Effect of hypoxia on modulation of pressor responses to isometric handgrip with low atmospheric pressure

Subjects & Protocol

18 healthy male volunteers, in the age of 20-40 years served as subjects. Female subjects and individuals with middle ear disease, Eustachian tube dysfunction, upper respiratory tract infection, allergic rhinitis and those on any medication(s) were excluded from the study. A written consent was obtained after the experimental protocol was explained. Ability to ventilate middle ears was ensured through the examination of the mobility of tympanic membranes with the Valsalva manoeuvre on ground and a standard ‘ear clearance’ run in the Decompression Chamber.

Experimentation was conducted between 1400h-2000h in a post absorptive state in an altitude chamber installed in the Department of High Altitude Physiology and Hyperbaric Medicine at Institute of Aerospace Medicine (IAM), Bangalore situated at an elevation of approximately 963 meters (3,159 ft) from the mean sea level. Subjects were asked not to consume alcoholic beverages the previous night and avoid tea, coffee and smoking from 1000h in the morning on the day of experimentation.

For experimentation in normoxic normobaria i.e. while breathing air at ground level (3,159 ft AMSL; 0.92 ATA), doors of the chamber were closed and measurements taken after about 15 min in order to maintain the thermal ambience comparable with that during hypobaric evaluations at ‘high altitude’. Experimentation in normoxic normobaria was followed by that in hypoxic or nonhypoxic hypobaria. One of these experiments was done on the same day and the other was accomplished on a separate occasion. Minimum separation between the two hypobaric experiments was two days.

For hypobaric conditions, the subject was decompressed to 15,000 ft in the altitude chamber with an ‘ascent’ rate of 3000 ft/min. After completion of the test procedure, the subject was ‘brought down’ to ground level at the rate of 3000 ft/min. For evaluation in hypoxic hypobaria, the subjects breathed chamber air and for evaluation in non-hypoxic hypobaria they were made to breathe air oxygen mixture using an oxygen mask (ABEU MK II) and regulator (MK-17E) assembly.

To prevent any ‘carry over’ effects, the sequence of evaluation was counterbalanced as :-

In each of the above conditions, handgrip test was performed as described below

A continuous recording of electrocardiogram (in Standard Lead-II configuration) was made for 2 min in resting sitting before the grip and for the same duration from second to fourth minute of the grip. It was accomplished for the analysis of heart rate variability (HRV).

Handgrip Test

Subject was given instruction for the proper use of dynamometer and not to hold breath during the procedure. Maximal voluntary contraction (MVC) was evaluated in the dominant hand for a few seconds. The best of the 3 readings was taken as MVC. Thereafter, subjects maintained a hand grip at 30% of their MVC for 4 min. Heart rate (HR) and blood pressure (BP) were recorded, on non exercising arm, at the end (last 30s) of first, second to fourth minutes during the contraction (i.e. before the release of handgrip) and sixth min (i.e. during recovery after release of the grip). Measurements of BP were made using a standard sphygmomanometer.

Analysis of Heart Rate Variability

Sequence of beat to beat R-R intervals was manually examined and edited for artifacts or ectopics following which this discrete event series (DES) was transformed into an evenly sampled time signal at 4 samples/sec by linear interpolation using Matlab® (Version 6.5.0.180913a Release 13) software. This transformation into a real time signal allowed the power spectra of R-R interval to be represented as a function of true frequency in units of cycles per second (Hz) as opposed to cycles per beat. Subsequent to removal of mean from the series, estimates of HRV were made using Matlab® through parametric (maximum entropy method) technique. Parameters of autoregressive model were calculated with ‘AR’ command using ‘forward-backward’ approach. Model order was determined by minimizing Akanke information criterion. ‘PMEM’ command was employed to calculate power spectrum estimates via Maximum Entropy Method. It permitted a frequency resolution of 0.002 Hz

Power in spectra was integrated as follows:-

Low Frequency (LF) - 0.040 - 0.150 Hz.

High Frequency (HF) - 0.150 - 0.500 Hz

To account for the variation in the R-R interval during experimentation, power was normalized to square of mean R-R interval as suggested by Taha et al [7]. Thereafter, for appropriate scaling, the normalized values were multiplied by 10⁶.

