OUP user menu

Carotid sinus hypersensitivity: disease state or clinical sign of ageing? Insights from a controlled study of autonomic function in symptomatic and asymptomatic subjects

Maw Pin Tan, Rose Anne M. Kenny, Tom J. Chadwick, Simon R.J. Kerr, Steve W. Parry
DOI: http://dx.doi.org/10.1093/europace/euq317 1630-1636 First published online: 7 September 2010


Aims This study sought to improve the currently limited understanding of the pathophysiology of carotid sinus hypersensitivity (CSH) by comparing autonomic function measured by heart rate variability (HRV) and baroreflex sensitivity inpatients with symptomatic CSH and asymptomatic individuals with and without CSH.

Methods and results Twenty-two patients with symptomatic CSH, 18 individuals with asymptomatic CSH, and 14 asymptomatic older individuals without CSH were recruited to our study. Non-invasive measurements of heart rate and blood pressure were obtained during 10 min of supine rest. Low frequency (LF), high frequency (HF), and total power spectral density (PSD) for HRV were determined using the autoregressive method. The baroreflex slope (BRS) and baroreflex effectiveness index (BEI) were determined using the sequence method for baroreflex sensitivity. There were significant increases in the LF-HRV (P = 0.014), total PSD (P = 0.031), LF:HF (P = 0.047), normalized (nu) LF-HRV (0.049), down ramp BEI (P = 0.017), and total BEI (P = 0.038) in the symptomatic CSH group compared with non-CSH controls. The asymptomatic CSH group had significantly higher LF-HRV (P = 0.001), total PSD (P = 0.002), nuLF-HRV (P = 0.026), and LF:HF (P = 0.030), as well as up, down, and total BRS (P = 0.012, P = 0.015, and P = 0.011, respectively) and BEI (P = 0.049, P = 0.001, and P = 0.006, respectively) than non-CSH control participants.

Conclusion This study has demonstrated an association between CSH with increased resting sympathetic activity and baroreflex sensitivity regardless of the presence of symptoms, indicating the presence of autonomic dysregulation in individuals with CSH. Our findings therefore suggest that CSH is part of a generalized autonomic disorder but do not differentiate between asymptomatic and symptomatic individuals.

  • Carotid sinus hypersensitivity
  • Syncope
  • Autonomic nervous system
  • Heart rate variability
  • Baroreflex sensitivity


Carotid sinus hypersensitivity (CSH) is characterized by at least 3 s of asystole and/or a systolic blood pressure (SBP) reduction of 50 mmHg or greater in response to carotid sinus massage (CSM), and is commonly associated with syncope, unexplained falls and drop attacks in older adults.13 It was first described by Roskam in 1930, and was previously assumed to be rare.4 This condition is now being diagnosed with increasing frequency alongside the development of specialist syncope units with recent reports of a positivity rate of 14–54% of individuals investigated with CSH.57 Individuals diagnosed with CSH for symptoms of falls or syncope have reported high injury and fracture rates of 47 and 25%, respectively5,8 in comparison with those diagnosed with vasovagal syncope, another common neurally mediated condition.9 There also appears to be a high prevalence of CSH in individuals with dementia.10

The pathophysiological processes underlying CSH remain poorly understood. Polvikovsky et al.11 recently published a case report of a neurodegenerative process centred on the medullary autonomic nuclei in an individual with CSH. This finding has since been substantiated by Miller et al.12 who found an increase in hyperphosphorylated-tau protein deposition in baroreflex-associated medullary nuclei in individuals with CSH. The above evidence suggests the possibility that CSH is associated with a neurodegenerative process and may be part of more widespread autonomic dysfunction, rather than a localized disorder within the carotid sinus. In order to investigate the possibility that CSH may be associated with a generalized autonomic disorder, we conducted a controlled assessment of heart rate variability (HRV) and baroreflex sensitivity inpatients with symptomatic CSH, asymptomatic individuals with CSH, and asymptomatic controls demonstrably without CSH.



