Journal of Neurology

, Volume 256, Issue 5, pp 803–809

Quantitative autonomic testing in the management of botulism


    • Department of NeurologyAcademic Teaching Hospital Wagner-Jauregg
  • Christoph Heibl
    • Department of Internal Medicine IKlinikum Kreuzschwestern Wels
  • Karl Stieglbauer
    • Department of NeurologyAcademic Teaching Hospital Wagner-Jauregg
  • Bettina Dreer
    • Department of Internal Medicine IKlinikum Kreuzschwestern Wels
  • Markus Nagl
    • Division of Hygiene and Medical Microbiology, Department of Hygiene, Microbiology, and Social MedicineInnsbruck Medical University
  • Peter Knoflach
    • Department of Internal Medicine IKlinikum Kreuzschwestern Wels
  • Franz Thaddäus Aichner
    • Department of NeurologyAcademic Teaching Hospital Wagner-Jauregg
Original Communication

DOI: 10.1007/s00415-009-5022-9

Cite this article as:
Topakian, R., Heibl, C., Stieglbauer, K. et al. J Neurol (2009) 256: 803. doi:10.1007/s00415-009-5022-9


Even with mild neurological signs, patients with botulism frequently complain of autonomic symptoms. This study aimed at the evaluation of sudomotor and cardiovascular reflex functions by quantitative autonomic testing (QAT), which may identify patients with autonomic involvement but otherwise benign clinical presentation. Five patients with food-borne botulism were subjected to a structured questionnaire on autonomic symptoms, cardiac and neurological examination, and QAT after a median of 2 weeks (baseline) and 12 weeks (follow-up) post intoxication. For calculation of haemodynamic and cardiovascular autonomic parameters, we used the Task Force® Monitor (Version 2.1, CNSystems, Graz, Austria). Cardiovagal function was assessed by Ewing’s test battery. Autonomic complaints were more pronounced than neurological symptoms. Baseline tests revealed widely abnormal sudomotor function and marked impairment of heart rate variation and blood pressure response to standing. Prominent features of cardiovascular failure were high resting heart rate, supine hypertension, orthostatic hypotension, and impaired baroreflex function. Three patients reported inability to keep up with their routine amount of physical work. Based on the baseline QAT results, these three patients were instructed to engage in physical activity but avoid physical strain until there was considerable improvement. On follow-up, fatigue was the most frequent residual complaint, sympathetic skin responses were present, and cardiovascular QAT results were significantly improved and did not differ from those of ten control subjects. QAT identified autonomic involvement in botulism patients with otherwise benign neurological presentation. Comprehensive evaluation of autonomic failure may provide useful information for the management of botulism.


BotulismAutonomic nervous systemAutonomic failureBaroreflex function


Food-borne botulism is caused by the ingestion of preformed toxin produced by the spore-forming, obligate anaerobic bacillus Clostridium botulinum. The disease typically starts with gastrointestinal symptoms followed by impaired motor function and autonomic symptoms [3, 11]. Neurological symptoms can be relatively mild with predominantly autonomic involvement, and even pure autonomic failure has been described [12]. Because severe impairment of cardiovascular autonomic functions carries the risk of cardiac complications, patient management might benefit from a systematic evaluation of the autonomic derangement in botulism. However, only few reports focus on the assessment of cardiovascular reflex functions in botulism [2, 20].

The aim of this study was to characterise and systematically evaluate cardiovascular and sudomotor autonomic involvement in five patients with botulism.

Patients and methods

We report on five previously healthy patients (four male, one female) who developed clinical signs of food-borne botulism 4 days after a barbecue. Patients were admitted to two departments of internal medicine at two local hospitals 6–7 days post intoxication. Clinically, food-borne botulism was in the differential diagnosis, and the patients underwent cardiac and neurological evaluation, laboratory and stool culture tests. Attempts to isolate C. botulinum from faecal samples of the patients were unsuccessful. Despite thorough investigations, neither the exact source of outbreak nor the type of botulinum toxin could be determined. However, botulinum toxin was detected in the sera from all patients and was neutralised by addition of commercial polyvalent antiserum in the mouse bioassay, the most sensitive, simple and widely used method for detecting toxin [13]. Sera drawn from the patients on the day of admission were diluted twofold with saline (test group) and polyvalent antitoxin against toxin types A, B and E (Chiron-Behring, Marburg, Germany) (control group), respectively. Two Balb/c mice each were injected intraperitoneally with 0.5 ml of these dilutions. The two mice challenged with sera from four patients, and one of two mice challenged with serum from the fifth patient all developed a typical wasp waist and palsy after 18–24 h and died from respiratory paralysis within 30–48 h. All control mice survived without any symptoms of paralysis [13].

