Effect of TENS on Sensory Parameters in Diabetic Polyneuropathy

BACKGROUND AND RATIONALE

Distal symmetric polyneuropathy is the most common chronic complication of diabetes and the principal cause of the sensory loss that precedes foot ulceration. Current management is directed almost entirely at symptomatic control of neuropathic pain and is not expected to modify sensory function. Transcutaneous electrical nerve stimulation (TENS) has been studied in this population, but trials have used pain intensity or composite symptom scores as endpoints, and cutaneous sensory function has rarely been treated as a prespecified outcome.

Stimulation frequency provides a mechanistic rationale for outcomes that may differ in kind. Conventional high-frequency, sensory-intensity TENS preferentially recruits large-diameter, low-threshold A-beta afferents, whereas low-frequency, higher-intensity stimulation is generally attributed to endogenous opioid-mediated descending inhibition. The A-beta population engaged by the former is the same population that mediates detection of weak mechanical stimuli and that monofilament testing assesses. Converging evidence from the repetitive sensory stimulation literature indicates that prolonged, passive peripheral input can lower tactile thresholds through long-term-potentiation-like mechanisms, with a frequency dependence paralleling the induction of synaptic potentiation and depression.

DESIGN

Single-centre, randomised, double-blind, sham-controlled, parallel-group trial with three arms allocated in a 1:1:1 ratio. Participants were recruited from the Endocrinology and Metabolism outpatient clinic of Akdeniz University Faculty of Medicine; assessments were performed at the Akdeniz University Sports Sciences Research and Application Centre.

ALLOCATION AND BLINDING

Allocation was performed by a lot-drawing procedure after written informed consent and before baseline assessment. Simple randomisation was used, without blocking or stratification. Participants and outcome assessors were blinded to group assignment. The sham arm followed an identical procedure at a sub-therapeutic intensity, so participants were unaware of whether they received active or sham stimulation. All sensory, nociceptive, and questionnaire-based assessments were performed by evaluators blinded to allocation.

INTERVENTIONS

TENS was delivered with a dual-channel PRIMERA TENS/NMES stimulator (DJO, LLC, USA) using standard 2-inch self-adhesive electrodes, one placed on the calf and one on the plantar surface of the foot. A zero-direct-current, asymmetric biphasic rectangular waveform was used. Intensity was individually adjusted to each participant's sensory tolerance and maintained at a comfortable level throughout each session. Sessions were delivered five days per week for eight weeks in all arms.

ASSESSMENTS

Assessments were performed at baseline and repeated in the week following the eight-week programme. The primary outcome was the tactile detection threshold at the plantar surface of the great toe, measured with Semmes-Weinstein monofilaments (index 1.65-6.65), with the corresponding buckling force recorded in grams; a higher value indicates a more impaired threshold. All monofilament measurements were obtained by a single blinded investigator.

Secondary outcomes were exploratory and comprised electrical pain detection and tolerance thresholds obtained by transcutaneous sural nerve stimulation using an ascending method of limits; the nociceptive flexion reflex recorded by surface electromyography from the short head of the biceps femoris; pressure pain threshold and tolerance measured by digital algometry at the dominant forearm and leg; and patient-reported measures of neuropathic pain, pain quality, depressive symptoms, and quality of life.

STATISTICAL APPROACH

Non-parametric methods were planned throughout because of small, unequal group sizes and the expected right-skew of several sensory variables. Continuous variables are summarised as median with interquartile range. Within-group change was tested with the Wilcoxon signed-rank test and between-group differences at post-treatment with the Kruskal-Wallis test, followed for the primary outcome by prespecified Mann-Whitney contrasts of each active arm against sham. Benjamini-Hochberg false-discovery-rate correction was applied across the exploratory secondary family.

Inclusion criteria

• Type 1 or type 2 diabetes mellitus
• Clinically confirmed diabetic peripheral polyneuropathy of at least five years' duration, diagnosed by an endocrinologist during routine outpatient follow-up on the basis of distal symmetric sensory symptoms together with abnormal bedside sensory testing
• Leeds Assessment of Neuropathic Symptoms and Signs (LANSS) score of 9 or above
• Standardized Mini-Mental State Examination score above 24
• Glycated haemoglobin (HbA1c) below 8.5%
• Wagner grade 0 feet (intact plantar skin without ulceration)
• Willing to continue prescribed diet and hypoglycaemic treatment unchanged throughout the study

Exclusion criteria

• Use of a cardiac pacemaker
• Alcohol dependence
• Peripheral arterial disease
• Renal failure
• Hepatic failure
• Hypothyroidism
• Vitamin B12 deficiency
• Folate deficiency
• Sarcoidosis
• Active infection
• Malignancy
• Coagulopathy
• Skin lesion, discolouration, or allergic reaction at the electrode sites