Cardiovascular autonomic nervous system dysfunction in. rheumatoid arthritis (RA) and systemic lupus erythematosus

Transcription

1 Q J Med 1999; 92: Cardiovascular autonomic nervous system dysfunction in patients with rheumatoid arthritis and systemic lupus erythematosus W. LOUTHRENOO, P. RUTTANAUMPAWAN, A. ARAMRATTANA1 and W. SUKITAWUT From the Division of Rheumatology, Department of Medicine, and 1Department of Family Medicine, Faculty of Medicine, Chiang Mai University, Chiang Mai, Thailand Received 1 July 1998 and in revised form 3 December 1998 Summary Although peripheral and central nervous system involvement have been well recognized in patients with rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE), autonomic nervous system (ANS) involvement has rarely been studied, and has shown conflicting results. We performed cardiovascular ANS assessment in 34 RA and 37 SLE patients, using standard cardiovascular reflex tests. The results in each patient were compared with ageand sex-matched healthy controls. Forty-seven percent of the RA patients and 19% of the SLE patients Introduction Although involvement of the peripheral nervous system (PNS) and central nervous system (CNS) in rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE) has been well described, autonomic nervous system (ANS) involvement in RA and SLE has rarely been described in standard textbooks of rheumatology.1 3 The ANS can be assessed by several tests including cardiovascular, sweating, pupillary reflex and skin tests. However, cardiovascular reflex tests have been most widely used as they are non- invasive, and results are easy to reproduce. The purpose of this study was to evaluate cardio- vascular ANS function in patients with RA and SLE, and to correlate the ANS function with clinical features. As the ANS function has been reported to deteriorate with age,4 we used age- and sex-matched healthy controls. had symptoms suggesting ANS dysfunction. The heart rate variation in response to deep breathing was significantly decreased in both the RA and SLE patients ( p=0.001). This diminished heart rate response showed no correlation with the disease duration, the number of swollen joints, the Ritchie articular index, ESR, or rheumatoid factor in the RA group, or the disease duration, the SLEDAI score or ESR in the SLE group. The clinical significance of the diminished cardiovascular ANS response needs to be investigated. Methods The patients studied were in- and out-patients with RA and SLE at the Division of Rheumatology, Department of Medicine, Faculty of Medicine, Chiang Mai University. The diagnoses of RA and SLE were made by using the criteria developed by the American College of Rheumatology.5,6 Control sub- jects were selected from healthy hospital staff who did not have symptoms of ANS dysfunction and were not taking any medications. These controls had never been studied before. All patients were checked for symptoms and signs of possible ANS dysfunction, including orthostatic hypotension (lightheadedness, blurred vision, sensation of weakness and unsteadi- ness, fainting or syncope upon standing), perspira- tion, palpitation and Raynaud s phenomenon. The Address correspondence to Dr W. Louthrenoo, Associate Professor of Medicine, Division of Rheumatology, Department of Medicine, Faculty of Medicine, Chiang Mai University, Chiang Mai, 50200, Thailand Association of Physicians 1999

2 98 W. Louthrenoo et al. criterion for age matching was an age difference of considered statistically significant. Statistical analysis <3 years for each matched pair. used the SPSS for Windows program, release 7.0. Patients were excluded from the study if they (i) had <10 g/dl haemoglobin; (ii) were pregnant; (iii) had diseases interfering with the autonomic nervous system, including diabetes mellitus, renal Results and liver diseases, Parkinson s disease, porphyria Thirty-four RA and 37 SLE patients were included in and amyloidosis; (iv) had cardiovascular diseases this study. The characteristics of these patients are including hypertension, ischaemic heart disease, con- given in Table 1. Nineteen RA patients (55.8%) gestive heart failure, valvular