Brain Stimulation 13 (2020) 840e849

Contents lists available at ScienceDirect

Brain Stimulation
journal homepage: http://www.journals.elsevier.com/brain-stimulation

A randomized, double blind, sham-controlled trial of repetitive

transcranial magnetic stimulation (rTMS) in the treatment of negative
symptoms in schizophrenia
Nand Kumar d, *, Sreenivas Vishnubhatla a, Ashima Nehra Wadhawan b, Sujata Minhas c,
Prashant Gupta c
a
Department of Bio Statistics, AIIMS, Delhi, India
b
Cardio Neuro Centre, AIIMS, Delhi, India
c
Department of Psychiatry AIIMS, Delhi, India
d
All India Institute of Medical Sciences, Delhi, India

a r t i c l e i n f o a b s t r a c t

Article history: Background: Research has implicated hypofrontality in the pathogenesis of Negative symptoms of
Received 17 July 2019 schizophrenia.These symptoms are often resistant to treatment. Repetitive Transcranial Magnetic
Received in revised form Stimulation (rTMS) has been shown to reverse this hypofrontality. Higher frequency rTMS has shown
13 February 2020
better promise, but so far there has been very little research in this area.
Accepted 17 February 2020
Objective: We aimed to evaluate the efficacy of high-frequency (20 Hz) unilateral rTMS over the left
Available online 29 February 2020
Dorso-Lateral Pre-frontal Cortex (DLPFC) in the improvement of Negative symptoms in Schizophrenia.
Methods:100 patients of schizophrenia with predominantly negative symptoms, were enrolled for this
Keywords:
Schizophrenia
randomized, sham-controlled, double-blind trial.Each participant received 20 sessions of rTMS at 20 Hz
Negative symptoms frequency and 100% motor threshold, via either the active or the sham coil, over 4 weeks. A total of 2000
rTMS pulses were imparted in 10 trains per session. Negative symptoms were assessed with the SANS and
Repetitive transcranial magnetic PANSS. CDSS was used to rule out depressive symptoms. Assessments were carried out at baseline, post-
stimulation intervention, and 1-month, 2-months, 3-months and 4-months follow ups.
Results: The improvement in the negative symptoms (Anhedonia, Alogia, Avolition, Attention impair-
ment) in active group was statistically significant at 0.01 and 0.05 (p-value) as compared to sham group.
Conclusions: These results suggest that high-frequency rTMS may lead to improvement in negative
symptoms of schizophrenia. It may be worth considering its use as an adjunct to pharmacological
treatment of negative symptoms.
© 2020 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Introduction indicated that pre-frontal brain dysfunction, especially in the

dorso-lateral pre-frontal cortex (DLPFC) may be involved in the
Negative symptoms in schizophrenia are a core feature of the aetiology of negative symptoms of schizophrenia [3,4]. Neuro-
disease and contribute to a substantial burden associated with the imaging studies have shown hypometabolism and hypoperfusion
illness [1]. They include affective flattening or apathy, alogia, avo- of pre-frontal regions in patients suffering from negative symptoms
lition or amotivation, anhedonia, impaired attention and social [5]. Studies have suggested that this hypofrontality is more pro-
withdrawal [2]. Despite the substantial research on schizophrenia nounced on the left side [6e8].
over the years, the treatment for negative symptoms remains Repetitive transcranial magnetic stimulation (rTMS) is a non-
elusive. Unlike the positive symptoms, they are often resistant to invasive brain stimulation technique which has been shown to be
antipsychotic drugs and psychological treatments [1]. Research has useful in the treatment of several psychiatric and neuro-psychiatric
disorders, including depression, PTSD, migraine, etc. [9e11]. More
recently, some authors have used it for the treatment of negative
symptoms as well, with varying outcomes [12e15]. It has been
* Corresponding author. hypothesized that rTMS may lead to reversal of hypofrontality
E-mail address: nandkm2001@gmail.com (N. Kumar).

https://doi.org/10.1016/j.brs.2020.02.016
1935-861X/© 2020 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/
).
 N. Kumar et al. / Brain Stimulation 13 (2020) 840e849 841

