doi: /brain/awm174 Brain (2007), 130,2425^2432 RETRACTED

doi:10.1093/brain/awm174 Brain (2007), 130,2425^2432 Characteristics of orthostatic hypotension in Parkinson s disease Hisayoshi Oka, Masayuki Yoshioka, Kenji Onouchi, Masayo Morita, Soichiro Mochio, Masahiko Suzuki, Toshiaki Hirai, Yasuhiko Ito and Kiyoharu Inoue Department of Neurology, Jikei University School of Medicine, 3-25- 8, Nishi-shinbashi, Minato-ku, Tokyo, Japan Correspondence to: Hisayoshi Oka, MD, Department of Neurology, Jikei University School of Medicine, 3-25- 8, Nishi-shinbashi, Minato-ku, Tokyo, 105-8461, Japan E-mail: h.oka@jikei.ac.jp Clinical symptoms of Parkinson s disease (PD) include not only motor distress but also autonomic dysfunction. Orthostatic hypotension (OH) occurs in one-fifth to one-half of all patients with PD. We examined the relation of this type of hypotension to clinical features and cardiovascular parameters such as cardiac 123 I-meta-iodobenzylguanidine (MIBG) uptake, changes on the Valsalva maneuver, and plasma norepinephrine concentrations on head-up tilt-table testing (HUT). We performed HUT in 55 patients with PD and divided them into two groups according to the presence or absence of OH, defined as a drop in systolic blood pressure (SBP mmhg) by 20 mmhg or more on standing. We evaluated cardiac sympathetic function by 123 I-MIBG scintigraphy and assessed cardiovascular autonomic function by using the Valsalva maneuver in all subjects. We also performed HUT, 123 I-MIBG scintigraphy and assessed cardiovascular autonomic function by using the Valsalva maneuver in 20 controls.the results of HUTshowed that 20 patients had OH and 35 did not. The hypotension was associated with gender, older age, longer disease duration, posture and gait instability phenotype, low mini-mental state examination scores and visual hallucinations. Cardiac 123 I-MIBG uptakes were lower in patients with OH. SBP fell further during early second phase in patients with OH than in patients without the condition and their increase in SBP during the late second phase and the overshoot of SBP during the fourth phase were lower. The blood pressure recovery time during the fourth phase on the Valsalva maneuver was longer in patients with OH than in those without OH. There was, however, no association between the fall in SBP on HUTand baroreflex sensitivity or the plasma norepinephrine concentrations, adjusted by age, disease duration, disease severity and dopaminergic medication using multiple regression analyses. Patients without OH already had impaired cardiac sympathetic and baroreceptor reflex functions as early abnormalities of cardiovascular autonomic control. Our results suggest that pronounced vasomotor and cardiac sympathetic dysfunction is the primary cause of OH in PD, although baroreceptor reflex failure may also make a minor contribution. It was unclear whether vasomotor and cardiac sympathetic dysfunction in patients with PD was caused primarily by the impairment of preganglionic or postganglionic lesions. Keywords: Parkinson s disease; cardiovascular sympathetic dysautonomia; cardiac radioiodinated metaiodobenzylguanidine ( 123 I-MIBG) scintigraphy, Valsalva maneuver; baroreflex sensitivity Abbreviations: HUT = head-up tilt-table testing; MIBG = 123 I-meta-iodobenzylguanidine; OH = orthostatic hypotension; PD = Parkinson s disease; SBP = systolic blood pressure Received March 6, 2007. Revised June 2, 2007. Accepted July 5, 2007. Advance Access publication August 2, 2007 Introduction The main motor symptoms of Parkinson s disease (PD) are resting tremor, rigidity, bradykinesia and gait disturbance. PD is also associated with autonomic dysfunction, characterized by signs and symptoms such as constipation, sweating disturbance and orthostatic and postprandial hypotension in progressive disease (Rajput et al., 1984). Autonomic dysfunction in PD is milder than that in multiple system atrophy (Sandroni et al., 1991a, b; Colosimo et al., 1995; Magalhaes et al., 1995). However, sympathetic noradrenergic dysfunction is clinically important, because orthostatic hypotension (OH) occurs in 20 50% of patients with PD, potentially affecting their daily activities and quality of life (Micieli et al., 1987; Turkka, 1987; Hillen et al., 1996; Senard et al., 1997, 2001; Goldstein et al., 2003; Allcock et al., 2004; Goldstein, 2006). ß The Author (2007). Published by Oxford University Press on behalf ofthe Guarantors of Brain. Allrights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

2426 Brain (2007), 130,2425^2432 H.Okaet al. Previous studies have found that OH is associated with advancing age, male sex and disease duration or severity (Piha et al., 1988; Sandyk and Awerbuch, 1992; Martin et al., 1993; Magalhaes et al., 1998). OH also correlates with the clinical subtype of motor symptoms in PD (Haapaniemi et al., 2001; Devos et al., 2003). Whether dopaminergic medication influences OH remains controversial. The loss of postganglionic sympathetic nervous fibres involved in cardiac and peripheral sympathetic innervation may cause OH in PD, because norepinephrine concentrations are lower in patients with OH than in those without OH (Senard et al., 1990, 1993; Galinier et al., 1994; Niimi et al., 1999; Goldstein, 2003). Baroreflex cardiovagal, sympathoneural or peripheral postganglionic dysfunction may cause OH in PD, because the response of RR intervals to increases or decreases in blood pressure is low, blood pressure decreases progressively