Amphetamine

Amphetamine is an organic molecule that is an indirectly acting sympathomimetic, with its d-isomer, dexamphetamine, being three to four times more active than the l-isomer. Its mechanism of action is thought to be the increasing of noradrenaline and dopamine by blocking their reuptake and facilitating their release at central presynaptic neurons. It is also a weak monoamine oxidase inhibitor (Reference Chiarello and ColeChiarello 1987; Reference Masand and TesarMasand 1996). The stimulation of the dopaminergic mesocorticolimbic and nigrostriatal pathways may be responsible for the motor and behavioural effects (Reference Homsi, Walsh and NelsonHomsi 2000; Medsafe 2007).

Dexamphetamine

Dexamphetamine is rapidly absorbed and readily crosses the blood–brain barrier. Its central effects include anorexia, vasoconstriction and hyperthermia. Peripherally, it can elevate diastolic and systolic blood pressure and has a weak bronchodilatory action (Reference Homsi, Walsh and NelsonHomsi 2000). In healthy people, it increases wakefulness, alertness, motivation and initiative, and elevates mood (Reference Masand and TesarMasand 1996). Given its enhanced potency, the d-isomer is most commonly used in clinical practice and nearly all of the studies of psychostimulant use for depression have used dexamphetamine.

Methylphenidate

Methylphenidate is a piperidine derivative and its exact mechanism of action is unknown. It binds to dopamine transporters on the presynaptic membrane, particularly in the striatum, resulting in an increase in extrasynaptic dopamine levels. It may also regulate dopamine homeostasis at the nerve terminal and has some effect on noradrenaline reuptake and binds weakly to the serotonin transporter. Unlike dexamphetamine, it does not facilitate dopamine release at the nerve terminal (Reference Challman and LipskyChallman 2000). Clinically, methylphenidate is considered a milder psychostimulant than the amphetamine derivatives.

Modafinil

Modafinil is a mild psychostimulant derived from benzhydrylsulfinylacetamide and is structurally unrelated to methylphenidate and amphetamine. Its mechanism of action is unknown and evidence suggests that it is dissimilar to older psychostimulants. Animal studies indicate that modafinil's primary activity may be related to a decrease in gamma-aminobutyric acid-mediated neurotransmission. Other in vitro studies have demonstrated that modafinil also interacts with other neurotransmitter pathways such as glutamine in the hypothalamus, serotonin in the prefrontal cortex and hypothalamus, dopamine in the prefrontal cortex and noradrenaline in the ventrolateral preoptic nucleus. It also increases neuronal activity selectively in the hypothalamus, activating tuberomammillary nucleus neurons, which release histamine and neurons in the perifornical area, in turn releasing orexin/hypocretin. These cell groups are associated with the control of wakefulness (Reference Keating and RaffinKeating 2005). Modafinil is proposed to have less potential for misuse compared with methylphenidate (Reference JasinskiJasinski 2000).

Nearly all of the studies discussed below have used standard-release formulations of the psycho-stimulants. Table 1 shows the pharmacokinetics of dexamphetamine, methylphenidate and modafinil.

Monotherapy

There have been several reviews of psychostimulants as monotherapy for primary unipolar depression (Reference Satel and NelsonSatel 1989; Reference Orr and TaylorOrr 2007). Most RCTs have compared the older psychostimulants with placebo or anti-depressants such as imipramine and phenelzine. There have been no RCTs with modafinil. Most of the trials are now more than 20 years old and have many methodological shortcomings, such as varying diagnostic criteria and the use of unvalidated rating scales or global impression alone (Reference Orr and TaylorOrr 2007). Most studies in non-elderly adults (under 65 years of age) were negative and, in the positive studies, stimulants were only modestly beneficial to certain subgroups compared with placebo.

In contrast to this there have been some early, but less than methodologically robust, positive trials of monotherapy psychostimulant use to treat depression in elderly people (Reference JacobsonJacobson 1958; Reference KaplitzKaplitz 1975). The most recent RCT (Reference Wallace, Kofoed and WestWallace 1995) was positive but involved only 16 medically ill patients with depression in a very short trial.

Overall, this treatment option has a poor evidence base. It may have some utility for elderly people or for treating depression in the medically unwell; this will be discussed below.

