Content uploaded by Jean-Philippe Miron
Author content
All content in this area was uploaded by Jean-Philippe Miron on Nov 17, 2021
Content may be subject to copyright.
Content uploaded by Jean-Philippe Miron
Author content
All content in this area was uploaded by Jean-Philippe Miron on Nov 16, 2021
Content may be subject to copyright.
Journal of Aective Disorders Reports 6 (2021) 100270
Available online 11 November 2021
2666-9153/© 2021 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
The relationship between pre-treatment heart rate variability and response
to low-frequency accelerated repetitive transcranial magnetic stimulation
in major depression
Jean-Philippe Miron
a
,
b
,
c
,
*
, Jack Sheen
c
,
d
, Tony Panzarella
e
, Molly Hyde
c
,
d
,
Farrokh Mansouri
c
,
d
, Linsay Fox
d
, Helena Voetterl
f
, V´
eronique Desbeaumes Jodoin
a
,
Paul Lesp´
erance
a
,
b
, Christophe Longpr´
e-Poirier
a
,
b
, Robert-Paul Juster
b
,
Zaris J. Daskalakis
c
,
g
,
h
,
i
, Daniel M. Blumberger
c
,
g
,
i
,
1
, Jonathan Downar
c
,
d
,
g
,
1
a
Centre Hospitalier de l’Universit´
e de Montr´
eal (CHUM) et Centre de Recherche du CHUM (CRCHUM), Universit´
e de Montr´
eal, QC, Canada
b
D´
epartement de Psychiatrie et d’Addictologie, Facult´
e de M´
edecine, Universit´
e de Montr´
eal, QC, Canada
c
Institute of Medical Science, Faculty of Medicine, University of Toronto, ON, Canada
d
Krembil Research Institute, University Health Network, Toronto, ON, Canada
e
Dalla Lana School of Public Health, University of Toronto, ON, Canada
f
Research Institute Brainclinics, Nijmegen and Department of Psychiatry, University of Amsterdam, the Netherlands
g
Department of Psychiatry, Faculty of Medicine, University of Toronto, ON, Canada
h
Department of Psychiatry, University of California San Diego, California, USA
i
Temerty Centre for Therapeutic Brain Intervention at the Centre for Addiction and Mental Health, Toronto, ON, Canada
ARTICLE INFO
Keywords:
Heart rate variability
HRV
MDD
Prediction
Accelerated
arTMS
ABSTRACT
Background: Major depressive disorder (MDD) is now the rst cause of disability worldwide. So far, no validated
and scalable biomarker has been identied to help with response prediction to antidepressant treatment. Cardiac
biomarkers such as heart rate variability (HRV) have been studied in MDD, but few studies have examined its
potential use for outcome prediction to repetitive transcranial magnetic stimulation (rTMS).
Objective: We recorded pre-treatment HRV in MDD participants prior to an rTMS course. We hypothesized that
higher pre-treatment HRV would be correlated with better clinical outcomes.
Methods: HRV was recorded as part of a single-arm, open-label rTMS feasibility study. Pre-treatment HRV was
assessed in N =30 MDD participants before they underwent a one-week (5 days, 8 daily sessions, 40 sessions
total) accelerated rTMS (arTMS) course using a low-frequency 1 Hz course (600 pulses per session, 50-minute
intersession interval) over the right dorsolateral prefrontal cortex at 120% of the resting motor threshold.
Clinical outcomes were captured using the Beck Depression Inventory-II (BDI-II). We tested for an association
between pre-treatment HRV and clinical outcomes on the BDI-II using a linear mixed effects model.
Results: Although average BDI-II score signicantly changed over time, these changes were not signicantly
associated with pre-treatment HRV (p =0.60). This nding remained when adjusting for age, sex, and HR,
individually and collectively.
Conclusion: The current study did not nd a relationship between pre-treatment HRV and response to low fre-
quency rTMS. Other approaches using cardiac biomarkers may have potential for response prediction.
Introduction
Major depressive disorder (MDD) is a debilitating illness with serious
socioeconomic repercussions (Lam et al., 2016). Even though effective
treatments are available, up to a third of patients do not respond to
rst-line approaches, such as antidepressant medication and psycho-
therapy (Kennedy et al., 2016). Alternative treatment options are
therefore needed.
Repetitive transcranial magnetic stimulation (rTMS) has been stud-
ied for over 25 years for MDD and is well-established as an effective
* Corresponding author.
1
Co-senior authors
Contents lists available at ScienceDirect
Journal of Affective Disorders Reports
journal homepage: www.sciencedirect.com/journal/journal-of-affective-disorders-reports
https://doi.org/10.1016/j.jadr.2021.100270
Received 24 May 2021; Received in revised form 25 October 2021; Accepted 5 November 2021
Journal of Aective Disorders Reports 6 (2021) 100270
2
intervention (Milev et al., 2016). With an advantageous side effect
prole over medication, response and remission rates have been esti-
mated to be as high as 40–60% and 25–35%, respectively, in recent
meta-analyses (Milev et al., 2016). Unfortunately, rTMS is burdened by
several limitations that decreases accessibility (Miron et al., 2021a,
2021b, 2020; 2019b, 2019a): equipment costs are high, technical
operation complex, and treatment courses lengthy, requiring daily ses-
sions over several weeks. This reinforces the need for careful patient
selection to maximize treatment outcomes and avoid futile treatment
courses.
Treatment response prediction through biomarkers, allowing more
objective and personalized care, has been dubbed “The Holy Grail of
Psychiatry” (Alhajji and Nemeroff, 2015; Nemeroff, 2015;bib442014).
MDD is increasingly recognized as a highly heterogeneous disorder
(GOLDBERG, 2011), with various clinical presentations requiring
different treatment approaches. Unfortunately, MDD subtyping using
clinical features has not yielded any benets regarding treatment
response prediction at the individual level. On the other hand, some
have favored a more agnostic approach, using machine-learning to
“‘biotype” and cluster MDD patients according to brain activity features
(Drysdale et al., 2016). Even though this approach is promising, it is
currently too complex and costly for routine clinical use. Biomarkers
that are both well-validated and practical for widespread use are
therefore needed.