during phase II, and blood pressure does not exceed the baseline value during phase IV on the Valsalva maneuver (Goldstein et al., 2005). However, it remains unclear which of type of dysfunction is primarily responsible for OH in PD. The aims of this study were (i) to examine differences in clinical features and phenotypes between PD with and that without orthostatic hypotension and (ii) to clarify the pathogenesis of orthostatic hypotension in PD on the basis of 123 I-metaiodobenzylguanidine (MIBG) uptake by the heart, hemodynamic autonomic testing on the Valsalva maneuver, and the response of plasma norepinephrine concentrations to head-up tilt-table testing (HUT). Methods Subjects One-hundred eight patients meeting the diagnostic criteria for PD proposed by Calne et al. (1992) were entered into a research database at Aoto Hospital, Jikei University School of Medicine. Fifty-five of these patients were recruited into this study, after excluding patients who were not able to perform HUT or the Valsalva maneuver because of motor impairment, cardiac disorders or other reasons. The subjects were 32 women and 23 men with a mean age of 68.1 4.6 years (range, 56 80 years) and a disease duration of 1 12 years (mean, 4.4 years). No patient had abnormal findings on magnetic resonance imaging, including evidence of brain ischaemia, cortical and brainstem atrophy, ventricular enlargement or cerebellar atrophy. All patients received levodopa or a dopa agonist and had good responses. No patient had signs or symptoms of cardiac disease or any abnormalities on chest radiography, electrocardiography or cardiac echography. We also excluded patients who were receiving medications with potential effects on autonomic function, such as beta-blockers, anticholinergic agents or antihypertensive drugs. No patient had previously received drugs with potential effects on 123 I-MIBG uptake at sympathetic nerve terminals (Hesse et al., 2005). Levodopa was given to 39 patients, ergot dopamine agonists to 22 and non-ergot dopamine agonists to 11. Treatment was continued during the study. Daily levodopa-equivalent-unit (LEU) doses were calculated on the basis of the theoretical equivalence to levodopa, i.e. ergot dopamine agonists such as cabergoline (mg) 67, pergolide (mg) 100, to facilitate direct comparison of dopaminergic medication between patients with and those without OH (Hobson et al., 1999; Deep Brain Stimulation for Parkinson s Disease Study Group, 2001). No patient was receiving bromocriptine. The only non-ergot dopamine agonist used in this study was pramipexole. We did not change the dosage or type of dopaminergic medication during the study. The severity of PD was assessed using the motor subsection of the Unified Parkinson s Disease Rating Scale (UPDRS motor score). Subjects were determined to have the tremor phenotype (TDT), posture and gait instability phenotype (PIGD) or indeterminate phenotype (INT) by calculating the ratio of the global tremor score to the global posture and gait score, derived from the motor subsection of the UPDRS score (Fahn et al., 1987). Subjects with a ratio of less than 1.0 were considered to have a PIGD phenotype, and those with a ratio of 1.0 or more were considered to have a TDT phenotype or INT phenotype. A Mini-Mental State Examination (MMSE) (Folstein et al., 1975) was performed in all patients. Descriptions of visual hallucinations (VH) were derived from clinical interviews with the patients and their caregivers (Holroyd et al., 1995). Patients with VH had to have experienced VH at least several times previously, irrespective of the relation of VH to medication. We excluded patients who had an MMSE score of less than 20, because it was unlikely they could adequately comprehend and respond to the interview questions (Oishi et al., 2005). As controls, 20 age-matched controls (12 men, 8 women, age 68.7 2.8 years, range 61 74 years) with no neurological disorders were studied. None of the controls had clinically significant illnesses potentially affecting the cardiac autonomic nervous system. Head-up tilt-table testing (HUT) All subjects underwent HUT in a silent room, maintained at an ambient temperature of 23 to 26 C. Most subjects were studied in the morning with no medication after they had fasted overnight, except for non-caloric liquids. The subject was tilted to a 60 upright position within 15 s by means of a head-up tilt table. Subjects were defined as having OH if the standing systolic blood pressure fell by 20 mmhg or more (Kaufmann, 1996). Systolic blood pressure (SBP), diastolic blood pressure (DBP), RR intervals and plasma norepinephrine concentrations (NE, pg/ml) were measured after 20 min of rest in the supine position and after 10 min in a tilted position on a tilt table. Venous blood was drawn through an indwelling catheter. Plasma NE concentrations were assayed by high-performance liquid chromatography according to methods validated by SRL Inc. (Oka et al., 2006a, b). Differences in the plasma NE concentrations between the two positions (ine) were also calculated. Cardiac 123 I-MIBG scintigraphy The subjects were given an intravenous injection of 111 MBq 123 I-MIBG (Daiichi Radioisotope Laboratories Co., Tokyo, Japan). Relative organ uptake of 123 I-MIBG was determined by regionof-interest (ROI) analysis in the anterior view. The ratio of the average pixel count in the heart (H) to that in the mediastinum (M) was calculated (H/M ratio) after 15 min (early) and after 3 h (delayed).

