This week, JAMA Internal Medicine published online an interesting paper, “Cumulative Use of Strong Anticholinergics and Incident Dementia: A Prospective Cohort Study.” They found that exposure to anticholinergic drugs significantly increased the risk of developing dementia. This study has important implications for those who prescribe and take psychiatric drugs but it is worthwhile describing the study in some detail before discussing those implications.
First, some context. Acetylcholine (ACh) is a neurotransmitter (a chemical that propagates a nerve signal from one cell to the next) that is found throughout the body. It is involved in gut motility, visual acuity, heart rate, and secretions. In the brain, its activity is linked to memory and movements. There are many drugs that blocked ACh receptors (the proteins that allow nerve transmission to occur). In psychiatry, anticholinergic drugs are used to combat some of the neurological effects of the neuroleptic drugs such as muscle stiffness and tremor. But there are also many drugs used to treat depression and psychosis that have anticholinergic effects.
Drugs that block the action of Ach – anticholinergics – often cause dry mouth, constipation, rapid pulse, urinary retention (difficulty emptying one’s bladder), and blurred vision. But these drugs can impair memory. What this study addressed was the extent to which these drugs are associated with a new diagnosis of dementia and also, of importance, whether individuals who discontinue these drugs remain at higher risk of developing dementia.
They evaluated adults 65 years or older who received their health care within one integrated system in Seattle. This allowed them to access records for all drugs the individuals had been prescribed for 10 years leading up to and during the study. They excluded individuals who had a prior diagnosis of dementia and they screened everyone for signs of dementia at the outset of the study and then biennially for the duration of the study. In total they followed 3434 individuals. They calculated total anticholinergic exposure in a way that took into account both the daily dose prescribed and the total number of days the dose was prescribed to determine a total standardized daily dose.
This was a careful study. To avoid what they call protopathic bias – the possibility that one of these drugs was prescribed to treat depression or insomnia that in retrospect was an early sign of dementia – they excluded from their calculations all drugs prescribed in the year prior to the onset of dementia. They statistically controlled for factors that could contribute to dementia: age, sex, smoking status, years of education, body mass index. They accounted for health care status and several specific health conditions that could impact the results – coronary heart disease, Parkinson’s disease, diabetes, stroke, hypertension, depression, and concurrent benzodiazepine use.
There were two major findings. The first is that total exposure to anticholinergic drugs increases the risk of developing dementia. But of further concern, these effects were seen even if the drugs had been stopped years before the onset of dementia.
They found that there was a dose response risk for developing dementia, i.e. those with greater exposure had an increased risk. The Hazard Ratio for those in the highest exposure group was 1.54 with a 95% confidence interval of 1.21-1.96. This means that those who had the highest exposure to the drug had a 1.5 greater likelihood of developing dementia than those who had no exposure.
This study has general implications since there are so many drugs on the market that have anticholinergic effects. In our age of increasing polypharmacy, especially as people age, this is a warning call to be judicious in our prescribing.
But the study has particular implications for psychiatry. First of all, many psychiatric drugs have anticholinergic effects. The older tricyclic antidepressants as well as some older neurolpetics were on the researchers’ list. But there were also some newer ones – paroxetine (Paxil), olanzapine (Zyprexa) and clozapine (Clozaril).
The paper is of particular interest to me because it relates to one of the very first research studies in which I participated. It was a study that tested the hypothesis that the dose of haloperidol that caused minimal neurological side effects – the neuroleptic threshold (NTH)– as measured with a simple, brief examination was also the effective dose. Going higher on the dose yielded more discomfort but no further reduction of psychotic symptoms. PET studies have supported this hypothesis: 70% blockade of dopamine receptors yields an anti-psychotic effect and 80% yields significant neurological side effects. Anticholinergic drugs are used to reduce the neurologic symptoms. But they should never be needed. If someone has these symptoms the thing to do is to lower the dose not add another drug. If the person still has psychotic symptoms on the lower dose, increasing it is not likely to yield any further improvement.
The NTH paper was published just before the newer neuroleptics were coming to the market. Whereas in the late 80’s, academic psychiatrists were talking about moderating dose, by the late 90’s everyone was talking about how safe the newer drugs were. Polypharmacy – once shunned in academic circles – came back into favor. Most psychiatrists in practice today were trained in the 1990’s or 2000’s.
Beyond that, my one gripe with this paper is that the list of drugs with anti-cholinergic effects did not seem complete. Another site, which appears to do a good job of tracking drugs’ anticholinergic burden, includes many more psychiatric drugs which are in common use. A recent study, reported the risk of dementia among people diagnosed with schizophrenia to be about double the risk of the general population. We know that many people are being exposed to psychiatric drugs at very young ages and then are often on them for many years. Although there is much discussion in the psychiatric literature about cognitive impairment in people who are diagnosed with schizophrenia it is most often discussed as a result of the psychiatric disorder not of the drugs the person is taking.
I sometimes wonder how we know what we know. For instance, a psychiatrist goes through extensive training and ends up with a sense of having a general grasp on the field. But no one can read everything that has been published. Much of our learning is still in the apprentice mode – we learn from those around us. Most residents never learn about the neuroleptic threshold. And many emerge with the belief that the cognitive impairments associated with schizophrenia are due to the “disease” rather than the drugs we give.
I have come back to this theme again and again. This is not an anti-drug discussion. It is a respect-the-drug discussion. We need to use these drugs with caution. Dose matters. Length of time a person is on them matters. Polypharmacy matters.
Acknowledgment: I want to thank the always insightful and incisive Nev Jones for discussing this paper with me and directing me to some helpful papers and websites pertinent to this topic.