← Back

tDCS-pDPN

Home-Based Transcranial Direct Current Stimulation in Painful Diabetic Peripheral Neuropathy: A Double Blinded, Randomized, Low-Intensity Sham-Controlled Crossover Trial

Share Share Copied! Compare

Year of Publication: 2026

Authors: Muthulingam JA, Lyng AB, Hostrup SN, et al; Aalborg University Hospital

Journal: Neuromodulation

Citation: Neuromodulation. 2026 Jul 1 (articles in press). DOI: 10.1016/j.neurom.2026.06.467

Link: https://doi.org/10.1016/j.neurom.2026.06.467

Clinical Question

Does 4 weeks of home-based high-intensity (2 mA) anodal M1 tDCS reduce pain more than low-intensity (0.3 mA) sham tDCS in adults with type 1 diabetes and painful diabetic peripheral neuropathy?

Bottom Line

In 22 adults with T1DM and pDPN completing this home-based, double-blind, crossover trial, 4 weeks of 2-mA tDCS over left M1 did not reduce NRS pain more than a 0.3-mA low-intensity sham (mean difference in average pain 0.03; 95% CI –0.40 to 0.46; p=0.89). Both interventions reduced pain from baseline within-group, and during treatment the low-intensity sham actually outperformed active tDCS, raising the possibility that 0.3-mA sham is not biologically inert. Active tDCS may still help a subset of patients with baseline central sensitization (higher temporal summation, lower pressure-pain thresholds).

        Major Points
        Investigator-initiated, randomized, double-blind, low-intensity (0.3 mA) sham-controlled crossover trial at Aalborg University Hospital, Denmark, Aug 14, 2023 – Mar 28, 2025 (NCT06152887).
35 adults with T1DM and clinically confirmed pDPN enrolled; 22 completed both 4-week treatment periods (37% dropout).
Intervention: Sooma tDCS, anode over left M1, cathode over contralateral supraorbital, 20 min/day, 5 days/week for 4 weeks; active 2 mA vs sham 0.3 mA; 6-week washout.
Primary outcome: average daily NRS pain in the week after 4-wk treatment — active vs sham mean difference 0.03 (95% CI –0.40 to 0.46; p=0.89). Maximum pain mean difference 0.09 (95% CI –0.42 to 0.60; p=0.73). No carryover, no period effect.
Within-group vs baseline: active tDCS reduced average pain 5.3±1.6→4.4±1.6 (p=0.03) and maximum pain 6.7±1.5→5.7±1.8 (p=0.01); sham showed no significant reduction (avg 4.6→4.4 p=0.4; max 6.1→5.8 p=0.3).
During treatment (wk 1-4), the low-intensity sham significantly out-reduced active on both average and maximum pain (all p<0.046) — opposite of the hypothesized direction.
Secondary outcomes (BPI, HADS, NTSS-6) all NS; SF-36 physical functioning favored sham (mean diff –5.43; 95% CI –10.52 to –0.33; p=0.04).
Responders (≥22.6% NRS reduction): 5/22 (23%) with active vs 3/22 (14%) with sham. Active responders showed lower foot pressure-pain threshold (222.7 vs 403 kPa; p=0.02) and trend to higher temporal summation.
Safety: 22 AE reports on active, 17 on sham (all p>0.05); AEs included nausea, skin irritation, headache, dizziness, dry mouth. 3 SAEs (1 hospitalization, 2 mental-health), all unrelated to tDCS.
Compliance: registered stimulations 19.04±1.4 (period 1) vs 17.7±4.8 (period 2), p=0.056; ~9.7% less use in period 2.
Authors caution that 0.3-mA "sham" may itself be neurophysiologically active (≈15% of the active current sustained for 20 min/day for 4 weeks) and recommend a truly inactive sham (0 mA with ramp) in future trials.

      

Design

Study Type: Investigator-initiated, randomized, double-blind, low-intensity sham-controlled, crossover clinical trial

Randomization: 1

Blinding: Double-blind: participants and outcome assessors/analysts blinded to sequence; blinding success not formally evaluated

Enrollment Period: August 14, 2023 – March 28, 2025

Follow-up Duration: Total study duration 20 weeks per participant (2 wk baseline + 4 wk treatment 1 + 6+ wk washout + 4 wk treatment 2 + 4 wk follow-up); primary endpoint assessed in the week after each 4-wk treatment

Centers: 1

Countries: Denmark

Sample Size: 35

Power Calculation: Assumed baseline NRS 5.3±2.0 SD; to detect a 1.2-point reduction (22.6% difference) between sham and active with 80% power at α=0.05 required 22 participants in a crossover design; up to 40 planned

Analysis: Intention-to-treat; linear mixed model for primary/secondary endpoints (assessing carryover and period effects); paired t-tests for within-group baseline vs treatment; McNemar for dichotomous side-effect variables; exploratory responder analysis with t-test/Mann–Whitney

Inclusion Criteria

Age ≥18 years
Clinical diagnosis of type 1 diabetes mellitus
Clinically confirmed painful diabetic peripheral neuropathy
Score of at least 4 on the Douleur Neuropathique 4 (DN4) questionnaire
Daily numeric rating scale (NRS) pain score >3
Able to read and understand Danish
Able to comply with scheduled visits and procedures
Signed informed consent

Exclusion Criteria

Ongoing alcohol dependence or drug abuse
Participation in other studies involving investigational drugs
History of brain surgery or significant intracranial abnormalities likely to increase trial risk
Active cancer
Major mental or psychiatric disorders
Known neural damage or previous chemotherapy
Active herpes simplex virus
Limb pain (unrelated to pDPN)
Pregnancy or lactation
Contraindications to MRI

