Anticholinergic antidepressant drugs and long-term risk of dementia

Study type
Protocol
Date of Approval
Study reference ID
19_098
Lay Summary

Many commonly used drugs are called “anticholinergic,” because they block acetylcholine, a compound that brain and nerve cells use to transmit impulses. Among such medications are drugs for depression, bladder control problems, digestive tract problems, and many other conditions. Side effects of anticholinergic drugs may include dry mouth, difficulty urinating, and constipation, among others. Recent epidemiologic studies have raised concerns that long-term use of drugs with anticholinergic properties, particularly by older individuals, may promote the development of dementia. However, it is not clear if this is a true effect of the anticholinergic drugs or just a reflection of the fact that people taking these drugs may not be in good general health. The aim of this study will be to compare the risk of dementia between patients using drugs for the same indication where one of the drugs is anticholinergic and the other is not. Specifically, we will compare the risk of dementia between long-term users of paroxetine (an anticholinergic antidepressant) and other antidepressants that are not anticholinergic.

Technical Summary

Anticholinergic drugs induce cognitive impairments, and accordingly their use has been discouraged in older patients. Beyond these acute effects, some observational studies have pointed to an association between chronic anticholinergic drug exposure and an increased risk of developing dementia. However, as the comparison groups for these studies have generally been non-users of anticholinergic drugs, confounding by indication remains a possible explanation for these associations. Accordingly, we will undertake a study to compare the risk of dementia with exposure to anticholinergic drugs for depression (amitriptyline and paroxetine) in comparison with drugs for depression that are not anticholinergic (fluoxetine and sertraline). We will conduct a time-to-event analysis for the outcome of a diagnosis of dementia with a Cox proportional hazards model, analysing cumulative duration of drug exposure as a time-varying covariate, and using propensity score methods to adjust for potential confounders.

Health Outcomes to be Measured

• Alzheimer’s disease
• Vascular dementia
• All-cause mortality

Collaborators

Andrew Mosholder - Chief Investigator - Food and Drug Administration - FDA
Andrew Mosholder - Corresponding Applicant - Food and Drug Administration - FDA
Benedict Wong - Collaborator - Food and Drug Administration - FDA
Catherine Callahan - Collaborator - Food and Drug Administration - FDA
David Graham - Collaborator - Food and Drug Administration - FDA
Kira Leishear White - Collaborator - Food and Drug Administration - FDA
Marie Bradley - Collaborator - Food and Drug Administration - FDA
Yuegin Zhao - Collaborator - Food and Drug Administration - FDA