Effects of Experimental Intermittent Theta Burst Stimulation on Negative Symptoms of Schizophrenia: A Double-Blind Sham-Controlled Study

Schizophrenia is a severe and chronic psychiatric disorder characterized by a heterogeneous constellation of symptoms that can be broadly classified into positive, negative, and cognitive domains.¹ While positive symptoms such as hallucinations and delusions often respond to antipsychotic medications, negative symptoms—including affective flattening, alogia, avolition, anhedonia, and social withdrawal—are less amenable to current pharmacotherapy and contribute substantially to long-term disability and poor functional outcomes.^2–4

Conventional antipsychotic medications have limited efficacy for negative symptoms, underscoring the imperative for novel therapeutic interventions.^3,5 Noninvasive brain stimulation techniques, particularly repetitive transcranial magnetic stimulation (rTMS), have garnered attention as promising adjunctive therapies,⁶ using magnetic pulses to modulate cortical excitability and synaptic plasticity by inducing electric currents in targeted brain regions.^7,8

Among rTMS paradigms, theta burst stimulation (TBS) is a relatively novel, patterned form of stimulation characterized by bursts of high-frequency pulses delivered at theta rhythm (5 Hz) intervals. TBS can be administered in intermittent (iTBS) or continuous (cTBS) protocols.⁹ The shorter duration (5–10 min) and lower stimulation intensities required for TBS make it a potentially more efficient and tolerable intervention.^10,11

Dysfunction in the dorsolateral prefrontal cortex (DLPFC) and associated frontostriatal circuits has been consistently implicated in the pathophysiology of negative symptoms.^6,12,13 Given the pivotal role of the DLPFC, it has become a primary target for neuromodulation interventions.^5,14 It is observed that iTBS applied to the left DLPFC may potentiate neural circuits, offering a novel approach to treating the resistant negative symptoms in schizophrenia.^15,16 Despite promising mechanistic rationale and preliminary findings, the clinical efficacy of iTBS in schizophrenia remains underexplored, with few double-blind, sham-controlled trials published to date.^4,16

The present study aims to rigorously evaluate the efficacy and safety of experimental iTBS applied to the left DLPFC on negative symptoms in patients with schizophrenia using a double-blind, sham-controlled design. Furthermore, this study seeks to provide valuable insights into the tolerability and feasibility of experimental iTBS in the treatment of negative symptoms of schizophrenia.

METHODS

Aims

To study the effect of experimental intermittent theta burst stimulation (iTBS) on predominantly negative symptoms in schizophrenia patients in a sham-controlled double-blind trial.

Study Setting

Patients with schizophrenia, with predominant negative symptoms, were recruited from the Psychiatry Outpatient Department at All India Institute of Medical Sciences, New Delhi, India, as well as 3 other tertiary care institutes in Delhi/NCR. Diagnoses were made according to the International Statistical Classification of Diseases, Tenth Revision (ICD-10) clinical criteria. The study has been registered under Clinical Trial Registry-India (CTRI; identifier: CTRI/2019/05/019099).

Sample Selection

Following the diagnosis, patients were referred to the Neuromodulation Centre for evaluation by research investigators. Enrollment was based on predefined inclusion and exclusion criteria, with written informed consent obtained from both participants and their caregivers or legally authorized representatives.

Study Design

The prospective randomized controlled trial (RCT) was done using computer-generated random numbers. The treatment protocol was blinded to the clinical team, investigators, research staff, and patients. Only the TMS therapists were aware of active or sham groups.

Sample Size Calculation

Bation et al¹⁷ evaluated rTMS efficacy on negative symptoms of schizophrenia in a double-blind sham-controlled pilot RCT. Postintervention, mean (SD) Scale for the Assessment of Negative Symptoms (SANS) scores were 67.10 ± 24.56 (sham) and 57.83 ± 20.09 (rTMS), showing higher variability. For sample size calculation, an intermediate SD of 15 was used. Anticipating a 10-point difference in SANS scores, with α=0.05 and 90% power, the required sample size was 56 per group. Accounting for 15% attrition, the sample size was 66 per group.