OH in Parkinson s disease Brain (2007), 130, 2425^2432 2427 Valsalva maneuver The Valsalva maneuver was done by having the subjects exhale into a mouthpiece at an expiratory pressure of 40 mmhg for 15 s. Blood pressure and RR intervals were measured during the Valsalva maneuver by tonometry, using a non-invasive blood pressure monitoring system (CBM3000, Nihon Colin Co., Ltd, Komaki, Japan). The results were analysed as described previously (Oka et al., 2006a, b, 2007a, b, c). SBP decreases in early phase II because of reduced cardiac output, in turn decreasing venous return and stroke volume, despite tachycardia caused by the withdrawal of cardiovagal control (phase II_E, mmhg). The decrease in SBP is arrested within at least 8 s. Late phase II is associated with an increase in blood pressure, reflecting the activation of vasomotor sympathetic nerves (phase II_L, mmhg). A transient fall in BP (phase III), lasting 1 to 2 s, occurs at the end of the Valsalva maneuver because of sudden drops in intrathoracic and abdominal pressures (Korner et al., 1976). Phase IV is the overshoot of blood pressure (phase IV, mmhg) due to the activation of cardiac sympathetic nerves (Sandroni et al., 1991a, b). Time intervals were determined for two periods of the Valsalva maneuver, from the end of phase III to the complete return of SBP to the baseline value (blood pressure recovery time: PRT, s) (Vogel et al., 2005). Baroreceptor reflex sensitivity (BRS, ms/mmhg) was derived from the correlation of RR with SBP during the early second and fourth phases of the Valsalva maneuver (Vogel et al., 2005; Oka et al., 2006a, b, 2007a, b, c). Statistical analyses Statistical analyses were performed using a statistical data analysis system (Esumi Co., Ltd, Tokyo, Japan). Pairwise comparisons were made using 2 tests for binary variables such as gender, motor phenotype, VH and the number of medications. Differences between groups were compared with the use of Welch s t-test for continuous variables. Significant differences among controls, PD with OH and PD without OH were determined by the two-tailed multiple t-tests with Bonferroni correction following analysis of variance (ANOVA). Relations between MMSE score or VH and the fall in SBP were evaluated by multiple logistic regression analysis, adjusted by risk factors such as age, disease duration, UPDRS motor score and dopaminergic medication. Correlations of the fall in SBP on HUT with cardiac 123 I-MIBG uptake and various indices of autonomic activity, such as the responses on the VM and the changes in NE on HUT, were evaluated by multiple regression analysis, adjusted by risk factors such as age, disease duration, UPDRS motor score and dopaminergic medication. P-values of less than 0.05 were considered to indicate statistical significance. The ethics committee of Jikei University School of Medicine approved all procedures. Results Clinical characteristics of patients with and without OH Twenty patients had OH, and thirty-five did not. A few patients experienced a floating sensation, but complaints during tilt table testing were minimal. Patients with OH were likely to be male and those without OH were likely to be female. Many patients with OH had PIGD phenotype. Table 1 Clinical characteristics of patients who had PD with or without orthostatic hypotension PD without OH PD with OH P value Number 35 20 Sex (M:F) 11/24 12/8 0.039 Age (range) Years 67.0 4.1 70.1 5.2 0.029 (55^75) (58^80) Symptom duration 3.9 1.4 5.6 2.1 0.025 (range) years (1^10) (2^12) UPDRS motor 23.1 5.3 26.5 5.9 NS (15^ 45) (18 ^ 45) MMSE (range) 27.4 1.9 24.7 3.2 0.001 (24 ^30) (20 ^30) Subtype (PIGD/ 15/20 15/5 0.021 TDT or INT) VH 7 15 0.002 Levodopa(number) 21 18 NS (mg/day) 276 48 274 62 NS (100 ^ 400) (100^400) Ergot dopamine 12 10 NS agonist(number) LEU (mg/day) 151 67 183 85 NS (67^335) (75^393) Non-ergot dopamine 7 4 NS agonist (number) (mg/day) 1.6 0.6 (0.5^3) 1.9 0.4 (1.5^2.5) NS Note: PD = Parkinson s disease; OH = orthostatic hypotension; UPDRS motor = motor subsection of Unified Parkinson s Disease Rating Scale; MMSE = mini-mental state examination. Tremor phenotype (TDT) or indeterminate phenotype (INT) and PIGD (posture and gait instability) phenotype were determined by calculating the ratio of the global tremor score to the global posture and gait score; derived from the motor subsection of the UPDRS score. Subjects with a ratio of less than 1.0 were defined as PIGD phenotype; and those with a ratio of 1.0 or more were defined as TDT or INT phenotype. VH = visual hallucinations; LEU = daily levodopa-equivalent-unit dose, calculated on the basis of the theoretical equivalence to levodopa, i.e. cabergoline (mg) 67; pergolide (mg) 100. NS = not significant; data = means SD. Patients with OH had a significantly longer duration of disease than patients without OH. VH was more frequent among patients with OH than among those without OH. The MMSE score was lower in patients with OH than in those without OH. There was no significant difference in the UPDRS motor score or dopaminergic medication between patients with and those without OH (Table 1). Baseline hemodynamic data, cardiac 123 I-MIBG scintigraphy, autonomic parameters on the Valsalva maneuver and HUT There were no significant differences in SBP, DBP or RR interval at rest among the controls, patients with and those without OH. SBP and DBP were decreased in patients with OH, while there was no difference among the groups was in

