Repetitive Transcranial Magnetic Stimulation for Elderly Patients with Cognitive Impairment: A Meta Analysis of Randomized Controlled Trials

Article

Repetitive Transcranial Magnetic Stimulation for Elderly Patients with Cognitive Impairment: A Meta Analysis of Randomized Controlled Trials

ZhiGang Cao^*, Mi Yang

Department of Neurology, Sichuan Taikang Hospital, Chengdu, Sichuan, China

Objective: This meta-analysis of randomized controlled trials (RCTs) examines the effects of repetitive transcranial magnetic stimulation (rTMS) on cognitive function in elderly patients (aged 55 or older) and identifies potential influencing factors.

Methods: Searches of Embase, Web of Science, Cochrane Library, PubMed, CNKI, and Wanfang identified RCT studies evaluating the Impact of rTMS on cognition in older adults. The search spanned the period from the database's inception to 15 December 2024. Following the study identification and data extraction by two independent reviewers, the risk of bias was assessed with the Cochrane RoB tool. Data synthesis and subgroup analysis were performed via SMD, WMD, and 95% CIs to evaluate the effects of rTMS and its influencing factors. The review protocol was prospectively registered in PROSPERO (CRD42020187336)

Results: A total of 13 studies, including 14 trials with 655 patients, were analyzed. Compared to the control group, real rTMS significantly improved cognition in elderly patients (Global Cognitive Function: SMD, 0.56;95% CI 0.33,0.79; p<0.0001; MMSE: WMD,2.04;95% CI 1.54,2.53; p<0.0001; ADAS-Cog: WMD, -2.59;95% CI -3.63, -1.54; p<0.0001; MoCA: WMD,2.46; 95% CI 1.47, 3.45; p<0.0001). Subgroup analysis revealed that participants receiving rTMS combined with cognitive training demonstrated more significant improvements than those receiving rTMS alone (WMD: 0.51; 95% CI: 0.26,0.76; p < 0.0001). Additionally, other subgroup analyses—including rTMS stimulating sites (multiple vs. single sites), education years (≥10 vs. <10 years), and rTMS frequency (5/10/20 Hz vs. iTBS (intermittent theta-burst stimulation)—showed no significant differences (p > 0.05).

Conclusion: Both standalone rTMS and rTMS combined with cognitive training—which varied across studies and included both computerized paradigms (face-name associative memory training, mirror neuron system therapy, VR interactions) and individualized one-on-one training—demonstrated significant efficacy in improving global cognitive function in elderly patients with mild to moderate symptoms. As a safe and well-tolerated intervention, rTMS demonstrates considerable potential for reliable clinical application.

Introduction

With a rapidly ageing global population, globally, over 60 million people aged 50 years and above are living with cognitive impairment(Primary neurodegenerative cognitive disorders—such as Alzheimer's disease (AD), frontotemporal dementia (FTD), dementia with Lewy bodies (DLB), and Parkinson's disease dementia (PDD)), and the projection indicates a doubling every five years¹. The global economic burden of dementia is estimated at approximately US $1.3 trillion annually, with unpaid caregiving constituting the largest component of these costs². Common cognitive impairment risk factors include both non-modifiable and modifiable elements, such as advanced age, lifestyle choices (sedentary behaviours and poor diet), physical inactivity, and lower educational level attainment³. Among these, ageing is a major contributor. Specifically, as age increases, the brain undergoes progressive structural and functional changes, including decreased synaptic plasticity in memory-related neural circuits and volumetric atrophy of the hippocampus⁴. Additionally, long-term epidemiological evidence shows that age-related cerebrovascular and cardiovascular diseases, such as stroke and ischemic heart disease, are closely connected to a higher risk of cognitive decline⁵. Cognitive impairment has effects that extend beyond individual patients, resulting in significant socioeconomic impacts and burdens⁶. In elderly populations, impaired cognition signals a critical shift from normal age-related decline to dementia. Without intervention, this condition often progresses to dementia, characterized by worsening memory, deficits in executive function, and a decline in language and problem-solving skills⁷. Early detection, diagnosis, and treatment are the most cost-effective cognitive impairment management methods⁸. Pharmacological interventions remain the cornerstone of clinical management for cognitive impairment. Current therapeutic approaches predominantly feature cholinesterase blockers (e.g., donepezil) and NMDA receptor inhibitors (memantine), calcium channel blockers (nilvadipine), and β-secretase inhibitors, which are used to reduce cognitive decline³. However, current evidence shows limited or negligible therapeutic efficacy of these agents, and their long-term safety profiles during extended treatment regimens remain poorly understood and defined⁹^,¹⁰.

As an emerging non-invasive neuromodulation modality, TMS holds significant promise in managing age-related cognitive disorders, especially for cases refractory to conventional drug therapies. Its targeted approach and excellent safety margin offer a new direction for addressing this growing clinical challenge¹¹. Cognition depends on the dynamic interplay of large-scale networks—default-mode, salience, and central-executive systems—that show aberrant connectivity in cognitive impairment. rTMS can be precisely delivered to key network nodes, such as the dorsolateral prefrontal cortex. The induced effects then propagate along white-matter tracts to distal hubs (e.g., the hippocampus and parietal cortex), normalizing network excitability, strengthening inter-network coupling, and increasing information transfer efficiency¹². At the neurochemical level, rTMS acutely modulates glutamate release and N-methyl-D-aspartate receptor function, while attenuating γ-aminobutyric acid–mediated inhibition, thereby restoring the excitatory–inhibitory balance required for accurate encoding and filtering of information¹³. Rodent studies further suggest that rTMS may accelerate β-amyloid clearance and reduce tau hyperphosphorylation, presumably by enhancing activity-dependent glymphatic flow and microglial phagocytosis¹⁴.

TMS works through Faraday's Law of Electromagnetic Induction. Specifically, a high-intensity electric current (around 8000 A) of short duration (100–300 μs) flows through a stimulation coil, generating a magnetic pulse field (ranging from 1 to 5 Tesla) in the vicinity. This time-varying magnetic field induces currents within conductive biological tissues, such as neuronal membranes, through electromagnetic processes coupling¹⁵. The interventions trigger synaptic activation, producing excitatory or inhibitory effects in the cerebral cortex. This localized neuromodulation then spreads to neighboring cortical and subcortical areas through long-range connections, thereby enhancing the efficacy of network-wide neural activity¹⁶^,¹⁷. rTMS enables the modification of large brain networks, thereby improving their function in patients with specific network alterations or certain conditions. Therefore, applying rTMS to modulate neural activity in targeted brain regions of older adults with cognitive impairment may reorganize dysfunctional network interactions and restore impaired functional connectivity, thereby enhancing cognitive function performance¹⁸^,¹⁹^,²⁰. Numerous published studies have shown that rTMS treatment has positive effects on cognitive function in elderly patients with cognitive impairment, including clinical behavioral aspects, functional connectivity, and cortical activity excitability²¹^,²²^,¹². Despite accumulating evidence supporting the therapeutic potential of rTMS in cognitive impairment, recent studies have raised concerns about its clinical effectiveness and the significance of translational efforts, including the non-persistent efficacy, heterogeneous responses among disease subtypes (AD, MCI) and individuals, non-standardized stimulation parameters, and a lack of personalized protocols for distinct etiologies and disease progression stages. These uncertainties mainly arise from methodological limitations in RCTs, such as inadequate sample sizes, incomplete behavioural assessments, and a lack of neuroimaging or biomarker-based outcome measures²³^,²⁴^,²⁵.

Notable limitations of current research include: existing studies on cognitive impairment mainly focus on specific cognitive functions (episodic memory, executive function), with little attention given to comprehensive assessments of overall cognitive outcomes. For example, Marra et al. (2015) solely examined the Impact of rTMS on the daily memory abilities of seniors with mild cognitive decline, while Wang et al. (2021) limited their analysis to rTMS-induced improvements in executive functions²⁶^,²⁷. This narrow focus highlights the importance of standardized, multi-faceted cognitive assessments to measure the widespread impacts of neuromodulatory interventions. Building on the mechanism by which rTMS modulates large-scale brain network connectivity, future clinical trials should prioritize the following directions: First, elucidate the specificity of network modulation. The effects of rTMS are target-dependent. Stimulation of the left DLPFC may preferentially improve executive functions and working memory²⁸, whereas modulation of the precuneus (a key DMN node) might enhance episodic memory¹¹. A comprehensive cognitive assessment is required to map these specific brain-behavior relationships. Second, assess translational value and ecological validity. Since the goal is to improve daily life, outcomes must extend beyond laboratory tests. Integrating patient and caregiver reports on functional abilities, along with performance-based measures that simulate real-world tasks, is crucial for demonstrating clinically meaningful benefits. Third, discover biomarkers for treatment individualization. multidimensional data (behavioral, neuroimaging, EEG, and genetic). This will help answer the critical question: which baseline characteristics predict optimal response to a given rTMS protocol? The answer is key to developing precision rTMS therapies. The high variability in study populations, which often includes different diagnostic groups. For example, Pagali et al. (2023) included patients with AD, MCI, schizophrenia, depression, bipolar disorder, and other neurological conditions in their rTMS trials. This broad inclusion criterion introduces confounding variability, thereby reducing the clinical specificity and translational relevance of reported therapeutic effects²⁹. Furthermore, the effectiveness of rTMS greatly relies on protocol parameters such as stimulation frequency, target sites, and treatment duration. Conducting systematic studies of these parameter-dependent effects is essential for optimizing rTMS protocols and advancing personalized treatment approaches³⁰.

Although recent meta-analyses have examined the therapeutic effectiveness of rTMS for cognitive disorders, these studies mainly focus on AD, MCI, and post-stroke cognitive impairment (PSCI)³¹^,³²^,³³. The review by Calvin Pak Wing Cheng and colleagues (2017) identified the cognitive benefits of rTMS in cognitively impaired elderly populations but also highlighted critical methodological shortcomings: ①Limited Quantity and Quality of Included Studies: The analysis relied on a relatively small pool of RCTs, resulting in compromised statistical power and increased vulnerability to the influence of any single study's results. ②Significant Clinical and Methodological Heterogeneity: The inclusion of patients with different etiologies (AD, MCI, vascular) and disease severities, alongside substantial variations in rTMS protocols (frequency, target), obscured the identification of optimal patient populations and stimulation parameters. ③Incomplete Outcome Assessment and Potential for Publication Bias: The scope of assessed cognitive outcomes may have been narrow, and the analysis was potentially susceptible to publication bias, thereby limiting the comprehensiveness and reliability of the findings. This study systematically evaluates the efficacy of rTMS for cognitive impairment in adults aged 55 and older by synthesizing data from recent, large-scale RCTs³⁴^,³⁵^,³⁶^,³⁶. We employed rigorous inclusion criteria to define a specific patient population, thereby enabling a targeted analysis. The incorporation of a substantial number of high-quality RCTs has established a robust evidence base, enhancing the reliability of the overall treatment effect estimate. Our analysis aims to optimize rTMS treatment protocols for this demographic. Furthermore, pre-specified subgroup analyses are designed to clarify the impact of key variables—including stimulation parameters, cortical targets, patient education, and cognitive training. The interaction between rTMS and cognitive training appears to be influenced by the complex dynamics of brain network organization, which involve fundamental properties such as topological configuration and oscillatory synchronization. These insights aim to inform personalized rTMS strategies and advance precision medicine³⁷. Finally, we will explore the clinical mechanisms of rTMS and contrast them with established pharmacological pathways to guide standardized clinical practice.

