| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Departments of Psychosomatics and Psychotherapy (C.HL., M.K.), Cardiology (B.B., D.VL., B.P.), and Clinical Chemistry (L.B., J.S., W.S.), University of Göttingen, Göttingen, Germany.
Address reprint requests to: Christoph Herrmann-Lingen, MD, Department of Psychosomatics and Psychotherapy, University of Göttingen, von-Siebold-Strasse 5, D-37075 Göttingen, Germany. Email: cherrma{at}gwdg.de
Received for publication May 10, 2002; revision received August 7, 2002.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
METHOD: This was a cross-sectional study of 119 patients (46 with CHF, 76 controls with cardiovascular risk factors) in a tertiary care center. The study included assessment of CHF severity, ejection fraction, pro-ANP (microtiter assay), and psychosocial status (self-rating questionnaires for anxiety, depression, vital exhaustion, and quality of life).
RESULTS: The diagnosis and severity of CHF was significantly related to pro-ANP levels, bad physical quality of life, vital exhaustion, and depression. However, there was no significant effect of disease severity on anxiety. In CHF patients, pro-ANP was negatively correlated with anxiety (
= -0.30, p = .041). In the whole group, anxiety was independently predicted by vital exhaustion, depression, and younger age (overall adjusted R2 = 0.48). Pro-ANP plasma levels showed an additional, inverse association with anxiety (ß = -0.17, p = .013, adjusted R2 = 0.50). Predicted mean anxiety scores derived from this model showed a good fit with anxiety scores observed in subgroups defined by CHF severity.
CONCLUSION: Pro-ANP plasma levels are independently and inversely related to anxiety. Even in severe CHF with severely compromised quality of life, anxiety tends to decrease with high pro-ANP levels. This might be part of a negative feedback loop limiting psychological distress and its adverse autonomic consequences in severe heart failure.
Key Words: Pro-ANP, • congestive heart failure, • anxiety.
Abbreviations: ANOVA = analysis of variance;; ANP = atrial natriuretic peptide;; CHF = congestive heart failure;; HADS = Hospital Anxiety and Depression Scale;; LVEF = left ventricular ejection fraction;; MQ = Maastricht Questionnaire;; NYHA = New York Heart Association;; pro-ANP = N-terminal Pro-Atrial Natriuretic Peptide;; PTCA = percutaneous transluminal coronary angioplasty;; QoL = quality of life;; SF36 = Short Form 36 Health Survey.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
On the other hand, natriuretic peptide receptors have not only been identified in the cardiovascular system and kidneys but also in several brain regions. ANP binds mainly to natriuretic peptide receptor A. Stimulation of this receptor in the brain inhibits the activation of the hypothalamic-pituitary-adrenal axis and the subjective reactions to anxiety-inducing stimuli (6). Studies in rodents and in patients with panic disorder have shown that infusions of ANP suppress the psychological and sympathetic response to behavioral and biochemical triggers of panic and nonpanic anxiety (7, 8). It has therefore been suggested that ANP may be the inhibitory component of a feedback loop involved in the physiology of anxiety.
This might also be important in heart failure, because patients with heart failure suffer from considerable psychological distress, vital exhaustion, and depression. In these patients, high endogenous levels of ANP might help limit emotional arousal, which could otherwise result in a further increase in sympathetic activity and adverse cardiac outcomes.
However, the effect of ANP on emotional functioning in CHF has only been studied in one sample: Murberg et al. (9) examined correlations of subjective well-being and depression with plasma levels of pro-ANP but found no such association. To date, no data are available on possible associations between natriuretic peptides and anxiety in cardiac patients.
The aim of the present study was to test the hypothesis that high plasma levels of pro-ANP are associated with relatively low anxiety in CHF patients.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
As part of the study, the patients completed a set of psychological questionnaires, including the Hospital Anxiety and Depression Scale and the Maastricht Questionnaire for vital exhaustion. The Hospital Anxiety and Depression Scale (HADS) (10–12) is a widely used short self-assessment questionnaire, especially developed for physically ill patients. Its items mainly ask for psychological manifestations of (generalized) anxiety and depressive mood. Each of the two subscales consists of seven items. Possible subscale scores range from 0 to 21. The Maastricht Questionnaire (MQ) (13) asks for levels of vital exhaustion, which have been shown to predict adverse outcomes in patients with coronary heart disease. The 21-item version used for this study has a possible range from 0 to 42. One hundred six patients also completed the SF-36 Health Survey, the most frequently used generic multidimensional quality of life (QoL) scale (14, 15) , which was introduced into the study shortly after its beginning. T-standardized scores of the physical and psychosocial components will be reported for this instrument.
