OUP user menu

Dexamphetamine treatment in stroke

D. Long, J. Young
DOI: http://dx.doi.org/10.1093/qjmed/hcg113 673-685 First published online: 18 August 2003


Reducing disability and dependency after a stroke is an important clinical objective. We examine what is known about the use of dexamphetamine in patients recovering from an acute stroke, and consider whether further clinical studies should be undertaken. Dexamphetamine has repeatedly been shown to enhance recovery after experimental brain injury in animals, the best effects being seen when dexamphetamine is combined with lesion-specific motor training or sensory stimulation. Postulated mechanisms for these beneficial effects in animals are in keeping with contemporary theories of neurophysiological rehabilitation in man. There have been few clinical studies of dexamphetamine during rehabilitation after an acute stroke. Four controlled trials demonstrated a tendency to an improved outcome when dexamphetamine was paired with therapy and administered 3–30 days after an ischaemic stroke. However, clinical studies to date have been small, included only highly selected patients, and have not addressed possible confounding effects of the drug on mood and untreated depression. Dexamphetamine has previously been used under supervision in medically ill patients and appears to be safe and well-tolerated. There is a need for well-designed studies to assess further the safety and efficacy of dexamphetamine in rehabilitation after stroke.


The amphetamines were first introduced into medical practice in 1935 for the treatment of narcolepsy. Subsequently, clinical indications escalated rapidly, and a 1946 review cited 39 possible uses.1 Besides being potent psycho-stimulant and anorectic agents,2 the drugs found a role in rehabilitation. Benefits have been reported in elderly patients responding poorly to physical therapy,3 in post-operative care,4 and in severe head injury when persistent deficits in initiation and attention limit progress.5 Dexamphetamine was also thought to improve functional deficits in patients with cerebrovascular disease.6 It was one constituent of a pill—the ‘Moran’—prescribed by Churchill’s personal physician for the ‘muzzy feelings’ he experienced after his first major stroke in 1953. Churchill is believed to have taken this tablet prior to public speaking in order to obtain short-term improvement in residual symptoms.6

There is evidence from animal models that the administration of dexamphetamine can facilitate recovery after experimental brain injury. The aim of this review is to establish what is known about the use of dexamphetamine in patients recovering from an acute stroke and to consider whether, and under what conditions, further clinical studies should be undertaken. The information presented results from a literature search based on Medline, Embase and the Cochrane register of controlled trials. Pharmaceutical sources were included and references cited in relevant articles traced.

Effects of dexamphetamine

Amphetamines are indirect sympathomimetic drugs that exert their action by stimulating the release of endogenous noradrenaline and other biogenic amines from storage sites in central and peripheral nerve terminals.7 Although the peripheral effects are mediated by noradrenaline, the central nervous effects are more complex and involve also the release of dopamine and possibly 5-hydroxytryptamine. The most commonly prescribed compound is the d-isomer, dexamphetamine. This is the most potent psycho-stimulant, and has fewer systemic effects than the racemic mixture. Dexamphetamine is rapidly absorbed from the gastrointestinal tract, and is widely distributed into most tissues with high concentrations in the brain and cerebrospinal fluid. The onset of drug action following oral administration of two 5 mg tablets is between 30 min and 2 h: subsequent clinical effects resolve rapidly because of the drug’s short plasma half-life (12–13 h) and would be expected to last 5–6 h.8

The central nervous effects of an oral dose within the usual therapeutic range of 5–30 mg include appetite suppression, increased arousal, elevation of mood, increased self-confidence, and improved concentration.7 Studies of healthy volunteers have demonstrated increased motor activity9 and improved learning10,,11 at similar doses. The current therapeutic indications for dexamphetamine are narcolepsy and attention-deficit hyperactivity disorder in children and young adults.

Experimental studies of dexamphetamine in brain injury

Investigations of mechanisms of recovery from experimental cerebral ischaemia identified a persisting deficit of noradrenaline in homogenized whole brain tissue in the rat12 and randomly selected cortical samples in the cat.13 Direct intra-ventricular infusion of noradrenaline, but not dopamine, facilitated recovery of hemiplegia in the rat.14 The possibility that pharmacological repletion of noradrenaline might alter outcome triggered a series of experiments to examine the effects of the sympathomimetic drug dexamphetamine in recovery from brain injury.

