Psychosomatic Medicine
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Published online before print February 6, 2008, 10.1097/PSY.0b013e31816477a1
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kojima, M.
Right arrow Articles by Tokudome, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kojima, M.
Right arrow Articles by Tokudome, S.
Related Collections
Right arrow Depression
Right arrow Psychophysiology
Right arrow Sympathetic Nervous System
Right arrow Renal
Right arrow Other Cardiovascular Medicine
Psychosomatic Medicine 70:177-185 (2008)
© 2008 American Psychosomatic Society

ORIGINAL ARTICLES

Loss of Fractal Heart Rate Dynamics in Depressive Hemodialysis Patients

Masayo Kojima, MD, PhD, Junichiro Hayano, MD, PhD, Hidekatsu Fukuta, MD, PhD, Seiichiro Sakata, MD, PhD, Seiji Mukai, MD, PhD, Nobuyuki Ohte, MD, PhD, Hachiro Seno, MD, Takanobu Toriyama, MD, Hirohisa Kawahara, MD, Toshiaki A. Furukawa, MD, PhD and Shinkan Tokudome, MD, MPH, PhD

From the Department of Health Promotion and Preventive Medicine (M.K., S.T.), Nagoya City University Graduate School of Medical Sciences, Nagoya, Japan; Department of Medical Education (J.H.), Nagoya City University Graduate School of Medical Sciences, Nagoya, Japan; Center for Mental Health and Care (J.H., T.A.F.), Nagoya City University Hospital, Nagoya, Japan; Department of Cardio-Renal Medicine and Hypertension (H.F., S.S., S.M., N.O.), Nagoya City University Graduate School of Medical Sciences, Nagoya, Japan; Nagoya Kyoritsu Hospital (H.S., T.T., H.K.), Nagoya, Japan; Department of Psychiatry and Cognitive-Behavioral Medicine (T.A.F.), Nagoya City University Graduate School of Medical Sciences, Nagoya, Japan.

Address correspondence and reprint requests to Masayo Kojima, Department of Health Promotion and Preventive Medicine, Nagoya City University Graduate School of Medical Sciences, 1 Kawasumi, Mizuho-cho, Mizuho-ku, Nagoya 467-8601, Japan. E-mail: masayok{at}med.nagoya-cu.ac.jp


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 NOTES
 REFERENCES
 
Objective: To assess the relationship between depression, reduced heart rate (HR) variability, and altered HR dynamics among patients with end-stage renal disease who are receiving hemodialysis (HD) therapy.

Methods: We analyzed the 24-hour electrocardiograms of 119 outpatients receiving chronic HD. HR variability was quantified with the standard deviation of normal-to-normal R-R intervals, the triangular index, and the powers of the high- (HF), low- (LF), very-low (VLF), and ultra-low frequency (ULF) components. Nonlinear HR dynamics was assessed with the short-term ({alpha}1) and long-term ({alpha}2) scaling exponents of the detrended fluctuation analysis and approximate entropy. The depression level was assessed using the Beck Depression Inventory, Second Edition (BDI-II). HR variability and dynamics measurements were compared by gender, diabetes, and depression with adjustment for age and serum albumin concentration.

Results: Most indices of HR variability and dynamics were negatively correlated with age, serum albumin concentration, depression score, and were lower in women and patients with diabetes. The {alpha}2 was inversely associated with these variables. Depressed men had significantly lower HF, LF, VLF, and marginally lower ULF than nondepressed persons after adjustment for diabetes and other covariates; no difference in depression was observed in women. The {alpha}2 showed marginally significant difference in depression independent from gender and diabetes.

Conclusions: Among the patients who received HD, depression is associated with reduced HR variability and loss of fractal HR dynamics. However, the influence of depression on HR variability may vary by gender and physiological backgrounds. Further prospective studies are necessary to confirm their association with poor prognosis.

Key Words: depression • fractal • heart rate variability • hemodialysis • nonlinear

Abbreviations: CHD = coronary heart disease; HR = heart rate; DFA = detrended fluctuation analysis; ApEn = approximate entropy; HD = hemodialysis; ESRD = end-stage renal disease; NKC = Nagoya Kidney Center; AMI = acute myocardial infarction; ECG = electrocardiography; PCR = protein catabolic rate; SDNN = standard deviation of normal-to-normal R-R intervals; HF = high-frequency band; LF = low-frequency band; VLF = very-low-frequency band; ULF = ultra-low-frequency band; SD = standard deviation; mNN = mean normal-to-normal R-R intervals; BDI = Beck Depression Inventory; DSM = Diagnostic and Statistical Manual of Mental Disorders; ANCOVA = analysis of covariance; GLM = general linear model.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
METHODS
 RESULTS
 DISCUSSION
 NOTES
 REFERENCES
 
Depression is a risk factor for the etiology and prognosis of coronary heart disease (CHD) (1,2). Although the underlying mechanisms are not yet fully understood, an imbalance of the autonomic nervous system in depressed patients is one possible explanation (3,4). Heart rate (HR) variability analysis is commonly used as an index of cardiac autonomic function (5), and reduced HR variability is known to predict mortality in patients with CHD (6,7). In addition, a significant association between depression and low HR variability among this population has been consistently reported (8–12).