Adjunctive treatment

There are several positive open trials, audits and case reports regarding the use of psychostimulants to augment standard antidepressant treatment, including tricyclic antidepressants (Reference Gwirtsman, Szuba and TorenGwirtsman 1994), monoamine oxidase inhibitors (Reference FeinbergFeinberg 2004), selective serotonin reuptake inhibitors (Reference Ninan, Hassman and GlassNinan 2004), venlafaxine (Reference Masand, Anand and TanquaryMasand 1998) and mirtazapine (Reference Schillerstrom and SeamanSchillerstrom 2002).

The two negative RCTs using methylphenidate as an augmentation agent compared with placebo in adults with major depression showed no statistically significant improvement or accelerated response (Reference Postolache, Rosenthal and HellersteinPostolache 1999; Reference Patkar, Masand and PaePatkar 2006). Reference Patkar, Masand and PaePatkar and colleagues (2006) noted that in 60 out-patients with treatment-resistant depression, more participants responded to extended-release methylphenidate compared with placebo (40% v. 23.3%). However, this difference was not statistically significant. There were no differences in the rate of adverse events or the number of participants who dropped out of the trial.

Reference Lavretsky, Park and SiddarthLavretsky and colleagues (2006) conducted an RCT involving 16 elderly participants with major depression, comparing methylphenidate with placebo augmentation, commenced simultaneously with citalopram. There was an accelerated anti-depressant response by week 3 and a greater reduction in depression scores in the participants on the combination treatment. However, side-effects and intolerability caused a number of participants to drop out of the trial.

There have been two well-designed RCTs comparing modafinil with placebo as an augmentation agent to antidepressants in the treatment of non-elderly people with major depression. Participants in these trials had experienced a partial response to an antidepressant, usually a selective serotonin reuptake inhibitor, and had significant and persistent fatigue or sleepiness. At the end of both trials, there was no significant difference in depression scores between the two groups. A 6-week trial (Reference DeBattista, Doghramji and MenzaDeBattista 2003) involving 168 patients reported a significant improvement compared with placebo for sleepiness at week 1 and fatigue at week 2, but not at the end of the trial. An 8-week trial (Reference Fava, Thase and DeBattistaFava 2005) including 311 patients reported a non-significant trend towards greater mean reductions on depressive scores at the end of the trial for the modafinil group. Participants on modafinil experienced more adverse events, notably nausea and feeling jittery.

The evidence base for using psychostimulants as augmentation agents for major unipolar depression is limited (Reference Orr and TaylorOrr 2007). It remains to be seen whether psychostimulants may benefit a subgroup of people with depression who also have significant fatigue or sleep problems. A systematic review found inadequate evidence for modafinil as an augmentation agent in the treatment of fatigue in individuals with major depression who have had a partial response to antidepressants (Reference Lam, Freeman and CatesLam 2007).

Post-stroke depression

There have been positive retrospective audits and open trials of psychostimulants, predominantly methylphenidate, in post-stroke depression (Reference KrausKraus 1995). Partial or full remission was reported in 40 to 80% of patients, with doses ranging from 5 to 40 mg daily, over 2–4 weeks. Stimulants were well tolerated and treatment response usually occurred within 1–3 days.

Reference Lazarus, Moberg and LangsleyLazarus and colleagues (1994) noted no difference in the degree of response to methylphenidate compared with nortriptyline. However, the average speed of response was significantly different (2.4 days for methylphenidate compared with 27 days for nortriptyline). Reference Grade, Redford and ChrostowskiGrade and colleagues (1998) noted a significant improvement with 3 weeks of methylphenidate in Hamilton Rating Scale for Depression scores. In contrast, a Cochrane review (Reference Martinsson, Hårdemark and EksborgMartinsson 2007) found no benefit for dexamphetamine for depression in the context of post-stroke rehabilitation.

An important consideration in using stimulants in the post-stroke period is the potential for an elevation in blood pressure. In the absence of robust post-stroke data, it is worth considering a brain-injury study (Reference Alban, Hopson and LyAlban 2004) that reported minimal adverse effects of methylphenidate on pulse rate and blood pressure (which increased by 7 bpm and 2.5 mmHg respectively).

Stroke rehabilitation

Reference Martinsson, Hårdemark and EksborgMartinsson and colleagues' (2007) Cochrane review on the use of dexamphetamine given over 1–5 weeks in the post-stroke period, usually in combination with physical or cognitive training, showed benefits of treatment on motor and language functioning, but not on neurological functioning, performance of activities of daily living or depression. A non-significant increase in mortality was observed with dexamphetamine. The authors concluded that dexamphetamine could not be recommended routinely in stroke rehabilitation.