Dexamphetamine has repeatedly been shown to enhance recovery, the best effects being seen when the drug is combined with lesion-specific training or sensory stimulation. Initial experiments involved rats with a motor deficit induced by unilateral removal of motor cortex by suction.15 The administration of a single intra-peritoneal dose of dexamphetamine produced a statistically significant acceleration of motor recovery as assessed by walking along a narrow beam. A single dose of haloperidol—an antagonist at central noradrenergic and dopaminergic (D2) receptors—halted spontaneous recovery and reduced the response to dexamphetamine when the two drugs were given simultaneously. A similar benefit was shown in the recovery of locomotor function after frontal cortex ablation in the cat.16 Multiple doses increased the effect, but the absence of task-specific practice during dexamphetamine treatment delayed benefit. A closer experimental model for stroke, using photo-chemically induced thrombotic infarction in a restricted area of the rat somato-sensory cortex, demonstrated improvement in both the rate and completeness of re-learning of a task requiring sensory-motor integration of tactile information.17 In cats with complete loss of binocular depth perception following visual cortex ablation by suction, the administration of multiple doses of dexamphetamine with visual experience allowed complete recovery: this did not occur in controls or in those given dexamphetamine but housed in the dark.18

Current theories of recovery after brain injury emphasize the need for lesion-specific training as the stimulus for re-organization of activity within neural networks and re-instatement of lost function.19 Using intra-cortical microstimulation techniques in squirrel monkeys with a lesion in the hand motor cortex induced by bipolar electrocoagulation, Nudo et al. demonstrated that movements represented in the infarcted area did not re-appear spontaneously. When the experiment was repeated with specific training in skilled hand use, the loss of hand territory adjacent to the infarct was prevented, and in some cases hand representations extended into regions previously occupied by the elbow and shoulder.20 In animal models, repeated, spaced training sessions are required to achieve enduring changes in cortical function.19 This is consistent with proposed mechanisms of learning at a cellular level. Studies in invertebrates suggest that changes in synaptic function are initiated by a high frequency train of identical stimuli.21 This results in co-ordinated biochemical changes in the pre- and post-synaptic receptors so that the efficiency of future transmissions at the synapse is increased. In the presence of further spaced episodes of repetitive stimulation, structural modifications occur, again involving the pre- and post-synaptic receptors and their mechanism of interaction. Identified changes include increased dendrite length and greater numbers of dendritic branches and spines, thus increasing sites available for synaptic contact with other neurones. These changes permit a permanent change in function, a process known as long-term potentiation (LTP), the cellular basis of memory and learning.

Enhancement of the benefits of lesion-specific training by dexamphetamine is therefore feasible if noradrenergic stimulation facilitates adaptive plasticity within the central nervous system. There is experimental evidence that LTP is augmented in the rat by both noradrenaline and amphetamine.22 In addition, a recent study has provided evidence of neurite growth followed by synaptogenesis in a pattern corresponding both spatially and temporally with behavioural recovery accelerated by dexamphetamine treatment.23 Using immuno-histochemical techniques, Stroemer studied rats with reproducible cortical ischaemia produced by permanent ligation of the distal middle cerebral artery. The administration of repeated, spaced doses of dexamphetamine caused a significant increase in the expression of a protein (GAP43) found on axonal growth cones in the neocortex surrounding the infarction during the first 14 days after brain injury. Synaptophysin is present in presynaptic vesicles, and its concentration can be used to quantify nerve terminals during neural development. The distribution of immunoreactivity to this protein was increased in cortical regions both ipsilateral and contra-lateral to the infarcted area between 14 and 60 days after injury. This suggests that neural changes also occur in the undamaged cortex, and Stroemer includes a review of animal studies supporting a key role for the contra-lateral cortex in sustained recovery from brain injury. Behavioural sensitization after prolonged exposure to amphetamine provides a further example of an experience-dependent change in cerebral function. Repeated administration of escalating doses of dexamphetamine produced long-lasting (> 1 month ) changes in the length of dendrites, in the density of dendritic spines and the number of branched spines in neurones of the nucleus accumbens and prefrontal cortex, areas involved in the behavioural consequences of amphetamine abuse.24