Neurological presentation was mild, and all patients were hospitalised for only a few days. For autonomic evaluation, patients were referred to the Department of Neurology, Wagner-Jauregg Hospital, Linz. All patients underwent detailed history taking with questionnaires on autonomic symptoms, a complete neurological examination, sudomotor tests, and a comprehensive quantitative autonomic testing (QAT) of cardiovascular functions. Our questionnaire on autonomic symptoms has been developed in-house to support structured history taking in patients with autonomic symptoms. It consists of 49 items to assess seven domains of autonomic symptoms (orthostatic and reflex syncope, sudomotor, pupillomotor, gastrointestinal, urinary, and male sexual dysfunction) and an appendix with 11 questions on six domains of quality of life (physical functioning, energy/fatigue, pain, social functioning, emotional well-being, and role limitations due to emotional or physical stresses).

Baseline tests were carried out after a median of 14 (range 11–17) days and follow-up tests after a median of 12 (range 11–14) weeks post intoxication. Cardiovascular QAT results of patients were compared to those of ten healthy controls matched for age, sex, and body mass index. None of the patients had a history of significant comorbidity or regular use of medication.

Sympathetic skin responses, nerve conduction studies, and electromyography

Sympathetic skin responses (SSRs) were examined at baseline and follow-up using standard EMG equipment (Viking IV D, Nicolet Biomedical Inc., Madison, WI). We recorded presence/absence of hand and feet SSR using electrical stimulation randomly delivered to the contralateral median and tibial nerves. In addition, three patients had nerve conduction studies in ≥3 peripheral nerves and electromyography in ≥1 muscle.

Cardiovascular quantitative autonomic testing

For monitoring and automatic calculation of haemodynamic and autonomic parameters, we used the Task Force® Monitor, Version 2.1 (CNSystems, Graz, Austria). This integrated system includes a six-channel electrocardiogram (ECG) and allows continuous measurement of beat-to-beat blood pressure (BP) using the vascular uploading technique on the finger [1, 8]. These BP values are automatically and continuously corrected to oscillometric BP values obtained on the contralateral arm.

Heart rate variability

Power spectral analysis of heart rate variability (HRV) offers insight into the fluctuations in autonomic tone by splitting biosignals into various underlying frequency components [19]. Beat-to-beat values of the R–R interval derived from the ECG were used for real-time calculation of HRV by an autoregressive model, displayed as three-dimensional sliding power spectra [1, 8]. HRV in the high frequency (HF) range (0.15–0.4 Hz) is considered as an index of parasympathetic activity, whereas HRV in the low frequency (LF) range (0.04–0.15 Hz) is the result of predominantly sympathetic activity. The LF/HF ratio of heart rate can be calculated as a parameter of sympathovagal balance [19].

Baroreceptor function

The automatic evaluation of baroreceptor sensitivity (BRS) using the sequence method has been described before [1, 8, 16]. In brief, the algorithm searches for episodes of spontaneous activation of the baroreceptor reflex. Events of baroreceptor reflex activation are identified when a progressive increase in systolic BP of ≥1 mmHg/beat over at least three consecutive heart beats is followed by a progressive lengthening in R–R interval of ≥4 ms/beat with a one-beat delay or, vice versa, progressive decreases in systolic BP are followed by a progressive shortening of R–R intervals. The number of detected events is referred to as the “event count”. A second BRS index is the “BRS slope”, which describes the steepness of the regression line between BP change in mmHg and the RR-interval change in ms. Only episodes with correlation coefficients >0.95 were selected, and from all regressions a mean slope of BRS was calculated for each steady-state period.