This review and an RCT (Reference Sonde and LökkSonde 2007) suggest that psychostimulants should not be used routinely in post-stroke rehabilitation. However, they might be tried in individual patients to target symptoms such as depression, fatigue and sleepiness, if other treatments have been ineffectual or are contra-indicated.

Acute and subacute recovery

A meta-analysis of controlled single-subject trials involving individuals included at a mean of 7.5 days after traumatic brain injury showed no effect of methylphenidate on command-following or yes/no communication abilities (Reference Martin and WhyteMartin 2007). Reference Kaelin, Cifu and MatthiesKaelin and colleagues (1996) reported an improvement in a number of attentional domains in a prospective study of 11 adults who received 10 days of methylphenidate, at a mean duration of 20 days after traumatic brain injury.

In an RCT of methylphenidate over 30 days in the subacute phase of traumatic brain injury, significant differences were seen on measures of attention, motor performance and disability after treatment, but this was not the case at 90 days (Reference Plenger, Dixon and CastilloPlenger 1996). Reference Lee, Kim and KimLee and colleagues (2005) described a 4-week RCT involving 30 patients who had experienced a major depressive episode, comparing methylphenidate with sertraline or placebo in treatment that began a mean of 35 days after traumatic brain injury. The study showed significant improvements in post-concussional symptoms, daytime wakefulness and reaction time with methylphenidate compared with placebo. Sertraline led to impaired performance in these domains.

Cognition

The rationale for the use of psychostimulants in cognitive rehabilitation after traumatic brain injury is derived from their use in ADHD, where methylphenidate has been shown to reduce hyperactivity, increase on-task behaviour and enhance working memory (Reference Whyte, Hart and SchusterWhyte 1997).

A number of small crossover RCTs of methylphenidate in adults with traumatic brain injury have been reported. Treatment duration varied between 10 days and 6 weeks, with a mean time since injury of 1.4–3.9 years. Two of the studies were negative (Reference Gualtieri and EvansGualtieri 1988; Reference Speech, Rao and OsmonSpeech 1993). However, Reference Whyte, Hart and SchusterWhyte and colleagues (1997, Reference Whyte, Hart and Vaccaro2004) reported significant improvements in domains related to cognitive speed and response rate, but not in attentional domains or motor speed. They concluded that methylphenidate improved ‘initial performance’, but may not enhance patients' ability to maintain performance over time.

One parallel-group RCT has been conducted, with cognitive functioning as a secondary outcome measure. Reference Mooney and HaasMooney & Haas (1993) studied the effects of methylphenidate over 6 weeks at a mean of 2 years after traumatic brain injury. No benefits were noted regarding attention or memory, but participants with higher pre-treatment anger scores did show a significant improvement on a verbal learning task.

Depression

Reference Lee, Kim and KimLee and colleagues (2005) noted significant improvements in mood with methylphenidate over 4 weeks; sertraline had a similar efficacy but a higher total number of adverse events. Thus, methylphenidate may be both beneficial and tolerable in the early stages of treatment for a depressive episode after traumatic brain injury, but it is unclear whether these results would be sustained over a longer period of time.

Behaviour

Reference Mooney and HaasMooney & Haas (1993) reported significant improvements in anger control and psychopathology with 6 weeks of methylphenidate treatment. Patients with higher pre-treatment anger scores showed the greatest benefit. Reference Speech, Rao and OsmonSpeech and colleagues (1993) found no effects of methylphenidate for differences in ‘social personality’ functioning on the Katz Adjustment Scale, despite the majority of patients and observers reporting improvements during the drug condition.

Depression

Psychostimulants have been used to treat depression in people with neurological disorders, including dementia, Parkinson's disease, epilepsy, neuropathy, normal pressure hydrocephalus and multiple sclerosis. However, data have been restricted to case series or large retrospective audits involving medically ill patients.

Apathy

Apathy is a common and disabling symptom seen in a range of neuropsychiatric and mental disorders. A number of case reports have noted the successful use of methylphenidate, dexamphetamine and modafinil for treating apathy in a range of neurological conditions (Reference Padala, Burke and BhatiaPadala 2007).

Early studies suggested that psychostimulants may be useful for apathy in people with varying degrees of cognitive impairment or dementia, with or without depression, but it was unclear whether the improvements were due to changes in affect, motivation or cognition, or a combination of the three (Reference MooreMoore 1981). Reference Galynker, Ieronimo and MinerGalynker and colleagues (1997) prospectively involved 17 individuals with dementia and found a significant improvement in negative symptoms with methylphenidate treatment.