Inclusion Criteria

Right-handed patients (18–60 years of age) meeting ICD-10 criteria for schizophrenia with predominant negative symptoms for at least 2 years were included. Eligibility required clinical stability on unchanged antipsychotics (eg, olanzapine, risperidone, clozapine) for ≥4 weeks, with Scale for the Assessment of Positive Symptoms (SAPS) ≤50 and SANS ≥60. Patients or their caregivers/legally authorized representatives also needed the capacity and willingness to provide informed consent.

Exclusion Criteria

Individuals with comorbid depressive disorder, bipolar affective disorder, or substance abuse/dependence within the past year (excluding nicotine and caffeine) as assessed by the Mini-International Neuropsychiatric Interview (MINI)¹⁸ were excluded. Additional exclusions included Axis II personality disorders interfering with protocol adherence, standard TMS contraindications, and known or suspected pregnancy.

Procedure

A total of 442 patients meeting ICD-10 criteria for schizophrenia were screened. Of these, 168 failed to meet inclusion criteria (mostly due to SANS scores <60), 17 were excluded for logistic reasons, and 116 declined participation (caregiver/patient apprehension, preference for medication, or inability to attend daily). The remaining 141 eligible patients provided informed consent (from both participants and caregivers/relatives) and were randomized by software to active (real iTBS) or sham iTBS groups. All procedures were conducted in accordance with the Declaration of Helsinki.

A baseline assessment was done by trained research staff using the MINI,¹⁸ SANS,¹⁹ SAPS,²⁰ Calgary Depression Scale for Schizophrenia,²¹ and Clinical Global Impression.²²

Each iTBS session lasted 5 minutes 50 seconds, delivered 5 days a week for 3 weeks (15 sessions). Sessions began with motor threshold (MT) determination and site localization; sham iTBS was applied using a sham coil. Patients continued their prescribed antipsychotics, which were recorded. Assessments were conducted postintervention and at 1-, 2-, 3-, 4-, 5-, and 6-month follow-ups using the same tools.

Of the 141 patients, 71 were assigned to active iTBS and 70 to sham. Two in the active group and 3 in the sham group discontinued before the fifth session due to attendance issues. One patient in the active group stopped at the 12th session because of worsening aggression and insomnia. Eight more (2 active, 6 sham) discontinued before completing 15 sessions, citing lack of benefit or logistical difficulties. In total, 14 patients dropped out, leaving 127 completers—66 in the active group and 61 in the sham group.

In the active arm, 69 of 71 patients and in the sham arm, 65 of 70 patients completed the postintervention assessment after 15 sessions. At the 1-month follow-up, 65 patients from the active group and 63 from the sham group were assessed. In the second month, 65 active and 61 sham patients participated; in the third month, 66 and 60, respectively; in the fourth month, 65 and 62; in the fifth month, 63 and 58; and at the 6-month follow-up, 59 patients in the active group and 57 in the sham group completed assessments. These figures illustrate gradual attrition across follow-ups, though the majority of patients remained engaged throughout the 6-month period (Figure 1).

Protocol Administered

Instead of the traditional 3-, 5-, or 6-pulse TBS protocols,²³ a novel 9-pulse iTBS paradigm was used, delivering 720 pulses per session in 8 trains of 90 pulses. Intertrain intervals were extended to prevent coil overheating and reduce patient discomfort. The total number of sessions was limited to 15, as our previous finding⁶ showed that dropout rates increase significantly beyond this point due to difficulties with daily attendance.

The stimulation site was localized 5.5 cm anterior and 0.5 cm lateral to the motor cortex and applied uniformly across both groups. Resting MT was determined by recording motor evoked potentials from the hand area (M1), and iTBS was delivered at 100% of the individual MT. Each session lasted 5 min 52 sec using a Magstim Rapid2 with a 70-mm figure-of-8 coil. Weekly follow-ups were conducted in the TMS clinic.

Statistical Analysis

Comparative analyses were conducted to evaluate the effects of TBS on negative symptoms between active and sham arms over time, as well as within subjects across different intervals. Student t test was applied to compare mean scores, while the Mann–Whitney U test was used for median comparisons when the standard deviation exceeded half the mean. A general linear model was employed to assess changes both within and between subjects over time.