Methods

Eligibility Criteria

Inclusion criteria were met if the studies involved: (1) individuals aged 55 or older patients with a diagnosis of MCI, AD, or PD-related cognitive impairment^38,39; (2) The group receiving rTMS; (3) The control group received sham rTMS (inactive or weak stimulation); (4) The results of cognitive outcome measures from the Mini-Mental State Examination (MMSE), the Alzheimer's Disease Assessment Scale-Cognitive Subscale (ADAS-cog), and other cognitive function assessment methods (such as mood and behavior) were reported or provided upon request; (5) Publications in English or Chinese.

Exclusion Criteria

Studies will be excluded for these reasons: (1) not randomized controlled trials (RCTs), (2) not reporting cognitive outcomes, (3) original data is unavailable, (4) conference abstracts, editorials, reviews, meta-analyses, theses/dissertations, case reports/series, or presenting duplicate data.

Search Strategy

Our meta-analysis followed PRISMA standards for systematic reviews⁴⁰ (Figure 1) and is registered with PROSPERO (registration number CRD42020187336), ensuring transparency and research compliance. We conducted a comprehensive search for clinical RCTs on rTMS treatment for cognitive defects across six databases: Embase, Web of Science, Cochrane Library, PubMed, China National Knowledge Infrastructure (CNKI), and Wanfang Database from their inception to 15 December 2024. The search criteria included MeSH terms, free-text keywords, and their equivalents, such as transcranial magnetic stimulation, cognitive dysfunction, and randomized controlled trial. Supplementary Material 1 outlines the search terms and the search strategies. EndNote X9 is used to manage the retrieved literature.

Data Extraction

EndNote (version X9) was used to manage the retrieved literature, and two reviewers removed irrelevant articles and selected articles of interest by reading titles and abstracts. Two reviewers, FJJ and YM, carefully evaluated the studies against the predetermined inclusion criteria and then selected those that met the criteria. After thoroughly reviewing the full texts, the chosen articles were compiled and examined to reach a final conclusion. Disagreements were resolved through discussion or by seeking input from a third reviewer, CZG. The data extracted from all included studies were independently organized into a pre-designed spreadsheet. Extracted data included (1) sample size; (2) study characteristics (such as study design, population, intervention duration, group configuration, TMS parameters); (3) outcome measures; (4) adverse effects. When available, the MMSE or ADAS-cog was selected as the cognitive outcome measure; otherwise, the primary outcome measure was used. Where accessible data proved insufficient, primary researchers were consulted for supplemental information. We included RCTs with parallel group or cross-over designs, analyzing only data from the first period of cross-over trials to prevent carry-over effects. The studies included compared sham treatments with active controls and assessed changes in overall cognitive function or specific cognitive areas.

Quality Assessment

The two reviewers assessed the methodological rigor of the trials included in the meta-analysis, drawing on the six key areas outlined in version 5.1.0 of the Cochrane Handbook for Systematic Reviews of Interventions' risk of bias tool. In cases of disagreement, we actively involved a third investigator (C-ZG) to secure a clear consensus. The assessment of bias risk and evidence quality is categorized as low, unclear, or high, following the standards outlined in the Cochrane Handbook. The evaluation screened for (1) selection bias, which was assessed based on the adequacy of randomization and allocation concealment methods; (2) execution bias, including blinding of participants, investigators, and assessors, evaluated by the quality of the blinding method. (3) Follow-up bias: assessed by outcome data completeness and missing data handling; (4) Publication bias: evaluated by selective outcome reporting, and (5) any other bias. The level of bias risk for each category was classified as high, low, or indeterminate.

Evidence Quality Assessment

Two reviewers independently used the GRADE (Grading of Recommendations Assessment, Development, and Evaluation) tool to assess the quality of evidence⁴¹. It considers five aspects: study limitations, result inconsistency, evidence indirectness, imprecision, and report bias. We classify the quality of evidence into four levels: "High," "Moderate," "Low," and "Very low."

Statistical Analysis

Data synthesis and analysis

The procedures for meta-analysis were conducted using Review Manager 5.3 software (Cochrane Collaboration, Oxford, UK). The effect of TMS on cognitive performance was assessed by comparing the average change in cognitive metrics—specifically, the mean difference (MD) between pre- and post-treatment scores—across both the experimental and control groups. Considering that a study reported multiple effect sizes from the same patient group, such as the MMSE, ADAS cog, MoCA, and CDR (Clinical Dementia Rating), the aggregated Impact of pertinent studies was assessed using SMD, presented with 95% CI. SMD facilitates comparison of inter-group mean variations across diverse measurement scales in meta-analytic studies. Data from multimodal biomarkers such as EEG and functional connectivity were excluded from the quantitative synthesis. This decision was based on the scarce data availability and the fact that the reported results were presented solely in graphical form, which was deemed inappropriate for robust statistical analysis due to the absence of standardized numerical metrics. For the studies in which the original data were displayed as mean ± SEs, SD could be calculated using the following formula: SEs = SD/√n (n indicated the number of participants), Then, input these data in Review Manager 5.3 for the subsequent meta-analysis. To assess the Impact of different grouping methods on outcome changes and result stability, we analyzed outcomes for individual scales. We used Weighted Mean Difference (WMD) to calculate effect sizes and estimate treatment effects, applying 95% CI. We assessed heterogeneity using Cochran's Q statistic and the I² test. Employed a fixed-effects analysis for I² < 50%, and a random-effects model for I² ≥ 50%.

Publication bias and sensitivity analysis

We checked for publication bias by creating funnel plots and using Egger's regression tests with Stata 17.0 version (Stata Corp, College Station, TX, USA) for outcomes that included 10 or more studies. We considered a p-value of less than 0.05 significant when comparing the outcome variables. The robustness of the meta-analysis findings was assessed through sensitivity analyses. This involved sequentially excluding each study (the 'leave-one-out' method) and conducting an analysis excluding studies deemed to have a high risk of bias.

Subgroup analyses

To explore other possible factors affecting the recovery of cognitive function, we conducted four main subgroup analyses of rTMS frequency (5-20 Hz vs. iTBS), rTMS stimulation sites (single site vs. multiple sites), education, years (10 or fewer vs. more than 10), and rTMS combined with cognitive training versus rTMS alone. The subgroup analysis for education used a 10-year cutoff, chosen because it aligns with epidemiological classifications, marks the completion of secondary education as a cognitive reserve indicator, and offers a balanced distribution for reliable analysis.

Results

Study Selection

The search across six databases revealed 1495 studies. After eliminating 520 duplicates, 975 distinct studies were retained for analysis. After reviewing the titles and abstracts, we excluded 804 studies for various reasons, including those not related to the topic (n = 712), reviews, case reports, meta-analyses (n = 33), studies not written in English or Chinese (n = 2), and conference abstracts (n = 57). After thoroughly reviewing the full texts, 13 studies, numbered 1 through 13, were ultimately selected for inclusion in the meta-analysis and review. Figure 1 displays the study selection methodology.

Figure 1: PRISMA

Flow diagram showing the search and selection procedure that was used for this meta-analysis. Diagram adapted from PAGE M J et al. (2021). Abbreviations: AD = Alzheimer’s disease; PD = Parkinson's disease; CI = mild cognitive impairment; rTMS = repetition transcranial magnetic stimulation

Study Characteristics

The 13 RCT studies included 14 trials involving a total of 655 patients with cognitive impairments, with an average age of 70.84 years, and 56% of the participants were women. Among these studies, 12 studies involved patients with AD, while only one involved patient with MCI. Among the included studies, the longest rTMS intervention duration was 24 weeks, while the shortest was 2 weeks. Of these 13 studies, six studies (1, 3, 4, 7, 8, 10) used a frequency of 20 Hz, two studies (11, 13) used 10 Hz, two studies (5, 9) used 5 Hz, and three studies (2, 6, 12) used iTBS. Eight studies used single-site rTMS (DLPFC, left temporoparietal, precuneus, left lateral parietal) (1, 3, 4, 5, 7, 8, 11, 12), while the other five used multi-site rTMS (cerebellum, parietal/temporal, and left DLPFC). In four studies, no concomitant psychotropic medication was used (4, 5, 7, 8). In nine studies, participants took cognition-enhancing drugs, including Donepezil hydrochloride, dementia medications, Memantine, and cholinesterase inhibitors. Overall cognitive function was evaluated using scales such as MoCA, MMSE, ADAS-Cog, and CDR. The individual study characteristics are shown in Table 1.

