Osteoarthritis and neurological pain in dogs: does acupuncture versus sham needling improve pain scores? — MAYBE (weak evidence)

In dogs with osteoarthritis or neurological pain, does acupuncture vs sham needling improve validated pain scores?

Shelf #605 · gate-verified and displayed under the developer's blinded-trial rule

MAYBE (weak evidence): The sham/placebo-controlled evidence carded here shows no benefit on any objective outcome, with isolated improvements confined to single owner-completed instruments within otherwise null trials. Baker-Meuten 2020 (PMID: 32977836), a prospective, randomised, placebo-controlled, blinded crossover trial in 32 dogs with naturally occurring osteoarthritis, found no evidence of differences between acupuncture and placebo for ground reaction forces, activity counts or subjective orthopaedic scoring; the Client Specific Outcome Measures (CSOM) questionnaire improved for acupuncture versus placebo (P = 0.035) and versus baseline (P = 0.0002), but not for placebo versus baseline (P = 0.221). Kapatkin 2006 (PMID: 16649937), a randomised controlled crossover trial in 9 dogs with elbow osteoarthritis, found electrostimulated acupuncture versus sham treatment made no difference to ground reaction forces or owner visual analog pain scores (the treatment made no difference). Teixeira 2016 (PMID: 27767433), a randomised blinded trial in 47 dogs with hip dysplasia, found owner assessments did not differ significantly among acupuncture, carprofen and placebo capsule groups, and kinetic evaluation did not differ among groups or over time. Design class (a) for the osteoarthritis question: randomisation and blinding are established by a carded record answering the shelf's own comparison and population (Baker-Meuten 2020, whose title and abstract state 'randomized, placebo-controlled, blinded'). Class (a) design does not rescue the stamp: with objective outcomes null in all three trials and one positive owner-questionnaire contrast among many outcomes tested, the honest reading is MAYBE (weak evidence). For neurological pain specifically, no sham-controlled record was found at drafting: the electroacupuncture record in thoracolumbar disc disease (Hayashi 2007, PMID: 17867976) compared electroacupuncture plus standard medical treatment with standard treatment alone, with no sham arm, and is not carded. Drafted Nov 10, 2026 (VetMed build, topic #605); all figures from the abstracts fetched and verified at drafting.

Baker-Meuten 2020: acupuncture did not improve objective function in 32 dogs with osteoarthritis — no differences vs placebo in ground reaction forces, activity counts or orthopaedic scoring; one owner questionnaire (CSOM) improved vs placebo (P = 0.035)

Prospective, randomized, placebo-controlled, blinded clinical trial, cross-over design (record's own wording, in its title and abstract; publication types: Journal Article, Randomized Controlled Trial, Veterinary)

Who: Dogs: 32 client-owned dogs with naturally occurring osteoarthritis that completed a prospective, randomized, placebo-controlled, blinded clinical trial using a cross-over design. Dogs were assigned to receive placebo or acupuncture treatment once weekly for 4 weeks in random order, with a two-week wash-out period between treatment phases. Outcome measures were ground reaction forces (GRF), subjective orthopedic scoring (SOS), activity counts (AC) and owner-completed clinical metrology instruments (the Canine Brief Pain Inventory [CBPI] and Client Specific Outcome Measures [CSOM]), compared between baseline and 1 week after each treatment phase (activity counts compared over the final week of each phase).

Result: Evidence of differences between baseline versus acupuncture and placebo treatments was not identified for GRF, AC or SOS. Evidence of differences was identified for some of the clinical metrology instrument scores, including the CSOM questionnaire, which showed evidence of improvement when comparing baseline versus acupuncture (P = 0.0002) and placebo versus acupuncture (P = 0.035), but not between baseline versus placebo (P = 0.221). The authors conclude that the applied acupuncture protocol did not show improvement in function when using objective outcome measures for osteoarthritis in dogs, although certain metrology measurements recorded some degree of treatment response.

Watch out: The single positive contrast (CSOM, P = 0.035 for placebo vs acupuncture) is one owner-completed instrument among many outcomes tested (GRF, activity counts, orthopaedic scoring, CBPI and CSOM), and every objective measure was null — a pattern consistent with a chance or expectation effect in owner reporting rather than a functional benefit. The abstract does not specify what the placebo treatment consisted of, the trial is a crossover in 32 dogs with a two-week wash-out, and treatment lasted only 4 weeks per phase.