Cardiac function abnormalities in MDD have been extensively
described, with a particular focus on decreased vagal control, which
could represent a failure of the prefrontal cortex (PFC) to inhibit sym-
pathetic activity (Sgoifo et al., 2015). At the neurobiological level,
subcortical hyperactivity is thought to be responsible for increased
sympathetic tone in response to stress (Wayne C Wayne C. Drevets et al.,
2008; Sgoifo et al., 2015). The PFC regulates sympathetic and para-
sympathetic activity by top-down mechanisms through connections
with the vagus nerve, via the subgenual anterior cingulate cortex
(sgACC) (Sgoifo et al., 2015). In MDD, these mechanisms could be
compromised, given the known dysfunctions of the PFC in the depressed
state. Since rTMS is theorized to normalize PFC activity (Wayne C.
Wayne C. Drevets et al., 2008; Koenigs and Grafman, 2009), its effects
could therefore be captured by cardiac activity recording (Wayne C
Wayne C Drevets et al., 2008; Koenigs and Grafman, 2009). Cardiac
activity has the advantage of being simple to monitor, cost-effective, and
made possible through a plethora of simple, robust and well-validated
devices.
One cardiac index in particular has been the subject of several
studies: heart rate variability (HRV). HRV is thought to be linked to
autonomic nervous system (ANS) balance. Indeed, lower HRV is asso-
ciated with increased sympathetic or decreased parasympathetic tone in
chronic stress or diseased state (Thayer and Lane, 2009). By contrast,
higher HRV is linked to better emotional regulation, social engagement,
and decrease in negative bias (Geisler et al., 2013). Decreased HRV has
been observed in MDD in several studies (Hartmann et al., 2019; Kemp
et al., 2012, 2010; Koch et al., 2019; Yeh et al., 2016). The presence of
melancholic features seems to increase this imbalance, suggesting that
the level of autonomic disturbance might be associated with the severity
of the depressive state (Kemp et al., 2014). HRV has also been used to
distinguish between unipolar and bipolar depression (Chang et al.,
2015).
The effects of rTMS on HRV has also been assessed. Preclinical
studies have conrmed a modulatory effect of rTMS on the ANS (Hong
et al., 2002). High-frequency (HF) rTMS over the left dorsolateral pre-
frontal cortex (DLPFC) in healthy controls has been shown to signi-
cantly increase HRV over sham (Choi and Jeon, 2020; Remue et al.,
2016; Vernieri et al., 2014; Yoshida et al., 2001). In another study,
treatment-naïve MDD participants were treated with either rTMS (N =
27) or antidepressants (N =25) (Udupa et al., 2007). The group treated
with antidepressants did not experience any HRV changes, whereas a
signicant increase was observed in the rTMS group by the end of the
treatment course. This was later replicated by the same group (Udupa
et al., 2011), but not observed in another, more recent rTMS study
(Tabitha A Tabitha A Iseger et al., 2019).
The use of HRV for response prediction to rTMS in MDD has not been
well studied. We therefore assessed the role of pre-treatment (baseline)
resting HRV in MDD participants who received an accelerated rTMS
(arTMS) course of 1 Hz stimulation delivered to the right DLPFC over a
5-day period (Miron et al., 2021a). We hypothesized that higher
pre-treatment HRV would be associated with greater improvement.
Methods
Participants and procedures
We conducted a prospective, single-arm, open-label feasibility study
in N =30 participants; complete methods are described elsewhere
(Miron et al., 2021a). In brief, adult outpatients age 18–85 were
included for study participation if they 1) had a Mini International
Neuropsychiatric Interview (MINI) conrmed MDD diagnosis (single or
recurrent episode) and 2) maintained a stable medication regimen from
4 weeks before treatment start to the end of the study. Exclusion criteria
were: 1) history of substance dependence or abuse within the last 3
months; 2) concomitant major unstable medical illness; 3) cardiac
pacemaker or implanted medication pump; 4) active suicidal intent; 5)
diagnosis of any personality disorder as assessed by a study investigator
to be primary and causing greater impairment than MDD; 6) diagnosis of
any psychotic disorder; 7) any signicant neurological disorder or insult
(including, but not limited to: any condition likely to be associated with
increased intracranial pressure, space occupying brain lesion, any his-
tory of seizure conrmed diagnostically by neurological assessment
[except those therapeutically induced by electroconvulsive therapy -
ECT], cerebral aneurysm, Parkinson’s disease, Huntington’s chorea,
dementia, stroke, neurologically conrmed diagnosis of traumatic brain
injury, or multiple sclerosis); 8) if participating in psychotherapy must
have been in stable treatment for at least 3 months prior to entry into the
study (with no anticipation of change in the frequency of therapeutic
sessions, or the therapeutic focus over the duration of the study); 9) any
clinically signicant laboratory abnormality in the opinion of the
investigator; 10) a dose of more than 2 mg daily lorazepam (or equiv-
alent) currently (or in the last 4 weeks) or any dose of an anticonvulsant
due to the potential to limit rTMS efcacy; 11) any non-correctable
clinically signicant sensory impairment and 12) any signicant car-
diovascular or metabolic disorder or insult including, but not limited to:
coronary artery disease, abnormal heart rhythms, heart failure, cardiac
valve disease, congenital heart disease, cardiomyopathy, vascular dis-
ease, dyslipidemia, diabetes, or hypertension. All participants provided
informed consent and this study was approved by the Research Ethics
Board of the University Health Network. rTMS was delivered with a
MagPro R20 stimulator equipped with a MC-B70 coil (MagVenture,
Farum, Denmark). Resting motor threshold (rMT) was determined ac-
cording to standard techniques (McClintock et al., 2017). Treatment
consisted of an arTMS course of 8 hourly sessions per day over 5
consecutive weekdays (Monday through Friday), totaling 40 sessions.