which is seen in the brains of patients having negative symptoms Negative Syndrome Scale (PANSS), with willingness to give
[5,16]. Most of the studies till date have evaluated the effects of informed consent were only included in the study. We did not
10 Hz pulses, in shorter treatment protocols which deliver up to include patients who had received rTMS treatment in the past for a
20,000 pulses to the brain [15,17e20]. But, higher frequency similar condition. Patients with Co morbid ICD-10 Axis I diagnosis,
treatments over longer times and higher number of total magnetic or Axis II Personality Disorder or who qualify any other exclusion
pulses administered have tended to show better improvements in criteria common to every TMS protocol were excluded.
research [18,21,22]. rTMS protocol: We used the Magstim Rapid2 and Magstim
Literature on effects of rTMS on negative symptoms of schizo- Rapid2 plus machine for this study. For intervention group, rTMS
phrenia is composed mainly of the studies with very small sample Magstim’s 70 mm figure-of-8 Double Rapid2 Air Cooled Coil (P/N
sizes and have inadequate power to detect inter-group differences 3910-00 S/NO 728) was used, while for sham group, Magstim’s
[15,18,20]. Moreover, methodological rigour has been a problem in sham coil (P/N 3950-00 S/NO105) was used. The coils were iden-
a number of these studies. E.g., the number of randomized, double- tical in external appearance which ensured that the patients
blind, sham-controlled trials has been rather limited, the sham remained blind to the intervention.
rTMS controls in a number of these studies may themselves induce After clinical assessment and before starting the rTMS inter-
some voltage in the brain, and the different approach to their vention, motor threshold of all the patients was determined fol-
administration, as compared to the active treatment may affect the lowed by the localization of stimulation site. Left dorso-lateral
blinding. Further, the time durations for which the patients in these prefrontal cortex (DLPFC) was taken as the site for stimulation.
studies have been followed up is rather less, mostly up to one We defined the site of stimulation as a point, 5.5 cm anterior and
month. Hence, little is known about the sustainability of the effects then 0.5 cm lateral from the point where optimum stimulation of
of rTMS in negative symptoms. the right abductor pollicis brevis muscle was observed on stimu-
Hence, given the need for an effective treatment of negative lation of cortical representation of left cerebral cortex. Same pro-
symptoms of schizophrenia and the need to explore potential role cedure was followed for both groups of patients.
for rTMS in this using robust study designs, we aim to evaluate the After marking site of stimulation and eliciting motor threshold,
efficacy of 20 sessions of high-frequency (20 Hz) unilateral rTMS following stimulation parameters to left DLPFC were used. The
intervention over the left DLPFC in the improvement of negative study subjects were administered 20 sessions of high frequency
symptoms in schizophrenia and compare it to sham controls, in a rTMS per day (5 consecutive sessions per week for 4 weeks) at
randomized, double-blind trial. 20 Hz frequency, 100% motor threshold (MT), for 10 s with inter-
train time interval ranging from 90 s to 120 s (as calculated auto-
Methods matically by the TMS machine for in built safety measures)
depending on the protocol intensity, total of 2000 pulses imparted
Study Design:The study was conceptualized as a randomized in 10 trains. Both sham and active group patients received the same
sham-controlled double-blind trial. stimulation parameters. Each session lasted for approximately
The study was conducted at the All Indian Institute of Medical 30 min.
Sciences (AIIMS), New Delhi which has dedicated rTMS unit with a Assessment: The assessments were carried out by a trained
fully functioning rTMS lab since 2009. Along with this, we have psychologist appointed for the project. The following instruments
weekly running TMS clinic. were used in assessment:
Sample size calculation: Barr and colleagues (2012) evaluated Mini International Neuropsychiatric Interview (MINI) version
the efficacy of 20 Hz RTMS on negative symptoms of schizophrenia 5.0.0 [23] Positive and Negative Syndrome Scale (PANSS) [24,25].
against sham controls in a double-blind RCT [12]. This study yielded Scale for Assessing Negative Symptoms in schizophrenia (SANS)
a mean SANS total score of 31.67 ± 13.94 in the sham group post- [2,26].
intervention. Anticipating 8 units of difference in SANS total score Clinical Global Impressions scale-severity index (CGI-S) [27]
post-intervention between the groups with a combined SD of 13.97, Calgary Depression Scale for Schizophrenia (CDSS) [28,29].
alpha error of 5% and power of 80%, the estimated sample size was Procedure: Patients who were receiving treatment for schizo-
34 per group. Our study included a sample of 50 per group ac- phrenia and stable on their medications for at least 4 weeks, but
counting for about 30% losses in follow up. who continued to have substantial negative symptoms, were
Randomization Procedure:For Randomization of enrolled pa- referred by their respective clinicians to our research team for
tients, 100 sequentially numbered opaque envelopes (with either participation in the study. Those who fulfilled the inclusion criteria
‘sham’ or ‘active’ written on it) were prepared as per a computer- and agreed to provide a written informed consent were random-
generated random sequence by the Biostatistics Department of ized to either of the study arms. In case a patient lacked the capacity
AIIMS, New Delhi. Only two rTMS technicians had access to them. to consent, their legal guardians were invited to provide consent.
As per the random sequence they would allot a number to the After baseline assessment on the abovementioned instruments,
randomized patient and provide the intervention as mentioned on rTMS intervention was provided as per mentioned protocol. Same
the card inside. assessments were also carried out at the completion of the inter-
The subjects were blinded regarding the treatment being given, vention (completion of 20 sessions of rTMS) and monthly follow-
also the rater was blinded for the allocation, only the rTMS trained ups for four months subsequently.
Technician who administered rTMS were aware of the allocation Analysis: Data was analysed using Stata version 14.2 (StataCorp.
from the time of randomization to the time of data analysis. A 2015). Descriptive analysis was done for socio-demographic data
trained doctor, unrelated to the research project, was present at all (Table 1). Two-sample t-test with equal variances and Chi-square
times during the intervention to tackle any emergent conditions. test were used to compare the two groups at baseline. Signifi-
Patients who were right-handed, clinically diagnosed as having cance value was set at 0.05. We performed an Intention to Treat
schizophrenia as per ICD-10 criteria for at least one year and (ITT) analysis, i.e., all individuals, who were randomized to exper-
receiving psychiatric treatment at the psychiatry outpatient imental arms at the baseline, were included in the analysis irre-
department of AIIMS, New Delhi, been on stable doses of medicines spective of whether they continued with the study protocol or
(if receiving) for the last 4 weeks, but continued to have significant dropped out. Missing values were imputed as per last observation
negative symptoms, i.e., score
15 on N scale of Positive and carried forward. Generalized Estimating Equations (GEE) 30 were
842 N. Kumar et al. / Brain Stimulation 13 (2020) 840e849

Table 1
Baseline socio-demographic and clinical characteristics of the sample population.

Active rTMS arm (n ¼ 50) Sham rTMS arm (n ¼ 50) p

Age in years, (mean ± SD) 32.4 ± 9.20 30.8 ± 9.34 0.38

Sex, n(%) 0.84

Male 29 (58.0) 28 (56.0)
Female 21 (42.0) 22 (44.0)

Marital status, n(%) 0.67

Never married 23 (46.0) 28 (56.0)
Currently married 19 (38.0) 18 (36.0)
Others 8 (16.0) 4 (8.0)

Education, n(%) 0.98

Intermediate 12 (24.0) 14 (28.0)
Graduate 9 (18.0) 10 (20.0)
Matric 9 (18.0) 9 (18.0)
Middle 9 (18.0) 9 (18.0)
Others 11 (22.0) 8 (8.0)

Employment status, n(%) 0.56

Employed 6 (12.0) 8 (16.0)
Unemployed 44 (88.0) 42 (84.0)

