Comparative Efficacies of Electroconvulsive Therapy (ECT) and Repetitive Transcranial Magnetic Stimulation (rTMS) in Treatment-Resistant Depression: A Randomized Controlled Trial in Nigeria

Abstract

Background:

Treatment-resistant depression (TRD) poses significant challenges globally, particularly in low-resource settings. While electroconvulsive therapy (ECT) remains the standard for TRD, repetitive transcranial magnetic stimulation (rTMS) offers a non-invasive alternative. This study compared short- and long-term efficacy and safety of ECT versus rTMS among Nigerian patients with TRD.

Methodology:

In a randomized, rater-blinded trial, 40 adults with TRD received either ECT (n=20; bilateral, twice weekly for 3 weeks) or high-frequency rTMS (n=20; left dorsolateral prefrontal cortex, 3000 pulses/session, 5 sessions/week for 6 weeks). Depression severity (HAM-D) was assessed at baseline, weeks 2, 4, 6, and at 6 and 12 months.

Results:

Both interventions significantly reduced HAM-D scores (p<0.001). ECT showed faster early improvement (Week 4: ECT=10.30±2.84, rTMS=12.25±2.99; p<0.001), with comparable outcomes at Week 6 (ECT=4.30±2.23, rTMS=4.40±2.35; p=0.89). However, rTMS demonstrated superior long-term control at 6 months (4.05±2.11 vs. 6.60±2.97; p<0.001) and 12 months (5.25±2.89 vs. 10.55±3.45; p<0.001). Remission rates at 12 months were 65% for rTMS versus 20% for ECT. Adverse events, particularly cognitive impairment, were more frequent with ECT (60% vs. 0%; p<0.001).

Conclusion:

While ECT provides faster acute symptom relief, rTMS offers superior sustained benefits with fewer adverse effects, suggesting its potential as a first-line neurostimulation for TRD in resource-constrained settings. Larger studies with cost-effectiveness analyses are warranted.

Introduction

Major depressive disorder (MDD) remains a leading cause of global disability, with over 280 million people affected worldwide according to the World Health Organisation [1,2]. A significant subset of patients with MDD does not achieve remission despite adequate pharmacological intervention. This condition, termed treatment-resistant depression (TRD), is reported to affect approximately 20% to 30% of patients [3,4]. Various neurostimulation techniques have emerged to address this challenge, with electroconvulsive therapy (ECT) historically regarded as the most effective intervention for TRD [5,6].

Despite its high efficacy, ECT is associated with side effects, such as transient cognitive impairment, and requires general anaesthesia, thereby limiting its acceptability and accessibility in some settings [7-9]. Repetitive transcranial magnetic stimulation (rTMS), a non-invasive neuromodulation technique, has gained popularity in recent years. It involves the application of magnetic pulses to specific brain regions to modulate neuronal activity and has shown comparable antidepressant effects in some studies [10-12].

Several meta-analyses and systematic reviews have attempted to compare rTMS with ECT. Ren et al. reported that while ECT was more effective for psychotic depression, rTMS had fewer side effects and better tolerability [13]. Eranti et al. found that both interventions were effective, with rTMS offering advantages in outpatient settings [14]. Meanwhile, studies by Magnezi et al. and Loo et al. demonstrated that patient preference often favours rTMS due to its non-invasive nature [15,16]. These findings underscore the need for head-to-head trials, particularly in low- and middle-income countries where contextual factors such as stigma, cost, and access to anaesthesia significantly influence treatment decisions.

In Nigeria, the availability of ECT remains limited to tertiary centres, and there is increasing interest in rTMS as an alternative. However, few studies have compared the two treatments locally. This study aimed to compare the efficacy, time course of response, safety profile, and durability of symptom remission between rTMS and ECT in a psychiatric setting of a Nigerian tertiary facility. We hypothesised that while ECT would show faster early response, rTMS would offer more sustained long-term remission with fewer adverse effects.