Table 1: Characteristics of the 13 included studies, which encompassed 14 trials

Study Name	Sample Size (Real/Sham)	Mean age (years) (Real /Sham)	Gender (M/F)	Disease type	Education (years) (Real /Sham)	Basic features of rTMS	Treatment duration	Cognitive Train	Drug, intervention	Adverse effects	Main cognitive outcome measure
1. Dong et al. (2023)	(38/38)	(79.8/81.2)	(28/48)	AD (mild)	(11.3/10.7)	20Hz, Left DLPFC	6weeks	Yes	Donepezil hydrochloride	No	MMSE; ADAS-COG;
2. Jiang et al. (2022)	(26/26)	(68.1/70.3)	(29/23)	AD (mild; moderate)	(5.3/4.6)	50Hz(iTBS), Bilateral DLPFC	4weeks	Yes	Donepezil hydrochloride	No	MMSE; MoCA
3. Xia et al. (2023)	(49/51)	(70.4/70.9)	(37/63)	AD (mild; moderate)	(10.1/10.3)	20Hz, Left DLPFC\Left temporoparietal	4weeks	Yes	Donepezil; Olanzapine	Headache; Head discomfort; tinnitus	MMSE; ADAS-COG
3. Xia et al. (2023)	(49/50)	(69.3/70.9)	(39/61)	AD (mild; moderate)	(10.2/10.3)	20Hz, Left DLPFC\Left temporoparietal	4weeks	Yes	Donepezil; Olanzapine	Headache; Head discomfort; tinnitus	MMSE; ADAS-COG
4. Chiara et al. (2020)	(27/23)	(73.5/73.3)	(29/21)	AD (mild; moderate)	(8.8/7.9)	20Hz, Left DLPFC	4weeks	Yes	No	No	MMSE; Associative memory
5.Roque GY et al. (2021)	(12/12)	(66.1/67.2)	(10/14)	MCI	(15.2/14.7)	5Hz, Left DLPFC	10weeks	Yes	No	Headache	MMSE; MoCA
6.Hoy et al. (2023)	(29/27)	(75/75.8)	(38/18)	AD (mild; moderate)	(12.7/13)	50Hz(iTBS), Bilateral DLPFC/PPC	6weeks	Yes	Donepezil；Memantine	Headache; Head discomfort	ADAS-COG; Resting State-EEG
7.Koch et al. (2022)	(25/25)	(75/72.3)	(24/26)	AD (mild; moderate)	(10.2/8.6)	20Hz, Precuneus	24weeks	No	No	No	MMSE; ADAS-COG; CDR
8.Mencarelli et al. (2024)	(8/8)	(68.5/70.6)	(10/6)	AD (mild; moderate)	(12.4/10.1)	20Hz, Precuneus	24weeks	No	No	No	MMSE; ADAS-COG; CDR
9.Yao et al. (2022)	(15/12)	(63.8/67.6)	(14/13)	AD (mild; moderate)	(10.5/9.4)	5Hz, Bilateral cerebellum	4weeks	NR	Memantine	No	MMSE; ADAS-COG; Moca; Functional connectivity
10.Zhao et al. (2017)	(17/13)	(69.3/71.4)	(14/16)	AD (mild; moderate)	(4.8/4.9)	20Hz, Bilateral Parietal/Temporal	6weeks	No	Cholinesterase Donepezil	No	MMSE; ADAS-COG; MoCA;
11.Wei et al. (2022)	(29/27)	(70/71.6)	(16/40)	AD	(7.3/6.6)	10Hz, Left lateral parietal	2weeks	No	cholinesterase	Transient fatigue	MMSE; CDR
12.Wu et al. (2024)	(20/22)	(66.8/65.3)	(13/29)	AD	(10.2/10)	50Hz(iTBS), Left DLPFC	8weeks	NR	cholinesterase inhibitors	uncomfortable scalp	MMSE; MoCA; CDR
13. Zhang et al. (2019)	(15/13)	(69/68.5)	(6/22)	AD (mild; moderate)	(12.4/11.5)	10Hz, Left DLPFC/LTL	4weeks	Yes	Sertraline Memantine	slight tingling in the scalp	MMSE; ADAS-COG

AD, Alzheimer’s disease, MCI, mild cognitive impairment, rTMS, repetitive transcranial magnetic stimulation, iTBS, intermittent theta-burst stimulation, Real rTMS group, Sham sham group, MMSE Mini-Mental State Examination, ADAS-cog Alzheimer Disease Assessment Scale-cognitive subsection, CDR Clinical Dementia Rating, L left, R right, DLPFC dorsolateral prefrontal cortex, LTL lateral temporal lobe, NR Not Reported, M, man, F, female

Adverse Events

Of the 13 included studies, five patients in the six project studies (3, 5, 6, 11, 12, 13) found adverse effects related to rTMS, including headache, head discomfort, and slight tingling, while seven studies reported no adverse effects. One had tinnitus that disappeared after withdrawing from the trial, and one had transient fatigue symptoms, with no reports of seizures or epilepsy. The side effects caused by TMS were brief and resolved quickly after the procedure, TMS is considered safe and effective²⁹.

Risk of Bias Assessment

Bias potential in the incorporated Research was evaluated according to the standards detailed in the Cochrane Handbook. Figure 2 summarizes the quality assessment. Of the 13 studies reviewed, 8 were rated as low risk, 4 were unclear, and 1 was categorized as high risk. All 13 studies used a random allocation design and specified the randomization method, including proper sequence randomization with a random number generator or table. Nine studies described sufficiently concealed allocation methods. Nine papers were blinded to the study subjects or the interventionists and also blinded to the outcome measures, while other studies had unclear risk assessments. Zhang et al. (2019) were judged to have a high risk due to incomplete outcome data and selective reporting, while other risks of bias were considered low. Additionally, 12 studies reported full results without selective outcome reporting and had minimal risk of bias.

Evidence quality evaluation results

The quality of evidence was assessed using the GRADE approach for five outcomes across the 13 included studies (Table 2). The GRADE assessment results demonstrated that among the outcomes, one (11.11%) was rated as high-quality evidence, five (55.56%) as moderate, two (22.22%) as low, and one (11.11%) as very low. Overall quality of evidence was predominantly moderate. Several limitations may contribute to the low-quality evidence rating for some outcomes. First, the risk of bias introduced by methodological shortcomings (e.g., in randomization or blinding) is a key concern. Second, the findings are limited by the imprecision of the estimates, which stems from small sample sizes and inconsistencies across studies. Third, the significant statistical heterogeneity observed compromises the reliability of the pooled results.

Table 2: Results of evidence quality

Certainty assessment							№ of patients		Effect		Certainty	Importance
№ of studies	Study design	Risk of bias	Inconsistency	Indirectness	Imprecision	Other considerations	[Real]	[Sham]	Relative (95% CI)	Absolute (95% CI)	Certainty	Importance
Global Cognitive Function
13	randomised trials	not serious	not serious	not serious	serious^a	none	359	347	-	SMD 0.56 SD higher (0.33 higher to 0.79 higher)	⨁⨁⨁◯ Moderate^a
MMSE
13	randomised trials	not serious	not serious	not serious	serious^a	none	330	320	-	MD 2.04 SD higher (1.54 higher to 2.53 higher)	⨁⨁⨁◯ Moderate^a
ADAS-COG
8	randomised trials	not serious	not serious	not serious	serious^a	none	231	225	-	MD 2.59 lower (3.63 lower to 1.54 lower)	⨁⨁⨁◯ Moderate^a
MoCA
5	randomised trials	not serious	not serious	not serious	very serious^a,b	none	89	84	-	MD 2.46 higher (1.47 higher to 3.45 higher)	⨁⨁◯◯ Low^a,b
CDR
4	randomised trials	not serious	serious^c	not serious	very serious^a,b	none	82	82	-	MD 0.41 lower (0.94 lower to 0.12 higher)	⨁◯◯◯ Very low^a,b,c
rTMS frequency (MMSE,ADAS-COG)
13	randomised trials	not serious	not serious	not serious	serious^a	none	359	347	-	SMD 0.52 SD higher (0.27 higher to 0.78 higher)	⨁⨁⨁◯ Moderate^a
rTMS sites (MMSE,ADAS-COG)
13	randomised trials	not serious	not serious	not serious	not serious	none	257	256	-	SMD 0.55 SD higher (0.4 higher to 0.71 higher)	⨁⨁⨁⨁ High
Cognitive train (MMSE,ADAS-COG)
13	randomised trials	not serious	serious^c	not serious	serious^a	none	324	313	-	SMD 0.51 SD higher (0.26 higher to 0.76 higher)	⨁⨁◯◯ Low^a,c
Education (MMSE,ADAS-COG)
13	randomised trials	not serious	not serious	not serious	serious^a	none	359	347	-	SMD 0.56 SD higher (0.33 higher to 0.79 higher)	⨁⨁⨁◯ Moderate^a

CI: confidence interval; MD: mean difference; SMD: standardised mean difference

Explanations

The research analysis has a relatively wide confidence interval
The sample size included in the study was less than 400
The heterogeneity of the research is greater than 50%

Figure 2: The risk of bias of included studies based on the Cochrane’s handbook

Effects of rTMS on Cognitive Impairment

Global Cognitive Function

A total of 13 studies and 14 trials involving 655 patients were selected to assess the combined effects of rTMS on patients' cognitive function. The pooled results showed that rTMS could significantly enhance cognitive functions (SMD, 0.56; 95% CI 0.33, 0.79; p < 0.0001) in the random-effects model analysis (P = 0.01, I² = 53.0%) In terms of the absolute difference, this corresponds to a mean improvement of 2.04 points (95% CI: 1.54 – 2.53) on the MMSE (Figure 3A). The funnel plot's visual inspection suggested a minor inclination toward publication bias (Figure 4A). However, Egger's regression showed no significant asymmetry (p = 0.131), indicating that our results are robust and reliable.

MMSE

Thirteen trials involving 599 participants evaluated the Impact of rTMS on overall cognition using the MMSE. A significant difference was observed between the real rTMS and the sham rTMS groups. Combined data showed notably improved MMSE results within the active rTMS cohort (WMD: 2.04; 95% CI: 1.54, 2.53; p < 0.0001), and no significant heterogeneity was observed (I² = 42%, p = 0.06) (Figure 3B). A funnel plot inspection indicated minor reporting bias (Figure 4B). Nevertheless, Egger's findings failed to demonstrate statistical significance. (p = 0.686), indicating that our results are robust and reliable.

ADAS-COG

Compared to the Sham rTMS group, eight studies (1, 3, 6, 7, 8, 9, 10, 13) with 405 participants (n = 231 in the real rTMS group and n = 174 in the sham rTMS group) assessed the effects of rTMS using the ADAS-COG. The real rTMS group showed a statistically significant effect (WMD: -2.59;95% CI: -3.63, -1.54; p<0.0001), and heterogeneity was negligible (I²=0%, p=0.53) (Figure 3C). The funnel plot showed no significant asymmetry (Figure 4C).

MoCA

Using data from five studies (2,5,9,10,12) with 173 participants (n=89 in the real rTMS group and n=84 in the sham rTMS group), Real rTMS significantly enhanced the MoCA score over sham rTMS (WMD: 2.46, 95% CI: 1.47,3.45, p<0.0001); heterogeneity was non-significant (Figure 3D). The funnel plot displayed no substantial asymmetry (Figure 4D).

CDR

Four trials (7, 8, 11, 12) with 164 participants (n = 82 in the real rTMS group and n = 82 in the Sham rTMS group) evaluated the Impact of rTMS on global cognitive function via the CDR. The findings revealed no notable distinction between active rTMS and sham stimulation. (WMD: -0.41,95% CI: -0.94,0.12, p=0.13) (Figure 3E). The heterogeneity of the included studies exceeds what was anticipated. (I²=67%, p=0.03), suggesting that the results of the included studies showed significant heterogeneity. The sensitivity analysis of the trials indicated that after removing the highly weighted study (Koch et al., 2022), the I² statistics among the studies was 0% (p = 0.90). Furthermore, the distinction between the two groups remained statistically insignificant (WMD: -0.14, 95% CI: -0.37, 0.09, p = 0.22). The funnel plot showed symmetry across the three studies (Figure 4E).

Publication Bias and Sensitivity Analysis

All eligible studies were assessed for publication bias. Figure 3 shows a visually symmetrical funnel plot, indicating a balanced distribution of studies. Additionally, the effectiveness of rTMS on Global Cognitive Function and MMSE, as shown by Egger's tests, indicated no significant publication bias with p-values of 0.13 and 0.68 for each test. We then performed a sensitivity analysis using a one-by-one exclusion method; removing one study at a time did not significantly change the pooled results, supporting the reliability of our findings Figure 5.

