Dysregulated Cardiovascular Autonomic Responses to Orthostatic Challenge in Mild Traumatic Brain Injury

Introduction

Mild traumatic brain injury (mTBI; aka concussion) can present with symptoms of dizziness, headache, and brain fog.^1,2 Impairment of the cardiovascular autonomic nervous system has been proposed as a potential mechanism for such symptoms.^1–5 More specifically, symptoms of dizziness or headache that are provoked during physiological stress (such as exercise and upright activity) have been attributed to a so-called “autonomic” symptomatic subtype^6,7—an emerging clinical profile that has recently been added to the previously established subtype framework^8,9—presumed to reflect impaired cerebral blood flow and/or autoregulation,^10,11 perhaps via arterial baroreflex function.^12–16 However, while prior work has indeed demonstrated altered baroreflex function after concussion,^{12,13,15–18} the variable methodology and inclusion criteria of prior work has limited the certainty of these associations,^17,19 and to our knowledge none have directly evaluated orthostatic control of heart rate (HR) and blood pressure (BP) during head-up tilt table testing (HUT) in a relatively acute, adult cohort of post-mTBI participants.

A key advantage of HUT testing is its role as a “stress test” of orthostatic tolerance, with direct clinical implications relative to treatment approaches and symptom burden during activity and other upright behaviors. When paired with heart rate deep breathing (HRDB) and the Valsalva maneuver—the resulting testing battery represents the current clinical standard for evaluation of cardiovascular autonomic function. This battery helps isolate distinct aspects of autonomic physiology, where HRDB evaluates cardiac parasympathetic (vagal) function via the respiratory sinus arrhythmia; the Valsalva maneuver probes both sympathetic and parasympathetic baroreflex-mediated responses through defined hemodynamic phases; and HUT specifically challenges the sympathetic arm of the baroreflex by imposing a sustained orthostatic stress. Despite prior studies evaluating orthostatic tolerance in sport-related and postconcussive populations, most have focused on pediatric/young adult sport concussion and/or relied mainly on active standing testing, as opposed to formal HUT testing.^{5,14–16,18,20–23} To the best of our knowledge, no prior work has evaluated tilt table responses in relatively acute, adult mTBI. Notably, of the few that did evaluate adults, these were typically focused on a very broad presenting sample (ranging from acute to those with more chronically persisting symptoms after concussion),^14,16,23 limiting our understanding of whether the reported changes could be interpreted as an acute versus chronic physiological feature following mTBI.

The purpose of this prospective, observational cohort study was to evaluate orthostatic responses to HUT testing, using beat-to-beat monitoring, alongside standardized cardiovascular autonomic testing in a cohort of adult, all-comer participants with symptomatic mTBI within 14 days of injury. Building from prior work, we hypothesized that mTBI participants would exhibit altered sympathetically mediated responses to baroreflex stimuli, compared to controls. Additionally, we examined the relationship between cardiovascular autonomic control and symptoms to test the prevailing, yet unresolved, premise that mTBI-related symptom burden is associated with underlying autonomic impairment.^1,5,19,24

Materials and Methods

Results

A total of 70 participants, 35 with mTBI and 35 healthy control subjects, completed the symptom and laboratory assessments between February 2021 and January 2025 (19 participants, 6 controls and 13 mTBI did not complete a VM due to restrictions around this procedure during the COVID-19 pandemic). The mechanism of concussion injury is categorized by the four most common types of contextual situations leading to a concussion injury. ⁵⁵ Six participants did not disclose their mechanism of injury. There were no statistically significant differences in sex or age between groups, nor in BP/HR at baseline (supine) or during HUT. However, postconcussion symptom severity score (RPQ) and previous lifetime history of mTBI and symptom scores were significantly higher in the mTBI group (<0.001). (Table 1).

OPEN IN VIEWER

Table 1.

