ORIGINAL PAPER

Patterns of Cardiovascular Control During Repeated Testsof Orthostatic Loading

Jerry Joseph Batzel • Nandu Goswami •

Helmut Karl Lackner • Andreas Roessler •

Mostafa Bachar • Franz Kappel • Helmut Hinghofer-Szalkay

! Springer Science+Business Media, LLC 2009

Abstract To investigate patterns of cardiovascular con-trol, a protocol of head up tilt (HUT) followed by lower

body negative pressure (LBNP), which represents a sig-

nificant cardiovascular control challenge, was employed.Linear regression of beat-to-beat heart rate (HR) and mean

blood pressure (MBP) data collected over repeated tests

was used to analyze control response during the LBNPphase of the combined HUT ? LBNP protocol. Four runs

for each of 10 healthy young males reaching presyncope

were analyzed. Subjects were classified into 2 groups basedon the consistency of MBP regulation in response to central

hypovolemia induced by LBNP. The consistent group

tended to exhibit consistent HR slope (rate of change ofHR over time as calculated by linear regression) whereas

subjects in the inconsistent group could not be easily

classified. Subjects with consistent MBP maintenanceexhibited patterns suggesting a consistency of response in

cardiovascular control whereas subjects less successful in

maintaining MBP exhibited less clearly defined patternsover four runs.

Keywords Graded orthostatic stress ! Presyncope !Heart rate ! Blood pressure ! Linear regression

Introduction

Head up tilt (HUT) followed by lower body negative

pressure (LBNP) represents a significant challenge to thecontrol mechanisms that act to maintain blood pressure

stability (e.g. Lelorier et al. 2003). As LBNP increases

during such a combined protocol (HUT ? LBNP), thestress can drive the control system to maximal imple-

mentation of cardiovascular compensatory mechanisms.

Indeed, when HUT ? LBNP is employed, extensivecentral hypovolemia leading to increased stresses on

blood pressure will nearly always lead to presyncope (e.g.

Lelorier et al. 2003). Thus, such a challenge provides aunique vantage point from which to observe control

response.

While much research has investigated hemodynamicresponses to various forms of orthostatic stress (Laszlo

et al. 1998; Goswami et al. 2008; Hinghofer-Szalkay et al.

2008), less is known about regularity of the controlresponse (Fink et al. 2004). The question arises whether the

response to a fixed stress will always trigger in a given

individual the same combination of control responseoptions. Furthermore, does the cardiovascular control sys-

tem have the ability to adjust the mix of responses to novel

stresses and can it vary its normal response combinationwhen this combination seems to be failing? That is to say,

must an individual who tends to respond to orthostatic

stress with increased heart rate (HR) continue to increaseHR in the same way if the stress increases or can a switch

be made to increased modulation of say Rs (systemic

arterial resistance)? We have recently shown that HR and

N. Goswami ! A. Roessler ! H. Hinghofer-SzalkayInstitute of Physiology, Center of Physiological Medicine,Medical University Graz, Graz, Austria

N. Goswami ! H. K. Lackner ! H. Hinghofer-SzalkayInstitute of Adaptive and Spaceflight Physiology,Wormgasse 9, Graz, Austria

J. J. Batzel (&) ! F. KappelInstitute for Mathematics and Scientific Computing,University of Graz, Heinrichstrasse 36, 8010 Graz, Austriae-mail: jerry.batzel@uni-graz.at

M. BacharDepartment of Mathematics, King Saud University,Riyadh, Saudi Arabia

123

Cardiovasc Eng

DOI 10.1007/s10558-009-9086-z

stroke volume (SV) responses are reproducible across

HUT ? LBNP using the method of surrogate data(Goswami et al. 2009c). Using experimental data from

that study, we investigated in this current work whether

blood pressure regulation and quantitative cardiovascularresponses in subjects during graded LBNP were repro-

ducible or consistent in pattern across four runs. We con-

sidered data from the LBNP phase because the HUT phasetypically generates initial, partly transient cardiovascular

responses, while graded LBNP, gradually applied post-HUT provides a repeatable, relatively smooth increase in

stress that should in a regular way increase the challenge to

the control system. In this context, the HUT phase (5 min)provides time to generate a stable reference stress level to

which the following consecutively added LBNP phases

could be added. We, therefore, examined the systemresponse only during the LBNP phase to assess patterns of

cardiovascular control. We additionally investigated whe-

ther orthostatic tolerance times to graded LBNP after a 70"upright tilt were correlated to the degree of regularity in

control responses.