Each rTMS session consisted of low-frequency (LF) 1 Hz stimulation
delivered over a 10 min period (1 single train, 600 pulses per session,
50-minute intersession interval) at 120% of rMT over the right dorso-
lateral prefrontal cortex (R-DLPFC), localized according to a previously
published heuristic approximating the F4 EEG site (Mir-Moghtadaei
et al., 2017).
Clinical outcomes were measured using the self-rated Beck Depres-
sion Inventory-II (BDI-II) (BECK et al., 1961). Participants completed the
BDI-II at baseline on the week prior to rTMS initiation, daily on the
morning on each treatment day, at the end of the last day of treatment,
and 1 and 4 weeks after treatment (9 assessments for each participant).
J.-P. Miron et al.
Journal of Aective Disorders Reports 6 (2021) 100270
3
Data acquisition
Cardiac data acquisition was captured through electrocardiogram
(ECG) recordings. ECG acquisition was performed one week prior to
treatment through a single-lead attached to the left wrist. This lead was
part of a 16-lead EEG system (EasyCap, Brain Products, Munich, Ger-
many), as the acquisition was done during an EEG session for which the
results are reported elsewhere (Voetterl et al., 2021). The reference
electrode was at the FCz location and ground electrode at the AFz
location. Impedances of <5kΩ were achieved. Each ECG recording
consisted of a 5-minute resting-state condition, prior to a TMS-EEG
session. During the recording, participants sat comfortably in a chair,
eyes-closed, and were asked to relax. Patients were asked not to change
anything regarding their usual routine the morning of the recording.
Data processing
Data was sampled at 512 Hz. Any potential linear trend was
removed. A highpass butterworth lter was applied at 0.5 Hz, ltering
data twice, once forward and once backward. Five seconds of data were
removed at the beginning and the end of recordings to decrease signal
drift. All data were visually inspected to correctly detect R peaks in the
QRS complex. Using the times corresponding to the visually detected R-
peaks, HRV parameters were then calculated with the python package
pyHRV (Gomes et al., 2019). We used the root mean square of the suc-
cessive differences (RMSSD) as the representative HRV index. The
RMSSD is a time-domain index that reects vagal tone and is relatively
free of respiratory inuences, compared to other HRV indexes; it is also
highly correlated with high-frequency (HF) frequency-domain index,
which is most representative of the vagal tone in that index category
(Laborde et al., 2017). The a priori decision of limiting ourselves to only
one HRV index also decreased the number of statistical tests needed,
decreasing the odds of chance ndings.
Statistical analysis
Given the longitudinal nature of the data collection for BDI-II, we
tested for a statistically signicant association between pre-treatment
RMSSD and the BDI-II outcome using a linear mixed effects model,
where the correlation of BDI-II measurements within the same subject
are dealt with by modeling random effects for each subject (i.e.
intercepts and/or slopes). The main xed effects include pre-treatment
RMSSD, and time (dened as continuous). Although the study design
was balanced (i.e., the BDI-II measurements were taken at xed times for
each subject and there were no missing data), given 9 repeated measures
over the course of treatment, we chose to consider time as a continuous
variable to reduce the number of estimated parameters required. In
addition, given that the average BDI over time was curvilinear, we chose
to employ a polynomial trend over time instead of a linear trend.
Results
Baseline demographics and clinical characteristics are located in the
clinical outcomes paper (Miron et al., 2021a). Average pre-treatment
resting heart rate (HR) was 76.9 ±14.0 beats per minute (bpm) and
average RMSSD was 25.0 ±9.8 SD milliseconds (ms) (Fig. 1). As dis-
played in Fig. 2, average BDI-II score decreased from mean (SD) 35.2
(9.2) to 23.6 (10.5) over the 5 days of treatment, and then leveled off
during follow-up, with 24.0 (11.7) at 1 week and 23.5 (13.3) at 4 weeks.
In most cases, subjects showed a decreasing trajectory, while there was
little to no changes in others. BDI-II at the start of treatment varied
considerably across subjects, as did the trajectory of response over time.
In most cases subjects showed a decreasing trajectory, while in other
subjects there was no change or a slight increase over time (Fig. 3).
A t of a linear mixed model with random intercepts and slopes
found that although average BDI-II changed signicantly over time,
these changes were not signicantly associated with pre-treatment
RMSSD. This model assumes the random intercepts and slopes have an
unstructured variance-covariance matrix. Both the variability of in-
tercepts (Z =3.61; p =0.0002) and slopes (Z =2.88; p =0.002) across
subjects was statistically signicant. The between subjects variability
represented 70.7% of the total variability. Time was represented by a
polynomial of degree 2. Both terms were highly statistically signicant.
In other words, the average BDI-II changed signicantly over time. A
cubic term for time was also introduced in the model tting, but it did
not signicantly improve the model t. The effect of pre-treatment
RMSSD was not statistically signicantly associated with average
changes in BDI-II over time (p =0.60). This nding remained when
adjusting for age, sex, and HR, individually and collectively. We provide
a visual representation of the correlation between RMSSD and percent
improvement at each time point in Fig. 4 (highest R
2
value of 0.0533 at
Day 2).
Fig. 1. Box and whisker plot of pre-treatment HR and RMSSD, bpm =beats per minute, ms =milliseconds.
J.-P. Miron et al.
Journal of Aective Disorders Reports 6 (2021) 100270
4
Fig. 2. Average observed BDI-II scores over time.
Fig. 3. Linear regression ts of BDI-II scores over time for individual subjects.
J.-P. Miron et al.
Journal of Aective Disorders Reports 6 (2021) 100270
5
Lastly, the main effect of age on BDI-II improvement reached sta-
tistical signicance (p =0.026). Using a model with time and time
squared for every one-year increase over the average age of 41.3, the
average BDI-II dropped by 0.3. This suggests that older participants had
a greater response to rTMS.
Discussion
This is the rst study to explore the ability of HRV to predict response
to 1 Hz arTMS of the right DLPFC in MDD. Using a linear mixed effects
model, we did not nd any association between pre-treatment resting
RMSSD and clinical outcomes.