As the benefits of dexamphetamine may be of rapid onset, they are unlikely to result solely from accelerated structural repair. Initial benefit may result from reversal of the widespread metabolic depression (diaschiasis) found in viable neurons close to, and remote from, an infarcted area: this is thought to be an important cause of immediate neurological deficit.25 In the rat, dexamphetamine increases glucose utilization in cortical areas surrounding acute brain injury over a time period similar to the observed effects on behavioural recovery.26 The locus caeruleus in the pons consists mainly of noradrenergic neurones, and has extensive projections throughout the cortex and to the cerebellum, brain stem and spinal cord.27 The region functions to maintain vigilance and responsiveness to novel stimuli, and may facilitate the response of cortical neurones to suboptimal signals when necessary. Its widespread connections provide a network able to mediate global changes in alertness and neuroplastic changes in areas remote from the site of injury. The importance of noradrenergic pathways in maintaining altered function in newly recruited pathways is supported by the observation in rats that administration of drugs that antagonize noradrenaline temporarily unmasks the symptoms of previous, and apparently resolved, experimental injury.28,,29

Many experimental models have been studied, but differences in the animal species used, and variation in the behavioural requirements of the task under investigation, make interpretation of the literature complex. The benefits of dexamphetamine after cortical injury depend on the behavioural requirements of the task to be re-learned.30 Schmanke found dexamphetamine facilitated the recovery of impaired beam-walking, a task requiring mainly motor activity: once locomotion had been initiated the animals successively added other behaviours enabling them to return to pre-operative performance. However, dexamphetamine had no effect on re-instatement of the ability to walk accurately over an elevated grid: the re-learning of this task requires primarily the use of somato-sensory and proprioceptive cues to inhibit incorrect placing responses. The site of injury is also important: dexamphetamine was unable to improve motor deficits produced by lesions in the cerebellar nuclei31 or substantia nigra.32 In a rat model using proximal middle cerebral artery occlusion, which produces extensive sub-cortical as well as cortical damage, the administration of dexamphetamine impaired performance on beam-walking and other neurological tests throughout the first month after injury.33 Finally, rats pre-treated with methamphetamine developed more extensive cerebral infarction after ischaemic cortical injury.34 This may explain the potential for neural toxicity in amphetamine abuse, but its relation to dexamphetamine’s beneficial effects after brain injury is not known.

Clinical studies of dexamphetamine in stroke

There have been few clinical studies in stroke primarily because of the potential for significant cardiovascular side-effects. The reported studies are small, of variable design quality, and have used different dosage schedules.

Motor recovery (Table 1)

Seven studies have investigated motor improvement. Crisostomo conducted a randomized double-blind trial in 8 subjects within 3 to 10 days of an acute ischaemic stroke.35 A single dose of dexamphetamine or placebo was paired with at least 45 min physical therapy. Two of the four subjects receiving active treatment showed noticeable clinical improvement on re-assessment 24 h later. Walker-Batson studied 10 hemiplegic patients enrolled between 16 and 30 days after an ischaemic stroke, and used 10 doses of 10 mg dexamphetamine at 4-day intervals, the doses again being closely paired with physical therapy.36 There was a statistically significant improvement in motor function in the treatment group 1 week after completion of dexamphetamine therapy, and this difference was still evident at 12 months. However, this was a single-blind study with the investigating neurologists and neuroscience nurse aware of treatment groups. The patients and physiotherapists were blinded during the treatment phase, but it is not clear who assessed final outcome, or whether any patients received further therapy during the follow-up period. Subjects were recruited over > 3 years from a pool of 400 patients, indicating considerable selection bias.

View this table:
Table 1

Studies of dexamphetamine in rehabilitation after ischaemic stroke (motor recovery)

