Tracheal Sound Spectra Depend on Body Height

Tracheal sounds originate from turbulent flow in upper and central airways. Turbulent flow characteristics are influenced by conduit dimensions. Because tracheal dimensions are a function of body height, we hypothesized that there should be a correlation between sound spectra and body length. We recorded tracheal sounds at standardized airflows in 21 healthy children 9.1 ± 0.6 yr of age (mean ± SE) and in 24 healthy adults 30.2 ± 0.8 yr of age. A contact sensor was attached at the suprasternal notch of the sitting subject, and airflow was measured at the mouth with a calibrated pneumotachograph. Tracheal sounds were low-pass-filtered at 2.4 kHz and digitized at 10 kHz. A 2048 point FFT was applied at a successive 100-ms intervals, using a Hanning data window. Resulting spectra were normalized to a reference power of 0.1 (mV)2/5 Hz. We applied a gating algorithm to extract sounds at inspiratory flows of 1 L/s (± 10% tolerance), and we computed average power spectra from the collected samples. We calculated the average spectral power (Pavg), the quartile frequencies below which 25% (Q1, 50% (Q2), and 75% (Q3) of the power in the range of 50 to 2,000 Hz was contained, the spectral edge frequency (SE95) below which 95% of the power was found, and the frequency where spectral power rolled off sharply (Fcut). We found that children had significantly louder sounds than did adults (Pavg, 27 ± 1 versus 17 ± 1 dB; p < 0.001) and higher frequencies (Q1365 ± 10 versus 196 ± 11 Hz; Q2, 747 ± 18 versus 431 ± 23 Hz; Q3, 1,127 ± 22 versus 711 ± 33 Hz; SE95, 1,627 ± 27 versus 1,028 ± 51 Hz; p < 0.001). There was a negative correlation between Fcut and individual height (r = −0.88). We conclude that tracheal sounds at given flow have higher frequency components in children than in adults, and they depend on body length. The difference is likely a function of upper and central airway dimension.