Eur Psychiatry (1995) 0 Elsevier, Paris
189
10, 189-194
Original
article
Decreased heart rate variability parameters in amitriptyline treated depressed patients: biological and clinical significance T Rechlin, Departments
of Psychiatry
D Claus, M Weis, WP Kaschka
and Neurology, University of Erlangen-Niimberg, Psychiatrische Schwabachanlage 6, 91504 Erlangen, Germany (Received
15 November
1993;
accepted
5 May
Universitiitsklinik,
1994)
Summary - One hundred-four depressed patients treated with amittiptyline (mean dosage: 163 mg/d; mean plasma level: 239 rig/ml) and 52 normal control subjects matched for age and sex underwent a standardized cardiovascular test battery (various autonomic cardiac parameters, which are largely independent from heart rate, namely the coefficients of variation (CV) while resting and during deep respiration, a spectral analysis of heart rate, the Valsalva ratio, and a posture index were determined). The tests included the determination of time- and frequency-derived measurements of heart rate variability (HRV), which is rather independent from heart rate. As compared to the controls the patients showed a significant plasma concentration-dependent decrease of R-R variation in the electrocardiogram @ < O.OOOl), while their heart rate was significantly elevated (p < 0.0001). The markedly reduced parameters of sinus arrhythmia in amitriptyline treated patients are suggested to be mainly due to the anticholinergic effect of this drug, although it can not be excluded that the affective disorder itself might be associated with low heart rate variability. The results indicate that autonomic heart rate parameters are a valuable tool for the detection of tricyclic antidepressant (TCA) intake in unconscious patients, especially in intensive care and emergency wards. heart
rate analysis
/ heart
rate variability
I amitriptyline
I cardiovascular
INTRODUCTION Autonomic cardiac dysfunction has been investigated in a broad variety of diseases, particularly in diabetes, alcoholism and cardiac diseases (Low, 1993). Meanwhile, a number of noninvasive cardiovascular reflex tests have been evaluated either based on the determination of heart rate variability (HRV) or based on the responses of blood pressure to provocation methods (eg change of posture). The regulation of HRV is carried out by the autonomic nervous system varying the intervals between consecutive beats according to physiological requirements. It is known that HRV, which is only poorly associated with heart rate, mainly results from the action of the parasympathetic vagus nerve, using acetylcholine as a neurotransmitter (Bannister and Mathias, 1988). Considering psychiatric diseases, it is still an unsettled issue whether affective disorders are associated with a decrement of HRV parameters or not. Although
autonomic
neumpathy
Jakobsen et al (1984) showed that treatment with tricyclic antidepressants (TCA) is associated with decreased HRV, little is known about the biological and clinical significance of this finding. To our knowledge, Low and Opfer-Gehrking (1992) are the only authors who have investigated the effects af amitriptyline on autonomic functions (75 mg given on two consecutive days to six subjects, respectively, resulting in TCA plasma levels below 85 rig/ml). This dosage significantly influenced the sudomotor (M3) acetylcholine receptor, but HRV during deep respiration, representing M2 receptor function, remained unaffected. This investigation led us to examine autonomic heart rate parameters in all inpatients treated with amitriptyline in a therapeutic dosage. METHODS Standardized out over one
investigations year in 104
of heart rate amitriptyline-treated
were
carried patients
190
T Rechlin
BPM
Deep Respiration
120
I
100~ 80 60 I 0
I
BPM
min
Restlng
120. IOO-
Pi
80
.
60
.
,N
u
9) !2
0
1
0
2
20 40
3
0
20 40
4
min
Hzxld2
1. Representative results of a 62-year-old patient before ther(a) and after 14 days (b) of treatment with 175 mg of amitip tyline per day (TCA plasma concentration: 210 rig/ml): heart rate Fig apy
increased from 72.3 to 94.2 Bpm, CVrest decreased from 3.15% to 1.05%, the CVdr from 5.83% to 2.55%. LF band decreased from 1.07 to 0.33, MF band from 0.09 to 0.02, and the HF baud
from 0.20 to 0.03 (lv
x Hz*).
