- Email: [email protected]
Transcranial direct current stimulation (tDCS) and the cardiovascular responses to acute pain in humans
Clinical Neurophysiology 126 (2015) 1039–1046
Contents lists available at ScienceDirect
Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph
Transcranial direct current stimulation (tDCS) and the cardiovascular responses to acute pain in humans J.W. Hamner, Mauricio F. Villamar, Felipe Fregni, J. Andrew Taylor ⇑ Department of Physical Medicine and Rehabilitation, Harvard Medical School and Spaulding Rehabilitation Hospital, Boston, MA, United States
a r t i c l e
i n f o
Article history: Accepted 14 August 2014 Available online 16 September 2014 Keywords: tDCS Brain stimulation Acute pain Cardiovascular Sympathetic Autonomic
h i g h l i g h t s We tested whether transcranial direct current stimulation (tDCS) alters both pain perception and its
cardiovascular responses and if it impacts basal hemodynamics. Active tDCS did not affect resting hemodynamics or autonomic outflow. Only with the least painful stimulus, tDCS modestly reduced perceived pain and the peak cardiovas-
cular and autonomic responses.
a b s t r a c t Objective: To determine if transcranial direct current stimulation (tDCS) reduces both acute pain perception and the resultant cardiovascular responses. Methods: Data were acquired on 15 healthy subjects at rest and in response to three cold pressor tests: 0, 7, and 14 °C. Subsequently, single sessions of sham and active anodal tDCS (2.0 mA for 40 min) were delivered to the left primary motor cortex (M1). Results: Perceived pain was reduced only after active tDCS with the 14 °C cold pressor test. This was accompanied by tendency for lesser increases in heart rate (2 beats/min, p = 0.09) and blood pressure (3 mm Hg, p = 0.06). The effect size of tDCS on peak heart rate and blood pressure responses at 14 °C was 0.47 and 0.54, respectively. On the other hand, baseline heart rate, blood pressure, leg blood flow, and leg vascular resistance were unaffected by tDCS. No other responses were affected. Conclusions: Our results demonstrate that M1 anodal tDCS has no effect on basal hemodynamics or cardiovascular autonomic outflow and has only modest effects on the responses to acute pain in healthy humans. Significance: Application of tDCS shifts the pain perception threshold in healthy individuals but does not significantly modulate efferent cardiovascular control at rest or in response to pain. Ó 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction In humans, acute painful stimuli increase heart rate, blood pressure and sympathetic nervous activity to the vasculature (Lewis, 1942). If one could modulate the responses to acute pain, this might have broader application to not only reduce the somatosensory burden of pain but also mitigate against its possible long-term cardiovascular effects (Goodson et al., 2013). One approach to treat pain is direct brain stimulation. However, until the 1990s only surgically ⇑ Corresponding author at: Cardiovascular Research Laboratory, Spaulding Hospital Cambridge, 1575 Cambridge Street, Cambridge, MA 02138, United States. Tel.: +1 617 758 5503; fax: +1 617 758 5514. E-mail address: [email protected] (J.A. Taylor).
implanted electrodes to achieve deep brain stimulation had been systematically tested and shown to induce significant decreases in pain. More recently, non-invasive, easy to administer approaches have generated increased interest as a potential therapeutic intervention (Fregni et al., 2007; Jensen et al., 2013). Transcranial direct current stimulation (tDCS) has long lasting modulatory effects on cortical function and allows a reliable sham-stimulation condition to assure specificity of effects. Therefore, tDCS might provide a testable avenue to both reduce pain perception and the resultant cardiovascular responses. There is evidence that tDCS can effectively modulate pain perception threshold in healthy individuals (Boggio et al., 2008; Csifcsak et al., 2009), indicating that its analgesic effects do not depend on aberrant neural activity. Beyond altering afferent pain
http://dx.doi.org/10.1016/j.clinph.2014.08.019 1388-2457/Ó 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
1040
J.W. Hamner et al. / Clinical Neurophysiology 126 (2015) 1039–1046
perception, there is data suggesting that brain stimulation could modulate efferent cardiovascular control (Schestatsky et al., 2013), and hence has been proposed as a potential tool for the management of hypertension, even independent of pain (Cogiamanian et al., 2010; Schestatsky et al., 2013). For example, high-frequency transcranial magnetic stimulation (TMS) reduces blood pressure and heart rate in rats, apparently through sympathoinhibition (Hong et al., 2002). tDCS-mediated reductions in heart rate and blood pressure have also been observed in humans (Binkofski et al., 2011), though this is not consistently found (de Vries et al., 2010; Floel et al., 2008; Raimundo et al., 2012; Vandermeeren et al., 2010). Even though there is also evidence of the contrary (Vandermeeren et al., 2010), some data suggest potential alterations in cardiac autonomic control with tDCS (Brunoni et al., 2013; Montenegro et al., 2011). However, this work relied on heart rate variability which can be a poor surrogate for direct measures of autonomic control (Taylor and Studinger, 2006). Directly measured muscle sympathetic outflow via microneurography has shown inhibition of pulse-synchronous activity after TMS (Macefield et al., 1998). In addition, microneurography data indicate that TMS alters sympathetic activity to skin (Silber et al., 2000; Niehaus et al., 1998). Hence, cortical stimulation may alter not only pain perception and its resultant cardiovascular responses but also basal homeostatic hemodynamics not related to pain perception. Nevertheless, despite evidence from TMS studies, direct measurements of the effects of tDCS on muscle sympathetic outflow via microneurography are still lacking. We tested the hypothesis that non-invasive cortical stimulation alters pain perception and the autonomic responses to acute pain. We chose M1 anodal tDCS to modulate cortical excitability since this montage has shown analgesic effects in the setting of experimental pain in healthy subjects (Boggio et al., 2008; Csifcsak et al., 2009) as well as in chronic pain patients (O’Connell et al., 2014), and because the M1 has been involved in the control of bulbar cardiovascular nuclei and vasomotor spinal preganglionic neurons (Ba-M’Hamed et al., 1996; Viltart et al., 2003). In addition to evaluating any potential changes induced by tDCS on baseline cardiovascular measures, we used the nociceptive response to immersion of the hand in cold water (cold pressor test) at three different temperatures as a provocative maneuver. The cold pressor test increases perceived pain, blood pressure, heart rate, vascular resistance, and sympathetic activity (Fagius et al., 1989), and the responses are directly related to decreasing water temperature (Kregel et al., 1992). A strong link between the perceived pain and the pressor responses is suggested by work showing that partial sedation proportionally reduces both perceived pain and cardiovascular responses to the cold pressor test (Noseir et al., 2003). Hence, by measuring hemodynamic variables (heart rate, blood pressure, leg blood flow, leg vascular resistance), as well as both indirect (RR interval, mean pressure variabilities) and direct indices (muscle sympathetic nerve recordings) of autonomic outflow, this design allowed us to examine whether tDCS alters perceived pain and its reflex cardiovascular responses.