RESULTS

Sociodemographic Characteristics

Table 1 shows the sociodemographic profile of the participants. The mean age was significantly lower in the active group compared to the control group (28.87 ± 7.35 vs 31.75 ± 9.72 years). Gender distribution was comparable (49 males and 22 females in the active group vs 43 males and 27 females in the sham group; P=.34). Most participants had education up to high school or graduation. Employment status was similar, with 61 (85.92%) unemployed in the active group and 60 (85.71%) in the sham group. Marital status was also comparable (70.42% married in the active group vs 64.29% in the sham group). The median duration of illness was 7 years in both groups (P=.22).

Table 2 presents intention-to-treat (ITT) and per-protocol (PP) analyses of active versus sham treatment on negative symptoms measured by SANS. The PP analysis included participants who completed postintervention assessments. Both groups showed significant within-group reductions in SANS scores from pre-to posttreatment (P<.001), indicating symptom improvement over time. However, between-group comparisons were not significant.

In ITT analysis, the mean difference in symptom change was 0.75 (95% CI: –2.91 to 4.42, P=.68) for the active group and –0.84 (95% CI: –6.76 to 5.07, P=.77) for the sham group. In PP analysis, differences remained insignificant: –0.04 (95% CI: –3.77 to 3.76, P=.99) for the active group and –0.84 (95% CI: –6.76 to 5.07, P=.77) for the sham group. These findings indicate that iTBS did not produce superior effects compared with sham.

Table 3 shows no significant baseline differences between active and sham groups across SANS domains (affective flattening, alogia, avolition-apathy, anhedonia, attention) or overall mean score (73.14 ± 11.97 vs. 72.38 ± 9.95; P=.68). This similarity persisted postintervention and at all follow-ups up to 6 months. Both groups showed significant within-group declines over time (P<.01), indicating symptomatic improvement; however, between-group differences and group×time interactions remained nonsignificant, suggesting comparable improvements across conditions.

Table 4 shows that baseline median SAPS scores indicated mild severity—8 (0–46) in the active group and 10 (0–49) in the sham group. Mann–Whitney U tests revealed no significant differences in total or domain scores between groups at any time point. Although follow-up score differences were minimal, both groups showed a gradual decline in symptom severity over time.

Table 5 shows that at treatment completion, 8 participants in the active group and 2 in the sham group showed significant improvement, though this difference was not statistically significant (χ² test). The number of improved participants gradually increased in both groups during follow-up, with no significant differences at any time point. By 6 months, 18 active and 15 sham participants demonstrated significant improvement, indicating similar patterns of symptomatic progress across groups.

Adverse Effects

A higher proportion of participants in the active group (17/71; 23.9%) reported local adverse effects compared to the sham group (4/70; 5.7%), a statistically significant difference (P = .02). Headache was reported by 8 (11.2%) active and 4 (5.7%) sham participants, though this difference was not significant (P = .23). Fatigue and sleepiness occurred in 6 (8.4%) participants in the active group but none in the sham group, showing significance (P = .01). Most adverse effects were mild, transient, and managed with psychoeducation. However, 1 participant in the active group discontinued follow-up due to worsening aggression and insomnia attributed to treatment.

DISCUSSION

Currently, no effective pharmacologic treatment exists for patients with schizophrenia who predominantly present with negative symptoms. Consequently, rTMS protocols, particularly iTBS, are being investigated, though no FDA-approved standard yet exists for schizophrenia. Previous studies using conventional rTMS have shown mixed efficacy.^6,23–25 The DLPFC has been the primary target due to its role in the pathophysiology of negative symptoms and its superficial cortical location.^26,27 More recently, a few studies have evaluated iTBS, providing preliminary but inconclusive evidence of benefit.^28,29 The present study was undertaken to further examine this evidence base while also assessing the feasibility and tolerability of iTBS in patients with negative symptoms of schizophrenia.