Baker-Meuten A, Wendland T, Shamir SK, Hess AM, Duerr FM. Evaluation of acupuncture for the treatment of pain associated with naturally-occurring osteoarthritis in dogs: a prospective, randomized, placebo-controlled, blinded clinical trial. BMC Vet Res. 2020;16(1):357. PMID: 32977836. Registry ID: none in the record. MeSH: Acupuncture Therapy/veterinary* · Animals · Dog Diseases/therapy* · Dogs · Female · Lameness, Animal/therapy · Male · Osteoarthritis/therapy, veterinary* · Pain/veterinary · Pain Measurement/veterinary

Kapatkin 2006: electrostimulated acupuncture versus sham made no difference in 9 dogs with elbow osteoarthritis — no significant effect on ground reaction forces or owner pain scores (the treatment made no difference)

Randomized, controlled, crossover clinical trial (publication types: Journal Article, Randomized Controlled Trial, Research Support, Non-U.S. Gov't)

Who: Dogs: 9 dogs with chronic forelimb lameness and radiographic evidence of elbow joint osteoarthritis secondary to elbow joint dysplasia, in a randomized, controlled, crossover clinical trial. After a 3-week control acclimation period with weekly gait analysis, dogs received electrostimulated acupuncture (ESA) once weekly for 3 weeks followed by a sham treatment once weekly for 3 weeks, or the sham treatment followed by ESA. Gait analysis was repeated prior to each treatment, and owners provided pain scores using a visual analog scale.

Result: Treatment (control, acupuncture, or sham) did not have a significant effect on ground reaction forces for any limb (the treatment made no difference). Owners of 8 of the 9 dogs were able to correctly guess the time period when ESA was delivered. The authors conclude that ESA did not have any significant effects on severity of lameness, as determined by measurement of ground reaction forces, or severity of pain, as determined by visual analog scale pain scores, in dogs with chronic elbow joint osteoarthritis secondary to elbow joint dysplasia.

Watch out: Only 9 dogs, and owner blinding was effectively broken — owners of 8 of 9 dogs correctly guessed when the real treatment was delivered — so the owner pain scores carry a high risk of expectation bias; the objective ground-reaction-force result, which is not subject to that bias, was null. The intervention is electrostimulated acupuncture, a specific modality, and the population is elbow osteoarthritis only.

Kapatkin AS, Tomasic M, Beech J, Meadows C, Boston RC, Mayhew PD, Powers MY, Smith GK. Effects of electrostimulated acupuncture on ground reaction forces and pain scores in dogs with chronic elbow joint arthritis. J Am Vet Med Assoc. 2006;228(9):1350-4. PMID: 16649937. Registry ID: none in the record. MeSH: Acupuncture Therapy/methods, veterinary* · Animals · Cross-Over Studies · Dog Diseases/therapy* · Dogs · Electric Stimulation Therapy/methods, veterinary* · Forelimb · Lameness, Animal/therapy · Linear Models · Osteoarthritis/therapy, veterinary* · Pain/epidemiology, veterinary* · Pain Measurement/veterinary · Severity of Illness Index · Treatment Outcome

Teixeira 2016: acupuncture was not significantly different from placebo in 47 dogs with hip dysplasia — owner assessments and gait kinetics did not differ among acupuncture, carprofen and placebo groups

Randomized, controlled clinical study with blinded treatment (record's own wording; publication types: Journal Article, Randomized Controlled Trial)

Who: Dogs: 54 dogs with hip dysplasia enrolled (7 removed; 47 treated and evaluated) plus 16 healthy dogs used to assess the evaluation protocol. Dogs with hip dysplasia were treated in a blinded manner for 30 days with acupuncture (once weekly for 5 sessions; n = 15), carprofen (4.4 mg/kg orally once daily; n = 16), or placebo capsules containing lactose (1 mg/kg orally once daily; n = 16). Dogs were evaluated 2 weeks and immediately before treatment (baseline) and at 2, 4 and 6 weeks after onset of treatment. Owners evaluated pain intensity with two validated questionnaires and a visual analogue scale (VAS) for pain, and lameness with a VAS for locomotion; kinetics of the hind limbs were also evaluated.

Result: Owners' assessments revealed that outcomes of the 3 treatments did not differ significantly. The Canine Brief Pain Inventory and VAS pain intensity assessments were decreased from baseline at weeks 4 and 6, respectively, but only in acupuncture-treated dogs. Locomotion VAS values were decreased at week 4 in acupuncture-treated and carprofen-treated dogs. Kinetic evaluation findings did not differ among the groups or over time. The authors conclude that neither acupuncture nor carprofen was significantly different from placebo; acupuncture and carprofen reduced subjectively evaluated lameness, and acupuncture was associated with a decrease in validated chronic pain scores (within-group changes from baseline).

