| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
LETTERS TO THE EDITOR |
Department of Psychiatry, Washington University School ofMedicine, St. Louis, MO63108
Watkinset al. (1) recently published aninteresting study in which vagal control correlated with state anxiety, butnot with depression severity in depressed psychiatric patients. Without anondepressed control group, they could not determine whether vagal control islower in depressed than nondepressed patients. Nevertheless, the associationof anxiety with vagal control is an important finding that may help to clarifyhow depression increases the risk of cardiac events in patients with coronaryheart disease (CHD).
Although this paper makes avaluable contribution, it does not adequately address some of the limitationsof the study. Furthermore, some of its conclusions are not well supported. Thescatter plot (refer to Figure 1 on page 336 of Watkins et al.(1) suggests that the correlationbetween baroreflex activity (BRC spec) and state anxiety may be largely due tothree outliers. The correlation might not be significant if these cases wereremoved. Also, the sample included 39 women and 17 men. Women may be morelikely to report anxiety and to havelower vagal tone than men. Thus, it would be of interest to determine thecorrelation after adjustment for gender.
The authorsattempt to reconcile their findings with two reports of reduced heart ratevariability (HRV) in depressed patients with CHD(2, 3). They state thatalthough depression has been associated with reduced HRV in CHD patients, ithas not been in "healthy volunteers," citing three studies ofpsychiatric patients with depression to support this contention. However, theyomit four studies which did find arelationship in psychiatric patients with depression(4–7),and they do not critically review the studies cited as supporting theirposition.
One of these studies compared unmedicateddepressed patients with depressed patients on tricyclic antidepressants andcontrols who were not depressed (8). Tricyclic antidepressants(TCAs) increase heart rate, and the TCA group had higher heart rates and lowervagal tone than the unmedicated depressed patients and the controls. Althoughthe unmedicated patients did not have lower vagal tone, they did havesignificantly higher heart rates than the controls. A second study compareddepressed patients with high- vs. low-trait anxiety and with nondepressedcontrols (9). The depressed/highanxiety patients had lower HRV than the controls. However, only sevendepressed patients had low anxiety, and their HRV was between that of thecontrols and the depressed/high anxiety patients. This report by Tulen et al.(9) correctly concluded that largerstudies are needed to determine whether depression is associated with lowerHRV, independent of anxiety.
Watkins et al. attempt toreconcile studies of HRV in depressed psychiatric patients with those indepressed CHD patients. They note that whereas HRV has been measured overbrief intervals during rest in depressed psychiatric patients, it has insteadbeen measured over 24-hour periods in depressed CHD patients. This is true forthe studies they cited, but other studies of psychiatric patients have foundeffects of depression on short-term measures of HRV(4, 5). They argue thatthe association of depression with low HRV probably is related to reducedphysical activity in depressed patients. However, one study controlled foractivity level by assessing HRV in activity-restricted inpatients on the dayafter cardiac catheterization (2).HRV differed significantly between depressed and age- and sex-matchednondepressed CHD patients.
In addition, we question theauthors’ implicit assumption that HRV is determined only by vagalcontrol. This is the basis for their effort to reconcile their findings withthe two studies that found reduced HRV in depressed CHD patients. In fact, HRVis determined both by parasympatheticand sympathetic modulation of theintrinsic cardiac pacemakers, as well as by neuroendocrine and circadiancontributions (10). Thus, low HRVreflects inadequate cardiac parasympathetic and/or excessive cardiacsympathetic tone. Some HRV indices primarily reflect vagal tone, but othersreflect both sympathetic and parasympathetic control, as well asneuroendocrine and circadian modulation of heart rate(11).
Somestudies of depressed psychiatric patients have found that vagal control tendsto be relatively low in this population(4, 6, 7).However, the two studies of depression in CHD patients that found significantdifferences between depressed and nondepressed patients used an HRV index thatrepresents all sources of variability, including parasympathetic, sympathetic,and circadian. In fact, the one study that included HRV indices of vagal tone,in addition to the broader index of HRV, did not find any differences betweendepressed and nondepressed patients on vagally modulated indices of HRVdepressed and nondepressed patients on vagally modulated indices of HRV(2). Interestingly, most studies ofHRV predictors of malignant arrhythmias or mortality in post-MI patients havefound that decreases in mixed parasympathetic/sympathetic and circadianindices are the strongest predictors(11). Vagal indices more oftentend to be weak or nonsignificant predictors(12). Clearly, the relationshipsamong depression, anxiety, HRV, vagal tone, and baroreflex sensitivity, andcardiac events in depressed CHD patients are more complex than Watkins et al.havesuggested.