Occupation, n(%) 0.97

Not known 27 (54.0) 26 (52.0)
Home maker 6 (12.0) 9 (18.0)
Student 6 (12.0) 5 (10.0)
Skilled/unskilled 9 (18.0) 8 (16.0)
Others 2 (4.0) 2 (4.0)

Religion, n(%) 0.53

Hindu 46 (92.0) 47 (94.0)
Others 4 (8.0) 3 (6.0)

Residential Area, n(%) 0.22

Delhi 18 (36.0) 24 (48.0)
Outside Delhi 32 (64.0) 26 (52.0)

Duration of illness, in months (mean ± SD) 125.7 ± 76.43 101.5 ± 84.67 0.13

Family History of psychotic illness in 1st degree relative,n(%) 0.71

Yes 17 (34.0) 15 (30.6)
No 33 (66.0) 34 (69.4)

PANSS (mean ± SD)

Total 65.66 ± 8.43 67.02 ± 10.07 0.49
Positive 9.68 ± 2.44 9.94 ± 2.70 0.62
Negative 24.34 ± 4.32 24.9 ± 5.01 0.55
General 31.64 ± 5.01 32.18 ± 5.66 0.61

SANS (mean ± SD)

Total 60.66 ± 11.75 61.46 ± 13.69 0.74
Affective flattening 15.76 ± 6.57 17.76 ± 6.01 0.12
Alogia 11.3 ± 3.26 11.06 ± 3.61 0.73
Anhedonia 16.62 ± 2.42 16.42 ± 2.80 0.70
Attention impairment 6.7 ± 3.23 5.76 ± 3.44 0.16
Avolition 10.22 ± 2.82 10.66 ± 2.92 0.45
CGI-S (mean ± SD) 4.62 ± 0.70 4.52 ± 0.74 0.49

Antipsychotic drugs, n(%) 0.33

Risperidone 15 (30.0) 18 (36.0)

Olanzapine 8 (16.0) 12 (24.0)
Others (Amisulpiride, Clozapine, Quetiapine, Aripiprazole, No antipsychotic) 14 (28.0) 7 (14.0)
Combinations of two or more antipsychotic drugs 13 (26.0) 13 (26.0)

Anticholinergic drugs (Trihexyphenidyl),n(%) 0.42

Yes 23 (46.0) 27 (54.0)
No 27 (54.0) 23 (46.0)

Sedative/Hypnotic drugs (Clonazepam, Lorazepam), n(%) 0.37

Yes 5 (10.0) 8 (16.0)
No 45 (90.0) 42 (84.0)

Antidepressant drugs (Fluoxetine, Escitalopram, Sertraline), n(%) 0.13

Yes 13 (26.0) 7 (14.0)
No 37 (74.0) 43 (86.0)
 N. Kumar et al. / Brain Stimulation 13 (2020) 840e849 843

used to compare the outcome measures between the two groups treatment arm (n ¼ 50) (Fig. 1). An attrition rate of 8% (n ¼ 4) in the
over time from pre-intervention to 4-months follow-up active rTMS arm and 6% (n ¼ 3) in the sham treatment arm was
assessment. seen during the intervention. This further increased to 32% in active
Ethics: An ethical approval of the study was taken from the and 38% in sham groups. The number of dropouts in the two groups
Institute Ethics Committee of AIIMS, New Delhi. The trial protocol were statistically similar (c2 ¼ 0.396, p ¼ .53).
was registered with the Clinical Trials Registry- India (CTRI)
athttp://ctri.nic.in/Clinicaltrials/login.php with the number CTRI/ Demographic and clinical characteristics
2019/05/019099.
Table 1 shows the baseline demographic and clinical charac-
Results teristics of the participants. No significant differences were found
between the active and the sham treatment groups in demographic
A total of 497 patients were screened for inclusion, and 100 of characteristics at the baseline. Also, clinical variables (illness
them were randomized to active treatment arm (n ¼ 50) or sham duration, family history, medication use and scores on PANSS& its

Fig. 1. CONSORT flow diagram depicting the flow of participants through the study.
844 N. Kumar et al. / Brain Stimulation 13 (2020) 840e849

subscales, SANS& its subscales, CGI-S and CDSS) were statistically SANS total and domain scores seemed to be maintained (Fig. 2). The
similar between the two groups at baseline. change in scores from post intervention to 4-month follow up was
not statistically different between the two groups, except in
rTMS safety and tolerability anhedonia domain. Statistically significant difference in the scores
on anhedonia, alogia, avolition, attention impairment was seen at
One patient in the active treatment group developed a partial the end of study (Pre-interventionto 4 months follow-up).
seizure during treatment in the 5th session after which the inter- The results from PANSS negative scale painted a similar picture.
vention was discontinued. Another patient in the active arm had The scores were statistically similar at the baseline, but a significant
aggravation of positive psychotic symptoms, and her intervention difference was seen between the groups after the intervention,
had to be stopped after 6 sessions. Other than these two events, with active rTMS group having lower mean score (Table 3). The
both active intervention and sham were well tolerated. drop in scores was significantly more in the active rTMS group after
the intervention. This difference was maintained throughout the
follow-up duration, whereby the change in scores from post-
Outcome measures
intervention to the end of 4-months follow-up was not signifi-
cantly different between the groups.
Negative symptoms of schizophrenia: The total SANS score
The PANSS positive scale did not show any statistically signifi-
reduced significantly after the intervention in both the active
cant difference between active and sham groups either at the
(60.6 ± 11.75 to 43.9 ± 12.67, p < .01) as well as sham (61.5 ± 13.69
baseline, after the intervention or through the follow-up period
to 50.5 ± 14.11, p < .01) rTMS arms on paired sample t-test. Though
(Table 3& Fig. 3). Though the total PANSS scores and the general
the baseline total SANS scores were statistically similar in both the
psychopathology scale scores or the change in these scores did not
groups, the post-intervention scores were significantly lesser
show difference between the two groups immediately after the
among the subjects who received active rTMS as compared to those
intervention, there was a significant difference between the groups
who received sham (Table 2). Among the SANS sub-domains, the
during the follow-ups.
difference in post-intervention scores between active and sham
Depressive symptoms: Since depression was already ruled out
groups was statistically significant in affective flattening and avo-
among the participants during screening, the CDSS scores were
lition domains, but not others. Statistically significant differences
negligible at baseline. The two groups of participants did not differ
were seen in the change in SANS domain scores (pre-intervention
in terms of depressive symptoms at any time points of measure-
vs post intervention) between active and sham groups in anhe-
ment (Table 4).
donia, alogia and attention impairment domains. During the four
months follow up of the study participants, the initial reduction in

Table 2
Comparison of SANS scores between the two study groups across the assessment time points.

