Background

Intradialysis hypotension is a common problem among hemodialysis patients. Some studies have shown that autonomic neuropathy could be a major cause of intradialysis hypotension, whereas others have not. Furthermore, whether there are parallel changes in the autonomic nervous system and somatic nerves remains unclear.

Methods

We investigated the autonomic and peripheral nervous functions of 12 chronic hemodialysis patients who suffered from intradialysis hypotension, and of 12 age- and sex-matched hemodialysis patients who had stable blood pressure during hemodialysis. We used spectral analysis of their heart rate variability and systemic vascular resistance to evaluate autonomic functions. Vibrameter and nerve conduction studies to assess peripheral nervous function were also performed. Low-frequency/high-frequency ratio power index was used as a surrogate of sympathovagal balance.

Results

The power index rose progressively in the control group and reached significantly high levels at hour 4 compared to the basal values (3.7 0.5 vs. 2.1 0.3; P < 0.05). However in the group prone to hypotension, the power index remained almost unchanged. In addition, their systemic vascular resistance was lower than that in the control group (13.7 1.8 vs. 22.3 2.6 Wood units; P < 0.05). The vibration perception thresholds of the index finger and great toe were significantly higher in the group prone to hypotension (4.7 0.7 vs. 2.2 0.3 vibration units and 3.1 0.4 vs. 1.5 0.2 vibration units, respectively; both P < 0.05).

Conclusion

We found that more severe damage to autonomic and peripheral nervous system occurred in patients prone to hypotension.

METHODS

Patients

Two groups of chronic hemodialysis patients were enrolled in this study. The first group consisted of 12 hypotension-prone patients, 4 women and 8 men, with a mean age of 61.6 5.1 years (28–83 years). The underlying diseases of renal failure were chronic glomerulonephritis in 6 patients, chronic interstitial nephritis in 5 patients, and nephrosclerosis in 1 patient. Intradialysis hypotension was defined as a fall of systolic arterial blood pressure to below 90 mm Hg with a drop of blood pressure of more than 20 mm Hg, and the appearance of nausea, vomiting, or muscle cramps. Hemodialysis patients who suffered from intradialysis hypotensive episodes in at least 50% of the dialysis sessions during the 2 months before the study were enrolled as study group. The control group consisted of 12 age- and sex-matched hemodialysis patients who had not experienced any hemodynamic events during the same period. The underlying diseases of renal failure in the control group were chronic interstitial nephritis in 7 patients, chronic glomerulonephritis in 3 patients, and gouty nephropathy in 2 patients. All of the patients were treated 3 times a week for 4 hours with diacetyl-cuprophan dialyzers of a 1.7 m² surface area, blood flow rates of 200 to 250 mL/min, and a dialysate flow rate of 500 mL/min. The dialysate composition was Na⁺ 140 mEq/L; K⁺ 2 mEq/L; Ca²⁺ 3.5 mEq/L; HCO₃^- 35 mEq/L; and glucose 200 mg/dL. The dialysate temperature was kept at 36.5°C. The dry weights of our patients were set by clinical assessment and by trial and error. All the intradialytic hypotensions of the hypotension-prone group could not be resolved by increasing dry weight. The ultrafiltration rate was held constant during the treatment. Patients with diabetes mellitus, congestive heart failure (more severe than New York Heart Association Functional Class II), and on medications that may affect the cardiovascular or autonomic nervous system were excluded from this study. The protein catabolic rate and general nutritional states were similar between the two groups. All patients complied well with the fluid restriction recommendations and maintained their interdialysis weight gain of less than 1 kg per day and less than 5% of post-dialysis dry weight between two consecutive hemodialysis sessions. All procedures were performed in the afternoon from 1:00 PM to 5:00 PM. The local ethics committee approved the study protocols and all patients gave written informed consent.