Materials and Methods

Study Design

This was a randomized, controlled, single-blind (rater-blinded), parallel-group clinical trial conducted at the Psychiatric Department of Ahmadu Bello University Teaching Hospital (ABUTH), Zaria, Nigeria, over 18 months (January 2024 to July 2025).

Participants

Eligible participants were adults aged ≥18 years with a diagnosis of major depressive disorder according to DSM-5 criteria, confirmed by the Mini International Neuropsychiatric Interview (MINI). TRD was defined as failure to respond to adequate doses (therapeutic doses for ≥6 weeks) of at least two different classes of antidepressants, as determined by the Antidepressant Treatment History Form (ATHF; score ≥3 for each trial). Additional inclusion criteria were baseline 17-item Hamilton Depression Rating Scale (HAM-D) score ≥20, stable medical status, and provision of written informed consent.

Exclusion criteria included: neurological disorders (e.g., epilepsy, traumatic brain injury, brain tumour); comorbid psychiatric disorders (e.g., schizophrenia, bipolar disorder, substance use disorder); previous exposure to ECT or rTMS; contraindications to anaesthesia (for ECT) or magnetic stimulation (for rTMS); pregnancy or breastfeeding; and presence of metallic implants in the head or neck.

Sample size determination

Sample size was calculated using G*Power 3.1 software for a repeated-measures ANOVA with between-within interaction. Assuming a moderate effect size (f = 0.25; equivalent to d = 0.5), α = 0.05, power = 0.80, correlation among repeated measures = 0.5, and 6 measurement time points, the required sample size was 34. To account for 15% attrition, we aimed to enroll 40 participants (20 per group). This aligns with sample sizes in prior RCTs comparing neurostimulation therapies in TRD [7,14].

Randomization and Blinding

Participants were randomly assigned (1:1) to either ECT or rTMS using a computer-generated random sequence (Stat Trek Random Number Generator) prepared by an independent statistician. Allocation concealment was maintained using sequentially numbered, sealed, opaque envelopes opened by the treating clinician after baseline assessment. The outcome assessor (a trained psychiatrist administering HAM-D) was blinded to treatment allocation throughout the study. Blinding integrity was assessed at study completion by asking the rater to guess group allocation (kappa = 0.85, indicating successful blinding).

Instrument for the Study

Sociodemographic questionnaire

Key variables collected include age, marital status, gender, and tribe. Hamilton Depression Rating Scale (HAM-D)

The HAM-D (also known as the HDRS) is the most widely used clinician-administered depression assessment scale. The original version contains 17 items (HDRS17) pertaining to symptoms of depression experienced over the past week. Although the scale was designed for completion after an unstructured clinical interview, there are now semi-structured interview guides available. The HAM-D was originally developed for hospital inpatients, thus the emphasis on melancholic and physical symptoms of depression. A later 21-item version (HDRS21) included 4 items intended to subtype the depression, but which are sometimes incorrectly used to rate severity. A limitation of the HAM-D is that atypical symptoms of depression (e.g., hypersomnia, hyperphagia) are not assessed. Method for scoring varies by version. For the HDRS17, a score of 0–7 is generally accepted to be within the normal range (or in clinical remission), while a score of 20 or higher (indicating at least moderate severity) is usually required for entry into a clinical trial [14].

Study Procedure

Following enrollment and baseline assessments, participants were randomly assigned to either ECT or rTMS. The treating clinician opened sequentially numbered, sealed opaque envelopes containing group allocation after all baseline assessments were completed. The outcome assessor remained blinded to treatment assignment throughout the study.

The ECT group received bilateral (bitemporal) ECT using a MECTA Spectrum 5000Q device under general anaesthesia (propofol 1.5–2.5 mg/kg) and muscle relaxation (succinylcholine 0.5–1.0 mg/kg). Seizure threshold was titrated at the first session using the empirical titration method; subsequent treatments were delivered at 1.5–2.5 times the threshold. ECT was administered twice weekly (Tuesdays and Fridays) for three weeks, totalling six sessions.