2. Methods 2.1. Subjects Fifteen healthy young individuals aged 21–30 participated in this study (7 female). All subjects had a body mass index between 18.5 and 29.9 kg/m2 and a normal resting electrocardiogram (ECG). None of them had any signs or symptoms of cardiovascular or neurological diseases, recent weight change, regular use of tobacco, or current pregnancy. All subjects gave written informed consent prior to participating. This study was approved by the Institutional
Review Board at Spaulding Rehabilitation Hospital (Protocol #2011-P-001879/1) and conformed to the Declaration of Helsinki. 2.2. Procedures Subjects visited the laboratory on two separate mornings (around 8 a.m.) to receive either active or sham tDCS. Study visits were separated by a minimum of 7 days to a maximum of 8 weeks. Subjects abstained from vigorous exercise for 2 days prior to each study visit to avoid autonomic and neuroendocrine effects of exercise. In addition, subjects refrained from caffeine and alcohol for the previous 24 h. Upon arrival, participants were instrumented for tDCS and physiologic assessments, as detailed below. On each visit, assessments were performed first at baseline while no brain stimulation was being delivered, and then again during either active or sham tDCS. Over the two study visits, participants underwent one session of active and one session of sham tDCS. These two interventions were delivered in random and counterbalanced order. Throughout the protocol, subjects were supine on a laboratory table and were instrumented for measurement of standard lead II of the ECG, beat-by-beat blood pressure in a finger of the left hand (Portapres, Finapres Medical Systems), brachial oscillometric blood pressure (Dash 2000, GE), respiratory excursions from a respiratory bellows placed around the chest, and popliteal artery blood flow velocity at the popliteal fossa of the left leg (Multi-Dop T2 4-MHz Doppler probe; Compumedics DWL, Singen, Germany). After instrumentation and calibration, multiunit postganglionic muscle sympathetic nerve recording from the common peroneal nerve was successfully obtained in a single session (either sham or active tDCS) in 12 subjects. For all subjects in both sessions, data were acquired during a 5-min period of quiet rest and in response to three cold pressor tests (0, 7, and 14 °C) performed in random order. For each cold pressor test, a 1-min baseline period was followed by 3-min immersion of the right hand in cold water. During the immersion, subjects rated their perceived pain on a Visual Analog Scale for pain every 30 s (0–10, with 0 corresponding to absence of pain and 10 corresponding to the worst imaginable pain). Cold pressor tests were separated by 10 min, allowing hand temperature to normalize between trials. Subsequently, either sham or active tDCS was applied. Active and sham tDCS were delivered using 35 cm2 sponge electrodes. The anode was placed over the left primary motor cortex (M1), corresponding to C3 in the International 10–20 Electroencephalography System, and the cathode over the right supraorbital area. An ActivaDoseÒ II Iontophoresis Delivery Unit (ActivaTek Inc., Salt Lake City, UT) was used in all experiments. In active tDCS, current was ramped up over a period of 30 s until reaching 2.0 mA, which were applied for the remainder of the testing session (40 min). Current density was 0.057 mA/cm2. For the sham condition, the same instrumentation was used but direct current was only applied for 30 s (Gandiga et al., 2006). tDCS was administered by an unblinded investigator who was not involved in data analysis. Subjects were blinded to the type of stimulation and the tDCS device was kept out of their sight for the duration of the study. Once 5 min of stimulation (active or sham) had elapsed, data were again acquired during a 5-min period of quiet rest and in response to the three cold pressor tests performed in random order while the stimulation continued to be delivered. 2.3. Data and statistical analysis Values were derived for mean blood pressure ([2 diastolic + systolic]/3) and for leg vascular resistance (mean blood pressure/leg blood flow). Resting baseline values pre- and post- sham and active tDCS were derived from 5-min averages
1041
J.W. Hamner et al. / Clinical Neurophysiology 126 (2015) 1039–1046
for all variables (heart rate, blood pressure, leg blood flow, leg vascular resistance, and muscle sympathetic activity). In addition, the magnitude of variabilities in RR interval and mean pressure were derived from power spectral analysis. Though these measures may not accurately reflect cardiac autonomic or vascular sympathetic outflows (Taylor and Studinger, 2006), prior work suggests that there may be alterations in cardiovascular variabilities after tDCS (Brunoni et al., 2013; Montenegro et al., 2011). For power spectral analysis, the ECG and arterial pressure waveforms were peak detected, edited for artifact, and linearly interpolated to a frequency of 4 Hz before being divided into 5 equal segments that overlapped by 50%. Each segment was detrended, Hamming filtered, and fast-Fourier transformed to its frequency representation squared. These periodograms were then averaged to produce the spectrum estimate. Average power was calculated in the low (0.04–0.15 Hz) and high (0.15–0.4 Hz) frequency ranges. Peak responses to each cold pressor test were derived from the difference between the average baseline value preceding each test and the highest 30-s average during each test. Average pain scores and peak pain scores within each cold pressor test were determined. Repeated measures ANOVAs were used to determine differences in pain and cardiovascular responses within each temperature level (0, 7, and 14 °C) of cold pressor test. This provided significance levels for the effect of time (pre- versus post-tDCS), the effect of condition (sham versus active tDCS), and their interaction. A p < 0.05 was considered significant. All data are presented as means with standard error of the mean. 3. Results All tDCS sessions proceeded uneventfully and were not associated with adverse effects other than a mild-to-moderate tingling or itching sensation and transient erythema in the area of stimulation. These occurred with both active and sham stimulation. 3.1. Resting hemodynamics and autonomic outflow Resting hemodynamics were unaffected by tDCS (Table 1). Systolic pressure tended to be higher after both sham and active tDCS (Time p = 0.04), but this effect was not different between stimulation conditions (Time Condition p = 0.47). All other hemodynamic variables demonstrated no differences with time, no differences between condition, and no interaction between time and condition (lowest p = 0.55). Unpaired comparisons of those with muscle sympathetic nerve recordings (n = 12) showed that there was no effect of time or condition, and no interaction between time and condition (Table 2). Indices of RR interval and mean pressure variability also showed no pronounced effects of active tDCS on either low or high frequency powers (Table 2). 3.2. Responses to pain Perceived pain in response to the cold pressor test was graded with temperature (p < 0.05) and tended to be reduced with
repeated trials (p < 0.05), especially for the two greater cold pressor stimuli (Fig. 1). At 14 °C, the reduction in perceived pain was greater after active tDCS than sham tDCS (p < 0.05) (Fig. 2). This reduction in perceived pain was reflected in a tendency for a lesser peak increase in heart rate and blood pressure after active tDCS (time condition; Fig. 3). Post-hoc t-tests showed that the peak increases in heart rate and blood pressure tended to be lower by 2 beats/min (p = 0.09) and 3 mm Hg (p = 0.06) with effect sizes of 0.47 and 0.54, respectively. However, no other hemodynamic response to the 14 °C cold pressor test was affected by active (or sham) tDCS. With the 7 °C cold pressor test, all hemodynamic responses tended to be lower with repeated trials (Time p < 0.10) and active tDCS did not magnify this effect (Time Condition all p > 0.20; Fig. 4). With the 0 °C cold pressor test, hemodynamic responses were similar across repeated trials and were unaffected by active tDCS (Time Condition all p > 0.40; Fig. 5). Muscle sympathetic nerve responses to the cold pressor tests were highest with the 0 °C test and minimal with both the 14 °C and 7 °C tests (Fig. 6). At all temperatures, responses were similar across repeated trials and were unaffected by active tDCS (Fig. 7). 4. Discussion Our results demonstrate that non-invasive brain stimulation via anodal M1 tDCS has no significant effects on basal hemodynamics or cardiovascular autonomic outflow. In addition, this form of cortical stimulation had only a slight effect on the responses to acute pain in healthy humans. At the lowest level of the nociceptive cold stimulus, when peak perceived pain was low (a reported value of 1 or less in 45% of trials), there was a modest decrease in perceived pain with active compared to sham tDCS. This translated into a tendency for slightly lower peak responses in heart rate (2 beats/min) and blood pressure (3 mm Hg), but no other significant hemodynamic effects. It should be noted that this level of nociceptive cold stimulus did not produce a discernable sympathoexcitation. With the colder stimuli, active tDCS had no effect on either hemodynamic or sympathetic nervous pressor responses. Hence, these data support prior findings of a shift in the pain perception threshold in healthy individuals, but do not support the postulate that non-invasive cortical stimulation via tDCS significantly modulates efferent cardiovascular control at rest or in response to pain. There is some prior work suggesting that non-invasive brain stimulation can alter efferent cardiovascular control. In anesthetized rats, TMS produced decreases in heart rate and mean pressure; however these effects were evanescent, lasting 4 and 11 s, respectively (Hong et al., 2002). Nonetheless, the transient blood pressure responses appeared to be due to inhibition of sympathetic vasomotor activity since they were only attenuated after alphaadrenergic blockade with prazosin and unaffected by cardiac sympathetic or parasympathetic blockade. In humans, longer duration TMS (6 min) resulted in significantly greater low-frequency heart rate variability as compared with a sham-stimulation group (Yoshida et al., 2001). However, both sham and active groups demonstrated increased high-frequency variability and changes
Table 1 Resting hemodynamics before and after sham and active anodal tDCS.