Procedures and Methods

Subjects

Gender and age may affect orthostatic and stress responses(Goswami et al. 2008). To reduce overall variability, data

was collected in 10 healthy, non-obese, non-medicated,

non-smoking males who were free from any somatic ormental condition with age 25 ± 3 years, weight 75 ±

12 kg, height 179 ± 6 cm and with a supine HR of

69 ± 10 bpm. Subjects were familiarized with the testprotocol and gave written informed consent to participate

in the study. The study was approved by the Graz Uni-

versity Ethics Board, and was performed in accordancewith the 1989 WMA Declaration of Helsinki.

Protocol

The subjects were fasting before each study. Experimentswere carried out between 9 and 11 a.m. in a semi-dark,

quiet room maintained at 23–24"C, 50–55% humidity.

Each subject was tested four times with a minimum of a2-week interval between test runs.

Each experimental test started with a 30 min supine

rest period to acquire cardiovascular steady state condi-tions. At minute zero of the stress protocol, the tilt table

was brought to 70" head-up position. After five more

minutes, -20 mmHg LBNP was added. LBNP wasincreased by 10 mmHg every 3 min, particularly as we

wanted to drive the subjects to presyncope quickly (Fig. 1).

We chose this protocol, as LBNP stress duration influencesstress responses and we also wanted to minimize possible

effects of rapid adaptation. As soon as presyncopal signs or

symptoms occurred, the table was brought back to 0" andLBNP was stopped at once. The criteria of presyncope

were taken as in Grasser et al. 2008: (a) blood pressure

fell below systolic 80 mmHg, or that it dropped rapidly(systolic (SBP) by C20 mmHg/min, diastolic (DBP) by

C10 mmHg/min), or HR dropped by C15 bpm; (b) light-

headedness, dizziness, visual disturbances, nausea, stom-ach awareness, clammy skin, excessive sweating, or skin

pallor. These were the criteria of termination but all the

subjects went through all the protocols with no problems.Test subjects were given guidance to avoid during the test

unduemovements of the lower limbs and to breathe normally.

Test persons were secured and had access to an emergencyshutdown (automatic return to supine and pressure neutral-

ization) at all times.Aphysicianwas in attendance to evaluate

subject reaction to the increasing LBNP stress.

Test Apparatus and Measurements

All tests were carried out at the Institute of Adaptive and

Spaceflight Physiology (www.meduni-graz.at/iap/AHST.

Fig. 1 Graded orthostatic stress protocol. After 30 min of supine rest, 5 min HUT was followed by 3-min increments of LBNP until presyncopalsigns or symptoms occurred

Cardiovasc Eng

123

htm) using a combined tilt-table-LBNP chamber device

equipped with a footrest. Care was taken to maintain thesealing at the iliac crest, as sealing position has been shown

to affect hemodynamic responses (Goswami et al. 2009a).

A transition from supine to upright position as well asnegative pressure buildup was complete within 5 s. The

execution of the pre-programmed test protocol and syn-

chronous recording of all data from the cardiovascularmonitoring system was done by LabView#.

Electrodes were placed at the neck and thoracic regions,the latter specifically at the midclavicular line at the

xiphoid process level (Fortin et al. 2006). Continuous

hemodynamic monitoring included systolic and diastolicblood pressure, HR (3-lead ECG), and thoracic impedance

using the Task Force Monitor#, CNSystems, Graz, Austria.