Cardiovascular QAT protocol

On days of autonomic assessment, patients were asked to stop any medication and restrain from consuming nicotine, caffeine, or alcohol. They had no meal at least 3 h before testing. Patients were tested between 1:00 and 3:30 p.m. in a quiet, dimly lit room with comfortable ambient temperature. After an adjustment period of at least 15 min with the patient supine on a tilt table with foot support, haemodynamic parameters were continuously recorded for 10 min in the supine position (to calculate mean resting heart rate, mean resting blood pressure, and baroreceptor sensitivity) and for 10 min after 60° tilt table testing. Following this, cardiovagal function was evaluated by three established indices derived from metronomic breathing, Valsalva manoeuvre, and active standing (Ewing’s test battery) [6]. Metronomic breathing was performed on two occasions with a breathing rhythm of 5-s inspiration and 5-s expiration over 80 s. We analysed five consecutive breathing cycles and calculated the mean ratio of the longest R–R interval during expiration to the shortest R–R interval during inspiration (E/I ratio). The Valsalva maneuver was performed on three occasions by blowing into a mouthpiece at a pressure of 40 mmHg for 15 s. The Valsalva ratio was calculated as the ratio of the longest R–R interval after blowing into the tube to the shortest R–R interval during the expiratory strain. The largest of three ratios was used for analysis. The heart rate response to active standing, the 30:15 ratio, was obtained from the maximum R–R interval between beat 20 and 40 and the minimum R–R interval between beat 5 and 25 after the start of active standing, which was carried out over 10 min to test BP control.

Statistical analyses

Distribution of data was skewed due to the small number of patients. Continuous data were expressed as median values and range minimum–maximum. The two-sided Mann–Whitney U test was used for intergroup comparisons. The level of significance was set at P < 0.05. All analyses were carried out with SPSS for Windows, Version 14.0 (SPSS; Chicago, IL).


Symptoms and signs

At baseline, autonomic complaints were more pronounced than neurologic symptoms (Table 1). All five patients reported blurred vision, dry mouth, new onset constipation, fatigue and reduced sweating. Neurological examination revealed pupillary abnormalities in all patients. Three patients had mild bulbar signs, with one also having reduced head control. None of the patients had extremity weakness.
Table 1

Clinical characteristics and sympathetic skin responses at baseline

Age (years)/sex

Botulism patient






Autonomic and other symptoms

 Blurred vision






 Dry mouth






 Constipation (new onset)






 Bladder voiding difficulty











 Orthostatic intolerance











Neurological findings

 Pupils fixed or dilated













 Facial paresis









 Reduced head control


 Extremity weakness

Sympathetic skin response

 No. of extremities where absent






On follow-up, pupillary abnormalities and accommodation problems were still present in two patients, with one of them still having mild dysarthria. No other neurological signs were found. Fatigue, though improved, was the most frequent residual complaint.

Sudomotor tests, nerve conduction studies, and electromyography

At baseline, SSRs could not be elicited in two patients, and another two patients had absent SSRs in at least one extremity (Table 1). One patient had SSRs in both hands and feet. On follow-up, all patients had SSRs in both hands and feet.

Amplitudes of compound muscle action potentials and motor unit potentials were mildly reduced at baseline in all three patients undergoing nerve conduction studies and electromyography. Repetitive nerve stimulation, performed only in the patient with the most prominent neurologic signs, did not show abnormalities at 2, 5, and 10 Hz. On follow-up, all parameters were in the range of the laboratory’s normative data.