All patients received routine cardiological workup. In addition, detailed echocardiograms were performed for assessment of systolic and diastolic function, including left ventricular ejection fraction (LVEF). Clinical CHF severity was rated by an experienced cardiologist, who was blind to the patients’ pro-ANP levels and questionnaire scores. Ratings were performed according to New York Heart Association (NYHA) criteria.
Venous blood samples were drawn and centrifuged. Heparinated plasma aliquots were then frozen and stored at -80°C, until they were analyzed in a specialized laboratory. Pro-ANP was determined using microtiter immunoassays (Immundiagnostik, Freiburg, Germany). These measurements were performed according to the recommendations of the manufacturer, with values 
1320 fmol/ml considered normal.
All data were coded and entered into an anonymized database on a personal computer. For the statistical analyses, SPSS for Windows (V.10) standard software was used. Because of considerably skewed distributions of hormone levels, nonparametric statistics (Spearman-Brown correlations, Mann-Whitney U tests, Wilcoxon tests, Kruskal-Wallis ANOVA) were used if possible. For multivariate prediction of anxiety, we performed stepwise linear regressions. The analyses including pro-ANP levels as a predictor were computed twice: once with pro-ANP raw values and alternatively with log-transformed pro-ANP values.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
On the psychological questionnaires, CHF patients scored significantly worse than the control subjects on the physical components of quality of life, vital exhaustion, and depression. In contrast, there was only a minimal and statistically insignificant difference in anxiety scores.
Associations of Pro-ANP, Disease Severity, and Questionnaire Scores
In the whole group, pro-ANP showed high correlations with ejection fraction ( = -0.75) and NYHA functional class (Spearman-Brown’s = 0.80), thus reflecting severity of cardiac dysfunction and clinical disease severity.
Among all measures of cardiac disease severity, plasma levels of pro-ANP were most closely related to (bad) self-rated physical QoL ( = -0.48, p < .0005). In addition, pro-ANP showed significant positive associations with vital exhaustion ( = 0.32, p < .0005) and depression ( = 0.23, p = .011). Slightly weaker associations were observed between the other cardiac severity markers and physical QoL whereas both ejection fraction and NYHA functional class showed associations similar to pro-ANP with vital exhaustion and depression.
Interestingly, none of the physical severity markers was associated with anxiety ( = 0.05–0.09, all p > .30). This lack of correlation was due to a nonlinear association between pro-ANP and anxiety, which was clearly distinct from the continuous associations of pro-ANP with QoL, depression, and vital exhaustion. On the subgroup level, there was no correlation between anxiety and pro-ANP in the control subjects ( = 0.09, p = .46) but a significant negative correlation existed between anxiety and pro-ANP ( = -0.30, p = .041) in the CHF group.
In the next step, we classified patients by functional CHF severity. Patients without known heart disease and those with a history of CHF but no current symptoms (NYHA I, N = 7) were classified as asymptomatic (NYHA 0-I, N = 80); those with dyspnea at moderate exercise (NYHA II, N = 19) as moderate heart failure; and those with dyspnea at minimal exercise or at rest (NYHA III-IV, N = 20) as severe heart failure (Table 2). The resulting three subgroups had median pro-ANP plasma levels of 541 fmol/ml, 5885 fmol/ml, and 11342 fmol/ml, respectively (p < .0005). There were also differences in several of the baseline variables (Table 2). Of these, most were also significantly related to pro-ANP levels. However, of all baseline variables that differed significantly among subgroups, age was the only variable which was also significantly associated with anxiety, whereas anxiety was independent of ejection fraction, cardiac or psychotropic medications, and the presence of ischemic vs. nonischemic heart failure.
|
|
Of these, vital exhaustion, depression, and age were retained as significant independent predictors of anxiety in a stepwise multivariate regression model (adjusted R2 = 0.48). Patients had higher anxiety scores if they were vitally exhausted (ß = 0.45, p < .0005); younger (ß[age] = -0.24, p = .001); and more depressed (ß = 0.29, p = .003).
Z-standardized mean anxiety scores observed in the three disease severity subgroups and estimated anxiety scores derived from the regression model are shown in Figure 2. Although observed scores were fairly well predicted by the regression model in the first two subgroups with no or only moderate CHF, observed anxiety scores were significantly lower (p = 0.015) than predicted in patients with severe CHF.
|
|
This result was virtually unchanged, when log-transformed pro-ANP values were used instead of raw values (ß = -0.16, p = .020) or when the eight patients taking psychotropic drugs were excluded (ß[pro-ANP] = -0.19, p = .009). There was no additional effect of sex, diagnosis of heart failure, NYHA class, or ejection fraction.
