Non-inferiority of Accelerated vs. Repetitive Transcranial Magnetic Stimulation in Major Depressive Disorder: A Systematic Review

Abstract

Repetitive transcranial magnetic stimulation (rTMS) is a non-invasive neuromodulation technique used to treat major depressive disorder (MDD), particularly in patients with treatment resistant depression. More recently, accelerated transcranial magnetic stimulation (aTMS) has shown comparable efficacy while offering a faster treatment option. This review assesses whether aTMS is non-inferior to rTMS. This systematic review compared aTMS and rTMS protocols for MDD treatment. Studies were identified through PubMed, Cochrane Library, and Web of Science (September 2024). Non-MDD populations and non-aTMS techniques were excluded, such as studies that did not qualify as clinical trials, cohort or randomized controlled trial. Inclusion criteria required direct protocol comparisons and a minimum sample of 10 participants. Bias assessment followed the Cochrane Collaboration’s robvis tool. Four studies were included, encompassing a total sample of 219 patients. Results showed symptom improvement across all groups, with accelerated protocols demonstrating faster response in some cases. However, long-term efficacy varied, and no study provided conclusive superiority of one protocol over the other. This study reveals the variability in aTMS protocols and the inconsistency in the depression rating scales selected across studies, noting that the lack of standardised assessment methods limits meta-analytical potential. Short follow-up durations contrast with antidepressant studies, restricting long-term efficacy evaluation. Additionally, improvements in sham methodologies have increased placebo responses, further influencing treatment outcome interpretation. This systematic review highlights the efficacy, tolerability, and cost-effectiveness of aTMS, suggesting its non-inferiority to rTMS and underscoring the need for further research to define the optimal accelerated protocol.

INTRODUCTION

Major depressive disorder (MDD) is one of the most common psychiatric diseases worldwide [1]. Its primary symptoms include anhedonia and depressed mood, which may be accompanied by additional symptoms such as weight loss or gain, insomnia or hypersomnia, agitation, or fatigue. These symptoms must represent a significant change in patients’ baseline function, persist for at least two weeks and cannot be attributed to other medical conditions or substances. Thus, this condition has a profound impact in patients’ quality of life [2].

The treatment of MDD involves multiple approaches. Firstly, lifestyle modifications should be implemented, including regular exercise, establishing a healthy sleep routine, controlling addictions, making dietary adjustments and others [3,4]. If these measures are insufficient, psychotherapy and pharmacological treatment are considered the gold standard for managing depression [5]. However, some patients fail to respond to at least two distinct antidepressants, while others experience frequent relapses – a condition known as treatment resistant depression (TRD) [6]. TRD affects approximately one- third of MDD patients, leading to a poor prognosis, worse medical outcomes, increased disability, faster cognitive decline, and a higher risk of dementia. Given its impact, TRD is a major concern in psychiatry and public health [7,8]. Therefore, new ways to handle depression are rising due to the high incidence and comorbidities mentioned before.

Emerging therapies, such as psychedelics [9] and transcranial magnetic stimulation (TMS) protocols are being tested and adopted worldwide, showing promising results.

TMS is currently used to treat two main conditions: obsessive-compulsive disorder and MDD, targeting different brain regions with specific stimulation protocols. Furthermore, recent studies are exploring its applications in other conditions, including cognitive impairment, addictions, post-traumatic stress disorder, neuroimaging and in negative symptoms of schizophrenia [10-16].

Repetitive transcranial magnetic stimulation (rTMS) has been effective for MDD treatment [17]. The most used protocol of this technique involves stimulating the left dorsolateral prefrontal cortex (L-DLPFC) with 10 Hz daily for 4 to 6 weeks [18].

Despite its efficacy, this protocol has shown significant limitations. Notably, the length of the treatment (more than a month) before achieving therapeutic effects can pose risks for patients in acute crises, such as those with suicidal behaviour. Additionally, factors like full-time employment, childcare responsibilities, or even transport difficulties can severely impact treatment adherence and increase morbidity [19].