Watch out: The placebo here is an oral lactose capsule, not sham needling — so the weekly acupuncture sessions, handling and owner attention were not matched in the placebo group, and any session effect is uncontrolled. The improvements reported for acupuncture are within-group changes from baseline, not differences from placebo: the between-group comparisons, which are the ones that answer this shelf's question, were not significant, and 7 of 54 enrolled dogs were removed from the study.

Teixeira LR, Luna SP, Matsubara LM, Cápua ML, Santos BP, Mesquita LR, Faria LG, Agostinho FS, Hielm-Björkman A. Owner assessment of chronic pain intensity and results of gait analysis of dogs with hip dysplasia treated with acupuncture. J Am Vet Med Assoc. 2016;249(9):1031-1039. PMID: 27767433. Registry ID: none in the record. MeSH: Acupuncture Therapy/veterinary* · Animals · Anti-Inflammatory Agents, Non-Steroidal/therapeutic use · Carbazoles/therapeutic use · Chronic Pain/therapy, veterinary* · Dog Diseases/therapy* · Dogs · Gait · Hip Dysplasia, Canine/therapy* · Humans · Lameness, Animal/pathology, therapy · Ownership · Pain Measurement/veterinary*

Synthesis: sham/placebo-controlled trials show no objective benefit of acupuncture for canine osteoarthritis pain, and no sham-controlled record exists for neurological pain — MAYBE (weak evidence)

Synthesis

Who: Synthesis of the three cards above: Baker-Meuten 2020 (32 dogs, naturally occurring osteoarthritis, randomised placebo-controlled blinded crossover), Kapatkin 2006 (9 dogs, elbow osteoarthritis, electrostimulated acupuncture vs sham crossover) and Teixeira 2016 (47 dogs, hip dysplasia, acupuncture vs carprofen vs placebo capsules). All concern osteoarthritis/hip-dysplasia pain; no carded record concerns neurological pain.

Result: (1) Objective outcomes: no difference from sham/placebo in any trial — ground reaction forces (Baker-Meuten 2020; Kapatkin 2006), activity counts and orthopaedic scoring (Baker-Meuten 2020), and hind-limb kinetics (Teixeira 2016) were all null. (2) Owner-reported outcomes: one positive contrast among many — CSOM improved for acupuncture vs placebo (P = 0.035) in Baker-Meuten 2020 — while owner visual analog pain scores were null in Kapatkin 2006 (with owners of 8/9 dogs able to guess the real treatment period) and owner assessments did not differ significantly among groups in Teixeira 2016. (3) Neurological pain: no sham-controlled record was found; the question is unanswered for that population. (4) Design class (a) for osteoarthritis: randomised, blinded, placebo-controlled evidence answering the shelf's own comparison exists (Baker-Meuten 2020) — and it is null on objective outcomes, which is why the stamp is MAYBE (weak evidence) rather than a benefit stamp.

Watch out: One-home check at drafting: PMIDs 32977836, 16649937 and 27767433 do not appear in any other pack on disk. Records deliberately not carded as claim-carriers: Hayashi 2007 (PMID: 17867976), a randomised study of electroacupuncture plus standard medical treatment versus standard treatment alone in 50 dogs with thoracolumbar intervertebral disk disease (overall success 23/26 vs 14/24) — it has no sham or placebo arm, so it cannot separate a needling effect from added treatment and attention, and it homes on an adjunct-therapy question if built; Laim 2009 (PMID: 19405884), electroacupuncture as an adjunct after hemilaminectomy with alternate (not random) assignment and no sham; and the gold-implant records at acupuncture points — Hielm-Björkman 2001 (PMID: 11688748, double-blind, 38 dogs, no significant difference from sham skin-piercing at non-acupuncture points) and Jaeger 2006 (PMID: 16731702, double-blind, 78 dogs, significantly greater improvement with gold bead implantation than placebo) — permanent implants are a different intervention from needling acupuncture, their results conflict with each other, and they home on their own question if built. A 2024 evidence review (PMID: 42005359, not carded) reached a similar conclusion for chronic pain: three of four placebo-controlled trials found acupuncture not efficacious versus placebo.

Synthesis of Baker-Meuten A et al. PMID: 32977836; Kapatkin AS et al. PMID: 16649937; Teixeira LR et al. PMID: 27767433. Registry IDs as on the individual cards.

Back to the library index · VetMed home

VetMed is an educational resource, not veterinary advice.