This monitor estimates arterial pressure using finger cuffs(the principle of vascular unloading technique was used to

estimate the arterial blood pressure (Penaz principle) and

regular calibration to standard arm cuff measurements (byTask force monitor). Mean arterial blood pressure (MBP)

was calculated from diastolic and systolic pressures, with

the estimation given as: MBP = DBP ? 1/3 (SBP - DBP).Pulse pressure (PP) was calculated as SBP - DBP.

Changes in thoracic impedance indicate changes in tho-

racic fluid volume over time; impedance cardiography wasperformed based on the original Kubicek approach, using

an improved estimate of thoracic volume. Impedance car-

diography data correlate well with direct measures such asthose from echocardiography or echo Doppler flowmetry.

Thoracic fluid content (TFC) represents the inverse of

baseline impedance. See Goswami et al. 2009b, c foradditional details on protocol and experimental design.

Data Analysis and Interpretation

Using typical cardiovascular changes during orthostatic

loading from previous studies (Gao et al. 2008; Goswamiet al. 2009b) error probability (a) of 0.05, power (1 - b) of0.80, we estimated the number of subjects required to be

10.

Data Preparation

The data was preprocessed by applying a simple moving

average smoother to the data using a convolution scheme

CONVN given in MATLAB 7.01 utilizing a moving win-dow of 50 data points resulting in a vector of processed

data points with data points approximately 30 s apart.

Analysis was carried out only for the LBNP stagesending at the time point 30 s prior to when HR was max-

imum. This stopping time was chosen because maximum

HR was seen as a reasonable marker for delineating the endof the effectiveness of the control response.

Orthostatic tolerance times here refer to the time during

which LBNP was applied until presyncopal symptomsdeveloped. It was calculated as total orthostatic tolerance

time (see) minus HUT phase (300 s).

Linear Regression Models

Linear regression was employed to characterize systembehavior during the LBNP phase. Regression models were

employed defined by the general form:

Y " X#A$ E %1&

where Y is an n-by-1 vector of n observations, X is an n-by-p matrix of regressors, A is a p-by-1 vector of parameters,

and E is an n-by-1 vector of random normal disturbances.For this study, linear regression was carried out for each

of the four tests of each subject to find the slope of each

state variable against time for the following state variables:HR, TFC, SV, MBP, PP, SBP, and DBP. Thus for each

state variable Y, the calculated regression parameter A

provided an estimate of the slope of the change of the givenstate versus time over the stages of increase in LBNP. The

regression and goodness of fit was provided by the MatLab

function regress in terms of the residual vector of R2

values.

The standard deviation of the slope of the MBP over

four runs was used to classify the subjects into consistentand inconsistent groups. To observe intra-individual con-

sistency in other measured hemodynamic variables, the

standard deviations in the slopes were also used, with alower standard deviation indicating higher consistency.

The two-sided t-test was used for comparing the slopes

of MBP, HR, SV. Units for slopes are: HR (beats/min/min),TFC (1/kX/min), stroke index (SI) (mL/L/min), blood

pressures (mmHg/min).

Results

Table 1 provides an overview of the slopes of states versus

time for selected state variables (MBP, HR, SV, and

orthostatic tolerance times) during increasing LBNP phasefor 10 subjects (A–J). Consistency in slope over four trials

for these states was in some subjects very clear, while in

other cases some degree of variation in the slopes wasobserved. However, over four trials, two clear patterns

emerged when comparing the changes (slopes) of MBP

(Table 1): 5 subjects (A B H I J) showed high consistencywithin their slopes (with maximum standard deviation of

0.7 mmHg/min) while the remaining subjects (C D E F G)

exhibited low consistency (with minimum standard devia-tion of 1.2 mmHg/min). Table 1 also indicates that mean

values for MBP are not significantly different for each

Cardiovasc Eng

123

group (p = ns). Table 2 gives the hemodynamic variables,

slopes, and associated R2 values for the regression com-putations from which the slopes of the hemodynamic

variables were derived. R2 values for HR are very high for

all subjects. TFC also exhibited very high R2 values exceptfor two subjects from the inconsistent group F and C.