In recent years, shortened protocols have been developed to improve compliance and comfort for patients. One of them is accelerated transcranial magnetic stimulation (aTMS), which involves intermittent theta burst stimulation (iTBS) delivered more than one time a day over a period of five days. iTBS is a form of rTMS that mimics endogenous theta rhythms, significantly reducing session duration while maintaining therapeutic efficacy [20].

Thus, this systematic review aims to evaluate whether aTMS is non-inferior to rTMS in treating MDD. If proven effective, aTMS could allow the treatment of a greater number of patients using the same resources within a shorter timeframe, while simultaneously improving adherence to therapy.

METHODS

A systematic review of the literature on aTMS vs. rTMS protocols for MDD treatment was elaborated according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses [21].

The study selection process was conducted in September 2024 using three databases: PubMed, Cochrane Library, and Web of Science. Search terms used to refer to aTMS included “accelerated transcranial magnetic stimulation”; “accelerated repetitive transcranial magnetic stimulation”; “accelerated TMS”; and “aTMS”. For MDD, the terms “major depressive disorder”; “depression”; and “MDD” were used. Filters were applied to include only studies published in English and after 2010. This restriction was justified by the fact that aTMS is a relatively new technique, with its current protocols and applications becoming widely adopted only in recent years — the earliest study identified using an accelerated protocol dates back to 2010, thereby justifying the applied filter [22].

Following the removal of duplicates, two independent reviewers screened the titles and abstracts based on predefined exclusion criteria. Studies were excluded if they did not qualify as clinical trials, cohort or randomized controlled trial (RCT) studies; involved non-human subjects; focused on diseases other than MDD; used techniques other than aTMS; or had objectives unrelated to aTMS in MDD treatment.

Subsequently, the full texts of eligible studies were reviewed by the same two reviewers. Inclusion criteria required a direct comparison between aTMS and rTMS protocols for MDD treatment, with a minimum sample size of 10 participants. Prato et al. [23] is a preliminary study. Although it does not provide detailed information on every step, its inclusion was unanimously decided due to the presence of essential data relevant to protocol comparison and statistical analysis. Prodi et al. [24] with a sample size of 28, includes 8 patients with bipolar depression, which may introduce bias. However, its inclusion was unanimously agreed upon, as the primary diagnosis was unipolar depression.

Both researchers independently conducted the article selection process and data collection reaching full agreement on all decisions. In cases where one researcher had uncertainties regarding a particular study, the final decision was made following clarification provided by the other. All selections were thus made with mutual consensus, ensuring a rigorous and unbiased inclusion process.

For each study, data were extracted on study details (first author, publication year, study design and country), patient sample characteristics for rTMS, aTMS, and sham groups (sample size, mean age and male-to-female ratio), diagnosis scale or method, follow-up time, specific technique parameters for rTMS and aTMS (stimulation site, frequency and intensity, sessions’ schedule and pulses per session) and outcome assessment methods. All extracted data are provided in Tables 1 and 2.

Table 1.

Study details and demographic

Study	Study design	Country	rTMS - patient sample (N/mean age/M/F)	aTMS - patient sample (N/mean age/M/F)	Sham	Main diagnosis	Diagnosis scale/method	Follow-up time
Prodi et al. [24], 2024	Cohort	Italy and USA	9/55/2/7	19/55/8/11	NA	TRD	NA/interview	3 mo
Prato et al. [23], 2023	Single-blind RCT	Italy	15/NA/NA/NA	18/NA/NA/NA	NA	MDE	NA/NA	56 day
Kim et al. [26], 2021	Single-blind RCT	South Korea	22/43.9 ± 10.9/3/19	21/45.1 ± 13.5/4/17	8/49.1 ± 15.6/2/6	MDD	DSM-IV/interview	42 day
Fitzgerald et al. [27], 2018	Single-blind RCT	Australia	57/49.9/24/33	58/48.2/25/33	NA	MDD and TRD	MINI, MADRS/interview	56 day

Values are presented as number only, mean, or mean ± standard deviation.