Statistical Analysis

The data were, first, examined for normality using Shapiro Wilk’s ‘W’ statistic. Logarithmic transformation was done for HRV indices, since they were found to deviate significantly from normality. One way, repeated measures analysis of variance (ANOVA) was used for determination of statistical significance of HRV indices. For heart rate and blood pressure values during handgrip test in normobaria, hypoxic hypobaria and non-hypoxic hypobaria, a two way, repeated measures ANOVA with Greenhouse-Geisser correction was employed. The two factors were- experimental conditions (with three levels viz, normobaria, hypoxic hypobaria and non-hypoxic hypobaria) and isometric muscle contraction (with five levels viz. resting sitting, first, second to fourth minute during handgrip and recovery in 6^th min). Greenhouse-Geisser correction was employed due to significant departures from the assumption of sphericity. Level of significance was set at p<0.05.

It is pertinent to note that Blood pressure could not be recorded during handgrip for one subject during non-hypoxic hypobaria and for two subjects during hypoxic hypobaria due to chamber noise/ movement. Therefore, number of subjects was effectively reduced to 15 in case of BP data.

Degrees of freedom have been adjusted accordingly. Analysis of HRV was performed in 10 subjects only.

Changes in Heart Rate

Changes in HR were qualitatively different in hypoxic hpobaria from those in non-hypoxic hypobaria. HR increased in hypoxic hypobaric (p=0.017) and decreased in non-hypoxic hypobaria. However, the decrease in non-hypoxic hypobaria was just short of significance (p=0.053). Such an increase in HR during hypoxia is shown to be due to sympathetic activation directly via stimulation of peripheral chemoreceptors and vagal withdrawal indirectly via increases in ventilation, for a review, refer to Wolfel & Levine [6]. The extent of change in heart rate (~7%) is also comparable to what is reported after an acute exposure to 15,000 ft altitude [8].

During exposure to non-hypoxic hypobaria, the HR tended to be lower. It was because the subjects were maintained in mild hyperoxia, rather than strict normoxia. Such an effect of hyperoxia on HR is also well established [9]. An objective evidence of the maintenance of subjects in mild hyperoxia (rather than strict normoxia) during exposure to non-hypoxic hypobaria and hypoxia during exposure to hypoxic hypobaria was available from SaO₂ which was significantly higher (p=1.484E-18) in nonhypoxic hypobaria (98.1±0.2%) and significantly lower (p=0.00E+00) in hypoxic hypobaric (86±0%) compared to that in normobaria (97.1±0.2%).

Changes in Blood Pressure

Unlike HR, BP exhibited changes which were qualitatively and quantitatively similar in non-hypoxic and hypoxic hypobaria. As stated above, DBP and PP exhibited significant main effect of pressure condition. SBP was largely unaffected with hypobaria (F=0.43, p=0.652) and changes in MAP could not attain statistical significance (F=1.71, p=0.201). It is to be appreciated that, while examining pressor responses to isometric muscle contraction (handgrip), it is the DBP which merits greatest importance.

A significant reduction in DBP during hypoxic hypobaria could be conceived, at the first instance, to be a result of ‘functional sympatholysis’ which offsets the effects of sympathetic activation. Nonetheless, a comparable (p=0.868) reduction of DBP was noted also in non-hypoxic hypobaria. In view of the above, it will be un-parsimonious to conceive that the effect was secondary to hypoxia. Low atmospheric pressure, the stressor which is common to both the conditions, seems to be more probable candidate. Such an effect is consistently demonstrable in previous studies [1,2]. Mediation of vasodilatation through accumulation of local metabolites (the proposed mechanism underlying ‘functional sympatholysis’) seems to be less probable a mechanism also from the observations that high energy phosphates, measured in skeletal muscle of humans, show no significant decrement in hypoxia at rest or even after exhaustive exercise. Green et al [10] investigated adaptations in skeletal muscle in response to progressive hypobaria in eight male subjects over 40 days of progressive decompression to the stimulated altitude of the summit of Mount Everest. Samples of the vastus lateralis muscle extracted before decompression (SL-1), at 380 and 282 Torr and on return to sea level (SL-2) indicated that maximal activities of enzymes representative of the citric acid cycle, beta-oxidation, glycogenolysis, glycolysis, glucose phosphorylation, and high-energy phosphate transfer were unchanged (p>0.05) at 380 and 282 Torr compared with initial SL-1 values.