Cardiovascular autonomic tests

Results of cardiovascular QAT with intergroup comparisons are given in Table 2, while scatter diagrams with individual QAT results are shown in Fig. 1. Compared to follow-up and controls, baseline values of resting heart rate and resting BP were significantly higher (Table 2), with two patients being hypertensive in the supine position. Baroreceptor sensitivity was dramatically decreased, with absolute reduction of BRS events and the BRS slope. In addition, cardiovagal function was highly impaired throughout all tests of Ewing’s battery (Table 2).
Table 2

Cardiovascular autonomic function in patients with botulism at baseline and follow-up compared to controls matched for age, sex, and body mass index


Botulism (n = 5)

Controls (n = 10)



Age (years)

28 (24–55)

29 (20–57)

Male sex, n (%)

4 (80%)

8 (80%)


26.5 (21.8; 29.2)

25.1 (20.8; 31.0)

Resting parameters


83 (78–85)*,†

69 (58–82)

66 (53–84)

 HR LF/HF ratio

3.0 (1.2–5.8)

1.9 (0.8–3.1)

1.3 (0.3–3.9)

 BP systolic

136 (126–168)*,

119 (116–130)

118 (108–167)

 BP diastolic

94 (80–106)*,†

85 (73–98)

79 (69–98)

 BRS events/10 min

3 (2–16)**,††

42 (20–62)

58 (25–112)

 BRS slope (ms/mmHg)

3.1 (2.1–4.6)*,††

12.2 (3.9–32.6)

17.3 (4.3–27)

Cardiovagal tests

 30:15 ratio

1.07 (1.01–1.16)*,††

1.33 (1.15–1.37)

1.39 (1.1–1.78)

 E/I ratio

1.05 (1.03–1.12)**,††

1.41 (1.26–1.45)

1.27 (1.18–1.67)

 Valsalva ratio

1.28 (1.06–1.49)*,††

1.53 (1.34–1.82)

1.84 (1.33–2.62)

Orthostatic tests

 Patients with OH, n (%)

3 (60%)



 BP systolic after 3 min tilt

124 (103–152)

130 (114–136)

122 (104–161)

 BP diastolic after 3 min tilt

95 (78–100)

92 (78–105)

89 (78–100)

 BP systolic after 3 min stand

130 (89–136)

124 (118–134)

121 (101–156)

 BP diastolic after 3 min stand

90 (66–101)

88 (78–99)

90 (79–105)

Values are medians and range minimum–maximum in brackets unless given otherwise

BMI body mass index, HR heart rate, LF low frequency, HF high frequency, BP blood pressure, BRS baroreceptor sensitivity, E/I expiratory/inspiratory, OH orthostatic hypotension

Mann–Whitney U test was used for intergroup comparisons; *P < 0.05 versus follow-up, **P < 0.01 versus follow-up, P < 0.05 versus controls, ††P < 0.01 versus controls
Fig. 1

Scatter plots of quantitative cardiovascular autonomic test results in individuals with botulism at baseline and at follow-up (n = 5) compared to controls (n = 10) matched for age, sex, and body mass index. P values indicate levels of significance of the difference between two groups tested by Mann–Whitney U tests. E/I ratio expiratory/inspiratory, NS not significant (P ≥ 0.05), LF/HF low frequency/high frequency

Two patients were diagnosed with orthostatic hypotension (OH) during active standing following standard criteria [4] (fall in systolic BP of 33 and 29 mmHg ≤ 3 min), and another patient had OH during tilting (fall in systolic BP of 41 mmHg ≤ 3 min). However, median blood pressure after 3 min of active standing and 3 min of tilting did not differ significantly between groups (Table 2).

On follow-up, there was profound recovery of baroreflex and cardiovagal functions, and parameters of heart rate and blood pressure variability were no longer different between patients and controls (Table 2; Fig. 1).


The finding that botulism can markedly impair control of heart rate variability and blood pressure response is in line with two other reports on cardiovascular functions in botulism [2, 20]. However, our study protocol included not only well established tests for the evaluation of sympathovagal modulation of heart rate and blood pressure, but also the non-invasive measurement of baroreflex function—the most powerful reflex for cardiovascular stabilisation [5].

The non-invasive measurement of sensitivity of the arterial baroreflex has been increasingly used as an index of cardiac autonomic control [16]. Baroreflex function is related to a number of cardiovascular risk factors, such as age, hypertension, diabetes, and smoking, and has been investigated in various disorders. Severe baroreflex dysfunction carries the risk of haemodynamic and cardiac complications. Impaired baroreflex function is an independent predictor of cardiac mortality after myocardial infarction [9, 10] and has been hypothesised to contribute to the occurrence of sudden unexplained death in patients with temporal lobe epilepsy [5].