Reference KaplitzKaplitz (1975) remains the only RCT of psychostimulants for apathy. It reported a positive effect of methylphenidate v. placebo in ‘withdrawn, apathetic geriatric patients’.

In light of the above evidence, it might be worth considering prescribing a psychostimulant for treatment-resistant apathy, regardless of aetiology, particularly in older adults.

Chronic fatigue syndrome

Psychostimulants have been trialled in people with chronic fatigue syndrome. Reference Olson, Ambrogetti and SutherlandOlson and colleagues (2003) conducted a 6-week RCT in 20 patients, comparing dexamphetamine (up to 15 mg twice a day) with placebo. The medication group showed a statistically significant reduction on the Fatigue Severity Scale. Other outcome measures, including quality-of-life patient-determined outcomes, were negative. The authors concluded that dexamphetamine may be useful in managing chronic fatigue syndrome.

There have been two crossover trials since: one with modafinil, which was negative (Reference Randall, Cafferty and ShneersonRandall 2005), and one with methylphenidate, which was modestly positive (Reference Blockmans, Persoons and van HoudenhoveBlockmans 2006). These studies suggest that psychostimulants do not have a routine role in chronic fatigue syndrome.

HIV/AIDS

Evidence from RCTs provides support for the efficacy of psychostimulants in the treatment of HIV/AIDS-associated depression. Significant improvements in mood, energy levels and quality of life were seen in patients with depressive-spectrum disorders and fatigue treated with 2 weeks of dexamphetamine (Reference Wagner and RabkinWagner 2000). Reference Fernandez, Levy and SamleyFernandez and colleagues (1995) found no difference over 6 weeks between methylphenidate and desipramine in HIV-positive individuals with a major depressive episode. Both groups showed significant improvements in mood, although methylphenidate did not have a more rapid onset of efficacy than desipramine.

Subjective fatigue occurs in up to 50% of patients during the course of HIV/AIDS (Reference Breitbart, Rosenfeld and KaimBreitbart 2001). There is reasonable RCT evidence to support the use of methylphenidate, pemoline and modafinil for HIV-related fatigue, with associated improvements seen in mood, distress levels and quality of life (Reference Breitbart, Rosenfeld and KaimBreitbart 2001; Reference Rabkin, McElhiney and RabkinRabkin 2004).

Small prospective and crossover studies have reported positive effects of methylphenidate, dexamphetamine and modafinil on measures of cognition, including memory, processing speed and executive functioning in HIV/AIDS (Reference Rabkin, McElhiney and RabkinRabkin 2004). However, negative trials also exist (Reference Hinkin, Castellon and HardyHinkin 2001) and improvement has been observed in placebo groups, suggesting the possibility of carry-over and/or practice effects.

Positive effects have not yet been found in RCTs for methylphenidate (Reference Fernandez, Levy and SamleyFernandez 1995) or dexamphetamine (Reference Wagner and RabkinWagner 2000) for cognitive functioning in HIV. Therefore, psychostimulants cannot currently be recommended as a treatment for this symptom cluster.

Other areas of psychiatry

The successful use of modafinil and methylphenidate to treat sedation caused by psychotropic drugs has been noted in case reports and uncontrolled studies in relation to sodium valproate (Reference BeriganBerigan 2004), antipsychotics, including clozapine (Reference MillerMiller 1996), and selective serotonin reuptake inhibitors (Reference Schwartz, Azhar and ColeSchwartz 2004). The possibility of adverse effects must be considered whenever using psychostimulants in people with psychotic disorders, with case reports noting the exacerbation of psychotic symptoms with modafinil (Reference Narendran, Young and ValentiNarendran 2002) and movement disorder symptoms with methylphenidate (Reference MillerMiller 1996).

Reference Rosenthal and BryantRosenthal & Bryant (2004) described benefits of modafinil on clinical presentation, global functioning and fatigue in a prospective study of individuals with schizophrenia and schizoaffective disorder. However, an RCT of modafinil for the alleviation of fatigue and cognitive deficits in a similar population found no benefit of modafinil over placebo (Reference Sevy, Rosenthal and AlvirSevy 2005). One participant withdrew from each of these studies because of worsening psychosis.

One RCT reported benefits of modafinil on abstinence rates in cocaine dependence (Reference Dackis, Kampman and LynchDackis 2005), whereas another (Reference Grabowski, Roache and SchmitzGrabowski 1997) found no benefits with methylphenidate.