In 40 seconds
Cognition is the most-studied outcome in dementia neuromodulation research. The Wang 2022 meta-analysis (17 RCTs, n=437) showed rTMS improved MMSE and ADAS-cog with SMD around 0.7. The Yang 2023 meta-analysis (10 RCTs, n=408) showed rTMS plus cognitive training improved cognition with SMD 0.66 (95% CI 0.30–1.02), persisting at 1–3 months. The Chu 2024 systematic review (143 studies, n≈5,800) rated overall certainty as low-to-moderate. Effects are real but modest — typically a 1.5–3 point ADAS-cog improvement at the cohort level. Individual response is variable. The Wei 2024 multisite null trial without cognitive training is a reminder that protocol pairing with cognitive training is decisive. Low-intensity consumer PEMF has no human RCT showing cognitive improvement in dementia.
Quick facts
- Wang 2022 meta-analysis: 17 RCTs / n=437 — SMD ~0.7 cognitive benefit
- Yang 2023 meta-analysis: 10 RCTs / n=408 — SMD 0.66 with cognitive training
- Chu 2024 systematic review: 143 studies / n≈5,800 — low-to-moderate GRADE certainty
- Protocol that worked: High-freq, multi-site rTMS + cognitive training, 5d/wk × 6 weeks
- Protocol that failed: Wei 2024 — stand-alone DLPFC rTMS without training, null result
- UK regulatory: NICE NG97: rTMS for AD only inside a clinical trial
- Consumer PEMF cognitive evidence: No human RCT in dementia
What 'modest cognitive benefit' actually means
'Statistically significant cognitive improvement' is easy to read past. Here is what it represents in practice.
The measurement instruments
Dementia cognition trials use a small number of standardised instruments: ADAS-cog (range 0–70, higher is worse, change of 3+ points usually considered clinically meaningful), MMSE (0–30, lower is worse, change of 2+ points usually clinically meaningful), CDR-SB (0–18, higher is worse, used in Koch 2022 and the Alzheimer's antibody trials), and ADCS-ADL (functional measure).
What the trial signals translate to
An ADAS-cog improvement of 1.9–3.8 points (the range in the Rabey, Sabbagh-subgroup and Wu trials) is at or just above the conventional threshold for clinical meaningfulness. Across the Wang 2022 and Yang 2023 meta-analyses the standardised effect size is in the moderate range. This is real but it is not transformative. It is roughly comparable in size to the effects of cholinesterase inhibitors over 6 months.
Why cognitive training pairing is so important
The trials that paired rTMS with cognitive training (Rabey 2013, Sabbagh 2020 NeuroAD lineage, Yang 2023 meta) consistently outperformed those that did not. The Wei 2024 null trial isolated DLPFC stimulation without paired training and found no benefit. The mechanistic argument is that rTMS opens a window of cortical plasticity that has to be channelled into a relevant cognitive task to produce a learning effect — analogous to physical therapy after stroke.
Why the precuneus target matters
Koch 2022 targeted the precuneus, a default-mode-network hub that fails very early in Alzheimer's. The result — CDR-SB stability over 24 weeks while sham declined — is the strongest single trial in the field. The 2025 52-week extension sustained the effect. This implies that the choice of cortical target matters as much as the device or the frequency.
Where consumer PEMF sits
Consumer-grade low-intensity PEMF has no human RCT showing cognitive improvement in dementia. The mechanistic evidence (Capelli 2017 miRNA modulation; the 2024 microglial polarisation paper) is at the cell-culture level. Mouse work (Arendash 2010) used TEMT at 918 MHz, which is a different modality. The translation from cell-culture mechanism to clinical cognitive outcome is the step the consumer PEMF literature has not yet taken.
Individual response is variable
Group-level meta-analysis hides individual variability. Some people respond meaningfully; others do not respond at all. There is no robust pre-treatment biomarker for likely responders. Anyone considering rTMS should set a clear measurement framework (ADAS-cog, MMSE or CDR-SB at baseline and 12 weeks) and a stop-rule for non-response.
Frequently asked questions
How much cognitive improvement should we expect from rTMS?
Group-level cohort signal in the trials is a 1.5–3 point ADAS-cog improvement, roughly comparable to cholinesterase-inhibitor effects over 6 months. Individual response is highly variable.
Does rTMS work without cognitive training?
Apparently not, or much less. The Wei 2024 multisite trial of stand-alone DLPFC rTMS without paired cognitive training was null. The trials that worked all paired rTMS with cognitive training.
How quickly does cognition change?
In the trials, change is detected at the cohort level at 6–12 weeks of treatment. Some trials show effects persisting at 1–3 months after the course ends. Individual response timing varies.
Can rTMS help cognition in mild cognitive impairment?
Chou 2020 meta-analysis (12 RCTs, n=329) found high-frequency multi-site rTMS with 10+ sessions improved general cognition and memory in MCI. The MCI evidence is less mature than the Alzheimer's evidence.
Do home PEMF mats improve cognition?
There is no completed sham-controlled human RCT showing a home PEMF mat improves cognition in dementia. The biology is plausible; the clinical-trial evidence does not exist.
Is the cognitive improvement maintained after rTMS stops?
Several trials report some persistence at 1–3 months. Koch 2022 used weekly maintenance for 22 weeks after the daily phase to sustain the effect. Long-term maintenance protocols are still being studied.
Can we combine rTMS with cholinesterase inhibitors?
The published trials enrolled patients already on cholinesterase inhibitors. There is no evidence the combination is unsafe; the cognitive benefit appears additive to standard care.
What about cognitive stimulation therapy alone, without rTMS?
Cognitive stimulation therapy (CST) is the only non-pharmacological intervention NICE NG97 specifically endorses for cognition and wellbeing in mild-to-moderate dementia. It is available on the NHS. It is the first thing to ask the memory clinic about before considering anything else.
Related dementia pages on PEMF UK
Looking for a PEMF clinic in the UK?
We list every credible PEMF therapy provider in the UK. Please remember: no UK clinic can legally claim to treat dementia with PEMF or rTMS.