heart disease, cardiomyopathy had positive RF (1580). Their mean number of and cardiac arrhythmia; or (v) had swollen joints, the Ritchie articular index and neurological diseases including multiple sclerosis, the ESR were 15.50±10.46, 11.59±6.08, and polyneuropathy or Guillain-Barré syndrome. Patients 35.22±18.27 mm/h, respectively. None had rheumatoid taking drugs that interfered with the ANS including vasculitis or peripheral neuropathy. All RA antihypertensive, diuretic, adrenergic inhibitor, vaso- patients were taking non-steroidal anti-inflammatory dilator, anti-arrhythmic, sedative, hypnotic and antiepileptic drugs and disease-modifying anti-rheumatic drugs at drugs were also excluded from the study. the standard dosage. Cardiovascular ANS function assessment was performed The mean modified SLEDAI score and ESR in SLE by one of us, PR, using the standard technique patients were 4.65±5.77, and 27.65±11.67 mm/h, described by Ewing et al.7 Two tests, the heart rate respectively. A past history of vasculitis, CNS lupus response to deep breathing and the immediate heart and peripheral neuropathy was documented in 12, rate response to standing (or the R-R ratio), were six and two patients, respectively. Among the patients performed to measure heart rate changes, reflecting with CNS lupus, there were seizures in three, organic the parasympathetic nerve function. The other two brain syndrome in one, transverse myelitis in one tests, the systolic blood pressure response to standing and aseptic meningitis in one. Two patients who had and the diastolic blood pressure response to sustained peripheral neuropathy had had mononeuritis multi- handgrip, were done to measure changes in blood plex. At the time of the study, four patients had pressure, reflecting the sympathetic nerve function. cutaneous vasculitis and one had a residual ulnar All patients and controls were confirmed to have a neuropathy from her previous neuropathy. Antinuclear normal sinus rhythm, without evidence of a conduction antibody test, using the indirect immunofluonormal defect in a standard electrocardiograph (ECG). rescent method, was positive in 31 and was negative The tests were performed under standardized conditions, in six cases. in the morning, after a period of relaxation. There were 62 healthy controls (50 females and Tobacco, alcohol and medications were not allowed 12 males). Each RA and SLE patient was matched before the tests. for age with these controls, and for sex where RA activity was determined by using the number possible. The mean±sd ages of the control groups of swollen joints, the joint tenderness scores deter- for RA and SLE were 47.0±10.6 and 30.3±7.9 mined by the Ritchie articular index, the rheumatoid years, respectively ( p>0.05); the percentages of sex- factor (RF) determined by the latex agglutination test matched control patients in the RA and SLE groups and erythrocyte sedimentation rate (ESR). An RF titre were 76% and 81%, respectively. Results of the of 1580 or more was considered significant. The cardiovascular ANS function assessment in RA and activity of SLE was determined by ESR and the SLE patients and their age- and sex-matched controls SLEDAI score.8 The SLEDAI score was modified in are shown in Table 2. The heart rate response to this study, as the DNA binding assay was not deep breathing was significantly diminished in both available at our hospital, and its score was not RA and SLE groups when compared with controls counted as part of our SLEDAI score. Thus our ( p=0.001). There was no correlation between the maximum modified SLEDAI score was 103. diminished heart rate response to deep breathing and number of swollen joints, the Ritchie articular index or ESR or RF titre in RA patients, nor to the Statistical analysis modified SLEDAI score or ESR in SLE patients. There Test values are reported as means±sd. Student s t was no significant correlation between the extent to test for paired samples was used to compare these which the systolic blood pressure fell and the orthostatic means. Wilcoxon matched-pairs signed-ranks test hypotension symptoms, nor between the heart was used for non-parametric values. Regression ana- rate variation in response to deep breathing or the lysis was used to find the correlation between the cardiovascular ANS function test and disease activity or dosage of medication. A p value <0.05 was R-R ratio and palpitation in both RA and SLE groups. A significant correlation between the diminished heart rate response to deep breathing and the dosage

