Antipsychotic Use in Dementia Linked To Increased Mortality

Pauline Anderson

July 01, 2019

UPDATED July 2, 2019 // OSLO, Norway — More evidence is linking use of antipsychotics to an increased mortality risk in patients with dementia.

In a Danish registry study of almost 33,000 patients with dementia, those who received an antipsychotic drug had a 35% increased risk of mortality compared with their peers who did not receive this type of medication.

The patients with preexisting diabetes or cardiovascular disease who received an antipsychotic had an even higher mortality rate.

Dr Ane Nørgaard

Clinicians should have access to some sort of risk stratification to identify patients with dementia who would most benefit from an antipsychotic, study presenter Ane Nørgaard, MD, PhD, Danish Dementia Research Center, Rigshospitalet, Copenhagen, Denmark, told Medscape Medical News.

"I hope that in the future we can develop some guidelines that will help clinicians determine which patients can benefit from the treatment with the lowest risk, and also which patients they should be really cautious about," Nørgaard said.

The findings were presented here at the Congress of the European Academy of Neurology (EAN) 2019.

Complex Management

Most patients with dementia will experience neuropsychiatric symptoms. These symptoms increase caregiver burden, lead to early nursing home placement, and can negatively affect quality of life.

Management of these symptoms is complex and may include psychosocial interventions, treatment of medical conditions, and a medication review. In severe cases, antipsychotic medications may be appropriate, Nørgaard told meeting attendees.

However, the evidence for an effect of these medications on behavioral symptoms and psychosis is "limited" in patients with dementia, she said.

"The overall conclusion from clinical trials is that the effect might be outweighed by the adverse effects," which can include cardiovascular side effects such as QT prolongation, as well as sedation, Parkinsonism, orthostatic hypotension, and metabolic disturbances, she said.

For the current study, the researchers used Danish national health registries to identify patients between the ages of 65 and 95 years who had a first diagnosis of "full blown" dementia between 2009 and 2014, and not mild cognitive impairment (MCI), Nørgaard noted.

The investigators also gathered information on redeemed drug prescriptions. Patients who initiated antipsychotic treatment after a dementia diagnosis were defined as the "exposed" group.

These exposed patients were matched to up to three unexposed patients by sex, age at time of dementia diagnosis, and year of first dementia diagnosis.

The study included 8244 patients in the exposed group and 24,730 in the unexposed group. Median age at time of dementia diagnosis was 82 years. For exposed patients, the median time from dementia diagnosis to first antipsychotic prescription was 252 days.

The outcome was 180-day mortality using Cox regression models.

Increased Mortality Rate

Results showed that the crude mortality rate per 100 patient-years was 58.3 for the exposed group vs 36.3 for the unexposed group. After adjusting for sex and age at dementia diagnosis, the crude hazard ratio (HR) for the exposed vs unexposed groups was 1.49 (95% confidence interval [CI], 1.41 - 1.57).

After also adjusting for other predefined confounding factors such as calendar year, nursing home residency, heart disease, cerebrovascular disease, and diabetes, the HR for 180-day mortality rate was 1.35 (95% CI, 1.27 - 1.43)

"This suggests that mortality was increased by 35%" in exposed patients, Nørgaard said.

To assess the effect of comorbidity, the researchers stratified the analyses for cardiovascular diseases, cerebrovascular diseases, and diabetes.

They found that the crude mortality rate increased more for those with preexisting cardiovascular disease. In those with cardiovascular disease, the rate was 78.2 for the exposed patients and 50.9 for the unexposed. For those without cardiovascular disease, the rates were 58.3 and 36.3, respectively.

It was a similar story for patients with preexisting diabetes. "If patients have either type 1 or type 2 diabetes before the dementia diagnosis, they are at further increased risk of death when they're treated with an antipsychotic," Nørgaard said.

In many cases, patients with severe neuropsychiatric symptoms should be treated with an antipsychotic; but the new results "suggest that clinicians should be even more aware" when considering an antipsychotic prescription if the patient has preexisting diabetes or cardiovascular disease, she noted.

Interestingly, the opposite pattern arose for those with preexisting cerebrovascular disease. "Here, the mortality rate increased more among the patients without previous cerebrovascular disease," said Nørgaard.

She speculated to Medscape Medical News that this might be due to increased clinician awareness. "We know that antipsychotics are also associated with increased risk of stroke, so maybe clinicians are already aware that they should not prescribe antipsychotics for patients with a history of stroke," she said.

Norgaard noted that she believes this is the first analysis of its kind to use a national cohort of dementia patients.

No Dementia-Type Stratification

After the presentation, session co-chair Irena Rektorova, MD, PhD, professor of neurology, School of Medicine, Masaryk University, Brno, Czech Republic, asked whether the researchers stratified the groups based on dementia type.

"I think this is really important. Dementia patients are a huge group, so did you stratify by what was the cause of dementia?" Rektorova asked.

The Danish registry "is not valid enough" for such stratification analyses, although "It would be very interesting to see whether some patients were at a higher or lower risk," Nørgaard answered.

Addressing a query from an attendee about the impact of different types of antipsychotic drugs, she said the study did not separate first- and second-generation agents.

"We have done some analysis to see how use of first- and second- generation drugs has changed over time, and found that second-generation agents are being used more and more in the dementia population. But from this current study, we can't say anything about the different effects on mortality," she added.

In addressing another query, Nørgaard said the study also could not determine mortality risk of dementia patients with neuropsychiatric symptoms who were not given an antipsychotic — even if it was indicated.

No "Direct Effect"

Commenting for Medscape Medical News, Ludwig Kappos, MD, professor and chair in the Department of Neurology, University Hospital Basel, Switzerland, said the increased mortality uncovered in the study might not be a direct effect of antipsychotic use.

"People who need antipsychotic drugs may also be in worse condition and have a worse prognosis" compared to those deemed not to need these medications, said Kappos, who was not involved with the research.

"There are many unknown factors that can contributed to a worse prognosis," he added. For example, study participants taking antipsychotics may have had more myocardial infarctions or more falls and fractures.

A good choice of antipsychotic drug that sedates and reduces negative impulses but might lead to relatively few side effects is clozapine, Kappos noted.

"A key advantage of this older antipsychotic drug is that it does not have extrapyramidal symptoms, so does not induce Parkinson-like problems. It helps people keep their mobility with fewer falls and fractures," he said.

All too often, when elderly patients fall, they're hospitalized and end up dying, Kappos added.

The study was funded by the Research Fund of Rigshospitalet, the University of Copenhagen, and the Danish Ministry of Health. Nørgaard and Kappos have disclosed no relevant financial relationships.

Congress of the European Academy of Neurology (EAN) 2019: Abstract O1104. Presented June 29, 2019.

For more Medscape Neurology news, join us on Facebook and Twitter


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.