* **
Time
Condition
Heart rate, beats/ min
Systolic pressure, mm Hg
Diastolic pressure, mm Hg
Mean pressure, mm Hg
Leg blood flow, cm/s
Leg vascular resistance, units
Pre-tDCS Post-tDCS Pre-tDCS Post-tDCS
Sham Sham Active Active
61 ± 2 61 ± 2 60 ± 2 60 ± 2
119 ± 3* 128 ± 4* 117 ± 3* 122 ± 3*
66 ± 3 67 ± 4 64 ± 3 62 ± 2
82 ± 3 84 ± 4 79 ± 2 80 ± 2
10.4 ± 0.5** 10.2 ± 0.4** 12.2 ± 0.8** 11.4 ± 0.8**
6.45 ± 0.42 6.70 ± 0.46 5.66 ± 0.54 5.76 ± 0.44
Time p < 0.05. Condition p < 0.05.
1042
J.W. Hamner et al. / Clinical Neurophysiology 126 (2015) 1039–1046
Table 2 Indices of autonomic outflow before and after sham and active anodal tDCS. Time
Pre-tDCS Post-tDCS Pre-tDCS Post-tDCS * **
Condition
Sham Sham Active Active
Sympathetic activity, spikes/s
Sympathetic activity, spikes/100 beat
RR interval Low frequency power, m s2
High frequency power, m s2
Mean pressure Low frequency power, mm Hg2
High frequency power, mm Hg2
8.5 ± 2.5 9.1 ± 2.4 6.4 ± 1.3 8.0 ± 1.6
888 ± 261 972 ± 258 641 ± 130 788 ± 159
829 ± 249 706 ± 165 875 ± 194 856 ± 194
239 ± 46 378 ± 64 406 ± 119 373 ± 93
5.76 ± 1.19** 8.31 ± 1.91** 5.31 ± 1.86** 6.11 ± 1.87**
1.33 ± 0.31* 1.96 ± 0.38* 1.36 ± 0.37* 1.97 ± 0.56*
Time p < 0.05. Condition p < 0.05.
Fig. 1. Average pain scores (mean + standard error of the mean) in response to cold pressor tests before and after sham and active anodal tDCS.
in heart rate were unreported, so it remains unclear what the significance of the change in low-frequency variability might be to cardiac autonomic control. Data from Macefield et al. (1998) suggest that directly measured peroneal nerve muscle sympathetic activity can be inhibited by transient TMS over the cortex. They found that a single cortical stimulus, transiently applied after the occurrence of the R-wave of the ECG, reduced the amplitude of subsequent sympathetic bursts by up to 50%. However, there was a tendency for the inhibited burst to be followed by a burst that was larger. Hence, the total magnitude of effect on tonic sympathetic activity may not have been pronounced. On the other hand, Silber et al. (2000) found that focal stimulation of the M1 using single-pulse TMS predictably elicited increased sympathetic activity to skin but not to muscle. It should be noted that the intensity of these stimuli was above the motor threshold and elicited visible forearm and/or leg movements. In addition, the increase in cutaneous sympathetic activity has been inferentially observed via sweating responses and has been postulated to be a correlate of a nonspecific arousal reaction rather than direct modulation of autonomic centers of the brain (Niehaus et al., 1998). It should be noted that none of the above studies in humans provided any statistical assessment of the heart rate and blood pressure responses to cortical stimulation.
Fig. 2. Differences in pain scores in response to cold pressor tests after sham and active anodal tDCS.
Due to the equivocal nature of the findings above, we set out to determine if non-invasive cortical stimulation alters basal hemodynamics and cardiovascular autonomic outflow. Our results
J.W. Hamner et al. / Clinical Neurophysiology 126 (2015) 1039–1046
1043
Fig. 3. Peak increases in hemodynamic variables during 14 °C cold pressor test.
demonstrate that long-lasting neuromodulatory effects exerted by tDCS are not evidenced in changes in heart rate, blood pressure, vascular resistance, or sympathetic activity. We also were not able to replicate the prior finding that cortical stimulation has an effect on heart rate variability (Brunoni et al., 2013; Montenegro et al., 2011). It has been proposed that non-invasive brain stimulation might be useful for the management of arterial hypertension (Cogiamanian et al., 2010). The rationale is that nervous control is key to blood pressure control and is implicated in the pathogenesis of several forms of hypertension. Because the motor cortex is anatomically and functionally related to bulbar cardiovascular nuclei (Ba-M’Hamed et al., 1996; Viltart et al., 2003), it was suggested that non-invasive brain stimulation techniques that effectively modulate human cortical function in this area could influence blood pressure. Our data do not support this postulate. Prior work has shown that pain perception threshold in young healthy individuals is increased by anodal tDCS of the M1 (Boggio et al., 2008). Given the strong link between perception of pain and the cardiovascular responses to pain (Noseir et al., 2003), we sought to determine if lesser pain perception related to a smaller pressor response to a standard nociceptive stimulus, the cold pressor test. Our data do agree with the finding that tDCS alters pain perception
in healthy individuals at low levels of pain. After anodal tDCS, at the lowest level of nociceptive cold stimulus (14 °C), peak perceived pain was reduced from an average of 2.5 to 1.6 (both in the range of mild pain). This was accompanied by a tendency for a small reduction in the tachycardiac and pressor responses. However, the clinical significance of a change of such magnitude (2 beats/min and 3 mm Hg) is questionable. In addition, the reduction in perceived pain and the reduction in hemodynamic response were not correlated, perhaps due to the limited scale of the changes. At higher levels of pain with cold pressor tests at 7 and 0 °C, neither pain perception nor any cardiovascular variable were altered by tDCS. Thus, these data suggest that, in healthy subjects, anodal tDCS of the M1 modestly reduces the perception and pressor responses of mild pain but does not affect moderate to severe pain. 4.1. Limitations It should be noted that tDCS may exert greater effects with repeated sessions of stimulation. Hence, it is feasible that repeated sessions could result in neuroplastic effects that impact hemodynamic control and responses to pain. Further, while there is no evidence that the hemisphere stimulated alters the autonomic
1044
J.W. Hamner et al. / Clinical Neurophysiology 126 (2015) 1039–1046
Fig. 4. Peak increases in hemodynamic variables during 7 °C cold pressor test.
effects of tDCS (Schestatsky et al., 2013), direct cortical stimulation has demonstrated lateralization (Oppenheimer et al., 1992). Similarly, given that a number of cortical regions have been implicated in circulatory control (Verberne and Owens, 1998), different tDCS montages might produce different effects. Potential targets may include the dorsolateral prefrontal, parietal, occipital and frontal cortices (Schestatsky et al., 2013). During active tDCS, stimulation was delivered for 40 min until the end of the protocol. Throughout this period, baseline heart rate variability measurements were performed, followed by cold pressor tests applied in random order. However, baseline heart rate variability measurements were always made over a 5-min period starting 5 min after the beginning of the stimulation (i.e. from the fifth to the tenth minute of stimulation). Left M1 anodal tDCS at 1.0 mA for as little as 3 min, or stimulation at 0.6 mA for 5 min, have been shown to induce neuromodulatory effects in humans (Nitsche and Paulus, 2000). However, if 5–10 min of stimulation at 2.0 mA are insufficient for an observable autonomic effect, then this could explain why we saw no change in heart rate variability measures following active tDCS.
Although our data from microneurography provide direct measures of vascular sympathetic nervous outflow, our findings derive from a small subset of individuals and are not paired comparisons. However, if there were consistent effects of tDCS on sympathetic outflow that were not appreciated due to the limited number of successful recordings, this effect would be evidenced in leg vascular resistance, which is sympathetically controlled. However, there were no differences in vascular resistance caused by tDCS. Also, our data derive from healthy individuals with normal levels of blood pressure who have systems that may compensate for variations in blood pressure induced by autonomic effects of tDCS. However, the comprehensive nature of our measures (i.e., heart rate, blood pressure, vascular resistance, and directly measured sympathetic activity) mean that any compensation for the effects of non-invasive brain stimulation on blood pressure regulation would be unmasked. That is, if tDCS did preferentially alter either cardiac or vascular control of resting blood pressure and the other arm of blood pressure control compensated such that pressure were unchanged, our measures would have revealed this.
J.W. Hamner et al. / Clinical Neurophysiology 126 (2015) 1039–1046
1045
Fig. 5. Peak increases in hemodynamic variables during 0 °C cold pressor test.
Fig. 6. Average sympathetic nerve activity responses to cold pressor tests before tDCS.
Fig. 7. Differences in sympathetic nerve activity responses to cold pressor tests after sham and active anodal tD.
1046
J.W. Hamner et al. / Clinical Neurophysiology 126 (2015) 1039–1046
4.2. Conclusions Our results show that non-invasive brain stimulation via anodal tDCS of the M1 has no effect on basal hemodynamics or cardiovascular autonomic outflow. This form of cortical stimulation did have an effect on the responses to mild pain and so these data do agree with a possible shift in pain perception threshold in healthy individuals. However, they do not support the postulate that non-invasive cortical stimulation via tDCS significantly modulates efferent cardiovascular control at rest or in response to pain beyond the pain threshold. Future work should seek to clarify whether different electrode montages, stimulation parameters, or repeated sessions of stimulation can be more effective. Financial interests None. Acknowledgement We thank the subjects for their generous participation. Conflict of interest: The authors declare no conflicts of interest in relation to this article. References Ba-M’Hamed S, Roy JC, Bennis M, Poulain P, Sequeira H. Corticospinal collaterals to medullary cardiovascular nuclei in the rat: an anterograde and a retrograde double-labeling study. J Hirnforsch 1996;37:367–75. Binkofski F, Loebig M, Jauch-Chara K, Bergmann S, Melchert UH, Scholand-Engler HG, et al. Brain energy consumption induced by electrical stimulation promotes systemic glucose uptake. Biol Psychiatry 2011;70:690–5. Boggio PS, Zaghi S, Lopes M, Fregni F. Modulatory effects of anodal transcranial direct current stimulation on perception and pain thresholds in healthy volunteers. Eur J Neurol 2008;15:1124–30. Brunoni AR, Vanderhasselt MA, Boggio PS, Fregni F, Dantas EM, Mill JG, et al. Polarity- and valence-dependent effects of prefrontal transcranial direct current stimulation on heart rate variability and salivary cortisol. Psychoneuroendocrinology 2013;38:58–66. Cogiamanian F, Brunoni AR, Boggio PS, Fregni F, Ciocca M, Priori A. Non-invasive brain stimulation for the management of arterial hypertension. Med Hypotheses 2010;74:332–6. Csifcsak G, Antal A, Hillers F, Levold M, Bachmann CG, Happe S, et al. Modulatory effects of transcranial direct current stimulation on laser-evoked potentials. Pain Med 2009;10:122–32. de Vries MH, Barth AC, Maiworm S, Knecht S, Zwitserlood P, Floel A. Electrical stimulation of Broca’s area enhances implicit learning of an artificial grammar. J Cogn Neurosci 2010;22:2427–36. Fagius J, Karhuvaara S, Sundlof G. The cold pressor test: effects on sympathetic nerve activity in human muscle and skin nerve fascicles. Acta Physiol Scand 1989;137:325–34. Floel A, Rosser N, Michka O, Knecht S, Breitenstein C. Noninvasive brain stimulation improves language learning. J Cogn Neurosci 2008;20:1415–22.
Fregni F, Freedman S, Pascual-Leone A. Recent advances in the treatment of chronic pain with non-invasive brain stimulation techniques. Lancet Neurol 2007;6: 188–91. Gandiga PC, Hummel FC, Cohen LG. Transcranial DC stimulation (tDCS): a tool for double-blind sham-controlled clinical studies in brain stimulation. Clin Neurophysiol 2006;117:845–50. Goodson NJ, Smith BH, Hocking LJ, McGilchrist MM, Dominiczak AF, Morris A, et al. Cardiovascular risk factors associated with the metabolic syndrome are more prevalent in people reporting chronic pain: results from a cross-sectional general population study. Pain 2013;154:1595–602. Hong B, Kuwaki T, Ju K, Kumada M, Akai M, Ueno S. Changes in blood pressure and heart rate by repetitive transcranial magnetic stimulation in rats. Neurosci Lett 2002;329:57–60. Jensen MP, Sherlin LH, Askew RL, Fregni F, Witkop G, Gianas A, et al. Effects of nonpharmacological pain treatments on brain states. Clin Neurophysiol 2013;124:2016–24. Kregel KC, Seals DR, Callister R. Sympathetic nervous system activity during skin cooling in humans: relationship to stimulus intensity and pain sensation. J Physiol 1992;454:359–71. Lewis T. Pain. London: MacMillan; 1942. Macefield VG, Taylor JL, Wallin BG. Inhibition of muscle sympathetic outflow following transcranial cortical stimulation. J Auton Nerv Syst 1998;68:49–57. Montenegro RA, Farinatti PT, Fontes EB, Soares PP, Cunha FA, Gurgel JL, et al. Transcranial direct current stimulation influences the cardiac autonomic nervous control. Neurosci Lett 2011;497:32–6. Niehaus L, Meyer BU, Roricht S. Magnetic stimulation over different brain regions: no differential effects on the elicited sympathetic skin responses. Electroencephalogr Clin Neurophysiol 1998;109:94–9. Nitsche MA, Paulus W. Excitability changes induced in the human motor cortex by weak transcranial direct current stimulation. J Physiol 2000;527(Pt 3):633–9. Noseir RK, Ficke DJ, Kundu A, Arain SR, Ebert TJ. Sympathetic and vascular consequences from remifentanil in humans. Anesth Analg 2003;96:1645–50. table. O’Connell NE, Wand BM, Marston L, Spencer S, Desouza LH. Non-invasive brain stimulation techniques for chronic pain. Cochrane Database Syst Rev 2014;4:CD008208. Oppenheimer SM, Gelb A, Girvin JP, Hachinski VC. Cardiovascular effects of human insular cortex stimulation. Neurology 1992;42:1727–32. Raimundo RJ, Uribe CE, Brasil-Neto JP. Lack of clinically detectable acute changes on autonomic or thermoregulatory functions in healthy subjects after transcranial direct current stimulation (tDCS). Brain Stimul 2012;5:196–200. Schestatsky P, Simis M, Freeman R, Pascual-Leone A, Fregni F. Non-invasive brain stimulation and the autonomic nervous system. Clin Neurophysiol 2013;124: 1716–28. Silber DH, Sinoway LI, Leuenberger UA, Amassian VE. Magnetic stimulation of the human motor cortex evokes skin sympathetic nerve activity. J Appl Physiol 2000;88:126–34. Taylor JA, Studinger P. Counterpoint: cardiovascular variability is not an index of autonomic control of the circulation. J Appl Physiol 2006;101:678–81. Vandermeeren Y, Jamart J, Ossemann M. Effect of tDCS with an extracephalic reference electrode on cardio-respiratory and autonomic functions. BMC Neurosci 2010;11:38. Verberne AJ, Owens NC. Cortical modulation of the cardiovascular system. Prog Neurobiol 1998;54:149–68. Viltart O, Mullier O, Bernet F, Poulain P, Ba-M’Hamed S, Sequeira H. Motor cortical control of cardiovascular bulbar neurones projecting to spinal autonomic areas. J Neurosci Res 2003;73:122–35. Yoshida T, Yoshino A, Kobayashi Y, Inoue M, Kamakura K, Nomura S. Effects of slow repetitive transcranial magnetic stimulation on heart rate variability according to power spectrum analysis. J Neurol Sci 2001;184:77–80.
Contents lists available at ScienceDirect
Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph
Transcranial direct current stimulation (tDCS) and the cardiovascular responses to acute pain in humans J.W. Hamner, Mauricio F. Villamar, Felipe Fregni, J. Andrew Taylor ⇑ Department of Physical Medicine and Rehabilitation, Harvard Medical School and Spaulding Rehabilitation Hospital, Boston, MA, United States
a r t i c l e
i n f o
Article history: Accepted 14 August 2014 Available online 16 September 2014 Keywords: tDCS Brain stimulation Acute pain Cardiovascular Sympathetic Autonomic
h i g h l i g h t s We tested whether transcranial direct current stimulation (tDCS) alters both pain perception and its
cardiovascular responses and if it impacts basal hemodynamics. Active tDCS did not affect resting hemodynamics or autonomic outflow. Only with the least painful stimulus, tDCS modestly reduced perceived pain and the peak cardiovas-
cular and autonomic responses.
a b s t r a c t Objective: To determine if transcranial direct current stimulation (tDCS) reduces both acute pain perception and the resultant cardiovascular responses. Methods: Data were acquired on 15 healthy subjects at rest and in response to three cold pressor tests: 0, 7, and 14 °C. Subsequently, single sessions of sham and active anodal tDCS (2.0 mA for 40 min) were delivered to the left primary motor cortex (M1). Results: Perceived pain was reduced only after active tDCS with the 14 °C cold pressor test. This was accompanied by tendency for lesser increases in heart rate (2 beats/min, p = 0.09) and blood pressure (3 mm Hg, p = 0.06). The effect size of tDCS on peak heart rate and blood pressure responses at 14 °C was 0.47 and 0.54, respectively. On the other hand, baseline heart rate, blood pressure, leg blood flow, and leg vascular resistance were unaffected by tDCS. No other responses were affected. Conclusions: Our results demonstrate that M1 anodal tDCS has no effect on basal hemodynamics or cardiovascular autonomic outflow and has only modest effects on the responses to acute pain in healthy humans. Significance: Application of tDCS shifts the pain perception threshold in healthy individuals but does not significantly modulate efferent cardiovascular control at rest or in response to pain. Ó 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction In humans, acute painful stimuli increase heart rate, blood pressure and sympathetic nervous activity to the vasculature (Lewis, 1942). If one could modulate the responses to acute pain, this might have broader application to not only reduce the somatosensory burden of pain but also mitigate against its possible long-term cardiovascular effects (Goodson et al., 2013). One approach to treat pain is direct brain stimulation. However, until the 1990s only surgically ⇑ Corresponding author at: Cardiovascular Research Laboratory, Spaulding Hospital Cambridge, 1575 Cambridge Street, Cambridge, MA 02138, United States. Tel.: +1 617 758 5503; fax: +1 617 758 5514. E-mail address: [email protected] (J.A. Taylor).
implanted electrodes to achieve deep brain stimulation had been systematically tested and shown to induce significant decreases in pain. More recently, non-invasive, easy to administer approaches have generated increased interest as a potential therapeutic intervention (Fregni et al., 2007; Jensen et al., 2013). Transcranial direct current stimulation (tDCS) has long lasting modulatory effects on cortical function and allows a reliable sham-stimulation condition to assure specificity of effects. Therefore, tDCS might provide a testable avenue to both reduce pain perception and the resultant cardiovascular responses. There is evidence that tDCS can effectively modulate pain perception threshold in healthy individuals (Boggio et al., 2008; Csifcsak et al., 2009), indicating that its analgesic effects do not depend on aberrant neural activity. Beyond altering afferent pain
http://dx.doi.org/10.1016/j.clinph.2014.08.019 1388-2457/Ó 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
1040
J.W. Hamner et al. / Clinical Neurophysiology 126 (2015) 1039–1046
perception, there is data suggesting that brain stimulation could modulate efferent cardiovascular control (Schestatsky et al., 2013), and hence has been proposed as a potential tool for the management of hypertension, even independent of pain (Cogiamanian et al., 2010; Schestatsky et al., 2013). For example, high-frequency transcranial magnetic stimulation (TMS) reduces blood pressure and heart rate in rats, apparently through sympathoinhibition (Hong et al., 2002). tDCS-mediated reductions in heart rate and blood pressure have also been observed in humans (Binkofski et al., 2011), though this is not consistently found (de Vries et al., 2010; Floel et al., 2008; Raimundo et al., 2012; Vandermeeren et al., 2010). Even though there is also evidence of the contrary (Vandermeeren et al., 2010), some data suggest potential alterations in cardiac autonomic control with tDCS (Brunoni et al., 2013; Montenegro et al., 2011). However, this work relied on heart rate variability which can be a poor surrogate for direct measures of autonomic control (Taylor and Studinger, 2006). Directly measured muscle sympathetic outflow via microneurography has shown inhibition of pulse-synchronous activity after TMS (Macefield et al., 1998). In addition, microneurography data indicate that TMS alters sympathetic activity to skin (Silber et al., 2000; Niehaus et al., 1998). Hence, cortical stimulation may alter not only pain perception and its resultant cardiovascular responses but also basal homeostatic hemodynamics not related to pain perception. Nevertheless, despite evidence from TMS studies, direct measurements of the effects of tDCS on muscle sympathetic outflow via microneurography are still lacking. We tested the hypothesis that non-invasive cortical stimulation alters pain perception and the autonomic responses to acute pain. We chose M1 anodal tDCS to modulate cortical excitability since this montage has shown analgesic effects in the setting of experimental pain in healthy subjects (Boggio et al., 2008; Csifcsak et al., 2009) as well as in chronic pain patients (O’Connell et al., 2014), and because the M1 has been involved in the control of bulbar cardiovascular nuclei and vasomotor spinal preganglionic neurons (Ba-M’Hamed et al., 1996; Viltart et al., 2003). In addition to evaluating any potential changes induced by tDCS on baseline cardiovascular measures, we used the nociceptive response to immersion of the hand in cold water (cold pressor test) at three different temperatures as a provocative maneuver. The cold pressor test increases perceived pain, blood pressure, heart rate, vascular resistance, and sympathetic activity (Fagius et al., 1989), and the responses are directly related to decreasing water temperature (Kregel et al., 1992). A strong link between the perceived pain and the pressor responses is suggested by work showing that partial sedation proportionally reduces both perceived pain and cardiovascular responses to the cold pressor test (Noseir et al., 2003). Hence, by measuring hemodynamic variables (heart rate, blood pressure, leg blood flow, leg vascular resistance), as well as both indirect (RR interval, mean pressure variabilities) and direct indices (muscle sympathetic nerve recordings) of autonomic outflow, this design allowed us to examine whether tDCS alters perceived pain and its reflex cardiovascular responses.
2. Methods 2.1. Subjects Fifteen healthy young individuals aged 21–30 participated in this study (7 female). All subjects had a body mass index between 18.5 and 29.9 kg/m2 and a normal resting electrocardiogram (ECG). None of them had any signs or symptoms of cardiovascular or neurological diseases, recent weight change, regular use of tobacco, or current pregnancy. All subjects gave written informed consent prior to participating. This study was approved by the Institutional
Review Board at Spaulding Rehabilitation Hospital (Protocol #2011-P-001879/1) and conformed to the Declaration of Helsinki. 2.2. Procedures Subjects visited the laboratory on two separate mornings (around 8 a.m.) to receive either active or sham tDCS. Study visits were separated by a minimum of 7 days to a maximum of 8 weeks. Subjects abstained from vigorous exercise for 2 days prior to each study visit to avoid autonomic and neuroendocrine effects of exercise. In addition, subjects refrained from caffeine and alcohol for the previous 24 h. Upon arrival, participants were instrumented for tDCS and physiologic assessments, as detailed below. On each visit, assessments were performed first at baseline while no brain stimulation was being delivered, and then again during either active or sham tDCS. Over the two study visits, participants underwent one session of active and one session of sham tDCS. These two interventions were delivered in random and counterbalanced order. Throughout the protocol, subjects were supine on a laboratory table and were instrumented for measurement of standard lead II of the ECG, beat-by-beat blood pressure in a finger of the left hand (Portapres, Finapres Medical Systems), brachial oscillometric blood pressure (Dash 2000, GE), respiratory excursions from a respiratory bellows placed around the chest, and popliteal artery blood flow velocity at the popliteal fossa of the left leg (Multi-Dop T2 4-MHz Doppler probe; Compumedics DWL, Singen, Germany). After instrumentation and calibration, multiunit postganglionic muscle sympathetic nerve recording from the common peroneal nerve was successfully obtained in a single session (either sham or active tDCS) in 12 subjects. For all subjects in both sessions, data were acquired during a 5-min period of quiet rest and in response to three cold pressor tests (0, 7, and 14 °C) performed in random order. For each cold pressor test, a 1-min baseline period was followed by 3-min immersion of the right hand in cold water. During the immersion, subjects rated their perceived pain on a Visual Analog Scale for pain every 30 s (0–10, with 0 corresponding to absence of pain and 10 corresponding to the worst imaginable pain). Cold pressor tests were separated by 10 min, allowing hand temperature to normalize between trials. Subsequently, either sham or active tDCS was applied. Active and sham tDCS were delivered using 35 cm2 sponge electrodes. The anode was placed over the left primary motor cortex (M1), corresponding to C3 in the International 10–20 Electroencephalography System, and the cathode over the right supraorbital area. An ActivaDoseÒ II Iontophoresis Delivery Unit (ActivaTek Inc., Salt Lake City, UT) was used in all experiments. In active tDCS, current was ramped up over a period of 30 s until reaching 2.0 mA, which were applied for the remainder of the testing session (40 min). Current density was 0.057 mA/cm2. For the sham condition, the same instrumentation was used but direct current was only applied for 30 s (Gandiga et al., 2006). tDCS was administered by an unblinded investigator who was not involved in data analysis. Subjects were blinded to the type of stimulation and the tDCS device was kept out of their sight for the duration of the study. Once 5 min of stimulation (active or sham) had elapsed, data were again acquired during a 5-min period of quiet rest and in response to the three cold pressor tests performed in random order while the stimulation continued to be delivered. 2.3. Data and statistical analysis Values were derived for mean blood pressure ([2 diastolic + systolic]/3) and for leg vascular resistance (mean blood pressure/leg blood flow). Resting baseline values pre- and post- sham and active tDCS were derived from 5-min averages
1041
J.W. Hamner et al. / Clinical Neurophysiology 126 (2015) 1039–1046
for all variables (heart rate, blood pressure, leg blood flow, leg vascular resistance, and muscle sympathetic activity). In addition, the magnitude of variabilities in RR interval and mean pressure were derived from power spectral analysis. Though these measures may not accurately reflect cardiac autonomic or vascular sympathetic outflows (Taylor and Studinger, 2006), prior work suggests that there may be alterations in cardiovascular variabilities after tDCS (Brunoni et al., 2013; Montenegro et al., 2011). For power spectral analysis, the ECG and arterial pressure waveforms were peak detected, edited for artifact, and linearly interpolated to a frequency of 4 Hz before being divided into 5 equal segments that overlapped by 50%. Each segment was detrended, Hamming filtered, and fast-Fourier transformed to its frequency representation squared. These periodograms were then averaged to produce the spectrum estimate. Average power was calculated in the low (0.04–0.15 Hz) and high (0.15–0.4 Hz) frequency ranges. Peak responses to each cold pressor test were derived from the difference between the average baseline value preceding each test and the highest 30-s average during each test. Average pain scores and peak pain scores within each cold pressor test were determined. Repeated measures ANOVAs were used to determine differences in pain and cardiovascular responses within each temperature level (0, 7, and 14 °C) of cold pressor test. This provided significance levels for the effect of time (pre- versus post-tDCS), the effect of condition (sham versus active tDCS), and their interaction. A p < 0.05 was considered significant. All data are presented as means with standard error of the mean. 3. Results All tDCS sessions proceeded uneventfully and were not associated with adverse effects other than a mild-to-moderate tingling or itching sensation and transient erythema in the area of stimulation. These occurred with both active and sham stimulation. 3.1. Resting hemodynamics and autonomic outflow Resting hemodynamics were unaffected by tDCS (Table 1). Systolic pressure tended to be higher after both sham and active tDCS (Time p = 0.04), but this effect was not different between stimulation conditions (Time Condition p = 0.47). All other hemodynamic variables demonstrated no differences with time, no differences between condition, and no interaction between time and condition (lowest p = 0.55). Unpaired comparisons of those with muscle sympathetic nerve recordings (n = 12) showed that there was no effect of time or condition, and no interaction between time and condition (Table 2). Indices of RR interval and mean pressure variability also showed no pronounced effects of active tDCS on either low or high frequency powers (Table 2). 3.2. Responses to pain Perceived pain in response to the cold pressor test was graded with temperature (p < 0.05) and tended to be reduced with
repeated trials (p < 0.05), especially for the two greater cold pressor stimuli (Fig. 1). At 14 °C, the reduction in perceived pain was greater after active tDCS than sham tDCS (p < 0.05) (Fig. 2). This reduction in perceived pain was reflected in a tendency for a lesser peak increase in heart rate and blood pressure after active tDCS (time condition; Fig. 3). Post-hoc t-tests showed that the peak increases in heart rate and blood pressure tended to be lower by 2 beats/min (p = 0.09) and 3 mm Hg (p = 0.06) with effect sizes of 0.47 and 0.54, respectively. However, no other hemodynamic response to the 14 °C cold pressor test was affected by active (or sham) tDCS. With the 7 °C cold pressor test, all hemodynamic responses tended to be lower with repeated trials (Time p < 0.10) and active tDCS did not magnify this effect (Time Condition all p > 0.20; Fig. 4). With the 0 °C cold pressor test, hemodynamic responses were similar across repeated trials and were unaffected by active tDCS (Time Condition all p > 0.40; Fig. 5). Muscle sympathetic nerve responses to the cold pressor tests were highest with the 0 °C test and minimal with both the 14 °C and 7 °C tests (Fig. 6). At all temperatures, responses were similar across repeated trials and were unaffected by active tDCS (Fig. 7). 4. Discussion Our results demonstrate that non-invasive brain stimulation via anodal M1 tDCS has no significant effects on basal hemodynamics or cardiovascular autonomic outflow. In addition, this form of cortical stimulation had only a slight effect on the responses to acute pain in healthy humans. At the lowest level of the nociceptive cold stimulus, when peak perceived pain was low (a reported value of 1 or less in 45% of trials), there was a modest decrease in perceived pain with active compared to sham tDCS. This translated into a tendency for slightly lower peak responses in heart rate (2 beats/min) and blood pressure (3 mm Hg), but no other significant hemodynamic effects. It should be noted that this level of nociceptive cold stimulus did not produce a discernable sympathoexcitation. With the colder stimuli, active tDCS had no effect on either hemodynamic or sympathetic nervous pressor responses. Hence, these data support prior findings of a shift in the pain perception threshold in healthy individuals, but do not support the postulate that non-invasive cortical stimulation via tDCS significantly modulates efferent cardiovascular control at rest or in response to pain. There is some prior work suggesting that non-invasive brain stimulation can alter efferent cardiovascular control. In anesthetized rats, TMS produced decreases in heart rate and mean pressure; however these effects were evanescent, lasting 4 and 11 s, respectively (Hong et al., 2002). Nonetheless, the transient blood pressure responses appeared to be due to inhibition of sympathetic vasomotor activity since they were only attenuated after alphaadrenergic blockade with prazosin and unaffected by cardiac sympathetic or parasympathetic blockade. In humans, longer duration TMS (6 min) resulted in significantly greater low-frequency heart rate variability as compared with a sham-stimulation group (Yoshida et al., 2001). However, both sham and active groups demonstrated increased high-frequency variability and changes
Table 1 Resting hemodynamics before and after sham and active anodal tDCS.
* **
Time
Condition
Heart rate, beats/ min
Systolic pressure, mm Hg
Diastolic pressure, mm Hg
Mean pressure, mm Hg
Leg blood flow, cm/s
Leg vascular resistance, units
Pre-tDCS Post-tDCS Pre-tDCS Post-tDCS
Sham Sham Active Active
61 ± 2 61 ± 2 60 ± 2 60 ± 2
119 ± 3* 128 ± 4* 117 ± 3* 122 ± 3*
66 ± 3 67 ± 4 64 ± 3 62 ± 2
82 ± 3 84 ± 4 79 ± 2 80 ± 2
10.4 ± 0.5** 10.2 ± 0.4** 12.2 ± 0.8** 11.4 ± 0.8**
6.45 ± 0.42 6.70 ± 0.46 5.66 ± 0.54 5.76 ± 0.44
Time p < 0.05. Condition p < 0.05.
1042
J.W. Hamner et al. / Clinical Neurophysiology 126 (2015) 1039–1046
Table 2 Indices of autonomic outflow before and after sham and active anodal tDCS. Time
Pre-tDCS Post-tDCS Pre-tDCS Post-tDCS * **
Condition
Sham Sham Active Active
Sympathetic activity, spikes/s
Sympathetic activity, spikes/100 beat
RR interval Low frequency power, m s2
High frequency power, m s2
Mean pressure Low frequency power, mm Hg2
High frequency power, mm Hg2
8.5 ± 2.5 9.1 ± 2.4 6.4 ± 1.3 8.0 ± 1.6
888 ± 261 972 ± 258 641 ± 130 788 ± 159
829 ± 249 706 ± 165 875 ± 194 856 ± 194
239 ± 46 378 ± 64 406 ± 119 373 ± 93
5.76 ± 1.19** 8.31 ± 1.91** 5.31 ± 1.86** 6.11 ± 1.87**
1.33 ± 0.31* 1.96 ± 0.38* 1.36 ± 0.37* 1.97 ± 0.56*
Time p < 0.05. Condition p < 0.05.
Fig. 1. Average pain scores (mean + standard error of the mean) in response to cold pressor tests before and after sham and active anodal tDCS.
in heart rate were unreported, so it remains unclear what the significance of the change in low-frequency variability might be to cardiac autonomic control. Data from Macefield et al. (1998) suggest that directly measured peroneal nerve muscle sympathetic activity can be inhibited by transient TMS over the cortex. They found that a single cortical stimulus, transiently applied after the occurrence of the R-wave of the ECG, reduced the amplitude of subsequent sympathetic bursts by up to 50%. However, there was a tendency for the inhibited burst to be followed by a burst that was larger. Hence, the total magnitude of effect on tonic sympathetic activity may not have been pronounced. On the other hand, Silber et al. (2000) found that focal stimulation of the M1 using single-pulse TMS predictably elicited increased sympathetic activity to skin but not to muscle. It should be noted that the intensity of these stimuli was above the motor threshold and elicited visible forearm and/or leg movements. In addition, the increase in cutaneous sympathetic activity has been inferentially observed via sweating responses and has been postulated to be a correlate of a nonspecific arousal reaction rather than direct modulation of autonomic centers of the brain (Niehaus et al., 1998). It should be noted that none of the above studies in humans provided any statistical assessment of the heart rate and blood pressure responses to cortical stimulation.
Fig. 2. Differences in pain scores in response to cold pressor tests after sham and active anodal tDCS.
Due to the equivocal nature of the findings above, we set out to determine if non-invasive cortical stimulation alters basal hemodynamics and cardiovascular autonomic outflow. Our results
J.W. Hamner et al. / Clinical Neurophysiology 126 (2015) 1039–1046
1043
Fig. 3. Peak increases in hemodynamic variables during 14 °C cold pressor test.
demonstrate that long-lasting neuromodulatory effects exerted by tDCS are not evidenced in changes in heart rate, blood pressure, vascular resistance, or sympathetic activity. We also were not able to replicate the prior finding that cortical stimulation has an effect on heart rate variability (Brunoni et al., 2013; Montenegro et al., 2011). It has been proposed that non-invasive brain stimulation might be useful for the management of arterial hypertension (Cogiamanian et al., 2010). The rationale is that nervous control is key to blood pressure control and is implicated in the pathogenesis of several forms of hypertension. Because the motor cortex is anatomically and functionally related to bulbar cardiovascular nuclei (Ba-M’Hamed et al., 1996; Viltart et al., 2003), it was suggested that non-invasive brain stimulation techniques that effectively modulate human cortical function in this area could influence blood pressure. Our data do not support this postulate. Prior work has shown that pain perception threshold in young healthy individuals is increased by anodal tDCS of the M1 (Boggio et al., 2008). Given the strong link between perception of pain and the cardiovascular responses to pain (Noseir et al., 2003), we sought to determine if lesser pain perception related to a smaller pressor response to a standard nociceptive stimulus, the cold pressor test. Our data do agree with the finding that tDCS alters pain perception
in healthy individuals at low levels of pain. After anodal tDCS, at the lowest level of nociceptive cold stimulus (14 °C), peak perceived pain was reduced from an average of 2.5 to 1.6 (both in the range of mild pain). This was accompanied by a tendency for a small reduction in the tachycardiac and pressor responses. However, the clinical significance of a change of such magnitude (2 beats/min and 3 mm Hg) is questionable. In addition, the reduction in perceived pain and the reduction in hemodynamic response were not correlated, perhaps due to the limited scale of the changes. At higher levels of pain with cold pressor tests at 7 and 0 °C, neither pain perception nor any cardiovascular variable were altered by tDCS. Thus, these data suggest that, in healthy subjects, anodal tDCS of the M1 modestly reduces the perception and pressor responses of mild pain but does not affect moderate to severe pain. 4.1. Limitations It should be noted that tDCS may exert greater effects with repeated sessions of stimulation. Hence, it is feasible that repeated sessions could result in neuroplastic effects that impact hemodynamic control and responses to pain. Further, while there is no evidence that the hemisphere stimulated alters the autonomic
1044
J.W. Hamner et al. / Clinical Neurophysiology 126 (2015) 1039–1046
Fig. 4. Peak increases in hemodynamic variables during 7 °C cold pressor test.
effects of tDCS (Schestatsky et al., 2013), direct cortical stimulation has demonstrated lateralization (Oppenheimer et al., 1992). Similarly, given that a number of cortical regions have been implicated in circulatory control (Verberne and Owens, 1998), different tDCS montages might produce different effects. Potential targets may include the dorsolateral prefrontal, parietal, occipital and frontal cortices (Schestatsky et al., 2013). During active tDCS, stimulation was delivered for 40 min until the end of the protocol. Throughout this period, baseline heart rate variability measurements were performed, followed by cold pressor tests applied in random order. However, baseline heart rate variability measurements were always made over a 5-min period starting 5 min after the beginning of the stimulation (i.e. from the fifth to the tenth minute of stimulation). Left M1 anodal tDCS at 1.0 mA for as little as 3 min, or stimulation at 0.6 mA for 5 min, have been shown to induce neuromodulatory effects in humans (Nitsche and Paulus, 2000). However, if 5–10 min of stimulation at 2.0 mA are insufficient for an observable autonomic effect, then this could explain why we saw no change in heart rate variability measures following active tDCS.