2428 Brain (2007),130,2425^2432 H.Okaet al. Table 2 Comparisons of cardiac 123 I-MIBG uptake; cardiovascular autonomic parameters on the Valsalva maneuver; and head-up tilt-table testing in controls; PD patients with and those without orthostatic hypotension Controls (n =20) PD (n =55) OH(^) (n = 35) OH(+)(n =20) P value SBP at rest (mmhg) 125 11 125 13 137 20 NS DBP at rest (mmhg) 75 9 70 6 7310 NS RR interval at rest (s) 0.83 0.09 0.88 0.12 0.89 0.10 NS SBP after tilt (mmhg) 129 15 121 13 110 17 a 0.001 DBP after tilt (mmhg) 82 11 71 8 b 62 11 c 0.001 RR interval after tilt (s) 0.73 0.09 0.83 0.11 d 0.89 0.11 e 0.015 Early H/M 2.55 0.21 1.78 0.30 d 1.53 0.23 c 0.001 Delay H/M 2.46 0.26 1.63 0.35 d 1.33 0.18 f 0.001 phase II E (mmhg) 19.7 10.2 21.5 9.0 33.4 14.0 g 0.003 phase II L (mmhg) 13.6 5.7 9.1 8.1 ^1.9 6.2 c 0.001 BRS II (ms/mmhg) 4.4 2.3 2.3 1.3 d 1.8 1.1 d 0.003 BRS IV (ms/mmhg) 6.7 2.3 4.4 2.3 e 2.5 0.9 f 0.001 phase IV (mmhg) 20.1 8.4 13.9 7.7 b 3.2 6.7 c 0.001 PRT (s) 2.0 1.0 3.9 2.4 11.5 6.1 f 0.001 NE at rest (pg/ml) 232 69 229 103 173 63 NS NE after tilt (pg/ml) 384 100 354 141 247 76 h 0.009 NE (pg/ml) 152 42 125 51 74 27 c 0.0001 Note: PD = Parkinson s disease; OH = orthostatic hypotension; SBP = systolic blood pressure; DBP = diastolic blood pressure; H/M = the ratio of the average pixel count in the heart (H) to that in the mediastinum (M); Phase II E = SBP decrease during early phase II; phase II L = SBP decrease during late phase II; BRS II and BRS IV = baroreflex sensitivity in phase II and IV; Phase IV = the overshoot of SBP during the phase IV; PRT = blood pressure recovery time from the end of phase III to the complete return of SBP to the baseline value in phase IV; NE = norepinephrine concentration; NE = the difference in plasma NE concentrations between the supine and upright positions; NS = not significant; data = means SD. a P50.001 as compared with controls; P50.01 as compared with patients without OH. b P50.005 as compared with controls. c P50.001 as compared with controls; P50.005 as compared with patients without OH. d P50.001 as compared with controls. e P50.02 as compared with controls. f P50.001 as compared with controls; P50.001 as compared with patients without OH. g P50.005 as compared with controls; P50.005 as compared with patients without OH. h P50.005 as compared with controls; P50.02 as compared with patients without OH. RR interval. Patients without OH had a lower DBP as compared with the controls. Early and delayed H/M ratios were lower in patients without OH and those with OH than in the control group. Early and delayed H/M ratios in patients with OH were lower than those in patients without OH. Phase II_E, II_L and PRT in patients without OH did not differ from the control value. Phase II_E and PRT in patients with OH were larger than those in patients without OH and the controls. Phase II_L in patients with OH was smaller than that in patients without OH and the controls. BRSs and phase IV in patients without OH were significantly lower than those in the controls. BRS IV and phase IV in patients with OH were significantly lower than those in patients without OH, whereas there was no significant difference in BRS II between patients with and without OH. Plasma NE concentration in the supine position was slightly but not significantly lower in patients with OH than in those without OH. Plasma NE concentration in the upright position and the difference in the plasma NE concentrations between the two positions (NE) were significantly less in patients with OH than in those without OH (Table 2). Multiple regression analyses between the fall insbponhutandmmsescore After adjustment for age, disease duration, the UPDRS motor score and dopaminergic medication on multiple regression analyses, the fall in SBP on HUT was significantly associated with the decrease in MMSE score (Table 3). Multiple logistic regression analyses of the fall in SBP on HUTand motor phenotype or VH Patients with PIGD phenotype or VH had a greater decrease in SBP on HUT than did patients with TDT phenotype or patients without VH, adjusted by disease duration, the UPDRS motor score and dopaminergic medication on multiple logistic regression analyses (Tables 4 and 5).

OH in Parkinson s disease Brain (2007), 130, 2425^2432 2429 Table 3 Multiple regression analysis of the correlation between the fall in SBP on HUT and MMSE Table 5 Multiple logistic regression analyses of the correlation between VH and the fall in SBP on HUT Fall in SBP on HUT Parameter estimate MMSE ^1.647 ^ 0.371 0.024 Age 0.281 0.127 NS Duration 0.920 0.150 NS UPDRS motor score 0.015 0.008 NS Levodopa 0.007 0.068 NS Ergot dopamine agonist (LEU) 0.032 0.234 NS Non-ergot dopamine agonist 1.106 0.061 NS Note: SBP = systolic blood pressure; HUT = head-up tilt-table resting; MMSE = mini-mental state examination; UPDRS = motor subsection of the Unified Parkinson s Disease Rating Scale; LED = daily levodopa-equivalent-unit dose; SE = Standard Error; NS = not significant. Table 4 Multiple logistic regression analysis of the correlation between PIGD phenotype and the fall in SBP on HUT PIGD phenotype OR (95% CI) a P Fall in SBP on HUT 1.07 (1.12^1.14) 0.012 Age 1.06 (0.94 ^1.19) NS Duration 0.70 (0.48 ^1.01) NS UPDRS motor score 0.99 (0.90 ^1.09) NS Levodopa 1.00 (1.00 ^1.01) NS Ergot dopamine agonist (LEU) 1.00 (0.99^1.01) NS Non-ergot dopamine agonist 0.84 (0.35^2.01) NS Note: OR = odds ratio; CI = confidence interval. PIGD phenotype = posture and gait instability phenotype; SBP = systolic blood pressure, HUT = head-up tilt-table testing; LED = daily levodopaequivalent-unit dose; NS = not significant. a Risks were adjusted by age; disease duration; motor subsection of the Unified Parkinson s disease rating scale (UPDRS motor score) and dopaminergic medication. Multiple regression analyses of the fall in SBP on HUTand cardiac 123 I-MIBG scintigraphy or autonomic parameters on the Valsalva maneuver and HUT After adjustment for age, disease duration, the motor subsection of the UPDRS and dopaminergic medication on multiple regression analyses, the