Arms

Field	Active tDCS (2 mA)	Control
Intervention	Sooma tDCS device, anode over left primary motor cortex (M1), cathode contralateral supraorbital; 2 mA current, 20 min/session, 5 days/week for 4 weeks; ramp-up/down 0.1 mA/s. Self-administered at home after single in-hospital training session.	Same electrode montage and device; current ramped up to 2 mA over 20 s (to mimic scalp sensation), reduced to 0.3 mA over 17 s and sustained at 0.3 mA for the remainder of the 20-min session before ramp-down. 5 days/week for 4 weeks.
N	22	22

Outcomes

Outcome	Type	Control	Intervention	P-value
Average daily NRS pain score in the week (week 5) after the 4-week treatment period, active vs low-intensity sham tDCS; mixed model, intention-to-treat	Primary	Sham tDCS average pain wk 5 — 4.45 ± 1.59 (Active/Sham seq) / 4.23 ± 1.33 (Sham/Active seq)	Active tDCS average pain wk 5 — 4.60 ± 2.02 (Active/Sham seq) / 4.67 ± 2.13 (Sham/Active seq)	0.89
Maximum daily NRS pain (wk 5)	Secondary	6.08 ± 1.63 / 5.89 ± 1.57	5.70 ± 2.25 / 5.75 ± 2.26	0.73
Average daily pain during wk 1 of treatment	Secondary	n/a	n/a (mixed model)	0.004
Average daily pain during wk 2	Secondary	n/a	n/a	0.02
Average daily pain during wk 3	Secondary	n/a	n/a	0.03
Average daily pain during wk 4	Secondary	n/a	n/a	0.046
BPI Short-Form pain score	Secondary	4.64 ± 1.89 / 3.98 ± 1.3	4.41 ± 1.77 / 4.68 ± 2.06	0.59
BPI interference score	Secondary	3.97 ± 1.48 / 4.35 ± 2.24	3.36 ± 2.01 / 3.39 ± 2.25	0.56
SF-36 physical functioning	Secondary	44.01 / 44.62	67.86 / 67.63	0.04
SF-36 general health	Secondary	36.39 / 35.00	38.21 / 37.73	0.33
SF-36 bodily pain	Secondary	45.00 / 47.50	52.86 / 56.36	0.40
HADS depression	Secondary	4.20 / 3.77	4.20 / 4.82	0.58
HADS anxiety	Secondary	5.90 / 5.77	5.67 / 5.55	0.88
NTSS-6 total score	Secondary	23.67 / 24.08	19.43 / 21.09	0.14
Within-group active vs baseline: average pain	Secondary	—	Baseline 5.3 ± 1.6 → 4.4 ± 1.6	0.03
Within-group active vs baseline: maximum pain	Secondary	—	Baseline 6.7 ± 1.5 → 5.7 ± 1.8	0.01
Within-group sham vs baseline: average pain	Secondary	Baseline 4.6 ± 1.8 → 4.4 ± 1.7	—	0.4
Within-group sham vs baseline: maximum pain	Secondary	Baseline 6.1 ± 1.9 → 5.8 ± 1.8	—	0.3
Responders (≥22.6% NRS reduction)	Secondary	3/22 (14%)	5/22 (23%)	not tested
Total adverse-event reports	Safety	17 occasions	22 occasions	>0.05
Serious adverse events (all considered unrelated to tDCS)	Safety	—	—
Reported AE categories	Safety	—	—
Dropout during study	Safety	38% during sham period; 8% before first treatment	54% of dropouts during active period

Subgroup Analysis

Responder vs non-responder analysis (n=5 vs 18 for active tDCS): active responders had lower pressure pain threshold at the foot (median 222.7 [IQR 183.3–284] vs 403 [318.3–514.3] kPa; p=0.02), a trend to higher temporal summation at the forearm (median VAS 3 vs 0; p=0.056) and foot (2 vs 1; p=0.09). No difference in conditioned pain modulation (63.3% vs 26.03%; p=0.4), sural amplitude/velocity, forearm PPT, or DN4. Interpretation: patients with baseline central sensitization and lower pain thresholds may be more likely to respond to 2-mA M1 tDCS.

Criticisms

Very small sample: only 22 of 35 randomized completed both arms (37% dropout), reducing power to detect between-group differences.
Single-center Danish study limited to type 1 diabetes with pDPN — generalizability to type 2 diabetes or other populations uncertain.
Chosen "sham" delivered a sustained 0.3 mA (≈15% of the active current) for 20 min/day for 4 weeks; recent EEG data suggest this may itself be biologically active, potentially inflating placebo-equivalent response and masking active-tDCS effect.
Blinding integrity was not formally assessed (no post-hoc blinding questionnaire).
Home-based, unsupervised administration introduces real-world variability in electrode placement and adherence; period-2 compliance was ~9.7% lower than period 1.
Analgesic medication use recorded but not adjusted for as a covariate; 27% reported changes in analgesic use during the trial.
Responder subgroup analysis based on only 5 active responders and 3 sham responders — exploratory and hypothesis-generating only.
Crossover design with short 4-week treatment periods limits detection of long-term efficacy; 6-week washout is a pragmatic compromise not supported by strong evidence.
Trial did not include an inactive (0 mA) sham arm, so the study cannot distinguish placebo effect from a genuine low-intensity neuromodulatory effect.

Funding

Sooma Medical (Helsinki, Finland) provided the tDCS devices but did not influence the design or reporting; statistical support from Aalborg University Hospital's Unit of Clinical Biostatistics.

Based on: tDCS-pDPN (Neuromodulation, 2026)