Epidemiology
The side-effects of antipsychotic drugs have always been a major concern for clinicians, and the appreciation of their importance in the treatment of schizophrenia has increased steadily over the years.  Akathisia is probably the most distressing of the EPS. A review of the literature for antipsychotic-induced akathisia indicates a prevalence of
5.5–21.2% and an incidence of 25–75%.¹⁴

Figures for prevalence and incidence vary considerably, depending on the study. There are two main explanations: variability in defining akathisia, and variability in the area of differential diagnosis.

Definitions
Some authors base their definition of akathisia on subjective experience (a ‘sensation of restlessness’), and this leads to high incidence rates. More conservative authors insist on objective symptoms to reach a diagnosis (visible restlessness, shifting from foot to foot, and so on).  Overall, an incidence of 25%, confirmed by the majority of studies since the mid-1980s, seems a valid figure.¹⁵

A number of common errors complicate the differential diagnosis of neuroleptic-induced akathisia (NIA).  One of the major problems is trying to differentiate NIA from various states of restlessness (e.g. anxiety, agitation, restless legs, tremor, choreoathetosis). Of these states, anxiety and agitation pose the most difficulties, especially in acutely psychotic patients.  Tremor and choreoathetosis should not be confused with akathisia¹⁶, but often are.  For these reasons, NIA can be either under diagnosed or over diagnosed.

Placebo-controlled phase III clinical trials invariably show an astonishingly high rate of mixed EPS in the placebo groups. Preliminary analysis of one such study has yielded a rather surprising result (Table 6).
 

Criterion  EPS in placebo-treated patients  Of these:   Patients who also showed   excitement on BPRS   Patients who were also judged    to have agitation as a side-effect  Incidence (% )  14*     95**   59** *Refers to patients satisfying the criterion as a fraction  of all patients treated with placebo  **Refers to patients satisfying the criterion as a fraction  of patients showing EPS upon treatment with placebo

Table 6. Incidence of EPS in patients (n=144) treated with placebo in phase III trials

One of the reasons for this disconcerting finding may be that, although many clinicians are trained in the use of scales such as the Positive and Negative Symptom Scale (PANSS), little interrater-reliability training on the use of EPS scales in large studies is provided. The situation is complicated further by the fact that the ratings in these studies are often performed by young clinicians, who have limited experience of diagnosing movement disorders.

Differential diagnosis
While performing psychometric studies at two centers, New York and Innsbruck, looking at the validity and reliability of the Hillside Akathisia Scale¹⁷, we have routinely evaluated the effects of activation procedures in these patients (Table 7)¹⁸.
 

Activation task    Mental activation  Motor activation     Incidence of NIA  New York       Innsbruck      (n=37)             (n=34)  Increased     Unchanged Decreased      Decreased

Table 7. Influence of activation tasks on NIA incidence.

Basic motor activation (simple finger tapping), but not mental activation, reduced NIA in akathisic patients.  This contrasts with tardive dyskinesia, where the reverse usually occurs. No reduction in NIA was  found in anxious or agitated patients.  In short, very simple clinical measures can help to validate a suspected diagnosis of NIA.¹⁹

Management of NIA
Trials examining dose–response relationships have found that lower doses of antipsychotics show less propensity to induce NIA than do higher doses²⁰.  Rapid increases in dose have been identified as a risk factor for NIA.  The use of antipsychotics with serotonin antagonism seems to reduce the risk of NIA, as does slow dose increase.

Dose reduction is, wherever possible, the first option.  Beta-blockers have consistently been associated with improvement of akathisia.  Anticholinergics, although widely used to treat NIA, have proved less effective than beta-blockers in at least two comparative studies.  Anticholinergics appear to be most useful when akathisia is accompanied by parkinsonism. 5HT2 antagonists such as ritanserin have also been shown to reduce NIA levels considerably^21,22.  An important treatment option is also switching from one antipsychotic to another, preferably to one with a potent 5HT₂ antagonism.

Akathisia is probably the most distressing of the EPS and therefore impacts greatly on compliance²³.  It is extremely important that we recognize and manage akathisia.  We can only hope that some of the novel antipsychotics or new 5HT₂ antagonists will help to diminish this problem.

[back] [contents] [forward]