Hypermobile Ehlers-Danlos Syndrome: Cerebrovascular, Autonomic and Neuropathic Features

Highlights

• •Hypermobile Ehlers-Danlos syndrome affects multiple systems, however, a comprehensive analysis of cerebrovascular, autonomic, and neuropathic features in a larger sample is lacking.
• •This study provided evidence of cerebrovascular dysregulation with reduced orthostatic cerebral blood flow velocity associated with symptoms of cerebral hypoperfusion, frequent small fiber neuropathy, and a widespread but mild autonomic failure in hEDS.

Abstract

Background

Hypermobile Ehlers-Danlos syndrome (hEDS) affects multiple systems, but comprehensive evaluations of a larger sample of hEDS patients are lacking. The objective of this study was to describe cerebrovascular, autonomic, and neuropathic features of hEDS.

Methods

This retrospective case-control study was conducted at Brigham and Women’s Faulkner Hospital between 2016-2023. Data from hEDS patients who completed autonomic testing and skin biopsies were analyzed. Outcome measures include validated surveys (Survey of Autonomic Functions, Neuropathy Total Symptom Score-6 (SAS)) and autonomic function testing (Valsalva maneuver, deep breathing, head-up tilt and sudomotor), cerebrovascular (cerebral blood flow velocity (CBFv) in the middle cerebral artery), respiratory (capnography), and neuropathic (skin biopsies for assessment of small fiber neuropathy) testing and inflammatory/ autoimmune markers.

Results

Total 270 hEDS patients were analyzed and compared to 29 healthy controls. Common hEDS complaints (prevalence > 90% ) were orthostatic sudomotor, vasomotor, gastrointestinal, and pain. Orthostatic cerebral blood flow velocity was reduced in 79% of hEDS and correlated with orthostatic dizziness. The head-up tilt test revealed postural tachycardia syndrome (prevalence 33%), hypocapnic cerebral hypoperfusion (22%), orthostatic cerebral hypoperfusion syndrome (18%), and neurogenic orthostatic hypotension (9%). Widespread but mild autonomic failure was present in 90% of hEDS patients on autonomic testing. Small fiber neuropathy using structural criteria was detected in 64%, and using combined structural and functional criteria in 82%.

Conclusions

This study provided evidence of cerebrovascular dysregulation with reduced orthostatic cerebral blood flow velocity associated with symptoms of cerebral hypoperfusion, frequent small fiber neuropathy, and widespread but mild autonomic failure in hEDS.

Graphical Abstract

Introduction

Hypermobile Ehlers-Danlos Syndrome (hEDS) is a heritable connective tissue disorder, genetically undefined, with a broad phenotypic spectrum.¹ The estimated prevalence of hEDS is approximately 1:500 (0.2%), with 70% of diagnosed cases in Wales, UK, occurring in females.² The true prevalence is unclear, as only a fraction of patients are clinically diagnosed³, and causal molecular defects are unknown.4, 5, 6

The hEDS patients may present with chronic pain, fatigue, and recurrent joint dislocations⁵ and as a result, many patients report poor quality of life, symptoms of depression, anxiety, substance misuse and psychosocial distress.^7,8

Autonomic complaints are common. hEDS patients experience orthostatic intolerance and bladder, gastrointestinal, secretomotor, and pupillomotor complaints.9, 10, 11 Underlying cardiovascular autonomic manifestations of hEDS are postural tachycardia syndrome (POTS), neuro-cardiogenic syncope, and orthostatic hypotension.^9,11,12 Musculoskeletal and neuropathic pain, accompanied by central sensitisation, are key features of hEDS, although the exact cause of pain remains unclear.13, 14, 15 Small fiber neuropathy (SFN) could be a peripheral substrate of neuropathic pain, while central sensitization could be a central component of pain.16, 17, 18

Orthostatic intolerance is frequent, affects most hEDS patients, and is associated with impaired quality of life.⁹ The cause of orthostatic intolerance in hEDS is not fully explained. Orthostatic hypotension, as a physiological substrate of orthostatic intolerance is rare in EDS and POTS is detected only in 15-50% of patients.^9,19, 20, 21

Cerebral hypoperfusion is a common cause of orthostatic intolerance.²² However, there is limited understanding of whether the cerebral blood flow is affected in hEDS. A subgroup of patients with myalgic encephalomyelitis/chronic fatigue syndrome and joint hypermobility syndromes exhibited lower orthostatic cerebral blood flow, as indicated by reduced blood flow measured by Doppler ultrasound of the neck arteries.²³ A study using transcranial Doppler found reduced orthostatic cerebral blood flow in hEDS.²¹

In this retrospective study, we aimed to assess the cerebrovascular, autonomic, and neuropathic features of hEDS in a larger cohort.

Materials and Methods

Study Participants

This retrospective, single-center study included consecutive hEDS patients who underwent autonomic testing between 2016 and 2023 at the Brigham and Women’s Faulkner Hospital Autonomic Laboratory for evaluation of dysautonomia and related small fiber neuropathy.

We used historical healthy controls from our research database. The controls had no history of orthostatic intolerance, chronic pain, or joint hypermobility and had normal autonomic testing.

Patient’s electronic records were reviewed for past medical history, medication usage, and laboratory data, including inflammatory and autoimmune markers.²⁴ The accuracy of co-morbidities, and use of medication was verified with each patient at the time of the autonomic testing.

Ethics

The Institutional Review Board of the Brigham and Women’s Hospital, Harvard Medical School approved the study as a minimal-risk study, and the consent form signature was waived.

Clinical Definitions

Key features of hEDS are joint hypermobility, variable pain syndromes, hyperextensible skin, and autonomic dysfunction.²⁵ Usually, the diagnosis of hEDS was suspected by patient’s primary care or specialist physicians, then the patients were referred to a genetic specialist (JK, AM or JM) to confirm the diagnosis. As there are no reliable genetic markers, the diagnosis of hEDS is clinical and based on consensus clinical criteria. Patients diagnosed in 2016-2017 were scored by the Beighton-Villefranche criteria²⁶, afterwards by the 2017 international classification.⁴

Exclusion criteria were missing electronic records and patients with autonomic testing who were unable to tolerate discontinuation of medication that affects autonomic function.