The HRV literature in MDD is complex. Even though initial studies
have consistently demonstrated that MDD patients have reduced HRV
compared to healthy controls (Tabitha A. Tabitha A Iseger et al., 2019),
other ndings have been more heterogeneous. Subsequent studies
focused primarily on assessing if HRV parameters could be used as
treatment response monitoring markers. Some have reported HRV in-
crease after treatment (Udupa et al., 2011, 2007) while others have not
(Brunoni et al., 2013; Tabitha A Tabitha A Iseger et al., 2019). Studies on
R² = 2E-05
-40%
-20%
0%
20%
40%
60%
0 102030405060
Day 1
R² = 0.0041
-20%
0%
20%
40%
60%
80%
100%
0 102030405060
Day 5
R² = 0.0533
-40%
-20%
0%
20%
40%
60%
80%
0 102030405060
Day 2
R² = 0.0064
-40%
-20%
0%
20%
40%
60%
80%
100%
0 102030405060
End of Last Day
R² = 0.0047
-60%
-40%
-20%
0%
20%
40%
60%
80%
0 102030405060
Day 3
R² = 0.0105
-20%
0%
20%
40%
60%
80%
100%
0 102030405060
1 week
R² = 0.021
-40%
-20%
0%
20%
40%
60%
80%
100%
0 102030405060
Day 4
R² = 9E-06
-40%
-20%
0%
20%
40%
60%
80%
100%
0 102030405060
4 weeks
Fig. 4. Correlation between RMSSD in milliseconds (ms) on the horizontal axis and percent improvement on the vertical axis at each time points.
J.-P. Miron et al.
Journal of Aective Disorders Reports 6 (2021) 100270
6
the use of HRV as a predictive biomarker have had similar mixed results.
In a yoga study in MDD participants (N =17) with partial remission to
antidepressants, Shapiro and colleagues found that remitters had higher
high-frequency (HF) and lower low-frequency (LF) HRV pre-treatment
values compared to non-remitters (David et al., 2007). Subsequently,
Jain and colleagues found a correlation between pre-treatment lower
relative power of very low frequency (VLF) and improvement after
treatment with escitalopram and yoga in two separate trials, although
the signicance was lost after adjusting for age and gender (Jain et al.,
2014). The small sample size and heterogeneity of results make those
studies difcult to interpret. These studies also did not report correction
for multiple comparisons despite the numerous statistical tests examined
for the HRV parameters studied. More recently, two small rTMS studies
did not nd any associations between pre-treatment HRV values and
clinical improvement (Tabitha A Tabitha A Iseger et al., 2019; Udupa
et al., 2007). Lastly and more importantly, a recent large trial (N =
1008) also did not nd an association between pre-treatment RMSSD
and antidepressant response in MDD patients, although they did report
that RMSSD could predict clinical outcomes as a function of anxious
versus nonanxious depression (Kircanski et al., 2018). In our study, we
did not have the ability to reliably measure anxiety using the BDI-II.
Our results thus add to the negative ndings of recent literature in
term of pre-treatment resting HRV as a predictor of clinical outcomes in
MDD. Nonetheless, this does not mean that other cardiac indexes could
not prove useful in this area. For example, the concept of Neuro-Cardiac-
Guided TMS (NCG-TMS) has recently garnered attention and has shown
that rTMS applied to sites near the DLPFC can inuence cardiac activity
(Iseger et al., 2021, 2017; Tabitha A. Tabitha A Iseger et al., 2019;
Tabitha A Tabitha A Iseger et al., 2019; Kaur et al., 2020). In their initial
pilot study, Iseger and colleagues reported a greater HR decrease during
DLPFC (F3/F4) 10 Hz rTMS stimulation in N =10 healthy subjects,
compared to other prefrontal (PFC) (FC3, FC4) and non-PFC (C3, C4, Pz)
areas (Iseger et al., 2017). So far, these ndings have been twice repli-
cated in N =20 (Kaur et al., 2020) and N =28 (Iseger et al., 2021)
healthy participants. A follow-up study done in N =15 MDD partici-
pants showed that HR signicantly decreased more during the rst
minute of active intermittent theta-burst stimulation (iTBS) compared to
sham stimulation (Tabitha A Tabitha A. Iseger et al., 2019). Importantly,
a trend towards an association between HR deceleration and treatment
response was also observed, but the study was not powered to detect
small effects. A larger trial will be needed before any conclusions can be
drawn, but this might suggest that cardiac reactivity to rTMS could be
useful in response prediction. This could represent a paradigm shift,
where efforts should now focus on assessing potential biological effects
during treatment or during a task, instead of at rest. A ‘cardiac hotspot’
for rTMS would satisfy the criteria of being simple, practical and suitable
for widespread clinical use, if it can demonstrate robust predictive value
for treatment outcome in MDD.
Lastly, the nding of a positive correlation between age and response
is not new. Even though initial rTMS studies reported decreased effec-
tiveness of rTMS for MDD in late-life depression (LLD) (Fregni et al.,
2006; Manes et al., 2001; Mosimann et al., 2004), recent ndings using
more intensive protocols suggest otherwise (Jodoin et al., 2019; Trevi-
zol et al., 2019). Given the burden of depression in older adults and the
need for novel interventions, future work should explore this potential
marker and treatment in older adults.
Limitations
This study had several limitations. Given its open-label nature, there
was no sham group allowing for assessment of overall efcacy. Our
modest sample size also could have put us at risk of a type-II error.