Conventionally, HR variability is evaluated using time-domain and frequency-domain analyses. In healthy individuals, normal HR fluctuations are suggested to be neither strictly regular nor completely random, but have a fractal-like structure characterized by self-similarity and scale invariance, that is, the presence of similar dynamics operating over multiple time scales (13). Methods based on a nonlinear system theory have been developed to evaluate qualitative fluctuation characteristics, whereas the traditional time-domain and frequency-domain HR variability analyses assess fluctuation quantity. Detrended fluctuation analysis (DFA) describes fluctuation fractal correlation properties, whereas approximate entropy (ApEn) is an index of overall complexity and time-series predictability. Importantly, nonlinear analyses of HR dynamics have been suggested to provide more powerful prognoses and to detect valuable physiologic and pathophysiologic information that is not achieved by conventional HR variability analysis (5,14,15). Loss of fractal HR dynamics was associated with an increased risk of mortality in patients with (16–19) and without (20) structural heart disease, and in post stroke patients (21).

We previously reported a J-curve relationship between cardiac mortality and the long-term scaling exponent ({alpha}2) derived from DFA in hemodialysis (HD) patients with CHD (22). Both increases and decreases of {alpha}2, indicating loss of fractal HR dynamics, were significantly associated with the risk of cardiac death; these associations were independent of those of clinical variables. Patients with end-stage renal disease (ESRD) who receive HD are thought to be vulnerable to emotional disturbances because they suffer from chronic stress relating to dietary constraints, time restrictions, functional limitations, various illnesses, and adverse effects of medications (23). Moreover, depression has been suggested as a possible independent risk factor of increased mortality in this population (24). These observations indicate that altered HR dynamics might be related to the link between depression and increased mortality among ESRD patients who receive HD. However, to date, only a few studies involving limited numbers of subjects have reported an association between HR dynamics and depression in patients with major depression and in the elderly with cardiac diseases (24–26).

The present study examined whether altered HR dynamics, as assessed by fractal correlations and complexity together with conventional time-domain and frequency-domain HR variability analyses, were associated with depressive symptoms among HD patients. To our knowledge, this is the first study to investigate the association between depression and {alpha}2. Moreover, no previous study has examined the association between depression and nonlinear properties of HR dynamics in a population of >100 subjects.

Age, gender, nutrition index, and diabetes are suggested to influence autonomic function and HR dynamics as well as mental states. To estimate the independent association between depression and HR dynamics and variability, we included those factors which might confound the results in the analysis. To explore the possibility of nonlinear relationships (such as J-curves) between these measures and depression, we compared depression severity among quintiles of HR dynamics and variability.


   METHODS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
RESULTS
 DISCUSSION
 NOTES
 REFERENCES
 
Subjects
The subjects were recruited from the patients screened for the Nagoya Kidney Center (NKC) study between May 2001 and May 2002. The prospective NKC study was designed to explore the influence of psychosocial factors on the long-term prognosis of ESRD patients receiving chronic HD therapy. The protocol was approved by the Research Ethics Committee of Nagoya City University Graduate School of Medical Sciences, Japan. According to the selection criteria, the eligible individuals were ESRD patients receiving regular 4-hour HD therapy three times per week at the Nagoya Central Clinic, which is one of the three clinics involved in the NKC study in Japan. Individuals were also aged <70 years, could read and complete the self-administered questionnaire unaided, had not experienced episodes of acute myocardial infarction (AMI), stroke, or a major surgical procedure within the past 2 months, and had not experienced malignant neoplasm or any psychiatric diagnosis during the past 5 years. Patients were excluded if they had hemodynamically significant valvular or congenital heart disease, atrial fibrillation or flutter, high-grade heart block, or permanent pacemaker implantation. A trained research assistant interviewed the patients and confirmed whether they met the criteria and lacked cognitive impairments. Of the 379 patients registered, 249 met the criteria and were invited to participate in the study and complete the questionnaire. Among these 249 patients, 68 persons declined to participate; these individuals tended to be older and more likely to have physical problems than those who agreed to participate. An additional 14 patients were unable to participate because of their therapy schedules; four patients failed to complete the questionnaire due to their physical condition or for personal reasons. Hence, a total of 163 patients provided written informed consent and completed the questionnaire.

Procedures
At enrollment, all patients underwent 24-hour Holter electrocardiographic analysis (Fukuda Denshi, Tokyo, Japan) and at the same time performed their usual daily activities. The participants were asked to complete a battery of questionnaires ≤1 week after the electrocardiographic examination. Blood chemistry results, a chest roentgenogram, and an echocardiogram obtained ≤1 month before the 24-hour electrocardiography (ECG) were used to assess the baseline clinical parameters. Medical data were obtained from hospital charts, including serum albumin concentration, protein catabolic rate (PCR) and Kt/V (a measure of HD treatment adequacy); these nutritional and dialytic parameters are established markers associated with survival in HD patients (27–30).

