Delayed orthostatic hypotension in Parkinson’s disease and in the general ageing population

Abstract

Objective

Delayed orthostatic hypotension (dOH) is defined by a sustained blood pressure (BP) fall ≥20/10 mmHg occurring beyond 3 minutes in the upright position, whose clinical relevance is yet undetermined. We investigated frequency, associated features and treatment strategies of dOH in Parkinson’s Disease (PD) and ageing individuals with history of syncope or orthostatic intolerance.

Methods

We retrospectively studied 213 individuals with PD and 213 age-matched individuals without parkinsonism referred for tilt-table testing to the Innsbruck and Florence Dysautonomia centres. In both cohorts, we reviewed the medical records of the 6 months before testing for history of syncope and falls, and of the 6 months afterwards for the recommended dOH treatment and clinical outcome.

Results

dOH was twice as frequent in PD than in ageing individuals [18% versus 9%; OR = 2.9 (95 CI.: 1.3–6.5), P = .007]. Upon prolonged head-up tilt, PD individuals showed a more severe systolic BP fall (P < .001). PD individuals with history of syncope also exhibited a sustained systolic BP fall from the 3^rd minute of head-up tilt onward compared to those without (P = .014). Both non-pharmacological and pharmacological OH treatment strategies were associated with symptomatic improvement at follow-up.

Conclusions

In PD individuals referred to tertiary settings, tilt-table testing more frequently disclosed dOH compared to age-matched individuals without parkinsonism. In PD, dOH is mainly driven by a progressive orthostatic systolic BP fall, increasing the risk of syncope. Given its prevalence and the potential for effective treatment, dOH should be actively screened for in clinical practice.

Key Points

• Delayed orthostatic hypotension (OH) is more frequent in individuals with Parkinson’s disease (PD) than in the general ageing population.

• The fall in systolic BP drives the development of delayed OH in PD and is associated with previous history of syncope.

• Treatment strategies for classic OH may be equally beneficial in individuals with delayed OH.

Introduction

Classic orthostatic hypotension (cOH) consists of a sustained reduction in systolic and/or diastolic blood pressure (BP) within 3 minutes in the upright position [1]. The prevalence of cOH in the general population ranges from 6% in individuals below the age of 50 to 30% in those above 70 [2]. In individuals with Parkinson’s disease (PD), the estimated pooled prevalence of cOH is 30%, increasing with disease duration and age [3]. In PD, cOH develops following neurodegenerative changes affecting the peripheral autonomic nervous system (ANS) [4] and may be aggravated by non-neurogenic factors, such as treatment with CNS-active drugs with hypotensive effect [5, 6], dehydration, infections and anaemia [7]. In ageing individuals, multiple causes of OH may co-exist, including cardiovascular diseases, polypharmacy and age-dependent ANS neurodegenerative changes, precluding a binary distinction between neurogenic and non-neurogenic OH causes [8, 9]. COH may remain asymptomatic, but most commonly causes symptoms of orthostatic intolerance, including light-headedness, fatigue, coat-hanger pain, dyspnoea or blurred vision [10]. In severe cases, cOH may lead to orthostatic syncope and injurious falls, ultimately impacting on the quality of life, morbidity and mortality of affected individuals [10–12].

Delayed OH (dOH) is characterized by a sustained BP fall ≥20/10 mmHg developing beyond the classic 3 minutes in the upright position [1]. It is considered a prodromal form of cOH [13–15], equally predicting future morbidity and mortality [16], but remains frequently overlooked in clinical practice, particularly in ageing individuals.

Here we investigated the frequency of cOH and dOH in PD and ageing individuals with history of syncope and orthostatic intolerance and explored whether dOH is associated with any specific clinical-demographic characteristic, including recent history of falls and syncope. In individuals with dOH, we also evaluated the rate of symptomatic improvement under non-pharmacological and pharmacological OH treatment strategies.

Methods

Recruitment

We retrospectively screened individuals with PD referred for syncope and orthostatic intolerance between January 2008 and September 2016 to the Dysautonomia Centre of the Neurology Department of the Innsbruck Medical University, who underwent cardiovascular autonomic function testing for clinical and/or research purposes.

Inclusion criteria for PD individuals were:

• diagnosis of probable PD [17] or probable PD with dementia (PDD) [18];

• available clinical information concerning the 6 months prior to cardiovascular autonomic function testing.