3 Cardiovascular autonomic dysfunction in RA and SLE 99 Table 1 Characteristics of rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE) patients RA (n=34) SLE (n=37) Sex (F5M) Age (years) 47.2± ±8.1 Duration (years) 5.1± ±2.2 Medication: n (dosage) Prednisolone (mg/day) 9 (4.45±1.10) 35 (17.03±13.59) Chloroquine (mg/day) 20 (168.45±61.17) 21 (202.33±71.10) Methotrexate (mg/week) 11 (4.55±1.51) Cyclophosphamide (mg/day) 13 (50.00±17.68) ANS symptoms Orthostatic hypotension 16 7 Palpitation 12 7 Raynaud s symptom 5 5 Perspiration 4 5 Data are either numbers, or means±sd. Table 2 Results of cardiovascular autonomic assessment in rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE) Parameter RA patients RA controls p SLE patients SLE controls p (n=34) (n=34) (n=37) (n=37) HRmax (beats/min) 91.59± ± ± ±11.80 HRmin (beats/min) 77.38± ± ± ±9.92 HRvar (beats/min) 14.20± ± ± ± R-R15 (mm) 16.19± ± ± ±2.77 R-R30 (mm) 16.19± ± ± ±2.47 R-R ratio 1.08± ± ± ± BPs baseline (mmhg) ± ± ± ±13.04 BPs stimulated (mmhg) ± ± ± ±12.81 BPs decrease 10.41± ± ± ± BPd baseline (mmhg) 76.65± ± ± ±9.66 BPd stimulated (mmhg) 87.00± ± ± ±10.15 BPd increase 10.35± ± ± ± Values are means±sd. HRmax, maximum heart rate during deep breathing; HRmin, minimum heart rate during deep breathing; HRvar, the difference between HRmax and HRmin; R-R 15, R-R interval at 15th heart beat after standing; R-R 30, R-R interval at 30th heart beat after standing; R-R ratio, the ratio of R-R 30 to R-R 15; BPs baseline, systolic blood pressure while supine; BPs stimulated, lowest systolic blood pressure while standing; BPs decrease, BPs baseline BPs stimulated; BPd baseline, diastolic blood pressure before handgrip; BPd stimulated, highest diastolic blood pressure during handgrip exercise; BPd increase, BPd stimulated BPd baseline. Discussion In this study, we used non-invasive cardiovascular reflex tests to evaluate cardiovascular ANS function. The heart rate response to deep breathing was found to be significantly diminished in both RA and SLE groups compared to their age- and sex-matched healthy controls. This diminished heart rate response showed no correlation to the duration of the disease, of chloroquine used in RA patients was observed ( p<0.05), but this correlation disappeared on mul- tiple regression analysis. the disease activity, or the dosage of the medication used in both RA and SLE. Unfortunately, we did not study the correlation between the presence of ANS dysfunction and the degree of articular destruction in our RA patients. Forty-seven percent of our RA and 19% of our SLE patients had symptoms suggesting ANS dysfunction. The high incidence of ANS dysfunction symptoms seen in our RA patients might have been related to their advanced age group. In this study, control subjects were recruited from healthy hospital staff. They were in general health- conscious. Thus, they might not represent the popula- tion from which the patients were drawn. As the controls had no symptoms of ANS dysfunction, it is

4 100 W. Louthrenoo et al. not known whether the ANS dysfunction symptoms study is not clear. Liote et al.17 found that 88% of seen in both our RA and SLE patients are more their 17 SLE patients with mild disease activity had common than those of the general population. at least one abnormal cardiovascular reflex test. Both Cardiovascular ANS assessment in RA has been the standing heart rate ratio and the fall in standing studied by several investigators, and there have been systolic blood pressure were abnormal, reflecting conflicting results because of the way in which abnormalities in both the parasympathetic and the controls were selected and the criteria used to sympathetic nervous system. In contrast to Gledhill s determine ANS dysfunction. Tests