Analysis of Holter Electrocardiograms
The Holter electrocardiograms were digitized with 12-bit resolution at 128 Hz, using a scanner (SCM2000, Fukuda Denshi), which detected and labeled all QRS complexes automatically. The results of the automatic analysis were reviewed, and any errors in QRS detection or labeling were edited manually. Based on the edited QRS labeling, patients with frequent ventricular and supraventricular ectopic beats that accounted for >10% of the total recorded beats were excluded. The normal-to-normal R-R interval data obtained from the edited time sequences of the QRS complexes were transferred onto a Microsoft Windows-based personal computer (8187CKJ, IBM, New York).

The computations of HR dynamics and variability measurements were performed by a custom-made Fortran 95 program. Subroutine source codes of the DFA were obtained from the website of Physio Toolkit (available at http://www.physionet.org), which is open-source software for biomedical science and engineering including nonlinear analysis of time series as well as conventional time-domain and frequency-domain HR variability analysis (31).

Traditional Time-Domain and Frequency-Domain HR Variability Analysis
The time-domain and frequency-domain HR variability measurements were analyzed using the methods recommended by the Task Force of the European Society of Cardiology and the North American Society of Pacing Electrophysiology (5).

For time-domain HR variability, the mean normal-to-normal R-R intervals (mNN), standard deviation (SD) of the normal-to-normal R-R intervals (SDNN), and triangular index were calculated. For frequency-domain HR variability, the power spectrum was computed using a fast Fourier transformation and the following frequency bands: <0.0033 Hz (ultra-low frequency [ULF]); 0.0033 to <0.04 Hz (very-low frequency [VLF]); 0.04 to <0.15 Hz (low-frequency [LF]); and 0.15 to 0.4 Hz (high-frequency [HF]). The frequency-domain measurements of HR variability were transformed to natural logarithms because their distributions were skewed.

Analysis of Nonlinear Dynamics
Nonlinear HR dynamics were assessed using DFA and ApEn. The DFA technique was used to quantify the fractal-like correlation properties of the R-R interval time series, and the details of this algorithm have been reported elsewhere (32). The scaling exponent indicates the strength of the correlations with previous values in the time series. Scaling exponent values approximating 0.5 suggest random dynamics (no correlation); values close to 1.5 describe highly correlated behavior; whereas values close to 1.0 are characteristic of fractal-like processes, suggesting that the fluctuations are generated by complex systems with multiple feedback regulations (13). The scaling exponents for HR dynamics were calculated separately for short-term (4–11 beats, {alpha}1) and longer-term (>11 beats, {alpha}2) correlations.

ApEn is a measure that quantifies the unpredictability of fluctuations (16) and reflects the likelihood that similar patterns of observations will not be followed by additional similar observations. Smaller ApEn values imply a greater likelihood of this. If the time series is highly irregular, the occurrence of similar patterns will not be predictive for the following measurements, and ApEn will be larger. We computed ApEn according to the algorithm of Pincus (33) with a fixed set of parameters: m = 2 and r = 20% of the SD of the data (m denotes the length of compared runs; r denotes the tolerance of the filter).

Measure of Depression
The level of depressive symptoms was assessed using the Japanese version of the Beck Depression Inventory, Second Edition (BDI-II) (34). The BDI is a self-report tool that has been commonly used to evaluate depressive symptoms. Unique features of this inventory are its usefulness in screening for depression among the general population and measuring the severity of depression in clinical settings (35–37) and its application in ESRD patients (38,39). The original BDI was revised to correspond to the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) criteria in 1996, and was republished as the BDI-II. The version used in the current study has been translated into Japanese using the back-translation method and has been well validated in the general population (35) as well as in clinical settings (37). The BDI-II scores range from 0 to 63, and the cut-off score of 14 indicates at least a mild-to-moderate level of depression (34).

Statistical Analysis
The data were analyzed using SPSS for Windows (version 12.0). All statistical tests were two-sided. A p ≤ .05 was considered to be statistically significant; p ≤ .10 and >.05 were considered to be marginal.

To investigate the interrelationships among the HR variability measures as well as their linear associations with the BDI-II score and clinical markers, the Pearson's correlation coefficients between the variables were calculated.

The mean values of HR dynamics and variability measures and background characteristics were compared by gender (female/male), presence of diabetes (yes/no), and presence of depressive symptoms (BDI-II ≥14/<14), using the {chi}2 statistic for categorical variables and the Student's t test for continuous variables. Then, three-way analysis of covariance (ANCOVA) was conducted using the general linear model (GLM) (40). Gender, diabetes, and depression were entered as main factors; age and serum albumin concentration were included in the models as covariates. To explore the potential nonlinear associations between HR dynamics and variability measures and depression severity, the scores on the HR dynamics and variability measures were divided in quintiles for each subject. The mean differences in the BDI-II scores by subgroups, and according to the quintiles of HR dynamics and variability variables, were examined using the GLM. Tukey's post hoc tests were conducted successively.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
DISCUSSION
 NOTES
 REFERENCES
 
Of the 163 patients who completed the questionnaires, 19 patients did not undergo 24-hour ECG analysis during the study period, six failed to produce analyzable ECG data, and 15 were excluded because of atrial fibrillation or frequent ectopic beats (>2000 per day). In addition, four patients were excluded because they had been prescribed antidepressants. Thus, data including HR dynamics and variability variables were analyzed for a total of 119 patients. There were no significant differences in age, gender, BDI-II scores, or medical history of diabetes among the subjects at the time of data completion.