Indeed, there was no a priori consideration of statistical power, given the
exploratory nature of this study. We also used an EEG set-up that was not
designed per se for ECG recording. Still, our signal was of good quality
and we did not experience any data loss. We also did not make
participants undergo a clinical ECG to formally exclude any cardiac
abnormalities, even though patients with pre-existing severe cardiac
conditions were excluded. The use of RMSSD as the only representative
index of HRV could be seen as a limitation, given that other studies
reported several other measures. Still, we believe that limiting ourselves
to a single HRV parameter decreases the odds of chance nding. RMSSD
has been reported as a robust HRV parameter that can be captured with a
short recording (unlike the standard deviation of the IBI of normal sinus
beats - SDNN) and is considered the primary time-domain measure used
to estimate the vagally mediated changes reected in HRV (Shaffer
et al., 2014; Shaffer and Ginsberg, 2017). RMSSD is also highly corre-
lated with HF power, which also reect parasympathetic activity, while
being less inuenced by breathing (Shaffer and Ginsberg, 2017). Large
HRV studies in MDD have also used RMSSD as their main HRV param-
eter (Brunoni et al., 2013; Kircanski et al., 2018). Regarding clinical
outcomes, the BDI-II was not validated for daily use and therefore might
not have been sensitive enough to capture rapid clinical changes (Miron
et al., 2021a). In the future, it would be important to use a scale vali-
dated for daily use, such as the 6-item Hamilton Depression Rating Scale
(HAMD-6). Furthermore, the fact that our recording session preceded
the participants rst ever rTMS session could have increased anxiety
levels, and therefore may have had an effect on cardiac activity by
increasing sympathetic tone, although average HR remained in the
normal range. Additionally, we were not able to control for antide-
pressant use given our small sample size and the risk of overtting.
Antidepressants have been reported to decrease HRV, but not in all
studies. HRV decreases were mostly seen with older tricyclic antide-
pressants (TCA), while this was not consistently seen with SSRIs and
more recent antidepressants (Kemp et al., 2012, 2010). Finally, it is
important to note that the ndings of this study pertain specically to 1
Hz right DLPFC stimulation, and may not generalize to other common
protocols such as high-frequency/theta-burst left or bilateral
stimulation.
Conclusion
In conclusion, we did not nd support for the use of pre-treatment
HRV to help predict clinical outcomes to a 1 Hz arTMS course. Our
ndings add to the body of recent negative studies that have also failed
to demonstrate utility of pre-treatment HRV for response prediction in
MDD. However, there remains promise in the potential utility for HRV
with DLPFC rTMS interventions due to the connections with the un-
derlying neurocircuitry. As discussed, future studies should focus on
assessing HRV in response to stimulation or a task instead of at rest.
Another approach could also be to include HRV with other biomarkers in
order to create a composite index from multiples sources, such as what is
being done in the eld of allostatic load (Flier et al., 1998). Other cardiac
indices or approaches (e.g., NCG-rTMS) could also prove useful for
outcome prediction and monitoring. Given the high costs associated
with MDD and the need to prevent chronic depressive states, researchers
must identify reliable and scalable biomarkers so that clinicians can go
beyond the current trial-and-error approach to treating depression.
CONFLICTS of INTEREST
The authors declare no nancial interests relative to this work. JPM
reports research grants from the Brain & Behavior Research Foundation
NARSAD Young Investigator Award and salary support for his graduate
studies from the Branch Out Neurological Foundation. JS, TP, MH, FM,
LF, HV, VDJ, PL, CLP and RPJ do not report any conict of interest.
ZJD has received research and equipment in-kind support for an
investigator-initiated study through Brainsway Inc and Magventure Inc.
His-work was supported by the Canadian Institutes of Health Research
(CIHR), the National Institutes of Mental Health (NIMH) and the Tem-
erty Family and Grant Family and through the centre for Addiction and
Mental Health (CAMH) Foundation and the Campbell Institute. DMB
J.-P. Miron et al.
Journal of Aective Disorders Reports 6 (2021) 100270
7
receives research support from CIHR, NIH, Brain Canada and the Tem-
erty Family through the CAMH Foundation and the Campbell Family
Research Institute. He received research support and in-kind equipment
support for an investigator-initiated study from Brainsway Ltd. He is the
site principal investigator for three sponsor-initiated studies for
Brainsway Ltd. He also receives in-kind equipment support from Mag-
venture for investigator-initiated research. He received medication
supplies for an investigator-initiated trial from Indivior. JD reports
research grants from CIHR, the National Institute of Mental Health,
Brain Canada, the Canadian Biomarker Integration Network in Depres-
sion, the Ontario Brain Institute, the Weston Foundation, the Klarman
Family Foundation, the Arrell Family Foundation, and the Buchan
Family Foundation, travel stipends from Lundbeck and ANT Neuro, in-
kind equipment support for investigator-initiated trials from MagVen-
ture, and is an advisor for BrainCheck, TMS Neuro Solutions, and
Restorative Brain Clinics.
Contributions
JPM lead the project, designed the study, was responsible for data
collection and analyses and wrote the manuscript. TP conducted sta-
tistical analyses and reviewed the manuscript. JS, MH, FM, LF and HV
participated in the study design, data collection, analyses and reviewed
the manuscript. VDJ, PL, CLP, RPJ and ZJD reviewed the manuscript.
DMB and JD supervised the project and reviewed the manuscript.
Acknowledgement
JPM would like to thank the Brain & Behavior Research Foundation
and the Branch Out Neurological Foundation for their nancial support
of this project. We would also like to thank Terri Cairo, Julian Kwok,
Meaghan Todd, Nuno Ferreira, Thomas Russell and Eileen Lam for their
involvement and organizational support throughout this project.
References
Alhajji, L., Nemeroff, C.B., 2015. Personalized Medicine and Mood Disorders. Psychiatric
Clin. North Am. 38, 395–403. https://doi.org/10.1016/j.psc.2015.05.003.
BECK, A.T., Ward, C.H., Mendelson, M., Mock, J., ERBAUGH, J., 1961. An Inventory for
Measuring Depression. Arch. Gen. Psychiat. 4, 561–571. https://doi.org/10.1001/
archpsyc.1961.01710120031004.
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., Bense˜
nor, I.M., 2013. Heart rate variability is a trait
marker of major depressive disorder: evidence from the sertraline vs. electric current
therapy to treat depression clinical study. Int. J. Neuropsychoph. 16, 1937–1949.
https://doi.org/10.1017/s1461145713000497.
Chang, H.-.A., Chang, C.-.C., Kuo, T.B.J., Huang, S.-.Y., 2015. Distinguishing bipolar II
depression from unipolar major depressive disorder: differences in heart rate
variability. World J. Biol. Psychiatry 16, 351–360. https://doi.org/10.3109/
15622975.2015.1017606.