Nerve conduction studies

NCV studies were conducted using a Viking IV Electromyographer (Nicolet, Madison, WI, USA). We chose the peroneal and tibial nerves to represent the motor component, and the sural nerve to represent the sensory component. Motor conduction studies of the peroneal and tibial nerves were conducted with surface electrodes arranged in belly-tendon fashion. The electrodes were placed over the extensor digitorum brevis muscles for the peroneal nerve and over the abductor hallucis muscle for the tibial nerve. In antidromic sensory conduction studies, sural electrodes were placed over the lateral malleoli. Active and reference pickup electrodes were placed 3 cm apart. Supramaximal stimuli were applied percutaneously with a hand-held bipolar nerve stimulator. The stimulation sites of the peroneal, tibial, and sural nerves were at the ankle, about 8 and 14 cm from the recording electrodes, respectively. The amplitudes of the compound motor action potential (CMAP) and sensory nerve action potential (SNAP) were automatically measured from the baseline to the negative peak. We recorded distal latency (DL) to onset, motor nerve conduction velocities (MNCV) and CMAP amplitude of the peroneal and tibital nerves, as well as distal latency to onset of take-off and SNAP amplitude of the sural nerve.

Vibration perception thresholds

VPT determinations were performed immediately after completion of nerve conduction study using the Vibration II (Sensortec, Inc., Clifton, NJ, USA). The vibration stimulus was delivered at a constant frequency of 120 Hz. By decreasing the output voltage, the amplitude of the vertical excursions of the vibrating rod was simultaneously reduced. The output voltage was displayed in "vibration" units (VU) as a digital readout. VU determinations were carried out using the "two-alternative, forced-choice" method, in which the patient was required to identify which of the two rubber rods ("A" or "B") was vibrating after a randomized A/B sequence at progressively decreasing VU levels. The study was terminated upon the fifth error. The required time to perform the study, including patient instruction, was generally less than 15 minutes. VU measurements were taken from the pulp of the index finger and great toe, from the same extremity on which the NCV studies were performed. The VPT was calculated from the 10 VU values representing the 5 errors and 5 lowest correct scores. The highest and lowest VU values were eliminated, and the mean of the remaining 8 VU values represented the VPT.

Heart rate variability analysis and autonomic function assessment

We checked blood pressure 10 minutes before puncturing, at the beginning of dialysis, every 15 minutes during hemodialysis, and at the end of hemodialysis. We prepared a Holter ECG recorder in all patients while receiving hemodialysis. The Holter ECG signal was recorded during the whole duration of hemodialysis using an Oxford solid-state 3-channel recorder (Medilog® Holter recorders; Oxford Instruments, Fremont, CA, USA). By detecting the QRS complex using Excel software version 2.0 (Microsoft Corp., Redmond, WA, USA), the signals were automatically processed on the Oxford laser Holter scanner to perform heart rate variability analysis. Spectral analysis of HRV was performed every 5 minutes using the Welch method on short-lasting heart rate tracers. We carried out heart rate variability analysis according to the recommendations of the task force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology¹⁰. We performed power spectral analysis using Fast Fourier transformation for the Holter ECG signals. Low-frequency (LF, 0.040.15 Hz) components, high-frequency (HF, 0.150.4 Hz) components, and the ratio of LF/HF were measured. The ratio of LF/HF (power index) was considered to mirror sympathovagal balance and reflect the sympathetic modulations^7,11,12.

Systemic vascular resistances

ECGs were performed at the end of hemodialysis using a Sonos 5500 (Hewlett Packard; Agilent, Palo Alto, CA, USA) echocardiographic system. M-mode measurements were obtained according to American Society of Echocardiography standards. We measured aortic annulus dimension (Ao) and aortic Doppler velocity time integral (AoVTI). Systolic (SBP) and diastolic (DBP) blood pressures were obtained using arm cuff measurements of a sphygmomanometer immediately prior to the echocardiograms. The SVR were determined by the following equations:

Statistics

All data were expressed as mean SEM or proportions as appropriate. To analyze the data, we applied unpaired t tests to test the differences between groups and repeated measure of analysis of variance (ANOVA) to examine the differences among serial measurements within groups. Once a difference was found in repeated measure of ANOVA, the Student-Neuman-Keuls test was used for pair-wise comparisons. Fisher's exact test was used to analyze proportional data. All p levels were two-tailed and values less than 0.05 were considered significant. Statistics were conducted using SPSS 8.0.1C (SPSS, Inc., Chicago, IL, USA).