Correlations Between Variables
To explore the interrelationships between the HR dynamics and variability measures, Pearson's correlation coefficients were calculated (Table 1). Statistically significant correlations were observed between most of the HR dynamics and variability variables.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Pearson's Correlation Coefficient Between Heart Rate Dynamics and Variability Measures

 

The linear associations of the HR dynamics and variability measures with depression score, age, and clinical indices are shown in Table 2. Age was negatively correlated with most of the HR dynamics and variability measures, but positively correlated with {alpha}2. HD duration failed to show significant correlations with any of the HR variables. All of the HR dynamics measurements as well as LF and VLF were significantly correlated with serum albumin concentration and PCR. The depression score was not associated with age or any of the clinical markers, but was significantly correlated with all of the HR dynamics measures and most of the HR variability indices. A significant positive correlation was detected between the depression score and {alpha}2.


View this table:
[in this window]
[in a new window]

  		TABLE 2. Pearson's Correlation Coefficient Between Age, Medical Characteristics, BDI-II, Heart Rate Dynamics, and Variability Measures

 

Comparisons by Gender
The demographic, clinical, and psychosocial characteristics of the subjects by gender are shown in Table 3. Women had significantly higher Kt/V and total cholesterol levels and were less likely to be current smokers. The variables that showed statistically significant or marginally significant gender differences were all HR dynamics measures: {alpha}1 (1.00 ± 0.30 in women versus 1.15 ± 0.27 in men, p = .004), {alpha}2 (1.19 ± 0.06 in women versus 1.16 ± 0.06 in men, p = .01), ApEn. (0.81 ± 0.26 in women versus 0.90 ± 0.26 in men, p = .09), HF (4.02 ± 0.98 in women versus 4.15 ± 0.98 in men, p = .02), and LF (4.51 ± 1.12 in women versus 5.05 ± 1.32 in men, p = .02). There was no significant gender difference in any of the time-domain measures.


View this table:
[in this window]
[in a new window]

  		TABLE 3. Demographic, Clinical, and Psychosocial Characteristics of the Subjects

 

Comparisons by Diabetes
Diabetic patients were more depressed (total BDI-II = 17.4 ± 11.5 versus 12.1 ± 7.5, p = .03), had spent fewer years undergoing dialysis (3.4 ± 2.2 versus 9.9 ± 6.7 years, p < .001), had lower Kt/V values (1.32 ± 0.22 versus 1.47 ± 0.21, p = .003), were more likely to have undergone revascularization (17.9% versus 2.2%, p = .002), and were more likely to have been prescribed calcium blockers (75.0% versus 53.8%, p = .047) than nondiabetics. Most of the HR variables showed statistically significant differences according to the presence of diabetes, except for the mNN. The {alpha}2 was significantly higher in diabetics, whereas the other variables were significantly lower in diabetics than in nondiabetics.

Comparisons by Depression
The depressed patients were marginally less likely to be married (67.3% versus 82.9%, p = .05). There were no other statistically significant differences in age, gender, smoking habits, or clinical characteristics associated with the presence or absence of depressive symptoms. Among HR dynamics and variability measures, only LF showed a significant difference by depressive status. Depressed patients had a lower LF (4.5 ± 1.4) than nondepressed patients (5.0 ± 1.1, p = .03). Although there were no statistically significant differences in the other variables, they all tended to be lower in the depressed group than the nondepressed group, with the exception of the {alpha}2. The {alpha}2 values were slightly higher in the depressed than the nondepressed group.

Comparisons by Gender x Diabetes x Depression
To examine the independent association between HR dynamics and variability measures and depression, three-way (gender x diabetes x depression) ANCOVAs with adjustments for age and serum albumin concentrations were conducted. Table 4 presents the adjusted means, standard errors, and p values. As a significant gender x diabetes interaction was observed in SDNN, and gender x depression interactions were confirmed in frequency-domain indices, these variables were analyzed separately by gender using two-way ANCOVA (depression x diabetes) (Table 5). No significant three-way interaction was observed for any measure.


View this table:
[in this window]
[in a new window]

  		TABLE 4. Comparisons of Heart Rate Dynamics and Variability Measures by Depression, Diabetes and Gender With Adjustment for Age and Serum Albumin Concentration Among Patients With ESRD

 

View this table:
[in this window]
[in a new window]

  		TABLE 5. Comparisons of SDNN and Frequency-Domain Heart Rate Variability Measures by Depression and Diabetes With Adjustment for Age and Serum Albumin Concentration Among ESRD Patients by Gender

 

HR dynamics measures showed independent susceptibility to diabetes and gender, respectively; {alpha}1 and ApEn were significantly or marginally significantly higher in diabetics and in men, whereas {alpha}2 was significantly lower in these subject groups. Regardless of gender and diabetes, {alpha}2 showed marginally significant differences in depression, such that depressed patients had a higher {alpha}2 than nondepressed patients.

