J Pain Res. 2026 Oct 30;19:630783. doi: 10.2147/JPR.S630783. eCollection 2026.

ABSTRACT

PURPOSE: To compare the 12-month effects of heat-conduction dry needling (HCDN) and transcutaneous electrical nerve stimulation (TENS) on pain and physical function in early-to-mid-stage knee osteoarthritis (KOA).

PATIENTS AND METHODS: In this single-center, parallel-group randomized clinical trial, 47 participants were allocated to HCDN (n=24) or TENS (n=23) for 4 weeks. Primary outcomes were WOMAC total and subscale scores; numerical rating scale (NRS) pain was prespecified. Secondary outcomes were 30-second Chair Stand Test, safety, and exploratory MRI cartilage volume. Seven participants had baseline data only. Fifty multiply imputed datasets were generated by chained equations with predictive mean matching. Linear mixed-effects models included group, categorical time, and group-by-time fixed effects and a participant-specific random intercept; estimates were pooled by Rubin’s rules.

RESULTS: Group-by-time interactions were significant for NRS pain (pooled Wald χ²[4]=13.48, P=0.009), WOMAC pain (χ²[4]=46.97, P<0.001), WOMAC total (χ²[4]=46.68, P<0.001), and WOMAC activities of daily living (χ²[4]=40.41, P<0.001). At 12 months, HCDN-TENS differences were -1.58 points for NRS pain (95% CI -2.36 to -0.79; Holm-adjusted P<0.001) and -3.70 points for WOMAC pain (95% CI -5.01 to -2.38; Holm-adjusted P<0.001). A ≥50% NRS reduction occurred in 69.8% of HCDN and 27.3% of TENS participants (risk ratio 2.58, 95% CI 1.17-5.65). No serious adverse events occurred.

CONCLUSION: HCDN was associated with more sustained reductions in pain and improvements in patient-reported function than TENS. Larger multicenter trials with sham or usual-care comparators are required.

PMID:42979458 | PMC:PMC13729857 | DOI:10.2147/JPR.S630783