measuring heart study,16 no relationship between autonomic neuro- rate variation have been most commonly employed, pathy and peripheral neuropathy was found. Straub and these have usually given abnormal results. et al.18 found that three of their 31 SLE patients Edmonds et al.9 demonstrated abnormalities in para- (9.7%) had cardiovascular ANS dysfunction defined sympathetic cardiovascular reflexes in nine of their by abnormal results in two of the five cardiovascular 27 RA patients (33%). Four of nine patients with reflex tests. The systolic blood pressure response to abnormal tests had symptoms suggesting ANS dysfunction. standing was abnormal in 29%. A significant correla- However, more than 50% of their patients tion between disease activity and the number of with abnormal tests had evidence of peripheral abnormal cardiovascular reflex tests was observed. neuropathy. Tan et al.10 also found abnormal heart Recently Lagana et al.19 and Laversuch et al.20 used rate response to deep breathing in eight of their h ECG monitoring in addition to the standard RA patients (27%). Five of these eight had had cardiovascular reflex tests to evaluate cardiovascular clinical symptoms of dysautonomia. Nine patients ANS function and showed that a significant had evidence of peripheral neuropathy, of whom decreased heart rate variability occurred in SLE five had clinical symptoms of dysautonomia. Leden patients, especially at night when the parasympath- et al.11 studied 17 RA patients and found an increase etic nerve function is prominent, when compared in resting heart rate in all patients, and an abnormal with controls. Twenty-two percent of Laversuch s,20 heart rate response to standing in all seven patients but none of Lagana s19 patients had symptoms of who had severe RA. An increase in resting heart rate dysautonomia. The diminished heart rate variability in RA was confirmed by Piha et al.12 but they found showed no correlation with the disease duration, the no abnormalities in parasympathetic cardiovascular disease activity or serology. In contrast to the above reflex tests. This increased heart rate was independent studies, Omdal et al.21 found no abnormality in the of severity of the disease, pain and ESR. They cardiovascular ANS function in their 34 SLE patients concluded that the increase in resting heart rate when using age- and sex-matched controls. might have been related to vascular inflammation, The diminished heart rate response to deep breath- accelerated atherosclerosis and poor physical condi- ing in RA and SLE seen in this study is in line with tion. Parasympathetic cardiovascular ANS dysfunction previous studies. However, many of the previous was confirmed by Toussirot et al.13 who studies had some limitations, such as selection of demonstrated that 60% of their 50 RA patients had controls and statistical analyses. Although the ages ANS dysfunction, defined by abnormal results on of patients were matched to those of controls, they two of the three cardiovascular reflex tests. However, compared means of the group rather than the age- ANS dysfunction showed no correlation with the matched pairs used by Laversuch et al.,20 Omdal duration of the disease, inflammatory syndrome, RF et al.,21 Stein et al.,22 and our study. Moreover, some titre or articular destruction. Approximately 50% of studies used the standard-age reference values as their patients had severe disease and showed some controls. As the ANS has been shown to deteriorate degree of articular destruction. Geenen et al.14 found with age,4 it is crucial to correct for age, and if diminished ANS function in RA patients who had possible for sex, to strengthen results of the statist- had the disease for <1 year. This diminished ANS ical analysis. function was related to severity of pain and might Although 26% of our RA and 95% of our SLE have been related to the pathophysiological mechan- patients took low doses of corticosteroids, the diminished isms in RA. In contrast to previous studies, Bekkelund heart rate response to deep breathing seen in et al.15 found no cardiovascular ANS abnormality in our patients showed no correlation with the dosage