nAge range (years)Clinical syndromeStudy designPaired with therapyDexamphetamine dose and treatment scheduleOutcome measureResultLength of follow-up
Crisostomo35 1988847–73Stroke with hemiparesis 3–10 days post-onsetDouble-blind, randomized, placebo- controlledYes10 mg single dose or placeboFugl-Meyer Motor Score42 [FMS]2 of 4 actively treated patients improved24 h after receiving trial medication
Walker-Batson36 19951048–73 (mean 65)Stroke with hemiparesis 16–30 days post onset.Single-blind, randomized, placebo- controlled.YesSingle 10 mg dose every 4th day for ten doses (15 mg in one patient).Fugl-Meyer Motor ScoreFMS improved at 1 week off drug treatment and at 12 months post stroke12 months after stroke onset
Mazagri39 199525Not statedWithin 72 h of ischaemic stroke with motor weaknessDouble-blind, randomized, placebo- controlled.Yes10 mg single dose or placeboFMS, Barthel Index, Canadian Neurological Scale43No difference at 2 days or 3 months after dosing3 months after receiving trial medication
Reding40 199521Mean 66Stroke with mixed deficits > 30 days post-onset.Double-blind, randomized, placebo- controlled.No10 mg daily for 14 days, then 5 mg daily for 3 days.Fugl-Meyer Motor Score, Barthel IndexNo significant differences4 weeks from study baseline
Vachalathiti41 200127Mean 62Stroke with hemiparesis mean 5.7 days post onsetDouble-blind, randomized, placebo- controlled.Not stated10 mg daily for 7 daysFugl-Meyer Score (for motor function and sensation)No significant differencesEnd of study (day 7)
Sonde37 200140Mean 77Stroke with hemiparesis 5–10 days post onsetDouble-blind, randomized, placebo- controlledYesSingle dose 10 mg d,l-amphetamine twice a week for 10 dosesFugl-Meyer Motor Score, Barthel IndexNo significant differencesAssessed after 10th session and 3 months after stroke
Martinsson38 200345Mean 67Carotid territory hemiplegic stroke within 72 h of onsetDouble-blind, randomized, placebo- controlledNoDaily dose for 5 days. Dose escalation study with placebo and three treatment groups: 2.5 mg b.d., 5 mg b.d., 10 mg b.d.LMAC44 and SSS45 (motor scores), Activity Index (AI), Barthel IndexSignificant improvement in LMAC, SSS and AI motor score until day 7Assessed on days 1, 3, 5 ,7 and at 1 and 3 months post-stroke

Sonde used a similar protocol in a group of elderly subjects enrolled within 5–10 days of a new stroke-related hemiparesis.37 Spaced doses of d,l-amphetamine were given 60 min prior to physiotherapy, but subjects with intra-cerebral haemorrhage were not excluded, and some patients were discharged to complete the trial medication at home. The authors found no significant difference in the rate of motor recovery to final assessment at three months. Martinsson performed a placebo-controlled dose-escalation trial designed to assess the safety of 5 consecutive days treatment with twice-daily dexamphetamine (2.5–10 mg b.d.) given within 72 h of an ischaemic stroke.38 A significant improvement in motor scores was evident from day 1 to day 7; this difference was not maintained at 3 months.

Three further studies have been published only in abstract format: the results need to be interpreted with caution, therefore. Mazagri reported a double-blind study of 25 patients given a single dose of dexamphetamine or placebo within 72 hours of an ischaemic stroke.39 The authors found no difference in the recovery of motor function assessed at 48 h and at 3 months. Reding40 detected no benefit in a controlled study comparing 12 patients given placebo and nine patients receiving 17 consecutive daily doses of dexamphetamine: no change was found in scores of motor function or depression. The patients were entered more than a month after stroke onset, and therapy sessions were neither standardized nor co-ordinated with drug treatment. Finally, Vachalathiti found no benefit in a study of 27 patients randomized to receive once daily dexamphetamine for 7 consecutive days starting on average 6 days after stroke-onset.41 It is not stated whether drug treatment was timed to coincide with therapy.

Recovery of speech deficits (Table 2)

Walker-Batson reported benefit in a group of six patients with aphasia.46 Spaced doses of dexamphetamine were given 30 min before 1 h of intensive speech therapy. Progress was assessed using the Porch Index of Communicative Abilities47 (PICA), a reliable and sensitive measure of changes in language, with which recovery at 6 months can be predicted using scores obtained at 1 month after stroke. Baseline scores were used to give a ‘corrected’ 1-month value from which future recovery was projected. Five of the six study subjects had achieved their projected 6-month score by 3 months, but total speech therapy input varied widely between individuals (from 12 to 114 h). Although this study was open-label with no placebo group for comparison, Walker-Batson has since replicated the findings in a double-blind controlled study using the same dosage schedule and assessment measure.48 One week after completion of drug treatment, subjects taking dexamphetamine achieved a significantly greater gain in PICA score than those randomized to speech therapy alone. At 6 months, there was still a trend to benefit, although the differences did not reach statistical significance. The groups were well matched for severity of initial deficit, but again total speech therapy input varied between individuals (21 to 111 h). Subjects randomized to dexamphetamine had a mean total therapy time of 66 h, compared to a mean of only 51.5 h for the control group. The extent to which this contributed to the improved outcome is unknown.