Thermal ambience in the three experimental conditions

It is noteworthy that, in their study by Tripathi et al [1], there was a mild, yet statistically significant, increase in the ambient temperature during exposure to non-hypoxic hypobaria compared to normoxic normobaria (Tdb and Twb in the chamber varied between 27-30°C and 22-26°C in normobaria and 28-30°C and 22-27°C in hypobaria, respectively). Even though such a mild increase in temperature could not have affected the blood pressure during exposure to hypobaria, it was a desirable control. The present study maintained the thermal ambience (F=0.028, p=0.972 for Tdb; F=5.49E-29, p=1.000 for Twb and F=0.001, p=0.99 for Oxford Index; Table 2 refers for details) in all the experimental conditions and qualifies to be an ideal one from an academic point of view.

Changes in Spectral Indices of HRV

Absolute value of total power and that in the HF band showed significant main effect of hand grip (F=12.048; p=0.007 for total power and F=18.368; p=0.002 for HF power). The effect persisted even after normalization of the power to mean R-R² (F=4.982; p=0.053 for total power and F=11.628; p=0.008 for HF power). These observations are appreciable as a significant reduction in the vagal modulations in heart rate control during handgrip. No significant effect of either pressure condition or handgrip was observed on LF power. Table 3 refers for these details. The latter observation, even though seem to be intriguing at the firs instance, is supported from recent reports which have shown that LF may be a poor marker of sympathetic activity [11,12]. A few studies are worth discussion and are presented here. After selective sinoatrial parasympathectomy, Randall et al [13] observed that low-frequency power in dogs was a reflection of multiple components. The authors synthesised that the total low-frequency power was composed of 50% parasympathetic activity, 25% sympathetic activity, and 25% other ‘yet to be identified’ factors. Grasso et al [14] performed autoregressive transfer function analysis of RR interval, and systolic & diastolic arterial pressure in a group of 10 healthy human subjects during supine rest and while standing, both before and after â1-selective blockade (with atenolol). Beta-blockade failed to induce systematic changes on the power of the LF peak of R-R, in any condition. The coherence between R-R and systolic arterial pressure in the same region remained high and a constantly negative phase (approximately 50-60 degrees, corresponding to a delay of 1-2 heart beats of RR on SAP) was always seen. Beta-blockade decreased the power of the LF peak of systolic arterial pressure, increased the transfer function gain between systolic arterial pressure and R-R at LF, and the HF power of R-R. Authors concluded that LF oscillations of R-R were not directly generated by the sympathetic drive to the heart but reflect mainly the parasympathetic activity. The observations of the present study further indicate inadequacy of HRV spectral indices as markers of autonomic activity (at least in hypobaric conditions). This inference derives enough support from literature and is believed to be due to above effects and additional confounding effects of ventilatory volume, respiratory rate, cardiac volume & mechanical function, arterial stiffness, arterial baroreflex function and responsiveness of sinus node, for a review, refer to Wolfel & Levine [6].

Close

Figure 2:

Grand average of HRV power spectra (across subjects) during Normoxic Normobaria (NN), NonHypoxic Hypobaria (NH) and Hypoxic Hypobaria (HH)

Export to PPT

Present study further endorses the attenuating effect of low atmospheric pressure on the efficacy of AGSM. Even though the experiment design of the present study does not permit a direct visualization of the mechanisms involved in vasodilatation induced by hypobaria (measurements of muscle sympathetic nerve activity, forearm blood flow or autonomic blockade were not possible due to non availability of these facilities), it is clear, from the observations that hypobaria exerts an effect on peripheral vasculature and the effect remains unvarying whether the subjects are maintained in hypoxia or non-hypoxia. This is the singular strength of this study. It also provides a valuable insight into the mechanism(s) involved in vasodilatation after acute induction to high altitudes.