Similarly SV or SI have high R2 values with 3 exceptions.

However, the estimated values for SI may include suffi-cient error to obscure the results. As can be seen from the

table linear regression is not so clearly successful based onR2 values for MBP, SBP, and DBP. However, PP R2 values

were in general very high. This is partly to be expected

given that there is some nonlinearity in the behavior ofthese pressure components during orthostatic stress.

Figure 2 provides the time course of the average of all

subjects over the entire protocol for HR and MBP.In regards to MBP, Table 2 shows that the slopes are

close to zero with, in some cases, low R2. Figure 2 also

gives an illustration of the significant variation in thisvalue. We tried fitting the MBP data with a trend of 2nd

and 3rd order, but the residuum of the 2nd order was not

better then for the linear case and the polyfit of 3rd orderstopped with a ‘‘bad’’ result. However, these results are

likely due to several factors including the existence of very

many data points in the beat-to-beat measurements, and thefact that the slope is nearly flat, with noisy data scattered

around this flat slope. Therefore, in this case we accepted

as significant for comparison purposes the calculated MBPslopes even with a low R2 value and used these slopes as a

rough indicator for the ‘‘overall’’ behavior for the change

of the MBP. Figure 3a–c are representative examples ofregression for MBP with associated R2 values. These

figures depict how the slope of MBP can give a represen-

tation of the direction of the change in MBP.To get a sense of the variation in responses between

subjects, Fig. 4 provides relative changes in TFC, HR, and

SI from the baseline to the first occurrence of syncopalsigns or symptoms across the 4 runs in subject F and H (left

and right columns, respectively). These time courses indi-

cate the degree of variation in the regularity in slopes andthe potential influence of the TFC on the degree of stress

induced by the protocol.Figure 5a (depicting data from subject A) and 5b

(depicting data from subject H) illustrate the calculated

slopes of HR and SV across the four runs in two subjectswith high consistency in MBP regulation.

Figure 6a (subject D) and 6b (subject F) show similar

variables in two subjects exhibiting low consistency inMBP regulation. Subject D exhibits significant responses in

HR, and SV across the runs with some bias towards reli-

ance on HR. Subject F also shows significant responses inboth variables.

Table 1 also shows the duration of LBNP application

(in other words, the orthostatic tolerance during the LBNPphase), which is discussed below.

Discussion

LBNP applied to subjects who are already in HUT position,represents a significant orthostatic challenge. Table 1 pro-

vides calculated values of slopes of key hemodynamic

variables reflecting subject response to the LBNP phaseover four runs of the combined HUT ? LBNP protocol.

Table 1 Overview of the slopes of states versus time of selected state variables (MBP, HR, SV, and orthostatic tolerance) during increasingLBNP phase for the 10 subjects (A–J)

MBPSlope ± ns HRSlope ± p\ .05 SVSlope ± p\ .05 OT ± ns

Panel A

J -1.35 ± 0.12 4.31 ± 0.70 -0.99 ± 0.37 7.97 ± 1.04

A -0.25 ± 0.35 4.84 ± 1.34 -2.01 ± 0.60 21.21 ± 4.02

I 0.37 ± 0.37 3.11 ± 0.09 -0.78 ± 0.60 7.19 ± 2.77

H 0.52 ± 0.63 5.01 ± 0.59 -1.58 ± 0.50 16.33 ± 2.57

B -0.06 ± 0.70 4.66 ± 0.75 -1.55 ± 0.41 13.80 ± 1.62

Panel B

E -0.31 ± 1.16 5.94 ± 0.99 -2.27 ± 0.75 12.22 ± 1.67

D -1.57 ± 1.21 8.38 ± 2.02 -3.55 ± 0.97 8.43 ± 2.77

G -0.62 ± 1.44 3.77 ± 0.65 -1.79 ± 0.31 17.95 ± 5.11

F -1.66 ± 1.45 7.04 ± 1.30 -1.67 ± 0.70 5.07 ± 1.51

C -0.94 ± 1.90 8.26 ± 2.16 -3.11 ± 0.65 5.13 ± 1.53

Panel A describes the consistent group and Panel B the inconsistent group (in relation to MBP regulation). Statistical comparisons of the meansof the slopes between the two groups over all the states appear in the first row

