Accepted Manuscript

Title: A meta-analysis of non-invasive brain stimulation and

autonomic functioning: Implications for brain-heart pathways
to cardiovascular disease
Author: Elena Makovac Julian F. Thayer Cristina Ottaviani
PII:
DOI:
Reference:

S0149-7634(16)30036-7
http://dx.doi.org/doi:10.1016/j.neubiorev.2016.05.001
NBR 2426

To appear in:
Received date:
Revised date:
Accepted date:

20-1-2016
24-4-2016
4-5-2016

Please cite this article as: Makovac, Elena, Thayer, Julian F., Ottaviani, Cristina, A metaanalysis of non-invasive brain stimulation and autonomic functioning: Implications
for brain-heart pathways to cardiovascular disease.Neuroscience and Biobehavioral
Reviews http://dx.doi.org/10.1016/j.neubiorev.2016.05.001
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

A meta-analysis of non-invasive brain stimulation and autonomic functioning: Implications

for brain-heart pathways to cardiovascular disease
Elena Makovaca, Julian F. Thayer b, Cristina Ottaviania,*

IRCCS Santa Lucia Foundation, Rome, Italy.

Department of 3Department of Psychology, The Ohio State University, Ohio, USA.

Corresponding author at: Neuroimaging Laboratory, IRCCS Santa Lucia Foundation, via
Ardeatina 306 - 00142 Rome, Italy. E-mail address: [email protected] (C.
Ottaviani).

Highlights
Non-invasive Brain Stimulation is effective in reducing HR and increasing HRV.

Stimulation technique moderates results with TMS being more effective than tDCS.

Moderation analysis showed that the PFC is the most appropriate brain site to stimulate.

ABSTRACT
Given the intrinsic connection between the brain and the heart, a recent body of research emerged
with the aim to influence cardiovascular system functioning by non-invasive brain stimulation
(NIBS) methods such as repetitive transcranial magnetic stimulation and transcranial direct current
stimulation. Despite the implications of cardiovascular activity modulation for therapeutic
purposes, such effects of NIBS have not yet been quantified. The aim of this study was to metaanalyze studies on NIBS effects on blood pressure (BP), heart rate (HR) and its variability (HRV).
PubMed and Scopus databases were searched for English language studies conducted in humans.
Twenty-nine studies were eligible for the analyses. Pooled effect sizes (Hedges g) were compared.
Random effect models were used. NIBS was effective in reducing HR (g = .17) and enhancing
HRV (g = .30). A marginal effect emerged for BP (g = .21). Significant moderators were the
stimulation technique and the site of stimulation. Results show that NIBS affects cardiovascular
and autonomic nervous system activity, confirming a potential pathogenic brain-heart pathway to
cardiovascular disease.

Keywords: Meta-analysis, Non-invasive brain stimulation, Transcranial magnetic stimulation,

Transcranial direct current stimulation, Autonomic nervous system, Cardiovascular

Contents
1. Introduction
2. Methods
2.1. Literature search and study selection
2.2. Coding
2.3. Data analyses
2.4. Moderator Analyses
3. Results
3.1. Blood Pressure
3.1.1. Moderator analysis
3.2. Heart Rate
3.2.1. Moderator analysis
3.3. Heart Rate Variability
3.3.1. Moderator analysis
4. Discussion
5. Limitations and conclusion

1. Introduction
Cardiovascular disease (CVD) has been recognized as the most common cause of death in the
world, accounting for 30% of all mortality (Lee and Cooper, 2009). Hypertension is one of the
primary risk factors for heart disease and stroke, the leading causes of death worldwide. In fact, the
World Health Report 2002 identified hypertension, or high blood pressure, as the third ranked
factor for disability-adjusted life years. One-quarter of the world's adult population has
hypertension and modeled projections indicate an increase to 1.15 billion hypertensive patients by
2025 in developing countries. As a consequence, hypertension places a heavy burden on society
and overall economic activity. Moreover, as the population ages, the economic impact of CVD on
health care systems will become even greater (Bromfield and Muntner, 2013).
A growing body of research demonstrates that psychosocial stresses play a crucial role in the
development of CVD (e.g., Cuffee et al., 2014), mainly via its effects on the autonomic nervous
system. A non-invasive measure, resulting from the dynamic interplay between the fast acting
parasympathetic nervous system and the relatively slower sympathetic nervous system, is heart rate
variability (HRV). HRV has been proposed as an index of the degree to which the autonomic
nervous system provides flexible and adaptive regulation to environmental stress (Thayer and Lane,
2000; 2009). Indeed, acute and chronic psychosocial stress not only increases heart rate (HR) but
also significantly reduces HRV in both human (e.g., Jarczok et al., 2013; Schubert et al., 2009) and
animals (see Sgoifo et al., 2014 for a review).
Despite the role of the brain in determining what is stressful to the individual, neuroimaging
methods have only recently been directed at understanding these brain-body pathways (McEwen
and Gianaros, 2010). Existing findings suggest that the orbitofrontal, anterior cingulate, and insular
brain regions act as a network to process the motivational aspects of environmental stimuli and
support adaptive cardiovascular reactions to stressors (Gianaros and Sheu, 2009; Gianaros et al.,
2012). A more extensive body of research exists on the identification of the reciprocal
interrelations between the autonomic nervous system and brain areas, the so-called Central
5

Autonomic Network (Benarroch, 1993). Taken together, these findings pointed to the primary
importance of the brain in the pathophysiology of essential hypertension (e.g., Julius and
Majahalme, 2000). In fact, the brain is implicated in the initiation of high blood pressure and
changes in brain function, structure, and organization correlate with the presence of hypertension
early in its course (see Jennings and Zanstra, 2009 for a review).
Nevertheless, most research in the field is correlational, and experimental manipulation is
necessary to increase our insight in the causal relationship between cortical functioning and
cardiovascular responses. In this context, techniques for non-invasive brain stimulation (NIBS) are
an interesting tool to investigate such a relationship.
To date, the NIBS techniques that have been most intensively studied with reference to the
cardiovascular and autonomic systems are repetitive transcranial magnetic stimulation (rTMS) and
transcranial direct current stimulation (tDCS). High frequency rTMS and anodal tDCS are able to
enhance the excitability of the stimulated hemisphere, while low frequency rTMS and cathodal
tDCS can suppress it.
TMS is a non-invasive tool for the electrical stimulation of neural tissue (Barker et al., 1985)
that can be applied as single pulses of stimulation, depolarizing neurons and evoking measurable
effects. Repetitive TMS consisting of trains of stimuli has the potential to modify excitability of the
cerebral cortex at the stimulated site and at remote areas along functional anatomical connections
(see Rossini et al., 2010 for a review).
TDCS differs from TMS as it does not induce neuronal firing by supra-threshold neuronal
membrane depolarization but rather modulates spontaneous neuronal network activity via the
application of weak electrical currents to different cortical areas. At the neuronal level, the primary
mechanism of action is the induction of polarity-dependent changes in cortical excitability (Priori et
al., 1998). Animal and human studies have provided insight regarding the mechanisms underlying
tDCS effects on neuroplasticity and showed that tDCS could induce specific changes in
neuropsychological, psychophysiological and motor activity as a function of targeted brain areas
6

(see Brunoni et al., 2012 for a review).

Despite accumulating evidence on the effects of NIBS on the cardiovascular system, such
effects have not yet been systematically quantified. Up to now, only a systematic review has been
conducted on the topic, highlighting the heterogeneity of included studies and the impossibility to
draw conclusive evidence of the effects of NIBS on cardiovascular and autonomic systems
(Schestatsky et al., 2013). This is mainly due to the fact that most of existing studies have been
designed with the aim to understand the safety of NIBS using cardiovascular parameters and not to
study the brain-heart connections.
To overcome these limitations, the aim of the present work was to quantify existing evidence
supporting the effects of NIBS on the cardiovascular and autonomic nervous systems. In choosing
our outcome measures, we decided to focus on those physiological markers that play a role as
established risk factors for cardiovascular disease, such as blood pressure (BP), HR, and HRV.
Resting high BP (e.g., Wilson et al., 1998) and HR have been shown to be independent risk factors
for cardiovascular disease and mortality (Custodis et al., 2013) and prolonged low HRV has been
associated with cardiovascular disease (Thayer et al., 2010 for a review).
The research question of the present study is: Can NIBS techniques be effectively used to
significantly reduce BP and HR and increase HRV?. Given the role of these biomarkers in the
development of cardiovascular disease, these observations will allow us to construct theoretical
models for translational science, ultimately to inform the development of personalized prevention
strategies. To better inform therapeutic interventions, an integrative investigative approach is called
for, combining techniques including functional brain imaging and detailed autonomic monitoring.