The rTMS group received high-frequency stimulation using a Magstim Rapid²stimulator with a 70-mm figure-of-eight air-cooled coil positioned over the left dorsolateral prefrontal cortex using the standard 5-cm rule. Resting motor threshold was determined daily. Stimulation parameters were: 10 Hz frequency; 110% resting motor threshold intensity; 4-second train duration; 26-second inter-train interval; 75 trains per session; 3000 pulses per session. Treatment was administered five sessions weekly for six weeks (30 sessions total), with each session lasting approximately 30 minutes.

All participants continued standard pharmacotherapy as clinically indicated throughout the study. Depression severity was assessed using the 17-item Hamilton Depression Rating Scale at baseline, weeks 2, 4, and 6 (end of acute treatment), and at 6 and 12 months follow-up. Adverse events were monitored at each session using a structured checklist, and cognitive function was assessed using the Montreal Cognitive Assessment at baseline, week 6, and 6 months. Participants who showed clinical deterioration or experienced severe adverse events were withdrawn and offered alternative standard care.

Interventions

rTMS Group: rTMS was delivered using a Magstim Rapid²stimulator with a 70-mm figure-of-eight air-cooled coil (The Magstim Company Ltd, Wales, UK). The coil was positioned over the left dorsolateral prefrontal cortex (DLPFC) using the standard 5-cm rule (measured from the motor cortex hotspot for the right abductor pollicis brevis muscle). Resting motor threshold (RMT) was determined daily as the minimum stimulus intensity producing a motor evoked potential ≥50 μV in the relaxed contralateral abductor pollicis brevis in at least 5 of 10 trials. Stimulation parameters were: frequency 10 Hz; intensity 110% of RMT; train duration 4 seconds; inter-train interval 26 seconds; 75 trains per session; total pulses per session = 3,000. Treatment was administered 5 sessions per week for 6 weeks (30 sessions total). Each session lasted approximately 30 minutes.

ECT Group: ECT was administered using a MECTA Spectrum 5000Q device (MECTA Corporation, Oregon, USA) under general anaesthesia (propofol 1.5–2.5 mg/kg) and muscle relaxation (succinylcholine 0.5–1.0 mg/kg). Bilateral (bitemporal) electrode placement was used with brief pulse stimulation (pulse width 1.5 milliseconds). Seizure threshold was titrated at the first session using the empirical titration method; subsequent treatments were delivered at 1.5–2.5 times threshold. Seizure adequacy was monitored electroencephalographically (EEG) and motorically (cuff method); inadequate seizures (EEG duration <25 seconds) were managed by restimulation at higher intensity. ECT was administered twice weekly (Tuesdays and Fridays) for 3 weeks, totalling 6 sessions.

Outcome Measures

The primary outcome was depression severity measured using the 17-item Hamilton Depression Rating Scale (HAM-D) at baseline, weeks 2, 4, and 6 (end of acute treatment), and at 6 and 12 months follow-up. The HAM-D was administered by a trained, blinded rater. The scale has been validated in Nigerian populations, and a translated version (Hausa) was available as needed.

Secondary outcomes included:

• i.Response: ≥50% reduction in HAM-D score from baseline
• ii.Remission: HAM-D score ≤7
• iii.Adverse events: systematically monitored at each session using a structured checklist; cognitive function assessed using Montreal Cognitive Assessment (MoCA) at baseline, week 6, and 6 months

Follow-up Procedures

During the 12-month follow-up period, all participants continued to receive standard pharmacotherapy as clinically indicated (antidepressants at therapeutic doses). No maintenance ECT or rTMS booster sessions were administered. Concomitant psychotherapy was permitted and recorded. Relapse (defined as HAM-D increase ≥50% from week 6 score with score ≥16) was managed by the treating psychiatrist according to clinical guidelines, with treatment changes documented but not analysed as part of the primary outcome.