Consecutive older patients diagnosed with symptomatic CSH during their routine investigations for unexplained falls, drop attacks, and syncope at our specialist syncope facility were invited to participate in the study. Asymptomatic control participants without a previous history of syncope, presyncope, or unexplained falls were recruited from a predefined group of community-dwelling elders recruited for a recent study by our group on the community prevalence of CSH.13 Carotid sinus hypersensitivity was diagnosed using a previously published and widely accepted protocol.14 In brief, CSM was performed for five seconds in a bilateral sequential fashion, beginning on the right, followed by the left in the supine position, and repeated in the 70° head-up tilt position.

Individuals who were unable to provide informed consent or unable to discontinue cardiac medications for clinical reasons were excluded. Patients with permanent cardiac pacemakers were also excluded as cardiac pacing may lead to inaccuracies in the interpretation of HRV and baroreflex sensitivity.

Clinical assessment of autonomic function

All participants were asked to refrain from smoking, caffeine ingestion, and heavy meals on the day of the investigations. Cardiac medications were discontinued for five half-lives before the investigations. All measurements were conducted between the hours of 12.00 p.m and 2.00 p.m.

Continuous electrocardiogram (ECG) and beat-to-beat blood pressure measurements were recorded for 10 min in the supine position during spontaneous breathing. Beat-to-beat systolic, mean and diastolic blood pressure was measured with a vascular unloading device (Task Force™ Monitor, CNSystems, Austria). Artefact and ectopic-free segments of at least 256 beats in length were analysed. Power spectral plots for individual participants were also visualized to identify potential outliers. The power spectral densities were calculated using the autoregressive method to obtain the power in very low frequency (VLF): ≤0.04 Hz, low frequency (LF): 0.04–0.15 Hz, high frequency (HF): 0.15–0.4 Hz, and total power spectral density (PSD): <0.40 Hz.15 The normalized values (nu) for the LF and HF bands were then calculated using the predefined formulae: nuLF = LF/(PSD − VLF) × 100 or nuHF = HF/(PSD − VLF) × 100. The ratio for the absolute values for LF-HRV and HF-HRV (LF:HF) was also calculated.

Baroreflex sensitivity was determined using the sequence method using artefact-free segments of continuous heart rate and blood pressure recordings of at least 5 min in length.1617 The sequence method involves computerized detection of up sequences, which are sequences of increases in SBP corresponding to lengthening of the R–R interval over three or more consecutive heart beats and down sequences or decreases in SBP corresponding to shortening of the R–R interval. The slope of regression or baroreflex slope (BRS) was determined by the change in R–R interval in milliseconds (ms) over the change in SBP in millimetres of mercury (mmHg). The difference in SBP between each cardiac cycle was paired with R–R interval at which the change in SBP occurred (lag 0). The baroreflex effectiveness index (BEI) was also calculated as the ratio between the number of SBP ramps associated with corresponding changes in R–R interval within two heart beats (lag 0, lag 1, and lag 2) to the total number of SBP ramps.18 For further clarification, any increase or decrease in SBP over three or more consecutive heart beats is known as a ramp, while the SBP ramps corresponding to appropriate changes in heart rate are known as sequences. The mean BRS and BEI for up ramps and down ramps are presented separately, as well as cumulatively as total ramps.

A favourable ethical opinion had been obtained from the Newcastle and North Tyneside Local Research Ethics Committee 2 prior to commencement of the study. Written informed consent was obtained from all participants.