REFERENCES
Department of Psychiatry and Behavioral Sciences, Duke University Medical Center, Durham, NC 27710
Hebrew Rehabilitative Center for the Aged, Boston, MA 02131
Carneyand colleagues believe that our paper did not adequately address somelimitations of our study and suggested that our data might not support ourstudy conclusions. Specifically, they were concerned about the potentialinfluence of outliers and gender bias in our sample, considered our review ofliterature selective and nonrepresentative, and believed that our assumptionsabout the nature of heart rate variability (HRV) wereflawed.
Our results were not due to outliers nor to apreponderance of women in our sample. We reanalyzed our data, omitting threepotential "outliers" and adjusting for gender. Results of theseanalyses support our initial findings that anxiety is related to baroreflexcontrol (ß=0.27, p <.05).
Carney implied that our review of the literaturewas biased and noted four additional studies that were not cited in ourreview. However, two of these studies were not peer-reviewed empiricalstudies: one was an abstract published 10 years ago that was never reported asa full-length manuscript (1) andthe other simply reported pilot data in the context of a review article(2). A third study was published inJapanese (3) without Englishtranslation, and data from the fourth study were inconclusive insofar as lowHRV was observed in "melancholic," but not "reactive"depressed patients (4). Thesereports do not significantly alter our interpretation of the extant literaturein the field.
It is now well established that alteredcardiac autonomic function adversely affects prognosis in cardiac patients.Numerous studies have demonstrated that increased sympathetic and decreasedparasympathetic nervous system activity increase the risk of ventriculartachycardia, ventricular fibrillation, and sudden cardiac death(5, 6). Analysis of HRVhas emerged as a promising method for quantitating autonomic function. HRV isgenerally thought to reflect sympathetic and parasympathetic balance in theautonomic regulation of the heart(7), so that low levels of HRVsuggest increased risk of sympathetically mediated lethal arrhythmias(8). Low HRV has been shown to bean independent predictor of mortality in post-MI patients(9, 10) and in patientswith heart failure (11), withmortality best predicted by reductions in long-term indices of HRV such as24-hour standard deviation of R-R intervals(9) or very low frequency power(10). Furthermore, recent studieshave reported that reduced vagal baroreflex control and new power law measuresof HRV are stronger predictors of mortality than conventional HRV indices(12–15).These findings suggest that prognosis may be more closely related to theadaptability of the cardiovascular regulatory system than to the absolutelevel of autonomic activity.
Although several studies,including our own work, have shown that low HRV is associated with depression(16–18),we suggested that physical activity and anxiety may also contribute to thisassociation. Carney et al. question whether there is evidence to support ourcontention that depression-related physical inactivity may contribute to thelow HRV reported by studies that derive HRV indices from HR measured duringambulatory conditions. Physical activity has been shown to be a majordeterminant of HRV determined in an ambulatory setting(19) and physical conditioning hasbeen associated with higher HRV, even during quiet resting conditions(20). In addition, there isevidence that depressed cardiac patients have both lower resting baroreflexcontrol and engage in regular physical exercise less frequently thannondepressed patients(21).
Insum, we agree with Carney et al. that the relationship between emotions,autonomic regulation, and clinical outcomes is complex. However, we stand byour observation that among a sample of depressed older men and women, anxiety,but not depression severity, is related to impaired vagal control of the heart.
REFERENCES
This article has been cited by other articles:
 |
|
 |
|
  D. K. Moser, B. Riegel, S. McKinley, L. V. Doering, K. An, and S. Sheahan Impact of Anxiety and Perceived Control on In-Hospital Complications After Acute Myocardial Infarction Psychosom Med, January 1, 2007; 69(1): 10 - 16. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Zipfel, A. Schneider, B. Wild, B. Lowe, J. Junger, M. Haass, F.-U. Sack, G. Bergmann, and W. Herzog Effect of Depressive Symptoms on Survival After Heart Transplantation Psychosom Med, September 1, 2002; 64(5): 740 - 747. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