Statistical Analysis

Analyses were conducted using IBM SPSS Statistics version 29 (IBM Corp., Armonk, NY, USA). Primary analyses followed intention-to-treat (ITT) principles with last observation carried forward (LOCF) for missing data. Sensitivity analyses compared ITT results with complete-case analysis. Descriptive statistics summarised baseline characteristics. Between-group differences in continuous variables were assessed using independent t-tests or Mann-Whitney U tests as appropriate; categorical variables were compared using chi-square or Fisher's exact tests.

For the primary outcome (HAM-D scores over time), a split-plot repeated-measures ANOVA was conducted with time (6 levels) as within-subject factor and treatment group (ECT vs. rTMS) as between-subject factor. Mauchly's test assessed sphericity; Greenhouse-Geisser correction was applied when violated. Partial eta squared (ηp²) was calculated as a measure of effect size, with 95% confidence intervals. Post-hoc pairwise comparisons with Bonferroni correction examined between-group differences at each time point. Response and remission rates were compared using chi-square tests at each time point. Adverse event frequencies were compared using Fisher's exact tests. Statistical significance was set at p ≤ 0.05 (two-tailed).

Ethical Considerations

The trial was approved by the Health Research Ethics Committee of Ahmadu Bello University Teaching Hospital, Shika-Zaria (Ref: ABUTHZ/HREC/H48/2024). All participants provided written informed consent after a full explanation of study procedures, risks, and benefits. Confidentiality was maintained throughout. The study was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines. Participants who showed clinical deterioration or experienced severe adverse events were withdrawn and offered alternative standard care.

Results

A total of 78 patients were screened for eligibility between January and May 2024. Of these, 38 were excluded (25 did not meet the inclusion criteria, 13 declined participation). The remaining 40 participants were randomly assigned to ECT (n=20) or rTMS (n=20). Three participants discontinued during the 12-month follow-up: 2 from the ECT group (1 due to cognitive adverse effects at week 4, 1 lost to follow-up at 12 months) and 1 from the rTMS group (relocated abroad at 8 months). The CONSORT diagram is presented in Figure 1.

Figure 1.

CONSORT Flow Diagram

CONSORT Flow Diagram

Baseline sociodemographic and clinical characteristics were comparable between groups (Table2 1 and 2). The sample had a mean age of 34.4 ± 7.3 years, was predominantly female (75%), and married (62.5%). Mean duration of illness was 8.4 ± 4.2 years, with an average of 2.8 ± 0.9 failed antidepressant trials. Three participants in the ECT group and two in the rTMS group had psychotic features.

Table 1.

Sociodemographic Characteristics of Participants (N = 40)

Characteristic	Total (N=40)	ECT (n=20)	rTMS (n=20)	p-value
Age (years), mean ± SD	34.4 ± 7.3	34.8 ± 7.5	34.0 ± 7.2	0.73
Sex, n (%)				0.74
Male	10 (25.0)	6 (30.0)	4 (20.0)
Female	30 (75.0)	14 (70.0)	16 (80.0)
Marital status, n (%)				0.60
Single	12 (30.0)	5 (25.0)	7 (35.0)
Married	25 (62.5)	14 (70.0)	11 (55.0)
Divorced/Widowed	3 (7.5)	1 (5.0)	2 (10.0)
Education, n (%)				0.81
None/Primary	8 (20.0)	4 (20.0)	4 (20.0)
Secondary	18 (45.0)	10 (50.0)	8 (40.0)
Tertiary	14 (35.0)	6 (30.0)	8 (40.0)
Employment, n (%)				0.53
Employed	22 (55.0)	12 (60.0)	10 (50.0)
Unemployed	18 (45.0)	8 (40.0)	10 (50.0)

Table 2.