Statistics and data analysis

Baseline demographics and haemodynamic indices were compared between groups with analysis of variance for continuous data, and χ2 test for categorical data. A P-value of <0.05 was considered statistically significant. Comparisons were made between non-CSH controls vs. asymptomatic CSH groups, and non-CSH controls vs. symptomatic CSH groups using dummy variables. Hierarchical linear regression analyses were performed with forced entry for the dummy variables, and stepwise entry for the potential confounders of baseline heart rate, blood pressure, age, gender, and medical history. The dependent variables of LF-HRV, HF-HRV, total-HF, and LF:HF were non-normally distributed and therefore natural logarithmically (ln) transformed before further analyses. The natural logarithmic mean with 95% confidence intervals (CIs) were presented for the transformed variables. The remaining dependent variables of nuLF-HRV, nuHF-HRV, baroreflex slope, and BEI were normally distributed and therefore analysed untransformed. The B coefficient represents the estimated difference of the adjusted mean between groups. All statistical analyses were performed using SPSS version 15.0.



Ninety-seven patients were diagnosed with CSH with a positive response to CSM at a specialist falls and syncope service from 1 January 2004 to 30 June 2008. Forty-five (46%) could not be contacted, met exclusion criteria, or had died. Fifty-two (54%) potential participants were contacted about the study of which 21 (40%) agreed to take part. Thirty-two participants were recruited from the community-dwelling cohort of older people with established CSH status. Eighteen (56%) participants fulfilled the diagnostic criteria for CSH following CSM, and the remaining 14 (44%) participants did not have CSH. The baseline characteristics of subjects in the three groups are summarized in Table 1. There were statistically significant differences between the three groups in the use of lipid-lowering drugs (P = 0.039) and baseline heart rate (P = 0.004). There were no other significant differences group characteristics. Eight (36%) participants from the symptomatic CSH group had asystole ≥3 s (cardioinhibitory) during CSM and the remaining 14 (64%) participants had vasodepressor responses, while 9 (50%) of participants had the cardioinhibitory subtypes and 9 (50%) had the vasodepressor subtype for the asymptomatic CSH group.

View this table:
Table 1

Characteristics of Participants

CharacteristicsSymptomatic CSH (n = 22)Asymptomatic CSH (n = 18)Non-CSH controls (n = 14)P-valuea
Age (years), mean (95% CI)73.0 (68.1,77.8)77.7 (75.6,79.7)76.6 (73.3,80.0)0.154b
Male sex, n (%)11 (50)14 (78)8 (57)0.188
Medical History, n (%)
 Atrial Fibrillation2 (9)1 (6)2 (14)0.699
 Arthritis6 (27)9 (50)8 (57)0.155
 Asthma/COPD3 (14)3 (17)3 (21)0.829
 TIA or stroke3 (14)1 (6)1 (7)0.647
 Diabetes1 (5)1 (6)00.685
 Hypertension9 (41)9 (50)5 (36)0.705
 Thyroid disease2 (9)1 (6)1 (7)0.913
 Angina5 (23)5 (28)4 (29)0.905
 Myocardial infarction2 (9)4 (22)4 (29)0.302
 Malignancy3 (14)2 (11)00.367
 Osteoporosis2 (9)03 (21)0.116
Medications, n (%)
 Psychoactive drugsc2 (9)2 (11)1 (7)0.928
 Antidepressants3 (14)1 (6)2 (14)0.655
 α-adrenoceptor antagonists2 (9)000.211
 β-adrenoceptor antagonists2 (9)2 (11)5 (36)0.083
 ACE-inhibitors7 (32)1 (6)4 (29)0.111
 Angiotensin-II antagonists2 (9)2 (11)1 (7)0.928
 Calcium-channel antagonists5 (23)6 (33)4 (29)0.755
 Diuretics6 (27)4 (22)3 (21)0.900
 Lipid-lowering drugs13 (59)4 (22)4 (29)0.039
 Steroids2 (9)000.221
 Proton-pump inhibitors5 (23)3 (17)2 (14)0.792
Haemodynamic indicesd, mean (95% CI)
 Heart rate, bpm75 (70,79)67 (62,71)63 (56,71)0.004b
 Systolic blood pressure, mmHg126 (118,135)130 (122,138)126 (116,136)0.737b
 Diastolic blood pressure, mmHg81 (74,88)85 (79,90)79 (70,88)0.522b
  • aχ2 test unless otherwise indicated.