Impact of iTBS on Negative Symptoms

This prospective longitudinal study evaluated the efficacy of high-frequency, high-density iTBS applied to the left DLPFC in reducing negative symptoms of schizophrenia. Although both active and sham groups showed significant improvement over 6 months, no between-group differences were observed on SANS scores. These findings contribute to the ongoing discourse on rTMS modalities for treating negative symptoms.³⁰

Given these neutral findings, it is important to consider sample-related factors that may have influenced treatment responsiveness. A recent meta-analysis³¹ reported that left DLPFC iTBS produced meaningful reductions in negative symptoms across multiple RCTs. Notably, the meta-analysis included heterogeneous samples, often comprising individuals with less chronic illness and fewer primary negative symptoms—characteristics associated with greater neuroplastic responsiveness. In contrast, the present study enrolled patients with more persistent, treatment-resistant symptoms, which may have limited the efficacy of iTBS. This contrast in sample profiles underscores the potential role of illness chronicity and baseline symptom severity in modulating treatment outcomes.

Some studies have reported delayed improvements in negative symptoms following iTBS. For instance, an RCT observed significant reductions in SANS scores at 1, 3, and 6 months posttreatment, suggesting a delayed therapeutic effect of iTBS on negative symptoms.²⁸ It was demonstrated that iTBS intervention over left DLPFC significantly improves negative symptoms in individuals with schizophrenia, particularly at 4-week follow-up.³²

A systematic review and meta-analysis³³ focusing on iTBS intervention found significant effects on general negative symptoms and specific domains like anhedonia and avolition, particularly with protocols involving more than 10 sessions and once-daily stimulation. Meta-analysis³¹ pooled studies with shorter follow-up durations. Our 6-month follow-up, though methodologically rigorous, may have allowed nonspecific symptom fluctuation to obscure true treatment effects. These variations in follow-up duration and stimulation schedules highlight methodological diversity across studies, which may explain inconsistent outcomes.

The present findings contrast with a prior study²⁸ that examined iTBS effects on negative symptoms using both SANS and the Positive and Negative Syndrome Scale (PANSS). It showed significant improvement on the SANS, but not on the PANSS negative subscale, even after 6 months, suggesting greater sensitivity of the SANS in detecting therapeutic effects. Similarly, another study³⁴ reported improvements in negative symptoms on the SANS, while the PANSS failed to capture such changes. In line with this, a large multicenter, double-blind, sham-controlled trial of high-frequency rTMS over the DLPFC¹⁴ found no significant effects on negative symptoms when assessed with the PANSS, further contributing to inconsistent findings. In the meta-analysis,³¹ several positive trials used the PANSS negative subscale, which may detect broader but more nonspecific improvements. The present study’s use of the SANS, while more detailed and symptom-specific, may be less sensitive to subtle therapeutic effects. Collectively, these studies offer partial support for our results and indicate that outcome detection in neuromodulation trials may depend on the scale employed.

Thus, while our findings align with reports of limited efficacy of rTMS/iTBS for negative symptoms, they underscore the need to critically consider psychometric sensitivity in assessment. Taken together, these considerations position the current study within the broader context of heterogeneous methodologies, patient profiles, and outcome measures that shape iTBS research.

Masked Positive Symptoms and Depressive Symptoms

The discrepancy in findings may stem from several factors, including unequal distribution of secondary negative symptoms between active and sham groups. Although SAPS was applied to reduce the influence of positive symptoms, rating scales are only indicative and cannot fully exclude overlapping states. An important question is whether noninvasive interventions benefit both primary and secondary negative symptoms, or whether effects are confined to primary symptoms, with secondary symptoms responding better to pharmacologic or psychosocial treatments. This view is supported by evidence of distinct underlying mechanisms for primary versus secondary negative symptoms.³⁵ Given that these often coexist, differentiation and quantification remain clinically challenging.³⁶ Although speculative, these considerations highlight the need to examine iTBS effects separately on primary and secondary symptoms, along with their neurobiological correlates.

Role of Placebo Effect

Another important consideration in interpreting the findings is the potential influence of placebo effects. Traditionally, negative symptoms of schizophrenia have been regarded as relatively resistant to placebo responses due to their persistence and chronicity. However, recent research challenges this assumption. It was emphasized that several factors—including symptom fluctuation, short trial durations, and high active-to-placebo group ratios—can contribute to significant placebo responses, even in populations with enduring symptoms.³⁶ Even though these factors were not present in the present study, the placebo effect was still observed. This raises the need to further explore methodological contributors that may have shaped the treatment response profile observed here. One such contributor may be the absence of a formal assessment of blinding and participant expectancy. Although only the TMS therapists were aware of treatment allocation, participants were not systematically asked to indicate whether they believed they had received active or sham stimulation. Consequently, the effectiveness of blinding cannot be stated with certainty, and expectancy effects could not be quantified. In the meta-analysis,³¹ positive trials frequently utilized protocols extending over multiple weeks. Our condensed schedule, although feasible, may not have allowed sufficient cumulative stimulation to induce durable neuroplastic changes. Given this limitation, it becomes important to consider additional trial design characteristics that may have influenced placebo responsiveness.