When predicted z-standardized anxiety scores derived from the four-predictor model were compared with the observed mean anxiety scores, there was a clearly better fit, with no significant difference between observed and predicted anxiety scores in any of the subgroups (Figure 3). Especially the decrease of anxiety scores observed in the patients with severe CHF is at least partially reflected by the model, indicating that high pro-ANP was most probably related to this decrease.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This finding is important for several reasons.
First, it adds to our knowledge about physiological factors affecting emotional well-being and quality of life in heart disease. The literature on this subject has, until now, mainly focused on psychosocial risk factors and – to a lesser degree – physical correlates of bad quality of life. In this tradition, centrally stimulated levels of neurohormones have been studied with respect to their potentially hazardous cardiovascular effects. More recently, adverse effects of peripheral mediators (such as proinflammatory cytokines) on emotional function have also been suspected in coronary patients (16). However, none of the previous studies has examined the neuropsychological consequences of hormones secreted by the heart itself nor has there been a full physiological understanding of the negative feedback loops limiting the increase in emotional arousal occurring as a result of heart disease. Although successful coping will play a major part in this adaptation process, direct physiological effects might also be involved. The heart might hereby not only be the target organ of psychoneuroendocrine efferents but rather play an active part in the regulation processes – not only of body fluid content and vascular resistance but also of emotional and neurohumoral activation. This would make sense from a biological point of view: the emotional and physiological arousal often found in anxiety states would impose an additional burden on the failing heart and thereby exert a deleterious effect; and a possible anxiolysis produced by ANP in response to atrial overload could partly reverse, or at least limit, this effect.
Second, this finding also helps one get a more differentiated impression of the different psychological and psychosomatic states experienced by patients with heart failure.
The subjective reports of physical quality of life show an impressively high correlation with CHF severity. Interestingly, the correlation is highest with pro-ANP, as a marker of functional CHF severity, and much lower with LVEF. On the other hand, there is no, or even a negative, correlation of CHF severity with anxiety. In terms of the magnitude of their correlations with pro-ANP (or CHF severity in general), vital exhaustion, depression, and overall psychosocial quality of life lie somewhere between self-rated physical status and anxiety: vital exhaustion is closest to the physical impairment of CHF. Depression, as measured by the HADS depression scale, which does not rely on physical indicators of depression, is closer to anxiety, although the associations of anxiety and depression scores with pro-ANP were clearly distinct. This may help to understand why depression, but not generalized anxiety, is independently associated with adverse cardiovascular outcomes in patients with documented heart disease (17) and why some authors (18, 19) even found opposite effects of anxiety and depression scores on prognosis. Despite the known adverse effects of anxiety on, eg, autonomic cardiac control (20), higher anxiety may in part also reflect lower ANP levels, which mean less severe heart failure and a better prognosis.
In addition, some unexpected earlier findings from our group found a possible explanation: patients with coronary artery disease and a history of PTCA (which has been shown to reduce ANP levels during exercise) (21) had higher anxiety than their counterparts without previous PTCA (22). They might also have had lower ANP levels. Furthermore, in that study, 1242 men with documented CHD and recent ventriculograms showed associations between ejection fraction and anxiety that were very similar to the association found between clinical heart failure severity and anxiety in the present study. Anxiety was somewhat higher in patients with slightly reduced ejection fraction compared with those who had normal systolic function. However, the lowest anxiety scores were observed in patients with moderate to severe left ventricular dysfunction (22). Finally, in the same study, the subgroup that also showed left ventricular hypertrophy had surprisingly low anxiety. The impaired ventricular compliance associated with left ventricular hypertrophy typically leads to atrial dilatation and increased ANP secretion (23), which could serve as a possible explanation of the low anxiety scores. However, these possible associations are still speculative.
Some additional caveats must also be kept in mind when interpreting the present results.
This is the first study describing a statistical association of pro-ANP and anxiety in cardiac patients. Further research is needed before our findings can be generalized. It will also be necessary to study the interplay of pro-ANP (and ANP) with other neurohumoral markers such as brain natriuretic peptide, endothelin, catecholamines, etc.
Our study should not be interpreted as demonstrating a causal effect of pro-ANP on anxiety. The cross-sectional evaluation performed in this study suggests a biologically plausible mechanism, although it doesn’t prove it. In addition, pro-ANP is a cleavage product of ANP activation without known cerebral effects and, rather, should be considered an indicator for medium-term plasma levels of (C-terminal) ANP. ANP, however, has a definite potential for exerting an anxiolytic effect. It has been shown to pass the blood-brain barrier and reduce anxiety even after peripheral application (7, 8, 24).