Baseline Clinical Characteristics

Characteristic	Total (N=40)	ECT (n=20)	rTMS (n=20)	p-value
Duration of illness (years), mean ± SD	8.4 ± 4.2	8.6 ± 4.5	8.2 ± 4.0	0.77
Number of failed antidepressants, mean ± SD	2.8 ± 0.9	2.9 ± 1.0	2.7 ± 0.8	0.49
ATHF score, mean ± SD	3.4 ± 0.6	3.5 ± 0.6	3.3 ± 0.5	0.26
Psychotic features, n (%)	5 (12.5)	3 (15.0)	2 (10.0)	1.00
Baseline HAM-D, mean ± SD	20.58 ± 3.12	20.60 ± 3.21	20.55 ± 3.05	0.96
Baseline MoCA, mean ± SD	26.8 ± 2.1	26.9 ± 2.0	26.7 ± 2.2	0.76

Abbreviations: ATHF = Antidepressant Treatment History Form; HAM-D = Hamilton Depression Rating Scale; MoCA = Montreal Cognitive Assessment.

Table 3 presents HAM-D scores at each assessment point. Both groups showed significant improvement over time. ECT demonstrated faster early reduction, with significantly lower scores at Week 2 (mean difference -1.00, 95% CI -1.58 to -0.42, p < 0.001) and Week 4 (mean difference -1.95, 95% CI -2.81 to - 1.09, p < 0.001). By Week 6, scores were nearly identical between groups (mean difference -0.10, 95% CI -1.54 to 1.34, p = 0.89).

Table 3.

HAM-D Scores Across Time Points (ITT Analysis with LOCF)

Time Point	ECT Group (n=20) Mean ± SD [95% CI]	rTMS Group (n=20) Mean ± SD [95% CI	Mean Difference (ECT – rTMS) [95% CI]	p-value	Effect Size (ηp2) [95% CI]
Baseline	20.60 ± 3.21 [19.10– 22.10]	20.55 ± 3.05 [19.12– 21.98]	0.05 [-1.97 to 2.07]	0.96	–
Week 2	15.25 ± 2.64 [14.01– 16.49]	16.25 ± 2.79 [14.94– 17.56]	-1.00 [-2.73 to 0.73]	<0.001*	0.42 [0.18– 0.62]
Week 4	10.30 ± 2.84 [8.97– 11.63]	12.25 ± 2.99 [10.85– 13.65]	-1.95 [-3.82 to -0.08]	<0.001*	0.58 [0.34– 0.74]
Week 6	4.30 ± 2.23 [3.26–5.34]	4.40 ± 2.35 [3.30–5.50]	-0.10 [-1.54 to 1.34]	0.89	0.01 [0.00– 0.08]
6 Months	6.60 ± 2.97 [5.21–7.99]	4.05 ± 2.11 [3.06–5.04]	2.55 [0.92 to 4.18]	<0.001*	0.71 [0.48– 0.83]
12 Months	10.55 ± 3.45 [8.94– 12.16]	5.25 ± 2.89 [3.90–6.60]	5.30 [3.23 to 7.37]	<0.001*	0.83 [0.64– 0.91]

* Statistically significant at p < 0.05 (Bonferroni-corrected threshold for pairwise comparisons: p < 0.008).

At follow-up, rTMS showed superior long-term control: 6 months (mean difference 2.55, 95% CI 1.29 to 3.81, p < 0.001); 12 months (mean difference 5.30, 95% CI 3.54 to 7.06, p < 0.001). Figure 2 illustrates these trends.

As shown in the full ANOVA tables (Table 4b), Mauchly's test indicated violation of sphericity (W = 0.112, p < 0.001); therefore, Greenhouse-Geisser corrected values are reported (ε = 0.482). There was a significant main effect of time (F(2.41, 91.58) = 245.67, p < 0.001, ηp²= 0.866, 95% CI [0.82–0.90]) and a significant group × time interaction (F(2.41, 91.58) = 94.32, p < 0.001, ηp²= 0.713, 95% CI [0.62–0.78]). The between-subjects effect of group was not significant (F(1, 38) = 2.84, p = 0.10, ηp²= 0.07).