  • bAnalysis of variance.

  • cIncludes antipsychotics and tranquilizers.

  • dDuring 10 min of supine rest.

  • CSH, carotid sinus hypersensitivity; TIA, transient ischaemic attacks; COPD, chronic obstructive pulmonary disease; ACE, angiotensin converting enzyme.

Heart rate variability

Heart rate variability data were excluded for eight participants, five for atrial fibrillation, and three for multiple ectopics. The natural logarithmic means are presented for LF-HRV, HF-HRV, total PSD, and LF:HF for each group, together with the actual means for nuLF-HRV and nuHF-HRV (Table 2).

View this table:
Table 2

Unadjusted Means with 95% confidence interval for heart rate variability measures

Symptomatic CSH (n = 20)Asymptomatic CSH (n = 15)Non-CSH controls (n = 11)
lnLF-HRV (95% CI), ms24.21 (3.76,4.67)4.79 (4.23,5.35)4.08 (3.43,4.71)
lnHF-HRV (95% CI), ms23.59 (3.06,4.12)4.29 (3.78,4.79)3.96 (2.96,4.96)
lnPSD (95% CI), ms25.06 (4.63,5.49)5.76 (5.31,6.20)5.28 (4.55,6.01)
lnLF:HF, (95% CI)0.69 (0.36,1.01)0.57 (0.22,0.93)0.08 (−0.34,0.49)
nuLF-HRV, (95% CI)63.1 (56.4,69.8)61.5 (53.6,69.4)50.1 (40.3,59.8)
nuHF-HRV, (95% CI)36.9 (30.2,43.6)38.5 (30.6,46.4)49.9 (40.1,59.7)
  • CSH, carotid sinus hypersensitivity; CI, confidence interval; ln, natural logarithmic transformation; HRV, heart rate variability; LF, low frequency; PSD, total power spectral density; nuLF, normalized LF; nuHF, normalized HF.

There were significant differences in LF-HRV (P = 0.014), total PSD (P = 0.031), LF:HF (P = 0.047), and nuLF-HRV (P = 0.049) between the symptomatic CSH group and the non-CSH control group (Table 3). The asymptomatic CSH group had significantly higher LF-HRV (P = 0.001), HF-HRV (P = 0.041), total PSD (P = 0.002), LF:LF (P = 0.030), and nuLF-HRV (P = 0.026) compared with the non-CSH control group. Normalized HF-HRV (P = 0.027) was, conversely, significantly lower in the asymptomatic CSH group compared with the non-CSH control group. The values presented for nuLF, nuHF, and LF:HF in Table 3 were adjusted for age differences only. However, once adjusted for heart rate as well as age, the differences in LF: HF (B = 0.46, 95% CI = −0.10 to 1.03; P = 0.109), nuLF (B = 9.7, 95% CI = −2.9 to 22.2; P = 0.127), and nuHF (B = −9.5, 95% CI = −22.1 to 3.0; P = 0.132) were no longer statistically significant between the symptomatic CSH and non-CSH control groups. Similarly, following adjustments for heart rate and age for comparisons between the asymptomatic CSH and non-CSH control groups, the differences in LF: HF (B = 0.54, 95% CI = −0.01 to 1.10; P = 0.054), nuLF (B = 12.0, 95% CI = −0.2 to 24.2; P = 0.053), and nuHF (B = −11.9, 95% CI = −24.09 to 0.29; P = 0.055) were no longer statistically significant.