2. Methods
2.1. Literature search and study selection
Two search strategies were used to systematically collect empirical studies of the effects of
NIBS on BP, HR, and HRV. First, Medline (http://www.pubmed.com) and Scopus databases were
7

searched for English-language publications through January 10th 2016. Separate searches were
conducted for the following keywords: Transcranial Magnetic Stimulation OR TMS combined with
blood pressure, heart rate, heart rate variability, cardiovascular, and autonomic nervous system.
The same searches were then repeated for Transcranial Direct Current Stimulation OR tDCS.
Second, the reference lists of previous systematic reviews were searched for relevant studies.
The search was limited to English-language publications and human samples. Inclusion criteria
for our analysis were as follows: a) use of TMS or tDCS, b) recorded BP, HR, and/or HRV, and c)
utilized a design suitable for calculating one or more effect sizes. Reasons for exclusion were: a)
review articles, b) case reports; c) articles examining the effects of NIBS applied to other parts of
nervous system apart from the brain.
A total of 730 results were retrieved. Comparison of the retrieved titles identified 264 studies
that were duplicates, thus leaving 466 abstracts for further evaluation (see Figure 1). The current
series of meta-analyses is based on data extracted from 29 studies that met the inclusion criteria (see
Table 1 and references marked with an asterisk in the reference list) and had BP, HR, and HRV as
outcomes. Among the 29 studies, additional data (not published in the reviewed article but needed
to calculate effect sizes or to run moderator analyses) were received for 4 studies.
Guidelines from the Task Force of the European Society of Cardiology and the North
American Society of Pacing and Electrophysiology (Task Force, 1996) were used to define the
HRV measurements to include in the pertinent meta-analysis. We decided to focus on vagallymediated HRV because, to date, this is the only HRV measure that has definite structural and
functional brain correlates (Thayer et al., 2012; Winkelmann et al., in press). Studies that reported
respiratory sinus arrhythmia (RSA), root-mean-square successive RR-interval difference (RMSSD);
mean successive difference (MSD); or any spectral measure in the high frequency (HF) range of
0.15 - 0.4Hz (natural log transform (lnHF), normalized (HF n.u.) or absolute power (absHF)) were
included in our meta-analysis that had HRV as the outcome measure. Given the sequential inclusion
of HRV measures, low-frequency HRV (LF-HRV) was not considered because it is not strictly a
8

measure of vagally-mediated HRV and its interpretation is controversial (e.g., Reyes Del Paso et al.,
2013). For this reason, LF-HRV would require a separate meta-analysis comprising the other nonvagal HRV measures, which would be beyond the scope of the present meta-analysis.

2.2. Coding
A standardized data coding form was developed to extract the following information from each
study: a) authors and publication year; b) study design; c) characteristics of the study sample (age,
sex, size, subgroups); d) NIBS method used; e) outcomes of interest; f) adjusted covariates; and g)
brief results. Following the indications of Cooper (2009), each study (and each participant) was
included only once in one of our meta-analyses.
Each research article was read and analyzed by 2 members of the research team. Intercoder
agreement was 100%.
For tDCS, when both anodal and cathodal stimulation were present (and compared with sham
stimulation), we opted for the anodal versus sham for two reasons: a) we were interested in
stimulation (and not in inhibition) of cortical areas; b) whereas it is well established that anodal
tDCS increases cortical excitability, the effects of cathodal tDCS are still a matter of debate (e.g.,
Monte Silva et al., 2010).
For TMS, when high- and low-frequency forms of stimulation were used, we chose the highfrequency stimulation due to findings from the existing systematic review suggesting that highfrequency TMS might be the optimal strategy to evoke cardiovascular and autonomic system
responses (Schestatsky et al., 2013).
When both left- and right-sided stimulation were used, we based our choice on theoretical and
empirical assumptions on the lateralized properties of cortical structures.
A study conducted by Pascual-Leone et al. (1993) has been excluded from the analyses because
it had 9 participants, each of them stimulated at different frequencies and in different scalp
positions.
9

The quality of evidence was assessed with a modified version of the Newcastle-Ottawa Scale
(NOS; Wells et al., 2011). We modified the criteria so that the subcategories would be applicable
to psychophysiological studies. In our modified NOS scale the maximum score that could be
achieved by a study was 7 stars (see Appendix A for the methodological quality scores attributed to
each study included in the present meta-analyses).

2.3. Data analyses

For each study (or subsample of a study), we calculated a Hedges g effect size. Based on
conventional standards, effect sizes of g equal to .20, .50, and .80 were considered small, medium,
and large, respectively (Cohen, 1988).

The g coefficient represents the difference between

responding to a NIBS induction and responding to a control condition (baseline or sham) divided by
the pooled standard deviation (Hedges & Olkin, 1985). Effect sizes indicating bigger decreases in
BP and HR and increases in HRV (change scores from baseline) associated with NIBS compared to
a control condition got a positive sign.
Calculation of effect sizes was based on means, standard deviations, difference in mean scores,
P values, and sample sizes of the groups. When not presented explicitly, the mean change in each
group was obtained by subtracting the final mean from the baseline mean. When the standard
deviation (SD) of the changes was not provided, we imputed a change-from-baseline standard
deviation using a correlation coefficient as indicated by Higgins and Green (2011): SDchange =
(SD2baseline + SD2final (2 * Corr * SDbaseline * SDfinal)). When only standard errors (SE) were
provided, standard deviations were obtained by applying the following formula SD = SE * n
(Higgins & Green, 2011). When studies did not provide raw data to calculate effect sizes and
instead provided statistics (e.g., t), we applied transformation formulas to convert to g (Lipsey &
Wilson, 2001). When a paper reported p < .05 or n. s., we computed Hedges g with p-values of
.045 and 1 (one-tailed), respectively, which likely yielded a highly conservative estimate of the
effect size. For studies reporting RR interval values (duration of 1 heartbeat in milliseconds), HR
10

was calculated and its corresponding variance was estimated proportionally to the RR interval
variance. Where citations reported multiple vagally-mediated HRV indices, a hierarchical inclusion
was implemented to prevent conflation of effect size estimates as follows: HF power only, next
RSA only, next RMSSD only, next MSD only, or else pNN50 only.
The effect sizes were computed in ProMeta Version 2.0 (Internovi). Random-effects models
were used in all the analyses as they account for the amount of variance caused by differences
between associations as well as differences among participants within associations. ProMeta also
computed 95% confidence intervals (CI) around the point estimate of an effect size. The Q and I2
statistics were used to assess heterogeneity among studies. A significant Q value indicates a lack of
homogeneity of findings among studies. I2 values of .25, .50, and .75 correspond to low, moderate,
and high between-trial heterogeneity, respectively.
The problem of publication bias was estimated informally by using a funnel plot of effect size
against standard error for asymmetry and formally by using Begg and Mazumdars rank
correlations, and Eggers regression intercept test. We did not rely on failsafe N because, despite its
popularity, it has been considered inadequate and problematic as a method to assess publication bias
(e.g., Becker, 2005). In fact, as reported by Higgins and Green (2011), the estimate of fail-safe N is
highly dependent on the assumed mean effect for the unpublished studies and runs against the
principle to concentrate on the size of the estimated effect and the associated confidence intervals,
rather than on whether the P value reaches a particular, arbitrary threshold. Publication bias is
mostly due to the tendency of journals and authors to publish studies with positive results rather
than those with negative or non-significant results. Moreover, studies with small sample sizes need
to detect larger effects to be published compared with studies with large samples.
As the main aim of a meta-analysis is to aggregate over all data, we first run the analyses
including the entire set of studies and then subsequently re-run them without some potential
outliers, to examine the impact of these specific studies. Potential outliers were excluded if they
had statistically significant standardized residuals (Ellis, 2010).
11

Statistics reported in this meta-analysis conformed to the Preferred Reporting Items for
Systematic Reviews and Meta-Analyses (PRISMA; Moher et al., 2009) statement (see Appendix B
for PRISMA checklist).

2.4. Moderator analyses

For each outcome we examined how the size of the association varied as a function of mean
age (years), sex (% of women), stimulation intensity (high versus low), stimulation duration (min),
number of sessions, presence of a sham condition (yes versus no), site of stimulation (motor cortex,
PFC, or other), sample (healthy versus pathological), presence of a stressful task/condition (yes
versus no), if the aim was to test the effects of NIBS on the cardiovascular system (yes versus no),
and methodological quality (07).
Sex was examined as a moderator in light of reported prominent sex differences in the
neural control of the autonomic nervous system (e.g., Nugent et al., 2011). Stimulation intensity,
duration, and number of sessions are crucial factors affecting the outcome of TMS (Robertson et al.,
2003) and tDCS (Kekic et al., 2015). The presence of sham stimulation is crucial for analyzing the
effects of NIBS on cardiovascular responses, since the parasympathetic nervous system rapidly
responds to almost every type of stimuli, including internal thoughts (Ottaviani et al., 2016). As to
sample status, it is important to establish if TMS and tDCS can be considered as powerful tools for
prevention only or also for therapy. It would be particularly relevant to establish if NIBS is
effective for pathological conditions that are critically characterized by autonomic nervous system
dysfunctions. Lastly, given the above-mentioned role of stress for CVD risk, it is important to
examine whether NIBS is effective in reducing not only cardiovascular activation at rest but also in
response to stressors.
A minimum of 4 studies for each subgroup was required for the moderation analysis.
Continuous moderators were evaluated using meta-regression (i.e., age, sex, stimulation
duration, number of sessions, methodological quality), while categorical moderators were entered
12

as grouping variables in the effect size calculations. For continuous moderators, the sign of the
coefficient indicates the direction of the relation between NIBS and cardiovascular response.