Slope refers to changes calculated per minute. All values are mean ± SD

MBP mean blood pressure (mmHg), HR heart rate (bpm), SV stroke volume (ml/l), OT orthostatic tolerance during LBNP phase (min)

Cardiovasc Eng

123

Stabilization of blood pressure is implemented via

negative feedback loops that respond to sensory informa-tion about an increase or decrease in systemic arterial

pressure (via the arterial baroreflex) as well as changes in

central venous and pulmonary blood pressures (via thecardiopulmonary baroreflex). These control loops adjust

heart rate, heart contractility, venous compliance, unstres-

sed volume (local mechanisms may also play a role here)

and systemic arterial resistance to counter perturbations in

blood pressure (e.g. Cooper and Hainsworth 2001; Kappelet al. 2007). There need not be, in principle, a unique

combination of responses to achieve blood pressure sta-

bilization. In this study, we observed that the degree ofconsistency in MBP regulation over the four test runs

classifies subjects into two distinct categories with an equal

number of members in each group. We will refer to the

Table 2 Slopes and the correlation coefficients of the changes within heart rate (HR), thoracic fluid content (TFC), stroke index (SI), pulsepressure (PP), systolic blood pressure (SBP) and mean blood pressure (MBP) of all subjects

HR slope R2 TFC slope R2 SI slope R2

J 4.31 ± 0.70 0.86 ± 0.08 -0.12 ± 0.09 0.75 ± 0.32 -0.49 ± 0.18 0.54 ± 0.27

A 4.84 ± 1.34 0.95 ± 0.01 -0.08 ± 0.02 0.86 ± 0.07 -0.99 ± 0.30 0.92 ± 0.03

I 3.11 ± 0.09 0.90 ± 0.04 -0.10 ± 0.05 0.84 ± 0.16 -0.40 ± 0.31 0.45 ± 0.28

H 5.01 ± 0.59 0.95 ± 0.04 -0.12 ± 0.04 0.96 ± 0.02 -0.91 ± 0.29 0.87 ± 0.09

B 4.66 ± 0.75 0.93 ± 0.04 -0.19 ± 0.01 0.98 ± 0.01 -0.87 ± 0.23 0.90 ± 0.05

E 5.94 ± 0.99 0.93 ± 0.02 -0.19 ± 0.05 0.96 ± 0.03 -1.17 ± 0.39 0.83 ± 0.08

D 8.38 ± 2.02 0.91 ± 0.04 -0.14 ± 0.02 0.91 ± 0.04 -1.95 ± 0.53 0.94 ± 0.04

G 3.77 ± 0.65 0.93 ± 0.03 -0.09 ± 0.01 0.96 ± 0.02 -0.87 ± 0.15 0.93 ± 0.03

F 7.04 ± 1.30 0.79 ± 0.08 -0.06 ± 0.13 0.49 ± 0.40 -0.75 ± 0.32 0.41 ± 0.25

C 8.26 ± 2.16 0.88 ± 0.10 -0.09 ± 0.05 0.67 ± 0.19 -1.72 ± 0.36 0.92 ± 0.03

PP slope R2 SBP slope R2 MBP slope R2

J -1.03 ± 0.17 0.69 ± 0.21 -2.00 ± 0.19 0.71 ± 0.19 -1.35 ± 0.12 0.62 ± 0.17

A -0.70 ± 0.23 0.79 ± 0.12 -0.69 ± 0.50 0.37 ± 0.26 -0.25 ± 0.35 0.16 ± 0.22

I -1.14 ± 0.52 0.94 ± 0.01 -0.45 ± 0.06 0.18 ± 0.11 0.37 ± 0.37 0.10 ± 0.08

H -1.71 ± 0.63 0.90 ± 0.04 -0.67 ± 0.87 0.49 ± 0.25 0.52 ± 0.63 0.32 ± 0.32

B -0.46 ± 0.92 0.58 ± 0.27 -0.41 ± 1.28 0.48 ± 0.29 -0.06 ± 0.70 0.32 ± 0.26

E -1.01 ± 0.79 0.66 ± 0.40 -1.09 ± 1.17 0.57 ± 0.32 -0.31 ± 1.16 0.39 ± 0.34

D -1.76 ± 0.78 0.94 ± 0.02 -2.71 ± 1.70 0.73 ± 0.30 -1.57 ± 1.21 0.58 ± 0.38

G -0.49 ± 0.12 0.71 ± 0.13 -0.93 ± 1.46 0.36 ± 0.44 -0.62 ± 1.44 0.46 ± 0.32