Pre- Post- 1 month 2 months 3 months 4 months Reduction (95% CI)

Intervention intervention follow up follow up follow up follow up
Pre-intervention to Post-intervention to 4 Pre-intervention to 4
post-intervention months follow-up months follow-up

N
Sham 50 50 50 50 50 50
Active 50 50 50 50 50 50

SANS total
Sham 61.5 ± 13.69 50.5 ± 14.11 48.3 ± 16.33 45.7 ± 18.13 44.6 ± 17.81 43.7 ± 18.46 11.0 (8.25, 13.67) 6.8 (3.26, 10.26) 17.7 (13.40, 22.04)
Active 60.6 ± 11.75 43.9 ± 12.67 38.9 ± 13.73 36.6 ± 14.72 35.2 ± 14.14 33.8 ± 15.79 16.6 (13.86, 19.42) 10.2 (7.75, 12.65) 26.8 (22.79, 30.89)
p 0.73 0.02 <0.01 <0.01 <0.01 <0.01 <0.01 0.12 <0.01

SANS affective flattening

Sham 17.8 ± 6.01 12.7 ± 6.82 11.9 ± 6.54 11.3 ± 6.86 10.8 ± 6.68 10.6 ± 6.83 5.1 (3.10, 6.12) 2.1 (0.64, 3.64) 7.2 (5.52, 8.88)
Active 15.8 ± 6.57 10.3 ± 5.91 8.8 ± 5.93 7.9 ± 5.84 7.4 ± 5.36 7.2 ± 5.97 5.5 (4.31, 6.61) 3.1 (2.25, 3.99) 8.6 (7.06, 10.10)
p 0.12 0.05 0.01 <0.01 <0.01 <0.01 0.62 0.27 0.24

SANS anhedonia
Sham 16.4 ± 2.80 15.2 ± 3.62 14.6 ± 3.42 14.1 ± 3.94 14.0 ± 3.91 13.8 ± 4.24 1.2 (0.53, 1.87) 1.4 (0.74, 2.06) 2.6 (1.64, 3.56)
Active 16.6 ± 2.42 14.3 ± 3.27 12.9 ± 3.60 12.4 ± 3.93 12.1 ± 4.16 11.6 ± 4.63 2.3 (1.71, 2.97) 2.7 (1.75, 3.61) 5.0 (3.88, 6.16)
p 0.70 0.18 0.02 0.04 0.02 0.01 0.02 0.03 <0.01

SANS alogia
Sham 11.1 ± 3.61 8.8 ± 4.16 8.3 ± 4.51 7.7 ± 4.89 7.5 ± 5.13 7.4 ± 4.87 2.3 (1.51, 3.02) 1.4 (0.27, 2.53) 3.7 (2.37, 4.95)
Active 11.3 ± 3.26 7.2 ± 4.33 6.3 ± 4.38 6.0 ± 4.50 5.6 ± 4.33 5.3 ± 4.60 4.1 (3.05, 5.11) 1.9 (1.13, 2.71) 6.0 (4.85, 7.15)
p 0.73 0.07 0.03 0.08 0.06 0.03 <0.01 0.46 <0.01

SANS avolition
Sham 10.7 ± 2.92 8.9 ± 2.83 8.6 ± 3.20 8.1 ± 3.58 8.0 ± 3.29 7.9 ± 3.18 1.7 (1.15, 2.33) 1.0 (0.35, 1.73) 2.8 (1.98, 3.58)
Active 10.2 ± 2.82 7.9 ± 2.28 6.9 ± 2.45 6.5 ± 2.90 6.4 ± 2.84 6.2 ± 3.15 2.3 (1.65, 2.99) 1.7 (0.94, 2.50) 4.0 (3.09, 4.99)
p 0.44 0.05 <0.01 0.02 <0.01 <0.01 0.21 0.20 0.05

SANS attention impairment

Sham 5.8 ± 3.44 4.9 ± 3.19 4.9 ± 3.18 4.5 ± 3.27 4.3 ± 3.47 4.1 ± 3.46 0.9 (0.24, 1.56) 0.8 (0.18, 1.38) 1.7(0.97, 2.39)
Active 6.7 ± 3.23 4.3 ± 3.46 4.1 ± 3.10 3.8 ± 3.88 3.7 ± 2.96 3.6 ± 3.00 2.4 (1.74, 3.14) 0.7 (0.03, 1.33) 3.1(2.29, 3.95)
p 0.16 0.37 0.18 0.25 0.34 0.44 <0.01 0.82 0.01
 N. Kumar et al. / Brain Stimulation 13 (2020) 840e849 845

Fig. 2. Comparison of SANS scores at baseline, post-intervention and follow-ups between the two study groups.
*Average difference in scores between the active and the sham rTMS groups, over all time points (pre-intervention, post intervention i. e after 20 sessions of rTMS, subsequent
assessments for 4 consecutive months post intervention).