fall in SBP on HUT was associated with a delayed H/M ratio of cardiac 123 I-MIBG uptake, larger phase II_E, smaller phase II_L and phase IV and longer PRT. The early H/M ratio, baroreflex sensitivity in phase II or IV, and plasma NE concentrations did not correlate with the fall in SBP on HUT (Table 6). Discussion In this study, OH in patients with PD was significantly associated with sex, age, disease duration, disease severity, PIGD motor phenotype, MMSE score and VH, but not SE P VH OR (95% CI) a P Fall in SBP on HUT 1.09 (1.02^1.17) 0.015 Age 1.02 (0.91^1.15) NS Duration 0.96 (0.66^1.42) NS UPDRS motor score 1.12 (0.98 ^1.27) NS Levodopa 1.01 (1.00 ^1.13) 0.029 Ergotdopamineagonist(LEU) 1.00(0.29^1.01) NS Non-ergot dopamine agonist 0.83 (0.29^1.17) NS Note: OR = odds ratio; CI = confidence interval; VH = visual hallucinations, SBP = systolic blood pressure; HUT = head-up tilt-table testing; LED = daily levodopa-equivalent-unit dose; NS = not significant. a Risks were adjusted by age; disease duration; motor subsection of the Unified Parkinson s disease rating scale (UPDRS motor score) and dopaminergic medication. Table 6 Multiple regression analysis of the correlation of the fall in SBP on HUT with cardiac 123 I-MIBG uptake, cardiovascular autonomic parameters on the Valsalva maneuver and NE concentration on head-up tilt-table testing Fall in SBP on HUT Parameter estimate SE P Early H/M ^7.840 ^ 0.209 NS Delay H/M ^10.550 ^ 0.312 0.035 phase II E (mmhg) 0.219 0.241 NS phase II L (mmhg) ^0.687 ^0.527 0.001 BRSII (ms/mmhg) ^0.363 ^0.048 NS BRS IV (ms/mmhg) ^1.436 ^ 0.245 NS phase IV (mmhg) ^ 0.450 ^ 0.371 0.007 PRT (s) 0.843 0.401 0.011 NE at rest (pg/ml) ^ 0.015 ^ 0.128 NS NE after tilt (pg/ml) ^0.015 ^0.177 NS NE (pg/ml) ^ 0.052 ^ 0.222 NS Note: H/M = the ratio of the average pixel count in the heart (H) to that in the mediastinum (M); Phase II E = SBP decrease during early phase II; phase II L = SBP decrease during late phase II; BRS II and BRSIV=baroreflex sensitivityinphaseii and IV; phaseiv=the overshoot of SBP during the phase IV; PRT =blood pressure recovery time from the end of phase III to the complete return of SBP to the baseline value in phase IV; NE = norepinephrine concentration; NE = the difference in plasma NE concentrations between the supine and upright positions; SE = standard Error; NS = not significant; data = means SD. with dopaminergic medication. Patients with OH were likely to be male and those without OH were likely to be female. Previous studies showed that OH was associated with increasing age and male sex, as well as with disease duration or severity (Piha et al., 1988; Sandyk and Awerbuch, 1992; Martin et al., 1993; Magalhaes et al., 1998). Some investigators have reported that the frequency of OH in PD suggests that OH is occasionally already present around the time of onset of movement disorders (Turkka, 1987; Senard et al., 1997; Goldstein et al., 2005; Allcock et al., 2006). Men with PD are considered to have a more rapid decline in motor functions than women.

2430 Brain (2007),130, 2425^2432 H. Oka et al. Patients who had PD with OH had a longer duration of disease than those without OH in our study. Orthostatic parameters were reported to be more sensitive to the duration rather than the severity of disease (Mesec et al., 1993). These findings may be related to treatment with higher doses of dopaminergic medications, in turn possibly accounting for the orthostatic drop in blood pressure (Jankovic and Kapadia, 2001). However, Goldstein et al. reported that OH was unrelated to levodopa treatment (Goldstein et al., 2005). In our study, OH was not associated with dopaminergic medication, including levodopa as well as ergot and non-ergot dopamine agonists. Our study demonstrated a significant association between VH and OH. Autonomic dysfunction has been reported to be more strongly related to PD with VH than without VH (Williams and Lees, 2005). The cardiovascular system is thought to be regulated by centres responsible for autonomic nervous control, such as the amygdala, hippocampal or paraventricular region and the frontobasal cortex (Mathias and Bannister, 1999). Lesions in these centres are apparently associated with VH (Harding et al., 2002). We have previously reported that cardiac and vasomotor sympathetic dysfunction is more severe in patients with VH than without VH (Oka et al., 2007b). These dysfunctions are probably attributed to Lewy-body lesions or neuronal loss in sympathetic ganglia, the central autonomic system, or both (Oka et al., 2007a). Our findings indicated that OH was associated with PIGD motor phenotype, but not with tremor motor phenotype. Some studies have found that the severity of autonomic deficits correlates with the clinical type of motor symptoms in PD. A significant negative correlation has been found between heart-rate variability and the severity of limb and axial bradykinesia, but not tremor (Haapaniemi et al., 2001). Sympathetic impairment on autonomic function tests has been considered more pronounced in akinetic type patients than in tremor type patients (Mesec et al., 1993; Saiki et al., 2004). Our study also showed that patients with OH had a lower MMSE score than those without OH. Dementia often complicates the course of PD. Age and disease severity have been associated with dementia in some studies (Hughes et al., 2000; Levy et al., 2002; Marder et al., 2002). However, the association of OH with dementia in PD remains unclear (Idiaquez et al., 2007). Diffuse or transitional Lewy body disease has been considered the primary pathological substrate for the subsequent development of dementia in PD. Lewy bodies in patients with OH may be widely distributed, not only in the brain but also in peripheral sympathetic ganglia. The