F -1.83 ± 0.82 0.75 ± 0.14 -2.85 ± 1.46 0.60 ± 0.36 -1.66 ± 1.45 0.49 ± 0.37

C -2.42 ± 0.41 0.86 ± 0.11 -2.64 ± 1.49 0.66 ± 0.40 -0.94 ± 1.90 0.63 ± 0.23

Fig. 2 The influence of HUTapplication on the subsequentLBNP induced hemodynamicchanges (HR and MBP). Thevalues indicated here are themeans and the standarddeviations (SD) of all the 10subjects across the four runs.LBNP20 represents only theinitial phase of the gradedLBNP

Cardiovasc Eng

123

group with lower standard deviation in MBP slope as the

consistent MBP group and the other group with high

standard deviation as the inconsistent MBP group.

Mean Blood Pressure and Heart Rate Control

The differences between the two MBP groups seem to

indicate various degree of activation of different reflex

pathways of blood pressure control. For example, those in

the consistent group tended to use HR in a consistent way

as follows:

• The 5 subjects most consistent in MBP variation were

chosen by ranking standard deviation in slopes of MBP

over four runs. We also noted the clear break in thestandard deviation in MBP slope between the 5 subjects

Fig. 3 Three representativesubject sets of the 4 runs asexamples of regression of MBPwith varying R2 values:a subject F, b subject A,c subject B

Cardiovasc Eng

123

Fig. 3 continued

Fig. 4 Relative changes in HR,TFC and SI from the baseline tothe first occurrence of syncopalsigns or symptoms across the 4runs in subject F and H (left andright panels, respectively)

Cardiovasc Eng

123

that were most consistent and the five least consistent in

MBP slope.

• Ranking subjects by magnitude in HR slope wouldresult in the 5 most consistent in MBP slope having the

smallest HR slope. We noted again a clear breakbetween the lower and higher HR slopes (subject G

from the inconsistent MBP group would be an excep-

tion). Subjects in the inconsistent group generally hadHR slopes nearly twice that of the consistent group.

• Subjects could also be grouped by percentage variation

in slopes. Again, 4 of 5 subjects of the consistent MBPgroup had the lowest variation in slope of HR, while all

of the inconsistent MBP subjects were in the highest

ranking of variation (Subject A from the consistentgroup would be an exception). For some subjects (C, D,

F) in the inconsistent MBP group, variation in slopes

over the four runs were 20–25% of the average HRslope for these subjects. The division line in this form

of grouping, however, was not so clear compared to the

previous groupings given above, with the highest slopevariation in the low (consistent) MBP variation group

16.2% and the lowest HR slope variation in the

inconsistent MBP variation group 16.7%. It is strikingthough that within the inconsistent MBP group were

found all the very high HR slopes and all the higher

variations in HR slopes while in the consistent groupthe lowest variations in HR slope were observed.