Exclusion criteria were:

• other major neurological or psychiatric diseases;

• missing clinical documentation;

• low quality of tilt-test examination;

• interruption of tilt-test examination before 10 minutes of 60° head-up tilt-test, except if the premature interruption was due to syncope;

• The tilt-test examination was performed during the interventional phase of a clinical trial.

Individuals aged ≥65 without evidence of PD, PDD, secondary or atypical parkinsonism, who underwent tilt-table testing in the same time-frame at the Syncope Unit of the Department of Geriatrics of the Careggi University Hospital (Italy) were sequentially matched 1:1 for age with the recruited PD subjects, if detailed clinical information concerning the 6 months preceding autonomic function testing was available.

Clinical-demographic data set

We collected the following information from the medical records contemporary to the head-up tilt examination:

1. Demographic characteristics: age, sex.

2. Comorbidities: history of diabetes, cognitive impairment, cardiovascular disease and risk factors (e.g. coronary artery disease, cardiac arrhythmias, heart failure and hypertension).

3. Medications: total number of drugs/day, use of anti-hypotensive or anti-hypertensive medications, type of anti-hypertensive medication, use of other CNS-active drugs with BP lowering effect [i.e. benzodiazepines, antidepressants and neuroleptics [5]].

4. In PD individuals: parkinsonian characteristics including disease duration, Hoehn & Yahr (H&Y) stage, use and L-Dopa equivalent daily dose (LEDD) of L-Dopa or dopamine agonists [19], deep brain stimulation, history of L-Dopa fluctuations, gait abnormalities (freezing of gait, start- or turning hesitation).

We further reviewed the medical records for history of falls (i.e. events that resulted in the person coming to rest inadvertently on the ground, floor or other lower level [20], without concomitant loss of consciousness) and syncope (i.e. transient loss of consciousness due to global cerebral hypoperfusion [21]) occurred in the six months preceding the tilt-test examination.

Head-up tilt examination

A 60° head-up tilt was performed monitoring the heart rate (HR) and BP with non-invasive beat-to-beat finger-cuff plethysmography, with the left hand fixed at heart level and calibrated with oscillometric arm-cuff BP readings at the end of the supine phase and at regular intervals during the orthostatic challenge (Task Force Monitor, CNSystem, Graz-Austria). HR, systolic and diastolic BP values after 10 minutes in the supine position, and after 3 and 10 minutes upon 60° head-up tilt were calculated by averaging 15 values of the continuous HR and BP recordings at the above given time points.

Additional cardiovascular autonomic function tests in the PD population

In individuals with PD, additional cardiovascular autonomic function indices, including the inspiratory-expiratory difference during deep-breathing at 6 cycles/minute, the Valsalva ratio and the BP behaviour during phases II and IV of the Valsalva manoeuvre [22] were calculated whenever available, and compared to normative values [23].

Therapeutic measures for dOH

In individuals of both cohorts, whose tilt-test disclosed dOH, we reviewed the medical records following the cardiovascular autonomic assessment and collected information on the dOH-specific therapeutic recommendations and symptomatic outcome at follow-up.

Ethical statements

The study was conducted in accordance with the ethical standards of the Declaration of Helsinki, the directives of the Austrian Agency for Research Integrity, and the current European Data Protection Regulation. Due to its retrospective design and begin before July 2020, no written informed consent or approval by the Innsbruck or Florence Ethical Committees were needed.

Statistical analysis

Qualitative variables were summarized by frequency and percentage and tested with the Χ² or Fisher’s exact test, whenever appropriate. Quantitative variables were summarized by median (1^st quartile; 3^rd quartile) and tested for normality with the Kolmogorov–Smirnov test. Group-wise comparisons were performed using the Mann–Whitney-U or T-test, depending on the data distribution.

We first conducted a descriptive analysis of the clinical-demographic characteristics of the PD versus age-matched individuals without parkinsonism. We then assessed the frequency of cOH and dOH in both populations, performed a univariate, binary logistic regression and receiver operating characteristic (ROC) curve analysis of the clinical-demographic factors (beyond the diagnostic group) associated with dOH.

HR, systolic and diastolic BP behaviour upon prolonged head-up tilt were compared between PD and ageing individuals by means of an ANOVA for repeated measurements, followed by group-wise comparisons for each time point (i.e. 10^th minute supine, 3^rd and 10^th upon head-up tilt) and hemodynamic parameter (HR, systolic and diastolic BP). In the PD cohort, we compared the frequency of pathological cardiovascular autonomic function indices during deep breathing and Valsalva manoeuvre among individuals with cOH, dOH and no OH.