3. Results
Table 1 discloses the specific contrast or test that were used to extract effect sizes in the present
series of meta-analyses. Studies marked with an asterisk in the table and figures indicate potential
outliers.

3.1. Blood pressure

Analysis of 13 studies (300 subjects) showed a marginally significant association between
NIBS and diminished BP (g = .21, 95% CI (-.02, .44), p = .069) in a heterogeneous set of studies (Q
(12) = 32.7, p = .001; I2 = 63.3). Figure 2 shows the forest plot. We did not find evidence of
asymmetry using a classic funnel plot (see Appendix C, panel a), Beggs rank test (Z = 0.49; p =
.63), or Eggers regression test (intercept = -1.04, t = -0.37, p = .72). No potential outliers emerged
in this meta-analysis.
3.1.1. Moderation analysis
No significant moderators emerged in this meta-analysis.

3.2. Heart rate

The meta-analysis showed significant associations between NIBS and decreased HR (18
studies; 348 subjects; g = .17, 95% CI [.01, .33], p = .04), which was small in size. Figure 3
presents the forest plot. Significant heterogeneity was shown by the Q and I2 statistics, Q (17) =
31.2, p = .02; I2 = 45.5. Kendalls tau excluded the presence of a publication bias (Z = 0.11; p =
.91), and this was confirmed by Begg and Mazumdars rank correlation test (intercept = -0.47, t = 0.26, p = .80 and by the funnel plot (see Appendix C, panel b). No potential outliers were
identified.
13

3.2.1. Moderation analysis

To have a comparable number of studies in each subgroup, in this meta-analysis the site of
stimulation had to be recoded in PFC versus other, thus aggregating M1 and other.
Contrasting studies with and without a sham condition yielded significant differences, Q (1) =
5.08, = p = .02, with HR decreases after NIBS only in studies that did not include such a condition
(g = .36, 95% CI [.10, .62], k = 8, n = 140) compared with those that did (g = .01, 95% CI [-.15,
.16], k = 10, n = 208). It has to be noted that only studies without a sham condition showed
substantial heterogeneity, Q (7) = 15.7; p = .03; I2 = 55.3.
The site of stimulation played a marginally significant role as a moderator, Q (1) = 3.26, p =
.06, with only studies that stimulated the PFC showing effects of NIBS on HR (g = .37, k = 8, n =
148 vs g = .09, k = 9, n = 185) in two homogeneous sets of studies.
A trend toward significant differences was also found on the basis of the type of stimulation
used, Q (1) = 2.89, p = .08, with significant HR decreases to NIBS only in studies that used TMS (g
= .23, 95% CI [.04, .42], k = 14, n = 254) compared with studies using tDCS (g = -.04, 95% CI [.29, .21], k = 4, n = 94). Only studies using TMS showed substantial heterogeneity, Q (13) = 27.2;
p = .01; I2 = 52.2.

3.3. Heart rate variability

The overall combined effect size for the total set of 15 studies (328 subjects) was significant
(g .30, 95% CI [.02, .58], p = .04; see Figure 4 for the forest plot) in a heterogeneous set of studies,
Q (14) = 60.97, p < .0001; I2 = 77.04.
Evidence of publication bias was detected by an asymmetrical funnel plot (see Appendix C,
panel c). The bias was neither confirmed by Beggs rank test (Z = -0.74; p = .46), nor by Eggers
regression test (intercept = -3.63, t = -1.60, p = .13).
Exclusion of an extreme outlier (Hamner et al., 2015) significantly increased the effect size (g

14

= .42, 95% CI [.21, .63], p < .0001) but not heterogeneity, Q (13) = 30.5; p = 004; I2 = 57.4. After
outlier exclusion, the funnel plot was no longer asymmetrical.
3.3.1. Moderation analysis
Contrasting studies using TMS with studies using tDCS yielded a significant difference, Q (1)
= 6.36, p = .01, with only the first being able to detect effects on HRV in response to the
stimulation (g = .60, k = 8, n = 165 vs g = -.10, k = 7, n = 163). Both sets of studies presented
significant heterogeneity, Q (12) = 17.4, p = .02; I2 = 59.9 and Q (12) = 28.5, p < .0001; I2 = 78.9,
respectively. Outlier removal had no impact on this moderation analyses.
The site of stimulation emerged as a marginally significant moderator, Q (2) = 4.63, p = .08,
with studies that stimulated the PFC (g = .43, k = 6, n = 153) and other areas such as Cz, T3, painrelated areas and the scalp (5 cm anterior to the hand area on parasagittal plane) (g = .56, k = 5, n =
109) but not studies stimulating the M1 area (g = -.44, k = 4, n = 66). It has to be noted that only
studies stimulating other areas and M1 showed substantial heterogeneity, Q (4) = 13.2; p = .01; I2 =
69.7 and Q (3) = 27.6; p < .0001; I2 = 89.1, respectively.

4. Discussion
In a series of meta-analyses we investigated the association between NIBS and BP, HR, and
increased HRV. HR and (marginally) BP were diminished, whereas HRV was enhanced by NIBS.
Even if effects were only small, moderator analysis suggested that there is the potential for larger
effects if future studies take a series of precautions as described here below. Overall, effects were
bigger in studies that used TMS instead of tDCS and stimulated the PFC instead of the primary
motor cortex or other brain regions (mainly Cz, T3, and pain-related areas).
First, studies using TMS were more effective than those using tDCS with up to medium effects
on HRV. This result is not surprising if we consider the technical difficulties in performing tDCS
in humans, in which the electrical current induced into the cortex reaches only a few centimeters in
15

depth in a non-selective way. Theoretically, TMS is a more adequate tool to investigate ANS
function because of its focality over cortical autonomic areas (Tremblay et al., 2014).
Second, the PFC appeared to be the golden target to stimulate to obtain effects on the
cardiovascular system. Unfortunately, as shown in Table 1 most of the examined studies did not
have cardiovascular changes as their target and only recorded HR and BP for safety issues. In
general, most research on the effects of NIBS has focused on the motor cortex, because of the ease
of obtaining the motor-evoked potential as an objective measure of the effectiveness of brain
stimulation. This is an unfortunate circumstance given that this site of stimulation was found to be
the least effective to reduce cardiovascular activation. Not surprisingly, the PFC was associated
with the most effective stimulation in terms of HR reductions and HRV increases. The PFC plays a
critical role in the representation of both internal and external context in the brain and the use of
such information to regulate behavior and peripheral physiology. Animal and human
pharmacological blockade, lesion, functional and structural neuroimaging studies implicate the PFC
in the regulation of cardiovascular autonomic activity (Ahern et al., 2001; Buchanan et al., 2010;
Gianaros, Van Der Veen, & Jennings, 2004; Lane et al., 2009; Ter Horst & Postema, 1997;
Winkleman et al., in press). Moreover, a meta-analysis of brain regions associated with HRV
identified the PFC as a core area in the regulation of HRV (Thayer et al., 2012). The present
findings complement this literature and suggest that this region may be a fruitful site for potential
intervention in both psychological and physiological dysregulation. Future NIBS are needed to
more fully understand the precise stimulation parameters that might be most beneficial.
Although we made an attempt to consider sample composition in our moderator analyses, this
analysis is inadequate (healthy versus pathological) to fully clarify the role of such a crucial factor
on the effects of NIBS at a cardiovascular level. The lack of significant moderation effects should
not be taken as a sign of NIBS being equally effective in healthy and pathological subjects, given
the substantial heterogeneity in the pathological sample (major depressive disorder, autism
spectrum disorder, migraine, bulimia nervosa, stroke, and spinal cord injury). Given the promising
16

result of the present meta-analysis, it is indeed possible that NIBS would emerge to be particularly
effective for a specific disease or psychopathological condition in which the autonomic or
cardiovascular systems play an important role. From a clinical perspective, brain stimulation has
the aim to induce long-term changes in cortical excitability. Unfortunately, none of the examined
studies was prospective or had an adequate follow-up assessment. The present meta-analytic
findings highlight the need for prospective studies in order to test the stability of the effects of NIBS
techniques on the cardiovascular and autonomic nervous systems and further elucidate mediating
and confounding factors.
Another surprising result was the lack of differences between studies examining resting
cardiovascular parameters and studies having participants under a stressful task or condition.
Again, this may be due to the heterogeneity of the examined stressful conditions, which ranged
from pain induction, to emotional pictures viewing, strenuous exercise and even pharmacologically
induced panic attack. Moreover, being that cardiovascular functioning was only a secondary
outcome measure in most studies, it has to be noted that the assessment of BP, HR, and HRV was
often far from the gold standard.
The present series of meta-analyses quantified the efficacy of NIBS in modifying several
established risk factors for CVD and provide preliminary -although modest- support to the
suggested use of NIBS as a further therapeutic tool for arterial hypertension (Cogiamanian et al.,
2010).

