The major finding of this study was that the cardiovascular responses, except the MAP, were attenuated in the sinusoidal (PLC plus CLC) LBNP compared to the constant (CLC only) LBNP, although the average gauge pressure (CLC) during the LBNP period was set at −25 mmHg in all conditions. The mean Z0 values showed that the degree of the thoracic blood volume shift to LBNP was the largest in the constant LBNP condition, second largest in the 180-s period condition, and the smallest in the 30-s period condition (Figure 3). Z0 is inversely related to central blood volume . The order of the degree of the variation against each LBNP condition was the same in the decreases of the SV and CO (Figure 3) and in the rise of the TPR (Figure 3). As for the orthostatic regulation of blood pressure, the MAP results showed a large drop in the 30-s period condition (Figure 3), whereas the other cardiovascular indices showed relative small variation in that condition. These results confirmed the previous report  that the cardiovascular adjustability to sinusoidal LBNP was maintained at the period of slower than 50-s (that is, 0.02 Hz) oscillation. Levenhagen et al. (1994) revealed that cardiovascular adjustability to sinusoidal LBNP was maintained >50 s of the period of oscillation with an amplitude of 25 mmHg of LBNP. The amplitude of 25 mmHg of sinusoidal LBNP has a CLC of −25 mmHg. However, to our knowledge there is no previous study that compares sinusoidal LBNP with constant LBNP. The results of the present study revealed that the effect of the CLC of LBNP on cardiovascular adjustability was attenuated by the addition of PLC to LBNP.
Considering that the spectral leakages from the main periodic component of oscillatory LBNP directly contaminate the other spectral components , the waveform of LBNP should be a sinusoidal pattern in cases of simultaneous measurements of HRV during oscillatory LBNP. In the present study’s HRV results, the response of lnHF and ln(LF/HF) to LBNP were attenuated in the sinusoidal LBNP compared to the constant LBNP. Since there was no significant effect on lnLF, the ln(LF/HF) results reflect a dominant effect of lnHF, which is an index of vagal activity . Therefore, we surmise that the attenuated cardiovascular adjustability by the addition of periodic oscillation of LBNP was caused by the suppression of the vagal responsiveness to LBNP.
Regarding the whole-body heating, the baseline of Tre was raised by almost 0.1°C by this experimental protocol. Previous studies showed the increase of their subjects’ core temperature from 0.5°C to 1.5°C by a water perfusion suit with a relatively higher water temperature [10, 15]. Although the heating condition in the present study was very mild whole-body heating, the significant main effect of the Heating on baseline values was observed for Tre, Tsk-abdomen, Tsk-foot, HR, SV, CO, TPR, lnHF, and ln(LF/HF). We suspect that the significant main effect of the Heating on ln(LF/HF) is the result of relative sympathetic activation during whole-body heating. However, there was no significant effect on lnLF. Moreover, there is an argument that the LF of HRV is not a biomarker of sympathetic activity but rather is a measure of the modulation of cardiac autonomic outflows by the baroreflex . Generally, heat stress does not alter the baroreflex control of the heart , and therefore, the present study’s finding on lnLF could be adequate. The baseline values of lnHF and SV dropped significantly during the Heating condition. The decreases in HF of HRV and SV are known to indicate a reduction in venous return . However, since the CO baseline values were decreased by the heating although MAP was maintained in our study, TPR was estimated that was increased in the heating condition. In previous studies using whole-body heating, the TPR dropped markedly with heating . However, heat stress induces vasoconstriction in non-cutaneous beds (that is, splanchnic, renal, muscle, and cerebral) [14, 40]. The present finding of a relatively small increase in TPR might reflect non-cutaneous vasoconstrictions caused by the very mild whole-body heating. We should have measured cutaneous vascular resistance (CVR) in reference to the TPR.
