Electrical acupoint stimulation of the affected arm in acute stroke: a placebo-controlled randomized clinical trial.

Objectives: To determine whether adding electrical stimulation of upper limb acupoints to conventional rehabilitation during acute stroke could produce greater and longer lasting motor improvements of the arm.

Design: Double-blind, randomized, placebo-controlled trial.

Setting: Acute stroke wards, followed by rehabilitation hospitals and subjects' residences.

Participants: Seventy-three patients ≤ 46 hours post stroke onset with moderate to severe weakness in the arm contralateral to the side of stroke.

Intervention: All subjects received conventional rehabilitation. Twenty-nine received additional electrical stimulation, 21 received additional placebo-electrical stimulation and 23 received conventional rehabilitation only, as control. Electrical stimulation or placebo-electrical stimulation was applied to acupoints GB20, LI15, LI11, LI10 and LI4, 60 minutes a day, five days a week, for four weeks.

Measurements: Primary outcome measures were hand grip and pinch strength, with Action Research Arm Test (ARAT) as secondary outcome measure. These were assessed on the affected arm at recruitment, then 4 (W4), 12 (W12) and 24 weeks (W24) afterwards.

Results: Post-hoc analysis showed that the electrical stimulation group had greater improvements than the control group in hand grip (P = 0.015) and pinch strength (P = 0.007) at W4, with the gains maintained at W12 and W24. In contrast, the placebo-electrical stimulation group did not differ from either the control or the electrical stimulation group. Between-group improvements in ARAT scores from baseline to W24 (by 16.8 in control, 27.6 in placebo-electrical stimulation group and 26.3 in electrical stimulation group) were not significant.

Conclusions: Adding four weeks of electrical stimulation during acute stroke appears to produce greater and longer lasting hand grip and pinch strength improvements than administering conventional rehabilitation alone.