• Thus, considering both magnitude of HR slope and

variation in HR slope as a joint criterion, there can be

defined two groups: the first group with the lower jointvalues being exhibited by 4 of the consistent MBP

subjects and the group with the higher joint valuesrepresented by 4 of the inconsistent MBP group. There

were 2 exceptions (subject G and A).

• Clearly these observations are on a small set of subjects,and the existence of exceptions to the groupings

discussed above prevents any clear statistical inferences

being made. Nevertheless, it appears that patterns didemerge from quantifying the slopes of MBP and HR

states over the LBNP phase.

• Finally, it is worth noting that very high R2 values forHR control in both groups suggest that even under

extreme stress there is no shifting away from HR

control response.

General Considerations

To maintain blood pressure and especially adequate cere-bral perfusion during central hypovolemia caused by HUT,

LBNP, or hemorrhage, control reflex responses increase

HR and total peripheral resistance. During graded levels ofLBNP, progressive increases in muscle sympathetic nerve

activity (MSNA) and increased systemic vasoconstrictor

responses have also been reported.

Fig. 5 Linear regression of HRand SV across the four runs intwo subjects with highconsistency a subject A,b subject H

Cardiovasc Eng

123

The increases in HR as well as the decreases in SV

during the LBNP phase were comparable to what has been

reported by others. In our study, in which only the gradedLBNP phase was examined, consistent blood pressure

maintainers employed in a consistent way the HR (subjects

A, B, H, I, J) across the four runs. Subject A is a partialexception in that the HR slope is low but the variation over

four runs is not.Inconsistent blood pressure maintainers, on the other

hand, exhibited higher HR slopes and generally high var-

iation in slope suggesting that the system was operatingclose to the edges of stable operation. Thus over four runs

these subjects invoked variations in the HR response,

which resulted in inconsistent patterns of response over thefour runs in MBP.

Observations based on SI and TFC are less clear-cut

partly because of the way they are estimated (see below).

Orthostatic Tolerance in the Two Groups of Blood

Pressure Regulators

Syncope is the end-point of orthostatic stability and it is

characterized by a rapid drop in mean arterial pressureinduced by a sudden drop in heart rate, stroke volume or

total peripheral resistance (Grasser et al. 2009). We,

therefore, speculated that subjects who showed consistent

patterns in blood pressure regulation, suggestive of a more

stable and effective response strategy, across the four runs

would have greater orthostatic tolerance times. However,the orthostatic tolerance times during the LBNP phase were

not different between the two groups (Table 1).

Effects of Passive HUT on the LBNP Induced

Responses

The thoracic impedance, reflecting the fluid shift time

course, estimates changes in cardiac preload, which chan-ges considerably during the test situation and drives

accordingly cardiovascular reflexes that adjust HR and

peripheral vascular resistance.Additionally, variations in the thoracic fluid shifts dur-

ing the LBNP phase might account for the differences in

the consistency of blood pressure response between the twogroups. However, we observed no differences in the slopes

of the TFC between the two groups.

It is also possible that the initial HUT phase (not used inthis study) of the combined HUT ? LBNP could have

resulted in varying levels of central volume at the begin-

ning of the LBNP phase, which could affect the responsepatterns during the LBNP phase (Fig. 4). For example,

exploratory’ regulatory features might be called upon in

persons with lower functional central volume reserve,induced by the HUT, to maintain blood pressure during

Fig. 6 Linear regression of HRand SV in two subjects with lowconsistency a subject Db subject F

Cardiovasc Eng

123

graded LBNP. That is, such persons might more fully

employ the possible range of cardiac and peripheral-vas-cular response patterns available in order to compensate for

limited central volume reserve. We, however, observed no

differences in the TFC changes caused by the HUT phasebetween the two groups (p = ns).

It is possible that the TFC changes for a given individual

shifted the operating point of the control near to maximalresponse which could induce irregular patterns in control

actions across four runs. This factor would not be detectedby the slopes of TFC alone.