Clinical Global Impression (CGI): The mean CGI severity index completion of intervention, as exemplified by the significant dif-
(CGI-S) score was similar between the active and sham participants ference between the change in scores from pre-to post intervention
at the baseline, but a significantly higher reduction in this score was between the two groups, and no such difference from post-
seen among the active group participants after the intervention intervention to 4 months follow-up.
(Table 4). It appeared to follow the patterns of SANS total and PANSS
negative scale, whereby the change in scores seemed to be main-
Discussion
tained over time, i.e., significant (or tending to be significant) dif-
ference between the groups after the intervention and at each
We examined the effects of 20 sessions of 20 Hz rTMS over left
follow-up (Fig. 4). This difference seemed to primarily stem from
DLPFC in the treatment of negative symptoms in schizophrenia
the reduction in illness severity (CGI-S score) immediately on
among patients who were continuing treatment as usual and found
846 N. Kumar et al. / Brain Stimulation 13 (2020) 840e849

Table 3
Comparison of PANSS scores between the two study groups across the assessment time points.

Pre- Post- 1 month 2 months 3 months 4 months Reduction in scores (95% CI)
Intervention intervention follow up follow up follow up follow up
Pre-intervention to Post-intervention to 4 Pre-intervention to 4
post-intervention months follow-up months follow-up

N
Sham 50 50 50 50 50 50
Active 50 50 50 50 50 50

PANSS total
Sham 67.0 ± 10.97 57.5 ± 9.94 56.7 ± 9.97 55.5 ± 11.69 55.0 ± 12.13 55.2 ± 11.44 9.5 (7.26, 11.82) 2.3 (0.05, 4.47) 11.8 (8.95, 14.65)
Active 65.7 ± 8.43 54.1 ± 10.50 50.9 ± 10.10 50.4 ± 10.38 49.1 ± 10.39 48.4 ± 10.76 11.6 (9.47, 13.65) 5.7 (4.03, 7.45) 17.3 (14.65, 19.95)
P 0.48 0.10 <0.01 0.02 0.01 <0.01 0.20 0.02 <0.01

PANSS positive scale

Sham 9.9 ± 2.70 8.7 ± 2.12 8.6 ± 2.01 8.7 ± 2.26 8.7 ± 2.28 8.9 ± 2.41 1.2 (0.65, 1.83) 0.2 (0.72, 0.36) 1.1 (0.47, 1.65)
Active 9.7 ± 2.44 8.5 ± 1.85 8.3 ± 1.79 8.5 ± 2.31 8.3 ± 1.96 8.3 ± 2.09 1.2 (0.74, 1.61) 0.2 (0.2, 0.56) 1.4 (0.76, 1.97)
P 0.62 0.62 0.38 0.76 0.40 0.22 0.87 0.29 0.48

PANSS negative scale

Sham 24.9 ± 5.01 20.8 ± 4.89 20.1 ± 5.44 19.4 ± 5.66 19.3 ± 6.09 19.3 ± 5.92 4.1 (3.09, 5.07) 1.5 (0.31, 2.77) 5.6 (4.07, 7.18)
Active 24.3 ± 4.32 18.5 ± 5.40 17.0 ± 5.23 16.8 ± 5.18 16.2 ± 4.89 15.7 ± 4.93 5.9 (4.92, 6.80) 2.7 (1.86, 3.62) 8.6 (7.40, 9.80)
P 0.55 0.03 <0.01 0.02 <0.01 <0.01 0.01 0.12 <0.01

PANSS general scale

Sham 32.2 ± 5.65 27.9 ± 5.23 27.9 ± 4.79 27.4 ± 5.67 27.1 ± 5.66 27.1 ± 5.20 4.2 (3.10, 5.34) 0.9 (0.06, 1.85) 5.1 (3.75, 6.49)
Active 31.6 ± 5.01 27.1 ± 5.46 25.6 ± 5.15 25.2 ± 5.31 24.6 ± 5.49 24.3 ± 5.57 4.5 (3.38, 5.66) 2.8 (1.91, 3.73) 7.3 (5.93, 8.75)
p 0.62 0.43 0.02 0.05 0.03 0.01 0.71 <0.01 0.03

favourable effects of rTMS intervention. We used a novel protocol in sucha situation, whereby the mere setting of rTMS application
which each participant received a total of 2000 pulses (in 10 trains) showed some improvement among the subjects. Nevertheless, a
in one session, at 100% motor threshold and inter-train interval clear statistically significant advantage in the improvement in
between 90 and 120 s. Significant differences were seen in our negative symptoms was seen among participants randomized to
study, in negative symptoms, as measured by PANSS negative active group as compared to the control group.
symptom scale and SANS, between the active and sham treatment Though our results yielded significant beneficial effects of rTMS
groups after the rTMS intervention. Ours is one of the largest in negative symptoms along with corresponding changes in CGI
published sample size studies, evaluating any rTMS interventions severity index, the overall change in negative symptoms was only
for negative symptoms of schizophrenia, available till date, and modest. The difference in SANS total score (sham ¼ 50.5 ± 14.11 vs
hence is adequately powered to detect small differences. active ¼ 43.9 ± 12.67) and PANSS negative symptom scale score
Previous studies have used rather small samples; maximum of (sham ¼ 20.82 ± 4.89 vs rTMS ¼ 18.48 ± 5.40) was approximately 7
23 participants in the 20 Hz active intervention group [22]. Further, points and 2 points, respectively, post-intervention. Hence, the
we could find only four RCTs (published in English language) which translation of the statistical improvement to clinical improvement
assessed the effect of 20 Hz pulses on negative symptoms needs to be done with caution, and potentially invites further
[12,21,22,31]. There are suggestions from the literature that rTMS research on this issue.
interventions with frequencies greater than 10 Hz might lead to The importance of the results of our study is re-iterated by the
better improvement in negative symptoms [18]. Our results provide fact that we followed our patients for a substantially long duration.
further validation to these findings. But, given the long duration of follow up, the attrition rate in our
Novak and colleagues (2006) administered 20 Hz, total 20,000 study was substantial. One-third (33/100) of the sample had
pulses unilaterally [31], while Barr and colleagues (2012) admin- dropped out towards the end. The primary reason for most of the
istered 20 Hz, 15,000 pulses to each hemisphere [12]. Both failed to drop outs was logistic difficulty in commuting to the centre.
find any significant improvement in negative symptoms. On the Transport or monetary assistance may have helped in reducing the
other hand, Zhao et al. (2014) and Rabany et al. (2014) administered drop outs. The number of drop-outs did not differ significantly
a total of 30,000 and 33,600 at 20 Hz pulses, respectively, and both between the two arms. In our review, we came across a single study
reported at least some statistically significant improvements in the which had a follow-up period comparable to ours. Dlabac-Delange
negative symptoms. We administered a total of 40,000 pulses and colleagues (2015) found sustained improvement in SANS
unilaterally (highest reported till date) to each patient over 20 scores over a period of 3 months with 10 Hz, total of 30000 pulses
sessions, and found significant reductions in negative symptoms. to each hemisphere [13]. Similarly, the favourable effects of rTMS in
This possibly indicates towards a potential critical number of pulses our study also seemed to be sustained over 4 months of follow-up
beyond which the rTMS treatment becomes effective and warrants after the intervention. It remains to be seen in future research that
further investigation. Similar increase in efficacy with number of how long such beneficial effects may be sustained and whether
pulses has been observed in studies on depression [32,33]. there is any role for booster sessions. An important advantage in
It should be noted that the participants who received sham our study was the use of sham coil. Majority of the studies use the
rTMS also had statistically significant improvements after the intervention coil as sham by tilting it to 90  [12,13,31]. This method
intervention. We used a specially designed sham coil which ensures has been found to induce at least some voltage in the brain
that no magnetic pulses are delivered to the patients who receive [34],whereas, a sham coil is devoid of this effect. Only Prikryl and
this intervention. One may speculate a possible placebo-effect in colleagues have used a sham coil in evaluation of efficacy for
 N. Kumar et al. / Brain Stimulation 13 (2020) 840e849 847

