Heart rate variability: How to assess effects of mild therapies on autonomic control in small groups of mild and borderline hypertensives?
In their recent retrospective study [1] Singh et al. could demonstrate
on the basis of the impressively huge data set of the Framinghan Heart
Study that, first, short-term HRV is reduced in men and women with systemic
hypertension and, second, among normotensive men, lower HRV was associated
with greater risk for developing hypertension. The authors concluded that
autonomic dysregulation is present in the early stage of hypertension.
Their findings are important but were not very surprising because the reported
HRV reduction was generally in accordance with findings of earlier studies
[2]. However, what they inevitably left out of consideration was the analysis
of the diurnal variation of blood pressure, heart rate and their respective
variabilities. Particularly the asleep/awake ratios of blood pressure are
probably more specific and sensitive than temporary daytime parameters
[3]. From the methodical points of view, the study was characterized by
the strong and uncompromising use of statistics, but unfortunately without
showing any raw data, e.g. by using scatter or box plots of the blood pressure
and HRV data. Moreover, one important question remained unanswered: How
did the HRV parameters and the covariates change during 4 years of follow-up?
Independently and without knowledge of the results from the above study,
we also studied the differences of linear and nonlinear HRV parameters
in (only) 25 essential non-treated hypertensive subjects with respect to
their status of hypertension. The study was carried out from spring to
fall 1998 and the results are not yet published. The purpose of our study
was to gain experience in the collection and interpretation of HRV data
from hypertensives for further studies. When comparing our HRV mean values
with the values from the Singh study, we were very surprised: After log
transformation, the mean values of LF, HF, and LF/HF were approximately
identical with those of Singh and colleagues in Table 2 of their paper.
As a result of the small N, our SEM (standard error of the mean)
values were up to 10 times higher than those of the huge Framingham group.
Consequently, significant differences between subgroups could not be demonstrated
and both specificity and sensitivity of all HRV parameters were extremely
poor. The separation of subgroups was much better for the nocturnal BP
fall which could not be observed by Singh et al. by reason of the Framingham
study design. And we achieved better results analyzing the 24-hour blood
pressure course and using nonlinear HRV parameters instead of the linear
spectral HRV markers LF and HF. The most prominent correlation, for example,
could be observed between the relative nocturnal blood pressure fall and
the approximate entropy (ApEn) of daytime heart period dynamics [4]. As
the clinical relevance of our observations remains to be proven, it makes
sense if, in future, results like ours could be taken into consideration
when analyzing large clinical databases of heart beat and blood pressure
data. Particularly, it would seem to be very promising to analyze the time
course of 24-hour BP level, if available, and to include also nonlinear
measures in 24-hour HRV analysis.
Another problem is how to make use of subtle group differences of huge
cross sectional studies, like those of Singh and colleagues, when dealing
with only a few, but very individual subjects. And what does adjustment
of measures for clinical covariates (e.g. age, gender, body mass index,
alcohol consumption, and cigarette smoking) mean in the clinical practice?
These problems and others, occurring in clinical practice as well
as in many clinical research settings, are not new, but most studies,
e.g. in hypertension, have not adequately taken the constraints of daily
clinical routine into consideration.
We therefore propose to design in future preferentially longitudinal
sectional or single case HRV studies rather than cross sectional clinical
HRV studies. These studies could address the question: How do HRV parameters
change in individuals over longer periods of time with respect to the change
of their status of hypertension and with respect to clinical covariates?
These studies would not provide odds ratios or similar epidemiological
parameters, but clinicians would be enabled to judge an increase or decrease
of HRV parameters in individuals, e.g. during therapy, which may be more
informative than one single starting value. It is a well known phenomenon
that on the one hand sensitivity and specificity of 24-hour HRV measures
are generally poor, but on the other hand reproducibility in individuals
is excellent (cf. [5, 6]). Thus, small changes of autonomic control, e.g.
as an effect of a mild anti-hypertensive intervention, may be well demonstrated
in individuals, but may be smeared in large populations.
We suppose that, when following the above recommendations, HRV methods
may help to gain further insight into subtle rhythmic and individually
different regulatory processes in the human organism. All HRV parameters
are per se mirrors of the whole human time organism, reflecting a multitude
of internally and externally triggered physiological rhythms influencing
each other. Mild therapies, like sports activities or psychosomatic therapies,
are often individually conceptualised to stimulate rhythmical processes
in the human organism and to enforce self-regulatory processes. Their therapeutic
effects are naturally difficult to recognise because they are masked by
various clinical or daily life activities that spontaneously influence
many clinical parameters more than the therapy itself. The analysis of
HRV in individuals, including methods from nonlinear dynamics and taking
the 24-hour heart rate and BP variations into consideration, altogether
could well have the power to become a useful diagnostic tool, particularly
in mild and long-term anti-hypertensive treatments.