differences between PD and dementia with Lewy bodies remain controversial (Oka et al., 2007a, b, c). We also examined the association of OH with autonomic parameters such as cardiac 123 I-MIBG uptake, changes on the Valsalva maneuver and plasma NE concentrations on HUT. Even patients without OH had lower cardiac MIBG uptakes, BRSs in phase II and phase IV, and overshoot of SBP in phase IV as compared with the control values. Furthermore, patients with OH had reduced 123 I-MIBG uptake as compared with patients without OH. Vasomotor and cardiac sympathetic functions on the Valsalva maneuver were more impaired in patients with OH than in patients without OH. NE responses on standing were decreased in patients with OH as compared with patients without OH. OH in PD has been extensively studied. Baroreflex sensitivity has been reported to be variably decreased in PD (Szili-Torok et al., 2001; Oka et al., 2003). In PD without OH, the baroreflex-cardiovagal gain is significantly decreased as compared with normal values, but in PD with OH the baroreflex-cardiovagal gain is universally very low. Baroreflex sensitivity is regulated mainly by parasympathetic activity (James et al., 1996; Kautzner et al., 1996). A reduced baroreflex-cardiovagal gain might indicate dysfunction of the reflex arc consisting of the afferent glossopharyngeal or vagal nerve, solitary nucleus and ambiguus nucleus of the medulla and the efferent vagal nerve to the sinus node in the heart (Joyner and Shepherd, 1997). Baroreflex sensitivity may decrease if lesions impair any part of this circuit. In this study, the decline in baroreflex sensitivity was slight but statistically significant in patients with OH. However, multiple regression analyses after adjustment for age, disease duration, the UPDRS motor score and dopaminergic medication showed no significant correlation between OH and baroreflex sensitivity. Cardiac sympathetic denervation in itself is not thought to cause OH (Robertson et al., 1993a, b). These findings suggest that baroreflex sensitivity does not contribute to OH in PD. Multiple regression analyses also demonstrated that OH significantly correlated with the delayed H/M ratio of cardiac 123 I-MIBG uptake, but not with the early H/M ratio. Perhaps cardiac sympathetic denervation contributes slightly to OH in patients with PD, although Goldstein has reported that patients with cardiac transplants do not have persistent OH (Goldstein, 2003). These facts suggest that failure of baroreflex-cardiovagal gain does not contribute to OH, whereas cardiac sympathoneural failure may be involved. In our study, OH in PD significantly correlated with the decrease in SBP during the early phase II (phase II_E), the increase in SBP during the late phase II (phase II_L), the overshoot of SBP during phase IV (phase IV) and blood pressure recovery time in phase IV (PRT). Early phase II (phase II_E) is characterized by a fall in BP due to reduced cardiac output and shortening of RR intervals caused by the attenuation of cardiovagal control. SBP elevation during the late second phase (phase II_L) indicates vasomotor function, and that during the fourth phase (phase IV) indicates cardiac sympathetic function (mainly cardiac muscle contraction) (Sandroni et al., 1991a, b). PRT reflects the adrenergic component of baroreflex function, accompanied by muscle sympathetic

OH in Parkinson s disease Brain (2007), 130,2425^2432 2431 neural activation with norepinephrine release and binding to -adrenergic receptors, resulting in vasoconstriction and BP recovery (Vogel et al., 2005; Oka et al., 2007a, b, in press). Therefore, OH is apparently caused by impairment of baroreflex-cardiac and vasomotor sympathoneural gain. Many investigators have reported that patients with PD and OH have lower plasma NE concentrations than patients without OH (Turkka, 1987; Senard et al., 1990, 1993; Galinier et al., 1994; Niimi et al., 1999; Goldstein et al., 2003). In our study, the plasma NE concentration in the supine position was slightly but not significantly lower in patients with OH than those without OH. Nonetheless, the plasma NE concentration in upright position and the difference in plasma NE concentrations between the upright and supine positions were significantly less in patients with OH than in those without OH. However, after adjustment for age, disease duration, UPDRS motor score and dopaminergic medication on multiple regression analyses, there was no significant correlation between OH and the plasma NE concentrations. Perhaps proportional increases in NE release from reduced vesicular stores resulting from partial loss of sympathetic fibers led to the non-significant difference in plasma NE concentrations between patients with and those without OH in this study. Measurement of plasma NE concentrations might fail to detect real decreases in NE release, because denervation would concurrently decrease both the release and reuptake of noradrenaline (Goldstein, 2003). Decreased plasma NE concentrations have been suggest to indicate sympathetic dysfunction involving peripheral sympathetic innervation (Turkka, 1987; Senard et al., 1993; Goldstein et al., 2003). Unfortunately, we could not determine whether OH in patients with PD is caused by preganglionic or postganglionic dysfunction on multiple regression analyses adjusted by risk factors. We speculate that no apparent association between plasma NE concentrations and the fall in SBP on HUT may be attributed to the effects of dopaminergic medication or denervation supersensitivity of postganglionic sympathetic terminals; however, supporting evidence is lacking. Our study also showed that even patients without OH had lower cardiac MIBG uptakes (Amino et al., 2005; Orimo et al., 2005), BRSs and overshoot of