View this table:
Table 2

Studies of dexamphetamine in rehabilitation after ischaemic stroke (recovery of speech deficits)

nAge range (years)Clinical syndromeStudy designPaired with therapyDexamphetamine dose and treatment scheduleOutcome measureResultLength of follow-up
Walker-Batson46 1992634–71 (mean 54)Stroke with aphasia (single left middle cerebral artery [MCA] infarction) 10–30 days post-onsetOpen, uncontrolledYesSingle 10 mg dose every 4th day for ten doses. (15 mg in one patient)Porch Index of Communicative Ability47'Accelerated' recovery in 5 of 6 patients3 months after stroke onset
Walker-Batson48 20012141–71Stroke with aphasia (single left MCA infarction) 16–45 days post onsetDouble-blind randomized, placebo- controlledYesSingle 10 mg dose every 3rd or 4th day for ten doses.Gain in score of Porch Index of Communicative Ability (PICA) over baselineSignificant difference in PICA gain scores at 1 week off drug treatment: non-significant trend to benefit at 6 months after stroke onset6 months after stroke onset
Ziegler49 19933Not statedStroke with speech apraxia after left MCA infarction. Time interval not statedDouble-blind crossover studyYesSingle dose of 10 or 20 mg dexamphetamine or placebo on two separate daysSentence- and word-production tasks. Rapid syllable repetition'Differentia' effect on aspects of task. Improvement not sustainedRepeat assessment ‘several days' after drug dose
Muller50 19944Not statedStroke with speech apraxia after left MCA infarction. Time interval not statedDouble-blind crossover studyYes10 mg dexamphetamine or placebo on two separate daysMunich Intelligibility Profile51Improved articulation. Faster in choice reaction taskNone

Two small descriptive cross-over studies49,,50 published in abstract format have examined the effects of a single dose of dexamphetamine given to subjects with speech apraxia late after stroke. Speech performance was assessed using word and sentence production tasks and rapid syllable repetitions. The administration of dexamphetamine improved articulation and tended to produce faster reactions in a choice reaction task. The authors felt the changes reflected specific modulation of speech function rather increased arousal, but the benefits were not sustained.

Studies of other agents enhancing central monoaminergic neurotransmitters in recovery after stroke

Levodopa has been explored as an adjunct to physical rehabilitation in a recent double-blind, placebo-controlled study of 53 hemiplegic patients entered 3 weeks to 6 months after an ischaemic stroke.52 Levodopa is metabolized to dopamine and subsequently (at a rate of up to 5%) to noradrenaline. Three weeks of daily drug therapy given 30 min prior to physiotherapy resulted in a statistically significant improvement in motor recovery. This benefit was sustained to the end of the study period at 6 weeks. All subjects had severe motor deficit at entry, but the group receiving levodopa had greater initial symptom severity. The benefits of levodopa were independent of baseline motor function.

Stimulation of dopaminergic pathways has also been studied using bromocriptine given in combination with speech therapy in patients with stable non-fluent aphasia late after stroke (6 months to 8 years).53 Patients able to tolerate high doses of bromocriptine (30 mg t.d.s.) showed significant improvement, particularly in verbal latency and reading-comprehension tests. However, contraindications to bromocriptine excluded 14 of 25 eligible subjects from enrolment and a further five subjects were withdrawn because of intolerable side-effects.

Methylphenidate is an alternative psychostimulant used in attention-deficit hyperactivity disorder. It has clinical effects in common with dexamphetamine, but its mechanism of action and distribution within the central nervous system differ. Methylphenidate’s main action is to increase extracellular dopamine by binding to the dopamine transporter in the cell membrane of presynaptic vesicles to block reuptake,54 but it has appreciable effects on noradrenaline and possibly serotonin reuptake also. Grade reports the only prospective study of this medication in early post-stroke recovery.55 Twenty-one consecutive patients admitted to a rehabilitation unit following a new stroke received a 3-week course of either methylphenidate or placebo. Methylphenidate was given continuously, and the dose increased gradually from 5 mg daily to 15 mg twice daily, if tolerated. The results indicated a small treatment effect, with a tendency to improvement in both motor function and low mood.