Dirk Cysarz
Hans Christoph Kümmell
Department of Clinical Research
Gemeinschaftskrankenhaus
58313 Herdecke, Germany
- Singh JP, Larson MG, Tsuji H, Evans JC, Odonnell CJ, Levy D. Reduced heart rate variability and new-onset hypertension - Insights into pathogenesis of hypertension: the Framingham Heart Study. Hypertension. 1998; 32: 293-297
- Parati G, Frattola A, Omboni S, Mancia G, Di Rienzo M. Analysis of heart rate and blood pressure variability in the assessment of autonomic regulation in arterial hypertension. Clin Sci. (Colch) 1996; 91 Supl.: 129-132
- Kario K, Motai K, Mitsuhashi T, Suzuki T, Nakagawa Y, Ikeda U, Matsuo T, Nakayama T, Shimada K. Autonomic nervous system dysfunction in elderly hypertensive patients with abnormal diurnal blood pressure variation - Relation to silent cerebrovascular disease. Hypertension. 1997; 30: 1504-1510
- Bettermann H, Van Leeuwen P. Evidence of phase transitions in heart period dynamics. Biol Cybern. 1998; 78: 63-70
- Kleiger RE, Bigger JT, Bosner MS, Chung MK, Cook JR, Rolnitzky LM, Steinman R, Fleiss JL. Stability over time of variables measuring heart rate variability in normal subjects. Am J Cardiol. 1991; 68: 626-630
- Hohnloser SH, Klingenheben T, Zabel M, Schroder F, Just H. Intraindividual Reproducibility of Heart Rate Variability. PACE. 1992; 15: 2211-2214
Response:
We thank Bettermann et al for their comments and their interest in
our work[1]. We were unable to assess the diurnal variation of heart rate
variability (HRV) because our study was based on daytime ambulatory ECG
recordings. It is reassuring, however, to know that the mean values of
the LF, HF and LF/HF ratio (after log transformation) from our population-based
study (931 men and 1111 women) was nearly identical to values in their
study of 25 patients. This was despite the intermediate duration (2 hour)
of our recordings compared with their 24-hour recordings.
Bettermann et al raise an interesting point concerning the change in
HRV variables over time. Unfortunately, we did not have follow-up HRV data
to include in our report. Such data are being acquired now and will allow
us to address longitudinal changes in HRV in future studies. We agree that
it would be interesting to analyze the time course of 24-hour blood pressure
levels and nonlinear measures of HRV.
Power spectral measures of HRV are characterized by large inter- and
intra-subject variations [2], which probably reflect the dynamics of physiological
control mechanisms over time and even during steady-state conditions [2,
3]. This complicates the extrapolation of results from a population study
to an individual patient in clinical practice. As mentioned in our paper,
in theory, HRV, as a surrogate measure for autonomic tone, may aid in the
management of hypertension [4] by guiding the selection of an appropriate
drug. An extension of this concept by Bettermann et al, suggesting the
use of longitudinal/single case studies evaluating changes in the autonomic
profile of an individual patient over time and the effects of antihypertensive
therapy on the HRV parameters, deserves evaluation. This is reinforced
by recent data suggesting that postural changes in spectral profiles can
be recognized and forecast within individuals [4].
National Heart, Lung, and Blood Institute’s
Framingham Heart Study
Framingham, Massachusetts
Martin Larson, DSc
Boston University School of Medicine
Boston, Massachusetts
Daniel Levy, MD
Massachusetts General Hospital Harvard Medical School
Boston, Massachusetts
- Singh JP, Larson MG, Tsuji H, Evans JC, O’Donnell CJ, Levy D. Reduced heart rate variability and new-onset hypertension: insights into pathogenesis of hypertension: the Framingham Heart Study. Hypertension. 1998;32:293–297.
- Task force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Heart rate variability standards of measurement, physiological interpretation, and clinical use. Circulation. 1996;93:1043–1065.
- Malliani A, Pagani M, Furlan R, Guzzetti S, Lucini D, Montana N, Cerutti S, Mela GS. Individual recognition by heart rate variability of two different autonomic profiles related to posture. Circulation. 1997;96:4143–4145.
- Salo TM, Viikari JS, Antila KJ, Voipio-Pulkki LM, Jalonen JO, Valimaki IA. Antihypertensive treatment and heart rate variability in diabetic patients: role of cardiac autonomic neuropathy. J Auton Nerv Syst. 1996;60:61–70.
Last modified 17/06/2005 19:52 (GMT+2)
Time of request 23/08/2019 11:12 (GMT+2)
© Copyright 2005 Scientific African