Computerized respiratory sounds can differentiate smokers and non-smokers

Abstract

Cigarette smoking is often associated with the development of several respiratory diseases however, if diagnosed early, the changes in the lung tissue caused by smoking may be reversible. Computerised respiratory sounds have shown to be sensitive to detect changes within the lung tissue before any other measure, however it is unknown if it is able to detect changes in the lungs of healthy smokers. This study investigated the differences between computerised respiratory sounds of healthy smokers and non-smokers. Healthy smokers and non-smokers were recruited from a university campus. Respiratory sounds were recorded simultaneously at 6 chest locations (right and left anterior, lateral and posterior) using air-coupled electret microphones. Airflow (1.0–1.5 l/s) was recorded with a pneumotachograph. Breathing phases were detected using airflow signals and respiratory sounds with validated algorithms. Forty-four participants were enrolled: 18 smokers (mean age 26.2, SD = 7 years; mean FEV1 % predicted 104.7, SD = 9) and 26 non-smokers (mean age 25.9, SD = 3.7 years; mean FEV1 % predicted 96.8, SD = 20.2). Smokers presented significantly higher frequency at maximum sound intensity during inspiration [(M = 117, SD = 16.2 Hz vs. M = 106.4, SD = 21.6 Hz; t(43) = −2.62, p = 0.0081, d z = 0.55)], lower expiratory sound intensities (maximum intensity: [(M = 48.2, SD = 3.8 dB vs. M = 50.9, SD = 3.2 dB; t(43) = 2.68, p = 0.001, d z = −0.78)]; mean intensity: [(M = 31.2, SD = 3.6 dB vs. M = 33.7,SD = 3 dB; t(43) = 2.42, p = 0.001, d z = 0.75)] and higher number of inspiratory crackles (median [interquartile range] 2.2 [1.7–3.7] vs. 1.5 [1.2–2.2], p = 0.081, U = 110, r = −0.41) than non-smokers. Significant differences between computerised respiratory sounds of smokers and non-smokers have been found. Changes in respiratory sounds are often the earliest sign of disease. Thus, computerised respiratory sounds might be a promising measure to early detect smoking related respiratory diseases.