(69 female, 35 male; mean age: 46.3 years, SD: 13.3 years) with episodes of major depression (DSMIII-R) and in 52 healthy subjects (35 female, 17 male; mean age: 45.0, SD: 12.8 years). The patients had been treated
et
al
with amitriptyline for at least 7 days (mean dosage: 163.2 mg/d, minimum: 75 mg/d, maximum: 275 mg/d, SD: 41.7 mg/d). The TCA serum concentrations (amitriptyline and nortriptyline) were determined by a commercially available fluorescent polarisation immunoassay (FPIA; TDx, Abbott) (mean TCA concentration: 239.4 ng/mI; minimum: 98 rig/ml, maximum: 606 @ml, SD: 105.8 @ml). Patients treated with other TCA or neuroleptic drugs besides amitriptyline were excluded from the study, as were patients with diabetes mellitus, alcoholism, neurological and cardiac diseases. Heart rate analysis was performed in a supine position with the computer program ProSciCard (medical research & diagnostic computer system GmbH, 35440 Linden, Germany). The examination consisted of a 5 minute recording of heart rate while resting (fig l), a recording of 120 heart beats during deep respiration with six breathing cycles per minute (fig I), a Valsalva maneuver, and a posture test. The tests employed have been described and discussed in detail elsewhere (Ziegler et al, 1992a). After a resting period of 10 minutes, the coefficient of variation (CV) (the standard deviation of the distribution of R-R intervals divided by their mean) and the root mean square of successive differences (RMSSD) while resting were calculated from 150 artifact free heart beats. The spectral analyses were carried out by Fast Fourier Transformation. Three frequency peaks were automatically separated (low frequency (LF): 0.01-0.05 Hz; mid frequency (MF): 0.05-0.15 Hz: high frequency (HF): 0.15-0.50 Hz). During the deep respiration (dr) period the CV was determined, as well as the RMSSD, and the mean circulant resultant (MCR) (Weinberg and Pfeifer, 1984), which was obtained by vector analysis technique. In the breathing cycle with the maximal R-R variation, the longest R-R interval (R-R max) during expiration, and the shortest R-R interval (R-R min) during inspiration were determined to obtain the difference R-R max and R-R min (E-I difference), and the ratio RRmax/R-Rmin (E/I ratio). To determine the Valsalva ratio, the sitting patient was asked to blow into a mouthpiece connected to a manometer and to maintain a pressure. of 40 mm Hg for 15 seconds. The R-R intervals were recorded during the maneuver and for 15 seconds following release. The ratio of the longest R-R interval following release to the shortest R-R interval during the maneuver was defined as the Valsalva ratio. To calculate the posture index the patient rested in supine position while the electrocardiogram was recorded. The subject was then asked to stand up quickly. After standing up the ratio of the longest RR interval to the shortest R-R interval was automatically determined. As suggested by Ziegler et al (1992a), the 2.3 percentiles have been chosen as the limits of age related normal values using the ProSciCard program.
Heart
191
rate variability
It is important to know that the results of all tests significantly (p < 0.001) decline with increasing age in normal subjects (CVrest: r = - 0.50, CVdrz r = -0.33; LF: = - 0.30: MF: r = - 0.51; I-IF: r = - 0.59; Valsalva ratio: r = - 0.16, p c 0.05; posture index: r = - 0.48) and show log normal distributions (Ziegler et al, 1992a). Male and female healthy subjects did not differ significantly with regard to the mean results of any of the tests applied (Ziegler et al, 1992a). Most of the tests are significantly correlated with each other; especially the tests during deep respiration (r = 0.82 - 0.96), and during resting (r = 0.78 - 0.87), except the low frequency band of spectral analysis (Ziegler et al, 1992a). The Valsalva ratio was not significantly associated with the results of the tests performed during deep breathing. Correlations of the Valsalva ratio with the other tests were either weak or lacking (r = 0.01 - 0.25). The HRV parameters of normal subjects were largely independent of the heart rate (CVrest: r = - 0.23, p < 0.01; CVdrz r = - 0.07, ns; LF: r = 0.16, ns; MF: r = - 0.01, ns; HF: r = - 0.08, ns; Valsava ratio: r = 0.06, ns; posture index: r = - 0.21, p c 0.05) (Ziegler et al, 1992a). Cardiovascular autonomic neuropathy was diagnosed when at least three of six heart rate parameters (CVrest, posture index, Valsalva ratio, MCR, LF and MF peak of spectral analysis) showed abnormal results (Ziegler et al, 1992b). Statistical analysis of the data included Wilcoxon’s rank sum test and Pearson’s correlation coefficients and was carried out using SPSS.
