Objective: To systematically assess benefits and harm of non-pharmacologic interventions for diabetic peripheral neuropathy (DPN) symptoms.Methods: MEDLINE, Embase, and the Cochrane Central Register of Controlled Trials were searched from 1966 to May 24, 2016 for randomized controlled trials. Two reviewers evaluated studies for eligibility, serially abstracted data, evaluated risk of bias, and graded strength of evidence (SOE) for critical outcomes (pain and quality-of-life).Results: Twenty-three trials were included. For pain, alpha-lipoic acid was more effective than placebo (moderate SOE) and frequency-modulated electromagnetic stimulation was more effective than sham (low SOE) in the short-term but not the long-term. Electrical stimulation (including transcutaneous) was not effective for pain (low SOE). Spinal cord stimulation was more effective than usual care for pain (low SOE), but had serious complications, and studies had no sham arm. Evidence for cognitive behavioral therapy and acupuncture was insufficient; no exercise or physical therapy trials met inclusion criteria. No interventions reported sufficient evidence on quality-of-life. Most studies were short-term with unclear risk of bias.Conclusions: Alpha-lipoic acid and spinal cord stimulation were effective for pain; studies were short-term with quality deficits. Spinal cord stimulation had serious adverse events. Further research should address long-term outcomes and other non-pharmacologic treatments.
Keywords: Diabetes; Diabetic peripheral neuropathy; Pain.