The high active-to-placebo ratio used in many neuromodulation trials, including this study, has been linked to amplified placebo responses. When the likelihood of receiving active treatment is greater, participant expectations rise, potentially enhancing placebo effects even in sham groups.³⁷ In the absence of direct measures of expectancy, it remains unclear to what extent participants’ beliefs regarding treatment allocation may have contributed to the observed within-group improvements. This limitation is particularly relevant given the significant placebo response noted in the study.

Trial duration is another key factor, particularly in neuropsychiatric disorders where symptom fluctuations may obscure treatment effects. To reduce transient or nonspecific improvements, we incorporated a 6-month follow-up, consistent with pharmacologic guidelines emphasizing extended observation to stabilize placebo trajectories.³⁷ In addition, only clinically stable patients on unchanged antipsychotic regimens for at least 4 weeks prior to enrollment were included. Despite these measures, both groups demonstrated modest within-group improvements in negative symptoms, while between-group differences remained nonsignificant. This suggests that placebo effects persisted and may have contributed substantially to observed changes. Together, these observations indicate that even carefully controlled trial conditions may not fully mitigate placebo-driven symptom shifts, underscoring the need for refined strategies to disentangle true neuromodulatory effects from nonspecific improvement.

These findings highlight the importance of long-term follow-up and stringent patient selection to differentiate true therapeutic effects from nonspecific variability. Future studies should incorporate formal assessments of blinding integrity and participant expectancy to better characterize placebo mechanisms and strengthen causal inference in neuromodulation trials. In this context, researchers must prioritize methodological rigor to clarify whether the modest improvements seen in both groups reflect genuine neurobiological change or predominantly placebo-related dynamics.

Role of Concurrent Pharmacologic Treatment

A major challenge in long-term follow-up of short-term interventions is the uncertainty regarding treatment durability and the confounding influence of concurrent therapies. Close follow-up may improve adherence to ongoing treatments, potentially enhancing response rates even in sham groups. Benzodiazepine use—known to attenuate TMS/iTBS efficacy^38,39—was monitored in this study and prescribed occasionally for sleep or anxiety. Complete exclusion, however, was not feasible. Even limited use of short-or intermediate-acting benzodiazepines may have reduced responsiveness in the active group.

Although medication regimens remained unchanged during the intervention, adjustments during follow-up—such as adjunctive treatments or dose increases—are often clinically warranted and ethically necessary. However, we did not systematically monitor additional treatments participants might have pursued independently. Such unrecorded variables, despite initial stabilization, may have influenced outcomes, particularly in the sham group. Prior research^38,39 highlights that postintervention pharmacologic or psychosocial changes can confound neuromodulation trial results. Thus, observed improvements may reflect cumulative nonspecific therapeutic effects rather than placebo alone, complicating interpretation of iTBS-specific efficacy.

Safety of iTBS in Patients With Schizophrenia

Importantly, the novel high-density iTBS protocol was well tolerated, with no serious adverse effects such as seizures reported. However, its efficacy may have been constrained by suboptimal stimulation parameters. To deliver a higher pulse dose (720 pulses/session), pulses per burst were increased but sessions were reduced from 20 to 15. Intertrain intervals were also extended to prevent coil overheating and improve comfort. While enhancing feasibility, these modifications may attenuate neuroplasticity. Evidence^40,41 indicates that temporal dynamics—such as burst spacing and session duration—critically shape long-term potentiation–like effects, underscoring the need for protocol optimization. Thus, while safety was preserved, the trade-off between tolerability and maximal neuroplastic engagement must be carefully considered.

