Australian Journal of Physiotherapy 2005 Vol. 51
221
Introduction
Hemiparesis is common following stroke. Reduced upper
limb function impacts on ability to perform activities of daily
living (Page et al 2004), which is likely to reduce
independence and increase burden of care. Constraint-
induced movement therapy (CIMT) is a family of techniques
that have been implemented to increase the amount and
quality of function of an affected upper limb. These
techniques involve restraint of the intact limb over an
extended period, in combination with a large number of
repetitions of task-specific training of the affected limb.
CIMT evolved when it was observed that monkeys stopped
using an affected upper extremity immediately after unilateral
deafferentation by dorsal rhizotomy and never spontaneously
resumed use (Knapp et al 1958, Knapp et al 1963, Taub
1977). It was proposed that animals could learn ‘non-use’
when attempts to use an affected limb immediately following
injury are unsuccessful. Subsequent research identified that
use of a deafferented arm could be induced by either
immobilising the unaffected arm for a period of days or by
training the affected arm (Taub 1980). When movement was
restricted for a period of one to two days, the animal would
use the limb while the restriction was in place but revert to
non-use as soon as it was removed. However, when the
restriction was imposed for a period of one to two weeks, use
of the affected upper extremity was maintained after the
restriction was removed. This mechanism of learned non-use
is thought to apply in humans who suffer hemiparesis
following stroke, where the initial period of motor
incapacitation is due to cortical injury (Taub et al 1999).
The first investigation into the effects of CIMT on humans,
involving both training of the paretic upper limb (six hours a
day each weekday for two weeks) and restraint of the
contralateral upper limb (90% of waking hours for 14
consecutive days) was described by Taub (1980). CIMT has
continued to evolve and now constitu