Self-Reported Measures

Autonomic symptoms over the past 6 months were assessed using the psychometrically sound, internally consistent (Cronbach α = 0.76) 12-item Survey of Autonomic Symptoms (SAS, 11 items for women, 12 items for men).²⁷ The SAS assesses autonomic symptom severity across 6 domains (orthostatic, sudomotor, vasomotor, gastrointestinal, urinary, and sexual). Higher scores indicate greater symptom burden. The SAS range is 0-55 for women and 0-60 for men. The cutoff point > 7 in the SAS score was considered to be clinically significant (sensitivity >60%, specificity > 90%).²⁷ SAS differentiates symptoms into none (0), symptoms are not bothering (1), symptoms are bothering a little (2), symptoms are bothering some (3), symptoms are bothering a moderate amount (4), and symptoms are bothering a lot (5). The SAS assesses the following domains: orthostatic (question 1), sudomotor (questions 2,5,6,7), vasomotor (questions 3,4), gastrointestinal (questions 8,9,10), urinary (question 11) and sexual (man, question 12). For each question, we grouped the symptom intensity into none 0, mild 1-2, moderate 4 and severe 5. For the domains, the intensities were grouped as follows: orthostatic, urinary and sexual: none (score 0), mild (1-2), moderate (4), and severe (5); vasomotor: none (0), mild (1-3), moderate (4-6), severe (7-10); sudomotor: none (0), mild (1-6), moderate (7-14), severe (15-20); gastrointestinal; none (0), mild (1-4), moderate (5-9), severe (10-15), total: none (0); mild (1-7), moderate (8-32), severe≥33.

Sensory complaints including pain over the preceding 24 hours were assessed using the validated Neuropathy Total Symptom Score-6 (NTSS), internally consistent with Cronbach α = 0.7.²⁸ NTSS quantifies sensory symptoms into 6 types (1, numbness and/or insensitivity; 2, prickling and/or tingling; 3, burning sensation; 4, aching pain and/or tightness; 5, sharp, shooting, lancinating pain; and 6, allodynia and/or hyperalgesia). Patients grade frequency and intensity of symptoms.The intensity includes none, mild (does not interferes/restrict daily activities), moderate (symptoms sometimes interferes with daily activities or patient needs pain treatment) and severe (symptoms usually interferes with daily activities even with pain treatment). The frequency was graded as never, occasional (<1/3 of the time), often (1/3-2/3 of the time) and almost continuous (> 2/3 of the time). The NTSS scores range from 0 to 21.96 with scores >0 indicating the presence of sensory symptoms in proportion to the number. The NTSS-6 total score > 6 is clinically significant.²⁸ Pain for last 7 days was assessed using the NIH Toolbox’s numerical rating scale (0 = no pain, 10 = worst imaginable pain).²⁹ The scores ≤3 correspond to mild, scores 4-6 to moderate and scores ≥7 to severe pain.³⁰

Autonomic Tests

The Brigham protocol was used for the quantitative assessment of autonomic functions and SFN.³¹ We previously described the protocol in detail.^31,32 Medications that may affect autonomic function were discontinued for 5 half-lives or longer before the testing. Briefly, cardiovascular functional autonomic testing included deep breathing (a marker of parasympathetic cardiovagal functions), the Valsalva maneuver and the head-up tilt test (both markers of parasympathetic and adrenergic sympathetic functions), and sudomotor evaluation (a marker of postganglionic sudomotor functions). A deep breathing test was done for 1 minute at 6 breaths per minute, with inhalation/exhalation of 10 seconds each. The cardiovagal index was obtained as the average difference between expiratory and inspiratory heart rate. The Valsalva maneuver was performed with expiratory pressure equal to 40 mm Hg for 15 seconds. The adrenergic index was defined as the difference between the baseline and the end of phase 2 in mean blood pressure. Patients were resting for 10 minutes at baseline and then were tilted upright for 10 minutes.

Patients were monitored for the presence of symptoms of cerebral hypoperfusion during the head-up tilt test. In particular, patients were asked for the presence of orthostatic lightheadedness, which is the most frequent and most typical symptom of cerebral hypoperfusion.33, 34, 35 Because many patients use 'lightheadedness' and 'dizziness' interchangeably, we opted to use the term 'lightheadedness' to refer to both sensations. However, we excluded the vestibular lightheadedness characterized as lightheadedness associated with tilt-table up movement at the beginning of the head-up tilt test that resolves seconds after the tilt table assumes the upright position.

The extraction of autonomic, cerebral blood flow and respiratory variables from the tilt were previously described in detail.³⁶ The following signals were recorded throughout the testing: electrocardiogram, beat-to-beat (by finger cuff with photoplethysmographic signal using volume-clamping ) and intermittent (by brachial sphygmomanometry) blood pressure, end-tidal CO₂ (Nonin Respsense Capnography, Nonin Medical Systems, Plymouth, MN), and cerebral blood flow velocity in the middle cerebral artery using transcranial Doppler (MultiDop T, Multigon, New York, NY). The supine normative threshold for the lower limit of CBFv is (women/men) 82.2 to 0.45 (cm/s)*age (years) / 72.09 to 0.38 (cm/s)*age (years).³¹ The normal threshold in percent for the orthostatic reduction in CBFv was calculated using the following: 90.36 to 0.443 * minute of the tilt (Equation 1).

Sudomotor testing was done using an electrochemical skin conductance (ESC).³⁷ ESC correlates with loss of sweat gland nerve fibers and it is a reasonable proxy for sudomotor function.³⁸

Skin Biopsy

Epidermal nerve fiber density (ENFD) and sweat gland nerve fiber density (SGNFD) were obtained using established standards.^39,40 Skin samples were taken from the proximal thigh 20 cm distal to the iliac spine and from the calf 10 cm above the lateral malleolus using a 3-mm circular punch tool. Skin samples were immunoperoxidase-stained for the axonal marker PGP 9.5. Skin processing and fiber counting were done at Therapath (New York, NY).

SFN is a clinical diagnosis typically confirmed by abnormal skin biopsy or by functional testing.^39,41 Therefore, SFN was defined as a combination of clinical signs suggestive of small fiber dysfunction (pinprick and thermal sensory loss, allodynia, and hyperalgesia³⁹) and structural (obtained from skin biopsy) or functional (obtained from ESC) variables.³⁸ The following subtypes of SFN were used in this study: sensory SFN (abnormal ENFD, normal SGNFD), mixed SFN (abnormal ENFD and SGNFD), autonomic (normal ENFD, abnormal SGNFD), functional (abnormal electrochemical skin conductance (ESC), and combined functional-morphological (at least 1 abnormal: ENFD, SGNFD, and ESC).^39,42, 43, 44

Grading of Small Fiber Neuropathy, Autonomic Impairment, and Cerebral Blood Flow

Test results were graded using the Quantitative Scale for Grading of Cardiovascular Autonomic Reflex Tests and Small Fibers from Skin Biopsies (QASAT).³¹ QASAT is an objective instrument for grading the severity of autonomic dysfunction, small fiber neuropathy, and cerebral blood flow abnormalities. Each domain (heart rate, blood pressure, cerebral blood flow, end-tidal CO₂) is analyzed, where a score equal to 0 is normal, and above 0 is abnormal with the severity proportional to the number.

QASAT score grading autonomic failure(QASAT_af) is defined as follows³⁶:

$$\text{QASA}{\mathrm{T}}_{\text{af}}=\text{QASA}{\mathrm{T}}_{\text{cardiovagal}}+\text{QASA}{\mathrm{T}}_{\text{adrenergic}}+\text{QASA}{\mathrm{T}}_{\text{sudomotor}}.;$$

where the QASAT_cardiovagal score is obtained from heart rate responses in deep breathing test; the QASAT_adrenergic score was obtained as a summation of blood pressure responses to the Valsalva maneuver and the head-up tilt scores; the QASAT_sudomotor score was obtained from the ESC test. The range of QASAT_ai is 0-22. Results for each domain and combined domains were assigned as normal, mildly abnormal, and severely abnormal.

The QASAT ranges were defined as: Autonomic failure: normal/none 0, abnormal: mildly 1-3, moderately 4-12 and severely >12-22; cardiovagal failure: normal/none 0, abnormal: mildly 1, moderately 2 and severely 3; adrenergic failure—Valsalva maneuver: normal/none 0, abnormal: mildly 1, moderately 2 and severely 3; adrenergic failure—orthostatic hypotension: normal/none 0, abnormal: mildly 1, moderately 2-5 and severely 6-10; orthostatic tachycardia: normal/none 0, abnormal: mildly 1-2, moderately 3-5 and severely 6-10; sudomotor failure—ESC: normal/none 0, abnormal: mildly 1-2, moderately 3-4 and severely 5-6; sudomotor failure—ENFD: normal/none 0, abnormal: mildly 1-2, moderately 3-6 and severely 7-8; abnormal SGNFD: normal/none 0, abnormal: mildly 1-2, moderately 3-6 and severely 7-8; reduced orthostatic end-tidal CO₂: normal/none 0, abnormal: mildly 1-2, moderately 3-5 and severely 6-10; reduced orthostatic CBFv: normal/none 0, abnormal: mildly 1-2, moderately 3-5 and severely 6-10; Details of calculations and grading of the testing were published previously.³⁶

Definition of Head-Up Tilt Patterns

We classified head-up tilt patterns into normal response, POTS (defined by orthostatic tachycardia ≥30 beats per minute for age > 19 year old (≥40 beats per minute for age 18-19 years old) without orthostatic hypotension)²², hypocapnic cerebral hypoperfusion (defined by reduced orthostatic cerebral blood flow velocity, normative value are defined by Equation 1) and reduced orthostatic end-tidal CO₂ (<30 mmHg) without orthostatic tachycardia and orthostatic hypotension),⁴⁵ orthostatic cerebral hypoperfusion syndrome (defined by reduced orthostatic CBFv where normative value is defined by Equation 1 without orthostatic hypotension, orthostatic tachycardia, and orthostatic hypocapnia)⁴⁶, orthostatic hypotension⁴⁷ and neurally mediated syncope.⁴⁷ Neurogenic orthostatic hypotension was classified as a subset of orthostatic hypotension if the compensatory heart rate increment during the orthostatic hypotension was absent.⁴⁸

Statistics

hEDS and controls were compared by Welch's 2-sample t-test for continuous variables and by chi-squared test for categorical variables. The relationship between lightheadedness during head-up tilt (absent versus present) and QASAT domains, adjusted for age and sex, was evaluated using a binary logistic regression model. Missing data were ignored. The R software (www.r-project.org) was used for statistical analyses.

Results

Study Participants

From 3861 patients that underwent autonomic testing, a total of 270 hEDS patients were analyzed in this study, and they were compared to 29 age, sex, and body mass index matched healthy controls (Figure 1 and Table 1).

Figure 1.

A flowchart diagram of the study.

Table 1.

Demographic and Baseline Characteristics

Variable	Control (n = 29)	hEDS (n = 270)	P-Value†	Conf.int
Age, years	38.66 (11.91)	35.47 (12.44)	.189	−1.58, 7.95
Sex, female %	93.1	94.1	.689
BMI, m2/kg	24.13 (4.86)	25.49 (5.96)	.242	−3.60, 0.91
Symptoms duration, years	0*	11.56 (8.10)	<.001	−12.53, −10.59
Comorbid conditions
Diabetes mellitus, %	0	1.9
Lyme disease, chronic, %	0	3.7
Mast cell activation syndrome, %	0	41.1
Hereditary alpha tryptasemia, %	0	3.7
Depression, %	0	80.9
Fibromyalgia, %	0	26.5
Irritable bowel syndrome, %	0	45.6
Anxiety, %	0	77.9
Headaches, %	0	79.4
Medical therapy
Anti-histamine, %	0	60.7
Pain, %	0	63.0
Pressor, %	0	31.1
Psychiatric, %	0	53.0
Hypertension, %	0	6.3
Antitachycardic, %	0	17.8
Gastrointestinal, %	0	35.6
Immmunomodulators, %	0	5.6
Laboratory evaluations
C-reactive protein-high sensitivity, normal ≤3 mg/L		3.16 (5.84)
C-reactive protein high sensitivity, % abnormal		26.4
Interleukin 6, normal<7.1 pg/mL		3.01 (1.46)
Interleukin 6, % abnormal		2.6
Interleukin 1b, normal<0.1 pg/mL		0.36 (1.07)
Interleukin 1b, % abnormal		5.9
Tumor necrosis factor alpha, normal ≤2.8 pg/mL		3.16 (4.64)
Tumor necrosis factor alpha, % abnormal		22.7
Leptin, normal range = 3.3-18.3 ng/mL		12.00 (13.85)
Leptin, % abnormal		0.21 (0.42)
Tryptase, normal<11.5 ng/mL		4.91 (3.60)
Tryptase, % abnormal		8.9
Voltage gated potassium channel complex antibody, normal ≤0.02 nmol/L		0.01 (0.09)
Voltage gated potassium channel complex antibody, % abnormal		3.1
Calcium channel P/Q antibody, normal ≤0.02 nmol/L		0.00 (0.01)
Calcium channel P/Q antibody, % abnormal		1.5
Trisulfated heparin disaccharide antibody, normal titer<10000		7400.00 (11998.52)
Trisulfated heparin disaccharide antibody, % abnormal		30.0
Fibroblast growth factor receptor 3 antibody, normal titer<3000		896.47 (2015.03)
Fibroblast growth factor receptor 3 antibody, % abnormal		17.6
Neutrophil, normal range = 1.8-7.7 K/uL		4.30 (1.37)
Neutrophil % abnormal		100.0
Lymphocyte, normal range = 1.0-4.8 K/uL		1.90 (0.62)
Lymphocyte % abnormal		0.0
Platelet, normal range = 150-400 K/uL		268.25 (70.19)
Platelet % abnormal		1.8
Norepinephrine supine, normal range = 70-750 pg/mL,		503.48 (219.13)
Norepinephrine supine, % abnormal		6.7
Norepinephrine standing, normal range = 200-1700 pg/mL		659.77 (288.79)
Norepinephrine standing, % abnormal		14.0
Cortisol, normal range = 6.0-18.4 ug/dL		13.55 (6.70)
Cortisol, % abnormal		12.5
ACTH, normal range = 7.2-63 pg/mL		17.88 (12.76)
ACTH, % abnormal		16.7
Myoglobin, normal ≤71 ng/mL		26.81 (11.87)
Myoglobin, % abnormal		3.7
Ferritin, normal range = 20-300 ug/L		68.00 (65.34)
Ferritin, % abnormal		20.0

% = percent of abnormal findings.

^⁎Controls typically reported mild, intermittent and nonsignificant symptoms. Data are mean ± sd.

^†Calculated using t test or chi-squared test as appropriate.

The first 29 patients were initially assesed using the Beighton-Villefranche criteria, while the remaining patients were evaluated using the newer 2017 international criteria. The Beighton-Villefranche criteria do not differentiate between hEDS and hypermobile spectrum disorders (HDS), the latter is considered a milder form of hEDS.⁴ Therefore, the diagnosis of hEDS was confirmed retroactively in the patients 1-29 using international criteria.

The most common comorbidities in the hEDS group were depression, headaches, anxiety, irritable bowel syndrome, and symptoms of mast cell activation disorder. The most common medical therapy in hEDS was anti-histamine, pain, and psychiatric. The most common abnormal laboratory variables were abnormal C-reactive protein high sensitivity, tumor necrosis factor-alpha, trisulfated heparin disaccharide antibody, and ferritin.

Symptoms

The SAS score was higher in hEDS compared to controls (Tables 2, and 3). At least 1 autonomic symptom was present in all hEDS patients. A clinically significant abnormal SAS score was found in 98.1% of hEDS patients and 0% in controls. Symptoms commonly (>90%) affected all autonomic domains except urinary. The most common symptom in the orthostatic domain was lightheadedness.

Table 2.

Survey of Autonomic Symptoms Scores

Domain	Control (n = 29)	hEDS (n = 270)	P-Value†	Conf.int
Total score	1.41 (1.52)	29.79 (9.74)	<.001	−29.67, −27.08
Orthostatic	0.38 (0.56)	4.09 (1.06)	<.001	−3.95, −3.46
Sudomotor	0.41 (0.78)	8.83 (4.54)	<.001	−9.03, −7.80
Vasomotor	0.45 (0.83)	6.43 (2.89)	<.001	−6.44, −5.52
Gastrointestinal	0.03 (0.19)	8.59 (3.81)	<.001	−9.02, −8.09
Urinary	0.14 (0.44)	1.79 (1.70)	<.001	−1.91, −1.39

Data are mean ± sd.

^†Calculated using t test.

Table 3.

Survey of Autonomic Symptoms Scores, Frequency and Severity

Domain	Control (n = 29)	hEDS (n = 270)	P-Value†
Total-abnormal, %	65.5	100.0	<.001
none, %	34.5	0.0
mild, %	65.5	1.9
moderate, %	0.0	56.7
severe %	0.0	41.5
Significant, %	0.0	98.1	<.001
Orthostatic, %	34.5	99.6	<.001
none, %	65.5	0.4
mild, %	31.0	1.5
moderate, %	3.4	7.0
severe %	0.0	91.1
Sudomotor, %	27.6	97.4	<.001
none, %	72.4	2.6
mild, %	27.6	32.2
moderate, %	0.0	49.6
severe %	0.0	15.6
Vasomotor, %	27.6	95.9	<.001
none, %	72.4	4.1
mild, %	27.6	14.8
moderate, %	0.0	27.8
severe %	0.0	53.3
Gastrointestinal, %	3.4	95.9	<.001
none, %	96.6	4.1
mild, %	3.4	12.6
moderate, %	0.0	39.3
severe %	0.0	44.1
Urinary, %	10.3	60.7	<.001
none, %	89.7	39.3
mild, %	6.9	4.4
moderate, %	3.4	22.6
severe %	0.0	33.7

^†Calculated using chi-squared test.

% = percent of findings.

The NTSS-6 score was higher in hEDS compared to controls (Tables 4, and 5). A clinically significant score was found in 88.8% of hEDS patients and 0% in controls. At least 1 pain complaint was present in all hEDS patients, the most common complaint was aching pain and prickling pain. Using the 0-10 numerical rating scale, pain was more common and more severe in hEDS than in controls.

Table 4.

Neuropathy Total Symptom Score-6

Variable	Control (n = 29)	hEDS (n = 270)	P-Value†	Conf.int
Total	0.22 (0.58)	11.50 (4.52)	<.001	−11.86, −10.70
Aching frequency	0.00 (0.00)	2.57 (0.76)	<.001	−2.66, −2.48
Aching intensity	0.00 (0.00)	2.11 (0.71)	<.001	−2.20, −2.03
Allodynia frequency	0.00 (0.00)	1.36 (1.01)	<.001	−1.48, −1.23
Allodynia intensity	0.00 (0.00)	1.46 (1.05)	<.001	−1.59, −1.33
Burning frequency	0.03 (0.19)	1.60 (1.09)	<.001	−1.72, −1.42
Burning intensity	0.03 (0.19)	1.54 (1.05)	<.001	−1.65, −1.37
Lancinating frequency	0.07 (0.37)	1.69 (1.05)	<.001	−1.80, −1.43
Lancinating intensity	0.03 (0.19)	1.70 (0.98)	<.001	−1.80, −1.53
Prickling frequency	0.00 (0.00)	2.03 (0.91)	<.001	−2.14, −1.92
Prickling intensity	0.00 (0.00)	1.70 (0.79)	<.001	−1.79, −1.60
Numbness frequency	0.10 (0.56)	1.92 (1.03)	<.001	−2.06, −1.58
Numbness intensity	0.03 (0.19)	1.66 (0.88)	<.001	−1.75, −1.50
Pain, NRS	0.00 (0.00)	4.40 (2.38)	<.001	−4.69, −4.12

NRS = the 0-10 numerical rating scale.

Data are mean ± sd.

^†Calculated using t test.

Table 5.

Neuropathy Total Symptom Score-6, Frequency and Severity

Variable	Control (n = 29)	hEDS (n = 270)	P-Value†
Total-abnormal, %	13.8	100.0	<.001
none, %	86.2	0.0
mild, %	13.8	11.1
moderate, %	0.0	60.0
severe %	0.0	28.9
Total significant, %	0.0	88.9	<.001
Allodynia, %	0.0	75.6	<.001
none, %	100.0	25.9
mild, %	0.0	18.9
moderate, %	0.0	38.5
severe %	0.0	16.7
Aching, %	0.0	97.8	<.001
none, %	100.0	3.3
mild, %	0.0	10.0
moderate, %	0.0	58.9
severe %	0.0	27.8
Prickling, %	0.0	94.1	<.001
none, %	100.0	7.0
mild, %	0.0	30.0
moderate, %	0.0	49.3
severe %	0.0	13.7
Burning, %	3.4	76.3	<.001
none, %	96.6	23.7
mild, %	3.4	17.4
moderate, %	0.0	39.6
severe %	0.0	19.3
Lancinating, %	3.4	83.0	<.001
none, %	96.6	17.4
mild, %	3.4	15.6
moderate, %	0.0	47.0
severe %	0.0	20.0
Numbness, %	3.4	87.0	<.001
none, %	96.6	12.2
mild, %	3.4	24.4
moderate, %	0.0	48.5
severe %	0.0	14.8
Pain, NRS, %	0.0	87.4	<.001
none, %	100.0	14.1
mild, %	0.0	18.5
moderate, %	0.0	46.0
severe, %	0.0	20.4

NRS = the 0-10 numerical rating scale.

Data are % percent of findings.

^†Calculated using chi-squared test.

Autonomic Tests Results

Tables 6, and 7 and Figures 2, and 3 summarize autonomic testing and skin biopsy results. Supine heart rate, blood pressure and end tidal-CO₂ were similar between hEDS and controls. Supine systolic (P = .002) and mean (P = .027) CBFv was lower in hEDS compared to controls. Supine cerebrovascular resistance index was higher in hEDS (P = .004).

Table 6.

Autonomic Testing Results

Variable	Control (n = 29)	hEDS (n = 270)	P-Value†	Conf.int
Deep breathing, beats/minute	16.69 (9.75)	14.45 (7.67)	.147	−0.79, 5.28
Valsalva ratio, beats/minute	1.60 (0.29)	2.20 (8.27)	.233	−1.60, 0.39
Valsalva maneuver, end of phase 2 decline, mmHg	−9.06 (10.03)	6.69 (14.83)	<.001	−19.92, −11.57
Electrochemical skin conductance, µS	82.96 (7.79)	79.75 (12.20)	.349	−3.52, 9.95
Electrochemical skin conductance, µS/kg	1.41 (0.21)	1.21 (0.33)	.029	0.02, 0.39
Epidermal nerve fiber density at proximal thigh, fibers/mm	14.12 (3.77)	12.13 (4.60)	.027	0.22, 3.76
Epidermal nerve fiber density at calf, fibers/mm	10.18 (2.17)	8.11 (4.00)	<.001	1.12, 3.01
Sweat gland nerve fiber density at proximal thigh, % of grid	58.96 (7.90)	51.99 (16.63)	.002	2.58, 11.36
Sweat gland nerve fiber density at calf, % of grid	51.87 (10.91)	45.57 (19.70)	.02	1.04, 11.55
Heart rate supine, beats/minute	76.52 (10.89)	77.52 (13.50)	.7	−6.11, 4.10
Heart rate orthostatic, beats/minute	92.21 (14.15)	97.21 (18.86)	.167	−12.11, 2.10
Systolic BP supine, mmHg	113.83 (9.26)	115.86 (12.57)	.399	−6.76, 2.70
Systolic BP orthostatic, mmHg	115.48 (11.48)	114.71 (13.60)	.769	−4.39, 5.93
Mean BP supine, mmHg	87.34 (7.46)	88.30 (9.30)	.592	−4.47, 2.56
Mean BP orthostatic, mmHg	89.52 (8.55)	90.78 (10.65)	.538	−5.28, 2.77
Diastolic BP supine, mmHg	74.10 (6.87)	74.52 (8.41)	.796	−3.60, 2.76
Diastolic BP orthostatic, mmHg	76.55 (7.90)	78.82 (9.84)	.232	−5.99, 1.45
Systolic CBFv supine, cm/sec	108.90 (10.39)	101.60 (17.18)	.002	2.88, 11.71
Systolic CBFv orthostatic, cm/sec	101.41 (11.87)	87.31 (17.38)	<.001	9.17, 19.04
Mean CBFv supine, cm/sec	68.12 (6.26)	65.01 (11.93)	.027	0.36, 5.86
Mean CBFv orthostatic, cm/sec	65.33 (8.09)	55.93 (12.42)	<.001	6.01, 12.79
Diastolic CBFv supine, cm/sec	47.72 (6.15)	46.76 (10.21)	.463	−1.65, 3.58
Diastolic CBFv orthostatic, cm/sec	47.28 (7.46)	40.24 (11.18)	<.001	3.93, 10.15
Mean CBFv corrected for end-tidal CO2 orthostatic, cm/sec	71.44 (8.28)	62.90 (14.23)	<.001	5.00, 12.10
Maximal decline in orthostatic mean CBFv, cm/sec	−5.32 (3.16)	−14.36 (7.91)	<.001	−10.55, −7.53
Maximal decline in orthostatic mean CBFv, %	−7.78 (4.42)	−21.85 (11.20)	<.001	11.94, 16.20
Respiratory frequency supine, breaths per minute	15.17 (4.08)	15.42 (5.00)	.795	−2.14, 1.64
Respiratory frequency orthostatic, breaths per minute	14.83 (2.42)	15.57 (5.14)	.178	−1.84, 0.35
End-tidal CO2 supine, mmHg	35.83 (4.09)	35.20 (3.90)	.412	−0.88, 2.14
End-tidal CO2 orthostatic, mmHg	32.52 (3.94)	30.76 (5.90)	.037	0.11, 3.40
Minimal end-tidal CO2 orthostatic, mmHg	31.86 (3.75)	27.73 (6.28)	<.001	2.53, 5.73
Maximal decline in orthostatic end-tidal CO2, mmHg	−3.97 (0.91)	−7.46 (5.33)	<.001	−4.22, −2.78
Maximal decline in orthostatic end-tidal CO2, %	−11.09 (2.18)	−21.32 (15.34)	<.001	8.23, 12.23
CVRi supine, mmHg/cm/sec	1.29 (0.17)	1.41 (0.31)	.004	−0.19, −0.04
CVRi orthostatic, mmHg/cm/sec	1.08 (0.19)	1.33 (0.39)	<.001	−0.34, −0.17
Cerebrovascular reactivity, %/mmHg	1.37 (0.93)	2.70 (2.93)	<.001	−1.82, −0.84
QASAT-CBFv, tilt response (assessing orthostatic decline)	0.00 (0.00)	4.66 (3.69)	<.001	−5.10, −4.21
QASAT-ET-CO2, tilt response (assessing orthostatic decline)	0.00 (0.00)	2.94 (3.73)	<.001	−3.38, −2.49
QASAT-autonomic failure	0.00 (0.00)	3.46 (2.50)	<.001	−3.76, −3.16
QASAT-cardiovagal	0.00 (0.00)	0.45 (0.71)	<.001	−0.53, −0.36
QASAT-adrenergic	0.00 (0.00)	1.13 (0.96)	<.001	−1.25, −1.02
QASAT-orthostatic hypotension	0.00 (0.00)	0.41 (1.24)	<.001	−0.56, −0.26
QASAT-orthostatic tachycardia	0.00 (0.00)	1.80 (3.01)	<.001	−2.16, −1.44
QASAT-sudomotor	0.00 (0.00)	1.47 (1.69)	<.001	−1.68, −1.26
QASAT-ENFD	0.00 (0.00)	1.57 (2.06)	<.001	−1.82, −1.33
QASAT-SGNFD	0.00 (0.00)	0.99 (1.77)	<.001	−1.22, −0.77

BP = blood pressure; CBFv = cerebral blood flow velocity; CVRi = cerebrovascular resistance; QASAT = Quantitative Scale for Grading of Cardiovascular Autonomic Reflex Tests and Small Fibers from Skin Biopsies; ENFD = epidermal nerve fiber density; SGNFD- sweat gland nerve fiber density.

Data are mean ± sd.

^†Calculated using t test.

Table 7.

Autonomic Testing Results, Frequency of Abnormal Findings

Variable	Control (n = 29)	hEDS (n = 270)	P-Value†
Orthostatic lightheadedness/dizziness	20.7	72.6	<.001
Orthostatic dyspnea	0.0	19.6	.018
QASAT-CBFv, reduced during the tilt, %	0.0	79.3	<.001
QASAT-ET-CO2, reduced during the tilt, %	0.0	53.0	<.001
QASAT-Autonomic failure, %	0.0	90.0	<.001
QASAT-Cardiovagal, %	0.0	34.4	<.001
QASAT-Adrenergic, %	0.0	69.3	<.001
QASAT-Orthostatic hypotension, %	0.0	14.8	.052
QASAT-Orthostatic tachycardia, %	0.0	38.5	<.001
QASAT-Sudomotor, %	0.0	57.1	<.001
QASAT-ENFD, %	0.0	54.1	<.001
QASAT-SGNFD, %	NA	33.3	NA
SFN, mixed, %	0.0	24.1	<.001
SFN, any from biopsy, %	0.0	63.8	1
SFN, any, %	0.0	81.7	.006
Postural tachycardia syndrome (POTS), %	0.0	32.6	<.001
Hypocapnic cerebral hypoperfusion (HYCH), %	0.0	21.5	<.001
Orthostatic cerebral hypoperfusion syndrome (OCHOS), %	0.0	18.1	<.001
Neurogenic orthostatic hypotension, %	0.0	8.9	.011
Orthostatic hypotension with normal orthostatic CBFv, %	0.0	12.5	.025

% = percent of abnormal findings.

^†Calculated using chi-squared test.

Figure 2.

QASAT results. (A) Absolute scores, mean ± sd (B), Relative scores in percent, mean ± sd (C), Percentage of patients in which the QASAT score was abnormal (> 0). CBFv = cerebral blood flow velocity in the middle cerebral artery; ET-CO₂ = end-tidal CO₂; AF = autonomic failure; Cardiov = cardiovagal, Adren = adrenergic; OH = orthostatic hypotension; OT = orthostatic tachycardia (orthostatic heart rate increment ≥ 30 BPM); Sudo = sudomotor; SFN = small fiber neuropathy; ENFD = epidermal nerve fiber density; SGNFD = sweat gland nerve fiber density; SFN-mixed=noth ENFD and SGNFD are abnormal; SFN-any-b=SFN with ENFD or SGNFD abnormal; SFN-any-b+=SFN with ENFD or SGNFD or sudomotor abnormality.

Figure 3.

The head-up tilt test shows hemodynamic variables at supine baseline and at every minute of head-up tilt, mean ± conf.int. (A) heart rate; (B) systolic blood pressure; (C) mean blood pressure; (D) diastolic blood pressure; (E) systolic cerebral blood flow velocity (CBFv), (F) mean CBFv; (G) diastolic CBFv; (H) end-tidal CO₂. CBFv = cerebral blood flow velocity in the middle cerebral artery.

Orthostatic systolic/mean/diastolic CBFv were lower in the hEDS group (P < .001). Head-up tilt test induced a greater decline in CBFv in hEDS compared to controls (−21.73 ± 11.17% vs −7.87 ± 4.28%, P < .001). The orthostatic cerebrovascular resistance index and cerebrovascular reactivity were higher in hEDS (P < .001). Orthostatic end-tidal CO₂ was lower in hEDS (P = .037).

Autonomic failure was detected in 90% of hEDS patients, none in controls (Table 7). The most common grade was mild abnormality (Figure 2, Table 8). In hEDS patients, all 3 domains (cardiovagal, adrenergic and sudomotor) were affected.

Table 8.

Autonomic Testing Results, Frequency and Severity of Abnormal Findings

Variable	Control (n = 29)	hEDS (n = 270)	P-Value†
CBFv, response to tilt none, %	100.0	20.7	<.001
mild, %	0.0	15.2
moderate, %	0.0	23.3
severe %	0.0	40.7
ET-CO2, response to tilt none, %	100.0	47.0	<.001
mild, %	0.0	14.1
moderate, %	0.0	13.0
severe %	0.0	25.9
Autonomic failure, total score none, %	100.0	10.0	<.001
mild, %	0.0	50.4
moderate, %	0.0	39.6
severe %	0.0	0.0
Cardiovagal none, %	100.0	65.6	<.001
mild, %	0.0	26.3
moderate, %	0.0	5.9
severe %	0.0	2.2
Adrenergic none, %	100.0	46.7	<.001
mild, %	0.0	38.1
moderate, %	0.0	14.4
severe %	0.0	0.7
Orthostatic hypotension none, %	100.0	85.2	<.001
mild, %	0.0	8.5
moderate, %	0.0	4.1
severe %	0.0	2.2
Orthostatic tachycardia none, %	100.0	61.5	<.001
mild, %	0.0	14.8
moderate, %	0.0	6.7
severe %	0.0	17.0
Sudomotor none, %	100.0	42.9	<.001
mild, %	0.0	16.4
moderate, %	0.0	33.2
severe %	0.0	7.5
ENFD none, %	100.0	45.9	<.001
mild, %	0.0	36.7
moderate, %	0.0	13.7
severe %	0.0	3.7
SGNFD none, %		66.7
mild, %		21.4
moderate, %		9.1
severe %		2.9

% = percent of findings.

^†Calculated using chi-squared test.

QASAT scores were higher in hEDS compared to controls (Tables 6, and 7). Abnormal orthostatic CBFv score was detected in 79.3% and ET-CO₂ in 53 % of hEDS patients.

Lightheadedness/dizziness during the head-up tilt test was more common in hEDS compared to controls (72.6% versus 20.7%, P < .001) (Table 7). Orthostatic CBFv (P = .015, confidence interval 0.10-0.96) and age (P = .038, −0.03 to 0.00) predicted orthostatic lightheadedness/dizziness but other variables not, including autonomic failure (absent versus present, P = .277, −0.24 to 0.83), abnormal end-tidal CO₂ (P = .526, −0.25 to 0.49), abnormal tachycardic heart rate response to head-up tilt (P = .365, −0.20 to 0.54) and orthostatic hypotension (P = .336, −0.24 to 0.76).

During the head-up tilt, POTS was detected in 32.6%, hypocapnic cerebral hypoperfusion in 21.5% and orthostatic cerebral hypoperfusion syndrome 18.1% of hEDS patients. Orthostatic hypotension was present in 14.8%, and neurogenic orthostatic hypotension in 8.9% of hEDS patients. Orthostatic hypotension with normal orthostatic CBFv (eg, without abnormal decline) was found in 12.5% of hEDS patients.

SFN was detected in 54.1% using symptoms and abnormal ENFD criterion, in 33.3% using symptoms and abnormal SGNFD, 63.8% using symptoms and abnormal SGNFD or ENFD criterion and in 81.7% using symptoms and abnormal skin biopsy or sudomotor testing criterion (Table 3). Mixed SFN (positive symptoms and abnormal both SGNFD and ENFD) was detected in 24.1% of hEDS patients.

Discussion

This study shows cerebrovascular dysregulation, small fiber neuropathy, and mild but widespread dysautonomia in hEDS.

Cerebrovascular Dysregulation

Previous Doppler studies have implicated the presence of cerebrovascular dysregulation and cerebral hypoperfusion. Reduced orthostatic blood flow was detected in both extracranial (internal carotid and vertebral arteries in patients with hypermobility²³) and intracranial (middle cerebral artery in patients with hEDS²¹) arteries. Our study provided additional evidence that cerebrovascular dysregulation and related cerebral hypoperfusion may play a major role in orthostatic intolerance in hEDS. We have found reduced orthostatic cerebral blood flow velocity in the majority (79.3%) of hEDS patients.

Transcranial doppler measures blood flow and not brain tissue perfusion. However, we found that reduced orthostatic cerebral blood flow velocity was associated with orthostatic lightheadedness, a major symptom of cerebral hypoperfusion.⁴⁹ That association reinforces the interpretation that reduced cerebral blood velocity in hEDS reflects cerebral hypoperfusion. The link between reduced cerebral blood flow and cerebral hypoperfusion is also consistent with the invasive study showing that reduced cerebral blood flow can be associated with cerebral hypoxia.⁵⁰ Nevertheless, the more direct perfusion studies are desirable to confirm our findings.