N, sample size; M, male; F, female; rTMS, repetitive transcranial magnetic stimulation; aTMS, accelerated transcranial magnetic stimulation; NA, not available; TRD, treatment resistant depression; RCT, randomized controlled trial; MDE, major depressive episodes; MDD, major depression disorder; DSM-IV, Diagnostical and Statistic Manual of Mental Disorders; MINI, Mini-International Neuropsychiatric Interview; MADRS, Montgomery and Åsberg Depression Rating Scale.

Table 2.

Stimulation and outcome method

Study	rTMS					aTMS					Outcome method
	Stimulation site	Frequency intensity	Sessions' schedule	Pulses/session		Stimulation site	Frequency intensity	Sessions' schedule	Pulses/session	Intersession interval
Prodi et al. [24], 2024	L-DLPFC	10 Hz 120%	1/day × 5 days a week × 4 weeks	3,000		L-DLPFC	10 Hz 120%	2/day × 5 days a week × 2 weeks	3,000	NA	HAM-D, HAM-A, MADRS, YMRS, SDS, MoCA, CGI-S, C-SSRS; measured at baseline, end of treatment, 1 and 3 months after treatment
Prato et al. [23], 2023	NA	NA	1/day × 4 weeks	NA		NA	NA	4/day × 5 days	NA	NA	MADRS, BDI-II, SSI; measured on day 0, 1, 2, 3, 4, 5, 14, 21, 28, and 56
Kim et al. [26], 2021	L-DLPFC	10 Hz 110%	1/day × 5 days a week × 3 weeks	3,000		L-DLPFC	10 Hz 110%	5/day × 3 days	3,000	NA	KQIDS-SR, KQIDS-C, CGI-S, CGI-I; measured at baseline, 3 days, 3 and 6 weeks
Fitzgerald et al. [27], 2018	L-DLPFC	10 Hz 120%	1/day × 5 days a week × 4 weeks	3,150		L-DLPFC	10 Hz 120%	Week 1: 3/day × 3 days Week 2: 3/day × 2 days Week 3: 3/day × 1 day	3,500	15−30 min	MADRS, BDI, SSI, HAM-D; measured at baseline, 1, 2, 3, 4, and 8 weeks

rTMS, repetitive transcranial magnetic stimulation; aTMS, accelerated transcranial magnetic stimulation; L-DLPFC, left dorsolateral prefrontal cortex; NA, not available; HAM-D, Hamilton Rating Scale for Depression; HAM-A, Hamilton Anxiety Rating Scale; MADRS, Montgomery and Åsberg Depression Rating Scale; YMRS, Young Mania Rating Scale; SDS, Self-Rating Depression Scale; MoCA, Montreal Cognitive Assessment; CGI-S, Clinical Global Impression Scale - Severity of Illness; C-SSRS, Columbia Suicide Severity Rating Scale; BDI-II, Beck Depression Inventory II; SSI, Beck Scale of Suicide Ideation; KQIDS-SR, Korean Quick Inventory of Depressive Symptomatology – Self-Reported; KQIDS-C, Korean Quick Inventory of Depressive Symptomatology – Clinician Administered; CGI-I, Clinical Global Impression Scale – Global Improvement.

The risk of bias was assessed using the Cochrane Collaboration’s risk of bias tool – robvis [25]. This provides a qualitative assessment of the included studies based on five distinct domains, along with an overall judgement column for each study. The domains considered are: bias due to randomisation, bias due to deviations from intended intervention, bias due to missing data, bias due to outcome measurement and bias due to selection of reported result.

Ethical Considerations

This study is a systematic review of previously published data and did not involve the recruitment of human participants or the collection of original clinical data. As such, approval by a research ethics committee was not required. All included studies were published in peer-reviewed journals and, where stated, received appropriate institutional ethical approval. The ethical status of each included study was considered during the selection process.

RESULTS

The full search and selection strategies are listed in Figure 1. The risk of bias summary table is provided in Figure 2.