Possible mechanisms of action in recovery after stroke

Increased levels of central noradrenergic activity are thought to mediate the beneficial effects of dexamphetamine in recovery after brain injury,56 although the evidence for this is limited. Goldstein analysed data collected during a prospective multicentre study of a neuroprotective drug (GM1 ganglioside) in acute stroke.57 The progress of patients in the control group who were coincidentally receiving centrally-acting drugs known to impair outcome in animals was compared to that of control subjects not taking such medication. The drugs considered detrimental to stroke recovery were antihypertensives that reduced central noradrenergic activity, neuroleptics, anticonvulsants and, finally, benzodiazepines, which were given to the majority (72%) of the 37 patients in this group. Measures of motor recovery in the upper and lower limb showed a trend to improvement in the group not taking the ‘detrimental’ drugs, with a statistically significant improvement in Barthel score at 56 and 84 days post-stroke in this group. The use of haloperidol, an antagonist at central noradrenergic and dopaminergic receptors, is recognised to impair functional recovery after head trauma.58

Studies using transcranial magnetic stimulation (TMS) and neuroimaging with positron emission topography (PET) demonstrate that task-specific training results in re-organisation of neuronal activity in the human motor system.59 Also, as in animal models, the unaffected cortex is thought to play an important part in sustained recovery from brain injury.60–,62 Dexamphetamine has been shown to augment cortical plasticity in healthy volunteers. A 30-min period of training involving repetitive thumb movements produced a (temporary) change in the direction of the subjects’ motor response to TMS of the relevant cortex.63 In a double-blind controlled study using this technique, the administration of a single dose of dexamphetamine (10 mg) prior to motor training produced a statistically significant reduction in the length of time needed to develop the new response, and this alteration in cortical function persisted significantly longer.64 In addition, this model has been used to show that drugs known to interfere with synaptic plasticity and the development of LTP in animal models also interfere with the induction of use-dependent cortical plasticity in human subjects.65 This suggests that the mechanisms underlying these two processes are similar.

In a double-blind trial of dexamphetamine (20 mg) and placebo, Uftring used functional MRI (fMRI) to demonstrate an increase in the volume of brain activated during a finger tapping exercise and auditory stimulation. The results suggested dexamphetamine increased recruitment of neural pathways, but only in areas also active in the placebo group.66 Unfortunately, the rate of finger tapping in the placebo and treatment groups was not recorded, and a consequent confounding effect on the degree of cortical activation cannot be excluded. Using PET scanning, Mattay measured regional cerebral blood flow while subjects performed a variety of neuropsychological tasks after dexamphetamine (0.25 mg/kg) or placebo administered in a double-blind crossover trial.67 The administration of dexamphetamine did not have a global effect on cerebral blood flow, but selectively increased perfusion in the cortical areas related to the task being performed.

Unwanted effects of dexamphetamine

Adverse effects of dexamphetamine are primarily dose-related and rare at oral doses of < 15 mg. Toxic reactions may, however, be idiosyncratic and occur after only 2 mg.7 The behavioural and neurophysiological effects of therapeutic doses vary between subjects, and are influenced by an individual’s baseline level of cognitive functioning.68

Table 3 summarizes the results of a search for information on the likely incidence of side-effects in medical patients. The studies presented have given full details of unwanted effects and included at least 10 subjects, but are not all placebo-controlled. The most common central side-effects are of insomnia (which can be avoided by dosing early in the day) and agitation or anxiety.69,,70 Confusion, altered behaviour and delusions may occur, particularly in patients with cognitive impairment or previous psychiatric illness. Caution is necessary in individuals with a past history of drug or alcohol dependence, but abuse and dependence are unlikely in the context of short courses of supervised treatment.71 Severe rebound depression with suicidal intent may occur on withdrawal after extended use.