their 43 RA patients, who had no ANS symptoms. of prednisolone used. This diminished heart rate An abnormal cardiovascular ANS function in SLE variability has been previously demonstrated to be has also been found in several studies. Gledhill independent of the use of corticosteroids.22 Most of et al.16 found that 13 of their 14 SLE patients (93%) our SLE patients had mild disease activity, with an without autonomic symptoms had cardiovascular average modified SLEDAI score of 4.65, similar to ANS dysfunction. However, 12 of these patients had those of Loite et al.,17 Laversuch et al.,20 and Stein evidence of peripheral neuropathy. The reason for et al.22 Patients with more severe disease tended to the high incidence of autonomic neuropathy in their have renal involvement, be anaemic, have high

5 Cardiovascular autonomic dysfunction in RA and SLE 101 blood pressure or to have conditions or be taking In summary, cardiovascular ANS dysfunction, particularly drugs which were excluded from our study. It might the parasympathetic nervous system, occurs be of interest to study the cardiovascular ANS in RA and SLE. It is usually not evident unless one function in the more severe SLE patients. specifically tests for it, by using the standard cardio- There have been substantial advances in the vascular reflex tests, before overt symptoms of ANS assessment of autonomic function, particularly of dysfunction occur. The significance of asymptomatic the cardiovascular system.23 However, the Ewing cardiovascular ANS dysfunction in clinical practice approach, which has been used to assess ANS remains to be investigated. One should be cautious function in clinical research for many years remains in prescribing drugs that have effects on the cardiovascular satisfactory.7,9,15,17,20,21 A 24-h ECG record has been ANS in patients with RA and SLE with recently used to evaluate heart rate variability in possible autonomic dysfunction. combination with the standard cardiovascular reflex tests as it is more sensitive in detecting the parasympathetic nervous system dysfunction, but an analysis References of the time and frequency domain for an evaluation 1. Conn DL. Rheumatoid neuropathy. In: Utsinger PD, Zvaifler of the heart rate variation from a 24-h ECG recording NJ, Ehrlich GE. eds. Rheumatoid arthritis. Philadelphia, JB is mathematically complex and perhaps cannot be Lippincott, 1985: done at a general hospital. Moreover, there is a good 2. Bacon PA. Extra-articular rheumatoid arthritis. In: McCarty correlation between the high frequency 24-h heart DJ, Koopman WJ. eds. Arthritis and allied conditions, 12th rate variability, which is influenced by the parasym- edn. Philadelphia, Lea & Febiger, 1993: pathetic nervous system, and the heart rate variation 3. Wallace DJ, Metzger AL. Systemic lupus erythematosus and in response to deep breathing.20 the nervous system. In: Wallace DJ, Hahn BH. eds. Dubois lupus erythematosus, 5th edn. Baltimore, Williams & Note the fall in the systolic blood pressure, at a Wilkins, 1997: mean of 10 mmhg in RA and SLE groups versus 8 4. Wieling W, van Brederode JFM, de Rijk LG, Borts C, and 6 mmhg, respectively, in the controls. The large Dunning AJ. Reflex control of heart rate in normal subjects standard deviation indicates that there were great in relation to age. A data base for cardiac vagal neuropathy. variations, implying that our subjects were heterogen- Diabetologia 1982; 22: eous. The fall in systolic blood pressure showed no 5. Arnett FC, Edworthy SM, Bloch DA, et al. The American correlation with the orthostatic hypotension sympclassification Rheumatism Association 1987 revised criteria for the of rheumatoid arthritis. Arthritis Rheum 1988; toms in both groups. However, there is no agreement 31: for the normal range for the postural change, and a decrease of as much as 25 mmhg can be normal Tan EM, Cohen AS, Fries JR, et al revised criteria for the classification of systemic lupus erythematosus. Arthritis At the time of study, none of our RA patients had Rheum 1982; 25: vasculitis or numbness that was suggestive of sensory 7. Ewing DJ, Clarke BF. Diagnosis and management of diabetic neuropathy. Four lupus patients had cutaneous vas- autonomic neuropathy. Br Med J 1982; 285: culitis and one had residual sensory and motor ulnar 8. Bombardier C, Gladman DD, Urowitz MB, et al. Derivation neuropathy. As we did not do a peripheral nerve of the SLEDAI. A disease activity index for lupus patients. conduction study, it is difficult to rule out the Arthritis Rheum 1992; 35: absence of peripheral neuropathy in these patients. 