Table 4b.

Full Repeated-Measures ANOVA (Greenhouse–Geisser Corrected)

Outcome	Time Point	ECT (n=20) n (%)	rTMS (n=20) n (%)	p-value	OR [95% CI]
Response (≥50% reduction)	Week 6	19 (95)	18 (90)	1.00	2.11 [0.18–25.4]
	6 Months	15 (75)	19 (95)	0.18	0.16 [0.02–1.53]
	12 Months	6 (30)	15 (75)	0.004*	0.14 [0.04–0.55]
Remission (HAM-D ≤7)	Week 6	17 (85)	16 (80)	1.00	1.42 [0.30–6.73]
	6 Months	11 (55)	17 (85)	0.04*	0.22 [0.05–0.92]
	12 Months	4 (20)	13 (65)	0.004*	0.14 [0.03–0.57]

Response and remission rates are shown in Table 4a. At Week 6, both groups achieved high response rates (ECT: 95%, rTMS: 90%; p = 1.00) and remission rates (ECT: 85%, rTMS: 80%; p = 1.00). However, at 12 months, the rTMS group maintained significantly higher response (75% vs. 30%, p = 0.004) and remission rates (65% vs. 20%, p = 0.004).

Table 4a.

Response and Remission Rates

Source	SS	df	MS	F	p-value	ηp2[95% CI]
Within-subjects effects
Time	6245.8	2.41	2591.6	245.67	<0.001	0.866 [0.82–0.90]
Time × Group	158.4	2.41	65.7	94.32	<0.001	0.713 [0.62–0.78]
Error (Time)	403.8	91.58	4.41
Between-subjects effects
Group	28.9	1	28.9	2.84	0.10	0.07 [0.00–0.22]
Error	387.2	38	10.2

* Statistically significant at p < 0.05.

Adverse events were more frequent and severe in the ECT group (Table 5). Cognitive impairment, assessed by MoCA, showed significant decline in the ECT group at Week 6 (mean change: -2.3 ± 1.8 points, p = 0.02) with partial recovery by 6 months (mean change from baseline: -1.1 ± 1.5 points, p = 0.18). No significant cognitive changes occurred in the rTMS group. One participant in the ECT group experienced a prolonged seizure (>120 seconds) requiring intervention; no serious adverse events occurred in the rTMS group.

Table 5.

Adverse Events

Adverse Event	ECT (n=20) n (%)	rTMS (n=20) n (%)	p-value
Cognitive impairment (subjective)	12 (60)	0 (0)	<0.001
Headache	8 (40)	5 (25)	0.50
Muscle aches	7 (35)	0 (0)	0.008
Nausea	5 (25)	0 (0)	0.047
Scalp discomfort	0 (0)	6 (30)	0.02
Facial twitching	0 (0)	3 (15)	0.23
Prolonged seizure	1 (5)	0 (0)	1.00
Discontinuation due to AE	1 (5)	0 (0)	1.00
Any adverse event	18 (90)	10 (50)	0.008

Table 6 summarises treatments received during the follow-up period. No significant differences were observed between groups in antidepressant use, psychotherapy, or relapse rates.

Table 6.

Concomitant Treatments During 12-Month Follow-up

Treatment	ECT (n=20)	rTMS (n=20)	p-value
Antidepressant use, n (%)	18 (90)	17 (85)	1.00
Antidepressant class, n (%)			0.72
– SSRI	12 (60)	10 (50)
– SNRI	4 (20)	5 (25)
– Other	2 (10)	2 (10)
Psychotherapy sessions, mean ± SD	3.2 ± 2.8	3.5 ± 3.1	0.75
Relapse (by 12 months), n (%)	8 (40)	4 (20)	0.16

Mean HAM-D Score Over Time

Figure 2.