5. Limitations and conclusion

Several limitations need to be acknowledged. First, we did not include studies published in
languages other than English. Fortunately, analyses suggest that publication bias is unlikely to have
influenced our results. In fact publication bias appeared in only one of the examined physiological
variables (HRV) and seemed to be related to the presence of an extreme outlier.
Second, there was marked heterogeneity across studies, the sources of which subgroup analyses
17

were unsuccessful in identifying. To address this limitation, random-effects models were used in
all the analyses. Nevertheless, there are likely to be several other important moderators that we did
not consider.
Also, the effects of NIBS on other cardiovascular parameters, such as peripheral resistance or
cardiac output, could not be quantified by the present meta-analysis due to the lack of studies. The
assessment of multiple physiological markers is likely to increase the robustness of findings and
therefore should be taken as another target for future studies.
Taken together, our series of meta-analyses suggests promising evidence that NIBS could be
tested as a potential therapeutic tool to diminish HR and enhance HRV. The modest but significant
sizes of the effects of NIBS seem comparable with those of lifestyle interventions such as smoking
cessation or increased physical exercise (Baena et al., 2014). It is important to keep in mind,
however, that with the exception of one study conducted on patients after stroke, none of the other
examined studies had a sample composed by patients with cardiovascular problems. For this
reason, in the current stage these techniques may yield a potential for prevention rather than
therapy. Clearly, there is a need for studies using TMS on the appropriate brain site (PFC or even
better vmPFC), conducted on both at risk patients and controls, with a more sophisticated
cardiovascular assessment and lasting for more than the short assessment that characterizes basic
research.

18

Acknowledgements
C.O. was supported by a Young Investigator Research Grant from the Italian Ministry of Health
(GR-2011-02348232). J.F.T. was supported by the US National Science Foundation grant IIS
1344825.

19

Appendix
Appendix A. Adapted version of the Newcastle-Ottawa Scale used in the present study for
quality assessment.

Newcastle-Ottawa Scale adapted (maximum 7 stars)

Selection: (Maximum 2 stars)
1) Representativeness of the general population sample:
a) Truly representative of the average in the target population. * (all subjects or random
sampling)
b) Somewhat representative of the average in the target population. * (non-random sampling)
c) Selected group of users.
d) No description of the sampling strategy.

2) Sample size:
a) Justified and satisfactory. *
b) Not justified.

Confounds: (Maximum 2 stars)

1) The subjects in different outcome groups are comparable, based on the study design or analysis.
For within-subject or correlational studies, confounding factors are controlled for or the sample has
acceptable homogeneity.
a) The study controls for the most important factors (Age, gender, BMI, baseline physiology). *
b) The study control for any additional factor (sham condition). *

Outcome: (Maximum 3 stars)

1) Assessment of the outcome (physiological variable):
20

a) Acceptable length, method of recording, artifact correction. **

b) Not the gold standard measurement but described. *
c) No description.
2) Statistical test:
a) The statistical test used to analyze the data is clearly described and appropriate, and the
measurement of the association is presented, including confidence intervals and the probability
level (p value). *
b) The statistical test is not appropriate, not described or incomplete.

21

Appendix B. PRISMA checklist

# Checklist item

Reported
on page #

1 Identify the report as a systematic review, meta-analysis, or both.

2 Provide a structured summary including, as applicable: background; objectives; data sources; study
eligibility criteria, participants, and interventions; study appraisal and synthesis methods; results;
limitations; conclusions and implications of key findings; systematic review registration number.

2; no
limitations
due to
space
constraints;
no
registration
number

Rationale

3 Describe the rationale for the review in the context of what is already known.

4-6

Objectives

4 Provide an explicit statement of questions being addressed with reference to participants, interventions,
comparisons, outcomes, and study design (PICOS).

6-7

Protocol and registration

5 Indicate if a review protocol exists, if and where it can be accessed (e.g., Web address), and, if available,
provide registration information including registration number.

No
registered
review
protocol

Eligibility criteria

6 Specify study characteristics (e.g., PICOS, length of follow-up) and report characteristics (e.g., years
considered, language, publication status) used as criteria for eligibility, giving rationale.
7 Describe all information sources (e.g., databases with dates of coverage, contact with study authors to
identify additional studies) in the search and date last searched.

6-7

Section/topic
TITLE
Title
ABSTRACT
Structured summary

INTRODUCTION

METHODS

Information sources

6-7

22

Search

8 Present full electronic search strategy for at least one database, including any limits used, such that it could
be repeated.

6-7, Figure
1.

Study selection

9 State the process for selecting studies (i.e., screening, eligibility, included in systematic review, and, if
applicable, included in the meta-analysis).

6-7, Figure
1

10 Describe method of data extraction from reports (e.g., piloted forms, independently, in duplicate) and any
processes for obtaining and confirming data from investigators.
11 List and define all variables for which data were sought (e.g., PICOS, funding sources) and any
assumptions and simplifications made.

7-8

Risk of bias in individual

studies

12 Describe methods used for assessing risk of bias of individual studies (including specification of whether
this was done at the study or outcome level), and how this information is to be used in any data synthesis.

9-10

Summary measures

13 State the principal summary measures (e.g., risk ratio, difference in means).

Synthesis of results

14 Describe the methods of handling data and combining results of studies, if done, including measures of
consistency (e.g., I2) for each meta-analysis.

Data collection process

Data items

Section/topic

# Checklist item

7-11, Table
1

Reported
on page #

15 Specify any assessment of risk of bias that may affect the cumulative evidence (e.g., publication bias,
selective reporting within studies).
16 Describe methods of additional analyses (e.g., sensitivity or subgroup analyses, meta-regression), if done,
indicating which were pre-specified.

9-10

Study selection

17 Give numbers of studies screened, assessed for eligibility, and included in the review, with reasons for
exclusions at each stage, ideally with a flow diagram.

6-8, Figure
1

Study characteristics

18 For each study, present characteristics for which data were extracted (e.g., study size, PICOS, follow-up
period) and provide the citations.
19 Present data on risk of bias of each study and, if available, any outcome level assessment (see item 12).

Table 1

Risk of bias across studies

Additional analyses

10-11

RESULTS

Risk of bias within studies

10, 12-13

23

Results of individual
studies

20 For all outcomes considered (benefits or harms), present, for each study: (a) simple summary data for each
intervention group (b) effect estimates and confidence intervals, ideally with a forest plot.

Figure 2,
Figure 3,
Figure 4

Synthesis of results

21 Present results of each meta-analysis done, including confidence intervals and measures of consistency.

11-14

Risk of bias across studies

22 Present results of any assessment of risk of bias across studies (see Item 15).

Appendix
C

Additional analysis

23 Give results of additional analyses, if done (e.g., sensitivity or subgroup analyses, meta-regression [see Item 12-14
16]).

DISCUSSION
Summary of evidence
Limitations
Conclusions

24 Summarize the main findings including the strength of evidence for each main outcome; consider their
relevance to key groups (e.g., healthcare providers, users, and policy makers).
25 Discuss limitations at study and outcome level (e.g., risk of bias), and at review-level (e.g., incomplete
retrieval of identified research, reporting bias).
26 Provide a general interpretation of the results in the context of other evidence, and implications for future
research.

14-16

27 Describe sources of funding for the systematic review and other support (e.g., supply of data); role of
funders for the systematic review.