In men, all four frequency-domain indices were significantly or marginally significantly lower in depressed patients than in nondepressed patients, even after adjusting for diabetes and other covariates (Table 5). By contrast, no significant difference was observed in any of the HR variability factors by depression in women.

Depression by Quintiles of HR Dynamics and Variability Measures
To explore the nonlinear associations between depression severity and the levels of HR dynamics and variability measurement values, we examined the mean BDI-II scores for the quintiles of the HR dynamics and variability measures using the GLM (Figure 1). With the exception of {alpha}2, the members of the lowest quintiles of the HR dynamics and variability measures showed the highest average depression scores compared with other subjects. In {alpha}2, the highest quintile showed the highest depression scores, and those in the middle quintile showed the lowest depression scores.


Figure 17
View larger version (23K):
[in this window]
[in a new window]

  		Figure 1. Mean BDI-II scores with standard errors for the quintiles of the nonlinear and traditional HR variability measures. ANOVA tests showed significant overall differences in {alpha}2, ApEn, SDNN, HF, and LF, and a marginal difference in ULF. Statistically significant differences between quintiles confirmed by Tukey's post hoc test are indicated as *p < .05, **p < .01, and ***p < .001. BDI-II = Beck Depression Inventory, Second Edition; HR = heart rate; ANOVA = analysis of variance; {alpha}2 = the long-term scaling exponent; ApEn = approximate entropy; SDNN = standard deviation of normal-to-normal R-R intervals HF = high-frequency band; LF = low-frequency band; ULF = ultra-low-frequency band.

 


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
NOTES
 REFERENCES
 
This study confirmed that the fractal and complexity properties of HR dynamics, as well as conventional time-domain and frequency-domain measures of HR variability, tend to be associated with depression severity among ESRD patients.

Low HR variability measured by conventional methods has been associated with depression in many studies of patients with CHD (4). Reduced HR variability, as measured by conventional methods, is thought to mainly reflect excessive sympathetic and/or inadequate parasympathetic tones. However, cardiac regulations are determined by complex interactions of hemodynamic, electrophysiological, and humoral variables as well as by autonomic and central nervous regulation (5). HR dynamics indices are expected to reflect the capacity of physiologic systems to respond to unpredictable stress and stimuli, and to describe the total individual soundness of physiological regulatory systems. The present results suggest that depressed patients are not merely in a state where autonomic nervous systems are disrupted, but rather they experience a broader dysregulation of physiological systems. Further evidence, including the analysis of various background correlates and experimental studies, will be necessary to disclose the mechanisms of the associations between depression and HR dynamics and variability.

The association between {alpha}2 and depression has not previously been examined. Earlier we reported adverse effects of both high and low values of {alpha}2 on cardiac death rates among 81 HD patients with coronary artery disease (22). In our present study, we confirmed a J-curve relationship between {alpha}2 and depression severity. Theoretically, reduced exponent values indicate random dynamics (that is, no correlation), whereas increased values describe highly correlated HR behaviors (32). Therefore, it seems reasonable that the favorable level of {alpha}2 in relationship to depression lies in the middle range. Although it is not possible to further discuss the extent to which depression explains the associations between {alpha}2 and cardiac deaths from the available data, ongoing prospective studies will disclose the linkage or independent influences of depression and HR dynamics on the prognosis of this population.

In relationship to depression severity, similar trends were observed in {alpha}1 and ApEn in the present subjects; the lowest depression scores were not apparent in the lowest or highest quintiles, but rather in the third or the fourth. Only one previous study has reported an association between depression and {alpha}1 and ApEn (26). This study found that in patients >60 years of age with recent unstable angina pectoris or AMI (n = 52, women = 52%), {alpha}1 was positively correlated (r = .31, p = .02) and ApEn was negatively correlated (r = –.28, p = .046) with the total score of the Hamilton Depression Scale (41). HR dynamics is influenced by aging, gender, and various physiological factors (42). Therefore, the inconsistency between these previous results and our own might be partly explained by differences in background characteristics. It is also possible that simple dichotomization of the variables might offset the effects of high and low values and yield misleading results. The possible nonlinear associations with health indices and HR dynamics should be examined carefully to establish favorable ranges for positive health outcomes.

Aging, gender, and physical disorders such as severe heart failure have previously been suggested to influence HR dynamics (42), although their interactions have not been investigated. We confirmed that the influences of gender and diabetes on HR dynamics were independent of each other as well as depression. Frequency-domain indices revealed significant gender interactions with depression or/and with diabetes. In men, frequency-domain indices were altered by depression, but only the LF was altered by diabetes. Low LF was independently associated with depression and diabetes in men. By contrast, depression failed to show a significant association with any frequency-domain indices in women, whereas clear differences were associated with diabetes. From the present results, however, it is not possible to conclude that men are more sensitive to depression than women in relationship to frequency-domain HR variability. According to the evidence from a cohort of 500,868 diabetic patients in Taiwan, female patients were suggested to be more vulnerable to autonomic neuropathy than male patients (43). In addition, the influence of the menstrual cycle on HR variability indices has been suggested (44,45). One previous study reported gender influences on fractal dynamics and the complexity of HR dynamics including {alpha}2 (46). Significant gender differences were not found in {alpha}2, whereas values of {alpha}1 and conventional HR variability measures were significantly higher and ApEn was lower in men among the healthy population (46). The possible interactions between gender, age, and physiological factors need to be considered to clarify the properties of HR dynamics and variability measurements.

Our current study had some limitations. First, our subjects were ESRD patients aged <70 years, who received HD therapy without severe physical problems and who could complete the questionnaire unaided. So, our study group was healthier than the general HD patient population. Moreover, we excluded those who had received a psychiatric diagnosis to avoid the influence of psychotropic medications (47–49). Severely depressed patients were not included, which might explain why we failed to observe significant differences in {alpha}1 and ApEn in the presence of depressive symptoms.

Second, we used self-report measures to evaluate depression. Depressed patients can have a distorted cognitive style and perceive themselves negatively (50). This would overestimate depressive symptoms, thus attenuating the association between depression and HR variables. Depressed patients often complain of physical symptoms as well as emotional disturbances, and it is especially difficult for patients suffering from chronic diseases to distinguish them. The depression score might therefore only reflect a patient's somatic disease severity, which invites a loss of HR dynamics and reduced HR variability. We compared the correlation coefficient between HR dynamics and variability measures and the two subscales of BDI-II, namely, the somatic-affective and the cognitive subscales (34). However, no differences were apparent (data not shown).

Third, depression as defined in the present study was not equivalent to major depression as defined by the DSM criteria. Thus, although the BDI has been well validated and used to measure depression severity in a number of studies of ESRD patients, our findings should be verified using different measurement tools.

In conclusion, loss of fractal HR dynamics and reduced HR variability measures are associated with depression in ESRD patients undergoing HD therapy. To confirm an association between depression and altered HR dynamics or their independent influences on prognosis, further studies with a prospective design will be necessary. Our findings should also be validated by examining other populations with different backgrounds.

We thank K. Ibuki for his contribution to the ECG data processing; H. Takahashi and K. Watanabe for coordinating the study; and K. Asada, H. Tomizawa, A. Nakata, K. Kobayashi, C. Asakura, E. Inoue, C. Yamauchi, E. Yamashiro, and the staff of the Kaikoukai Central Clinics at Nagoya, Anjoh, and Hekikai, Japan, for collecting data. We also wish to thank all of the participants in the study.


   NOTES
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 NOTES
REFERENCES
 
Received for publication January 11, 2007; revision received September 26, 2007.

This study was supported by a Grant-in-Aid B15790301 for the Encouragement of Young Scientists from the Japan Society for the Promotion of Science.