Figure 3: Forest plot: Meta-analysis of rTMS on different cognitive outcomes (a-e)

Figure 4: Funnel plot: rTMS on different cognitive outcomes (a-e)

Figure 5: Sensitivity analysis of rTMS on Global cognition and MMSE

Subgroup Analyses

Table 3 outlines our subgroup analyses, which investigated how different related factors influence clinical outcomes.

rTMS frequency

Among the included studies, the stimulation frequency of TMS was divided into 5, 10, 20, and 50 Hz(iTBS). Subgroup analysis of rTMS frequency indicated that 5, 10, 20 Hz rTMS and iTBS were (SMD, 0.49; 95% CI 0.24,0.75; P=0.0002) and (SMD, 0.61; 95% CI -0.24,1.45; P=0.16), respectively. The results revealed that rTMS at 5, 10, and 20 Hz had a positive effect on patients' global cognitive function (Figure 6A).

rTMS sites

The analysis indicated that stimulating multiple sites (SMD, 0.77; 95% CI 0.45,1.08; P<0.001) and single-site stimulation did produce similar benefits (SMD, 0.49; 95% CI 0.31,0.66; P<0.001). Both significantly improved cognitive outcomes (Figure 6B).

Cognitive train

In the subgroup analysis of cognitive training, trials without cognitive training failed to demonstrate any meaningful enhancement in cognition post-real rTMS (SMD, 0.25; 95% CI -0.07,0.57; P=0.12). rTMS yielded notable results within the cognitive training group. (SMD, 0.64; 95% CI 0.30, 0.98; P=0.003) (Figure 6C).

Education

Patients' education levels ranged from 4 to 15 years, with 10 years identified as a key cutoff. Notably, patients with education of ≥ 10 years (SMD, 0.58; 95% CI 0.29, 0.87; P<0.0001) and those with < 10 years of education (SMD, 0.53; 95% CI 0.09, 0.97; P=0.02) showed significant improvements in cognitive scores in both groups (Figure 6D).

Table 3: Results for subgroup analysis

Subgroup	No. trials	Sample Size (Real /Sham)	SMD (95% CI)	Heterogeneity (I²) (%)	P value
rTMS frequency
5, 10, 20 Hz	11	(284/272)	0.49[0.24,0.75]	51	0.0002
iTBS	3	(75/75)	0.61[-0.24,1.45]	84	0.16
rTMS sites
multiple sites	5	(86/91)	0.77[0.45,1.08]	63	<0.001
single-site	9	(257/256)	0.49[0.31,0.66]	13	<0.001
Cognitive train
Yes	8	(245/240)	0.64[0.30,0.98]	68	0.003
No	4	(79/73)	0.25[-0.07,0.57]	0	0.12
Education(years)
≥ 10	10	(260/258)	0.58[0.29,0.87]	57	0.0001
< 10	4	(99/89)	0.53[0.09,0.97]	54	0.02

SMD, standardized mean difference, iTBS, intermittent theta-burst stimulation

Figure 6: Forest plot: Subgroup Analyses of rTMS on different cognitive outcomes (a-d)

Discussion

We have thoroughly and systematically searched for studies related to rTMS treatment and cognitive impairment in elderly patients. This search resulted in 13 studies with a total of 655 participants, and A comprehensive meta-analysis was performed on these studies to assess the therapeutic effect of rTMS on cognitive impairment. After completing our analysis, we identified the following key primary outcomes: 1) The rTMS group demonstrated significantly greater improvement in overall cognitive function compared to the sham-stimulation group, as measured by the MMSE, MoCA, and ADAS-Cog. In subgroup analyses: 2) HF-rTMS showed consistent cognitive benefits in elderly patients, while iTBS shortened treatment duration with similar effectiveness, supporting its clinical use utility; 3) rTMS combined with cognitive training yielded significantly greater therapeutic effects than rTMS alone; 4) four studies employed multi-site stimulation, while nine used single-site rTMS. The most common target was the left DLPFC, featured in eight studies. Other targeted areas included the precuneus, parietal, temporal, and cerebellum. Despite comparable cognitive gains, single- and multi-site stimulation showed no statistically significant difference; the latter showed a clear advantage, indicating that further Research is warranted in this area. No serious adverse effects were observed in any study, suggesting that rTMS is a safe treatment option for elderly patients with cognitive impairment. These results not only support but also reinforce previous meta-analyses in the field.

Our meta-analysis reveals that rTMS provides both short-term and long-term benefits for older adults with cognitive impairment, consistent with the findings of Chen et al. (2018)⁴². Our findings collectively advocate for a paradigm shift in rTMS therapy, transitioning from a broadly effective approach to a more refined and stratified treatment method. This approach recommends prioritizing the left DLPFC with either HF-rTMS or the more time-efficient iTBS protocol, integrated with cognitive training, for most elderly patients with cognitive impairment. Additionally, the choice of protocol can be tailored based on individual factors, such as time availability and treatment response. For complex cases resistant to standard therapy, exploring sequential multi-target stimulation—such as targeting the precuneus for memory or the cerebellum for executive function—offers a promising and personalized direction for future research and clinical practice.

Current pharmacological strategies for cognitive impairment target various mechanisms, including neurotransmitter regulation, core pathology, and metabolic modulation. This is exemplified by acetylcholinesterase inhibitors, which enhance cholinergic transmission; NMDA receptor antagonists like memantine, which protect neurons from excitotoxicity; anti-Aβ monoclonal antibodies (lecanemab, donanemab) that clear amyloid plaques to modify disease progression; and GLP-1 receptor agonists (tirzepatide), which reduce Aβ production and improve cerebral metabolis⁴³. However, these approaches face significant limitations. The complex, multifactorial nature of cognitive disorders means single-target drugs often fail to address the entire disease network. Furthermore, many agents cannot repair neuronal damage or halt the progression of pathology, are associated with side effects (dizziness, confusion, microhemorrhages), and have a narrow therapeutic window, often showing efficacy only in early-stage patients at a high financial cost. Finally, several promising drug candidates remain in preclinical or clinical trials, with their safety and efficacy in humans yet to be fully established. Compared to pharmacotherapy, rTMS appears to offer a more comprehensive therapeutic profile for various conditions, including cognitive impairment, major depressive disorder, pain syndromes, and schizophrenia. Its efficacy is attributed to multiple neurophysiological effects: 1) Regulation of neural circuitry: rTMS modulates cortical excitability and functional connectivity. For instance, HF-rTMS can induce long-term potentiation (LTP)-like effects, enhance ipsilateral cortical activity, improve cerebrovascular function, and suppress neuronal apoptosis⁴⁴. 2) Enhancement of cerebral metabolism: rTMS stimulates cerebral cells, leading to increased phosphorylation levels, which play a critical role in mediating cell function, proliferation, survival, and neural signal transduction⁴⁵. 3) Neuroprotective and plasticity-enhancing effects: These include preserving dopaminergic neuron function, reducing glutamate-induced excitotoxicity, upregulating neurotrophic factors, diminishing oxidative stress and neuroinflammation, and inhibiting astrocyte proliferation⁴⁶. Enhances local functional integration in the brain through synaptic plasticity modulation, improves neural information processing efficiency by rebalancing interhemispheric inhibition, and reduces grey matter volume loss in prodromal AD cohorts⁴⁷. In rat models of cerebral ischemia, HF-rTMS was shown to activate calcium influx via the p-Akt/GSK3β/β-catenin signalling pathway, promoting the release of neurotrophic factors and enhancing the proliferation of neural stem cells in the peri-infarct region⁴⁸.

Overall, rTMS is a well-tolerated intervention for elderly patients with cognitive impairment, with a very low rate of serious adverse events reported in the literature. Its proven ability to enhance cognitive function highlights its safety and reinforces its therapeutic value⁴⁹^,¹²^,⁵⁰. These findings support the hypothesis that rTMS therapy may improve cognitive function in older adults, offering a more objective overview of its possible mechanisms and influencing factors.

In the Subgroup Analysis

Education

Advanced education correlates with decreased AD susceptibility. This protective effect is connected to enhanced neural plasticity and compensatory capacity, which can delay the progression of neuropathological changes such as amyloid-beta deposition⁵¹. Research indicates that greater educational attainment could significantly influence rTMS treatment outcomes, possibly because of a stronger underlying neurophysiological mechanism⁵². The study revealed no notable cognitive outcome disparities between patients with 10 or more years of schooling and those with less than 10 years, the possible explanations are as follows: first this discrepancy with previous findings may be due to an uneven sample size across subgroups, especially a much larger number in the ≥10 years group (n = 260) compared to the <10 years group (n = 99); Secondly the influence of education extends beyond early formal schooling to include lifelong participation in cognitively demanding activities (reading, social engagement, skill acquisition), these activities help sustain cognitive reserve by continuously stimulating neural circuits, which is essential for optimal brain function⁵³.

rTMS frequency

All studies in our meta-analysis employed TMS techniques with high frequency. The treatment protocols consisted of conventional rTMS in ten studies and iTBS in three studies. Initial subgroup analyses suggested a potential superiority of rTMS over iTBS, accompanied by significant heterogeneity among iTBS studies. This variability likely stems from unequal sample sizes (rTMS: n = 556; iTBS: n = 150) and differences in cognitive assessment tools, as seen in Hoy et al. (2023). When the latter study was removed in sensitivity analysis, heterogeneity decreased markedly, and both intervention modalities then demonstrated significant cognitive benefits. This finding supports earlier Research by demonstrating that selecting the appropriate treatment protocol is crucial for obtaining consistent therapeutic outcomes. iTBS simulates the brain's natural firing patterns and produces strong synaptic plasticity through its high-frequency, short-burst approach. Additionally, its clinical use is very efficient, with each session lasting about 3 minutes. Large-scale randomized controlled trials, such as the THREE-D study published in The Lancet, have shown that iTBS is not inferior to traditional rTMS for treatment-resistant depression, with similar remission (32%) and response (49%) rates⁵⁴. A key benefit of iTBS is the significant reduction—about 90%—in session duration. Although side effects like headaches occur at similar rates between protocols, iTBS is generally better tolerated due to its shorter treatment times. Ivan J. Torres and colleagues (2023) found that memory enhancement after iTBS was linked to a simultaneous increase in hippocampal volume, supporting evidence for iTBS-induced neuroplastic changes within the prefrontal-temporal circuits⁵⁵. Despite its promise as an efficient neuromodulation method, the current evidence supporting iTBS remains limited, requiring more rigorous RCTs to definitively confirm its effectiveness and mechanisms.