18

16
16-17

FUNDING
Funding

From: Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and MetaAnalyses: The PRISMA Statement. PLoS Med 6(6): e1000097. doi:10.1371/journal.pmed1000097

24

Appendix C. Funnel plots for meta-analysis on a) blood pressure, b) heart rate; and c), heart
rate variability.
a)

25

b)

26

c)

27

References
Accornero, N., Li Voti, P., La Riccia, M., Gregori, B., 2007. Visual evoked potentials modulation
during direct current cortical polarization. Exp. Brain. Res. 178, 261266.
Ahern, G.L., Sollers, J.J., Lane, R.D., Labiner, D.M., Herring, A.M., Weinand, M.E., .Hutzler, R.,
Thayer, J.F., 2001. Heart rate and heart rate variability changes in the intracarotid sodium
amobarbital test. Epilepsia 42, 912921.
*Aslaksen, P.M., Vasylenko, O., Fagerlund, A.J., 2014. The effect of transcranial direct current
stimulation on experimentally induced heat pain. Exp. Brain. Res. 232, 18651873.
Baena, C.P., Olandoski, M., Younge, J.O., Buitrago-Lopez, A., Darweesh, S.K., Campos, N.,
Sedaghat, S., Sajjad, A., van Herpt, T.T., Freak-Poli, R., van den Hooven, E., Felix, J.F., FariaNeto, J.R., Chowdhury, R., Franco, O.H., 2014. Effects of lifestyle-related interventions on
blood pressure in low and middle-income countries: systematic review and meta-analysis. J.
Hypertens. 32, 961973.
Barker, A.T., Jalinous, R., Freeston, I.L., 1985. Non-invasive magnetic stimulation of human motor
cortex. Lancet. 1, 11601167.
Becker, B.J., 2005. Failsafe N or file-drawer number, in Rothstein, H.R., Sutton, A.J., Borenstein,
M. (Eds.), Publication bias in meta-analysis: Prevention, assessment and adjustments. Wiley,
Hoboken, NY, pp. 111125.
Benarroch, E.E., 1993. The central autonomic network: functional organization, dysfunction, and
perspective. Mayo Clin. Proc. 68, 9881001.
*Berger, C., Domes, G., Balschat, J., Thome, J., Hppner, J. in press. Effects of prefrontal rTMS on
autonomic reactions to affective pictures. J. Neural. Transm. (Vienna).
Bromfield, S., Muntner, P., 2013. High blood pressure: the leading global burden of disease risk
factor and the need for worldwide prevention programs. Curr. Hypertens. Rep. 15, 134136.
*Brunoni, A.R., Kemp, A.H., Dantas, E.M., Goulart, A.C., Nunes, M.A., Boggio, P.S., Mill, J.G.,
Lotufo, P.A., Fregni, F., Benseor, I.M., 2013. Heart rate variability is a trait marker of major
28

depressive disorder: evidence from the sertraline vs. electric current therapy to treat depression
clinical study. Int. J. Neuropsychopharmacol. 16, 19371949.
Brunoni, A.R., Nitsche, M.A., Bolognini, N., Bikson, M., Wagner, T., Merabet, L., Edwards, D.J.,
Valero-Cabre, A., Rotenberg, A., Pascual-Leone, A., Ferrucci, R., Priori, A., Boggio, P.S.,
Fregni, F., 2012. Clinical research with transcranial direct current stimulation (tDCS):
challenges and future directions. Brain. Stimul. 5, 175195.
*Brunoni, A.R., Vanderhasselt, M.A., Boggio, P.S., Fregni, F., Dantas, E.M., Mill, J.G., Lotufo,
P.A., Benseor, I.M., 2013. Polarity- and valence-dependent effects of prefrontal transcranial
direct

current

stimulation

on

heart

rate

variability

and

salivary

cortisol.

Psychoneuroendocrinology 38, 5866.

Buchanan, T.W., Driscoll, D., Mowrer, S.M., Sollers, J.J. 3rd, Thayer, J.F., Kirschbaum, C., Tranel,
D., 2010. Medial prefrontal cortex damage affects physiological and psychological stress
responses differently in men and women. Psychoneuroendocrinology 35, 5666.
Buharin, V.E., Butler, A.J., Shinohara, M., 2014. Motor cortical disinhibition with baroreceptor
unloading induced by orthostatic stress. J. Neurophysiol. 111, 26562664.
*Cabrerizo, M., Cabrera, A., Perez, J.O., de la Rua, J., Rojas, N., Zhou, Q., Pinzon-Ardila, A.,
Gonzalez-Arias, S.M., Adjouadi, M., 2014. Induced effects of transcranial magnetic
stimulation on the autonomic nervous system and the cardiac rhythm. Scientific World Journal
349718.
*Casanova, M.F., Hensley, M.K., Sokhadze, E.M., El-Baz, A.S., Wang, Y., Li, X., Sears, L., 2014.
Effects of weekly low-frequency rTMS on autonomic measures in children with autism
spectrum disorder. Front. Hum. Neurosci. 8, 851.
*Clancy, J.A., Johnson, R., Raw, R., Deuchars, S.A., Deuchars, J., 2014. Anodal transcranial direct
current stimulation (tDCS) over the motor cortex increases sympathetic nerve activity. Brain.
Stimul. 7, 97104.

29

*Clarke, B.M., Upton, A.R., Kamath, M.V., Al-Harbi, T., Castellanos, C.M., 2006. Transcranial
magnetic stimulation for migraine: clinical effects. J. Headache Pain 7, 341346.
Cogiamanian, F., Brunoni, A.R., Boggio, P.S., Fregni, F., Ciocca, M., Priori, A., 2010. Noninvasive brain stimulation for the management of arterial hypertension. Med Hypotheses 74,
332336.
Cohen, J., 1988. Statistical power analysis for the behavioral sciences, second ed. Erlbaum,
Hillsdale, NJ.
Cooper, H., 2009. Research synthesis and meta-analysis: A step by step approach, fourth ed. Sage,
Thousand Oaks, CA.
Cuffee , Y., Ogedegbe, C., Williams, N.J., Ogedegbe, G., Schoenthaler. A., 2014. Psychosocial risk
factors for hypertension: an update of the literature. Curr. Hypertens. Rep. 16, 483.
Custodis, F., Reil, J.C., Laufs, U., Bhm, M., 2013. Heart rate: a global target for cardiovascular
disease and therapy along the cardiovascular disease continuum. J. Cardiol. 62, 183187.
*De Putter, L.M.S., Vanderhasselt, M.A., Baeken, C., De Raedt, R., Koster, E.H.W., 2015.
Combining tDCS and working memory training to down regulate state rumination: a singlesession double blind sham-controlled trial. Cognitive. Ther. Res. 6, 754765.
Reyes del Paso, G.A., Langewitz, W., Mulder, L.J., van Roon, A., Duschek, S., 2013. The utility of
low frequency heart rate variability as an index of sympathetic cardiac tone: a review with
emphasis on a reanalysis of previous studies. Psychophysiology. 50, 477487.
Filippi, M.M., Oliveri, M., Vernieri, F., Pasqualetti, P., Rossini, P.M., 2000. Are autonomic signals
influencing cortico-spinal motor excitability? A study with transcranial magnetic stimulation.
Brain. Res. 881, 159164.
*Flel, A., Rsser, N., Michka, O., Knecht, S., Breitenstein, C., 2008. Noninvasive brain
stimulation improves language learning. J. Cogn. Neurosci. 20, 14151422.

30

*Foerster, A., Schmitz, J.M., Nouri, S., Claus, D., 1997. Safety of rapid-rate transcranial magnetic
stimulation: heart rate and blood pressure changes. Electroencephalogr. Clin. Neurophysiol.
104, 207212.
Gianaros, P.J, Sheu, L.K., 2009. A review of neuroimaging studies of stressor-evoked blood
pressure reactivity: emerging evidence for a brain-body pathway to coronary heart disease risk.
Neuroimage. 47, 922936.
Gianaros, P.J., Onyewuenyi, I.C., Sheu, L.K., Christie, I.C., Critchley, H.D., 2012. Brain systems
for baroreflex suppression during stress in humans. Hum. Brain. Mapp. 33, 17001716.
Gianaros, P.J., Van Der Veen, F.M., Jennings, J.R., 2004. Regional cerebral blood flow correlates
with heart period and high-frequency heart period variability during working-memory tasks:
Implications for the cortical and subcortical regulation of cardiac autonomic activity.
Psychophysiology 41, 521530.
Giorli, E., Tognazzi, S., Briscese, L., Bocci, T., Mazzatenta, A., Priori, A., Orlandi, G., Del Sette,
M., Sartucci, F., 2015. Transcranial direct current stimulation and cerebral vasomotor reserve: a
study in healthy subjects. J. Neuroimaging. 25, 571574.
Giovannelli, F., Banfi, C., Borgheresi, A., Fiori, E., Innocenti, I., Rossi, S., Zaccara, G., Viggiano,
M.P., Cincotta, M., 2013. The effect of music on corticospinal excitability is related to the
perceived emotion: a transcranial magnetic stimulation study. Cortex. 49, 702710.
*Gulli, G., Tarperi, C., Cevese, A., Acler, M., Bongiovanni, G., Manganotti, P., 2013. Effects of
prefrontal repetitive transcranial magnetic stimulation on the autonomic regulation of
cardiovascular function. Exp. Brain. Res. 226, 265271.
*Hamner, J.W., Villamar, M.F., Fregni, F., Taylor, J.A., 2015. Transcranial direct current
stimulation (tDCS) and the cardiovascular responses to acute pain in humans. Clin.
Neurophysiol. 126, 10391046.
Higgins, J.P.T., Green, S., 2011. Cochrane handbook for systematic reviews of interventions,
version 5.1.0. Retrieved from www.cochrane- handbook.org
31