Choi, K.W., Jeon, H.J., 2020. Heart Rate Variability for the Prediction of Treatment
Response in Major Depressive Disorder. Front. Psychiatry 11, 607. https://doi.org/
10.3389/fpsyt.2020.00607.
David, S., A, C., Ian, M., D, Dmitry, Cristina, O., F, L., Andrew, Michelle, A., 2007. Yoga
as a Complementary Treatment of Depression: effects of Traits and Moods on
Treatment Outcome. Evid.-based Compl. Alt. 4, 493–502. https://doi.org/10.1093/
ecam/nel114.
Drevets, Wayne W.C., Price, J.L., Furey, M.L., 2008a. Brain structural and functional
abnormalities in mood disorders: implications for neurocircuitry models of
depression. Brain Struct. Funct. 213, 93–118. https://doi.org/10.1007/s00429-008-
0189-x.
Drevets, Wayne C, Savitz, J., Trimble, M., 2008b. The subgenual anterior cingulate
cortex in mood disorders. Cns Spectrums 13, 663–681. https://doi.org/10.1017/
s1092852900013754.
Drysdale, A.T., Grosenick, L., Downar, J., Dunlop, K., Mansouri, F., Meng, Y., Fetcho, R.
N., Zebley, B., Oathes, D.J., Etkin, A., Schatzberg, A.F., Sudheimer, K., Keller, J.,
Mayberg, H.S., Gunning, F.M., Alexopoulos, G.S., Fox, M.D., Pascual-Leone, A.,
Voss, H.U., Casey, B.J., Dubin, M.J., Liston, C., 2016. Resting-state connectivity
biomarkers dene neurophysiological subtypes of depression. Nat. Med. 23, 1–16.
https://doi.org/10.1038/nm.4246.
Flier, J.S., Underhill, L.H., McEwen, B.S., 1998. Protective and Damaging Effects of Stress
Mediators. New Engl. J. Med. 338, 171–179. https://doi.org/10.1056/
nejm199801153380307.
Fregni, F., Marcolin, M.A., Myczkowski, M., Amiaz, R., Hasey, G., Rumi, D.O., Rosa, M.,
Rigonatti, S.P., Camprodon, J., Walpoth, M., Heaslip, J., Grunhaus, L.,
Hausmann, A., Pascual-Leone, A., 2006. Predictors of antidepressant response in
clinical trials of transcranial magnetic stimulation. Int. J. Neuropsychoph. 9,
641–654. https://doi.org/10.1017/s1461145705006280.
Geisler, F.C.M., Kubiak, T., Siewert, K., Weber, H., 2013. Cardiac vagal tone is associated
with social engagement and self-regulation. Biol. Psychol. 93, 279–286. https://doi.
org/10.1016/j.biopsycho.2013.02.013.
GOLDBERG, D., 2011. The heterogeneity of “major depression. World Psychiatry 10,
226–228. https://doi.org/10.1002/j.2051-5545.2011.tb00061.x.
Gomes, P., Margaritoff, P., Silva, H.P.da, 2019. pyHRV: development and evaluation of
an open-source python toolbox for heart rate variability (HRV). In: Proc. Int’l Conf.
on Electrical, Electronic and Computing Engineering (IcETRAN), pp. 822–828.
Hartmann, R., Schmidt, F.M., Sander, C., Hegerl, U., 2019. Heart Rate Variability as
Indicator of Clinical State in Depression. Front. Psychiatry 9, 735. https://doi.org/
10.3389/fpsyt.2018.00735.
Hong, B., Kuwaki, T., Ju, K., Kumada, M., Akai, M., Ueno, S., 2002. Changes in blood
pressure and heart rate by repetitive transcranial magnetic stimulation in rats.
Neurosci. Lett. 329, 57–60. https://doi.org/10.1016/s0304-3940(02)00592-x.
Iseger, Tabitha A, Arns, M., Downar, J., Blumberger, D.M., Daskalakis, Z.J., Vila-
Rodriguez, F., 2019a. Cardiovascular differences between sham and active iTBS
related to treatment response in MDD. Brain Stimul. 13, 167–174. https://doi.org/
10.1016/j.brs.2019.09.016.
Iseger, Tabitha T.A., Bueren, N.E.R.vVan., Kenemans, J.L., Gevirtz, R., Arns, M., 2019b.
A frontal-vagal network theory for Major Depressive Disorder: implications for
optimizing neuromodulation techniques. Brain Stimul. 13, 1–9. https://doi.org/
10.1016/j.brs.2019.10.006.
Iseger, T.A., Padberg, F., Kenemans, J.L., Dijk, H.van, Arns, M., 2021. Neuro-Cardiac-
Guided TMS (NCG TMS): a replication and extension study. Biol. Psychol. 162,
108097 https://doi.org/10.1016/j.biopsycho.2021.108097.
Iseger, T.A., Padberg, F., Kenemans, J.L., Gevirtz, R., Arns, M., 2017. Neuro-Cardiac-
Guided TMS (NCG-TMS): probing DLPFC-sgACC-vagus nerve connectivity using
heart rate - First results. Brain Stimul. 10, 1006–1008. https://doi.org/10.1016/j.
brs.2017.05.002.
Jain, F.A., Cook, I.A., Leuchter, A.F., Hunter, A.M., Davydov, D.M., Ottaviani, C.,
Tartter, M., Crump, C., Shapiro, D., 2014. Heart rate variability and treatment
outcome in major depression: a pilot study. Int. J. Psychophysiol. 93, 204–210.
https://doi.org/10.1016/j.ijpsycho.2014.04.006.
Jodoin, V.D., Miron, J.-.P., Lesp´
erance, P., 2019. Safety and Efcacy of Accelerated
Repetitive Transcranial Magnetic Stimulation Protocol in Elderly Depressed Unipolar
and Bipolar Patients. Am. J. Geriatric Psychiatry Ofcial J. Am. Assoc. Geriatric
Psychiatry 27, 548–558. https://doi.org/10.1016/j.jagp.2018.10.019.