RESULTS

The general characteristics of the hypotension-prone and control groups are shown in Table 1. There were no significant differences in age, sex, dialysis duration, body weight, body height, ultrafiltration rate of hemodialysis, dialysis dosage, serum prealbumin, hematocrit, and plasma intact parathyroid hormone (iPTH) values between these two groups.

Table 1 - General characteristics of patients.

Full table

Table 2 shows the results of VPT measurements and NCV studies. According to NCV studies, there were no significant differences in the prevalence of peripheral neuropathy (55% in hypotension-prone group and 36% in the control group). There were also no significant differences in the parameters, including MNCVs, CMAPs, and DLs of the peroneal nerve and tibial nerve, and SNAP and DL of the sural nerve, between the hypotension-prone and the control groups. However, the VPT measurements of the index finger and the great toe were significantly higher in the hypotension-prone group than that in the control group (2.2 0.3 vs. 1.5 0.2 VU, 4.7 0.7 vs. 3.1 0.4 VU, respectively; P < 0.05).

Table 2 - Results of vibration perception threshold measurements and nerve conduction studies.

Full table

Figures 1a and 1b show the changes of mean arterial pressure and heart rate during hemodialysis, respectively. Compared with the control group, the hypotension-prone group had a significantly lower mean arterial pressure during hemodialysis, except for the basal value; the levels measured between hour 2 and 3 were significantly lower than that at hour 0 (70 3 and 78 4 vs. 90 4 mm Hg; both P < 0.05). On the contrary, the control group showed no change in mean arterial pressure throughout the whole dialysis session. The data for frequency domain of spectral analysis of heart rate variability are shown in Table 3. There were no significant differences of HF or LF between the two groups, except for the basal level of HF. The power index of heart rate variability is shown in Figure 1c. The hypotension-prone group showed a significantly lower LF/HF ratio than that in the control group from the beginning to the end of hemodialysis. The serial measurements of the LF/HF ratio rose progressively and reached a significantly higher level at hour 4 in the control group (3.7 0.5 vs. 2.1 0.3; P < 0.05), whereas in the hypotension-prone group, the LF/HF ratio did not change significantly throughout the whole course of hemodialysis, even when the mean arterial pressure was significantly decreased. In addition, we also found that the SVR was significantly lower in the hypotension-prone group than in the control group (13.7 1.8 vs. 22.3 2.6 Wood units; P < 0.05) Table 4.

Figure 1.

Power index (LH/HF ratio), mean arterial pressure, and heart rate during dialysis of hypotension-prone patients and controls.*P < 0.05 and **P < 0.005 for hypotension-prone patients vs. controls; ***P < 0.05 vs. baseline values of respective groups. LF/HF ratio is the ratio between low-frequency and high-frequency power of heart rate variability.

Full figure and legend (27K)

Table 3 - Results of the heart rate variability.

Full table

Table 4 - Hemodynamic characteristics.

Full table

DISCUSSION

Spectral analysis of the heart rate variability has been proven to be a useful and noninvasive tool for monitoring the variations of sympathovagal balance controlling heart rate and vasomotor tone^7,11,12,13. The HF components are of purely vagal origin, but there is controversy in the interpretation of the LF components, which some assumed to be markers of sympathetic modulation, but others as reflections of both sympathetic and parasympathetic controls^12,14,15,16. Thus, we used the LF/HF ratio, the power index, as a surrogate of the sympathovagal balance¹⁷. Although the VPT test has not been commonly used to assess peripheral nervous function, it has been reported to have a good correlation with the clinical score of uremic neuropathy^12,15. In addition, the intra-individual variation of the VPT test has also been demonstrated to be acceptable and comparable between normal and uremic patients¹². Because the VPT test is easily performed and easily tolerated by the patient, the normal variation, as well as the reproducibility, can be properly defined, and the elevated values are relevant and sufficiently specific for peripheral nerve function, we used the VPT test in addition to the NCV study to assess peripheral nervous function.