We subsequently performed a univariate analysis of the clinical-demographic features associated with history of syncope or falls in PD individuals, followed by a binary logistic regression (including those variables with P < .2), ROC curve calculation, and ANOVA of the hemodynamic changes upon prolonged head-up tilt in PD patients with versus without history of syncope or falls. The same statistical approach was repeated for the ageing individuals.

Finally, we performed a descriptive analysis of the therapeutic measures applied to treat dOH in both cohorts and their clinical outcome (i.e. symptoms improved or not). All analyses were done with SPSS®, version 29.0. A 2-sided P value < .05 was considered statistically significant. A Benjamini-Hochberg correction [24] was applied to multiple comparisons.

Results

Study population

We included 213 PD (35% women, n = 75, P = .048) and 213 age-matched individuals without parkinsonism (45% women, n = 95, P = .234). The clinical-demographic characteristics of the study population are summarized in Table 1. Compared to age-matched individuals, PD individuals suffered more frequently from cognitive impairment (P < .001) and were on a higher number of daily medications (P < .001), including anti-hypotensive (P < .001) and CNS-active drugs (P < .001). Conversely, ageing individuals suffered more frequently from cardiovascular comorbidities (P < .001) and were more often on anti-hypertensive regimens (P < .001). In the months preceding autonomic testing, ageing individuals had suffered more syncope episodes (P < .001), while the frequency of a history of recent falls was similar in both cohorts. The clinical-demographic features of the PD versus ageing population considered for the dOH-specific research questions (Table 1, right part) showed comparable results to the overall study population.

Table 1.

Clinical-demographic characteristics of the study population

	Overall study population			dOH endpointa
	PD (n = 213)	Ageing individuals (n = 213)	P Value	PD (n = 163)	Ageing individuals (n = 183)	P Value
Demographic characteristics
Age, years	73 (69; 76)	73 (69; 77)	0.771	73 (69, 77)	73 (69, 77)	0.771
Female	75 (35)	95 (45)	0.048	61 (37)	80 (44)	0.234
Clinical features
Cognitive impairment	61 (29)	14 (7)	<0.001 *	42 (26)	12 (7)	<0.001
Diabetes	15 (7)	31 (15)	0.012	11 (7)	26 (14)	0.025
Cardiovascular comorbidities	69 (32)	119 (56)	<0.001 *	57 (35)	100 (55)	<0.001
Medications
Total number of drugs/day	6 (4; 9)	5 (2; 7)	<0.001 *	6 (3,9)	5 (2,7)	<0.001
Anti-hypotensive drugs	19 (9)	0 (0)	<0.001 *	9 (6)	0 (0)	0.001
Anti-hypertensive drugs	90 (42)	144 (68)	<0.001 *	72 (44)	123 (67)	<0.001
Diuretics	42 (20)	60 (28)	0.041	34 (21)	54 (29)	0.65
ACE inhibitors	30 (14)	74 (35)	<0.001	24 (15)	63 (34)	<0.001
AT2R antagonists	20 (9)	45 (21)	<0.001	14 (9)	39 (21)	0.001
Ca2+ blockers	33 (16)	49 (23)	0.049	28 (17)	43 (23)	0.146
Β-blockers	33 (16)	55 (26)	0.008	46 (28)	25 (14)	0.24
Nitrates	8 (4)	11 (5)	0.481	6 (4)	9 (5)	0.573
α1 antagonists	29 (14)	28 (13)	0.887	20 (12)	21 (11)	0.819
α2 agonists	2 (1)	0 (0)	0.499	2 (1)	0	0.221
CNS-active drugs	108 (51)	60 (28)	<0.001 *	82 (51)	51 (28)	<0.001
Benzodiazepines (PD = 212)	52 (25)	31 (15)	0.010	39 (24)	27 (15)	0.28
Neuroleptics	26 (12)	7 (3)	<0.001	19 (12)	5 (3)	0.001
Antidepressants	77 (36)	41 (19)	<0.001	57 (35)	36 (20)	0.001
PD-specific characteristics
Disease duration, years	6 (3; 10)	–		5 (3,9)	–
Hoehn & Yahr Stage	2.5 (2; 3)	–		2.5 (2;3)	–
Use of L-Dopa	179 (84)	–		133 (82)	–
L-Dopa dose, mg/d	450 (300; 750)	–		400 (300;700)	–
Use of dopamine agonists (PD = 212)	83 (39)	–		69 (42)	–
DA LED, mg/d	0 (0; 130)	–		0 (0;480)	–
LEDD, mg/d	600 (333; 945)	–		555 (300; 908)	–
Deep brain stimulation	5 (2)	–		3 (2)	–
L-Dopa fluctuations (PD = 210)	72 (34)	–		50 (31)	–
Gait abnormalities (PD = 209)	66 (31)	–		47 (29)	–
Clinical readouts
Syncope in the 6 months before tilt test (PD = 211)	39 (18)	155 (73)	<0.001	25 (15)	133 (73)	<0.001
Falls in the 6 months before tilt test (PD = 211)	103 (48)	119 (56)	0.146	74 (45)	101 (55)	0.75

Abbreviations: ACE inhibitors: Angiotensin-converting-enzyme inhibitors; AT2R: Angiotensin type II Receptor; Ca²⁺ blockers: Calcium channel blockers; CNS: Central Nervous System; cOH: classic OH; dOH: delayed OH; DA: dopamine agonists; LEDD L-dopa equivalent daily dose; PD: Parkinson’s disease.

Qualitative values are reported by frequency (%) and quantitative values are reported by median (1^st quartile, 3^rd quartile). Values in bold: statistically significant after Benjamini-Hochberg correction.

^aIncluded all individuals without cOH, whose head-up tilt test was prolonged beyond 3 minutes.

^*Included in the binary logistic regression model.

Classic and delayed orthostatic hypotension in PD versus ageing individuals

Upon head-up tilt, 47 (22%) individuals with PD and 24 (11%) ageing individuals (P = .006) developed cOH. Prolonging the orthostatic stress, additionally disclosed dOH in 18% (n = 29) of the PD and 9% (n = 17) of the age-matched individuals (P = .013, Fig. 1).

Figure 1.

(a) Frequency of cOH and dOH in PD and ageing individuals; (b) ANOVA of HR, systolic and diastolic BP upon 10 minutes head-up tilt test in PD and ageing individuals. In the supine position, PD individuals had significantly higher HR [67 (61; 76) vs. 63 (57; 70) bmp, P = .002] and diastolic BP values [77 (69; 86) vs. 74 (65; 82) P ≤ .001]. Upon 10 minutes of head-up tilt-test, PD individuals showed a moderate, yet significantly higher HR increase [+8 (+3; +12) vs. + 6 (+2; +10) bpm, P = .003], a smaller increase in diastolic BP [+1 (−7; +7) vs. 8 (−3; +11) mmHg, P ≤ .001] and a fall in systolic BP [−4 (−15;+6) vs. +4 (−9; +13) mmHg, P ≤ .001] compared to ageing individuals. Created with Microsoft PowerPoint 2016.

In the binary logistic regression model, including the diagnostic group (i.e. PD vs. age-matched individuals) and other clinical features associated with a dOH diagnosis at univariate analysis (Table 1, right panel), the presence of dOH remained positively associated with a diagnosis of PD [OR = 2.9 (95% CI: 1.3–6.5), P = .007] and negatively with use of CNS-active drugs with BP lowering effect [OR = 0.4 (95% CI: 0.2–1.0), P = .039]. The area under the ROC curve was 0.786 (95% CI: 0.744–0-828, P < .001).

At ANOVA of supine to 10 minutes head-up tilt test hemodynamic changes, PD individuals showed different HR, systolic and diastolic BP patterns from ageing individuals (Fig. 1). PD individuals had significantly higher supine HR (P = .002) and diastolic BP values (P ≤ .001). Upon 10 minutes head-up tilt, PD individuals exhibited a moderate, yet significantly higher HR (P = .003), a sustained, greater fall in systolic BP (P ≤ .001), and a smaller increase in diastolic (P ≤ .001) BP.

Cardiovascular autonomic function indices in PD individuals

Deep breathing was performed in 131 PD individuals, Valsalva manoeuvre in 107 (Fig. 2). Subgroups comparison among the individuals with cOH, dOH or without OH upon head-up tilt test disclosed no significant differences in the frequency of pathological age-adjusted expiratory-inspiratory difference, Valsalva ratio or missing BP overshoot in the phase II and IV of the Valsalva manoeuvre.

Figure 2.

Frequency of pathological cardiovascular autonomic function indices in the PD cohort. Among the 131 individuals with PD who performed a deep breathing, 18 exhibited cOH, 13 dOH and 76 no OH upon head-up tilt test. A Valsalva manoeuvre was conducted in 107 individuals with PD, including 23 with cOH, 19 with dOH and 89 without OH. Created with Microsoft PowerPoint 2016.

History of syncope

Thirty-nine (18%) PD individuals had a recent history of syncope (Table 2). Univariate analysis pinpointed an association between syncope events and pre-established anti-hypotensive treatment, confirmed by a binary logistic regression [OR = 4.6 (95% CI: 1.5–13.5), P = .006]. The area under the ROC curve was 0.669 (95% CI: 0.569–0-769, P = .001). Upon head-up tilt, PD individuals with history of syncope showed a higher HR increase (P = .005) and had a more pronounced systolic BP fall (P = .014) from the 3^rd minute of head-up tilt onward (Fig. 3).

Table 2.

Univariate analysis of clinical and demographic characteristics associated with falls and syncope in PD (left) and ageing individuals without parkinsonism (right)

	PD						Ageing individuals
	Syncope + (n = 39)	Syncope − (n = 172)	P value	Falls + (n = 103)	Falls −(n = 108)	P value	Syncope + (n = 155)	Syncope − (n = 58)	P value	Falls + (n = 119)	Falls − (n = 94)	P value
Demographic characteristics
Age, years	73 (69; 76)	73 (69; 76)	0.671	72 (69;76)	73 (69;77)	0.466	72 (69; 77)	74 (69; 77)	0.802	73 (69; 76)	73 (69; 79)	0.440
Female	15 (36)	59 (35)	0.905	42 (41)	32 (30)	0.090*	68 (44)	27 (47)	0.726	61 (51)	34 (36)	0.028 *
Clinical features
Cognitive impairment	10 (28)	51 (29)	0.697	32 (31)	29 (27)	0.500	11 (7)	3 (5)	0.763	7 (6)	7 (7]	0.647
Diabetes	4 (10)	11 (6)	0.397	9 (9)	6 (6)	0.369	26 (17)	5 (9)	0.133*	12 (10)	19 (20)	0.037 *
Cardiovascular comorbidities	17 (45)	52 (30)	0.108*	34 (33)	35 (32)	0.926	87 (56)	32 (55)	0.900	61 (51)	58 (62)	0.128*
Medications
Total number of drugs/day	7 (5;11)	6 (3;9)	0.130*	6 (5;9)	6 (3;9)	0.860	5 (2; 7)	3,5 (2; 7)	0.256	4 (2; 6)	5 (2;7)	0.616
Use of anti-hypotensive	9 (23)	10 (6)	<0.001 *	13 (13)	6 (6)	0.073*	0 (0)	0 (0)	–	0 (0)	0 (0)	–
Use of anti-hypertensive	17 (44)	72 (42)	0.843	41 (40)	48 (44)	0.495	109 (70))	35 (60)	0.166*	79 (66	65 (69)	0.669
Diuretics	11 (28)	31 (18)	0.150*	23 (22)	19 (18)	0.389	47 (30)	13 (22)	0.253	30 (25)	30 (32)	0.280
ACE inhibitors	5 (13)	25 (14)	0.782	14 (14)	16 (15)	0.799	61 (39)	13 (22)	0.021 *	41 (34)	33 (35)	0.921
AT2R antagonists	3 (8)	17 (10)	0.673	5 (5)	15 (14)	0.025 *	34 (22)	11 (19)	0.636	23 (19)	22 (23)	0.469
Ca2+ blockers	5 (13)	27 (16)	0.651	13 (13)	19 (18)	0.314	39 (25)	10 (17)	0.221	25 (21)	24 (25)	0.436
Β-blockers	4 (10)	28 (16)	0.344	16 (15)	16 (15)	0.884	38 (24)	17 (29)	0.477	29 (24)	26 (27)	0.586
Nitrates	1 (3)	7 (4)	1.000	2 (2)	6 (6)	0.281	8 (5)	3 (5)	1.00	5 (4)	6 (6)	0.475
α1 antagonists	5 (13)	24 (14)	1.000	15 (15)	14 (13)	0.736	18 (15)	8 (14)	0.864	20 (13)	19 (11)	0.336
α2 agonists	1 (3)	1 (1)	0.336	1 (1)	1 (1)	1.000	0 (0)	0 (0)	–	0 (0)	0 (0)	–
CNS active drugs	23 (59)	84 (49)	0.267	58 (57)	49 (45)	0.096*	39 (36)	21 (25)	0.111*	39 (33)	21 (23)	0.093*
Benzodiazepines (PD = 210)	14 (36)	37 (22)	0.061*	28 (27)	23 (22)	0.336	18 (12)	13 (22)	0.047	20 (17)	11 (12)	0.294
Neuroleptics	7 (18)	19 (11)	0.473	17 (16)	9 (8)	0.071	6 (5)	1 (2)	0.677	6 (4)	1 (1)	0.137
Antidepressants	15 (41)	61 (35)	0.471	43 (42)	33 (31)	0.090	27 (17)	14 (24)	0.268	28 (23)	13 (14)	0.075
PD-related characteristics
Disease duration, years	6 (3;11)	6 (3;9)	0.899	7 (4; 12)	5 (2; 8)	0.008 *	–	–	–	–	–	–
Hoehn & Yahr Stage	2,5 (2;3)	2,5 (2;3)	0.813	3 (2;3)	2 (2;3)	<0.001 *	–	–	–	–	–	–
Use of L-Dopa	35 (20)	144 (80)	0.168	93 (90)	86 (80)	0.031 *	–	–	–	–	–	–
L-Dopa dose, mg/d	500 (338; 800)	450 (300;700)	0.445	550 (300;800)	400 (150; 600)	0.001	–	–	–	–	–	–
Use of dopamine agonists (PD = 212)	13 (33)	70 (41)	0.395	41 (40)	42 (39)	0.892	–	–	–	–	–	–
DA LED, mg/d	0 (0;86)	0 (0; 140)	0.595	0 (0; 140)	0 (0; 120)	0.975	–	–	–	–	–	–
LEDD, mg/d	737 (404; 957)	600 (300; 945)	0.188*	775 (410; 1075)	510 (257; 800)	<0.001 *	–	–	–	–	–	–
Deep brain stimulation	2 (5)	3 (2)	1.000	4 (80)	1 (20)	0.204	–	–	–	–	–	–
L-Dopa fluctuations (PD = 210)	12 (32)	59 (35)	0.713	42 (41)	29 (27)	0.036 *	–	–	–	–	–	–
Gait abnormalities (PD = 209)	11 (31)	55 (32)	0.886	45 (44)	21 (20)	<0.001 *	–	–	–	–	–	–
Clinical outcomes
cOH	12 (31)	35 (20)	0.158*	27 (26)	20 (18)	0.179*	18 (12)	6 (10)	0.794	12 (10)	12 (13)	0.539
dOH	5 (18)	23 (17)	0.827	15 (19)	13 (15)	0.400	14 (10)	2 (4)	0.242	8 (8)	8 (10)	0.662

Abbreviations: ACE inhibitors: Angiotensin-converting-enzyme inhibitors; AT2R: Angiotensin type II Receptor; Ca²⁺ blockers: Calcium channel blockers; DA: dopamine agonists; CNS: Central Nervous System; cOH: classic OH; dOH: delayed OH; LEDD L-dopa equivalent daily dose; PD: Parkinson’s disease.

Qualitative values are reported by frequency (%) and quantitative values are reported by median (1^st quartile, 3^rd quartile).

Values in bold: statistically significant after Benjamini-Hochberg correction.

^*Included in the binary logistic regression model.

Figure 3.

ANOVA of HR, systolic and diastolic BP measured upon head-up tilt test in PD and age-matched individuals without parkinsonism with positive versus negative history of syncope (upper panel) and falls (lower panel). Created with Microsoft PowerPoint 2016.

Among the age-matched individuals, 155 (73%) had a recent history of syncope (Table 2). Univariate analysis pinpointed an association between history of syncope and intake of ACE-inhibitors and benzodiazepines, not confirmed by the binary logistic regression (entailing a diagnosis of diabetes, the use of anti-hypertensive and other CNS-active drugs as covariates). ANOVA also did not reveal any difference in orthostatic hemodynamic changes between ageing individuals with versus without history of syncope (Fig. 3).

History of falls

One-hundred and three (48%) PD individuals reported a history of falls, associated at univariate analysis with Angiotensin type II Receptor (AT2R) antagonists, disease duration, H&Y Stage, LEDD, presence of L-Dopa fluctuations and gait abnormalities (Table 2). The binary logistic regression confirmed an association between history of falls and more advanced H&Y stages [3 (2; 3) in the fallers versus 2 (2; 3) in the non-fallers; OR = 2.2 (95% CI: 1.3–3.7), P = .002; area under the ROC curve = 0.617 (95% CI: 0.518–0-716, P = .025)] (Fig. 3).

One-hundred nineteen (56%) ageing individuals without parkinsonism had a history of recent falls (Table 2). At multivariable analysis, falls were more frequent in women [OR = 1.9 (95% CI: 1.1–3.5), P = .024] and individuals with diabetes [OR = 2.6 (95% CI: 1.1–5.8), P = .024; area under the ROC curve =0.649 (95% CI: 0.575–0-723, P = .001)].

In both studied cohorts, ANOVA disclosed no significant differences in the orthostatic HR, systolic and diastolic BP behaviour between fallers and non-fallers (Fig. 3).

Therapeutic approaches in PD and ageing individuals with dOH

Information on the treatment strategy was available for 86% (n = 25/29) of PD and all the ageing individuals (n = 17) diagnosed with dOH (see the Appendix S1  in the Supplementary Data section). Sixty percent (n = 15/25) of PD and all (n = 17) ageing individuals with dOH received recommendations for behavioural measures (i.e. increasing the daily fluid and salt intake, sleeping in a head-up position overnight). In addition, 11 patients (n = 4 PD, n = 7 ageing individuals) were advised to wear elastic stockings or elastic abdominal binders (n = 6 PD) [25]. Forty percent of the treated individuals with PD (n = 10/25) and 29% of the ageing individuals (n = 5/17) received pharmacological interventions, including reduction or withdrawal of previously established anti-hypertensive regimens (n = 5 PD, and n = 5 ageing individuals) or the introduction of sympathomimetic drugs, such as midodrine or etilefrine (n = 5 PD). A combination of pharmacological and non-pharmacological measures was recommended to 24% of PD (n = 6/25) and 29% of ageing individuals (n = 5/17).

Follow-up data were available for 13 out of 25 PD and four out of 17 ageing individuals. Symptom improvement was reported by 11 out of 13 PD and all ageing individuals (n = 4), without significant differences between pharmacological, non-pharmacological or combined therapeutic approaches (P = .531; Appendix S1).

Discussion

Here we found that both cOH and dOH were more frequent in PD than in age-matched individuals without parkinsonism referred for syncope and orthostatic intolerance. This indicates that OH, in its classic and delayed presentation, represents a specific cause of orthostatic intolerance in PD individuals [26]. The presence of dOH was also negatively associated with the intake of CNS-active drugs. Restricting such medications is in fact indicated in clinical practice to minimize the risk of hypotension and falls in individuals with autonomic failure [6].

In PD individuals, the orthostatic fall in systolic BP drove the development of dOH. This reflects observations in individuals with cOH, in whom systolic BP falls are predominantly observed [27], and suggests that neurodegenerative changes affecting the efferent baroreflex branch primarily impact systolic rather than diastolic BP regulation. The orthostatic HR increase was also more pronounced in the PD compared to the age-matched individuals, indicating that mechanisms of baroreflex compensation, and specifically the noradrenergic stimulation to the heart, are partially preserved at the stage when individuals suffer from dOH [28]. Losing such residual baroreflex buffering may ultimately determine the transition to cOH, in which an insufficient orthostatic HR counter-regulation is typically observed [13, 15, 29–32]. The frequency of pathological cardiovagal (i.e. Valsalva ratio, expiratory-inspiratory difference) and vascular sympathetic noradrenergic (i.e. BP regulation during and upon conclusion of the Valsalva manoeuvre) function indices did not eventually differ between PD individuals with or without cOH or dOH. Cardiovascular autonomic function indices alone may therefore not sufficiently stratify the severity of ANS impairment in PD individuals or other compensatory mechanisms may determine the occurrence of dOH rather cOH. Compared to the age-matched cohort, PD individuals also displayed higher diastolic BP values in the supine position. This may equally indicate an incipient baroreflex failure, predisposing PD individuals to developing supine hypertension [33, 34].

PD individuals with history of syncope were more frequently treated with anti-hypotensive medications. Notwithstanding, they displayed a progressive fall in systolic BP from the 3^rd minute of head-up tilt onward. Failing to stabilize the BP within the first minutes in the upright position may be critical for triggering syncope episodes [20], while slower and less steeper falls in BP, like in the case of dOH, may enable affected individuals to recognize symptoms of orthostatic intolerance and prompt behavioural adaptations (like counter-pressure manoeuvres, sitting or lying down) to prevent syncope occurrence. The severity of orthostatic BP falls may however variate from day-to-day and during the day. In clinical practice, the integration of a 24 h-ambulatory BP monitoring may therefore better estimate the degree of hypotensive susceptibility and syncope odds in single individuals with dOH [35]. In the age-matched controls, history of syncope was more frequent than in PD individuals, but it was not associated with any specific clinical-demographic characteristic or orthostatic hemodynamic pattern. This suggests that other mechanisms that remained undetected upon 10 minutes head-up tilt testing, such as reflex syncope or syncope due to cardiopulmonary causes, may have determined syncope occurrence in this patient population.