View this table:
Table 3

Studies reporting the incidence and nature of side-effects experienced by patients prescribed dexamphetamine for depression or as an adjunct in rehabilitation

Number of patients receiving dexamphetamine and dose-range.Age range (years)Clinical syndromeStudy designTotal incidence of side-effects leading to discontinuationNumbers reporting specific side-effects*
Overal70 196629 patients, 15 mg dailyNot statedPsychiatric in-patients with depressionOpen prospective compared with tranylcypromineNot statedInsomnia (7), nausea (2), restlessness (7), constipation (3), dry mouth/throat (6), dizziness (3), palpitations (3), hypotension (1)
Clark3 198888 patients, stepwise increase from 2.5 mg b.d. to 10 mg b.d.66–94Mixed conditions with 'poor motivation'Open, uncontrolled.26%Aggressive/uncooperative behaviour (10), confusion or delusions (7), hypomania (1), atrial tachycardia (1), vomiting (1), constipation (1)
Masand81 1991154 patients, 2.5–30 mg daily (mean 9 mg)Mean 65 (78% > 60)Medical and surgical in-patients with secondary depressionRetrospective case note reviewApprox 10% (side- effects reported in 15% but not all led to discontinuation)Anxiety/agitation (6), hypomania (3), confusion (3), delusions (3), hallucinations (1), hypertension (1), atrial fibrillation (1), sinus tachycardia (2), insomnia (1), spasticity (1), nausea (1)
Hornstein5 199627 patients, 5–30 mg daily15–75 (mean 36)Mixed brain injury (trauma or cerebrovascular disease)Retrospective case note reviewNoneAnxiety (1), insomnia (1), agitation (1), restlessness (2), tachycardia (1)
Wagner72 200020 patients, 5–40 mg18–65HIV-positive men with depression and fatique.Double-blind for 2 weeks, then open with placebo non- responders included.20%Overstimulation (5), insomnia (5), palpitations (3), loss of appetite or weight (5)
Unwin82 200028 patients, 10 mg every 3rd or 4th day for 10 dosesMean 612–6 weeks after onset of thrombo-embolic strokeRandomized, placebo controlledNoneBlood pressure recordings during and following administration of dexamphetamine showed no significant change (patients with uncontrolled hypertension excluded).
Martinsson38 2003Total of 45 patients in study. 10 patients given each of 2.5 mg b.d., 5 mg b.d. or 10 mg b.d. for 5 daysMean 67Within 72 h of acute ischaemic strokeDouble-blind, randomized, placebo controlled2/28 treated patients withdrawn: icterus (1) and hallucinations (1)Dose-dependent significant increases in systolic BP (up to 16 mmHg), diastolic BP (up to 10 mmHg) and pulse rate (up to 17 bpm) at high dose vs. placebo. Insomnia, headache, dry mouth and nausea more frequent in treated groups.
  • *Individual patients reported more than one symptom in some studies.

Hypertension and cardiac arrhythmias are the peripheral side-effects of concern but, because of the drug’s short half life, these resolve within 24 h of discontinuation or respond to dose reduction.3,,72 Nausea, abdominal discomfort, altered bowel habit and urinary retention may occur. Amphetamines may raise intra-ocular pressure by pupillary dilatation, and are contra-indicated in glaucoma.7 Although psychostimulants are potentially epileptogenic, they have been safely prescribed in head-injured patients with seizures.73 Dexamphetamine has been used in patients with resistant seizures to overcome the sedative effects of high doses of anti-convulsant medication.74 Stroke is a rare consequence of amphetamine abuse75 and can occur in infrequent users—it has been reported after ingestion of a single dose of 40 mg of a racemic mixture of amphetamine.76 Haemorrhagic stroke is more common, but infarction is also recognised.77,,78

Despite the potential for serious toxicity from prolonged or high-dose illegal use, reviewers have consistently concluded that lower doses of dexamphetamine (5–30 mg daily) given under supervision in medically-ill adults seldom cause clinical problems.71,79–,81 Significant cardiotoxicity has not been reported in studies of relatively unselected patients.3,,81 Unwin and Walker-Batson reported no events in their experience of 50 patients post-stroke, but subjects with known cardiac disease were excluded.82 Martinsson demonstrated a dose-dependent increase in systolic and diastolic blood pressure and pulse rate.38 Hypertensive patients were included in this study, but there appeared to be no increase in adverse events attributable to these drug-related haemodynamic changes.

Recommendations for future studies

It seems inappropriate to abandon the possibility that the beneficial effects of dexamphetamine after experimental brain injury in animals may be reproducible in man. Further well-designed, placebo-controlled trials are required, and it is reassuring that psychostimulants have previously been used safely under supervision in medically ill patients.