Kaur, M., Michael, J.A., Hoy, K.E., Fitzgibbon, B.M., Ross, M.S., Iseger, T.A., Arns, M.,
Hudaib, A.-.R., Fitzgerald, P.B., 2020. Investigating high- and low-frequency neuro-
cardiac-guided TMS for probing the frontal vagal pathway. Brain Stimul. https://doi.
org/10.1016/j.brs.2020.03.002.
Kemp, A.H., Quintana, D.S., Felmingham, K.L., Matthews, S., Jelinek, H.F., 2012.
Depression, comorbid anxiety disorders, and heart rate variability in physically
healthy, unmedicated patients: implications for cardiovascular risk. PLoS One 7,
e30777. https://doi.org/10.1371/journal.pone.0030777.
Kemp, A.H., Quintana, D.S., Gray, M.A., Felmingham, K.L., Brown, K., Gatt, J.M., 2010.
Impact of Depression and Antidepressant Treatment on Heart Rate Variability: a
Review and Meta-Analysis. Biol. Psychiat. 67, 1067–1074. https://doi.org/10.1016/
j.biopsych.2009.12.012.
Kemp, A.H., Quintana, D.S., Quinn, C.R., Hopkinson, P., Harris, A.W.F., 2014. Major
depressive disorder with melancholia displays robust alterations in resting state
heart rate and its variability: implications for future morbidity and mortality. Front.
Psychol. 5, 1387. https://doi.org/10.3389/fpsyg.2014.01387.
Kennedy, S.H., Lam, R.W., McIntyre, R.S., Tourjman, S.V., Bhat, V., Blier, P.,
Hasnain, M., Jollant, F., Levitt, A.J., MacQueen, G.M., McInerney, S.J., McIntosh, D.,
Milev, R.V., Müller, D.J., Parikh, S.V., Pearson, N.L., Ravindran, A.V., Uher, R.,
Group, the C.D.W., 2016. Canadian Network for Mood and Anxiety Treatments
(CANMAT) 2016 Clinical Guidelines for the Management of Adults with Major
Depressive Disorder. Can. J. Psychiatry 61, 540–560. https://doi.org/10.1177/
0706743716659417.
Kircanski, K., Williams, L.M., Gotlib, I.H., 2018. Heart rate variability as a biomarker of
anxious depression response to antidepressant medication. Depress. Anxiety 62,
187–189. https://doi.org/10.1002/da.22843.
Koch, C., Wilhelm, M., Salzmann, S., Rief, W., Euteneuer, F., 2019. A meta-analysis of
heart rate variability in major depression. Psychol. Med. 49, 1948–1957. https://doi.
org/10.1017/s0033291719001351.
Koenigs, M., Grafman, J., 2009. The functional neuroanatomy of depression: distinct
roles for ventromedial and dorsolateral prefrontal cortex. Behav. Brain Res. 201,
239–243. https://doi.org/10.1016/j.bbr.2009.03.004.
Laborde, S., Mosley, E., Thayer, J.F., 2017. Heart Rate Variability and Cardiac Vagal
Tone in Psychophysiological Research – Recommendations for Experiment Planning,
Data Analysis, and Data Reporting. Front. Psychol. 08 https://doi.org/10.3389/
fpsyg.2017.00213, 89–18.
Lam, R.W., McIntosh, D., Wang, J., Enns, M.W., Kolivakis, T., Michalak, E.E., Sareen, J.,
Song, W.-.Y., Kennedy, S.H., MacQueen, G.M., Milev, R.V., Parikh, S.V.,
Ravindran, A.V., Group, the C.D.W., 2016. Canadian Network for Mood and Anxiety
Treatments (CANMAT) 2016 Clinical Guidelines for the Management of Adults with
Major Depressive Disorder. Can. J. Psychiatry 61, 510–523. https://doi.org/
10.1177/0706743716659416.
Manes, F., Jorge, R., Morcuende, M., Yamada, T., Paradiso, S., Robinson, R.G., 2001.
A Controlled Study of Repetitive Transcranial Magnetic Stimulation as a Treatment
of Depression in the Elderly. Int. Psychogeriatr. 13, 225–231. https://doi.org/
10.1017/s1041610201007608.
J.-P. Miron et al.
Journal of Aective Disorders Reports 6 (2021) 100270
8
McClintock, S.M., Reti, I.M., Carpenter, L.L., McDonald, W.M., Dubin, M., Taylor, S.F.,
Cook, I.A., O’Reardon, J., Husain, M.M., Wall, C., Krystal, A.D., Sampson, S.M.,
Morales, O., Nelson, B.G., Latoussakis, V., George, M.S., Lisanby, S.H., 2017.
Consensus Recommendations for the Clinical Application of Repetitive Transcranial
Magnetic Stimulation (rTMS) in the Treatment of Depression. J. Clin. Psychiatry 79,
1–16. https://doi.org/10.4088/jcp.16cs10905.
Milev, R.V., Giacobbe, P., Kennedy, S.H., Blumberger, D.M., Daskalakis, Z.J., Downar, J.,
Modirrousta, M., Patry, S., Vila-Rodriguez, F., Lam, R.W., MacQueen, G.M.,
Parikh, S.V., Ravindran, A.V., Group, C.D.W., 2016. Canadian Network for Mood and
Anxiety Treatments (CANMAT) 2016 Clinical Guidelines for the Management of
Adults with Major Depressive Disorder. Can. J. Psychiatry 61, 561–575. https://doi.
org/10.1177/0706743716660033.
Mir-Moghtadaei, A., Dunlop, K., Mansouri, F., Blumberger, D.M., Daskalakis, Z.J.,
Downar, J., 2017. Scalp-based heuristics for locating the nodes of the salience
network for use in neurostimulation. Presented at the 2nd International Brain
Stimulation Conference, p. P479. https://doi.org/10.1016/j.brs.2017.01.404.