Fig. 3. Comparison of PANSS scores at baseline, post-intervention and follow-ups between the two study groups.
*Average difference in scores between the active and the sham rTMS groups, over all time points (pre-intervention, post intervention i. e after 20 sessions of rTMS, subsequent
assessments for 4 consecutive months post intervention).

negative symptoms previously [35]. Moreover, the sham coil in our Numbered Opaque Sealed Envelopes (SNOSE), which were carefully
study was identical to the active coil in appearance and the same handled.
stimulation parameters were used in the sham as the active coil. Depression is a common confounder for negative symptoms of
This helped in ensuring more stringent patient blinding. The schizophrenia due to their overlapping conditions. Moreover,
blinding was further strengthened by the use of Sequentially depression is known to respond to rTMS interventions over left
DLPFC. Hence, we made an efforts to control for this confounder by

Table 4
Comparison of CGI-S and CDSS scores between the two study groups across the assessment time points.

Pre- Post- 1 month 2 months 3 months 4 months Reduction$ (95% CI)

Intervention intervention follow up follow up follow up follow up
Pre-intervention to Post-intervention to 4 Pre-intervention to 4
post-intervention months follow-up months follow-up

N
Sham 50 50 50 50 50 50
Active 50 50 50 50 50 50

CGI-S
Sham 4.5 ± 0.73 4.2 ± 0.68 3.9 ± 0.65 3.9 ± 0.76 3.9 ± 0.77 3.8 ± 0.79 0.3 (0.17, 0.43) 0.4 (0.20, 0.57) 0.7 (0.44, 0.91)
Active p- 4.6 ± 0.69 3.9 ± 0.70 3.7 ± 0.84 3.6 ± 0.92 3.4 ± 0.86 3.4 ± 0.90 0.7 (0.52, 0.92) 0.5 (0.36, 0.68) 1.2 (0.97, 1.51)
value 0.49 0.02 0.07 0.08 <0.01 <0.01 <0.01 0.27 <0.01

CDSS
Sham 0.3 ± 1.05 0.12 ± 0.72 0.22 ± 0.91 0.12 ± 0.44 0.24 ± 0.98 0.26 ± 0.97 0.2 (0.01, 0.37) 0.1 (0.42, 0.14) 0.04 (0.3, 0.4)
Active p- 0.3 ± 0.74 0.12 ± 0.44 0.1 ± 0.36 0.02 ± 0.14 0.12 ± 0.59 0.02 ± 0.14 0.2 (0.00, 0.36) 0.1 (0.01, 0.22) 0.3 (0.1, 0.5)
value 1 1 0.39 0.13 0.46 0.09 1 0.13 0.24
848 N. Kumar et al. / Brain Stimulation 13 (2020) 840e849

rTMS for regular clinical treatment of Negative symptoms in

schizophrenia with adequate safety measures.

Declaration of competing interest

The authors declare that they have no conflict of Interest.

Acknowledgement

The study is fully funded by Department of science and Tech-

nology under CSRI Initiative (vide grant No DST No:SR/CSRI/48/
2013(G). I sincerely acknowledge the contribution of technical staff
of the Neuromodulation lab for mental health for their assistance in
completing this project.

References

[1] Buchanan RW. Persistent negative symptoms in schizophrenia: an overview.

Schizophr Bull 2007;33(4):1013e22. https://doi.org/10.1093/schbul/sbl057.
Fig. 4. Comparison of CGI-S scores at baseline, post-intervention and follow-ups be- [2] Andreasen NC. Negative symptoms in schizophrenia: definition and reliability.
tween the two study groups. Arch Gen Psychiatr 1982;39(7):784. https://doi.org/10.1001/
*Average difference in scores between the active and the sham rTMS groups, over all archpsyc.1982.04290070020005.
time points (pre-intervention, post intervention i. e after 20 sessions of rTMS, subse- [3] Wolkin A, Sanfilipo M, Wolf AP, Angrist B, Brodie JD, Rotrosen J. Negative
quent assessments for 4 consecutive months post intervention). symptoms and hypofrontality in chronic schizophrenia. Arch Gen Psychiatr
1992;49(12):959e65. http://www.ncbi.nlm.nih.gov/pubmed/1360200.
[Accessed 20 May 2019].
[4] Potkin SG, Alva G, Fleming K, et al. A PET study of the pathophysiology of
screening to exclude the subjects having depression. Moreover, the negative symptoms in schizophrenia. Am J Psychiatr 2002;159(2):227e37.
CDSS, which is a scale for measuring depressive symptoms specif- https://doi.org/10.1176/appi.ajp.159.2.227.
ically in people having schizophrenia, was used to measure the [5] Hill K, Mann L, Laws KR, Stephenson CME, Nimmo-Smith I, McKenna PJ.
Hypofrontality in schizophrenia: a meta-analysis of functional imaging
severity of depressive symptoms at baseline and to evaluate studies. Acta Psychiatr Scand 2004;110(4):243e56. https://doi.org/10.1111/
changes if any through follow-ups. Both the groups had statistically j.1600-0447.2004.00376.x.
similar CDSS scores at baseline, and no significant changes were [6] Molina Rodríguez V, MontzAndre
e R, Pe
rez Castejo