Although our data from microneurography provide direct measures of vascular sympathetic nervous outflow, our findings derive from a small subset of individuals and are not paired comparisons. However, if there were consistent effects of tDCS on sympathetic outflow that were not appreciated due to the limited number of successful recordings, this effect would be evidenced in leg vascular resistance, which is sympathetically controlled. However, there were no differences in vascular resistance caused by tDCS. Also, our data derive from healthy individuals with normal levels of blood pressure who have systems that may compensate for variations in blood pressure induced by autonomic effects of tDCS. However, the comprehensive nature of our measures (i.e., heart rate, blood pressure, vascular resistance, and directly measured sympathetic activity) mean that any compensation for the effects of non-invasive brain stimulation on blood pressure regulation would be unmasked. That is, if tDCS did preferentially alter either cardiac or vascular control of resting blood pressure and the other arm of blood pressure control compensated such that pressure were unchanged, our measures would have revealed this.
J.W. Hamner et al. / Clinical Neurophysiology 126 (2015) 1039–1046
1045
Fig. 5. Peak increases in hemodynamic variables during 0 °C cold pressor test.
Fig. 6. Average sympathetic nerve activity responses to cold pressor tests before tDCS.
Fig. 7. Differences in sympathetic nerve activity responses to cold pressor tests after sham and active anodal tD.
1046
J.W. Hamner et al. / Clinical Neurophysiology 126 (2015) 1039–1046
4.2. Conclusions Our results show that non-invasive brain stimulation via anodal tDCS of the M1 has no effect on basal hemodynamics or cardiovascular autonomic outflow. This form of cortical stimulation did have an effect on the responses to mild pain and so these data do agree with a possible shift in pain perception threshold in healthy individuals. However, they do not support the postulate that non-invasive cortical stimulation via tDCS significantly modulates efferent cardiovascular control at rest or in response to pain beyond the pain threshold. Future work should seek to clarify whether different electrode montages, stimulation parameters, or repeated sessions of stimulation can be more effective. Financial interests None. Acknowledgement We thank the subjects for their generous participation. Conflict of interest: The authors declare no conflicts of interest in relation to this article. References Ba-M’Hamed S, Roy JC, Bennis M, Poulain P, Sequeira H. Corticospinal collaterals to medullary cardiovascular nuclei in the rat: an anterograde and a retrograde double-labeling study. J Hirnforsch 1996;37:367–75. Binkofski F, Loebig M, Jauch-Chara K, Bergmann S, Melchert UH, Scholand-Engler HG, et al. Brain energy consumption induced by electrical stimulation promotes systemic glucose uptake. Biol Psychiatry 2011;70:690–5. Boggio PS, Zaghi S, Lopes M, Fregni F. Modulatory effects of anodal transcranial direct current stimulation on perception and pain thresholds in healthy volunteers. Eur J Neurol 2008;15:1124–30. Brunoni AR, Vanderhasselt MA, Boggio PS, Fregni F, Dantas EM, Mill JG, et al. Polarity- and valence-dependent effects of prefrontal transcranial direct current stimulation on heart rate variability and salivary cortisol. Psychoneuroendocrinology 2013;38:58–66. Cogiamanian F, Brunoni AR, Boggio PS, Fregni F, Ciocca M, Priori A. Non-invasive brain stimulation for the management of arterial hypertension. Med Hypotheses 2010;74:332–6. Csifcsak G, Antal A, Hillers F, Levold M, Bachmann CG, Happe S, et al. Modulatory effects of transcranial direct current stimulation on laser-evoked potentials. Pain Med 2009;10:122–32. de Vries MH, Barth AC, Maiworm S, Knecht S, Zwitserlood P, Floel A. Electrical stimulation of Broca’s area enhances implicit learning of an artificial grammar. J Cogn Neurosci 2010;22:2427–36. Fagius J, Karhuvaara S, Sundlof G. The cold pressor test: effects on sympathetic nerve activity in human muscle and skin nerve fascicles. Acta Physiol Scand 1989;137:325–34. Floel A, Rosser N, Michka O, Knecht S, Breitenstein C. Noninvasive brain stimulation improves language learning. J Cogn Neurosci 2008;20:1415–22.
Fregni F, Freedman S, Pascual-Leone A. Recent advances in the treatment of chronic pain with non-invasive brain stimulation techniques. Lancet Neurol 2007;6: 188–91. Gandiga PC, Hummel FC, Cohen LG. Transcranial DC stimulation (tDCS): a tool for double-blind sham-controlled clinical studies in brain stimulation. Clin Neurophysiol 2006;117:845–50. Goodson NJ, Smith BH, Hocking LJ, McGilchrist MM, Dominiczak AF, Morris A, et al. Cardiovascular risk factors associated with the metabolic syndrome are more prevalent in people reporting chronic pain: results from a cross-sectional general population study. Pain 2013;154:1595–602. Hong B, Kuwaki T, Ju K, Kumada M, Akai M, Ueno S. Changes in blood pressure and heart rate by repetitive transcranial magnetic stimulation in rats. Neurosci Lett 2002;329:57–60. Jensen MP, Sherlin LH, Askew RL, Fregni F, Witkop G, Gianas A, et al. Effects of nonpharmacological pain treatments on brain states. Clin Neurophysiol 2013;124:2016–24. Kregel KC, Seals DR, Callister R. Sympathetic nervous system activity during skin cooling in humans: relationship to stimulus intensity and pain sensation. J Physiol 1992;454:359–71. Lewis T. Pain. London: MacMillan; 1942. Macefield VG, Taylor JL, Wallin BG. Inhibition of muscle sympathetic outflow following transcranial cortical stimulation. J Auton Nerv Syst 1998;68:49–57. Montenegro RA, Farinatti PT, Fontes EB, Soares PP, Cunha FA, Gurgel JL, et al. Transcranial direct current stimulation influences the cardiac autonomic nervous control. Neurosci Lett 2011;497:32–6. Niehaus L, Meyer BU, Roricht S. Magnetic stimulation over different brain regions: no differential effects on the elicited sympathetic skin responses. Electroencephalogr Clin Neurophysiol 1998;109:94–9. Nitsche MA, Paulus W. Excitability changes induced in the human motor cortex by weak transcranial direct current stimulation. J Physiol 2000;527(Pt 3):633–9. Noseir RK, Ficke DJ, Kundu A, Arain SR, Ebert TJ. Sympathetic and vascular consequences from remifentanil in humans. Anesth Analg 2003;96:1645–50. table. O’Connell NE, Wand BM, Marston L, Spencer S, Desouza LH. Non-invasive brain stimulation techniques for chronic pain. Cochrane Database Syst Rev 2014;4:CD008208. Oppenheimer SM, Gelb A, Girvin JP, Hachinski VC. Cardiovascular effects of human insular cortex stimulation. Neurology 1992;42:1727–32. Raimundo RJ, Uribe CE, Brasil-Neto JP. Lack of clinically detectable acute changes on autonomic or thermoregulatory functions in healthy subjects after transcranial direct current stimulation (tDCS). Brain Stimul 2012;5:196–200. Schestatsky P, Simis M, Freeman R, Pascual-Leone A, Fregni F. Non-invasive brain stimulation and the autonomic nervous system. Clin Neurophysiol 2013;124: 1716–28. Silber DH, Sinoway LI, Leuenberger UA, Amassian VE. Magnetic stimulation of the human motor cortex evokes skin sympathetic nerve activity. J Appl Physiol 2000;88:126–34. Taylor JA, Studinger P. Counterpoint: cardiovascular variability is not an index of autonomic control of the circulation. J Appl Physiol 2006;101:678–81. Vandermeeren Y, Jamart J, Ossemann M. Effect of tDCS with an extracephalic reference electrode on cardio-respiratory and autonomic functions. BMC Neurosci 2010;11:38. Verberne AJ, Owens NC. Cortical modulation of the cardiovascular system. Prog Neurobiol 1998;54:149–68. Viltart O, Mullier O, Bernet F, Poulain P, Ba-M’Hamed S, Sequeira H. Motor cortical control of cardiovascular bulbar neurones projecting to spinal autonomic areas. J Neurosci Res 2003;73:122–35. Yoshida T, Yoshino A, Kobayashi Y, Inoue M, Kamakura K, Nomura S. Effects of slow repetitive transcranial magnetic stimulation on heart rate variability according to power spectrum analysis. J Neurol Sci 2001;184:77–80.