SBP in phase IV as compared with the control values. Even patients who have PD without OH can show certain degrees of alteration in cardiovascular autonomic control in response to gravitational stimuli (Barbic et al., 2007). Our results suggested that cardiac sympathetic and baroreceptor reflex functions were impaired even in patients without OH as an early abnormality of cardiovascular autonomic control. Our results suggest that pronounced vasomotor and cardiac sympathetic dysfunction might cause OH in patients with PD. Baroreceptor reflex failure may also partially and minimally contribute to OH. Patients without OH had already impaired cardiac sympathetic and baroreceptor reflex functions as early abnormalities of cardiovascular autonomic control. It was unclear whether vasomotor and cardiac sympathetic dysfunction in patients with PD was caused primarily by the impairment due to preganglionic or postganglionic lesions. References Allcock LM, Ullyart K, Kenny RA, Burn DJ. Frequency of orthostatic hypotension in a community based cohort of patients with Parkinson s disease. J Neurol Neurosurg Psychiatry 2004; 75: 1470 1. Allcock LM, Kenny RA, Burn DJ. Clinical phenotype of subjects with Parkinson s disease and orthostatic hypotension: Autonomic symptom and demographic comparison. Mov Disord 2006; 21: 1851 5. Amino T, Orimo S, Itoh Y, Takahashi A, Uchihara T, Mizusawa H. Profound cardiac sympathetic denervation occurs in Parkinson disease. Brain Pathol 2005; 15: 29 34. Barbic F, Perego F, Canesi M, Gianni M, Biagiotti S, Costantino G, et al. Early abnormalities of vascular and cardiac autonomic control in Parkinson s disease without orthostatic hypotension. Hypertension 2007; 49: 120 6. Calne DB, Snow BJ, Lee C. Criteria for diagnosing Parkinson s disease. Ann Neurol 1992; 32 (Suppl): 125 7. Colosimo C, Albanese A, Hughes AJ, de Bruin VM, Lees AJ. Some specific clinical features differentiate multiple system atrophy (striatonigral variety) from Parkinson s disease. Arch Neurol 1995; 52: 294 8. Deep Brain Stimulation for Parkinson s Disease Study Group.Deep-brain stimulation of the subthalamic nucleus or the pars interna of the globus pallidus in Parkinson s disease. N Engl J Med 2001; 345: 956 63. Devos D, Kroumova M, Bordet R, Vodougnon H, Guieu JD, Libersa C, et al. Heart rate variability and Parkinson s disease severity. J Neural Transm 2003; 110: 997 1011. Fahn S, Elton RL. Members of the UPDRS committee: Unified Parkinson s Disease Rating Scale. In: Fahn S, Marsden CD, Calne DB, Goldstein M, editors. Recent developments in Parkinson s disease. Florham Park, NJ: Macmillan Healthcare information; 1987. pp. 153 63. Folstein MF, Folstein SE, McHugh PR. Mini-mental state. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975; 12: 189 98. Galinier M, Senard JM, Valet P, Doazan JP, Durrieu G, Tran MA, et al. Relationship between arterial blood pressure disturbances and alpha adrenoceptor density. Clin Exp Hypertens 1994; 16: 373 89. Goldstein DS, Pechnik S, Holmes C, Eldadah B, Sharabi Y. Association between supine hypertension and orthostatic hypotension in autonomic failure. Hypertension 2003; 42: 136 42. Goldstein DS. Dysautonomia in Parkinson s disease: neurocardiological abnormalities. Lancet Neurol 2003; 2: 669 76. Goldstein DS, Eldadah BA, Holmes C, Pechnik S, Moak J, Saleem A, et al. Neurocirculatory abnormalities in Parkinson disease with orthostatic hypotension: independence from levodopa treatment. Hypertension 2005; 46: 1333 9. Goldstein DS. Orthostatic hypotension as an early finding in Parkinson s disease. Clin Auton Res 2006; 16: 46 54. Haapaniemi TH, Pursiainen V, Korpelainen JT, Huikuri HV, Sotaniemi KA, Mylylla VV. Ambulatory ECG and analysis of heart rate variability in Parkinson s disease. J Neurol Neurosurg Psychiatry 2001; 70: 305 10. Harding AJ, Broe GA, Halliday GM. Visual hallucinations in Lewy body disease relate to Lewy bodies in the temporal lobe. Brain 2002; 125: 391 403. Hesse B, Tägil K, Cuocolo A, Anagnostopoulos C, Bardiés M, Bax J, et al. EANM/ESC procedural guidelines for myocardial perfusion imaging in nuclear cardiology. Eur J Nucl Med Mol Imaging 2005; 32: 855 97. Hillen ME, Wagner ML, Sage JI. Subclinical orthostatic hypotension is associated with dizziness in elderly patients with Parkinson disease. Arch Phys Med Rehabil 1996; 77: 710 2.

2432 Brain (2007), 130,2425^2432 H.Okaet al. Hobson DE, Pourcher E, Martin WR. Ropinirole and pramipexole, the new agonists. Can J Neurol Sci 1999; 26 (Suppl 2): S27 33. Holroyd S, Keller AS. A study of visual hallucinations in Alzheimer s disease. Am J Geriatr Psychiatry 1995; 3: 198 205. Hughes TA, Ross HF, Musa S, Bhattacherjee S, Nathan RN, Mindham RH, et al. A 10-year study of the incidence of and factors predicting dementia in Parkinson s disease. Neurology 2000; 54: 1596 602. Idiaquez J, Benarroch EE, Rosales H, Milla P, Rios L. Autonomic and cognitive dysfunction in Parkinson s disease. Clin Auton Res 2007; 17: 93 8. James MA, Robinson TG, Panerai RB, Potter JF. Arterial baroreceptorcardiac reflex sensitivity in the elderly. Hypertension 1996; 128: 953 60. Jankovic J, Kapadia AS. Functional decline in Parkinson disease. Arch Neurol 2001; 58: 1611 5. Joyner MJ, Shepherd JT. Autonomic regulation of circulation. In: Low PA, editor. Clinical autonomic disorders. 