rTMS sites

Subgroup analysis based on stimulation sites showed no significant advantage of multi-site rTMS over single-site stimulation (Hedges' g = 0.47, 95% CI [0.14, 0.79] vs. 0.24, 95% CI [-0.45, 0.92]), which is contrary to our initial hypothesis. Meta-analytic evidence (Hu et al., 2024; Wang et al., 2019) suggests that multi-site rTMS, which targets multiple functional brain networks, provides significantly greater cognitive benefits than single-site stimulation, thereby demonstrating superior efficacy²⁴^,⁵⁶. However, the meta-analysis by Wing Cheng et al. (2017) did not find a significant extra cognitive benefit from multi-site rTMS stimulation compared to more focused protocols⁴². Our findings align with these results. Several possible explanations may explain these observations：①in our meta-analysis, the selected protocols varied: eight trials used single-site rTMS (usually to the DLPFC), while five trials explored multi-site stimulation paradigms. Additionally, several newly targeted brain regions—including the temporoparietal junction (TPJ), posterior parietal cortex (PPC), cerebellum, lateral parietal cortex, and lateral temporal lobe (LTL)—are currently under investigation. While some studies report that TMS applied to different brain regions can increase precuneus activity and modulate the effective connectivity of the DLPFC, enhancing functional connectivity between the anterior cingulate cortex and medial prefrontal regions within the default mode network (DMN), other studies show that stimulation at different sites may have opposite effects on the same outcome measure. For example, Alexander et al. demonstrated that TMS applied to depression-related targets (F3 and 5 cm anterior to M1) caused heart rate deceleration via the frontal-vagal network, while stimulation at other sites led to heart rate acceleration^57,58^,⁵⁹. Therefore, multi-site rTMS should be used cautiously in clinical settings. ②the heterogeneity in the single-site rTMS group, caused by differences in assessment tools and a small sample size (n=73), was confirmed by a sensitivity analysis. Removing the study by Hoy et al. (2023) resulted in a significant reduction in heterogeneity.

Cognitive train

In our systematic review, seven studies used a combined protocol of rTMS plus cognitive training, while six studies administered rTMS alone. Comparative analysis showed that the combined intervention provided significantly greater clinical benefits. These findings support our hypothesis that cognitive training improves cognitive function by reactivating memory traces through guided rehearsal and recall, including emotional stimulation to strengthen consolidation, and encouraging bottom-up modulation of neural excitability³¹. TMS delivers induced currents that top-down modulate cortical neurons, altering their action potentials and excitability. This modulation enhances synaptic plasticity and connectivity, resulting in electrophysiological changes that boost metabolic and hemodynamic processes, collectively leading to symptomatic improvement³⁰. The meta-analysis by Eleni-Nefeli et al. (2024) showed that combined rTMS and cognitive training enhances overall cognition, specific neurocognitive areas, and quality of life in patients with AD⁶⁰. Sabbagh et al. (2019) found that using neuronavigated rTMS, along with cognitive training, is not only safe and well-tolerated by patients with mild Alzheimer's disease but also appears to be effective in managing the condition⁶¹. Despite strong evidence from meta-analyses and RCTs supporting the advantages of combining rTMS with cognitive training, some studies—such as the network meta-analysis by Chu et al. (2021) (n = 1,070)—did not show a significant overall cognitive benefit of the combined approach over individual therapies⁶². These conflicting findings should be approached with caution, especially since the mechanisms behind how combined rTMS and cognitive training enhance cognition are still not well understood. We therefore propose the following hypotheses to explain these observations: Primarily, the interaction between rTMS and CT on overall cognition can be affected by the complexity of the brain's functional networks. This complexity encompasses key dynamic features like topological structure and synchrony, which are greatly shaped by structural connectivity metrics⁶³. Heterogeneous neuroanatomical damage and different patterns of functional connectivity among patients may explain why some do not respond to both rTMS and cognitive training interventions. Additionally, the neuroprotective effects of non-invasive brain stimulation on cognition may vary depending on participant characteristics, target region choice, the use of standardized versus personalized cognitive training protocols, and potential antagonistic interactions between different therapeutic approaches. For example, De Marco et al. reported that cognitive training increases DMN connectivity in individuals with MCI. In contrast, recent evidence suggests that rTMS can temporarily deactivate DMN functional connectivity, thereby helping to reduce cognitive deficits in patients with amnestic MCI⁶⁴^,⁶⁵.

Due to the limited number of available studies, a formal comparison of stimulation sites was not possible. However, the DLPFC was the main target and served as the primary site in eight of the included trials, consistent with both current and past Research findings¹¹^,⁶⁶. Information processing in the brain is organized into large-scale interconnected networks that dynamically reconfigure in response to environmental stimuli and cognitive demands. By leveraging the regulatory role of cross-network interactions, rTMS targeting the dorsolateral prefrontal cortex enhances resting-state functional connectivity across distributed networks, as indicated by fMRI⁶⁷. Our meta-analysis also identified the cerebellum, left temporoparietal junction, and precuneus as new therapeutic targets; rTMS applied to these areas showed significant improvements in cognitive function. This phenomenon may reflect the synergistic interactions within functional brain networks described earlier. As Tong et al. (2024) demonstrated, rTMS remakes network architecture in Alzheimer’s disease by boosting functional connectivity between gray and white matter, with the corona radiata and superior longitudinal fasciculus playing crucial roles⁶⁸. The combined effects and interactions among stimulation sites require further research for a thorough understanding. The studies incorporated into this meta-analysis employed HF-rTMS (≥5 Hz), and three of them applied the iTBS protocol. This selection may, at least in part, explain the observed superiority of high-frequency rTMS in enhancing cognitive function. Although Turriziani et al. (2019) demonstrated that two weeks of LF-rTMS (1 Hz) applied to the right DLPFC improved recognition memory in patients with AD compared to sham stimulation, the lack of similar studies calls for independent Research replication⁶⁹. Research examined in this meta-analysis (1, 2, 3, 6, 9, 10, 11, 12, 13) explicitly documented concurrent pharmacotherapy in patients with AD—most commonly donepezil, memantine, or other cholinesterase inhibitors—indicating that medication status may influence rTMS effectiveness. In line with this, Li et al. (2023) demonstrated that combining rTMS with neurotransmitter-modulating agents enhances functional connectivity within the frontoparietal and default-mode networks more effectively than rTMS alone⁷⁰. Future Research should aim to evaluate potential interactions among these clinical factors.

Limitations and Future Research

The research has several limitations. First, the meta-analysis included only 13 studies (14 trials), leading to a modest sample size. Second, the cognitive tools examined were the MMSE, ADAS-Cog, MoCA, and CDR. Except for the CDR, all measures showed significant effect sizes for cognitive improvement. However, evidence for other clinically relevant outcomes remains limited, as only four studies have reported multimodal indicators, including functional connectivity, resting-state electroencephalography, serum inflammatory markers, and grey-matter volume. These initial findings still highlight promising biomarker candidates. Third, aside from the study by Roque GY et al. 2021, which enrolled patients with MCI, all RCTs enrolled exclusively mild-to-moderate AD patients. Fourth, the current analysis only compared outcomes at the initial post-treatment assessment. Strong recommendations for future research include: ①Large-scale multicenter RCTs are needed to strengthen these findings; ②the standardized stimulation protocol for the clinical treatment of rTMS still requires further exploration; ③Additional studies are necessary to understand the mechanism linking multimodal assessment parameters and the therapeutic effect of cognitive impairment; ④the long-term effectiveness of rTMS must be confirmed through longitudinal studies. ⑤Further investigation should encompass additional cognitive deficit variations and individuals suffering from advanced Alzheimer's, with dedicated subgroup analyses for each group.

Conclusion

This meta-analysis provides strong evidence that rTMS significantly improves overall cognitive function in individuals with mild-to-moderate AD. Combining cognitive training with rTMS offers additional cognitive benefits compared to rTMS alone. Among the limited studies using iTBS, its effectiveness was comparable to traditional rTMS, but the treatment duration was notably shorter. In summary, this meta-analysis includes the largest and most recent cohort (n = 655) studying rTMS for age-related cognitive impairment. Beyond the pooled effect size analysis, we conducted four targeted subgroup analyses and explored their potential moderators and underlying mechanisms, supporting rTMS as a safe and effective treatment for elderly patients with mild-to-moderate AD. A typical course involves HF-rTMS over the left DLPFC, with iTBS protocols serving as a time-efficient alternative. While the safety profile is favorable, these recommendations should be considered within the context of the need for larger and longer-term RCTs. These findings provide valuable guidance for developing future therapeutic strategies.

Conflicts of Interest

The authors declare no conflicts of interest.