Fig. 1.

PRISMA flow diagram.

aTMS, accelerated transcranial magnetic stimulation; rTMS, repetitive transcranial magnetic stimulation; MDD, major depression disorder; PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

Fig. 2.

Cochrane Collaboration’s risk of bias tool – robvis.

Included Studies

Prodi et al. [24] conducted a cohort study in Italy and the United States, including 9 patients in the rTMS group (one session per day, five days a week for four weeks) and 19 in the aTMS group (two sessions per day, five days a week for two weeks), all diagnosed with TDR. Outcome measures included multiple depression scales: Hamilton Rating Scale for Depression (HAM-D), Hamilton Anxiety Rating Scale (HAM-A), Montgomery and Åsberg Depression Rating Scale (MADRS), Young Mania Rating Scale (YMRS), Self-Rating Depression Scale (SDS), Montreal Cognitive Assessment (MoCA), Clinical Global Impression Scale - Severity of Illness (CGI-S), and Columbia Suicide Severity Rating Scale. Assessments were conducted at baseline (T0), end of treatment (T1), and at 1 and 3 months post-treatment (T2, T3). At baseline, no statistically significant differences were observed between groups. A progressive reduction in depressive symptoms was recorded across all timepoints, with significant improvements in HAM-D, MADRS, CGI-S, and SDS scores (p ≤ 0.001). Both standard and accelerated protocols led to clinical benefits, with statistically significant changes at T1, T2, and T3 compared to T0, as well as between T1 and T3 and T2 and T3 (p ≤ 0.05). HAM-A scores decreased significantly (p ≤ 0.001), particularly in the accelerated protocol group, where reductions were observed at all timepoints compared to T0 (p ≤ 0.01). In the standard protocol group, this decrease was significant up to T2 (p ≤ 0.01) but not at T3. Cognitive function, assessed by MoCA scores, showed a significant improvement over time (p ≤ 0.01), especially in patients receiving accelerated rTMS, though a minor decline was noted at T3 compared to T2 (p ≤ 0.05). Suicidal risk (C-SSR score) decreased significantly in the accelerated rTMS group, particularly at T3 compared to T0 (p ≤ 0.05), while no significant reductions were observed in the standard protocol group. YMRS scores remained stable, with no significant changes over time. The study concluded that both rTMS and aTMS led to improvements in anxiety, depression, and cognitive symptoms. When directly comparing the two protocols, their efficacy appeared similar at week one, but the accelerated protocol demonstrated a superior clinical response in the longer term.

Prato et al. [23] conducted a single-blind RCT in Italy, including 15 patients in the rTMS group (one session per day for four weeks) and 18 in the aTMS group (four sessions per day for five days), all diagnosed with major depressive episodes (MDEs) and treated with venlafaxine or fluvoxamine. Outcomes were assessed using MADRS, Beck Depression Inventory II (BDI-II), and Beck Scale for Suicide Ideation (SSI) at multiple timepoints (days 0, 1, 2, 3, 4, 5, 14, 21, 28, and 56). MADRS scores showed a significant improvement on day 5 (p = 0.001) and day 56 (p = 0.037), while BDI-II differences were not statistically significant, and SSI scores showed no significant differences between protocols. Response rates were higher with aTMS, at 84.6% vs. 45.5% on day 28 (p = 0.043) and 92.3% vs. 45.5% on day 56 (p = 0.012). Remission rates were also greater, reaching 76.9% vs. 18.2% on day 28 (p = 0.004) and 69.2% vs. 36.4% on day 56 (p = 0.107). The study concluded that aTMS is faster and more effective than standard rTMS in patients with MDEs.