In our study, we found that the hypotension-prone hemodialysis patients had lower basal levels and lesser change of the power index, as well as lower systemic vascular resistance and higher vibration perception threshold. Because fluid removal during ultrafiltration is a potent stimulus for the sympathetic activity, we kept a constant ultrafiltration rate in both groups throughout the whole course of hemodialysis. The results of lesser changes of the power index and lower systemic vascular resistance suggested that under the similar stimulation, the autonomic nervous system of the hypotension-prone patients could not compensate adequately and would lead to hemodynamic instability. This might be caused by a dysregulation of sympathetic activity or sympathovagal imbalance. Furthermore, a lower basal LF/HF ratio was found in the hypotension-prone patients in our study, which is consistent with the results of previous reports^18,19. This might imply that the hypotension-prone patients have chronically reduced efficiency of the autonomic system.

Around 10% to 80% of uremia patients have been reported to have peripheral polyneuropathy^20,21. However, the results of previous reports in the literature about damages that occurred in somatic and autonomic nervous systems of uremic patients are inconsistent^5,6. Most of these studies were performed using results of the NCV tests. Compared with the VPT measurement, results of the NCV studies are less sensitive to detect subtle changes in peripheral nervous system. Theoretically, any part of the tract, including the receptor, Pacinian corpuscles, the very distal portion of the A nerve fibers, the dorsal column, and the sensory cortex could be assessed efficiently using the VPT test, but the NCV study could only be used to evaluate more proximal portions of the nerves. In addition, normal vibration sensations depend on synchronous volleys of impulse trains in a large number of A nerve fibers, but the NCV study evaluates only one nerve at a time. Therefore, dysfunction of the axonal membrane or of the receptor might desynchronize the impulse trains without displaying abnormalities in a single nerve test. Because uremic polyneuropathy is predominantly dying-back axonal degeneration²², the test of the VPT seems to be a sensitive and easy method to detect peripheral neuropathy in hemodialysis patients. Therefore, it is not surprising that we found parallel damages of autonomic nervous system and vibration perception threshold, but not nerve conduction velocity.

Although we did not perform nerve biopsies among our patients, the impairment of the VPT due to segmental demyelinization of the afferent nerve fibers has been demonstrated in histopathologic studies in diabetics²³, as well as in uremic patients²⁴.

CONCLUSION

We found that more severe damage occurred in the autonomic nervous system in the hypotension-prone patients, which might be the cause of intradialytic hypotension. In addition to the autonomic dysfunction, the vibration perception threshold was higher in the hypotension-prone patients, which indicated a more severe generalized polyneuropathy in these patients.