View this table:
Table 3

Linear regression analysis comparing heart rate variability measures between non-CSH control vs. symptomatic CSH and non-CSH control vs. asymptomatic CSH groups

vs. non-CSH controls (n = 11)
Symptomatic CSH (n = 20)Asymptomatic CSH (n = 15)
Ba (95% CI)P-valueBa (95% CI)P-value
lnLF-HRVb, ms20.94 (0.20,1.68)0.0141.27 (0.55,1.99)0.001
lnHF-HRVb, ms20.65 (−0.22,1.53)0.1370.88 (0.04,1.73)0.041
lnPSDb (95%CI), ms20.70 (0.07,1.34)0.0311.04 (0.42,1.66)0.002
lnLF:HFc0.50 (0.01,0.99)0.0470.57 (0.06,1.08)0.030
nuLF-HRVc10.8 (0.1,21.6)0.04912.9 (1.6,24.0)0.026
nuHF-HRVc−10.8 (−21.5,0.00)0.051−12.8 (−24.0,−1.5)0.027
  • aB coefficient represents the adjusted mean difference between the two groups.

  • bStepwise linear regression adjusted for baseline differences in heart rate and age.

  • cStepwise linear regression adjusted for age.

  • CSH, carotid sinus hypersensitivity; B, parameter estimate; CI, confidence interval; HRV, heart rate variability; LF, low frequency; PSD, total power spectral density; nuLF, normalized LF; nuHF, normalized HF.

Baroreflex sensitivity

The unadjusted means for BRS and the BEI for up, down, and total ramps are summarized in Table 4. Following adjustments for heart rate and age differences, down (P = 0.017) and total BEI (P = 0.038) were significantly higher in the symptomatic CSH group compared with the non-CSH control group (Table 5). The variables up BRS, down BRS, total BRS, and up BEI also appeared non-significantly higher in the symptomatic CSH group compared with the non-CSH control group. The baroreflex sensitivity measures of the up slope (P = 0.012), down slope (P = 0.015), total slope (P = 0.011), down BEI (P = 0.001), up BEI (P = 0.049), and total BEI (P = 0.006) were statistically significantly higher in the asymptomatic CSH group compared with the non-CSH control group.

View this table:
Table 4

Unadjusted Mean with 95% confidence intervals for baroreflex slope and the baroreflex effectiveness index

Symptomatic CSH (n = 20)Asymptomatic CSH (n = 16)Non-CSH controls (n = 11)
Baroreflex Slope (ms/mmHg)
 Up slope, mean (95% CI)7.14 (5.69,8.60)9.21 (7.10,11.31)6.74 (4.87,8.60)
 Down slope, mean (95% CI)7.08 (5.37,8.78)9.19 (7.03,11.35)6.72 (3.81,9.63)
 Total slope, mean (95% CI)7.00 (5.58,8.43)9.19 (7.12,11.26)6.94 (4.95,8.92)
Baroreflex effectiveness index
 Up ramps, mean (95% CI)42.4 (31.2,53.5)41.2 (32.9,49.6)35.1 (20.3,50.0)
 Down ramps, mean (95% CI)40.1 (31.0,49.1)45.5 (36.9,54.0)29.3 (13.9,44.8)
 Total, mean (95% CI)41.0 (31.3,50.8)43.4 (35.7,51.2)33.8 (20.3,47.3)
  • CSH, carotid sinus hypersensitivity; CI, confidence interval; BEI, baroreflex effectiveness index.

View this table:
Table 5

Linear regression analyses for differences in baroreflex slope and baroreflex effectiveness index between symptomatic CSH vs. non-CSH control and asymptomatic CSH vs. non-CSH control groups

vs. non-CSH controls (n = 11)
Symptomatic CSH (n = 20)Asymptomatic CSH (n = 16)
Ba (95% CI)P-valueBa (95% CI)P-value
Baroreflex slope (ms/mmHg)
 Up slopeb2.45 (−0.34,5.24)0.0833.44 (0.81,6.08)0.012
 Down slopeb2.06 (−0.14,6.26)0.0613.79 (0.77,6.82)0.015
 Total slopeb2.31 (−0.40,5.03)0.0933.36 (0.80,5.93)0.011
Baroreflex effectiveness index
 Up rampsb14.2 (−1.9,30.3)0.08215.3 (0.1,30.5)0.049
 Down rampsb14.2 (3.5,33.2)0.01724.7 (10.7,38.8)0.001
 Totalb14.6 (0.9,28.4)0.03818.8 (5.8,31.8)0.006
  • aB coefficient represents the estimated mean difference between the two groups.