Isserles, M., Shalev, A.Y., Roth, Y., Peri, T., Kutz, I., Zlotnick, E., Zangen, A., 2013. Effectiveness
of deep transcranial magnetic stimulation combined with a brief exposure procedure in posttraumatic stress disorder--a pilot study. Brain. Stimul. 6, 377383.
*Jahanshahi, M., Ridding, M.C., Limousin, P., Profice, P., Fogel, W., Dressler, D., Fuller, R.,
Brown, R.G., Brown, P., Rothwell, J.C., 1997. Rapid rate transcranial magnetic stimulation--a
safety study. Electroencephalogr. Clin. Neurophysiol. 105, 422429.
Jarczok, M.N., Jarczok, M., Mauss, D., Koenig, J., Li, J., Herr, R.M., Thayer, J.F., 2013.
Autonomic nervous system activity and workplace stressors--a systematic review. Neurosci.
Biobehav. Rev. 37, 18101823.
*Jenkins, J., Shajahan, P.M., Lappin, J.M., Ebmeier, K.P., 2002. Right and left prefrontal
transcranial magnetic stimulation at 1 Hz does not affect mood in healthy volunteers. BMC
Psychiatry 2, 1.
Jennings, J.R., Zanstra, Y., 2009. Is the brain the essential in hypertension? Neuroimage. 47, 914
921.
Julius, S., Majahalme, S., 2000. The changing face of sympathetic overactivity in hypertension.
Ann. Med. 32, 365370.
Knotkova, H., Rosedale, M., Strauss, S.M., Horne, J., Soto, E., Cruciani, R.A., Malaspina, D.,
Malamud, D., 2012. Using transcranial direct current stimulation to treat depression in HIVinfected persons: The outcomes of a feasibility study. Front. Psychiatry. 3, 59.
*Kuppuswamy, A., Balasubramaniam, A.V., Maksimovic, R., Mathias, C.J., Gall, A., Craggs,
M.D., Ellaway, P.H., 2011. Action of 5 Hz repetitive transcranial magnetic stimulation on
sensory, motor and autonomic function in human spinal cord injury. Clin. Neurophysiol. 122,
24522461.
Lane, R.D., McRae, K., Reiman, E.M., Chen, K., Ahern, G.L., Thayer, J.F., 2009. Neural correlates
of heart rate variability during emotion. Neuroimage 44, 213222.

32

Lee, D.E., Cooper, R.S., 2009. Recommendations for global hypertension monitoring and
prevention. Curr. Hypertens Rep. 11, 444449.
Lipsey, M.W., Wilson, D.B., 2001. Practical meta-analysis. Thousand Oaks, CA, Sage.
Macefield, V.G., Taylor, J.L., Wallin, B.G., 1998. Inhibition of muscle sympathetic outflow
following transcranial cortical stimulation. J. Auton. Nerv. Syst. 68, 4957.
McEwen, B.S., Gianaros, P.J., 2010. Central role of the brain in stress and adaptation: links to
socioeconomic status, health, and disease. Ann. N. Y. Acad. Sci. 1186, 190222.
Moher, D., Liberati, A., Tetzlaff, J., Altman, D.G., 2009. Preferred reporting items for systematic
reviews and meta-analyses: The PRISMA statement. PLoS Medicine 6, e1000097.
Monte-Silva, K., Kuo, M.F., Liebetanz, D., Paulus, W., Nitsche, M.A., 2010. Shaping the optimal
repetition interval for cathodal transcranial direct current stimulation (tDCS). J. Neurophysiol.
103, 17351740.
*Montenegro, R.A., Farinatti, Pde T., Fontes, E.B., Soares, P.P., Cunha, F.A., Gurgel, J.L., Porto,
F., Cyrino, E.S., Okano, A.H., 2011. Transcranial direct current stimulation influences the
cardiac autonomic nervous control. Neurosci. Lett. 497, 3236.
Nakatani-Enomoto, S., Hanajima, R., Hamada, M., Mochizuki, H., Kobayashi, S., Enomoto, H.,
Sugiura, Y., Matsumoto, H., Furubayashi, T., Terao, Y., Sato, F., Ugawa, Y., 2011. Some
evidence supporting the safety of quadripulse stimulation (QPS). Brain. Stimul. 4, 303305.
Nsi, T., Mki, H., Kotilahti, K., Nissil, I., Haapalahti, P., Ilmoniemi, R.J., 2011. Magneticstimulation-related physiological artifacts in hemodynamic near-infrared spectroscopy signals.
PLoS One. 6, e24002.
Nugent, A.C., Bain, E.E., Thayer, J.F., Sollers, J.J., Drevets, W.C., 2011. Sex differences in the
neural correlates of autonomic arousal: a pilot PET study. Int. J. Psychophysiol. 80, 182191.
*Okano, A.H., Fontes, E.B., Montenegro, R.A., Farinatti, Pde T., Cyrino, E.S., Li, L.M., Bikson,
M., Noakes, T.D., 2015. Brain stimulation modulates the autonomic nervous system, rating of

33

perceived exertion and performance during maximal exercise. Br. J. Sports. Med. 18, 1213
1218.
Ottaviani, C., Thayer, J.F., Verkuil, B., Lonigro, A., Medea, B., Couyoumdjian, A. Brosschot, J.F.
(2016). Physiological concomitants of perseverative cognition: A systematic review and metaanalysis. Psychol. Bull. 142, 231259.
Pascual-Leone, A., Houser, C.M., Reese, K., Shotland, L.I., Grafman, J., Sato, S., Valls-Sol, J.,
Brasil-Neto, J.P., Wassermann, E.M., Cohen, L.G., M. Hallett, M., 1993. Safety of rapid-rate
transcranial magnetic stimulation in normal volunteers. Electroencephalogr. Clin.
Neurophysiol. 89, 120130.
*Pecuch, P.W., Evers, S., Folkerts, H.W., Michael, N., Arolt, V., 2000. The cerebral hemodynamics
of repetitive transcranial magnetic stimulation. Eur. Arch. Psychiatry. Clin. Neurosci. 250,
320324.
Priori, A., Berardelli, A., Rona, S., Accornero, N., Manfredi, M., 1998. Polarization of the human
motor cortex through the scalp. Neuroreport. 9, 22572260.
*Raimundo, R.J., Uribe, C.E., Brasil-Neto, J.P., 2012. Lack of clinically detectable acute changes
on autonomic or thermoregulatory functions in healthy subjects after transcranial direct current
stimulation (tDCS). Brain. Stimul. 5, 196200.
*Remue, J., Vanderhasselt, M.A., Baeken, C., Rossi, V., Tullo, J., Raedt, R., in press. The effect of
a single HF-rTMS session over the left dlPFC on the physiological stress response as measured
by heart rate variability. Neuropsychology.
Robertson, E.M., Thoret, H., Pascual-Leone, A., 2003. Studies in cognition: the problems solved
and created by transcranial magnetic stimulation. J. Cogn. Neurosci. 15, 948960.
Rossini, P.M., Rossini, L., Ferreri, F., 2010. Brain-behavior relations: transcranial magnetic
stimulation: a review. IEEE. Eng. Med. Biol. Mag. 29, 8495.

34

Santarnecchi, E., Feurra, M., Barneschi, F., Acampa, M., Bianco, G., Cioncoloni, D., Rossi, A.,
Rossi, S., 2014. Time Course of Corticospinal Excitability and Autonomic Function Interplay
during and Following Monopolar tDCS. Front. Psychiatry. 5, 86.
Schestatsky, P., Simis, M., Freeman, R., Pascual-Leone, A., Fregni, F., 2013. Non-invasive brain
stimulation and the autonomic nervous system. Clin. Neurophysiol. 124, 17161728.
Schubert, C., Lambertz, M., Nelesen, R.A., Bardwell, W., Choi, J.B., Dimsdale, J.E., 2009. Effects
of stress on heart rate complexity--a comparison between short-term and chronic stress. Biol.
Psychol. 80, 325332.
Sgoifo, A., Carnevali, L., Grippo, A.J., 2014. The socially stressed heart. Insights from studies in
rodents. Neurosci. Biobehav. Rev. 39, 5160.
*Sibon, I., Strafella, A.P., Gravel, P., Ko, J.H., Booij, L., Soucy, J.P., Leyton, M., Diksic, M.,
Benkelfat, C., 2007. Acute prefrontal cortex TMS in healthy volunteers: effects on brain 11CalphaMtrp trapping. Neuroimage. 34, 16581664.
Task Force of the European Society of Cardiology and the North American Society of Pacing and
Electrophysiology, 1996. Heart rate variability: Standards of measurement, physiological
interpretation and clinical use. Circulation. 93, 10431065.
Ter Horst, G.J., Postema, F., 1997. Forebrain parasympathetic control of heart activity: Retrograde
transneuronal viral labeling in rats. Am. J. Physiol. 273, H29262930.
Thayer, J.F., Ahs, F., Fredrikson, M., Sollers, J.J. 3rd, Wager, T.D., 2012. A meta-analysis of heart
rate variability and neuroimaging studies: implications for heart rate variability as a marker of
stress and health. Neurosci. Biobehav. Rev. 36, 747756.
Thayer, J.F., Lane, R.D., 2000. A model of neurovisceral integration in emotion regulation and
dysregulation. J. Affect. Disord. 61, 201216.
Thayer, J.F., Lane, R.D., 2009. Claude Bernard and the heart-brain connection: Further elaboration
of a model of neurovisceral integration. Neurosci. Biobehav. Rev. 33, 8188.

35

Thayer, J.F., Yamamoto, S.S., Brosschot, J.F., 2010. The relationship of autonomic imbalance,
heart rate variability and cardiovascular disease risk factors. Int. J. Cardiol. 141, 122131.
Tremblay, S., Lepage, J.F., Latulipe-Loiselle, A., Fregni, F., Pascual-Leone, A., Thoret, H., 2014.
The uncertain outcome of prefrontal tDCS. Brain Stimul. 7, 773783.
*Udupa, K., Sathyaprabha, T.N., Thirthalli, J., Kishore, K.R., Raju, T.R., Gangadhar, B.N., 2007.
Modulation of cardiac autonomic functions in patients with major depression treated with
repetitive transcranial magnetic stimulation. J. Affect. Disord. 104, 231236.
*Udupa, K., Thirthalli, J., Sathyaprabha, T.N., Kishore, K.R., Raju, T.R., Gangadhar, B.N., 2011.
Differential actions of antidepressant treatments on cardiac autonomic alterations in depression:
A prospective comparison. Asian. J. Psychiatr. 4, 100106.
*Van den Eynde, F., Claudino, A.M., Campbell, I., Horrell, L., Andiappan, M., Stahl, D., Schmidt,
U., 2011. Cardiac safety of repetitive transcranial magnetic stimulation in bulimic eating
disorders. Brain. Stimul. 4, 112114.
van Honk, J., Hermans, E.J., d'Alfonso, A.A., Schutter, D.J., van Doornen, L., de Haan, E.H., 2002.
A left-prefrontal lateralized, sympathetic mechanism directs attention towards social threat in
humans: evidence from repetitive transcranial magnetic stimulation. Neurosci. Lett. 319, 99
10.
van Honk, J., Schutter, D.J., Putman, P., de Haan, E.H., d'Alfonso, A.A., 2003. Reductions in
phenomenological, physiological and attentional indices of depressive mood after 2 Hz rTMS
over the right parietal cortex in healthy human subjects. Psychiatry. Res. 120, 95101.
Vandermeeren, Y., Jamart, J., Ossemann, M., 2010. Effect of tDCS with an extracephalic reference
electrode on cardio-respiratory and autonomic functions. BMC. Neurosci. 11, 38.
Vernieri, F., Altamura, C., Palazzo, P., Altavilla, R., Fabrizio, E., Fini, R., Melgari, J.M., Paolucci,
M., Pasqualetti, P., Maggio, P., 2014. 1-Hz repetitive transcranial magnetic stimulation
increases cerebral vasomotor reactivity: a possible autonomic nervous system modulation.
Brain. Stimul. 7, 281286.
36

*Vitor-Costa, M., Okuno, N.M., Bortolotti, H., Bertollo, M., Boggio, P.S., Fregni, F., Altimari,
L.R., 2015. Improving cycling performance: Transcranial direct current stimulation increases
time to exhaustion in cycling. PLoS One 10, e0144916.
Wells, G.A., Shea, B., OConnell, D., Peterson, J., Welch, V., Losos, M., Tugwell, P., 2011. The
Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in metaanalyses.

Ontario,

Canada:

Ottawa

Hospital

Research

Institute.

Retrieved

from

http://www.ohri.ca/ programs/clinical_epidemiology/oxford.asp
Wilson, P.W., D'Agostino, R.B., Levy, D., Belanger, A.M., Silbershatz, H., Kannel, W.B., 1998.
Prediction of coronary heart disease using risk factor categories. Circulation. 97, 18371847.
Winkelmann, T., Thayer, J.F., Pohlack, S., Nees, F., Grimm, O., Flor, H., in press. Structural brain
correlates of heart rate variability in a healthy young adult population. Brain Struct. Func.
*Yoshida, T., Yoshino, A., Kobayashi, Y., Inoue, M., Kamakura, K., Nomura, S., 2001. Effects of
slow repetitive transcranial magnetic stimulation on heart rate variability according to power
spectrum analysis. J. Neurol. Sci. 184, 7780.
*Yozbatiran, N., Alonso-Alonso, M., See, J., Demirtas-Tatlidede, A., Luu, D., Motiwala, R.R.,
Pascual-Leone, A., Cramer, S.C., 2009. Safety and behavioral effects of high-frequency
repetitive transcranial magnetic stimulation in stroke. Stroke. 40, 309312.
*Zwanzger, P., Eser, D., Vlkel, N., Baghai, T.C., Mller, H.J., Rupprecht, R., Padberg, F., 2007.
Effects of repetitive transcranial magnetic stimulation (rTMS) on panic attacks induced by
cholecystokinin-tetrapeptide (CCK-4). Int. J. Neuropsychopharmacol. 10, 285289.

37

Figure Captions
Fig. 1. Flow chart showing study selection for the meta-analysis.
Fig. 2. Forest plot for meta-analysis on NIBS effects on blood pressure
Fig. 3. Forest plot for meta-analysis on NIBS effects on heart rate.
Fig. 4. Forest plot for meta-analysis on NIBS effects on heart rate variability.

38

Table 1
Studies included in the meta-analysis and conditions/comparisons used to derive effect sizes.

Studies

W
Sti
o Ag m
Sh
Status
n
(% e typ
am
)
e

Aslakse
n et al.
(2014)

5
23. tD Healt Ye
51
1
1 CS
hy
s

Berger
et al.
(2015)

T
2 10 23.
M
0 0
3
S

M1

Healt Ye
dlPFC
hy
s

Brunoni
4 76. 43. tD
Ye
et al.
MDD
3
6 2 9 CS
s
(2013a)

Brunoni 2 85 24. tD

Stim
site

F3

Healt Ye dlPFC

Inten Sessio Go Dur

sity
ns
al
at

Task

2
mA

No

Pain
inductio
n (heat)

10
Hz

11

Ye
s

15

Emotio
nal
pictures

2
mA

12

Ye
s

30

No

1.5

Ye

33

Neutral

Contrast
Respons
e to
pain
(change
score
from
baseline
) in
sham vs
anodal
conditio
n
Respons
e to
high
negative
pictures
in the
HF
active
vs sham
conditio
ns
Pre to
posttreatme
nt
changes
in
complet
ers
assigne
d to
placebo
in sham
vs
anodal
conditio
n
Respons

Hedge
s g
(MA)

-0.12
(SBP);
-0.19
(DBP)

0.12
(HR2)

0.06
(lnHFHRV)

0.10
39

et al.
(2013b)

CS

hy

mA

Cabreriz
T
1
o et al.
30 253 M
0
(2014)
S

Healt
No
hy

Casanov
T
1 22. 13.
a et al.
M
8 2
1
(2014)
S

ASD No dlPFC

M1

Clancy
et al.
(2014)

1
34. tD Healt Ye
50
4
53 CS
hy
s

Clarke
et al.
(2006)

T
3 85. 41.
Migra
pain
M
No
3 7
4
ine
area5
S

M1

10
Hz

Ye
s

0.5
Hz

18

Ye
s

12

1
mA

Ye
s

15

Ye
s

Ye
s

25

No

20

De
Putter et 3 80. 23. tD Healt Ye
2
dlPFC
al.
8 33 13 CS
hy
s
mA
(2015)

Flel et
al.
(2008)

poster
ior
part
1 47. 25. tD Healt Ye
of the
9 4
6 CS
hy
s
perisy
lv.
area

1
mA

and
e to
(HFnegative negative HRV)
images stimuli
(minus
neutral)
in sham
vs
anodal
conditio
n
Baselin -0.25
e vs
(HR);
No
during
0.00
stimulat (HFion
HRV)
0.73
First vs (HR4);
No
last
0.77
session (HFHRV)
Change
score
from
No
baseline -0.22
(semi- in sham (HFsupine)
vs
HRV)
anodal
conditio
n
0.55
Pre vs
(HR);
post
No
1.16
stimulat
(HFion
HRV)
Sham vs
anodal
groups
0.24
(plus
Relax
(HFworking
HRV)
memory
training
)
Change
score
-0.00
Langua
from
(SBP);
ge
baseline
-0.00
learning in sham
(DBP);
paradig
vs
0.00
m
anodal
(HR)
conditio
n
40

Foerster
1
et al.
3
(1997)

T
27.
M
8
S

C3,
Healt Ye C4,
hy
s Fz, Pz

20
Hz

No

0.00
8

No

Sham vs 0.85
active (SBP);
conditio 0.85
(DBP)
ns
6

Gulli et
al.
(2013)

1 41.
35
2 7

T
M
S

Healt
No
hy

OFC

0.7
Hz

27

Ye
s

No

M1

2
mA

Ye
s

40

Pain
inductio
n (Cold
Pressor
Task)

