increased, which the researchers suggest may reflect the restoration of more normal speech mechanics.
The study also highlights a practical advantage of voice tracking. Participants were less likely to complete symptom questionnaires when feeling unwell, possibly preventing questionnaire-based detection of exacerbations. Conversely, recording a short voice sample may provide a simpler, less burdensome way to monitor respiratory health, making it a potentially inexpensive and discreet tool for future home monitoring.
Although the findings are promising, the authors stress that larger validation studies and predictive models will be needed before voice-based biomarkers can be incorporated into routine clinical care.
The observational study identifies opportunities for future research
To improve confidence that each flare represented a true flare, researchers verified events using multiple sources, including symptom questionnaires, electronic health records, medication use, and clinical review. However, they did not use the Rome criteria, a more recent framework for defining COPD exacerbations that incorporates physiological measurements such as heart rate, respiratory rate, oxygen saturation and inflammatory markers alongside symptoms.
Since the study did not collect these vital signs, the authors note that some exacerbations may have been classified differently according to the Rome definition. They also point out that heart rate itself can affect voice characteristics, making it an important factor to consider in future studies.
Also, the findings remain observational and show correlations rather than predictions. While the voice changed in tandem with exacerbations, the study cannot determine whether these changes can reliably predict an impending exacerbation. This distinction is particularly important because many other acute illnesses, including heart failure and common respiratory infections, can also affect the voice.
Several additional limitations should also be considered. The researchers did not systematically record participants’ daily inhaler use, even though inhaled corticosteroids can change voice quality. The study was also relatively small and included only native Dutch people, limiting the extent to which the findings can be applied. In addition, the large number of statistical comparisons raised the possibility that some significant associations were due to chance.
Future studies should validate these findings in larger, multilingual populations and further refine the prediction algorithms already in development. If voice-based prediction proves feasible, the authors suggest it could eventually support care pathways by prompting additional diagnostic tests when early signs of an exacerbation are detected.
Smartphone speech analysis could support future respiratory care
Overall, the study shows that measurable voice changes occur from the very onset of an asthma or COPD exacerbation.
These results document the potential of voice as a biomarker for detecting exacerbations in chronic respiratory disease and illustrate the utility of home monitoring by mobile phone.
However, the authors emphasize that larger validation studies and predictive models are needed before voice-based biomarkers can be used clinically for remote detection of exacerbations.