Clinical Characteristics of Sample

Participant characteristics	mTBI	Control	p
Number of participants	35	35
Female, n (%)	18 (51)	20 (57)	0.63
Age, median (IQR; range)	28 (8.5; 20–51)	26 (11; 18–68)	0.48
Days since injury, mean (SD), days	8.7 (3.4)	N/A
Rivermead Post-Concussion Symptom Severity, median (IQR)	21 (20)	0 (1)	<0.001*
Previous lifetime mTBI, median (IQR; range)	2 (2.5; 0–10)	0 (0; 0–2)	<0.001*
Supine systolic BP, mean (SD), mmHg	114 (15)	116 (16)	0.56
Supine mean BP, mean (SD), mmHg	85 (12)	87 (14)	0.62
Supine diastolic BP, mean (SD), mmHg	68 (13)	69 (13)	0.29
Supine mean HR, mean (SD), bpm	63 (11)	63 (12)	0.75
HUT systolic BP, mean (SD), mmHg	121 (23)	124 (27)	0.51
HUT mean BP, mean (SD), mmHg	93 (22)	97 (17)	0.38
HUT diastolic BP, mean (SD), mmHg	79 (15)	84 (20)	0.10
HUT mean HR, mean (SD), bpm	82 (12)	85 (13)	0.42
Mechanism of concussion injury, n (%)
Fall	6 (17)
Motor vehicle accident	4 (11)
Head impact with object	5 (14)
Sport	11 (32)
Other	3 (9)
Not disclosed	6 (17)

*Statistically significant difference.

BP, blood pressure; bpm, beats per minute; HR, heart rate; HUT, head-up tilt; IQR, interquartile range; mTBI, mild traumatic brain injury; SD, standard deviation.

The results of standardized autonomic testing and tilt table testing are summarized in Table 2. Supine-based HR and BP variability (SD and LF power) did not differ between groups, nor did MHRR, Valalva ratio, or total HF power of supine HR variability. However, HF power of HR during HUT did trend lower in mTBI than controls (p = 0.02). Total LF power in the HUT position for HR and BP was significantly reduced in individuals with mTBI compared to controls (p < 0.013; see Table 2 for summary).

OPEN IN VIEWER

Table 2.

Measures of Autonomic Function

Measures of autonomic function	mTBI group	Control group	p
HR/BP variability
Supine systolic BP, median (IQR), mmHg	6.0 (2.2)	6.3 (4.1)	0.52
Supine mean BP, median (IQR), mmHg	4.9 (1.7)	5.0 (2.3)	0.30
Supine diastolic BP, median (IQR), mmHg	4.8 (2.0)	5.1 (2.9)	0.11
Supine HR, median (IQR), bpm	4.2 (2.2)	5.0 (3.6)	0.40
HUT systolic BP, median (IQR), mmHg	8.2 (3.5)	10.5 (5.1)	0.02
HUT mean BP, median (IQR), mmHg	6.0 (2.2)	8.3 (3.9)	0.003 *
HUT diastolic BP, median (IQR), mmHg	5.8 (1.9)	7.8 (3.5)	0.005 *
HUT HR, median (IQR), bpm	5.9 (3.5)	8.7 (4.5)	0.003 *
Measures of sympathetic function
Supine systolic BP LF, median (IQR)	5.5 (6.1)	7.0 (10.1)	0.74
Supine mean BP LF, median (IQR)	4.3 (4.2)	6.1 (5.3)	0.25
Supine diastolic BP LF, median (IQR)	3.0 (3.8)	4.8 (6.3)	0.39
Supine HR LF, median (IQR)	6.0 (6.4)	6.6 (12.9)	0.20
HUT systolic BP LF, median (IQR)	14.0 (17.9)	22.2 (30.9)	0.004 *
HUT mean BP LF, median (IQR)	11.7 (8.3)	16.7 (15.7)	0.002 *
HUT diastolic BP LF, median (IQR)	9.5 (7.3)	15.8 (15.9)	0.002 *
HUT HR LF, median (IQR)	10.9 (18.5)	25.6 (34.4)	0.011 *
Barosensitivity (BRSa), median (IQR)	37.9 (25.9)	39.1 (28.8)	0.49
Measures of parasympathetic function
Supine HR HF, median (IQR)	3.6 (7.1)	4.1 (6.9)	0.67
HUT HR HF, median (IQR)	2.6 (5.3)	5.2 (8.6)	0.02
MHRR, mean (SD)	20.9 (7.5)	23.4 (8.2)	0.19
Valsalva ratio, mean (SD)	2.1 (0.3)	2.2 (0.4)	0.24

Total power is reported for LF and HF domains; the unit for HR calculations is bpm² and BP is mmHg². HR/BP variability was quantified by standard deviation.

*

Statistically significant difference, based on p < 0.013.

BP, blood pressure; BRSa, adrenergic barosensitivity; HF, high frequency; HR, heart rate; HUT, head-up tilt; LF, low frequency; MHRR, mean heart rate range; SD, standard deviation.