2nd edn. Philadelphia: Lippincott- Raven Publishers; 1997. pp. 61 71. Kaufmann H. Consensus statement on the definition of orthostatic hypotension, pure autonomic failure and multiple system atrophy. Clin Auton Res 1996; 6: 125 6. Kautzner J, Hartikainen JEK, Camm AJ, Malik M. Arterial baroreflex sensitivity assessed from phase IV of the Valsalva maneuver. Am J Cardio 1996; 78: 575 9. Korner PI, Tonkin AM, Uther JB. Reflex and mechanical circulatory effects of graded Valsalva maneuvers in normal man. J Appl Physiol 1976; 40: 434 40. Levy G, Schupf N, Tang MX, Cote LJ, Louis ED, Mejia H, et al. Combined effect of age and severity on the risk of dementia in Parkinson s disease. Ann Neurol 2002; 51: 722 9. Magalhaes M, Wenning GK, Daniel SE, Quinn NP. Autonomic dysfunction in pathologically confirmed multiple system atrophy and idiopathic Parkinson s disease - a retrospective comparison. Acta Neurol Scand 1995; 91: 98 102. Magalhaes MVM, Mendes A, Silva T, Bettencourt M, Correia AP, Silva MR, et al. Disautonomy in idiopathic Parkinson s disease. prevalence and natural history. Mov Disord 1998; 13 (Suppl 2): 273. Mathias CJ, Bannister RS. Autonomic failure: a textbook of clinical disorders of the autonomic nervous system. 4th edn. New York: Oxford University Press; 1999. Marder K, Jacobs DM. Dementia. In: Factor SA, Weiner WJ, editors. Parkinson s disease: diagnosis and clinical management. New York: Demos Medical Publishing, Inc.; 2002. pp. 125 35. Martin R, Manzanares R, Molto JM, Canet T, Ruiz C, Matias-Guiu J. Cardiovascular reflexes in Parkinson disease. Ital J Neurol Sci 1993; 14: 437 42. Mesec A, Sega S, Kiauta T. The influence of the type, duration, severity and levodopa treatment of Parkinson s disease on cardiovascular autonomic responses. Clin Auton Res 1993; 3: 339 44. Micieli G, Martignoni E, Cavallini A, Sandrini G, Nappi G. Postprandial and orthostatic hypotension in Parkinson s disease. Neurology 1987; 37: 386 93. Niimi Y, Ieda T, Hirayama M, Koike Y, Sobue G, Hasegawa Y, et al. Clinical and physiological characteristics of autonomic failure with Parkinson s disease. Clin Auton Res 1999; 9: 139 44. Oishi N, Udaka F, Kameyama M, Sawamoto N, Hashikawa K, Fukuyama H. Regional cerebral blood flow in Parkinson disease with nonpsychotic visual hallucinations. Neurology 2005; 65: 1708 15. Oka H, Mochio S, Yoshioka M, Morita M, Inoue K. Evaluation of baroreflex sensitivity by the sequence method using blood pressure oscillations and R-R interval changes during deep respiration. Eur Neurol 2003; 50: 230 43. Oka H, Mochio S, Onouchi K, Morita M, Yoshioka M, Inoue K. Cardiovascular dysautonomia in de novo Parkinson s disease. J Neurol Sci 2006a; 241: 59 65. Oka H, Mochio S, Yoshioka M, Morita M, Onouchi K, Inoue K. Cardiovascular dysautonomia in Parkinson s disease and multiple system atrophy. Acta Neurol Scand 2006b; 113: 221 7. Oka H, Morita M, Onouchi K, Yoshioka M, Mochio S, Inoue K. Cardiovascular autonomic dysfunction in dementia with Lewy bodies and Parkinson s disease. J Neurol Sci 2007a; 254: 72 77. Oka H, Yoshioka M, Onouchi K, Morita M, Mochio S, Suzuki M, et al. Impaired cardiovascular autonomic function in Parkinson s disease with visual hallucinations. Mov Disord 2007b (DOI: 10.1002/ mds.21581). Oka H, Yoshioka M, Morita M, Onouchi K, Suzuki M, Ito Y, et al. Reduced cardiac 123 I-MIBG uptake reflects cardiac sympathetic dysfunction in Lewy body disease. Neurology (2007c). Orimo S, Amino T, Itoh Y, Takahashi A, Kojo T, Uchihara T, et al. Cardiac sympathetic denervation precedes neuronal loss in the sympathetic ganglia in Lewy body disease. Acta Neuropathol 2005; 109: 583 8. Piha SJ, Rinne JO, Rinne UK, Seppanen A. Autonomic dysfunction in recent onset and advanced Parkinson s disease. Clin Neurol Neurosurg 1988; 90: 221 6. Rajput AH, Offord KP, Beard CM, Kurland LT. Epidemiology of parkinsonism: incidence, classification, and mortality. Ann Neurol 1984; 16: 278 2. Robertson D, Hollister AS, Biaggioni I. Arterial baroreflex failure in man. Clin Auton Res 1993a; 3: 212. Robertson D, Hollister AS, Biaggioni I, Netterville JL, Mosqueda-Garcia R, Robertson RM. The diagnosis and treatment of baroreflex failure. N Engl J Med 1993b; 329: 1449 55. Saiki S, Hirose G, Sakai K, Kataoka S, Hori A, Saiki M, et al. Cardiac 123 I-MIBG scintigraphy can assess the disease severity and phenotype of PD. J Neurol Sci 2004; 220: 105 11. Sandroni P, Ahlskog JE, Fealey RD, Low PA. Autonomic involvement in extrapyramidal and cerebellar disorders. Clin Auton Res 1991a; 1: 147 55. Sandroni P, Benarroch EE, Low PA. Pharmacological dissection of components of the Valsalva maneuver in adrenergic failure. J Appl Physiol 1991b; 71: 1563 7. Sandyk R, Awerbuch GI. Dysautonomia in Parkinson s disease: relationship to motor disability. Int J Neurosci 1992; 64: 23 31. Senard JM, Valet P, Durrieu G, Berlan M, Tran MA, Montastruc JL, et al. Adrenergic supersensitivity in parkinsonians with orthostatic hypotension. Eur J Clin Invest 1990; 20: 613 9. Senard JM, Rascol O, Durrieu G, Tran MA, Berlan M, Rascol A, et al. Effects of yohimbine on plasma catecholamine levels in orthostatic hypotension related to Parkinson disease or multiple system atrophy. Clin Neuropharmacol 1993; 16: 70 6. Senard JM, Rai S, Lapeyre-Mestre M, Brefel C, Rascol O, Rascol A, et al. Prevalence of orthostatic hypotension in Parkinson s disease. J Neurol Neurosurg Psychiatry 1997; 63: 584 9. Senard JM, Brefel-Courbon C, Rascol O, Montastruc JL. Orthostatic hypotension in patients with Parkinson s disease: pathophysiology and management. Drugs Aging 2001; 18: 495 505. Szili-Torok T, Kalman J, Paprika D, Dibo G, Rozsa Z, Rudas L. Depressed baroreflex sensitivity in patients with Alzheimer s and Parkinson s disease. Neurobiol Aging 2001; 22: 435 8. Turkka JT. Correlation of the severity of autonomic dysfunction to cardiovascular reflexes and to plasma noradrenaline levels in Parkinson s disease. Eur Neurol 1987; 26: 203 10. Vogel ER, Sandroni P, Low PA. Blood pressure recovery from Valsalva maneuver in patients with autonomic failure. Neurology 2005; 65: 1533 7. Williams DR, Lees AJ. Visual hallucinations in the diagnosis of idiopathic Parkinson s disease: a retrospective autopsy study. Lancet Neurol 2005; 4: 605 10.