Can Medication-Free Research Ever Be Ethical in Older People With Psychotic Disorders?

The history of medicine demonstrates that apparently logical ideas sometimes prove to be inaccurate when subjected to empirical research. In psychopharmacology, repeated antipsychotic discontinuations (drug-free intervals) used to be strongly recommended for treating patients with schizophrenia with the presumption that they would prevent tardive dyskinesia.¹ Yet, a subsequent empirical study found that intermittent drug treatment was associated with an increased risk of persistent tardive dyskinesia.² Now, Bola's thought-provoking review (this issue) shows that the relevant empirical data do not support the previous notion of untreated psychosis causing long-lasting neurobiological damage. (This is not to question the serious consequences of relapse that may follow antipsychotic withdrawal.)

How ethical is antipsychotic-free research in older persons with schizophrenia or other psychotic disorders? The article by Bola does not address this issue directly because the studies included in his review focused mostly on younger adults. There are some similarities, as well as several important differences, between the use and effects of antipsychotics in younger versus older adults. The issue of antipsychotic use in older patients is one of growing public health significance because of rapidly increasing numbers of older people with schizophrenia and other psychotic disorders.³

The limited published literature on antipsychotics in older patients suggests that these drugs have largely similar therapeutic effects across age groups in terms of reducing positive symptoms and preventing relapse, and also in terms of side effects.⁴ At the same time, there are several critical differences in antipsychotic pharmacology between younger and older adults. The required dosages tend to be considerably lower, and most side effects (except for acute dystonia) are much more prevalent in the elderly. The typical (first-generation) antipsychotics are also not usually recommended in older patients given the 6-fold elevated incidence of tardive dyskinesia in that age group.⁵ Unlike younger adults, the most common use of antipsychotics in the geriatric group is not for the 2 indications for which they are approved by the Food and Drug Administration (FDA)—schizophrenia and bipolar disorder—but is mostly off-label in persons with dementia complicated by psychosis and agitation. Moreover, older individuals tend to have multiple physical comorbidities, enhancing the probability of drug-drug interactions. Below I briefly review the data on antipsychotic use in people with dementia and schizophrenia in older persons.

In the dementia patients, efficacy of the antipsychotics compared with placebo is greater for agitation and aggression than for psychotic symptoms per se. In addition, the FDA has recently issued black box warnings about the risk of strokes and mortality with the atypical (second-generation) antipsychotics in elderly patients with dementia. The result is that the picture with respect to antipsychotic use is considerably more clouded in the elderly patients with psychotic disorders than it is for younger adults with schizophrenia. It is noteworthy that in the Clinical Antipsychotic Trials in Intervention Effectiveness (CATIE) study of atypical antipsychotics for schizophrenia (which was restricted to patients under age 65), the control group received typical antipsychotics but not placebo.⁶ In contrast, the CATIE trial of psychosis of Alzheimer disease included a control group that received placebo but not the typical antipsychoics.⁷ This is a clear indication of the view that placebo was considered safer than first-generation antipsychotics in elderly persons with dementia.

An important condition in the differential diagnosis of new-onset psychosis in the elderly is iatrogenic (drug-induced) psychosis. Clinically, it is recommended that patients with new-onset psychosis (including suspected late-onset schizophrenia) not be given pharmacotherapy until a specific diagnosis is made—in other words, starting antipsychotics may be delayed at least for some time (unless the psychotic symptoms are severe.)

A majority of older persons with schizophrenia have had the illness for several decades. Is discontinuation of antipsychotics in this group appropriate? The traditional thinking is that schizophrenia gets worse with age (“dementia praecox” of Kraepelin) and that these patients would need medications for the rest of their lives. The existing data challenge this notion. We, as well as other researchers, have found that with aging, the positive symptoms improve significantly.⁸ A likely neurobiological explanation for this phenomenon is an aging-associated decrease in dopaminergic and other monoamienergic activities. A small proportion of elderly patients has sustained symptomatic remission of schizophrenia and may not need antipsychotics.⁹ Indeed, continued use of antipsychotics is associated with an increased risk of long-term side effects such as tardive dyskinesia with typical neuroleptics and metabolic syndrome with the atypical agents. Nonetheless, given the possible risk of relapse following abrupt withdrawal of antipsychotics, sudden discontinuation of these drugs in the elderly patients is not recommended. A very slow, tapered reduction in dosage with close monitoring is suggested.⁵ In a minority of patients with very chronic schizophrenia, complete discontinuation may be feasible even after decades of use.

So, what can one say about medication-free research in older persons with psychotic disorders? I believe that such research is appropriate in specific circumstances. These include new-onset psychosis without a clear diagnosis, psychosis of Alzheimer disease, and even chronic schizophrenia with evidence of sustained symptomatic remission and improved functioning during progressive dose reductions. The contraindications to medication-free research may include presence of severe psychotic symptoms and/or agitation/aggression, a recent history (during the previous 5 years) of psychotic relapse following neuroleptic discontinuation, and a desire on the part of a patient or her or his care provider to continue medications.

There is no evidence that psychosis is neurobiologically toxic in older persons, and clinical course evidence does not support the neurotoxicity hypothesis in younger patients.^10,11 On the other hand, there are consistent data showing long-term adverse effects of antipsychotics on the brain and other organs in this age group. Therefore, well-monitored medication-free research in carefully selected older patients may be ethical, provided that necessary safeguards to prevent relapse and other adverse effects of antipsychotic discontinuation are in place.