References

Pérez Palmer N, Trejo Ortega B, Joshi P. Cognitive Impairment in Older Adults: Epidemiology, Diagnosis, and Treatment. The Psychiatric clinics of North America. 2022; 45(4): 639-661. DOI: 10.1016/j.psc.2022.07.010.
Frisoni G B, Aho E, Brayne C, et al. Alzheimer's disease outlook: controversies and future directions. Lancet. 2025. DOI: 10.1016/s0140-6736(25)01389-3.
Livingston G, Huntley J, Sommerlad A, et al. Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Lancet. 2020; 396(10248): 413-446. DOI: 10.1016/s0140-6736(20)30367-6.
Xia X, Jiang Q, Mcdermott J, et al. Aging and Alzheimer's disease: Comparison and associations from molecular to system level. Aging cell. 2018; 17(5): e12802. DOI: 10.1111/acel.12802.
Ganguli M, Fu B, Snitz BE, et al. Mild cognitive impairment: incidence and vascular risk factors in a population-based cohort. 2013; 80(23): 2112-2120. DOI: 10.1212/WNL.0b013e318295d776.
Wang YY, Zhang M, Wang XX, et al. Correlates of cognitive impairment in the elderly in China: A cross-sectional study. Frontiers in public health. 2022; 10: 973661. DOI: 10.3389/fpubh.2022.973661.
Global status report on the public health response to dementia: еxecutive summary. World Health Organization. 2021.
Farsai PS. Cognitive Impairment in Older Adults and Oral Health Considerations: Treatment and Management. Clinics in geriatric medicine. 2023; 39(2): 295-310. DOI: 10.1016/j.cger.2023.01.001.
Walsh S, King E, Brayne C. France removes state funding for dementia drugs. 2019; 367: l6930. DOI: 10.1136/bmj.l6930.
Lawlor B, Segurado R, Kennelly S, et al. Nilvadipine in mild to moderate Alzheimer disease: A randomised controlled trial. PLoS medicine. 2018; 15(9): e1002660. DOI: 10.1371/journal.pmed.1002660.
Lin Y, Jiang WJ, Shan PY, et al. The role of repetitive transcranial magnetic stimulation (rTMS) in the treatment of cognitive impairment in patients with Alzheimer's disease: A systematic review and meta-analysis. J Neurol Sci. 2019; 398: 184-191. DOI: 10.1016/j.jns.2019.01.038.
Chou YH, Sundman M, Ton That V, et al. Cortical excitability and plasticity in Alzheimer's disease and mild cognitive impairment: A systematic review and meta-analysis of transcranial magnetic stimulation studies. Ageing research reviews. 2022; 79: 101660. DOI: 10.1016/j.arr.2022.101660.
Bashir S, Uzair M, Abualait T, et al. Effects of transcranial magnetic stimulation on neurobiological changes in Alzheimer's disease. Molecular medicine reports. 2022; 25(4). DOI: 10.3892/mmr.2022.12625.
Wu C, Lin T, Ding Q, et al. Continuous Theta-Burst Stimulation Promotes Paravascular CSF-Interstitial Fluid Exchange through Regulation of Aquaporin-4 Polarization in APP/PS1 Mice. Mediators of inflammation. 2022; 2022: 2140524. DOI: 10.1155/2022/2140524.
Siebner HR, Funke K, Aberra AS, et al. Transcranial magnetic stimulation of the brain: What is stimulated? - A consensus and critical position paper. Clin Neurophysiol. 2022; 140: 59-97. DOI: 10.1016/j.clinph.2022.04.022.
D'ostilio K, Goetz S M, Hannah R, et al. Effect of coil orientation on strength-duration time constant and I-wave activation with controllable pulse parameter transcranial magnetic stimulation. Clin Neurophysiol. 2016; 127(1): 675-683. DOI: 10.1016/j.clinph.2015.05.017.
Pashut T, Magidov D, Ben-Porat H, et al. Patch-clamp recordings of rat neurons from acute brain slices of the somatosensory cortex during magnetic stimulation. Frontiers in cellular neuroscience. 2014; 8: 145. DOI: 10.3389/fncel.2014.00145.
Grady C, Sarraf S, Saverino C, et al. Age differences in the functional interactions among the default, frontoparietal control, and dorsal attention networks. Neurobiology of aging. 2016; 41: 159-172. DOI: 10.1016/j.neurobiolaging.2016.02.020.
Wang JX, Voss JL. Long-lasting enhancements of memory and hippocampal-cortical functional connectivity following multiple-day targeted noninvasive stimulation. Hippocampus. 2015; 25(8): 877-883. DOI: 10.1002/hipo.22416.
Wang JX, Rogers LM, Gross EZ, et al. Targeted enhancement of cortical-hippocampal brain networks and associative memory. Science (New York, NY). 2014; 345(6200): 1054-1057. DOI: 10.1126/science.1252900.
Senczyszyn A, Szcześniak D, Wieczorek T, et al. Improvement of working memory in older adults with mild cognitive impairment after repetitive transcranial magnetic stimulation - a randomized controlled pilot study. Front Psychiatry. 2023; 14: 1196478. DOI: 10.3389/fpsyt.2023.1196478.
Cristancho P, Kamel L, Araque M, et al. iTBS to Relieve Depression and Executive Dysfunction in Older Adults: An Open Label Study. The American journal of geriatric psychiatry: official journal of the American Association for Geriatric Psychiatry. 2020; 28(11): 1195-1199. DOI: 10.1016/j.jagp.2020.03.001.
Sanches C, Stengel C, Godard J, et al. Past, Present, and Future of Non-invasive Brain Stimulation Approaches to Treat Cognitive Impairment in Neurodegenerative Diseases: Time for a Comprehensive Critical Review. Front Aging Neurosci. 2020; 12: 578339. DOI: 10.3389/fnagi.2020.578339.
Hu M, Nitsche MA, Lv Y, et al. The effects of repetitive transcranial magnetic and transcranial direct current stimulation on memory functions in older adults with mild cognitive impairment: a systematic review and meta-analysis. Front Hum Neurosci. 2024; 18: 1436448. DOI: 10.3389/fnhum.2024.1436448.
Zhang Q, Zhu PP, Yang L, et al. Research hotspots and trends in transcranial magnetic stimulation for cognitive impairment: A bibliometric analysis from 2014 to 2023. World J Psychiatry. 2024; 14(10): 1592-1604. DOI: 10.5498/wjp.v14.i10.1592.
Drumond Marra HL, Myczkowski ML, Maia Memória C, et al. Transcranial Magnetic Stimulation to Address Mild Cognitive Impairment in the Elderly: A Randomized Controlled Study. Behavioural neurology. 2015; 2015: 287843. DOI: 10.1155/2015/287843.
Wang X, Ding Q, Li Y, et al. Repetitive transcranial magnetic stimulation impacts the executive function of patients with vascular cognitive impairment: a systematic review and meta-analysis. Front Neurol. 2024; 15: 1374395. DOI: 10.3389/fneur.2024.1374395.
Miller A, Allen RJ, Juma AA, et al. Does repetitive transcranial magnetic stimulation improve cognitive function in age-related neurodegenerative diseases? A systematic review and meta-analysis. Int J Geriatr Psychiatry. 2023; 38(8): e5974. DOI: 10.1002/gps.5974.
Pagali SR, Kumar R, Lemahieu AM, et al. Efficacy and safety of transcranial magnetic stimulation on cognition in mild cognitive impairment, Alzheimer's disease, Alzheimer's disease-related dementias, and other cognitive disorders: a systematic review and meta-analysis. Int Psychogeriatr. 2024; 36(10): 880-928. DOI: 10.1017/s1041610224000085.
Yang Y, Yan Z, Chang W, et al. Effect of different modalities of transcranial magnetic stimulation on Parkinson's patients cognitive impairment and long-term effectiveness: a systematic review and network meta-analysis. Front Neurosci. 2024; 18: 1354864. DOI: 10.3389/fnins.2024.1354864.
Yang T, Liu W, He J, et al. The cognitive effect of non-invasive brain stimulation combined with cognitive training in Alzheimer's disease and mild cognitive impairment: a systematic review and meta-analysis. Alzheimer's research & therapy. 2024; 16(1): 140. DOI: 10.1186/s13195-024-01505-9.
Zhu M, Huang S, Chen W, et al. The effect of transcranial magnetic stimulation on cognitive function in post-stroke patients: a systematic review and meta-analysis. BMC Neurol. 2024; 24(1): 234. DOI: 10.1186/s12883-024-03726-9.
Chou YH, Ton That V, Sundman M. A systematic review and meta-analysis of rTMS effects on cognitive enhancement in mild cognitive impairment and Alzheimer's disease. Neurobiology of aging. 2020; 86: 1-10. DOI: 10.1016/j.neurobiolaging.2019.08.020.
Yuan LQ, Zeng Q, Wang D, et al. Neuroimaging mechanisms of high-frequency repetitive transcranial magnetic stimulation for treatment of amnestic mild cognitive impairment: a double-blind randomized sham-controlled trial. Neural Regen Res. 2021; 16(4): 707-713. DOI: 10.4103/1673-5374.295345.
Taylor JL, Hambro BC, Strossman ND, et al. The effects of repetitive transcranial magnetic stimulation in older adults with mild cognitive impairment: a protocol for a randomized, controlled three-arm trial. BMC Neurol. 2019; 19(1): 326. DOI: 10.1186/s12883-019-1552-7.