DOI:10.1097/PSY.0b013e31816477a1


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 NOTES
 REFERENCES
 

  1. Hemingway H, Marmot M. Evidence based cardiology: psychosocial factors in the aetiology and prognosis of coronary heart disease. Systematic review of prospective cohort studies. BMJ 1999;318:1460–7.[Free Full Text]
  2. Frasure-Smith N, Lesperance F, Talajic M. Depression and 18-month prognosis after myocardial infarction. Circulation 1995;91:999–1005.[Abstract/Free Full Text]
  3. Glassman AH, Shapiro PA. Depression and the course of coronary artery disease. Am J Psychiatry 1998;155:4–11.[Abstract/Free Full Text]
  4. Carney RM, Freedland KE, Veith RC. Depression, the autonomic nervous system, and coronary heart disease. Psychosom Med 2005;67(Suppl 1):S29–S33.[Abstract/Free Full Text]
  5. 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;93:1043–65.[Free Full Text]
  6. La Rovere MT, Bigger JT Jr, Marcus FI, Mortara A, Schwartz PJ. Baroreflex sensitivity and heart-rate variability in prediction of total cardiac mortality after myocardial infarction. ATRAMI (Autonomic Tone and Reflexes After Myocardial Infarction) Investigators. Lancet 1998;351:478–84.[CrossRef][Medline]
  7. Kleiger RE, Miller JP, Bigger JT Jr, Moss AJ. Decreased heart rate variability and its association with increased mortality after acute myocardial infarction. Am J Cardiol 1987;59:256–62.[CrossRef][Medline]
  8. Rechlin T, Weis M, Spitzer A, Kaschka WP. Are affective disorders associated with alterations of heart rate variability? J Affect Disord 1994;32:271–5.[CrossRef][Medline]
  9. Carney RM, Saunders RD, Freedland KE, Stein P, Rich MW, Jaffe AS. Association of depression with reduced heart rate variability in coronary artery disease. Am J Cardiol 1995;76:562–4.[CrossRef][Medline]
  10. Krittayaphong R, Cascio WE, Light KC, Sheffield D, Golden RN, Finkel JB, Glekas G, Koch GG, Sheps DS. Heart rate variability in patients with coronary artery disease: differences in patients with higher and lower depression scores. Psychosom Med 1997;59:231–5.[Abstract/Free Full Text]
  11. Stein PK, Carney RM, Freedland KE, Skala JA, Jaffe AS, Kleiger RE, Rottman JN. Severe depression is associated with markedly reduced heart rate variability in patients with stable coronary heart disease. J Psychosom Res 2000;48:493–500.[CrossRef][Medline]
  12. Carney RM, Blumenthal JA, Stein PK, Watkins L, Catellier D, Berkman LF, Czajkowski SM, O'Connor C, Stone PH, Freedland KE. Depression, heart rate variability, and acute myocardial infarction. Circulation 2001;104:2024–8.[Abstract/Free Full Text]
  13. Iyengar N, Peng CK, Morin R, Goldberger AL, Lipsitz LA. Age-related alterations in the fractal scaling of cardiac interbeat interval dynamics. Am J Physiol 1996;271:R1078–84.[Medline]
  14. Perkiomaki JS, Zareba W, Ruta J, Dubner S, Madoery C, Deedwania P, Karcz M, Bayes de Luna. Fractal and complexity measures of heart rate dynamics after acute myocardial infarction. Am J Cardiol 2001;88:777–81.
  15. Makikallio TH, Huikuri HV, Hintze U, Videbaek J, Mitrani RD, Castellanos A, Myerburg RJ, Moller M. Fractal analysis and time- and frequency- domain measures of heart rate variability as predictors of mortality in patients with heart failure. Am J Cardiol 2001;87:178–82.[CrossRef][Medline]
  16. Ho KK, Moody GB, Peng CK, Mietus JE, Larson MG, Levy D, Goldberger AL. Predicting survival in heart failure case and control subjects by use of fully automated methods for deriving nonlinear and conventional indices of heart rate dynamics. Circulation 1997;96:842–8.[Abstract/Free Full Text]
  17. Makikallio TH, Hoiber S, Kober L, Torp-Pedersen C, Peng CK, Goldberger AL, Huikuri HV. Fractal analysis of heart rate dynamics as a predictor of mortality in patients with depressed left ventricular function after acute myocardial infarction. TRACE Investigators. TRAndolapril Cardiac Evaluation. Am J Cardiol 1999;83:836–9.[CrossRef][Medline]
  18. Schmidt G, Malik M, Barthel P, Schneider R, Ulm K, Rolnitzky L, Camm AJ, Bigger JT Jr, Schomig A. Heart-rate turbulence after ventricular premature beats as a predictor of mortality after acute myocardial infarction. Lancet 1999;353:1390–6.[CrossRef][Medline]
  19. Huikuri HV, Makikallio TH, Peng CK, Goldberger AL, Hintze U, Moller M. Fractal correlation properties of R-R interval dynamics and mortality in patients with depressed left ventricular function after an acute myocardial infarction. Circulation 2000;101:47–53.[Abstract/Free Full Text]
  20. Tapanainen JM, Thomsen PE, Kober L, Torp-Pedersen C, Makikallio TH, Still AM, Lindgren KS, Huikuri HV. Fractal analysis of heart rate variability and mortality after an acute myocardial infarction. Am J Cardiol 2002;90:347–52.[CrossRef][Medline]
  21. Makikallio AM, Makikallio TH, Korpelainen JT, Sotaniemi KA, Huikuri HV, Myllyla VV. Heart rate dynamics predict poststroke mortality. Neurology 2004;62:1822–6.[Abstract/Free Full Text]
  22. Fukuta H, Hayano J, Ishihara S, Sakata S, Ohte N, Takahashi H, Yokoya M, Toriyama T, Kawahara H, Yajima K, Kobayashi K, Kimura G. Prognostic value of nonlinear heart rate dynamics in hemodialysis patients with coronary artery disease. Kidney Int 2003;64:641–8.[CrossRef][Medline]
  23. Kimmel PL. Depression in patients with chronic renal disease: what we know and what we need to know. J Psychosom Res 2002;53:951–6.[CrossRef][Medline]
  24. Yeragani VK, Rao KA, Smitha MR, Pohl RB, Balon R, Srinivasan K. Diminished chaos of heart rate time series in patients with major depression. Biol Psychiatry 2002;51:733–44.[CrossRef][Medline]