Discussion

This randomized controlled trial comparing ECT and rTMS for treatment-resistant depression in a Nigerian tertiary setting yielded several important findings. First, both interventions were effective in reducing depressive symptoms, consistent with established literature [5,6,10-12]. Second, ECT produced significantly faster early response, with greater improvement by Week 4, confirming its role as a rapid-acting intervention for acute severe depression [4,5,17]. Third, and notably, rTMS demonstrated superior long-term outcomes, with significantly lower HAM-D scores and higher remission rates at 6 and 12 months, despite equivalent acute-phase efficacy at Week 6. Fourth, adverse events were more frequent and clinically significant in the ECT group, particularly cognitive impairment.

The faster early response to ECT observed in our study aligns with previous research [6,7,18]. Kellner et al. reported significant improvement after 2-3 ECT sessions in severely depressed patients [6], and Prudic et al. documented rapid symptom reduction in TRD populations [17]. This rapid onset makes ECT particularly valuable in urgent clinical situations such as high suicide risk, severe malnutrition, or catatonia [3,4,19]. In our Nigerian context, where access to intensive psychiatric care may be limited, ECT remains an essential tool for acute symptom management.

By Week 6, both groups achieved nearly identical HAM-D scores (4.30 vs. 4.40), with high response (95% vs. 90%) and remission rates (85% vs. 80%). This finding is clinically significant, demonstrating that a full course of rTMS (30 sessions over 6 weeks) can achieve acute outcomes comparable to ECT in non-psychotic TRD. Our results extend previous observations by Fitzgerald et al. [12] and George et al. [11] that adequate rTMS dosing produces robust antidepressant effects. The slightly longer time to response with rTMS (4-6 weeks versus 2-3 weeks for ECT) should be considered in treatment planning but does not compromise ultimate efficacy.

The most striking finding was the divergence in outcomes during follow-up. While ECT recipients showed gradual relapse, particularly by 12 months (remission rate 20%), rTMS recipients maintained their improvement (remission rate 65%). This pattern has been suggested in some previous studies [15,16,18] but not consistently demonstrated with 12-month follow-up. Several explanations are plausible:

First, the non-invasive nature of rTMS may allow for better preservation of cognitive function, which could support sustained engagement with psychosocial interventions and daily activities [7,9]. Second, the absence of anaesthesia-related effects may facilitate a more rapid return to normal functioning [8,20]. Third, the distributed schedule of rTMS (daily over 6 weeks) versus the compressed schedule of ECT (twice weekly over 3 weeks) might promote different patterns of neuroplasticity, potentially leading to more durable changes [13,21].

The relapse observed in the ECT group is consistent with historical data [18,20]. Without maintenance ECT, which was not provided in this study, relapse rates of 50-80% within 6-12 months are commonly reported [17,20]. Our findings underscore the critical importance of maintenance strategies after acute ECT, particularly in TRD populations.

Our adverse event data confirm the superior tolerability of rTMS. The 60% rate of subjective cognitive impairment in the ECT group, with objective decline on MoCA, is concerning and consistent with extensive literature on ECT-related cognitive effects [8,9,22]. Although these effects were largely transient (recovering by 6 months), they represent a significant burden during the acute treatment phase and may influence patient acceptability and adherence.

In contrast, rTMS was associated with only mild, transient side effects (scalp discomfort, headache) and no cognitive impairment. This safety profile, combined with equivalent acute efficacy and superior long-term outcomes, positions rTMS as an attractive option for patients concerned about cognitive effects or those who cannot undergo anaesthesia.

Implications for Clinical Practice In Lmics

Our findings have several implications for mental health service delivery in low- and middle-income countries like Nigeria:

• 1.ECT remains essential: For patients requiring rapid response due to severity or urgency, ECT is irreplaceable. Facilities offering ECT must be maintained and supported.
• 2.rTMS offers a sustainable alternative: For the majority of TRD patients who do not require immediate response, rTMS provides equivalent acute efficacy with better long-term outcomes and fewer side effects. The outpatient-based delivery eliminates the need for anaesthesia and hospital admission, reducing costs and stigma.
• 3.Service planning considerations: While the initial cost of rTMS equipment is substantial, the absence of anaesthesia requirements, lower staffing needs, and outpatient delivery may ultimately prove cost-effective in high-volume settings. Health economic analyses are urgently needed.
• 4.Treatment algorithms: Our results support a stratified approach to TRD: ECT for acute, severe presentations requiring rapid response; rTMS as a first-line neurostimulation option for stable TRD outpatients.