  • bAdjusted for baseline differences in heart rate and age using stepwise linear regression.

  • CSH, carotid sinus hypersensitivity; CI, confidence interval; BEI, baroreflex effectiveness index; B, parameter estimate.


Heart rate variability and non-invasive assessments of baroreflex sensitivity in individuals with CSH have never previously been reported. The assessment of HRV in participants with and without the symptomatic presentation of CSH suggests that CSH is associated with enhanced LF-HRV and total PSD compared with control subjects without CSH. Increases were also observed for LF:HF and nuLF-HRV for both CSH groups. In addition, the measurement of baroreflex sensitivity with the sequence method has demonstrated significantly higher down ramp and total BEI in individuals with CSH regardless of symptoms, with marginal increases in the up BEI. The baroreflex slopes were also significantly higher in the asymptomatic CSH group compared with non-CSH controls, whereas the apparent increases observed in the symptomatic CSH group did not achieve statistical significance.

Previous studies involving CSH were either uncontrolled or had included asymptomatic control subjects with unknown CSH status.1923 However, CSH is also present in 35% of asymptomatic older individuals with no symptoms of dizziness, falls, or blackouts.24 This suggests that the efforts of the above studies may have been compromised by the presence of undiagnosed CSH in the otherwise asymptomatic control groups.13 In addition, the high prevalence of CSH in asymptomatic older individuals also raises the question whether CSH is merely an age-related clinical sign with no genuine association with syncope or falls.7 Our study represents the first study to include a group of asymptomatic older adults with CSH. This approach has allowed us to define the autonomic changes associated with CSH, and to distinguish these changes from the pathophysiological changes associated with the symptomatic presentation of CSH.

The finding of increased LF-HRV, nuLF-HRV, and LF:HF during resting conditions suggests the presence of higher resting sympathetic activity in both CSH groups compared with control subjects without CSH. Heart rate variability is a computation-dependent advancement over the assessment of clinical autonomic cardiovascular reflexes, which allows the assessment of the influence of the sympathetic and parasympathetic systems on heart rate fluctuations at rest, and therefore does not require patient cooperation.25 Heart rate variability in the HF range reflects the heart rate fluctuation during respiration and is therefore an indication of parasympathetic function, while the HRV in the LF range is associated with sympathetic function. Normalized values for LF and HF-HRV, as well as the LF:HF ratio, are also regularly calculated. These derived measures provide a useful assessment of the relative influence of the sympathetic and parasympathetic systems on heart rate.15,26

The increases in both BRS and the BEI observed in participants with symptomatic CSH and asymptomatic CSH suggests that CSH is associated with increased baroreceptor sensitivity at rest. Dehn et al.27 evaluated the carotid sinus baroreflex gain in 10 subjects with CSH using the neck chamber suction method. They found that carotid sinus baroreflex gain was significantly higher in subjects with CSH compared with 16 age-matched controls.27 The stimulation of the carotid sinus to measure baroreflex sensitivity inpatients with known CSH will invariably result in an exaggerated response, limiting the validity of these findings. Morley et al.,28 however, measured baroreflex sensitivity using the phenylephrine pressor test in nine subjects with asystolic responses to CSH, and also found a significant increase in BRS in cases with CSH compared with individuals with sick sinus syndrome and control participants.28 Control participants in the latter two studies were not investigated with CSM to exclude the presence of asymptomatic CSH. Nevertheless, this body of evidence suggests that CSH is associated with increased baroreflex sensitivity.