Miron, J.-.P., Feffer, K., Cash, R.F.H., Derakhshan, D., Kim, J.M.S., Fettes, P.,
Giacobbe, P., Blumberger, D.M., Daskalakis, Z.J., Downar, J., 2019a. Safety,
tolerability and effectiveness of a novel 20Hz rTMS protocol targeting dorsomedial
prefrontal cortex in major depression: an open-label case series. Brain Stimul. 12,
1319–1321. https://doi.org/10.1016/j.brs.2019.06.020.
Miron, J.-.P., Hyde, M., Fox, L., Sheen, J., Voetterl, H., Mansouri, F., Jodoin, V.D.,
Zhou, R., Dees, S., Mir-Moghtadaei, A., Blumberger, D.M., Daskalakis, Z.J., Vila-
Rodriguez, F., Downar, J., 2021a. Evaluation of a 5 day accelerated 1Hz repetitive
transcranial magnetic stimulation protocol in major depression: a feasibility study.
J. Affect. Disord. Rep. 4, 100077 https://doi.org/10.1016/j.jadr.2021.100077.
Miron, J.-.P., Sheen, J., Mansouri, F., Blumberger, D.M., Daskalakis, Z.J., Vila-
Rodriguez, F., Downar, J., 2020. The role of low-frequency repetitive transcranial
magnetic stimulation in major depression: a call to increase the evidence base. Brain
Stimul. 13, 1296–1297. https://doi.org/10.1016/j.brs.2020.06.013.
Miron, J.-.P., Voetterl, H., Fox, L., Hyde, M., Mansouri, F., Dees, S., Zhou, R., Sheen, J.,
Jodoin, V.D., Mir-Moghtadaei, A., Blumberger, D.M., Daskalakis, Z.J., Vila-
Rodriguez, F., Downar, J., 2021b. Optimized repetitive transcranial magnetic
stimulation techniques for the treatment of major depression: a proof of concept
study. Psychiat. Res. 298, 113790 https://doi.org/10.1016/j.psychres.2021.113790.
Miron, J.-.P., Voetterl, H., Mansouri, F., Blumberger, D.M., Daskalakis, Z.J., Downar, J.,
2019b. A case series of a novel 1Hz right-sided dorsolateral prefrontal cortex rTMS
protocol in major depression. Brain Stimul. 13, 372–374. https://doi.org/10.1016/j.
brs.2019.11.006.
Mosimann, U.P., Schmitt, W., Greenberg, B.D., Kosel, M., Müri, R.M., Berkhoff, M.,
Hess, C.W., Fisch, H.U., Schlaepfer, T.E., 2004. Repetitive transcranial magnetic
stimulation: a putative add-on treatment for major depression in elderly patients.
Psychiat. Res. 126, 123–133. https://doi.org/10.1016/j.psychres.2003.10.006.
Nemeroff, C.B., 2015. The Holy Grail of Psychiatry. Cerebrum Dana Forum Brain Sci
2015.
Nemeroff, C.B., 2014. PSYCHOPHARMACOLOGY AND THE FUTURE OF
PERSONALIZED TREATMENT: commentary: psychopharmacology and the Future of
Personalized Treatment. Depress. Anxiety 31, 906–908. https://doi.org/10.1002/
da.22303.
Remue, J., Vanderhasselt, M.-.A., Baeken, C., Rossi, V., Tullo, J., Raedt, R.D., 2016. The
effect of a single HF-rTMS session over the left DLPFC on the physiological stress
response as measured by heart rate variability. Neuropsychology 30, 756–766.
https://doi.org/10.1037/neu0000255.
Sgoifo, A., Carnevali, L., Alfonso, M.de los D.L.A.P., Amore, M., 2015. Autonomic
dysfunction and heart rate variability in depression. Ann. NY Acad. Sci. 18, 343–352.
https://doi.org/10.3109/10253890.2015.1045868.
Shaffer, F., Ginsberg, J.P., 2017. An Overview of Heart Rate Variability Metrics and
Norms. Front. Public Heal 5, 258. https://doi.org/10.3389/fpubh.2017.00258.
Shaffer, F., McCraty, R., Zerr, C.L., 2014. A healthy heart is not a metronome: an
integrative review of the heart’s anatomy and heart rate variability. Front. Psychol.
5, 1040. https://doi.org/10.3389/fpsyg.2014.01040.
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,
81–88. https://doi.org/10.1016/j.neubiorev.2008.08.004.
Trevizol, A.P., Goldberger, K.W., Mulsant, B.H., Rajji, T.K., Downar, J., Daskalakis, Z.J.,
Blumberger, D.M., 2019. Unilateral and bilateral repetitive transcranial magnetic
stimulation for treatment-resistant late-life depression. Int. J. Geriatr. Psych. 34,
822–827. https://doi.org/10.1002/gps.5091.
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. Disorders 104,
231–236. https://doi.org/10.1016/j.jad.2007.04.002.
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. Psychiatry 4, 100–106.
https://doi.org/10.1016/j.ajp.2011.02.006.
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, 281–286. https://doi.org/
10.1016/j.brs.2013.12.014.
Voetterl, H.T.S., Miron, J.-.P., Mansouri, F., Fox, L., Hyde, M., Blumberger, D.M.,
Daskalakis, Z.J., Vila-Rodriguez, F., Sack, A.T., Downar, J., 2021. INVESTIGATING
EEG BIOMARKERS OF CLINICAL RESPONSE TO LOW FREQUENCY RTMS IN
DEPRESSION. J. Affect. Disord. Reports 100250. https://doi.org/10.1016/j.
jadr.2021.100250.
Yeh, T.-.C., Kao, L.-.C., Tzeng, N.-.S., Kuo, T.B.J., Huang, S.-.Y., Chang, C.-.C., Chang, .
H..-A., 2016. Heart rate variability in major depressive disorder and after
antidepressant treatment with agomelatine and paroxetine: ndings from the
Taiwan Study of Depression and Anxiety (TAISDA). Prog. Neuro-Psychopharmacol.
Biol. Psychiatry 64, 60–67. https://doi.org/10.1016/j.pnpbp.2015.07.007.
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, 77–80. https://doi.org/
10.1016/s0022-510x(00)00505-0.
J.-P. Miron et al.