To examine whether group-wise differences in HR and mean BP variability (SD) were associated with altered baroreflex function, estimated by LF total power, Spearman’s rho test was used. Here, we found a significant positive correlation between HR variability (SD) and total LF power in HR and mean BP during HUT (0.88 and 0.51, respectively; p < 0.001). This was confirmed using post hoc linear regression, controlling for age, sex, baseline systolic BP, lifetime mTBI history, and symptom score as covariates. The resulting model testing differences in LF power (of systolic BP) between mTBI and controls remained significant (Analysis of Variance [ANOVA] p = 0.001; 95% confidence interval 112.90–125.18).

Finally, to evaluate the relationship between postconcussive symptom scores (RPQ) and our primary physiological outcomes (LF power of BP/HR during HUT) across the full sample of controls and mTBI, Spearman’s rho revealed weak, nonsignficant correlations, between RPQ and LF of mean BP during HUT, as well as RPQ and LF of HR during HUT (rho −0.17 and −0.26 respectively for RPQ), as well as weak associations between secondary outcomes of BP/HR SD during HUT (rho −0.19 to −0.31). When evaluated specifically within the mTBI group of interest, correlations remained weak and nonsignficant (see Fig. 1).

OPEN IN VIEWER

FIG. 1.

Rivermead postconcussion symptom scores were only weakly associated with primary autonomic measures during HUT. (A) LF of mean BP during HUT, Spearman’s rho 0.26 (p = 0.07). (B) LF of HR during HUT, rho 0.02 (0.46). BP, blood pressure; LF, low frequency; HR, heart rate; HUT, head-up tilt; MBP, mean BP; RPQ-13, Rivermead Post-Concussion Questionnaire.

Discussion

The results of this study support previously published work showing abnormal autonomic function in mTBI by providing additional evidence of abnormal responses to orthostatic challenge and reduced HR and BP variability (and related LF and HF power) in individuals with symptomatic mTBI.^13,48,53 These data advance earlier reports by demonstrating blunted sympathetically mediated baroreflex responses to HUT—lower total LF power of HR and BP during HUT in individuals with mTBI compared to controls—in addition to reduced HR/BP variability.^13,48,53 Such state-dependent abnormalities further implicate deficiencies in stressor-evoked sympathetic (and to a lesser extent vagal) output, demonstrated in prior work, while confirming that these findings are indeed present during the early post injury phase (≤14 days) and in response to upright challenge (the latter being ecologically relevant to day-to-day activity).^{12–16,18,56,57}

Notably, while the current study did not find group differences in standardized cardiovascular autonomic testing metrics (namely, measures of BRSa, MHRR, VR; Table 2), our results are consistent with prior work suggesting that the most robust differences (mTBI vs. controls) are observed in response to stressor-evoked provocation.^{12–16,18,56,57} HUT was intentionally selected as the test of primary interest because it provides both a standardized and physiologically relevant orthostatic stressor, mimicking the postural transitions of daily activity (e.g., standing, walking). Orthostatic intolerance—including symptoms upon postural change—has been reported at higher rates in concussed adolescents compared to controls. Haider et al. found that concussed adolescents had smaller changes in HR when moving from supine to standing (n = 297 mTBI, 214 controls), consistent with cardioautonomic dysfunction upon postural change. ¹⁴ Heyer et al. similarly demonstrated orthostatic intolerance in youth with persistent postconcussion symptoms using HUT. ¹⁸ Thus, we hypothesized that abnormal arterial baroreflex-mediated control of HR and BP would be present. In support of this premise, Leddy et al. proposed that ANS dysfunction following concussion impairs the regulation of cerebral blood flow during exercise, with inappropriately elevated arterial CO₂ levels driving CBF out of proportion to exercise intensity—a mechanism that may also be operative during passive upright challenge. ⁵⁸ Our study advances prior work by expanding into a more generalizable patient population (beyond sports concussion in young adults/adolescents), while still limiting the scope to the subacute period post injury.

Throughout the literature, HR variability has been widely used to investigate the “balance” between parasympathetic or sympathetic mechanisms of cardiovascular autonomic regulation.^48,53 Reductions in HR variability, as well as in total LF and HF power, have been associated with constrained dynamic range and flexibility in autonomic responses relative to daily life, reducing physiological adaptability and resilience. Therefore, the decrease in HR/BP variability in the HUT position observed in the current study, combined with the significant differences in HUT-LF power in mTBI participants, supports our initial hypothesis of impaired autonomically mediated responses to orthostatic challenge. Such changes could plausibly contribute to altered cerebral blood flow after mTBI, which has been shown both acutely and chronically, and to hypoperfusion of the brain, potentially explaining some of the symptom burden seen during everyday tasks/daily routines.^11,59–61 Future studies might pair targeted physiological testing of cerebral blood flow with sympathetic autonomic evaluation to further confirm this relationship.^22,62