Moussavi Z, Uehara M, Rutherford G, et al. Repetitive transcranial magnetic stimulation as a treatment for Alzheimer's disease: A randomized placebo-controlled double-blind clinical trial. Neurotherapeutics: the journal of the American Society for Experimental NeuroTherapeutics. 2024; 21(3): e00331. DOI: 10.1016/j.neurot.2024.e00331.
Brem AK, Di Iorio R, Fried PJ, et al. Corticomotor Plasticity Predicts Clinical Efficacy of Combined Neuromodulation and Cognitive Training in Alzheimer's Disease. Front Aging Neurosci. 2020; 12: 200. DOI: 10.3389/fnagi.2020.00200.
Petersen RC. Mild cognitive impairment as a diagnostic entity. Journal of internal medicine. 2004; 256(3): 183-194. DOI: 10.1111/j.1365-2796.2004.01388.x.
Mckhann GM, Knopman DS, Chertkow H, et al. The diagnosis of dementia due to Alzheimer's disease: recommendations from the National Institute on Aging-Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease. Alzheimer's & dementia : the journal of the Alzheimer's Association. 2011; 7(3): 263-269. DOI: 10.1016/j.jalz.2011.03.005.
Page MJ, Mckenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Bmj. 2021; 372: n71. DOI: 10.1136/bmj.n71.
Schünemann HJ, Oxman Ad, Brozek J, et al. Grading quality of evidence and strength of recommendations for diagnostic tests and strategies. Bmj. 2008; 336(7653): 1106-1110. DOI: 10.1136/bmj.39500.677199.AE.
Cheng CPW, Wong CSM, Lee KK, et al. Effects of repetitive transcranial magnetic stimulation on improvement of cognition in elderly patients with cognitive impairment: a systematic review and meta-analysis. Int J Geriatr Psychiatry. 2018; 33(1): e1-e13. DOI: 10.1002/gps.4726.
Terao I, Kodama W. Comparative Efficacy, Tolerability, and Acceptability of Donanemab, Lecanemab, Aducanumab, Melatonin, and Aerobic Exercise for a Short Time on Cognitive Function in Mild Cognitive Impairment and Mild Alzheimer's Disease: A Systematic Review and Network Meta-Analysis. J Alzheimers Dis. 2024; 98(3): 825-835. DOI: 10.3233/jad-230911.
Garcia-Sanz S, Ghotme KA, Hedmont D, et al. Use of transcranial magnetic stimulation for studying the neural basis of numerical cognition: A systematic review. J Neurosci Methods. 2022; 369: 109485. DOI: 10.1016/j.jneumeth.2022.109485.
Kim SK, Lee GY, Kim SK, et al. Protective Effects of Repetitive Transcranial Magnetic Stimulation Against Streptozotocin-Induced Alzheimer's Disease. Mol Neurobiol. 2024; 61(3): 1687-1703. DOI: 10.1007/s12035-023-03573-8.
Wang N, Jia Y, Zhou X, et al. Effects of Repetitive Transcranial Magnetic Stimulation on Pallidum GABAergic Neurons and Motor Function in Rat Models of Kernicterus. Brain Sci. 2023; 13(9). DOI: 10.3390/brainsci13091252.
Liu C, Han T, Xu Z, et al. Modulating Gamma Oscillations Promotes Brain Connectivity to Improve Cognitive Impairment. Cereb Cortex. 2022; 32(12): 2644-2656. DOI: 10.1093/cercor/bhab371.
Luo J, Feng Y, Hong Z, et al. High-frequency repetitive transcranial magnetic stimulation promotes neural stem cell proliferation after ischemic stroke. Neural Regen Res. 2024; 19(8): 1772-1780. DOI: 10.4103/1673-5374.389303.
Begemann MJ, Brand BA, Ćurčić-Blake B, et al. Efficacy of non-invasive brain stimulation on cognitive functioning in brain disorders: a meta-analysis. Psychol Med. 2020; 50(15): 2465-2486. DOI: 10.1017/s0033291720003670.
Liu H, Wu M, Huang H, et al. Comparative efficacy of non-invasive brain stimulation on cognition function in patients with mild cognitive impairment: A systematic review and network meta-analysis. Ageing research reviews. 2024; 101: 102508. DOI: 10.1016/j.arr.2024.102508.
Clouston SAP, Smith DM, Mukherjee S, et al. Education and Cognitive Decline: An Integrative Analysis of Global Longitudinal Studies of Cognitive Aging. The journals of gerontology Series B, Psychological sciences and social sciences. 2020; 75(7): e151-e160. DOI: 10.1093/geronb/gbz053.
Pizzini FB, Ribaldi F, Natale V, et al. A visual scale to rate amygdalar atrophy on MRI. European radiology. 2025; 35(7): 4246-4256. DOI: 10.1007/s00330-024-11249-7.
Mai S, Cai J, Zheng W, et al. Path analysis of social engagement, loneliness and cognitive function among Chinese older adults. Psychology, health & medicine. 2023; 28(4): 1087-1099. DOI: 10.1080/13548506.2022.2151635.
Blumberger DM, Vila-Rodriguez F, Thorpe KE, et al. Effectiveness of theta burst versus high-frequency repetitive transcranial magnetic stimulation in patients with depression (THREE-D): a randomised non-inferiority trial. 2018; 391(10131): 1683-1692. DOI: 10.1016/s0140-6736(18)30295-2.
Torres IJ, Ge R, Mcgirr A, et al. Effects of intermittent theta-burst transcranial magnetic stimulation on cognition and hippocampal volumes in bipolar depression. Dialogues in clinical neuroscience. 2023; 25(1): 24-32.DOI: 10.1080/19585969.2023.2186189.
Wang X, Mao Z, Ling Z, et al. Repetitive transcranial magnetic stimulation for cognitive impairment in Alzheimer's disease: a meta-analysis of randomized controlled trials. J Neurol; 2020: 267(3): 791-801. DOI: 10.1007/s00415-019-09644-y.
Nardone R, Sebastianelli L, Versace V, et al. TMS-EEG Co-Registration in Patients with Mild Cognitive Impairment, Alzheimer's Disease and Other Dementias: A Systematic Review. Brain Sci. 2021; 11(3). DOI: 10.3390/brainsci11030303.
Koch G, Bonnì S, Pellicciari MC, et al. Transcranial magnetic stimulation of the precuneus enhances memory and neural activity in prodromal Alzheimer's disease. Neuroimage. 2018; 169: 302-311. DOI: 10.1016/j.neuroimage.2017.12.048.
Alexander TS, Jasmina P, Tara K, et al. Target Engagement and Brain State Dependence of Transcranial Magnetic Stimulation: Implications for Clinical Practice. Biol Psychiatr. 2023.
Georgopoulou EN, Nousia A, Martzoukou M, et al. Efficacy of rTMS Combined with Cognitive and Language Training in People Living with Alzheimer's Disease: A Systematic Review. Brain Sci. 2024; 14(9). DOI: 10.3390/brainsci14090891.
Sabbagh M, Sadowsky C, Tousi B, et al. Effects of a combined transcranial magnetic stimulation (TMS) and cognitive training intervention in patients with Alzheimer's disease. Alzheimer's & dementia: the journal of the Alzheimer's Association. 2020; 16(4): 641-650. DOI: 10.1016/j.jalz.2019.08.197.
Chu Cs, Li CT, Brunoni AR, et al. Cognitive effects and acceptability of non-invasive brain stimulation on Alzheimer's disease and mild cognitive impairment: a component network meta-analysis. Journal of neurology, neurosurgery, and psychiatry. 2021; 92(2): 195-203. DOI: 10.1136/jnnp-2020-323870.
Brem AK, Di Iorio R, Fried PJ, et al. Corticomotor Plasticity Predicts Clinical Efficacy of Combined Neuromodulation and Cognitive Training in Alzheimer's Disease. Front Aging Neurosci. 2020; 12: 200. DOI: 10.3389/fnagi.2020.00200.
De Marco M, Meneghello F, Pilosio C, et al. Up-regulation of DMN Connectivity in Mild Cognitive Impairment Via Network-based Cognitive Training. Current Alzheimer research. 2018; 15(6): 578-589. DOI: 10.2174/1567205015666171212103323.
Cui H, Ren R, Lin G, et al. Repetitive Transcranial Magnetic Stimulation Induced Hypoconnectivity Within the Default Mode Network Yields Cognitive Improvements in Amnestic Mild Cognitive Impairment: A Randomized Controlled Study. J Alzheimers Dis. 2019; 69(4): 1137-1151. DOI: 10.3233/jad-181296.
Hou Y, Liu F, Lin N, et al. Systematic review and meta-analysis of repetitive transcranial magnetic stimulation (rTMS) for activities of daily living in Alzheimer's disease. Neurol Sci. 2025; 46(1): 63-77. DOI: 10.1007/s10072-024-07709-z.
Beynel L, Powers JP, Appelbaum LG. Effects of repetitive transcranial magnetic stimulation on resting-state connectivity: A systematic review. Neuroimage. 2020; 211: 116596. DOI: 10.1016/j.neuroimage.2020.116596.
Qin T, Wang L, Xu H, et al. rTMS concurrent with cognitive training rewires AD brain by enhancing GM-WM functional connectivity: a preliminary study. Cereb Cortex. 2024; 34(1). DOI: 10.1093/cercor/bhad460.
Turriziani P, Smirni D, Mangano GR, et al. Low-Frequency Repetitive Transcranial Magnetic Stimulation of the Right Dorsolateral Prefrontal Cortex Enhances Recognition Memory in Alzheimer's Disease. J Alzheimers Dis. 2019; 72(2): 613-622. DOI: 10.3233/jad-190888.
Li M, Qin Z, Chen H, et al. Effects of Combined Intervention of rTMS and Neurotransmitter Drugs on the Brain Functional Networks in Patients with Cognitive Impairment. Brain Sci. 2023; 13(3). DOI: 10.3390/brainsci13030419.
Liao X, Li G, Wang A, et al. Repetitive Transcranial Magnetic Stimulation as an Alternative Therapy for Cognitive Impairment in Alzheimer's Disease: A Meta-Analysis. J Alzheimers Dis. 2015; 48(2): 463-472. DOI: 10.3233/jad-150346.