9. Edmonds ME, Jones TC, Saunders WA, Sturrock RD. Peripheral neuropathy, especially sensory, can affect Autonomic neuropathy in rheumatoid arthritis. Br Med J responses in some of the tests, in particular cold 1979; 2: stimulation. However, this test was not used in 10. Tan J, Akin S, Beyazova M, Sepici V, Tan E. Sympathetic skin response and R-R interval variation in rheumatoid our study. arthritis. Two simple tests for the assessment of autonomic The pathogenesis of the ANS dysfunction in function. Am J Phys Med Rehabil 1993; 72: patients with RA and SLE is not clearly understood. 11. Leden I, Eriksson A, Lilja B, Sturfelt G, Sundkvist G. Vasculitis of the vasa nervorum and secondary amyl- Autonomic nerve function in rheumatoid arthritis of varying oidosis has been proposed.13 The pathogenesis may severity. Scan J Rheumatol 1983; 12: have an immune component. This is supported by 12. Piha SJ, Voipio-Pulkki LM. Elevated resting heart rate in improvement of acute autonomic neuropathy after rheumatoid arthritis: possible role of physical treatment with immunosuppressive drugs in a patient deconditioning. Br J Rheumatol 1993; 32: with SLE.25 The presence of circulating autosystem involvement in rheumatoid arthritis. 50 cases. 13. Toussirot E, Serratrice G, Valentin P. Autonomic nervous antibodies against nerve growth factor, cervical J Rheumatol 1993; 20: ganglia and the vagus nerve has been recently 14. Geenen R, Godaert GLR, Jacobs JWG, Peters ML, Bijlsma demonstrated in RA and SLE patients who had JWJ. Diminished autonomic nervous system responsiveness cardiovascular ANS dysfunction.26,27 The significance in rheumatoid arthritis of recent onset. J Rheumatol 1996; of these auto-antibodies in the pathogenesis of ANS 23: dysfunction remains to be determined. 15. Bekkelund SI, Jorde R, Husby G, Mellgren SI. Autonomic

6 102 W. Louthrenoo et al. nervous system function in rheumatoid arthritis. A function in systemic lupus erythematosus. Lupus 1994; controlled study. J Rheumatol 1996; 23: : Gledhill RF, Dessein PHMC. Autonomic neuropathy in 22. Stein KS, McFarlane IC, Goldberg N, Ginzler EM. Heart rate systemic lupus erythematosus. J Neurol Neurosurg Psychiatr variability in patients with systemic lupus erythematosus. 1988; 51: Lupus 1996; 5: Loite F, Osterland CK. Autonomic neuropathy in systemic 23. Bannister R, Mathias CJ, Matthias C. Autonomic failure: a lupus erythematosus: cardiovascular autonomic function textbook of clinical disorders of the autonomic nervous assessment. Ann Rheum Dis 1994; 53: system, 3rd edn. Oxford, Oxford University Press, Thomas JE, Schirger A, Fealey RD, Sheps SG. Orthostatic 18. Straub RH, Zeuner M, Lock G, et al. Autonomic and hypotension. Mayo Clin Proc 1981; 56: sensorimotor neuropathy in patients with systemic lupus erythematosus and systemic sclerosis. J Rheumatol 1996; 25. Hoyle C, Ewing DL, Parker AC. Acute autonomic 23: neuropathy in association with systemic lupus erythematosus. Ann Rheum Dis 1985; 44: Lagana B, Tubani L, Maffeo N, et al. Heart rate variability and cardiac autonomic function in systemic lupus 26. Dicou E, Hurez D, Nerriere V. Natural autoantibodies erythematosus. Lupus 1996; 5: against the nerve growth factor in autoimmune diseases. J Neuroimmunol 1993; 47: Laversuch CJ, Seo H, Modarres H, Collins DA, McKenna W, 27. Maule S, Quadri R, Mirante D, et al. Autonomic nervous Bourke BE. Reduction in heart rate variability in patients dysfunction in systemic lupus erythematosus (SLE) and with systemic lupus erythematosus. J Rheumatol 1997; rheumatoid arthritis (RA): possible pathogenic role of 24: autoantibodies to autonomic nervous structures. Clin Exp 21. Omdal R, Jorde R, Mellgren SI, Husby G. Autonomic Immunol 1997; 110:423 7.