Although an individual’s symptom profile is often used to categorize impairment and guide treatment after mTBI,^30,63–65 we found relatively weak relationships between measures of HR/BP variability and overall symptom scores in our cohort (with nonsignificant correlations ranging from −0.17 to −0.26). A similar finding was also demonstrated between mean change in HR during HUT and symptom score. ²⁰ Together, these findings support the widely held view that multiple physiological impairments likely contribute to clinically overlapping symptom profiles after mTBI. ⁶⁶ This is critically important within clinical care as treatment protocols are typically based on the categorization of symptoms after mTBI. The understanding that multiple, potentially overlapping physiological impairments can contribute to symptom profiles should help clinicians to avoid over-isolating targeted treatment, and rather, guide them to assess the performance of multiple physiological systems regardless of symptom profile. For example, an mTBI patient where dizziness and headache are prominent, and who is not responding to vestibular rehab alone, may require additional orthostatic retraining in parallel (in addition to treatment of balance/oculomotor dysfunction). Future studies might attempt to parse domain-specific symptom assessments to differentiate variable sources of “dizziness” and related symptoms that are currently beyond the scope of the RPQ to discern (e.g., orthostatic vs. vestibular vs. proprioceptive symptoms), combined with mechanism-targeted physiological testing, to better understand the contributions of impaired autonomic function to clinical symptomatology.^22,62

To our knowledge, there are few, if any, prior studies that have reported complete standardized autonomic testing in mTBI participants compared to controls. Interestingly, a recent study that included moderate and severe traumatic brain injury (TBI) patients similarly failed to show definite group-wise differences on standard autonomic tests (e.g., HUT, Valsalva, and HRDB measures). ⁶⁷ Given those findings in more severe TBIs, the lack of abnormalities on standard laboratory autonomic measures in individuals with mTBI is perhaps not entirely surprising. It is important to note that the clinical battery of cardiovascular autonomic tests was chiefly developed to evaluate syndromes of autonomic failure. Thus, the standard clinical battery, and related clinical normative data ranges, may be insensitive to lesser injuries. ⁶⁸ Furthermore, the most widely used standardized autonomic testing battery (commonly referred to as the Ewing battery) is most sensitive and specific in the assessment of lesion-based diseases such as autonomic neuropathy and/or pure autonomic failure, where the neuroanatomical distribution of the pathology is fairly discrete (in contrast to the presumed network dysfunction that is present in TBI).

Conclusion

A blunted sympathetic response to orthostatic challenge (HUT), where a robust increase in sympathetic activity is expected, may contribute to symptoms after mTBI. Clinically, caution should be taken when using symptom reports in isolation, as commonly used symptom scores may lack the capacity to localize cardiovascular autonomic impairment after mTBI and to inform clinical treatment plans.

Transparency, Rigor, and Reproducibility Statement

Sample size was 70 participants, 35 with mTBI and 35 healthy control subjects, based on a priori power analysis using the effect size results in Dobson et al. for standing systolic BP (Cohen’s d = 0.86), at 80% power with an α = 0.05, a required group sample size of 44 subjects was needed to make group comparisons. All 70 participants completed study questionnaires and standardized autonomic testing within 14 days after injury (19 participants, 6 controls, and 13 mTBI did not complete a VM due to restrictions around this procedure during the COVID-19 pandemic). Further analysis of sympathetic and parasympathetic contributions to HR and BP variability during supine and HUT states was performed to identify low (LF) and HF domains via continuous wavelet transform. All data were screened for the assumptions of parametric statistical tests and reported according to the most appropriate metric based on normality. Accordingly, between-group comparisons were made using Student’s t-test or Mann–Whitney U tests, and variables were reported as mean/median and SD/interquartile range, respectively. Data from this study are available upon request.

Authors’ Contributions

R.P.: Conceptualization, methodology, formal analysis, data curation, and writing—original draft. G.L.: Data curation and writing—reviewing and editing. C.M.: Investigation, data curation, and writing—reviewing and editing. P.C.F.: Methodology, software, formal analysis, funding acquisition, and writing—reviewing and editing. L.E.D.: Conceptualization, methodology, and writing—reviewing and editing. D.J.P.: Writing—reviewing and editing. J.S.K.: Writing—reviewing and editing. M.M.C.: Conceptualization, methodology, formal analysis, validation, supervision, and writing—reviewing and editing.