Supplementary Material 1: Search strategy

Database	Date	Terms	Results
MEDLINE PubMed	2024-12-15	#1: Transcranial Magnetic Stimulation
		"Transcranial Magnetic Stimulation"[MeSH Terms] OR ("Transcranial Magnetic Stimulation"[MeSH Terms] OR ("transcranial"[All Fields] AND "magnetic"[All Fields] AND "stimulation"[All Fields]) OR "Transcranial Magnetic Stimulation"[All Fields] OR ("magnetic"[All Fields] AND "stimulations"[All Fields] AND "transcranial"[All Fields]) OR ("Transcranial Magnetic Stimulation"[MeSH Terms] OR ("transcranial"[All Fields] AND "magnetic"[All Fields] AND "stimulation"[All Fields]) OR "Transcranial Magnetic Stimulation"[All Fields] OR ("magnetic"[All Fields] AND "stimulation"[All Fields] AND "transcranial"[All Fields]) OR "magnetic stimulation transcranial"[All Fields]) OR ("Transcranial Magnetic Stimulation"[MeSH Terms] OR ("transcranial"[All Fields] AND "magnetic"[All Fields] AND "stimulation"[All Fields]) OR "Transcranial Magnetic Stimulation"[All Fields] OR ("stimulations"[All Fields] AND "transcranial"[All Fields] AND "magnetic"[All Fields])) OR ("Transcranial Magnetic Stimulation"[MeSH Terms] OR ("transcranial"[All Fields] AND "magnetic"[All Fields] AND "stimulation"[All Fields]) OR "Transcranial Magnetic Stimulation"[All Fields] OR ("stimulation"[All Fields] AND "transcranial"[All Fields] AND "magnetic"[All Fields]) OR "stimulation transcranial magnetic"[All Fields]) OR ("Transcranial Magnetic Stimulation"[MeSH Terms] OR ("transcranial"[All Fields] AND "magnetic"[All Fields] AND "stimulation"[All Fields]) OR "Transcranial Magnetic Stimulation"[All Fields] OR ("transcranial"[All Fields] AND "magnetic"[All Fields] AND "stimulations"[All Fields]) OR "transcranial magnetic stimulations"[All Fields]) OR ("Transcranial Magnetic Stimulation"[MeSH Terms] OR ("transcranial"[All Fields] AND "magnetic"[All Fields] AND "stimulation"[All Fields]) OR "Transcranial Magnetic Stimulation"[All Fields] OR ("transcranial"[All Fields] AND "magnetic"[All Fields] AND "stimulation"[All Fields] AND "paired"[All Fields] AND "pulse"[All Fields]) OR "transcranial magnetic stimulation paired pulse"[All Fields]) OR ("Transcranial Magnetic Stimulation"[MeSH Terms] OR ("transcranial"[All Fields] AND "magnetic"[All Fields] AND "stimulation"[All Fields]) OR "Transcranial Magnetic Stimulation"[All Fields] OR ("transcranial"[All Fields] AND "magnetic"[All Fields] AND "stimulation"[All Fields] AND "repetitive"[All Fields]) OR "transcranial magnetic stimulation repetitive"[All Fields]) OR ("Transcranial Magnetic Stimulation"[MeSH Terms] OR ("transcranial"[All Fields] AND "magnetic"[All Fields] AND "stimulation"[All Fields]) OR "Transcranial Magnetic Stimulation"[All Fields] OR ("transcranial"[All Fields] AND "magnetic"[All Fields] AND "stimulation"[All Fields] AND "single"[All Fields] AND "pulse"[All Fields]) OR "transcranial magnetic stimulation single pulse"[All Fields]))	25253
		#2：Cognitive Dysfunction
		"Cognitive Dysfunction"[MeSH Terms] OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("cognitive"[All Fields] AND "dysfunctions"[All Fields]) OR "cognitive dysfunctions"[All Fields] OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("dysfunction"[All Fields] AND "cognitive"[All Fields]) OR "dysfunction cognitive"[All Fields]) OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("dysfunctions"[All Fields] AND "cognitive"[All Fields]) OR "dysfunctions cognitive"[All Fields]) OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("cognitive"[All Fields] AND "disorder"[All Fields]) OR "cognitive disorder"[All Fields]) OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("cognitive"[All Fields] AND "disorders"[All Fields]) OR "cognitive disorders"[All Fields]) OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("disorder"[All Fields] AND "cognitive"[All Fields]) OR "disorder cognitive"[All Fields]) OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("disorders"[All Fields] AND "cognitive"[All Fields]) OR "disorders cognitive"[All Fields]) OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("cognitive"[All Fields] AND "impairments"[All Fields]) OR "cognitive impairments"[All Fields]) OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("cognitive"[All Fields] AND "impairment"[All Fields]) OR "cognitive impairment"[All Fields]) OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("impairment"[All Fields] AND "cognitive"[All Fields]) OR "impairment cognitive"[All Fields]) OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("impairments"[All Fields] AND "cognitive"[All Fields]) OR "impairments cognitive"[All Fields]) OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("mild"[All Fields] AND "cognitive"[All Fields] AND "impairment"[All Fields]) OR "mild cognitive impairment"[All Fields]) OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("cognitive"[All Fields] AND "impairment"[All Fields] AND "mild"[All Fields]) OR "cognitive impairment mild"[All Fields]) OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("cognitive"[All Fields] AND "impairments"[All Fields] AND "mild"[All Fields]) OR "cognitive impairments mild"[All Fields]) OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("impairment"[All Fields] AND "mild"[All Fields] AND "cognitive"[All Fields]) OR "impairment mild cognitive"[All Fields]) OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("impairments"[All Fields] AND "mild"[All Fields] AND "cognitive"[All Fields]) OR "impairments mild cognitive"[All Fields]) OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("mild"[All Fields] AND "cognitive"[All Fields] AND "impairments"[All Fields]) OR "mild cognitive impairments"[All Fields]) OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("cognitive"[All Fields] AND "decline"[All Fields]) OR "cognitive decline"[All Fields]) OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("cognitive"[All Fields] AND "declines"[All Fields]) OR "cognitive declines"[All Fields]) OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("decline"[All Fields] AND "cognitive"[All Fields]) OR "decline cognitive"[All Fields]) OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("declines"[All Fields] AND "cognitive"[All Fields]) OR "declines cognitive"[All Fields]) OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("mental"[All Fields] AND "deterioration"[All Fields]) OR "mental deterioration"[All Fields]) OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("deterioration"[All Fields] AND "mental"[All Fields]) OR "deterioration mental"[All Fields]) OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("deteriorations"[All Fields] AND "mental"[All Fields])) OR ("Cognitive Dysfunction"[MeSH Terms] OR ("cognitive"[All Fields] AND "dysfunction"[All Fields]) OR "Cognitive Dysfunction"[All Fields] OR ("mental"[All Fields] AND "deteriorations"[All Fields]) OR "mental deteriorations"[All Fields]))	345092
		#1 AND#2	2386
		Filters applied: Randomized Controlled Trial.	276
The Cochrane library	2024-12-15	Transcranial Magnetic Stimulation
		#1:MeSH descriptor: [Transcranial Magnetic Stimulation] explode all trees #2:(Transcranial Magnetic Stimulation):ti,ab,kw OR (Magnetic Stimulations, Transcranial):ti,ab,kw OR (Magnetic Stimulation, Transcranial):ti,ab,kw OR (Stimulations, Transcranial Magnetic):ti,ab,kw OR (Stimulation, Transcranial Magnetic):ti,ab,kw #3:(Transcranial Magnetic Stimulations):ti,ab,kw OR (Transcranial Magnetic Stimulation, Paired Pulse):ti,ab,kw OR (Transcranial Magnetic Stimulation, Repetitive):ti,ab,kw OR (Transcranial Magnetic Stimulation, Single Pulse):ti,ab,kw #4:#1 OR #2 OR #3	8775
		Cognitive Dysfunction
		#5:MeSH descriptor: [Cognitive Dysfunction] explode all trees #6:(Cognitive Dysfunctions):ti,ab,kw OR (Mild Cognitive Impairment):ti,ab,kw OR (Impairments, Mild Cognitive):ti,ab,kw OR (Impairment, Mild Cognitive):ti,ab,kw OR (Mild Cognitive Impairments):ti,ab,kw #7:(Cognitive Impairments, Mild):ti,ab,kw OR (Cognitive Impairment, Mild):ti,ab,kw OR (Mental Deteriorations):ti,ab,kw OR (Deteriorations, Mental):ti,ab,kw OR (Cognitive Decline):ti,ab,kw #8:(Decline, Cognitive):ti,ab,kw OR (Mental Deterioration):ti,ab,kw OR (Cognitive Declines):ti,ab,kw OR (Declines, Cognitive):ti,ab,kw OR (Deterioration, Mental):ti,ab,kw #9:(Disorders, Cognitive):ti,ab,kw OR (Dysfunction, Cognitive):ti,ab,kw OR (Cognitive Disorder):ti,ab,kw OR (Impairments, Cognitive):ti,ab,kw OR (Cognitive Impairments):ti,ab,kw #10:(Impairment, Cognitive):ti,ab,kw OR (Disorder, Cognitive):ti,ab,kw OR (Cognitive Dysfunctions):ti,ab,kw OR (Cognitive Disorders):ti,ab,kw OR (Cognitive Impairment):ti,ab,kw #11:(Dysfunctions, Cognitive):ti,ab,kw #12:#5 OR #6 OR #7 OR #8 OR #9 OR #10 OR #11	59721
		#13:MeSH descriptor: [Randomized Controlled Trial] explode all trees #14:(Randomized Controlled Trial):ti,ab,kw OR (Randomized Controlled Trials as Topic):ti,ab,kw #15:#13 OR #14
		#16:#4 AND #12 AND # 15	211
		Filters applied: Randomized Controlled Trial.	211
Embase	2024-12-15	Transcranial Magnetic Stimulation
		#1:'transcranial magnetic stimulation'/exp' OR repetitive transcranial magnetic stimulation':ab,ti OR 'magnetic stimulations, transcranial':ab,ti OR 'magnetic stimulation, transcranial':ab,ti OR 'stimulations, transcranial magnetic':ab,ti OR 'transcranial magnetic stimulation':ab,ti OR 'transcranial magnetic stimulations':ab,ti OR 'transcranial magnetic stimulation, paired pulse':ab,ti OR 'transcranial magnetic stimulation, repetitive':ab,ti OR 'transcranial magnetic stimulation, single pulse':ab,ti	37289
		cognitive defect
		#2: 'cognitive defect'/exp OR 'cognition disorders':ab,ti OR 'cognitive complaints':ab,ti OR 'cognitive decline':ab,ti OR 'cognitive deficiency':ab,ti OR 'cognitive deficit':ab,ti OR 'cognitive disability':ab,ti OR 'cognitive disorder':ab,ti OR 'cognitive disorders':ab,ti OR 'cognitive disturbance':ab,ti OR 'cognitive dysfunction':ab,ti OR 'cognitive impairment':ab,ti OR 'delirium, dementia, amnestic, cognitive disorders':ab,ti OR overinclusion:ab,ti OR 'response interference':ab,ti OR 'cognitive defect':ab,ti	710584
		#1 and #2	2890
		Filters applied: Randomized Controlled Trial.	259
Web of Science	2024-12-15	Transcranial Magnetic Stimulation
		#1: (TS=(repetitive transcranial magnetic stimulation) OR TS=(transcranial magnetic stimulation) OR TS=(magnetic stimulations, transcranial) OR TS=(magnetic stimulation, transcranial) OR TS=(stimulations, transcranial magnetic) OR TS=(stimulation, transcranial magnetic) OR TS=(transcranial magnetic stimulations) OR TS=(transcranial magnetic stimulation, paired pulse) OR TS=(transcranial magnetic stimulation, repetitive) OR TS=(transcranial magnetic stimulation, single pulse))	47480
		cognitive defect
		#2: (TS=(cognitive defect) OR TS=(cognition disorders) OR TS=(cognitive complaints) OR TS=(cognitive decline) OR TS=(cognitive deficiency) OR TS=(cognitive deficit) OR TS=(cognitive disability) OR TS=(cognitive disorder) OR TS=(cognitive disorders) OR TS=(cognitive disturbance) AND TS=(cognitive dysfunction) OR TS=(cognitive impairment) OR TS=(delirium, dementia, amnestic, cognitive disorders) OR TS=(overinclusion) OR TS=(response interference) OR TS=(cognitive defect))	961512
		#1 AND #2 Filters applied: Filters applied: Randomized Controlled Trial AND Clinical Trial	287
China National Knowledge Infrastructure	2024-12-15	重复经颅磁刺激
		#1：SU=('重复经颅磁刺激'+'经颅磁刺激'+'无创性脑刺激'+'非侵入性脑刺激'+'间歇性θ节律刺激'+'间歇性θ短阵快速脉冲'+'Theta爆发式磁刺激')	6731
		认知障碍
		#2：SU=('认知障碍'+'认知功能'+'阿尔兹海默病'+'痴呆')	144100
		#3：#1 AND #2	1637
		Filters applied: Randomized Controlled Trial.	17
Wanfang Database	2024-12-15	重复经颅磁刺激
		#1：（主题:(经颅磁刺激) or 题名或关键词:(重复经颅磁刺激 or 经颅磁刺激 or 无创性脑刺激 or 非侵入性脑刺激 or 间歇性θ节律刺激 or 间歇性θ短阵快速脉冲 or Theta爆发式磁刺激))	7981
		认知障碍
		#2：(主题:(认知障碍) or 题名或关键词:(认知障碍 or 认知功能 or 阿尔兹海默病 or 痴呆))	116347
		#3：#1 AND #2	1247
		Filters applied: Randomized Controlled Trial.	445

Article Info

Journal of Neurology & Neuromedicine
Article Type : Review Article
View/Download pdf
DOI : 10.29245/2572.942X/2025/4.1326

Article Notes

Published on: October 14, 2025

Keywords

Repetitive Transcranial Magnetic Stimulation
Elderly Patients
Cognitive Impairment
Meta-Analysis
Randomized Controlled Trial

*Correspondence:

Dr. Zhigang Cao,
Department of Neurology, Sichuan Taikang Hospital, Chengdu, Sichuan, China;
Email: caozhigang8158@163.com

Copyright: ©2025 Cao ZG. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License.

Article

Repetitive Transcranial Magnetic Stimulation for Elderly Patients with Cognitive Impairment: A Meta Analysis of Randomized Controlled Trials

Abstract

Introduction

Methods

Eligibility Criteria

Exclusion Criteria

Search Strategy

Data Extraction

Quality Assessment

Evidence Quality Assessment

Statistical Analysis

Results

Study Selection

Study Characteristics

Adverse Events

Risk of Bias Assessment

Evidence quality evaluation results

Effects of rTMS on Cognitive Impairment

Publication Bias and Sensitivity Analysis

Subgroup Analyses

Discussion

In the Subgroup Analysis

Limitations and Future Research

Conclusion

Conflicts of Interest

References

Article Info

Article Notes

Keywords

*Correspondence: