Clinical manifestations

Presentation and course

Akathisia is the abnormal state of excessive restlessness. It fundamentally has two components: (1) a sensory component, which includes a sensation of inner restlessness, an urge to move, and dysphoria and (2) a motor component, manifested as movements that result from the sensation (65). Thus, akathisia is one of several movement disorders manifested by both motor and sensory phenomena (84). Patients describe feeling an overwhelming inability to remain still. The sensation is often unbearable and unpleasant, and patients may feel an associated tension, anxiety, irritability, aggressiveness, and impatience (63; 27; 08; 48). The patients describe feeling jittery, fidgety, and nervous. Akathisia is often accompanied by depression, including possible suicide attempts (08; 48). This is true even with newer atypical antipsychotics. Some studies suggest that suicide attempts are more closely associated with subjective sensory symptoms than motor symptoms, and attentional impairment (100; 101). However, one systematic study indicated no clear association between akathisia and suicide risk (94). This requires further study. Worst of all, movement provides only brief relief. The sensory components can also be relieved by passive movement or a perception of movement, such as riding in a car (84). The movements secondary to restlessness are often stereotyped and specifically involve the lower limbs, including pacing, walking in place, body rocking, leg swinging, toe-tapping, and an inability to stay in one place (tasikinesia) (10; 13; 01). Akathisia can interfere with activities of daily living and can be disabling. The restless movements in the legs associated with akathisia must be differentiated from those involved in the leg stereotypy syndrome, which is a much more common movement disorder (46; 66). In contrast to akathisia, leg stereotypy syndrome is not associated with uncomfortable restlessness; it typically consists of stereotypical, repetitive, flexion-extension, abduction-adduction movement of the legs or tapping of the feet. The movements are suppressible and decrease with distraction.

Akathisia has been found during any stage of Parkinson disease (60) but may be associated with greater severity of motor symptoms (20). About 50% of patients experience akathisia as a fluctuating phenomenon that may be present when on or off (120; 05). Parkinson disease has been associated with restless leg syndrome, with prevalence as high as 20%, but some of those patients may have akathisia (82). The two entities are often difficult to differentiate.

In relation to dopamine receptor-blocking agents, akathisia may occur as an early acute or subacute side effect (80) or as a tardive phenomenon after chronic administration and a worsening with withdrawal of the agent (09). In the early acute form, patients occasionally develop akathisia abruptly at the onset of oral treatment, sometimes within an hour of receiving the drug (10), but most patients with neuroleptic-induced akathisia have a gradual onset after several weeks of continued therapy and 90% of cases occur within 90 days of onset or change of therapy (16). It appears to be dose-dependent. It is more commonly seen with initiation and rapid dose escalation of first-generation antipsychotics or with polytherapy. Akathisia may aggravate the psychopathology for which the neuroleptic is given, partially because of noncompliance with the medication (24). Tardive akathisia occurs after prolonged exposure to neuroleptic drugs (typical and atypical) (76; 41) and has a similar pharmacologic profile as tardive dyskinesia. It can occasionally evolve from acute to tardive akathisia, so-called acute persistent akathisia (also referred to as “chronic akathisia” in the literature) (105). In contrast to the early subacute variety, tardive akathisia typically appears or worsens with dosage reduction or drug withdrawal (59) and is suppressed by dosage increases. Diagnostic criteria include the occurrence during or within 3 months of cessation of neuroleptic therapy and must be persistent when the neuroleptic is stopped for at least 1 month (105). In most cases, tardive akathisia is associated with other tardive syndromes, particularly classical tardive dyskinesia. The persistent nature of tardive akathisia is similar to tardive dyskinesia in that it rarely remits (19; 125).

Some patients have the movements of akathisia without the sensory component, which is referred to as pseudoakathisia (01). This may occur in psychotic patients who cannot verbalize the sensory phenomena but also is associated with the movements of tardive dyskinesia or tardive stereotypy and leg stereotypy syndrome. These patients may be misdiagnosed with other psychiatric states such as mania or agitated depression.

Prognosis and complications

Akathisia found in association with Parkinson disease responds inconsistently to antiparkinsonian therapy (60). Subacute drug-induced akathisia will persist as long as the agent is taken; however, the severity may fluctuate. It will resolve on discontinuation of the drug. Tardive akathisia remits only rarely (13), usually persisting indefinitely without treatment. Particular importance should be connected with the risk of medication noncompliance and suicide.

Clinical vignette

A 68-year-old woman with Huntington disease for over 10 years had chorea, gait disorder with occasional falls, and cognitive dysfunction. She also had a history of obsessive-compulsive symptoms and anxiety treated with sertraline. For increasing chorea, she was placed on tetrabenazine with an escalating dose to 12.5 mg three times daily. After about 6 months, she developed what appeared to be increased movements and depression. The tetrabenazine was initially decreased but then in the emergency room was increased again because of concern that the movements represented increasing chorea. On admission, it was found that the movements were, in fact, severe restlessness. She was getting up constantly. She described dysphoria and a need to move. Because of this, her frequency of falling was increased substantially. The symptoms were so severe she was not sleeping and she had suicide ideation. The tetrabenazine was tapered off, and low-dose quetiapine was initiated along with an increase in her sertraline. Her chorea increased, but the akathisia, depression, and suicidality all improved, and she was discharged.

Biological basis

Etiology and pathogenesis

Because akathisia is a sensorimotor disorder, the causative abnormality involves complex somatosensory systems. The basal ganglia are central to most movement disorders, and although they do not directly receive sensory information, they process sensory information indirectly and serve as the gatekeeper for sensory input to motor and other nonmotor features (through their anatomically distinct loops) at various levels. Abnormal sensorimotor integration may be key in the pathogenesis of akathisia and other movement disorders that involve sensory features (84). The situations in which akathisia occurs indicate that loss of dopaminergic function is central to its development. Akathisia arises in three main clinical situations, all of which impact dopaminergic systems: (1) Parkinson disease, (2) dopamine receptor-blocking drug (neuroleptic) treatment, and (3) treatment with other medications that impact dopamine systems, including VMAT2 inhibitors (eg, tetrabenazine, valbenazine, deutetrabenazine, and reserpine) (81; 47; 78; 99). Although it is suggested that atypical antipsychotics have a lower propensity to cause akathisia than typical agents, it is known that both have the potential to cause it depending on dose and potency (62). Antiemetic dopamine receptor blockers (ie, metoclopramide and promethazine) can also cause akathisia (02). Subacute akathisia is a dose-related phenomenon. Individual susceptibility to anti-dopaminergic drug-induced akathisia may be determined by dopaminergic genetic factors (eg, in the dopamine D3 receptor gene or the A1+ variants of the DRD2/ANKK1 Taq1A allele) (25; 62).

The association with parkinsonism and pharmacological blockade of dopamine receptors suggests that akathisia reflects a state of hypoactivity of dopaminergic systems. Decreased dopamine signaling results in the reduced ability to generate automatic movements. On the other hand, increased dopaminergic signaling might result in repetitive and stereotyped movements such as seen in akathisia (84). This may be the case with tardive akathisia, suggesting hypersensitivity of dopamine receptors.

Pharmacologic studies have implicated several nondopaminergic pathways in the pathophysiology of akathisia. The development of akathisia with atomoxetine therapy and the response of many patients to beta-blockers suggests that noradrenergic hyperactivity may play a role (122). Serotonin mechanisms, and specifically 5-HT2 receptor stimulation, have been implicated by pharmacologic studies, demonstrating that SSRI agents such as fluoxetine can cause akathisia (64; 32), and relief may be seen with 5-HT2A/2C receptor antagonists (87; 61). Up to 25% of those treated with fluoxetine may develop akathisia (64; 35). Serotonin acts by inhibiting dopamine release.

It has been hypothesized that dopaminergic blockade occurring within the nucleus accumbens is important to the development of akathisia in relation to other nondopaminergic pathways (106). It has been suggested that the decreased dopaminergic stimulation of the core of the nucleus accumbens and increase in the adrenergic input from the locus ceruleus to the shell portion of the accumbens results in symptoms of akathisia. The increased noradrenergic outcome is the result of a D2 receptor blockade in the locus ceruleus. This mismatch between activities of the core and shell leads to the seemingly senseless and purposeless movements, as seen in obsessive-compulsive disorder. The movements are associated with activation of the shell region (which receives projections from the infralimbic cortex) and immediate reward-seeking behavior due to decreased activity of the core region, both related to akathisia. This may explain the response to beta-blocking agents.

Dopaminergic blockade also leads to downregulation of GABAergic activity, as well as increased release of acetylcholine, particularly from striatal interneurons. These changes could also play a role in the development of akathisia and have led to the study of medications that alter both systems (77).

An association between akathisia and abnormalities of iron metabolism has been proposed based on the similar association of iron and restless legs syndrome. Several studies have found low iron and ferritin levels and high total iron-binding capacity in schizophrenic patients with akathisia (55). In a study, schizophrenic patients without akathisia posted results on all three tests that were intermediate between those of akathitic patients and controls (55). However, this is not a consistent finding, and the role of iron in this disorder remains to be clarified (96).

Epidemiology

The prevalence of acute or subacute akathisia from first-generation neuroleptics over many studies and many years was found to be variable, from as low as 3% to as high as 75%, with an estimate 20% to 30% overall (01). The variation depended on the definition of akathisia as well as risk factors, including type, dose, potency of the neuroleptic, psychiatric diagnosis, female gender, age, ethnicity of the patient population, concomitant features, and recognition of the disorder. For second-generation agents, of the 372 patients with schizophrenia and schizoaffective disorder evaluated in a prevalence study through the FACE-SZ data set from 10 schizophrenia centers in France, the global prevalence was 18.5% (06). In a 9-month, cross-sectional study in Tunisia of 124 outpatients with schizophrenia, schizoaffective disorder and bipolar disorder treated with first- and second-generation drugs, on one month of stable therapy, 19.5% were found to have akathisia (48). In other studies, the frequency was thought to be equal or higher in patients with bipolar disorder than with schizophrenia, about 24% higher overall, although it depended on the drug (12; 50; 85; 21; 18). Controlled trials have demonstrated frequencies of akathisia in 13% to 24% (depending on dose) of bipolar I disorder patients treated with lurasidone (51), in 5% to 8% of patients acutely treated with brexpiprazole for bipolar mania (115), and in 17% of patients with bipolar I or II treated with cariprazine (121).

A meta-analysis of 58 studies demonstrated that the estimated weighted odds ratio of akathisia under second-generation antipsychotics, compared with placebo, was 2.43 (95% CI 1.91-3.10) (22). Nevertheless, in a review of over 70 trials involving atypical antipsychotics, the frequency was consistently lower than that seen with first-generation drugs (50; 56). The same conclusion was reached when accounting for only low doses of first-generation agents (67). One additional study demonstrated lower frequency of akathisia in first-episode psychosis (37). The frequency has been suggested to be about 15% overall with second-generation agents (21). On the other hand, the frequency of acute akathisia was actually high in a study of 129 severely ill patients with first-episode schizophrenia spectrum disorders who were naïve to antipsychotics: 42% with risperidone and aripiprazole (124). In a study from Estonia, the frequency of akathisia remained at 30% in an inpatient cohort that was initially on first-generation agents and then changed to second-generation agents 8 years later (83). The results of such comparison studies with first-generation agents depended on the comparator agent; for example, studies with haloperidol were different from those with perphenazine (16).

Several studies have demonstrated that the frequency of akathisia depends on the potency of the second-generation agents. A meta-analysis showed that olanzapine and quetiapine had a low frequency of 4%, whereas it was 20% with other second-generation agents and 50% with haloperidol (97; 27). In a study examining the incidence of acute akathisia, a pooled analysis was completed for three prospective, randomized, flexible-dose, and open-label clinical trials including 493 patients with first-episode nonaffective psychosis in Spain (49). Akathisia was determined using the Barnes Akathisia Rating Scale at 6 weeks after antipsychotic initialization. The overall incidence of akathisia was 19.5%, with significant differences between drugs: haloperidol (57%), risperidone (20%), aripiprazole (18.2%), ziprasidone (17.2%), olanzapine (3.6%), and quetiapine (3.5%). Another meta-analysis of 56 studies examined the frequency with specific second-generation agents. The estimated weighted mean incidence rate of akathisia was 3.9% (95% CI 2.4-6.3) for iloperidone, 6.8% (95% CI 5.1-9.0) for asenapine, 10.0% (95% CI 7.4-13.5) for brexpiprazole, 12.7% (95% CI 10.1-16.1) for lurasidone, and 17.2% (95% CI 13.4-22.1) for cariprazine (22). In this study, type of medication (p < 0.0001), diagnosis (p = 0.02), and race (p = 0.0003) significantly explained part of the heterogeneity of the incidence estimates of akathisia between studies. Similar results have been reported by Chow and colleagues, who examined 177 double-blind and open-label trials of the nine newest atypical agents (18). In a long-term study of asenapine, the figures ranged from 1% to 8% depending on duration and dose (53). Although lower figures have been reported with quetiapine and clozapine, they can still be implicated (72; 102). In a community-dwelling population, the frequency of akathisia was lower with quetiapine and amisulpride, with prevalence figures of 1% to 6% (06). A study in patients with bipolar disorder demonstrated a frequency of 5% with 300 and 600 mg per day of quetiapine (21). It should be noted that frequencies with haloperidol in these comparator studies was quite varied, as earlier studies noted. Although the dose appears to be important in relation to the frequency of akathisia, it does not appear to be a factor in the severity (18; 123). In other studies, age appeared to play a role. With aripiprazole, the prevalence was 17% in older and 26% in younger depressed patients (107).

The Genetic Risk and Outcome of Psychosis study is a longitudinal study of 1120 patients with nonaffective psychosis in a younger group with a mean age of 27. Prevalence, incidence, and persistence rates for akathisia were 9%, 5%, and 17%, respectively (69). A study in adolescents demonstrated an approximately 11% frequency with atypical antipsychotics (34). Risk factors for acute akathisia include age, ethnicity, higher antipsychotic doses, high-potency agents, or polytherapy with combinations of atypical agents with other psychotropic drugs (as high as 35% prevalence or 2- to 3-fold increased risk), and bipolar disorder (33; 56; 06).

In nonpsychiatric populations, the risk seems lower; 297 patients undergoing general anesthesia for ambulatory surgery were randomized to receive postoperative nausea/vomiting prophylaxis with droperidol (0.625 mg or 1.25 mg) or ondansetron 4 mg. One out of 18 (6%) in the ondansetron 4 mg group, 1 of 84 (1.2%) in the droperidol 0.625 mg group, and 3 of 87 in the droperidol 1.25 mg (3.4%) group developed akathisia (17). In another study, from a chart review of 592 oncology patients treated with dopamine-blocking antiemetics and referred to a psychiatry service, 4.7% were diagnosed with antiemetic-induced akathisia, which generally took 4 to 6 months to diagnose (92). Studies examining the frequency of akathisia in migraine patients who received single doses of prochlorperazine or metoclopramide had a frequency of 5% to 6% with a definitive diagnosis of akathisia, but 34% had symptoms suggesting a possible diagnosis of akathisia (113; 30). The rate of delivery of antiemetics (metoclopramide) may impact the frequency of akathisia (119). In a blinded trial, the incidence of akathisia was observed to be 26% with bolus therapy and 7% in those given a slow infusion (114).

The prevalence of tardive akathisia is estimated at approximately 20% to 30% of cases treated chronically with neuroleptics (09), a figure similar to that of classical tardive dyskinesia. In general, figures on tardive syndromes have been examined through meta-analyses of large numbers of patients to determine the frequency caused by first- and second-generation antipsychotics. In 41 trials (n=11,493), the prevalence rate of tardive syndrome was 21% for second-generation agents and 30% for first-generation agents (14). Annual incidence was 2.6% for second-generation agents and 6.5% for first-generation drugs (15). There have been reports of tardive akathisia caused by lurasidone (112) and others. Two centers that previously reported large numbers of patients suggested wide variations in the frequency of diagnosis (13). Studies have implied that tardive akathisia occurs more commonly in younger patients treated with higher doses, but it is unclear if there is a gender predilection.

Patients with Parkinson disease show a comparable rate of development of akathisia--26% in one series (60) and 45% in another (20). For up to half of the patients, akathisia relates to nonmotor fluctuations (120; 05).

There have been several studies implicating selective serotonin reuptake inhibitors, such as fluoxetine and sertraline. The frequency of akathisia with these agents is unknown, but it appears to be less than that seen with antipsychotics (01). One study, however, suggested a frequency as high as 25% (64). In a double-blind, placebo-controlled study of patients with Huntington disease treated with tetrabenazine, the frequency of akathisia was 9% of 54 patients (Huntington Disease Study Group 2006). The figures for the newer VMAT2 inhibitors, deutetrabenazine and valbenazine, which have been approved for the treatment of tardive dyskinesia and Huntington disease, appear to be lower (79; 43; 45; 40; 28; 78; 99).

Prevention

Akathisia complicating Parkinson disease cannot be prevented. Drug-induced akathisia is preventable by avoiding the use of the offending agents. This is particularly true when considering neuroleptic therapy for patients with nonpsychotic disorders and other medical uses. Unfortunately, for many psychotic disorders, no alternatives to anti-dopaminergic drugs are available. Some believe that atypical neuroleptics produce a lower incidence of akathisia, but even clozapine (36) and quetiapine (90; 102) cause severe akathisia. The lowest incidence of drug-induced akathisia can be achieved if physicians reserve antipsychotic drugs for situations of absolute necessity, prescribe the lowest dose required for the desired therapeutic effect, and withdraw the drug at the earliest possible opportunity. When used though, the treating physician needs to be vigilant.

Differential diagnosis

In patients taking neuroleptic drugs, akathisia is often difficult to distinguish from the psychotic psychomotor agitation caused by the underlying psychiatric condition for which the drugs are being given: anxiety, agitated depression, substance intoxication, or withdrawal. Akathisia also occurs in a number of other scenarios for which dopamine-blocking agents are used. Cancer patients also seem to be particularly prone as they often receive dopamine-blocking anti-emetic agents during chemotherapy (52). Patients treated for gastrointestinal problems with metoclopramide and other similar agents are also prone. Intravenous metoclopramide used in the treatment of migraines was associated with akathisia in 9.2% of cases (31). Failure to recognize akathisia may lead to an inappropriate increase in medication and, thus, compounds the problem (58). It is important to keep an index of suspicion. Schizophrenic stereotypies may be mistaken for restless movements of akathisia. Tardive dyskinesia must also be distinguished from akathisia in those patients who are distressed by their movements. Drug withdrawal is another condition that may be confused with akathisia.

Restless legs syndrome resembles akathisia in many respects and may also be a complication of Parkinson disease. However, in restless legs syndrome, symptoms are generally confined to the lower limbs and tend to be diurnal. Tics also result from a psychic urge to move (57) but should be relatively easy to identify when witnessed.

Diagnostic workup

Akathisia is recognized solely by its clinical features. Although nonspecific reduction of serum iron and ferritin and elevation of acute phase reactants have been reported, their clinical importance is questionable (42). No laboratory tests firmly support the diagnosis. The same is true for virtually all disorders that may be considered in the differential diagnosis.

Management

It is important that psychiatric patients be informed about the possible development of akathisia or other movement disorders if neuroleptic therapy is being prescribed. Recognition is important for patients and physicians, lest they misinterpret the condition as worsening agitation or psychosis; recognition is also important for maintenance of medication compliance and prevention of secondary suicide ideation (23). Should acute or subacute akathisia occur, the first step in therapy is to reduce or eliminate intake of the offending drug (89). Time to reversal after removal of the inciting agent is variable, and there may be a worsening before it improves (16). If withdrawal is not feasible, consider substituting the agent with a less potent one, preferably a second-generation agent (89). If the substitution is unsuccessful, medical therapy will be necessary. There are no treatments specifically FDA-approved for akathisia. However, akathisia may respond to a variety of medications for which there is ample evidence (89; 27).

Anticholinergic medications (including diphenhydramine) have historically been utilized to treat akathisia--perhaps because of their usefulness in drug-induced parkinsonism and acute dystonia, but evidence is scant. A Cochrane review in 2006 found no relevant randomized controlled trials and, hence, no reliable evidence to support or refute the use of anticholinergics for patients with acute or subacute akathisia (93). Several class III and IV trials have been reported (89). Several trials have demonstrated mixed results with various agents in the class. In 2007, a double-blind, placebo-controlled study of biperiden injections was completed in 30 patients (15 active versus controls) with assessments at baseline and three times after the first injection at 2-hour intervals. Response was defined as at least a 2-point decline in the global akathisia score. The numbers of responders in the two groups were not significantly different, and the changes in the Barnes Akathisia Rating Scale were similar (04). A review of five studies and 1094 patients demonstrated that prophylactic use of diphenhydramine had no impact on the development of akathisia (75). A systematic review of the use of anticholinergics prophylactically in the emergency department when neuroleptics or metoclopramide are used to treat headache found the same result (70). Anticholinergics do not appear to be an effective treatment for akathisia and have troublesome adverse effects associated with use. They should not be used. Overall, the level of evidence is 1, and the recommendation grade is B (89).

The current first choice for treating akathisia is the beta-blocker propranolol, which has been in use since 1983. Seven class II and III open-label and controlled trials have shown consistent results with doses of 120 mg/day or less and a 50% improvement (71; 01; 89). Propranolol has been reported to improve tardive dyskinesia as well and may be useful for tardive form of akathisia (39). Other nonselective beta-blockers such as nadolol and pindolol have been found to be effective, though less so because of diminished blood-brain barrier penetration. B1 selective agents, such as metoprolol and betahexol, have also been utilized with some success. These medications should be used with caution in those with depression, pulmonary disease, diabetes mellitus, or cardiac disease. Its side effects include orthostatic hypotension, bradycardia, and exercise intolerability. Overall, the level of evidence is 1, and the recommendation grade is B (89).

5-HT2a antagonists, including mianserin, mirtazapine, and trazodone, represent another approach to treating antipsychotic-induced akathisia, with accumulating evidence supporting their use (88). The basis of their use relates, in part, to the lower propensity of antipsychotics with strong inhibition of 5-HT2a receptors to cause akathisia. Data support their use in akathisia secondary to first- and second-generation antipsychotics. In a systemic review and evidence-based meta-analysis, six randomized controlled trials were found examining 5-HT2a antagonists, five of which included a placebo control group (61; 89). The overall effect size in the analysis is substantial, RR=7.10 with 95% CI 3.08–16.40 (p< 0.0001) based on change in the Barnes Akathisia Rating Scale. Because of the small number of trials, a difference between the three agents: mianserin, mirtazapine, and trazodone, could not be detected. These findings point to the class of 5-HT2a antagonists as effective and safe for neuroleptic-induced akathisia. Overall, the level of evidence for 5HT antagonists is 1, and the recommendation grade is B (89). Some of these trials are discussed.

Mianserin is a tetracyclic antidepressant that demonstrates notable 5-HT2a and 5-HT2c antagonism and also has antihistaminergic and alpha-2 antagonistic activity but limited anticholinergic effects. It is not approved for use in the U.S. One study compared the efficacy of vitamin B6, mianserin, and placebo in the treatment of acute akathisia. Sixty schizophrenia and schizoaffective in-patients with akathisia were randomized to receive vitamin B6 1200 mg/d, mianserin 15 mg/d (low dose), or placebo for 5 days, in a double-blind design. Compared with the placebo group, the vitamin B6-treated and mianserin-treated patients showed a significant improvement in the subjective symptoms (P < 0.0001), subjective distress (P < 0.0001), and global (P < 0.0001) subscales. A reduction of at least 2 points on the Barnes Akathisia Rating Scale global subscale was noted in the vitamin B6 group (13/23, 56%) as well as in the mianserin groups (13/20, 65%) and in only one patient in the placebo group (1/17, 6%; P < 0.0005) (73). There was one other study of B6 that was positive. It is recommended only for short-term use in those who fail other therapies. Overall, the level of evidence for vitamin B6 is 1+, and the recommended grade is A (89). B6 toxicity, particularly peripheral neuropathy, must be monitored.

Mirtazapine is a potent antagonist of central alpha-2 auto- and hetero-adrenergic receptors, as well as an antagonist of 5-HT2A/2C, 5-HT3, and histaminergic H1 postsynaptic receptors. The use of mirtazapine offers advantages over other anti-akathisia drugs in its better adverse effect profile as well as its ability to treat coexisting depression (91). In a 7-day double-blind trial, 90 antipsychotic-treated patients meeting DSM-IV criteria for akathisia were randomly assigned to mirtazapine (n = 30; 15 mg), propranolol (n = 30; 80 mg), or placebo (n = 30). Twenty-four patients (26.6%) did not complete the study (seven mirtazapine, eight propranolol, nine placebo) due to lack of response (n = 19) and adverse events (n = 5). Both mirtazapine and propranolol significantly reduced akathisia severity (Barnes Akathisia Rating Scale: -34% mirtazapine (p = .012) and -29% propranolol (p = .023) vs. placebo -11%). Thirteen (43.3%) mirtazapine- and nine (30.0%) propranolol-treated patients versus two (6.7%) placebo-treated patients responded (BARS global scale reduction ≥ 2; odds ratios 10.7 [95% confidence interval (CI), 2.1-53.3] and 6.0 [95% CI, 1.1-30.7]). Five (16.7%) of 30 propranolol-treated patients and none in the mirtazapine and placebo groups (p = .0195 for both) prematurely discontinued the study due to clinically significant hypotension or bradycardia (87). In a meta-analysis that included this trial and a smaller trial, the risk ratio for response was 6.67 (95% CI 2.14–20.78; p = 0.001) and for remission was 6.20 (95%CI 1.74–22.08; p = 0.005), respectively, indicating that mirtazapine is about six times more effective than placebo in ameliorating symptoms of acute akathisia (88). The most common adverse effect was mild sedation. Increased appetite and subsequent weight gain may also be an issue. There have been rare reports of mirtazapine causing akathisia. The mechanism by which this occurs is unknown, but treating physicians should be aware of this possibility (95).

Trazodone has potent serotonin 5-HT2a receptor agonist and alpha-1 adrenergic receptor antagonist effects and is a weak serotonin reuptake inhibitor and a weak histamine H1 receptor antagonist. A double-blind, crossover study of trazodone has also been found to improve akathisia. This was a 6-day study (3 days active, 3 days placebo with no washout) performed on 13 patients with acute akathisia diagnosed by DSM IV criteria. There was significant improvement of the Barnes Akathisia Rating scale subjective, objective, distress, and global parts with 100 mg at night (108). This study was followed by a double-blind, placebo-controlled, 5-day trial of trazodone 50 mg per day in 52 patients with mild to severe akathisia (103). The study showed a significant improvement in objective (P = 0.01) and subjective (P = 0.001) symptoms of akathisia, as well as significant change in global clinical assessment (P = 0.001). It was also well tolerated.

Gabapentin enacarbil, with a mean dose of 567 mg/day, was examined in an open-label prospective trial of eight schizophrenia patients with akathisia. The Barnes Akathisia Rating Scale global akathisia score significantly decreased after 1 and 2 weeks of treatment when compared to baseline (P=0.01 and P=0.01, respectively). There was no worsening of psychosis (110). Other agents with limited evidence for the treatment of akathisia include benzodiazepines (clonazepam), amantadine, and clonidine.

Based on results of the various studies, it is suggested that propranolol, mirtazapine, and perhaps trazodone be considered first-line agents (86; 89). The other classes listed above could be considered if these fail or are contraindicated (for instance, propranolol should not be utilized in patients with asthma or diabetes).

An updated evidence-based guideline for the treatment of tardive dyskinesia has been provided in 2018 (07). In tardive akathisia, aside from propranolol (39), the catecholamine depletors, valbenazine and deutetrabenazine, have level A evidence (28; 40). Reserpine and tetrabenazine have also been the effective agents but at level C (13; 47). The problem with these drugs is that they have the potential to cause subacute akathisia as well. For the newer VMAT2 agents, their pharmacokinetic profile seems to make such side effects less common. Zolpidem, which is a GABA-mimetic drug and a selective agonist of the benzodiazepine receptors, and gabapentin, which enhances the activity of GABA, have also been found to be helpful in some patients with akathisia, without causing significant drowsiness (47; 117; 109). Anticholinergic agents will likely make tardive akathisia worse as they do with classical tardive dyskinesia, but there are little data on this. A case report suggests a role for electroconvulsive therapy, but further experience is needed (26). Deep brain stimulation of the subthalamic nucleus has been reported to be effective in patients with Parkinson disease who experienced severe akathisia with or without a relationship to levodopa (104). Although there are increasing data for the use of deep brain stimulation of the globus pallidus interna for classical tardive dyskinesia and tardive dystonia (111), it is not clear whether deep brain stimulation would be a suitable treatment for tardive akathisia.

Special considerations

Anesthesia

Akathisia may be seen in surgical patients as dopamine receptor antagonists acquire a wider role in control of emesis and pain (11). The butyrophenone droperidol, and phenothiazine metoclopramide used as an adjunctive antiemetic with opiate analgesics, may cause akathisia when administered intravenously (118) or epidurally (03). Midazolam, a benzodiazepine often used for preoperative sedation, induced paradoxical akathisia in three patients who received epidural anesthesia (68).