Table I. Results of heart rate parameters treated jects.
patients
(mean plasma
Normal subjects (n = 52)
Heart rate CVr
RMSSdr CVdr RMSSDdr En-ratio FYI - difference MCR LF JHF HF Vakalva ratio Posture index
74.1 BpM 4.39% 26.89 ms 7.67% 37.57 ms 1.34 238 ms 0045 0.95 x lo-4 Hz2 0.89 x lO-’ Hz2 0.54 x lo” Hz2 1.60 1.25
level 239 r&ml)
in amitriptyline and normal sub-
Abnormal results in Amitriptyline treated patients amitriptyline (n = 104) treated patients W) 93.2 BpM > 100 BpM: 1.82% 82 7.41 ms 77 4.81% 37 14.49 Ins 45 1.20 38 122ms 45 0.020 38 0.55 x lcrc Hz2 31 0.15 x lWHz2 84 O.l2xl@Hz* 60 1.35 30 64 1.09
20
HR: heart rate in beats per minute; CVr: coefficient of variation while resting; RMSSDr: root mean square of successive differences while resting; CVdrz coefficient of variation during deep respiration; RMSSDdr: root mean square of successive differences during deep respiration; E/I: ratio of the longest R-R interval during expiration over the shortest R-R interval during inspiration; E/I: difference of the longest R-R interval during expiration and the shortest R-R interval during inspiration; MCR: mean circular resultant: LF: low freauencv band of soectral analysis (0.01-0.05 Hz); klF: mid freqienc; band of sbectral analysis (0.05-0.15 Hz); HF: high frequency band of spectral analysis (0.15-0.50 Hz).
RESULTS Included in the study were 104 amitriptyline treated patients (69 female, 35 male; mean age: 46.3 years; mean dosage: 163.2 mg/d; mean TCA plasma level: 239.4 r&ml) with episodesof major depression. As compared to 52 control subjects matched for age and sex, the patients had significantly decreased HRV-parameters (p c 0.0001) in all tests and a significantly increased (p < 0.0001) heart rate. Altogether, 67.3% of the patients (n = 70) fulfilled the criteria of a cardiovascular autonomic neuropathy. The results are presented in table I and figures 2 and 3. The correlation between CVrest and TCA plasma level was 0.50, and between CVdr and TCA plasma level - 0.54 (p < 0.0001). In the amitriptyline treated patients the HRV-parameters while resting were significantly correlated with each other (eg CVrest and RMSSrest: r = 0.80, p < 0.0001). The samewas true of the parametersduring deep respiration (eg CVdr and RMSSDdr: r = 0.86). The indices of HRV markedly decreasedwith age in the amitriptyline treated patients (CVrest: r = - 0.48; CVdr: r = - 0.53; LF:
r = - 0.39; MF: r = - 0.36; HF: r = - 0.36; p c 0.001). As in normal controls, the CV during deep respiration and the results of the spectral analyses were independent of the heart rate (CVdr: r = - 0.11, ns; LF: r = 0.04; ns; MF: r = -0.01, ns; HF: r = - 0.19; ns), while the CV at rest correlated sigticantly with the heart rate (r = - 0.34, p c 0.001).
DISCUSSION Autonomic dysfunction has been observed in affective disorders for a long time, particularly in melancholic depression. Dalack and Roose (1990), who used a 24 hour ECG recording system, reported diminished beat-to-beat R-R variation in depressedpatients and suggested this to be a hint at decreased parasympathetic activity. However, Yeragani et al (1992a) and Rechlin et al (1994a) using short observation periods (150 heart beats) failed to find significant differences between the HRV of unmedicated patients with major depression and mentally healthy subjects. Therefore, it is an unsettled issue whether major depression is
192
T Rechlin
et al
Sp@ctnl
Vabaln
0.5 1
ntb
.
In ptlmb
tmatecl with amltriptylln
pow
h thI YF band In patbnb amMptytlm
tr8at.d
with
SpactmlpowarInttl.HFbmdInpatbnbtreabdwith amltrlptylha
.
Fig 2. As the results of all indices of HRV show log normal distribution in normal subjects, the log CV while resting of the patients (upper diagram) and the normal age related range (+ 2SD/- 2SD) are demonstrated. It can clearly been seen that 82% of the patients (n = 85) reached abnormal values. CVs during deep respiration (mid diagram) and the Valsalva ratio are also presented (lower diagrams).
associated with decreased parasympathetic activity or not. Investigations of depressed patients treated with various TCA (Jakobsen et al, 1994; Kristensensen et al, 1989; Yeragani et al, 1992b; Rechlin et al, 1994a) and investigations in schizophrenic patients treated with clozapine (Zahn and Pickar, 1993; Rechlin et al, 1994b) showed reversibly decreased R-R variation, which has been suggested to be caused, at least in part, by the anticholinergic effects of these drugs. Interestingly, amitriptylinetreated patients with major depression showed lower
Fig 3. Extent
of abnormal
results in the spectral
analysis.
HRV parameters than equally treated patients with neurotic depression (Rechlin, 1994). The clinical and biological significance of all these findings has not yet been sufficiently elucidated. The results of the present study, obtained in a large group of patients with major depression, demonstrate that amitriptyline treatment is associated with a plasma concentration-dependent decrement of HRV parameters, including time and frequency derived measures, as well as HRV responses to different provocation procedures (deep breathing, Valsalva, change of posture) (figs l-3). The coefficient of variation while resting and the results of the spectral analysis are the most sensitive heart rate parameters with regard to TCA intake. Therefore, we have proposed that
Heart
193
rate variability
these tests should be widely used as a semi-quantitave tool to detect TCA-induced anticholinergic delirium, TCA overdosage, and intoxication (Rechlin et al, 1995). Previous studies suggest that vagal activity influences HRV at all frequencies up to 0.5 Hz, while the sympathetic nervous system affects HRV below 0.15 Hz (Akselrod et al, 1981; Pomeranz et al, 1985; Lishner et al, 1987; Weise et al, 1987; Freeman et al, 1991). The LF band seems to be related to vasomotor activity, the MF band to baroreceptor activity, and the HF band to respiratory sinus arrhythmia (Pagani et al, 1986). Interestingly, in the amitriptyline treated patients, the abnormal results of the spectral analyses were most pronounced in the MF bands, which was also found in a preliminary study by Yeragani et al (1994) in nortriptyline treated patients. This may be due to decreased activity in the oscillating neuronal network and the reticular neurones in the lower brainstem caused by amitriptyline, as these neurones have their preferential firing rate around 0.1 HZ (Langhorst et al, 1984). It has been demonstrated by others that atropine has a pronounced effect on the HF band (Pomeranz et al, 1985; Weise et al, 1987), which is a relatively pure measure of parasympathetic function. If baroreceptors were essentially influenced by TCA, this would be important in understanding TCAinduced postural hypotension, especially in elderly patients with reduced autonomic function and impaired reactivity. Therefore, it might be advisable to perform a heart rate analysis in each patient at the beginning of an anticholinergically active therapy. Since we could show that treatment with 20 mg/day of the selective serotonin reuptake inhibitor (SSRI) paroxetine does not cause alterations of cardiac autonomic function (Rechlin et al, 1994c), SSRI may be used in depressed patients with a pre-existing cardiovascular autonomic neuropathy to minimize the cardiovascular risk of antidepressant treatment. It is not known whether the massive changes of HRV (functional cardiovascular autonomic neuropathy in 67.3% of the patients) associated with amitriptyline treatment might have serious consequences. TCA have been used for a long time, and if the recommended precautions are observed, they are rather safe drugs. In diabetics, cardiovascular autonomic polyneuropathy has been observed in about 15-20% of insulin dependent patients (Ziegler et al, 1992b). At the same time, it has been shown that these patients have an elevated mortality risk (Ewing et al, 1980; O’Brien et al, 1991). However, analyses of the increased mortality risk revealed that most of the diabetic
patients with cardiovascular autonomic neuropathy died of renal complications and not of sudden death (Ewing et al, 1980; O’Brien et al, 1991). Thus, there might be a closer relationship between cardiovascular autonomic dysfunction and other severe complications caused by diabetes than between cardiovascular autonomic neuropathy and elevated mortality risk. Most of the patients treated with amitriptyline in therapeutic dosage showed normal HRV during deep respiration indicating that the baroreflex mechanisms were not totally disrupted as often is the case in diabetic polyneuropathy. These findings could support the notion that decreased HRV has no severe consequences for TCA treated patients. However, recently a 17-fold higher risk of death from cardiac infarction was reported in young women treated with antidepressants or benzodiazepines (Thorogood et al, 1992). Since benzodiazepines have no cardiac side effects, the reason for the high infarction risk in unknown. Nevertheless, it cannot be excluded that autonomic dysfunction in TCA treated patients might be a trigger factor for sudden death (Riddle et al, 1993) and severe cardiovascular complications in rare cases. Measurements of heart rate variability are increasingly applied to investigate autonomic cardiac function in psychiatric disorders and to assess the impact of psychopharmacological treatment on the mechanismsof baroreceptor reflexes and sinus arrhythmia. With regard to autonomic functions, long term cardiovascular risk studies are required in patients treated with TCA and other anticholinergic drugs in order to elucidate possible early and late adverse effects of these drugs. Furthermore, heart rate analyses may be used as a sensitive, semi-quantitative method to detect the intake of TCA in unconscious or delirious patients.
ACKNOWLEDGMENTS work wassupportedby the M and F Marohn Foundation, ‘Grant awardedto D Claus,MD’. This
REFERENCES Akselrod S, Gordon D, Ubel FA, Shannon DC, Barger AC, Chen RJ. Power spectrum analysis of heart rate fluctuation: a quantitative probe of beat-to-beat cardiovascular control. Science 198 I ;2 13:22%2 Bannister R, Mathias C. Testing autonomic reflexes. In: Bannister R, ed. Autonomic failure: a textbook of clinical disorders of rhe auronomic nervous system. London: Oxford University Press, 1988;289-307 Dalack GW, Roose SP. Perspectives on the relationship between cardiovascular disease and affective disorder. I C/in Psychiatry 1990;5 1 (suppl 7):4-9
194
T Rechlin