References

1. Schreiber, MJ Jr: Clinical case-based approach to understanding intradialytic hypotension. Am J Kidney Dis 2001 38 (Suppl 4): S1–S10,  | PubMed | ISI |
2. Daugirdas, JT: Pathophysiology of dialysis hypotension: An update. Am J Kidney Dis 2001 38 (Suppl 4): S11–17,  | PubMed | ISI | ChemPort |
3. Converse, RL Jr, Jacobsen, TN, Jost, CM, et al: Paradoxical withdrawal of reflex vasoconstriction as a cause of hemodialysis-induced hypotension. J Clin Invest 1992 90: 1657–1665,  | PubMed | ISI |
4. Ligtenberg, G, Blankestijn, PJ, Boomsma, F, Koomans, HA: No change in automatic function tests during uncomplicated haemodialysis. Nephrol Dial Transplant 1996 11: 651–656,  | PubMed | ISI | ChemPort |
5. Ogura, T, Makinodan, A, Kubo, T, et al: Electrophysiological course of uraemic neuropathy in haemodialysis patients. Postgrad Med J 2001 77: 451–454,  | Article | PubMed | ISI | ChemPort |
6. Vita, G, Bellinghieri, G, Trusso, A, et al: Uremic autonomic neuropathy studied by spectral analysis of heart rate. Kidney Int 1999 56: 232–237,  | Article | PubMed |
7. Malik, M: Heart rate variability. Standards of measurement, physiological interpretation, and clinical use. Circulation 1996 93: 1043–1065,  | PubMed | ISI |
8. Nielsen, VK: The peripheral nerve function in chronic renal failure. IV. An analysis of the vibratory perception threshold. Acta Med Scand 1972 191: 287–296,  | PubMed | ISI | ChemPort |
9. Tegner, R, Lindholm, B: Vibratory perception threshold compared with nerve conduction velocity in the evaluation of uremic neuropathy. Acta Neurol Scand 1985 71: 284–289,  | PubMed | ISI | ChemPort |
10. Klima, RR, Weigand, AH, DeLisa, JA: Nerve conduction studies and vibratory perception thresholds in diabetic and uremic neuropathy. Am J Phys Med Rehabil 1991 70: 86–90,  | PubMed | ISI | ChemPort |
11. Akselrod, S, Gordon, D, Madwed, JB, et al: Hemodynamic regulation: Investigation by spectral analysis. Am J Physiol 1985 249: H867–875,  | PubMed | ISI | ChemPort |
12. Askelrod, S, Gordon, D, Ubel, FA, et al: Power spectrum analysis of heart rate fluctuation: A quantitative probe of beat-to-beat cardiovascular control. Science 1981 213: 220–222,  | PubMed | ISI | ChemPort |
13. van Ravenswaaij-Arts, CM, Kollee, LA, Hopman, JC, et al: Heart rate variability. Ann Intern Med 1993 118: 436–447,  | PubMed | ChemPort |
14. Kamath, MV, Fallen, EL: Power spectral analysis of heart rate variability: A noninvasive signature of cardiac autonomic function. Crit Rev Biomed Eng 1993 21: 245–311,  | PubMed | ISI | ChemPort |
15. Montano, N, Ruscone, TG, Porta, A, et al: Power spectrum analysis of heart rate variability to assess the changes in sympathovagal balance during graded orthostatic tilt. Circulation 1994 90: 1826–1831,  | PubMed | ISI | ChemPort |
16. Appel, ML, Berger, RD, Saul, JP, et al: Beat to beat variability in cardiovascular variables: Noise or music? J Am Coll Cardiol 1989 14: 1139–1148,  | Article | PubMed | ISI | ChemPort |
17. Malliani, A, Lombardi, F, Pagani, M: Power spectrum analysis of heart rate variability: A tool to explore neural regulatory mechanisms. Br Heart J 1994 71: 1–2,  | PubMed | ISI | ChemPort |
18. Takahashi, H, Matsuo, S, Toriyama, T, et al: Autonomic dysfunction in hemodialysis patients with persistent hypotension. Nephron 1996 72: 418–423,  | PubMed | ISI | ChemPort |
19. Cavalcanti, S, Severi, S, Chiari, L, et al: Autonomic nervous function during haemodialysis assessed by spectral analysis of heart-rate variability. Clin Sci 1997 92: 351–359,  | PubMed | ISI | ChemPort |
20. Asbury, AK: Uremic neuropathy, in Peripheral Neuropathy, 1978, edited by Dyck PJ, Thomas PK, Lambert EH, Philadelphia, WB Saunders, pp 982–990
21. Bazzi, C, Pagani, C, Sorgato, G, et al: Uremic polyneuropathy: A clinical and electrophysiological study in 135 short- and long-term hemodialyzed patients. Clin Nephrol 1991 35: 176–181,  | PubMed | ISI | ChemPort |
22. Dyck, PJ, Johnson, WJ, Lambert, EH, O'Brien, PC: Segmental demyelination secondary to axonal degeneration in uremic neuropathy. Mayo Clin Proc 1971 46: 400–431,  | PubMed | ISI | ChemPort |
23. Thomas, PK: Diabetic neuropathy: Morphological aspects. Proc R Soc Med 1967 60: 145,  | PubMed | ISI | ChemPort |
24. Forno, L, Alston, W: Uremic polyneuropathy. Acta Neurol Scand 1967 43: 640–654,  | PubMed | ISI | ChemPort |

Acknowledgments

This work was supported by grants from Kaohsiung Veterans General Hospital VGHKS89-42 to Dr. Kang-Ju Chou, and VGHKS90-26 to Dr. Hua-Chang Fang.