Kim et al. [26] conducted a single-blind RCT in South Korea, including 22 patients in the rTMS group (one session per day, five days a week for three weeks), 21 in the aTMS group (five sessions per day for three days), and 8 in the sham group (following the aTMS protocol). All participants were diagnosed with MDD using MADRS. Outcome measures included the Korean Quick Inventory of Depressive Symptomatology – Self-Reported (KQIDS-SR) and Clinician-Administered (KQIDS-C), CGI-S, and Clinical Global Impression Scale – Global Improvement (CGI-I), assessed at baseline, three days, and at three and six weeks. The analysis showed a significant main effect of time, indicating symptom improvement across all groups. KQIDS-SR scores improved significantly in both accelerated and conventional rTMS groups compared to sham at week 6 (p = 0.015 and p = 0.027, respectively), though only accelerated rTMS showed a significant improvement at week 3 (p = 0.011). No significant differences were observed between the accelerated and conventional rTMS groups. KQIDS-C scores followed a similar trend, with significant differences at week 3 between accelerated rTMS and sham (p = 0.011), but not at week 6. For CGI-S scores, a significant group × time interaction effect (p = 0.044) was found. At week 3, the accelerated rTMS group had significantly lower CGI-S scores than the conventional rTMS group (p = 0.038) and showed a trend towards improvement over sham (p = 0.082). By week 6, CGI-S scores remained significantly different only between the accelerated rTMS and sham groups (p = 0.022). For CGI-I scores, there was a significant main effect of group (p = 0.017), with accelerated rTMS outperforming sham at both week 3 (p = 0.005) and week 6 (p = 0.020). Overall, accelerated rTMS led to a faster and more pronounced clinical improvement compared to sham, particularly in the early phase of treatment.

Fitzgerald et al. [27] conducted a single-blind RCT in Australia, including 57 patients in the rTMS group (one session per day, five days a week for four weeks) and in the aTMS group (first week: three sessions per day for three days; second week: two sessions per day for two days; third week: three sessions in one day). All patients were diagnosed with TDR and MDD using the Mini-International Neuropsychiatric Interview and MADRS. Outcome measures included MADRS, BDI, SSI, and HAM-D, assessed at baseline and weeks 1, 2, 3, 4, and 8. The analysis showed a significant main effect of time, indicating symptom improvement across all measures. However, there was no significant effect of treatment group or time-by-group interaction, suggesting similar efficacy between accelerated and standard protocols. In conclusion, both rTMS and aTMS were effective in treating depression, with no significant differences in clinical outcomes.

DISCUSSION

The efficacy of aTMS has been demonstrated in several RCTs comparing it with sham in recent years [20,28-33]. The protocols vary, ranging from two to ten sessions per day over four, five or ten days. Due to this variability, some other studies have directly compared different protocols [29,34], but the current evidence remains insufficient to draw definitive conclusions regarding the ambiguity of accelerated protocols.

Efficacy has been assessed both clinically, as in the aforementioned studies, and physiologically [20,28]. To understand the objectives of these studies, it is important to recognise that when TMS was first introduced, the optimal stimulation site was uncertain. Research later established the L-DLPFC as the gold standard [35-38]. Physiological findings indicate that three functional connectivity pairs were significantly altered compared with sham from baseline to the end of treatment. Among these, two showed clinically relevant effects, as evidenced by statistically significant changes in the MADRS — specifically, the left hemisphere amygdala with the left hemisphere default mode network and the right hemisphere amygdala with the left hemisphere salience network [28]. Prior to these findings, perfusion changes were assessed [20], revealing that clinical improvement correlated with increased perfusion in the left superior medial gyrus and right inferior parietal lobule, alongside decreased perfusion in the left para-hippocampal gyrus and right posterior cerebellar lobe.

Both aTMS and rTMS appear to be safe and well tolerated by most patients [26,27,30,34,39-41]. However, some adverse events should not be overlooked. The most common include headache, dizziness, nausea, fatigue and discomfort in the stimulated area. These findings are inconsistent across the literature. While one study reports mild fatigue in a single patient [39], another describes a case where treatment was discontinued due to severe headache and eye pain [34]. One study specifically noted fatigue and headache as the only adverse effects, both of which resolved with ibuprofen [30]. In Kim et al. [26], the Frequency, Intensity, and Burden of Side Effects Rating (FIBSER) scale was used to assess and compare patient-reported side effects across the aTMS, rTMS or sham groups. On day 3 of treatment, FIBSER scores were higher in the sham group than in the treated groups, with no significant difference between aTMS and rTMS. By week 3, no statistically significant differences were observed among any of the groups. At week 6, the findings mirrored those from day 3. These results suggest that aTMS is a safe treatment option and may facilitate higher adherence rates compared to rTMS, given the shorter protocol [40].