M1

20
Hz

No

No

Healt
No dlPFC 1 Hz
hy

60-s
bloc
ks/1
No
5-s
brea
ks

Kuppus
T Spinal
sensor
wamy et 1
39.
Ye
20
M
cord
i5 Hz
al.
5
7
s
S injury
motor
(2011)

Ye
s

15

Montene
2
gro et al.
0
(2011)

Ye
s

20

Hamner
et al.
(2015)*

1 46. 25. tD Healt Ye

5 6 53 CS
hy
s

Jahansha
T
33. 33.
hi et al. 6
M
3
3
(1997)
S

Jenkins
et al.
(2002)

T
2
24.
50
M
0
6
S

303

Healt
No
hy

tD Healt Ye
CS
hy
s

T3

2
mA

No9

No
(sitting)

No
(sitting)
11

-0.24
(SBP);
Baselin
-0.07
e vs
(DBP);
recover
0.80
y post
(HR8);
right0.68
stimulat
(HRV:
ion
HF
n.u.)
Respons
e to
pain
1.58
(change (SBP);
score
0.40
from (DBP);
baseline 0.00
) in
(HR);
sham
-2.38
and
(HFanodal HRV)
conditio
n
0.21
Pre vs
(SBP);
post
0.05
stimulat
(DBP);
ion (2
0.06
blocks)
(HR8)
Effect
of left
site
0.79
stimulat (MAP)
ion
; 0.83
(before (HR8)
vs
after)10
Pre- to
0.25
post- (SBP);
treatme 0.20
nt in (DBP);
active
-0.39
vs sham (HR)
Main
0.33
effect of (HFstimulat HRV)1
41

Healt
tD
hy
Ye
33
CS (cycli s
sts)

Okano
et al.
(2015)

1
0

Pecuch
et al.
(2000)

2
55
0

Raimun
do et al.
(2012)

5
30. tD Healt Ye
64
0
5 CS
hy
s

Remue
et al.
(2015)

T
1 10 21.
M
9 0
8
S

32

T
M
S

2
mA

Ye
s

20

Healt Ye Fronta 10/2

hy
s l lobe 0 Hz

No

1
mA

Ye
s

20

20
Hz

Ye
s

10

T3

C3

Healt Ye dlPFC
7
hy
s

2
ion in
the
entire
group
HRV
during
Increme
the test 0.44
ntal
in sham (HRV3
maxima
vs
ms
l cyclist
anodal threshold)
test
conditio
n
0.23
Pre vs (DBP);
post
0.37
No
stimulat (SBP);
ion
-0.09
(HR8)
Change
score
0.11
from
(SBP);
baseline
No
0.05
in sham
(sitting)
(DBP);
vs
-0.16
anodal
(HR)
conditio
n
Respons
e to
Stress
stress
inductio inductio
n
0.72
n (T2)
(negativ in the (rMSS
e
D)
active
feedbac vs sham
k)
conditio
n

Sibon et
T
1
24.
al.
50
M
0
73
(2007)
S

Healt Ye
dlPFC
hy
s10

10
Hz

No 303

No

Udupa
et al
(2007)

MDD No

15
Hz

12

Ye
s

No

T
2 29. 32.
M
7 6
6
S

M1

12

Change
score
from
baseline
0.36
in sham
(HR)
vs
active
conditio
n
Pre vs
0.40
post
(rMSS
stimulat
D)
ion
42

Udupa
et al.
(2011)

T
3 55. 30.
M
0 2
5
S

Van den
Eynde et 3 86.
al.
7 8
(2011)

VitorCosta et
al.
(2015)

T
M
S

1
1

26

Yoshida
1
et al.
6
(2001)

T
293 M
S

scalp
5 cm
anteri
or to
the
MDD No hand
area
on
parasa
gittal
plane

Bulim
ia
Ye
dlPFC
nervo s
sa

15
Hz

Ye
s

10
Hz

No

M1

2
mA

Ye
s

Healt Ye
hy
s14

Cz

0.2
Hz

Ye
s

M1

20
Hz

No

Ye
s

tD Healt Ye
CS
hy
s

Yozbatir
1 16.
an et al.
67
2 6
(2009)

T
M
S

Strok
No
e

Zwanzg
er et al.
(2007)

T
M
S

Healt Ye
dlPFC 1 Hz
hy
s

1 54.
26
1 5

12

12

No

0.35
(SBP);
Pre vs
0.60
post (DBP);
stimulat 0.14
ion
(HR);
0.19
(HRV)

Respons
e after
20
rTMS
-0.28
trains (SBP);
(minus
-0.41
203
No
baseline (DBP);
) in
0.40
sham vs (HR)
active
conditio
n
Respons
e after
Constan
exhausti
t load
on in
-0.04
13
test
anodal (HR)
(cycling
vs sham
)
conditio
ns
Before
0.74
No
vs after
5.8
(HF(sitting) stimulat
HRV)
ion
-0.56
Pre vs
(SBP);
immedi
-0.28
20
No
ately
(DBP);
post
0.00
TMS
(HR8)
Maximi
Pharma m HR
cologica respons
lly
e after
0.08
30
induced sham vs (HR15)
panic
active
attack conditio
ns

43

Note. MA = Meta-analysis; SBP = Systolic Blood Pressure; DBP = Diastolic Blood Pressure; MAP
= Mean Arterial Pressure; HR = Heart Rate; HF-HRV = High Frequency-Heart rate variability;
rMSSD = Root Mean Square of the Successive Differences; MDD = Major Depression Disorder;
Wo = Women; Stim = Stimulation; Durat = Stimulation duration; tDCS = Transcranial direct
Current Stimulation; TMS = Transcranial Magnetic Stimulation; ASD = Autism Spectrum
Disorder; M1 = Primary Motor Cortex; dlPFC = Dorsolateral prefrontal cortex; OFC =
Orbitofrontal cortex;
*

Potential outlier; 1 The study included two sessions but we only considered the HF stimulation; 2

Phasic cardiac response; 3 Estimated value; 4 RR intervals; 5 N = 19 in temporal, n = 24 in occipital,

and n = 8 in frontal areas; 6 The study also included HR as an outcome but insufficient data were
provided for ES computation; 7 Stimulation on right and left side; 8 Pulse rate; 9 Before and after the
stimulation participants underwent a series of tasks (Digit Symbol Substitution and Verbal Fluency
Task); 10 The study also had a right/left site stimulation that yielded non-significant effects; 11
Controlled breathing; 12 The study found a significant effect on HF-HRV in athletes only; 13 Left
occipital stimulation used as sham; 14 Not included in the analyses due to missing data; 15 Maximum
HR peak.

44

Table 2
Quality assessment performed using the Newcastle-Ottawa Scale adapted.

Studies
Aslaksen et al. (2014)
Berger et al. (2015)
Brunoni et al. (2013a)
Brunoni et al. (2013b)
Cabrerizo et al. (2014)
Casanova et al. (2014)
Clancy et al. (2014)
Clarke et al. (2006)
De Putter et al. (2015)
Flel et al. (2008)
Foerster et al. (1997)
Gulli et al. (2013)
Hamner et al. (2015)*
Jahanshahi et al. (1997)
Jenkins et al. (2002)
Kuppuswamy et al. (2011)
Lai et al. (2010)
Montenegro et al. (2011)
Okano et al. (2015)
Pecuch et al. (2000)
Raimundo et al. (2012)
Remue et al. (2015)
Sibon et al. (2007)
Udupa et al. (2007)
Udupa et al. (2011)
Van den Eynde et al. (2011)
Vitor-Costa et al. (2015)
Yoshida et al. (2001)
Yozbatiran et al. (2009)
Zwanzger et al. (2007)

Quality Score Selection Confounds Outcome

(0-7)
(0-2)
(0-2)
(0-3)
5
6
7
5
3
5
6
5
4
4
2
3
4
3
4
4
2
4
3
4
5
4
4
6
6
4
5
4
4
4

2
2
2
1
0
1
2
2
1
1
0
0
1
0
1
1
0
1
0
1
2
1
1
2
2
1
2
1
1
1

1
2
2
1
0
1
1
0
1
1
1
0
1
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
1

2
2
3
3
3
3
3
3
2
2
1
3
2
3
2
2
2
2
2
2
2
2
2
3
3
2
2
2
3
2

Note: * = potential outliers.

45

Records iden+ed through database

(PubMed, Scopus), and manual
reference searching to
January 10 2016 (n = 730)
Medline (n = 188), Scopus (n = 542)
Duplicate records removed
(n = 264)
Unique records screened
(n = 466)
Records excluded (Review ar+cles;
Methodological; No cardiovascular
measures; Case study or study protocol)
( n = 421)
Full-text ar+cles assessed for eligibility
(n = 45)
Full-text ar+cles excluded (missing data)
(n = 16)
Studies included in the Meta-Analysis
(n = 29)