In adults with botulism and mild neurological presentation, even pronounced autonomic dysfunction is unlikely to lead to life-threatening complications. However, botulism has also been reported in association with sudden infant death syndrome [15], when severe autonomic failure cannot be ruled out as the cause of death.

On the basis of the initial QAT results, we instructed three patients to engage in physical activity but avoid physical strain until there was considerable improvement. All three patients had marked baroreflex dysfunction, orthostatic hypotension, and widely absent SSRs. The previously healthy subjects reported reduction of physical working capacity due to recurrent orthostatic symptoms and inability to sweat. In these patients with only mild neurological signs, QAT helped objectify the extent of autonomic involvement, which otherwise might have remained unrecognised.

Use of QAT in botulism has not been frequently reported in the literature, while findings in neurophysiologic tests are well described [3]. Apart from small evoked muscle action potentials in response to a single supramaximal nerve stimulus in clinically affected muscles, there may be typical, though unspecific findings on repetitive nerve stimulation and single-fibre electromyography suggesting botulism. Neurophysiologic tests can be especially helpful in establishing the diagnosis of botulism in patients with a typical clinical presentation when bioassays and stool cultures are negative [3]. However, in the setting of serologic and toxicological confirmation and in patients with predominantly autonomic symptoms, routine neurophysiologic tests do not add further useful information influencing patient management.

All our patients reported hypohydrosis. Two patients showed absent SSRs in both hands and feet at baseline, and another two patients had SSRs absent in at least one extremity. In contrast, one patient had SSRs present in all four limbs. Abnormalities of SSRs in patients with botulism have been described before [2]. Hypohydrosis was probably more severe and generalised in the patients of Chen et al. [2], who reported absent SSRs at baseline in all patients.

All patients showed significant neurological and autonomic recovery after a median follow-up of 12 weeks post-intoxication. Botulinum toxin blocks the release of acetylcholine at peripheral cholinergic terminals by irreversibly inhibiting neurotransmitter exocytosis. At the neuromuscular junctions, this blockade is overcome by enlargement of the junctions and sprouting of nerve terminals, which takes weeks to months [3]. Neurological recovery is often complete, but in severe cases secondary changes such as critically illness neuropathy and myopathy patients may result in late symptoms [7]. Neurological symptoms may be outlasted by general fatigue [3, 7, 20]. In our patients, fatigue was the most frequent complaint on follow-up. A limitation of our study is that we did not use an internationally accepted and validated questionnaire, such as the “Autonomic Symptom Profile” [18].

In this context it is remarkable that a recent case-control study by Gottlieb et al. [7] on the long-term outcome of 217 botulism patients showed that several years after the infection a substantial number of patients was still reporting dry mouth, fatigue, limitation in vigorous activities and other complaints. The authors concluded that, rather than relying on subjective perception of symptoms, measurement of functional status with objective characterisation of symptoms could help to better define and prevent the long-term consequences of botulism [7].

Therapy in patients with predominantly autonomic involvement is symptom orientated and includes stool softeners and adequate hydration. Some patients with neurological symptoms show a beneficial response to cholinesterase inhibitors, such as edrophonium chloride [17]. Drugs that enhance the release of acetylcholine from nerve terminals such as guanidine can improve muscle strength [14]. However, experience with these drugs is limited and experimental, and patients should be carefully monitored, particularly in the presence of significant comorbidity or a history of cardiovascular events. Our QAT findings suggest that drugs that interfere negatively with baroreflex function should be avoided.

In conclusion, cardiovascular autonomic failure in botulism can manifest with marked baroreflex dysfunction and impaired heart rate variation and blood pressure responses to standing. Features are a higher resting heart rate, supine hypertension, and orthostatic hypotension. Our study suggests the use of QAT for the management of botulism, especially when neurological signs are mild and the severity of autonomic failure may be underestimated without comprehensive testing.

Conflict of interest statement

The authors report no conflicts of interest.

Copyright information

© Springer-Verlag 2009