Studies suggesting clinical benefit in ischaemic stroke have given dexamphetamine within 30 days of onset. However, dexamphetamine is able to enhance use-dependent plasticity in normal volunteers,64 and further trials of dexamphetamine in combination with intensive physical therapy more than a month after stroke-onset would be worthwhile. It is not clear whether spaced doses are necessary to avoid habituation during a course of treatment lasting only a few weeks. Multiple doses have been given at intervals in animal studies because of concern that prolonged daily administration may predispose to excitotoxic neuronal death.23,,24 However, in human subjects with narcolepsy, daily dexamphetamine is given safely over many years without evidence of cumulative toxicity, although the therapeutic dose required may increase slowly with time.83 Neurophysiological effects would be expected to peak up to 2–3 h after oral administration, and therapy sessions timed 60 min after dosing may miss peak drug levels. It may be that when rehabilitation occurs in a therapeutic environment such as a stroke unit, the precise co-ordination of peak drug levels with physical therapy sessions is less critical, and the effects of continuous twice daily dosing should be explored.

Study subjects will require careful medical supervision in a setting where co-ordinated multidisciplinary care and standardized physical therapy sessions can be provided. Full functional assessment using validated instruments should be performed before and during the trial, and for at least 3 months thereafter. Consideration should be given to assessment of mood, particularly with regard to depression and apathy. Apathy84 and depression85 both occur commonly after a stroke, and remission or treatment of post-stroke depression is associated with improved functional outcome.86–,88 Although dexamphetamine has no role in the modern treatment of depression, its acute mood-elevating effects have been used effectively in medically-ill patients72,80,81,,89 and are of rapid onset.80,,81 It is possible that dexamphetamine may coincidentally improve mood and so augment participation in—and gain from—physical therapy, and formal assessment of depression, apathy and cognition before, during and after drug treatment would be worthwhile.

Although the potential benefits of dexamphetamine in stroke-recovery have been attributed to augmentation of central noradrenergic pathways, stimulation of the release of other monoaminergic neurotransmitters, particularly dopamine and possibly 5-hydroxytryptamine (serotonin), also occurs.7 Scheidtmann’s study,52 showing an equivalent benefit in motor recovery when levodopa is paired with physical therapy, is important. In addition, there is preliminary evidence that serotinergic stimulation by the administration of the selective serotonin reuptake inhibitor fluoxetine also enhances motor performance after stroke.90 Future studies using newer drugs with more specific effects on predominantly a single monoaminergic neurotransmitter system may help clarify the underlying biochemical mechanisms and, because of a wider safety margin, allow the recruitment of greater numbers of subjects.

The National Institute for Neurological Disorders and Stroke (NINDS) is currently sponsoring two trials looking at dexamphetamine as an adjunct to rehabilitation after acute ischaemic hemiplegic stroke. The Amphetamine Enhanced Stroke Recovery trial (AESR), a multi-centre study hoping to recruit up 130 patients, will provide further information on the safety profile and potential benefit of dexamphetamine combined with physical therapy in promoting motor recovery.91 The second study hopes to identify the areas of the brain involved in enhanced recovery: all subjects will undergo fMRI scanning, EEG and transcranial magnetic stimulation studies at intervals up to 1 year post-treatment.92 Finally, a clinical and fMRI study looking at dexamphetamine coupled with physiotherapy for moderate to severe hemiparetic stroke, sponsored by the Canadian Institutes for Health Research at the University of Toronto, is nearing completion.


The only current indications for treatment with dexamphetamine are narcolepsy and hyperactivity disorders in children. Several types of animal study have suggested that dexamphetamine, a potent noradrenergic stimulant, may have a role in improving recovery after brain injury. The limited clinical data are inconsistently supportive, but studies to date have been small, used different dosage schedules and included patients with different pathologies at varying times post-stroke. Possible confounding effects on mood and untreated depression have not been assessed. There is a need for well-designed pilot studies to establish safety in less selected subjects, and to determine the patient characteristics and treatment schedules for a larger randomized controlled trial. The experimental evidence that dexamphetamine may enhance brain recovery through increased axonal sprouting and growth is especially interesting. These effects are consistent with the concept of neuroplasticity in brain recovery and the biological basis for contemporary neurophysiological rehabilitation. Synergy between the chemical and physical treatments is therefore plausible. However, which patients (if any) benefit most, and when, remains to be determined.


View Abstract