Many studies in the meta-analysis³¹ employed standard iTBS parameters (600 pulses/session, shorter intertrain intervals, ≥20 sessions). In contrast, our high-density protocol increased pulse dose but reduced total sessions and prolonged intertrain intervals—modifications known to influence synaptic potentiation. This comparison highlights how variations in protocol design can significantly affect treatment efficacy, even when overall pulse exposure is increased.

Evidence suggests that longer intertrain intervals disrupt synaptic consolidation and reduce excitatory effects.⁴⁰ Similarly, optimizing burst spacing is critical, as increased spacing may weaken or delay plastic changes.⁴¹ Thus, safety-driven modifications, though essential for tolerability, may have compromised efficacy by interrupting time-dependent mechanisms underlying synaptic potentiation.

Nonetheless, our findings add to evidence supporting iTBS safety in schizophrenia. Consistent with established guidelines and meta-analyses,^42,43 no serious adverse events, including seizures, were reported. Side effects were mild and transient, mainly headaches or scalp discomfort. Although the higher pulse count with a shortened schedule did not yield superior outcomes, the protocol maintained a robust safety profile. The subtherapeutic effects underscore the importance of calibrating stimulation parameters to balance neuroplastic engagement with feasibility, echoing prior work⁴³ stressing the safety and necessity of optimized rTMS (iTBS) protocols.

CONCLUSION

This study contributes to the field of neuromodulation by evaluating the efficacy, feasibility, and safety of high-density iTBS applied to the left DLPFC for negative symptoms of schizophrenia. Despite theoretical promise and earlier supportive findings, no significant differences were observed between active and sham groups across 6 months, consistent with mixed evidence. Possible confounders include placebo effects, overlapping secondary symptoms, and concurrent pharmacologic adjustments. Future research should optimize stimulation parameters, stratify patients by symptom subtype and responsiveness, and rigorously monitor adjunctive treatments. Extended follow-up will be crucial for distinguishing true therapeutic effects from nonspecific symptom fluctuations.

Article Information

Published Online: May 11, 2026. https://doi.org/10.4088/JCP.25m16093
© 2026 Physicians Postgraduate Press, Inc.
Submitted: August 22, 2025; accepted March 9, 2026.
To Cite: Kumar N, Prasad S, Srivastava MVP, et al. Effects of experimental intermittent theta burst stimulation on negative symptoms of schizophrenia: a double-blind sham-controlled study. J Clin Psychiatry 2026;87(2):25m16093.
Author Affiliations: Department of Psychiatry, All India Institute of Medical Sciences (AIIMS), New Delhi, India (N. Kumar, Madiha); Department of Psychiatry, Lady Hardinge Medical College, New Delhi, India (Prasad); Department of Neurology, Cardio Neuro Centre, AIIMS, New Delhi, India (Srivastava); Department of Biostatistics, AIIMS, New Delhi, India (Kalaivani); Neuropsychiatry C.N.C., AIIMS, New Delhi, India (Patil); Department of Psychiatry, Institute of Human Behaviour and Allied Sciences (IHBAS), New Delhi, India (M. Kumar).
Corresponding Author: Nand Kumar, MD, Professor, Department of Psychiatry, AIIMS, New Delhi, India 110029 ([email protected]).
Relevant Financial Relationships: Prof. Nand Kumar received funding from the Indian Council of Medical Research. The other authors report no relevant financial relationships.
Funding/Support: This research was funded by the Indian Council of Medical Research, New Delhi, India. No. Coord/7/(6)/CARE-NM/2018/NCD-11.
Acknowledgment: The authors thank the patients, clinics, staff, and colleagues who made this research possible.
ORCID: Nand Kumar: https://doi.org/0000-0002-2538-996X

Clinical Points

• Effective treatments for persistent negative symptoms in schizophrenia are limited, and previous intermittent theta burst stimulation (iTBS) studies have been inconsistent. This study explores whether left dorsolateral prefrontal cortex iTBS helps patients with chronic, treatment-resistant symptoms.
• Practitioners should recognize that iTBS may not provide meaningful benefit for longstanding or primary negative symptoms, particularly with stable medication and limited neuroplasticity.
• When using neuromodulation, patient characteristics, treatment duration, and protocol are important considerations. Some patients may need longer, tailored courses or alternative strategies if standard iTBS does not work.