This systematic review has some limitations. Although the total sample size of 219 patients provides some statistical power, the number of studies directly comparing accelerated and conventional TMS protocols remains very limited [23,24,26,27]. This highlights the need for further research on this relatively recent TMS methodology.

The included studies encompassed patient populations diagnosed with MDD, TRD, and MDEs. While these diagnostic entities are not entirely synonymous, they share substantial symptomatic and clinical overlap, particularly in the context of TMS treatment trials. Importantly, all included patients presented with significant depressive symptomatology confirmed by structured or semi-structured psychiatric interviews, which supports the clinical relevance of comparing these populations. Nevertheless, we acknowledge that such diagnostic heterogeneity may influence treatment response and introduces a potential limitation to the interpretation of efficacy comparisons between protocols.

Another important limitation is the relatively short follow-up duration in all included studies, particularly when compared to other therapies such as antidepressants, which often demonstrate more pronounced effects after six months. Additionally, the placebo effect is a critical factor in these interventions due to their inherent nature, with sham response rates in rTMS trials increasing over time. This trend is likely driven by methodological refinements, including improved blinding techniques, enhanced sham coils, and better control of sensory artifacts. Similar patterns have been observed in antidepressant and pain treatment trials, where heightened clinical attention and patient expectations may further strengthen placebo effects across different therapeutic approaches [42]. A major limitation of the current evidence base lies in the considerable heterogeneity of stimulation parameters—including frequency, intensity, number of sessions—and the wide variation in outcome measures employed across studies. Rating scales such as the MADRS, HAM-D, BDI, and QIDS differ in sensitivity, clinical utility, and cross-cultural validity, complicating direct comparison. While this variability reflects the evolving nature of aTMS protocols, it simultaneously limits the feasibility of meta-analytical synthesis. Harmonisation of stimulation protocols and the adoption of standardised, validated outcome measures would significantly enhance comparability and facilitate future pooled analyses.

This review is relevant and valuable for hospitals and clinics that perform TMS, helping them make informed decisions on the most appropriate protocol for patients with MDD. As previously mentioned, aTMS has been proven effective, but its efficacy compared to rTMS remains a key consideration. This study analyses all available data specifically comparing aTMS and rTMS, with results indicating that both approaches yield similar outcomes.

The findings of this review suggest that aTMS may represent a clinically viable and resource-efficient alternative to conventional rTMS. In clinical practice, aTMS protocols could be particularly suitable for patients requiring rapid therapeutic onset —such as those presenting with suicidal ideation — or for individuals for whom logistical constraints limit prolonged treatment courses. In contrast, conventional rTMS may be more appropriate in scenarios requiring extended clinical monitoring or gradual dose escalation. These distinctions underscore the importance of individualised treatment planning. Future research should explore patient-centred factors, such as urgency of treatment, occupational and caregiving responsibilities, and healthcare system capacity, to determine the optimal context for implementing aTMS in real-world settings.

CONCLUSION

To our knowledge, this is the first systematic review to directly compare accelerated and repetitive TMS protocols. In conclusion, this review provides a critical assessment of these approaches for treating MDD. It underscores the need for standardised follow-up scales and evaluation time points to facilitate quantitative comparisons. While this study focused on qualitative analysis, the findings suggest that aTMS and rTMS demonstrate comparable efficacy. Given the cost-effectiveness of accelerated protocols, they may offer a preferable alternative to conventional rTMS. However, further research is needed to establish the optimal accelerated protocol and evaluate long-term outcomes.