Our study adds hypocapnic cerebral hypoperfusion and orthostatic cerebral hypoperfusion syndrome to the orthostatic syndromes associated with hEDS. The common feature of POTS, hypocapnic cerebral hypoperfusion and orthostatic cerebral hypoperfusion syndrome is reduced orthostatic cerebral blood flow velocity and the absence of orthostatic hypotension. In POTS and hypocapnic cerebral hypoperfusion , reduced cerebral blood flow velocity results from hypocapnia-induced cerebral arteriolar vasoconstriction.⁴⁵ Orthostatic cerebral hypoperfusion syndrome differs from hypocapnic cerebral hypoperfusion and POTS by absent hypocapnic hyperventilation, hence the reduced orthostatic cerebral blood flow velocity results from the dynamic cerebral autoregulatory failure due to lack of compensatory orthostatic arteriolar vasodilation.⁴⁶ Dynamic cerebral autoregulation refers to high-pass filtering of cerebral blood flow in response to fast blood pressure changes induced by an upright position.⁵¹ Cerebral vasodilatory responses are determined by endothelial, myogenic, and metabolic factors, and to a lesser extent, by autonomic innervation.⁵² Orthostatic cerebral hypoperfusion syndrome can then be explained by an increased cerebral arteriolar stiffness caused by abnormal myogenic responses in the small cerebral vessels due to collagenosis in hEDS.⁵³ Why some patients have orthostatic cerebral hypoperfusion syndrome while others POTS/hypocapnic cerebral hypoperfusion, needs to be explored in further studies.

Autonomic Failure

Dysautonomia, a generic term for autonomic dysfunction, is frequently encountered in hEDS. For example, one study found that 74% of patients wih hEDS reported symptoms of orthostatic intolerance, 41% were diagnosed with POTS using head-up tilt testing, and 65% had evidence of distal sudomotor dysfunction.^54,55 Another study showed that 49% of hEDS patients met the POTS criteria, while orthostatic hypotension was detected in only 1 patient out of 35.¹⁹ Our study detected autonomic failure, a particular form of dysautonomia associated with a decline in autonomic functions, in 90% of hEDS patients, affecting several autonomic domains simultaneously, including parasympathetic, adrenergic, and sudomotor. Autonomic failure was mild in most of our patients. Neurogenic orthostatic hypotension, a marker of more advanced sympathetic adrenergic failure, was found only in 8.9% of our patients, which is consistent with previous studies.

Small Fiber Neuropathy

It was reported that SFN affects between 61%-100% of hEDS patients16, 17, 18 using either skin biopsies with ENFD or sudomotor measurements. Our results are similar to previous studies using either the ENFD or sudomotor criterion. Our study provides additional evidence of a link between hEDS and SFN. The new finding is the direct structural evidence of sudomotor autonomic fibers degeneration as reflected by abnormal SGNFD in 33.2% of hEDS patients, that is in addition to ENFD and sudomotor abnormalities. These results support previous observations indicating that the loss of small fibers might be a pathological substrate of autonomic failure and neuropathic pain,16, 17, 18 probably accompanied by central sensitization.¹⁵

An association between hEDS and mast cell disorders has been made in previous studies⁵⁶ but the true prevalence of mast cell disorders in hEDS remains unclear due to differing clinical criteria used in the definition of the mast cell disorders.57, 58, 59 In our study, 61% of patients used anti-histamine medications, and 41% had a history of symptoms associated with mast cell activation syndrome. The relationship between hEDS and mast cell activation syndrome should be evaluated in further studies as mast cell activation may damage small fibers. Mast cells localize to the connective tissue encircling and protecting small fibers, and their aberrant activation, resulting in release of vasoactive and inflammatory collagen damage, may lead to SFN.^60,61

Limitation

A referral bias may affect the selection of subjects and, therefore, the studied population may not be representative of the whole hEDS population. However, a relatively large number of hEDS patients that we have studied may still present a representative sample. The co-morbidities were obtained primarily from chart review, which can introduce an ascertainment bias. Nevertheless, the best effort has been made to verify the accuracy of the diagnoses with the study patients at the time of testing. Although the control healthy group was relatively small compared to the hEDS group, the study utilized validated tests with established normal and abnormal ranges. Therefore, the results are considered valid despite the smaller control sample. Other biases may influence the results. The cerebral blood flow was assessed indirectly using transcranial Doppler which measures flow velocity instead of blood flow. The blood flow velocity depends on the number of parameters⁶² including the diameter of the insonated vessel. The imaging study suggests that the middle cerebral artery diameter, the vessel we insonated in our study, does not change during orthostatic stress⁶³ and, therefore, blood flow velocity is a good proxy of flow. Transcranial Doppler is also sensitive to the angle of the Doppler probe. Although the actual angle depends on the skull anatomy, once we found the insonated vessel, the probe was kept at a constant angle throughout the testing using a 3-directional holder and therefore the blood flow changes induced by the tilting are deemed accurate. Another important variable that can affect the cerebral blood flow velocity is hematocrit. Although we did not adjust for the hematocrit, none of our subjects had clinically significant anemia that could explain orthostatic intolerance.

Conclusions

In summary, hEDS is a complex disorder with multisystem involvement. The most significant finding of this study is the presence of cerebrovascular dysregulation associated with reduced cerebral blood flow that can result in cerebral hypoperfusion, orthostatic intolerance, and impaired well-being. This is a potentially modifiable treatment target that may improve the quality of life of affected individuals. Pharmacological treatment options, however, may require further investigations. Our study provides additional evidence that SFN is an underlying cause of both dysautonomia and certain types of chronic pain in hEDS.

CRediT authorship contribution statement

Peter Novak: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Software, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. David M. Systrom: Writing – review & editing, Data curation. Sadie P. Marciano: Writing – review & editing, Data curation. Alexandra Witte: Writing – review & editing. Arabella Warren: Writing – review & editing. Donna Felsenstein: Writing – review & editing. Matthew P. Giannetti: Writing – review & editing. Matthew J. Hamilton: Writing – review & editing. Jennifer Nicoloro-SantaBarbara: Writing – review & editing. Mariana Castells: Writing – review & editing. Khosro Farhad: Writing – review & editing. David M. Pilgrim: Writing – review & editing. William J. Mullally: Writing – review & editing. Mark C. Fishman: Writing – review & editing. Jeff M. Milunsky: Writing – review & editing, Data curation. Aubrey Milunsky: Writing – review & editing, Data curation. Joel Krier: Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.