The finding of enhanced resting sympathetic activity in individuals with CSH is intriguing. It has been suggested that CSH may be associated with atherosclerosis,29 or up-regulation of central α2-adrenoreceptors.30 The former theory has been disputed by the presence of normal arginine vasopressin release,20 whereas the latter was called into question by the lack of response to yohimbine, a potent central α-adrenoreceptor agonist.22 Other studies have suggested reduced sternocleidomastoid electromyographic activity2324 and peripheral sympathetic hyporesponsiveness21,31 in individuals with CSH, while direct neural recordings in individuals with CSH have demonstrated that sympathetic withdrawal occurs during CSM.3233 Our findings therefore suggest that a reduction in sympathetic response in individuals with CSH occurs on a background of increased baseline sympathetic activity. This may be explained by the loss of peripheral adrenoreceptor function in individuals with CSH. It is may also be possible that the enhanced sympathetic activity may be due to the loss of central regulation, with desensitization of peripheral adrenergic receptors occurring through a local negative feedback mechanism. An exaggerated hypotensive response may then be observed when sympathoinhibition occurs during CSM in association with a blunted peripheral vasomotor response. An inverse relationship between sympathetic nerve traffic and adrenergic vascular responsiveness has previously been demonstrated in healthy, young adults,3435 while a separate study has demonstrated increased basal sympathetic activity in older individuals with larger reductions in blood pressure observed from ganglion blockade.36

The alteration in autonomic activity observed in individuals with CSH regardless of the presence of symptoms therefore prompts the suggestion that CSH is a clinical sign associated with ageing rather than a distinct pathologic entity, which may be premature. Miller et al.12 demonstrated an increase in hyperphosphorylated-tau protein deposition within the nucleus tractus solitarius, nucleus ambiguous, and cardiac reflex arc, which are key medullary nuclei associated with the baroreflex response, in a post-mortem study of patients with symptomatic CSH compared with control subjects without CSH.12 Tau protein is a precursor of the neurofibrillary tangle and hence an early indicator of Alzheimer's disease. Our study suggests that the pathological changes observed in the above study manifests as autonomic dysregulation rather than autonomic failure. It also gives credence to the hypothesis of dysfunctional central regulation contributing towards inappropriate resting sympathetic activation and baroreflex overactivity, although whether this is secondary to CSH or is causative in its development remains unknown.

Study limitations

Non-invasive assessments of HRV and baroreflex sensitivity are limited by the presence of atrial fibrillation and other ECG artefacts including multiple ectopics. This restricted our ability to assess HRV and baroreflex sensitivity in 20% of our subjects, a finding also noted by others.37 The exploratory nature of this study has to be emphasized, which will account for its relatively small sample size. However, the handful of studies evaluating autonomic function in CSH has had fewer subjects. Furthermore, our study has also been unique in the inclusion of asymptomatic control subjects with an established CSH status.


Individuals with CSH have higher resting LF-HRV, nuLF-HRV, and LF:HF regardless of symptoms compared with healthy older control participants without CSH. In addition, BRS and BEI were enhanced in both CSH groups, compared with non-CSH control subjects. The above findings suggest the presence of autonomic dysregulation in individuals with CSH, indicating that CSH is a generalized autonomic disorder. The observed alterations in autonomic activity underlying CSH, however, occur regardless of the presence of symptoms. Therefore, while there does appear to be a pathological process underlying this condition, the findings of our study do not explain the differences between individuals with asymptomatic and symptomatic CSH.


M.P.T. was funded by the Royal College of Physicians/Dunhill Medical Trust joint research fellowship. This study was also supported by the United Kingdom National Institute for Health Research Biomedical Research Centre for Ageing and Age-related Disease and awarded to the Newcastle upon Tyne Foundation Hospitals NHS Trust.


Many thanks to Dorothy Carman, Irene Blair, and Leanne Thompson from the Clinical Research Facility for their assistance during the study.

Conflicts of interest: none declared.


View Abstract