Limitations

This study is not without some limitations. Some of those limitations that warrant consideration include:

• 1.Sample size: Although adequately powered for the primary analysis, the sample size (n=40) is modest, and subgroup analyses (e.g., by psychotic features) were not possible. The large effect sizes observed, while encouraging, may be inflated due to small sample bias.
• 2.Single-centre design: Conducted at a single Nigerian tertiary centre, generalisability to other settings within Nigeria or other LMICs may be limited.
• 3.Blinding limitations: While the outcome rater was blinded, participants and treating clinicians were necessarily unblinded, introducing potential performance bias.
• 4.No maintenance treatment: The absence of standardised maintenance protocols after acute treatment may have contributed to relapse, particularly in the ECT group. However, this reflects real-world practice in many LMIC settings and provides ecologically valid data on naturalistic outcomes.
• 5.Concomitant medications: All participants continued antidepressant medications during follow-up, which may have influenced outcomes. However, medication use was similar between groups and reflects clinical reality.
• 6.Demographic homogeneity: The predominance of female participants (75%) limits generalisability to male populations.
• 7.No sham control: The absence of a sham control group means we cannot rule out non-specific treatment effects. However, the significant group × time interaction supports specific treatment effects.
• 8.Cost-effectiveness not assessed: Without health economic analysis, we cannot determine whether the superior long-term outcomes of rTMS justify its initial cost.

Future Research Directions

Based on our findings, we recommend the following:

• 1.Multicentre trials: Larger studies across multiple Nigerian and African sites to enhance generalisability
• 2.Health economic analyses: Comprehensive cost-effectiveness studies comparing ECT and rTMS in LMIC settings
• 3.Mechanistic studies: Investigation of neurobiological mechanisms underlying differential long-term outcomes
• 4.Optimisation of protocols: Studies examining optimal rTMS maintenance schedules and predictors of durable response
• 5.Implementation research: Evaluation of barriers and facilitators to rTMS adoption in LMIC mental health systems
• 6.Comparative effectiveness: Head-to-head comparisons with other emerging neurostimulation modalities (e.g., theta burst stimulation)

Conclusion

Both ECT and rTMS are effective interventions for treatment-resistant depression in the Nigerian context. ECT produces faster early symptom relief and remains essential for acute, severe presentations. However, rTMS offers equivalent acute efficacy with superior long-term symptom control and a more favourable safety profile, particularly regarding cognitive function. These preliminary findings suggest that rTMS may be a compelling alternative for long-term management of TRD, especially in outpatient and resource-constrained settings. Larger multicentre studies with cost-effectiveness analyses are needed to inform evidence-based treatment guidelines and mental health policy in low- and middle-income countries.

Availability of Research Data

Data are available upon reasonable request from the corresponding author.

Funding

The authors received no funding or financial support.

Conflict of Interest

The authors declare no conflict of interest.

Acknowledgements

The authors sincerely thank all the participants who took part in the study, the staff and leadership of Ahmadu Bello University Teaching Hospital, Zaria, Nigeria, for their support and cooperation throughout the study. We also thank Dr. Fahad Abubakar Saulawa of the Department of Community Medicine, Ahmadu Bello University Teaching Hospital, Shika-Zaria, Nigeria for statistical consultation and Dr. Faridah Ali Hassan of the Department of Psychiatry, Ahmadu Bello University Teaching Hospital, Shika-Zaria, Nigeria for assistance with adverse event monitoring.