High-amplitude oscillations in human skin blood flow are distinct from known cardiac or respiratory influences

C.D. Haley, C.J. Gordon, N.A. S. Taylor and A.B. Jenkins, Department of Biomedical Science, University of Wollongong, Wollongong, NSW 2522, Australia.

We have recently described the presence of high-amplitude skin blood flow ( Q-dot _sk) oscillations in humans (Haley et al., 2004a) and animals (Haley et al., 2004b). These oscillations had a characteristic frequency of ∼0.4 Hz, spanning ∼700-800 ms, and were comprised of high-amplitude peaks that are up to 7-fold greater than basal Q-dot _sk. We hypothesised that the high-amplitude oscillations may be related to previously identified lower-amplitude (±30 %) oscillations in Q-dot _sk caused by respiration (0.15-0.4 Hz) and cardiac frequency (0.4-1.6 Hz) (Stefanovska et al., 1999). Hence, we sought to compare the spectra of respiratory and cardiac activities, with the Q-dot _sk time-frequency spectrum.

Forearm Q-dot _sk (ventral aspect), cardiac frequency and respiration were measured simultaneously (20 Hz) from eight males (27.9 yr (SD 6.4), 181.1 cm (SD 4.8), 75.8 kg (SD 7.5), during semi-recumbent rest at 25°C (50% R.H.), on two separate days, for a 5-6 min period. Skin blood flow, was estimated using laser-Doppler flowmetry (TSI Laserflo BPM2, Vasamedics Inc., U.S.A.), cardiac frequency was collected using a three-lead electrocardiogram (ECG: model 100, Humtec, Australia) and respiration was measured from rib cage movement (mercury-in-silastic strain gauge: Hokansen EC-4SB, U.S.A.). Respiratory, ECG and Q-dot _sk data were analysed using a wavelet transform in the frequency domain of 0.05-2 Hz; the dominant frequency band for each variable was calculated as the central portion accounting for 95% of the total integrated power of the time-averaged frequency spectrum. Data are means ± standard errors.

The dominant frequency band for the Q-dot _sk spectrum (0.72 ±0.03) was significantly different (P<0.05, paired t-test) from both the respiratory (0.22 ±0.02) and ECG spectra (1.31 ±0.01). Variations between studies (n=16) in the Q-dot _sk dominant frequencies were not significantly correlated with between study variations in either cardiac (r=0.02, P =0.94) or respiration (r=-0.15, P=0.60) dominant frequencies. These results indicate that the high-amplitude Q-dot _sk oscillations are not directly related to either cardiac frequency or respiratory function. Instead, we propose that these oscillations are related to local factors, such as changes in transmural pressure, or the release of substances that alter vascular endothelial and smooth muscle function.

Haley, C.D., Zeyl, A., Taylor, N.A.S. & Jenkins, A.B. (2004a) Journal Thermal Biology, 29, 717-723.

Haley, C.D., Gordon, C.J., Taylor, N.A.S. & Jenkins, A.B. (2004b) Journal Thermal Biology. 29, 779-783.

Stefanovska, A., Bracic, M & Kvernmo, H.D. (1999) IEEE Transactions on Biomedical Engineering, 46(10), 1230-1239.