As for the mean values’ responses to LBNP, the ANOVA showed that the main effect of Heating was significant only in HR (Figure 3). However, in the gain of the transfer function, the significant main effect of the Heating was observed in HR, SV, and Z0 (Figure 4). As for the influence of temperature on the distribution of blood, which was examined by right heart catheterization , the measurement of Z0, and gamma camera imaging , heat stress induces a central blood volume reduction. Although the heating conditions used in the present study were very mild compared to those of previous studies, we were able to detect the significant increase in the gain of Z0 in the Heating condition. Considering that the beat-by-beat signal of the cardiovascular response contains other periodic fluctuations (that is, respiratory sinus arrhythmia and Mayer wave-related sinus arrhythmia), a transfer function analysis of a target sinusoidal period of LBNP would be advantageous to experimentally detect cardiovascular adjustability, with high repeatability.
Regarding the phase results and the transfer function, a significant main effect of the Period was observed in all variables (HR, SV, Co, and Z0; see Figure 4). These results confirmed those of , who showed the relatively large lag angle to the sinusoidal LBNP in a short period of condition. The phase angle of HR and Z0 were relatively large compared to SV and CO, reflecting the adverse response direction to LBNP. Within the variables of the same response direction to LBNP, the lag angle in HR was relatively large compared to Z0, reflecting the cascade reaction to LBNP (that is, the reduction of central blood volume could be preceding the HR response to LBNP). Moreover, the significant main effect of the Heating on the HR of phase of the transfer function indicated that the lag angle in the HR was decreased by heat stress. Considering the augmented gain in HR caused by whole-body heating, the HR response to LBNP in heat stress was characterized by a large amplitude and quick reaction. In this study, the AP was not measured continuously because of the limitations of our experimental facility. The dynamic responses of AP, TPR, and also CVR to sinusoidal LBNP should be investigated to increase our understanding of orthostatic intolerance under heat stress.
Our finding of distinctly high coherence of Z0 reflected the similarity to the sinusoidal LBNP. Z0 is inversely related to central blood volume . The similarity was close to a coefficient of 1.0 (Figure 4). The main effect of Period on the Z0 of phase indicated that the coherence in Z0 was relatively lower in the 180-s period condition compared to the 30-s period. This might not have been caused as a physiological consequence, because the set value of the 180-s period of sinusoidal LBNP curve has a relatively long duration to maintain above −0.6 mmHg, which was the upper limit of the controllable range of gauge pressure of the LBNP system we used. Nevertheless, sinusoidal LBNP caused a sinusoidal thoracic blood shift.
As for the association between the subjects’ lifestyle habits and their cardiovascular responses to LBNP, the linear regression analysis showed a significant correlation between HR and MEQ score, except for the 30-s period LBNP condition. The subjects’ dietary habits and daily physical activity were not significantly correlated with their cardiovascular responses to LBNP. Previous studies showed that individual differences in MEQ score were related to the subjects’ baseline HR , sleeping behavior , personality , and mental health . The previously reported diurnal variation in vascular function indicates that the vasoconstrictor response is lower in the morning than afternoon [47, 48]. The reduced vascular function could cause the higher cardiac responsiveness to maintain the AP against the LBNP . Although high HR responsiveness is not directly correlated with LBNP tolerance , we suspect, but have not proved, that the reduced physiological arousal in evening-type subjects might contribute to the high HR responsiveness to the LBNP. A previous study reported that higher HR values with a low MEQ score (Evening type) was associated with low vagal activity .
The present study has several limitations. There were marginal but significant differences in the gauge pressure between the LBNP conditions. The differences of 0.40 mmHg (30-s vs. constant) and 0.27 mmHg (30-s vs. 180-s) were both significant. The magnitude of LBNP is directly related to the reductions of CVP [2, 13]. These data suggest 2-mmHg decreases of CVP for every −10 mmHg LBNP. The physiological significance of the difference of 0.40 mmHg of LBNP is not known, but it might be negligible. These marginal but significant differences may have been caused by the combination of accurate repeatability of the sinusoidal LBNP and the mismatch of PID parameters in the LBNP control system. This issue merits further study.