  25. Nahshoni E, Aravot D, Aizenberg D, Sigler M, Zalsman G, Strasberg B, Imbar S, Adler E, Weizman A. Heart rate variability in patients with major depression. Psychosomatics 2004;45:129–34.[Abstract/Free Full Text]
  26. Vigo DE, Nicola Siri L, Ladrón De Guevara MS, Martínez-Martínez JA, Fahrer RD, Cardinali DP, Masoli O, Guinjoan SM. Relation of depression to heart rate nonlinear dynamics in patients > or = 60 years of age with recent unstable angina pectoris or acute myocardial infarction. Am J Cardiol 2004;93:756–60.[CrossRef][Medline]
  27. Hakim RM. Levin N. Malnutrition in hemodialysis patients. Am J Kidney Dis 1993;21:125–37.[Medline]
  28. Owen WF Jr, Lew NL, Liu Y, Lowrie EG, Lazarus JM. The urea reduction ratio and serum albumin concentration as predictors of mortality in patients undergoing hemodialysis. N Engl J Med 1993;329:1001–6.[Abstract/Free Full Text]
  29. Held PJ, Port FK, Wolfe RA, Stannard DC, Carroll CE, Daugirdas JT, Bloembergen WE, Greer JW, Hakim RM. The dose of hemodialysis and patient mortality. Kidney Int 1996;50:550–6.[Medline]
  30. Lowrie EG, Lew NL. Death risk in hemodialysis patients: the predictive value of commonly measured variables and an evaluation of death rate differences between facilities. Am J Kidney Dis 1990;15:458–82.[Medline]
  31. Goldberger AL, Amaral LA, Glass L, Hausdorff JM, Ivanov PC, Mark RG, Mietus JE, Moody GB, Peng CK, Stanley HE. PhysioBank, PhysioToolkit, and PhysioNet: components of a new research resource for complex physiologic signals. Circulation 2000;101:E215–E220.[Medline]
  32. Peng CK, Havlin S, Stanley HE, Goldberger AL. Quantification of scaling exponents and crossover phenomena in nonstationary heartbeat time series. Chaos 1995;5:82–7.[CrossRef][Medline]
  33. Pincus SM, Goldberger AL. Physiological time-series analysis: what does regularity quantify? Am J Physiol 1994;266:H1643–H1656.[Medline]
  34. Beck AT, Steer RA. Manual for the Beck Depression Inventory—II. San Antonio, TX: Psychological Corporation; 1996.
  35. Kojima M, Furukawa TA, Takahashi H, Kawai M, Nagaya T, Tokudome S. Cross-cultural validation of the Beck depression inventory—II in Japan. Psychiatry Res 2002;110:291–9.[CrossRef][Medline]
  36. Kojima M, Senda Y, Nagaya T, Tokudome S, Furukawa TA. Alexithymia, depression and social support among Japanese workers. Psychother Psychosom 2003;72:307–14.[CrossRef][Medline]
  37. Hiroe T, Kojima M, Yamamoto I, Nojima S, Kinoshita Y, Hashimoto N, Watanabe N, Maeda T, Furukawa TA. Gradations of clinical severity and sensitivity to change assessed with the Beck depression inventory—II in Japanese patients with depression. Psychiatry Res 2005;135:229–35.[CrossRef][Medline]
  38. Kimmel PL, Peterson RA, Weihs KL, Simmens SJ, Alleyne S, Cruz I, Veis JH. Multiple measurements of depression predict mortality in a longitudinal study of chronic hemodialysis outpatients. Kidney Int 2000;57:2093–8.[CrossRef][Medline]
  39. Watnick S, Wang PL, Demadura T, Ganzini L. Validation of 2 depression screening tools in dialysis patients. Am J Kidney Dis 2005;46:919–24.[CrossRef][Medline]
  40. SPSS Advanced Models 12.0. Chicago, IL: SPSS Inc; 2006.
  41. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychol 1960;23:56–62.[Free Full Text]
  42. Goldberger AL, Amaral LA, Hausdorff JM, Ivanov PC, Peng CK, Stanley HE. Fractal dynamics in physiology: alterations with disease and aging. Proc Natl Acad Sci U S A 2002;99(Suppl 1):2466–72.[Abstract/Free Full Text]
  43. Chen HF, Ho CA, Li CY. Age and sex may significantly interact with diabetes on the risks of lower-extremity amputation and peripheral revascularization procedures: evidence from a cohort of a half-million diabetic patients. Diabetes Care 2006;29:2409–14.[Abstract/Free Full Text]
  44. Guasti L, Grimoldi P, Mainardi LT, Petrozzino MR, Piantanida E, Garganico D, Diolisi A, Zanotta D, Bertolini A, Ageno W, Grandi AM, Cerutti S, Venco A. Autonomic function and baroreflex sensitivity during a normal ovulatory cycle in humans. Acta Cardiol 1999;54:209–13.[Medline]
  45. Minson CT, Halliwill JR, Young TM, Joyner MJ. Sympathetic activity and baroreflex sensitivity in young women taking oral contraceptives. Circulation 2000;102:1473–6.[Abstract/Free Full Text]
  46. Beckers F, Verheyden B, Aubert AE. Aging and nonlinear heart rate control in a healthy population. Am J Physiol Heart Circ Physiol 2006;290:H2560–H2570.[Abstract/Free Full Text]
  47. Yeragani VK, Meiri P, Balon R, Pohl R, Golec S. Effect of imipramine treatment on changes in heart rate and blood pressure during postural and isometric handgrip tests. Eur J Clin Pharmacol 1990;38:139–44.[CrossRef][Medline]
  48. Yeragani VK, Pohl R, Balon R, Ramesh C, Glitz D, Jung I, Sherwood P. Heart rate variability in patients with major depression. Psychiatry Res 1991;37:35–46.[CrossRef][Medline]
  49. Rechlin T, Weis M, Claus D. Heart rate variability in depressed patients and differential effects of paroxetine and amitriptyline on cardiovascular autonomic functions. Pharmacopsychiatry 1994;27:124–8.[Medline]
  50. Beck AT. Depression: Clinical, Experimental, and Theoretical Aspects. New York: Harper & Row; 1967.




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kojima, M.
Right arrow Articles by Tokudome, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kojima, M.
Right arrow Articles by Tokudome, S.
Related Collections
Right arrow Depression
Right arrow Psychophysiology
Right arrow Sympathetic Nervous System
Right arrow Renal
Right arrow Other Cardiovascular Medicine

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS