Antipsychotics were long considered the most dangerous medications used in psychiatry. Then their reputation got worse.

It is new evidence that antipsychotics can cause harm and even death in demented (most often elderly) patients that has led to articles like the one in this week’s Science Times about overprescribing. The problem appears to include older medicines, like Haldol and Thorazine, as well as newer ones, like Risperdal, Seroquel, and Zyprexa.

In Great Britain, the drugs’ standing is yet worse. In the Guardian, a commentary by a geriatric psychiatrist from Liverpool carried the headline “The scandal of dementia drugs must be addressed” and the subhead “Overprescribing antipsychotic drugs to people with dementia is a symptom of our neglect.” A Member of Parliament, Paul Burstow (pictured above), called the prescribing abusive and proposed that the doctors responsible be prosecuted. The Health Minister has promised an investigation.

All right, the British can go overboard when it comes to expressing disapproval of psychotherapeutic medication. Sometimes it seems that they don’t like behavioral treatments much better. It’s as if mental health care undermines the national character — the stiff upper lip should suffice. But even our own Food and Drug Administration has put a black box warning on antipsychotics, first on the new versions and now, last week, on the older ones.

Antipsychotics, starting with Thorazine in 1952, were the first of the new psychotherapeutic medications — and the first medications tested in placebo-controlled trials. The drugs turned out to be helpful in the treatment of schizophrenia, though the side effects, including late-appearing neurological syndromes, were horrific. By 1955, there were reports that antipsychotics were useful in what was called senile dementia, and by the 1980s it seemed solidly established that, as an overview paper put it, “antipsychotic medication has a definite but limited therapeutic role in the treatment of behavioral disturbances in nursing home patients with dementia.”

Doctors hated to use antipsychotics, but the diseases they treated were heartbreaking, too. In the 1990s, a new generation of antipsychotics seemed to promise similar efficacy with fewer neurological side effects. Then it emerged that the new drugs led to obesity and diabetes — and that the pharmaceutical houses had been suppressing relevant evidence. Still, as recently as five years ago, family doctors were being encouraged to use antipsychotics in the elderly.

The unexpected news about excess deaths began to break in 2003 when a placebo-controlled study that was summarized as favorable (it showed that low-dose Risperdal diminished agitation associated with dementia) found higher, but not statistically significant, rates of cardiovascular death in patients given the active drug. In 2005, a summary study found a “small, increased risk for death” in demented patients given the newer antipsychotics. The difference, drug versus placebo, was one-and-a-half times, 3.5% compared to 2.3%, over ten to 12 weeks of use. Doctors’ worry was that over a longer periods, the change in risk might be greater. The cause of the excess deaths was unknown, but in the elderly, the medications can cause delirium which, also for unknown reasons, is associated with increased mortality.

The risk-of-death finding was later extended to older antipsychotic medications. So far, it does not seem that the problem applies to the use of the medication in its more traditional use, the treatment of schizophrenia. There, for all their (very substantial) flaws, the medications can be life-altering and life-saving.

Ironically, the mortality risk in the elderly was discovered only because drug companies had undertaken new trials, in the effort to get an FDA “indication” for the use of antipsychotics in treating dementia. The finding did not emerge until quite extensive research had been undertaken; the whole story makes the case for follow-up studies on widely used medications. In truth, even now we do not know what the risks and benefits are, in demented patients, for very short-term and very long-term use of antipsychotics, both of which are common.

Parallel to the change in understanding of risk came discouraging evidence about the drugs' ability to help demented patients. A controversial 2006 study in the New England Journal of Medicine concluded that “adverse effects offset advantages in efficacy” of the new antipsychotic drugs in patients with Alzheimer’s disease. Only 32% of the patients did well on Zyprexa, and 21% improved on a placebo. Presumably the placebo response was due not to expectancy on the part of the demented patients but to the enhanced care that they received when participating in a drug trial. As I have suggested elsewhere, “clinical management” can be a powerful tool for good.

Despite the suggestion from the British parliament that geriatricians be disciplined for using antipsychotics, the consensus in the field, reflected in the Times piece, remains that (as the older research literature suggests) with careful management, the medications can occasionally be of use in the elderly. But the risk of death seems real. In this population, the drugs are almost certainly overprescribed. And in approaching dementia, everyone is in agreement that a thorough diagnostic workup looking for discrete causes should come first, along with behavioral approaches that reorient patients and add variety to their days.