n MJ, et al. Cerebral
perfusion correlates of negative symptomatology and parkinsonism in a
seen at any time points during the study. sample of treatment-refractory schizophrenics: an exploratory 99 mTc-
A major limitation of our study was that we located the DLPFC HMPAO SPET study. Schizophr Res 1997;25(1):11e20. http://www.ncbi.nlm.
by manual measurements over the scalp. Even though the “5 cm nih.gov/pubmed/9176923. [Accessed 20 May 2019].
[7] Moreno-In ~ iguez M, Ortun
~ o F, Arbizu J, Milla

n M, Soutullo C, Cervera-Enguix S.
rule” has been a standard practice, studies have shown that neuro
Regional cerebral blood flow SPECT study, at rest and during Wisconsin Card
navigational techniques are more accurate in localizing the stim- Sorting Test (WCST) performance, in schizophrenia naive patients or treated
ulation site [36]. Also, our study participants continued their with atypical neuroleptics. ActasEsp Psiquiatr 2015;33(6):343e51. http://
www.ncbi.nlm.nih.gov/pubmed/16292718. [Accessed 20 May 2019].
respective usual pharmacological interventions during the study
[8] Suzuki M, Kurachi M, Kawasaki Y, Kiba K, Yamaguchi N. Left hypofrontality
period as prescribed by their clinicians. One may argue that this correlates with blunted affect in schizophrenia. Jpn J Psychiatry Neurol
might lead to some confounding of the outcomes. But, we did not 1992;46(3):653e7. http://www.ncbi.nlm.nih.gov/pubmed/1487849.
find any difference in the medications administered to the partic- [Accessed 20 May 2019].
[9] Stilling JM, Monchi O, Amoozegar F, Debert CT. Transcranial magnetic and
ipants of active or sham group at the baseline. Moreover, we direct current stimulation (TMS/tDCS) for the treatment of headache: a sys-
included only those patients who were receiving stable doses of tematic review. Headache J Head Face Pain 2019;59(3):339e57. https://
medicines for at least one month, without any recent dose or doi.org/10.1111/head.13479.
[10] Gaynes BN, Lloyd SW, Lux L, et al. Repetitive transcranial magnetic stimula-
medication changes. Hence, the chances of medicines being a tion for treatment-resistant depression. J Clin Psychiatr 2014;75:477e89.
confounder are low. Continuation of treatment as usual is in https://doi.org/10.4088/JCP.13r08815. 05.
keeping with the ethical guidelines as rTMS is not yet a proven [11] Yan T, Xie Q, Zheng Z, Zou K, Wang L. Different frequency repetitive trans-
cranial magnetic stimulation (rTMS) for posttraumatic stress disorder (PTSD):
exclusive treatment modality for schizophrenia. a systematic review and meta analysis. J Psychiatr Res 2017;89:125e35.
We did not take an upper limit cut-off for PANSS positive scale https://doi.org/10.1016/j.jpsychires.2017.02.021.
scores. This was done to make the study inclusions more repre- [12] Barr MS, Farzan F, Tran LC, Fitzgerald PB, Daskalakis ZJ. A randomized
controlled trial of sequentially bilateral prefrontal cortex repetitive trans-
sentative of the real-world conditions. Nevertheless, since our
cranial magnetic stimulation in the treatment of negative symptoms in
sample included only the patients who were already on stable schizophrenia. Brain Stimul 2012;5(3):337e46. https://doi.org/10.1016/
treatment, the incidence of positive symptoms in our study was j.brs.2011.06.003.
[13] Dlabac-de Lange JJ, Bais L, van Es FD, et al. Efficacy of bilateral repetitive
quite low. The maximum PANSS positive scale score in our study
transcranial magnetic stimulation for negative symptoms of schizophrenia:
sample was 16. results of a multicenter double-blind randomized controlled trial. Psychol
Med 2015;45(6):1263e75. https://doi.org/10.1017/S0033291714002360.
[14] Hansbauer M, Wobrock T, Kunze B, et al. Efficacy of high-frequency repetitive
Conclusion transcranial magnetic stimulation on PANSS factors in schizophrenia with
predominant negative symptoms e results from an exploratory re-analysis.
Our study showed benefits of high frequency (20 Hz) rTMS at Psychiatr Res 2018;263:22e9. https://doi.org/10.1016/j.psychres.2018.02.030.
[15] Wang J, Zhou Y, Gan H, et al. Efficacy towards negative symptoms and safety
100% motor threshold with 2000 pulses per session administered
of repetitive transcranial magnetic stimulation treatment for patients with
through 10 trains over left DLPFC for 20 sessions over 4 weeks for schizophrenia: a systematic review. Shanghai Arch Psychiatr 2017;29(2):
patients suffering from schizophrenia with predominant Negative 61e76. https://doi.org/10.11919/j.issn.1002-0829.217024.
[16] Lefaucheur J-P, Andre
-Obadia N, Antal A, et al. Evidence-based guidelines on
symptoms. This study adds to the existing literature with its robust
the therapeutic use of repetitive transcranial magnetic stimulation (rTMS).
design and large sample size despite few limitations and paves the Clin Neurophysiol 2014;125(11):2150e206. https://doi.org/10.1016/
path for future research regarding administration of high frequency j.clinph.2014.05.021.
 N. Kumar et al. / Brain Stimulation 13 (2020) 840e849 849