Nevertheless, the use of pro-ANP instead of ANP in the present study was well justified. Pro-ANP is produced during activation of ANP in a 1:1 molar ratio. It has a longer plasma half-life and is therefore more likely to reflect medium-term ANP secretion. It is thereby less likely to be influenced by acute exercise, such as climbing stairs and walking throughout hospital floors. It is also more stable in blood specimens and has therefore been shown to be a better marker of—especially less severe—CHF than ANP itself (25).
Our results seem to be in contrast to those reported by Murberg et al. (9). In their study of 119 patients with stable CHF, depression scores showed no association with plasma pro-ANP, whereas in our study they did. However, the association observed in our study was weak and partly caused by the group difference between patients with and without CHF. The symptoms measured by the different depression scales used in the two studies might show a varying overlap with anxiety (which was not measured by Murberg et al.). A possible positive association between depression and pro-ANP might have been masked by confounding symptoms of anxiety in the Murberg et al. study. It is worth mentioning that Murberg et al. found an insignificant ( = -0.15) negative association of pro-ANP with neuroticism, which can be suspected to show a relevant correlation with anxiety. Interestingly, in our study, the global psychosocial score of the SF36 showed only an insignificant association with pro-ANP. It therefore seems necessary to use instruments that clearly differentiate between symptoms of anxiety and depression, such as the HADS, in order to find the suspected differential ANP effect. With a nonspecific measure of psychological distress, the different associations of anxiety and depression with pro-ANP might otherwise be counterbalanced, resulting in a lack of correlation.
A final point of concern are the high absolute plasma levels observed in the CHF patients. In patients with panic disorder, an infusion of ANP that raised plasma levels four- to five-fold to those of baseline levels seems to produce marked anxiolytic effects (8). In contrast, the patients from the present study, even those with moderate CHF, had mean pro-ANP plasma levels that were 10- to 25-fold above those of the control persons, some of whom also had levels above the normal range. One might expect a stronger anxiolytic effect of these extremely high levels. However, the chronic psychosocial distress caused by the experience of heart failure will be more difficult to antagonize than short anxiogenic stimuli. Second, earlier studies have shown blunted renal and cardiovascular responses to chronically elevated ANP levels in CHF (1). ANP resistance might also occur in the brain and account for the relatively modest effect observed in our sample as well as for the poor sympathoinhibitory effect of ANP observed in another study (5).
Future research will have to confirm or disprove our still preliminary findings. Should the results be confirmed, therapeutically administered neutral endopeptidase inhibitors, such as candoxatril (26) or synthetic natriuretic peptides, which have recently be shown to produce a short-term cardiovascular benefit and a reduction in dyspnea (27), might also be useful in treating anxiety in patients with heart failure. For example, patients with implanted defibrillators and frequent shock delivery who often suffer from severe heart failure as well as severe anxiety might be candidates for this new treatment option. However, it would be premature to recommend these treatments without further study. Our speculations are also not meant to minimize the importance of successful coping and psychosocial interventions in CHF patients but could possibly help obtain a more complete understanding of the processes involved and help develop a broader range of therapies.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J T Parissis, M Nikolaou, D Farmakis, V Bistola, I A Paraskevaidis, S Adamopoulos, G Filippatos, and D T Kremastinos Clinical and prognostic implications of self-rating depression scales and plasma B-type natriuretic peptide in hospitalised patients with chronic heart failure Heart, May 1, 2008; 94(5): 585 - 589. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Lin, J. Chang, V. Zemon, and E. Midlarsky Silent illumination: a study on Chan (Zen) meditation, anxiety, and musical performance quality Psychology of Music, April 1, 2008; 36(2): 139 - 155. [Abstract] [PDF]
|
|
|
|
 |
|
 T. Hillemacher and S. Bleich Neurobiology and Treatment in Alcoholism--Recent Findings Regarding Lesch's Typology of Alcohol Dependence Alcohol Alcohol., March 13, 2008; (2008) agn016v1. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. HILLEMACHER, K. BAYERLEIN, J. WILHELM, D. POLEO, H. FRIELING, M. ZIEGENBEIN, W. SPERLING, J. KORNHUBER, and S. BLEICH VOLUME INTAKE AND CRAVING IN ALCOHOL WITHDRAWAL Alcohol Alcohol., January 1, 2006; 41(1): 61 - 65. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