Ewing DJ, Campbell IW, Clarke BF. Assessment of cardiovascular effects in diabetic autonomic neuropathy and prognostic implications. Ann Intern Med 1980:92:308-l 1 Freeman k, Saul JP, Roberts MS, Berger RD, Broadbridge C, Cohen RJ. Spectral analysis in diabetic autonomic neuropathy. Arch Neuro 1991;48:185-90 Jakobsen J, Hauksson P, Vestergaard P. Heart rate variation in patients treated with antidepressants. An index of anticholinergic effects? Psychopharmacology 1984;84:544-8 Kristensen E, Jakobsen J, Bartels U, Vestergaard P. Cholinergic dysfunction of heart, pupil, salivary glands, and urinary bladder in healthy volunteers during long-term treatment with clomipramine. Psychopharmacology 1989;98: 398-402. Langhorst P, Schulz G, Lambertz M. Oscillating neuronal network of the “common brainstem system”. In: Miyakawa et al, eds. Mechanism of blood pressure waves. Berlin: Springer Verlag 1984;257-75 Lishner M, Akselrod S, Mor Avi V, Oz 0, Divon M, Ravid M. Spectral analysis of heart rate fluctuations. A non-invasive, sensitive method for the early diagnosis of autonomic neuropathy in diabetes mellitus. J Auton Nerv Syst 1987;19:119-25 Low PA, Opfer-Gehrking TL. Differential effects of amitriptyline on sudomotor, cardiovagal, and adrenergic function in human subjects. Muscle & Nerve 1992;15:134w Low PA. Autonomic nervous system function. J Clin Neurophysiol 1993;10:14-27 O’Brien IA, McFadden JP, Corral1 RJM. The influence of autonomic neuropathy on mortality in insulin-dependent diabetes. Q J Med, New Series 79,1991;290:495-502 Pagani M, Lombardi F, Guzetti S et al. Power spectral analysis of heart rate and arterial blood pressure variabilities as a marker of sympathovagal interaction in man and conscious dog. Circ Res 1986;59: 178-93 Pomeranz B. Macaulay RIB, Caudill MA, Kutz I, Adam D, Gordon D. Assessment of autonomic function in humans by heart rate spectral analysis. Am .I Physiol 1985;248:H151-3 Rechlin T. Decreased parameters of heart rate variation in amitriptyline treated patients: lower parameters in melancholic depression than in neurotic depression: a biological marker? Biol Psychiatry 1994;36:705-7 Rechlin T, Claus D, Weis M. Heart rate analysis in 24 patients treated with 150 mg amitriptyline per day. Psychophurmacology 1994a;116:110-14 Rechlin T, Claus D, Weis M. Heart rate variability in schizo-
et al phrenic patients and changes of autonomic heart rate parameters during treatment with clozapine. Biol Psychiatry 1994b;35:888-92 Rechlin T, Weis M, Claus D. Heart rate variability in depressed patients and differential effects of paroxetine and amitriptyline on cardiovascular autonomic functions. Pharmucopsychiutry 1994c;27: 124-8 Rechlin T, Weis M, Claus D, Kaschka WP. Identifying delirious states and autonomic cardiovascular dysfunction associated with amitriptyline treatment by standardized analysis of heart rate. Psychiatry Res 1995; in press Riddle MA, Geller B, Ryan N. Another sudden death in a child treated with desimipramine. J Am Acad Child Adolesc Psychiatry 1993;32:792-5 Thorogood M, Cowen P, Mann J, Murphy P, Vessey M. Fatal myocardial infarction and use of psychotropic drugs in young women. L.ancet 1992;340: 1067-8 Weinberg CR, Pfeifer MA. An improved method for measuring heart rate variability: assessment of cardiac autonomic function. Biometrics 1984;40:855-61 Weise F, Heydenreich F, Runge U. Contributions of sympathetic and vagal mechanisms to the genesis of heart rate fluctuation during orthostatic load: a spectral analysis. J Auton Nerv Sys 1987;21:127-34 Yeragani VK, Pohl R, Balon R et al. Heart rate variability in patients with major depression. Psychiatry Res 1992a;37:3546 Yeragani VK, Pohl R, Ramesh C et al. Effect of imipramine treatment on heart rate variability measures. Neuropsychobiology 1992b;26:27-32 Yeragani VK, Srinivasan K, Pohl R, Berger R, Balon R, Ramesh C. Effects of nortriptyline on heart rate variability in panic disorder patients: a preliminary study using power spectral analysis of heart rate. Neuropsychobiology 1994;29: l-7 Zahn TP, Pi&r D. Autonomic effects of clozapine in schizophrenia: comparison with placebo and fluphenazine. Biol Psychiatry 1993;34:3-12 Ziegler D, Laux G, Dannehl K et al. Assessment of cardiovascular autonomic function: agerelated normal ranges and reproducility of spectral analysis, vector analysis and standard tests of heart rate variation and blood pressures responses. Diabetic Med 1992a;9: 166175 Ziegler D, Gries FA, Sptiler M, Lessmann F. The epidemiology of diabetic neuropathy. J Diabetic Complications 1992b;6:49-57