[17] Shi C, Yu X, Cheung EFC, Shum DHK, Chan RCK. Revisiting the therapeutic [26] Andreasen NC. The scale for the assessment of negative symptoms (SANS):
effect of rTMS on negative symptoms in schizophrenia: a meta-analysis. conceptual and theoretical foundations. Br J Psychiatry 1989;155(S7):49e52.
Psychiatr Res 2014;215(3):505e13. https://doi.org/10.1016/ https://doi.org/10.1192/S0007125000291496.
j.psychres.2013.12.019. [27] Guy W. ECDEU assessment manual for psychopharmacology. Rockville MD.
[18] Kennedy NI, Lee WH, Frangou S. Efficacy of non-invasive brain stimulation on https://ci.nii.ac.jp/naid/10027491559/. [Accessed 11 May 2019].
the symptom dimensions of schizophrenia: a meta-analysis of randomized [28] Addington D, Addington J, Maticka-Tyndale E. Assessing depression in
controlled trials. Eur Psychiatr 2018;49:69e77. https://doi.org/10.1016/ schizophrenia: the calgary depression scale. Br J Psychiatr Suppl 1993;(22):
j.eurpsy.2017.12.025. 39e44. http://www.ncbi.nlm.nih.gov/pubmed/8110442. [Accessed 15 May
[19] Aleman A, Enriquez-Geppert S, Knegtering H, Dlabac-de Lange JJ. Moderate 2019].
effects of noninvasive brain stimulation of the frontal cortex for improving [29] Lako IM, Bruggeman R, Knegtering H, et al. A systematic review of in-
negative symptoms in schizophrenia: meta-analysis of controlled trials. struments to measure depressive symptoms in patients with schizophrenia.
Neurosci Biobehav Rev 2018;89:111e8. https://doi.org/10.1016/ J Affect Disord 2012;140(1):38e47. https://doi.org/10.1016/j.jad.2011.10.014.
j.neubiorev.2018.02.009. [31] Novak T, Horacek J, Mohr P, et al. The double-blind sham-controlled study of
[20] He H, Lu J, Yang L, et al. Repetitive transcranial magnetic stimulation for high-frequency rTMS (20 Hz) for negative symptoms in schizophrenia:
treating the symptoms of schizophrenia: a PRISMA compliant meta-analysis. negative results. Neuro Endocrinol Lett. 27(1-2):209-213. http://
Clin Neurophysiol 2017;128(5):716e24. https://doi.org/10.1016/ www.ncbi.nlm.nih.gov/pubmed/16648775. Accessed May 12, 2019.
j.clinph.2017.02.007. [32] Gershon AA, Dannon PN, Grunhaus L. Transcranial magnetic stimulation in the
[21] Rabany L, Deutsch L, Levkovitz Y. Double-blind, randomized sham controlled treatment of depression. Am J Psychiatr 2003;160(5):835e45. https://doi.org/
study of deep TMS add-on treatment for negative symptoms and cognitive 10.1176/appi.ajp.160.5.835.
deficits in schizophrenia. J Psychopharmacol 2014;28(7):686e90. https:// [33] George MS, Post RM. Daily left prefrontal repetitive transcranial magnetic
doi.org/10.1177/0269881114533600. stimulation for acute treatment of medication-resistant depression. Am J
[22] Zhao S, Kong J, Li S, Tong Z, Yang C, Zhong H. Randomized controlled trial of Psychiatr 2011;168(4):356e64. https://doi.org/10.1176/
four protocols of repetitive transcranial magnetic stimulation for treating the appi.ajp.2010.10060864.
negative symptoms of schizophrenia. Shanghai Arch Psychiatr 2014;26(1): [34] Lisanby SH, Gutman D, Luber B, Schroeder C, Sackeim HA. Sham TMS: intra-
15e21. https://doi.org/10.3969/j.issn.10020829.2014.01.003. cerebral measurement of the induced electrical field and the induction of
[23] Sheehan DV, Lecrubier Y, Sheehan KH, et al. The Mini-International Neuro- motor-evoked potentials. Biol Psychiatr 2001;49(5):460e3. http://www.ncbi.
psychiatric Interview (M.I.N.I.): the development and validation of a struc- nlm.nih.gov/pubmed/11274658. [Accessed 12 May 2019].
tured diagnostic psychiatric interview for DSM-IV and ICD-10. J Clin Psychiatr [35] Prikryl R, Ustohal L, PrikrylovaKucerova H, et al. A detailed analysis of the
1998;59(Suppl 20):22e33. quiz 34-57, http://www.ncbi.nlm.nih.gov/ effect of repetitive transcranial magnetic stimulation on negative symptoms
pubmed/9881538. [Accessed 6 March 2019]. of schizophrenia: a double-blind trial. Schizophr Res 2013;149(1e3):167e73.
[24] Kay SR, Fiszbein A, Opler LA. The positive and negative syndrome scale https://doi.org/10.1016/j.schres.2013.06.015.
(PANSS) for schizophrenia. Schizophr Bull 1987;13(2):261e76. http://www. [36] Fitzgerald PB, Hoy K, McQueen S, et al. A randomized trial of rTMS targeted
ncbi.nlm.nih.gov/pubmed/3616518. [Accessed 11 May 2019]. with MRI based neuro-navigation in treatment-resistant depression. Neuro-
[25] Kay SR, Opler LA, Lindenmayer JP. Reliability and validity of the positive and psychopharmacology 2009;34(5):1255e62. https://doi.org/10.1038/
negative syndrome scale for schizophrenics. Psychiatr Res 1988;23(1): npp.2008.233.
99e110. http://www.ncbi.nlm.nih.gov/pubmed/3363019. [Accessed 11 May
2019].