Conclusions

Young healthy males subjected to graded lower bodynegative, while in upright tilted position, could be classi-

fied into two groups based on consistency of MBP regu-

lation across four runs. Subjects in the consistent groupexhibited reliance on one cardiovascular variable to

maintain their pressure but the subjects in the inconsistent

groups did not exhibit such consistency and may have beenoperating near the edge of control system tolerance. These

results, obtained in healthy normal young males, provide

valuable insights into cardiovascular regulation as testpersons reach syncope. As syncope is multifactorial in

origin, further studies are needed to see if these results

apply also to persons with syncopal histories or diseaseconditions.

Acknowledgments We wish to thank the subject participants’ fortheir time and patience. Funding: Partially funded by AustrianResearch Funds under Project 18778-N13, and the Austrian ResearchPromotion Agency (FFG project 817086 ‘Orthocap’).

References

Cooper VL, Hainsworth R. Carotid baroreceptor reflexes in humansduring orthostatic stress. Exp Physiol. 2001;86:677–81.

Fink M, Batzel JJ, Kappel F. Modeling orthostatic stress in thecardiovascular respiratory system. Cardiovasc Eng. 2004;4:27–38.

Fortin J, Habenbacher W, Heller A, Hacker A, Gruellenberger R,Innerhofer J, et al. Noninvasive beat-to-beat cardiac outputmonitoring by an improved method of transthoracic bioimped-ance measurement. Comput Biol Med. 2006;36:1185–203.

Gao YF, Goswami N, Grasser EK, Roessler A, Stoger E, SchwabergerG, et al. Radix astragali and orthostatic response: a double-maskedcrossover study. Aviat Space Environ Med. 2008;79:94–8.

Goswami N, Grasser EK, Roessler A, Schneditz D, Hinghofer-Szalkay H. The cardiovascular response to lower body negativepressure in humans depends on seal location. Physiol Res.2009a;58:311–8.

Goswami N, Lackner HK, Grasser EK, Hinghofer-Szalkay H.Individual stability of orthostatic tolerance response. ActaPhysiol Hung. 2009b;96:157–66.

Goswami N, Roessler A, Lackner HK, Schneditz D, Grasser EK,Hinghofer Szalkay H. Heart rate and stroke volume responsepatterns to augmented orthostatic stress. Clin Auton Res.2009c;19:157–65.

Goswami N, Loeppky JA, Hinghofer-Szalkay H. LBNP: pastprotocols and technical considerations for experimental design.Aviat Space Environ Med. 2008;79:459–71.

Grasser EK, Goswami N, Hinghofer-Szalkay G. Presyncopal cardiaccontractility and autonomic activity in young healthy males.Physiol Res 2008 [Epub ahead of print].

Grasser EK, Goswami N, Roessler A, Vrecko K, Hinghofer-SzalkayH. Hemodynamic and neurohormonal responses to extremeorthostatic stress in physically fit young adults. Acta Astronaut.2009;64:688–96.

Hinghofer-Szalkay H, Goswam N, Roessler A, Grasser EK, SchneditzD. Reactive hyperemia in the human liver. Am J PhysiolGastrointest Liver Physiol. 2008;295:G332–7.

Kappel F, Fink M, Batzel JJ. Aspects of control of the cardiovascular-respiratory system during orthostatic stress induced by lowerbody negative pressure. Math Biosci. 2007;206:273–308.

Laszlo Z, Roessler A, Hinghofer-Szalkay H. Cardiovascular changesduring and after different LBNP levels in men. Aviat SpaceEnviron Med. 1998;69:32–9.

LeLorier P, Klein GJ, Krahn A, Yee R, Skanes A, Shoemaker JK.Combined head-up tilt and lower body negative pressure as anexperimental model of orthostatic syncope. J Cardiovasc Elec-trophysiol. 2003;14:920–4.

Cardiovasc Eng

123