Nearly half of the PD individuals had a recent history of falls. This was associated with a more advanced H&Y stage [36], indicating Parkinson-dependent postural instability as putative cause of falls in these individuals, but no specific hemodynamic pattern upon head-up tilt. In contrast with studies reporting an association between cOH and falls [37–39], this data indicates a greater role of motor impairment over dOH in determining the risk of falls in PD individuals. This may be due to dOH slowness in determining orthostatic intolerance, leaving affected individuals the time to counteract them.

Also every second ageing individual without parkinsonism had a recent history of falls, independent risk factors being female sex and diabetes [40–42]. While diabetes may increase the risk of falls due to its association with sight impairment, sensory-motor neuropathy, cognitive decline and polypharmacy [43], female sex has been associated with indoor [44] and outdoor falling [45], possibly due to women’s reduced physical activity and increased propensity to report falls to healthcare providers [46, 47].

Both in PD and ageing individuals with dOH, the therapeutic approaches reflected those indicated for cOH [7, 48, 49], with non-pharmacological measures most frequently recommended first [21]. Even though the number of treated individuals was small, most of them reported symptomatic improvement at follow-up, independently of the pharmacological versus non-pharmacological therapeutic approach. This observation provides preliminary evidence that cOH therapeutic strategies may be equally effective to manage dOH.

Limitations

Our study has limitations. Detailed cardiovascular autonomic function indices were not available for all participants, particularly in the age-matched controls, and further studies are needed to appraise the contribution of cardiovagal versus cardiovascular noradrenergic dysfunction to dOH development. On the other hand, experienced clinicians from the Innsbruck centre conducted a centralized data analysis, minimizing potential biases due to data comparison from different centres and populations. The retrospective, rather than prospective design, and the lack of follow-up information for some of the included cases might have determined a documentation bias, though clinical data were extracted from electronic medical records following a structured methodology. Our cohorts included individuals with orthostatic intolerance and syncope, possibly introducing a selection bias towards more severe cases. Finally, there was a delay between data collection and analysis due to personnel and COVID-19 constrains, affecting several autonomic centres across Europe [50]. However, given that the dOH definition remained unchanged over time, this did not likely impact on the timeliness of the results.

Conclusion

DOH is a form of orthostatic hypotensive susceptibility peculiar to individuals with PD and it is mainly driven by a progressive systolic BP fall in the upright position. This poses PD individuals at risk of syncope and should be actively sought for in clinical practice, given its prevalence and potential response to symptomatic treatment. Other hemodynamic mechanisms likely determine syncope occurrence in ageing individuals without parkinsonism.

Declaration of Conflicts of Interest:

J-P.N. reports Honoria from the Austrian Autonomic Society, the Austrian Parkinson Society, outside of the present work.

K.S. KS reports honoraria from the International Parkinson and Movement Disorders Society, grants from the FWF Austrian Science Fund, the Michael J. Fox Foundation, and the International Parkinson and Movement Disorder Society, as well as personal fees from Teva, UCB, Lundbeck, AOP Orphan Pharmaceuticals AG, AbbVie, Roche, and Grünenthal, outside the submitted work.

W.P. reports personal fees from Alterity, AbbVie, AC-Immune, AstraZeneca, BIAL, Biogen, Britannia, Lilly, Lundbeck, Neuroderm, Roche, Takeda, Teva, UCB and Zambon in relation to consultancy around clinical drug development programmes for PD and MSA, outside of the present work.

M.R. reports speaker fees from Boston Scientific.

A.F. reports royalties from Springer Verlag, speaker fees and honoraria from Bial, CNSystems, Elsevier, Theravance Biopharma, Medtronic, Sanofi, Austrian Autonomic Society, Austrian Neurology Society, International Parkinson Disease and Movement Disorders Society, Elsevier and research grants from the FWF-Austrian Science Fund, Medical University of Innsbruck, Mission MSA and Dr Johannes and Hertha Tuba Foundation, outside of the present work.

Declaration of Sources of Funding:

Dr Caliò is supported by the Austrian Science Fund [FWF FG2700-B], Dr Campese is supported by Mission MSA and Dr. Johannes and Herta Tuba Foundation. These financial sponsors played no role in the design, execution, analysis and interpretation of data or writing of the study.

Data Availability:

Any study-related data not included in the present manuscript can be made available in a de-identified way, upon reasonable request from any qualified investigator.