Media dementia scare over hay fever and sleep drugs

"Hay fever tablets raise risk of Alzheimer's," is the main front page news in the Daily Mirror. The Guardian mentions popular brand names such as Nytol, Benadryl, Ditropan and Piriton among the pills studied.

But before you clear out your bathroom medicine cabinet, you might want to consider the facts behind the (somewhat misleading) headlines.

The first thing to realise is although some of these drugs can be bought over the counter (OTC), in the US, OTC drugs are typically provided by a private health company. So the study was able to partially track the effects of OTC as well as prescription medication (which would be impossible in the UK).

These were medicines that have an "anticholinergic" effect, including some antihistamines, antidepressants and drugs for an overactive bladder.

If you have been prescribed these medicines, do not stop taking them without speaking to a doctor first. The harms of stopping might outweigh any potential benefits.

That said, this large, well-designed US study suggested those taking the highest levels of anticholinergic prescribed medicines were at a higher risk of developing dementia compared with those not taking any.

Importantly, the increased risk was only found in people who took these medicines at the equivalent of once every day for more than three years. No link was found at lower levels.

However, this shouldn't make us complacent. These are not unrealistic doses of medicines, so the results may be applicable to a significant proportion of older adults.

Furthermore, we can't say if reducing the amount of anticholinergic medicines will reduce the risk of dementia to normal.

The bottom line? Do not stop taking medications without a full consultation with a doctor. It may do more harm than good.

Where did the story come from?

The study was carried out by researchers from the University of Washington and Group Health Research Institute.

It was funded by the National Institute on Aging, the National Institutes of Health, and the Branta Foundation.

The study was published in the peer-reviewed medical journal, JAMA Internal Medicine.

A number of the study's authors reported receiving research funding from pharmaceutical companies, including Merck, Pfizer and Amgen.

The story made almost all the newspapers and many online and broadcast services, with front page "splash" stories in the Mirror and The Times.

This coverage lacked the necessary caution and has all the hallmarks of a media scare story.

The media reporting of the study generally took the findings at face value and did not highlight the potential risks associated with stopping medications suddenly.

Any changes to medicines should be made after full consultation with a medical professional and should factor in your individual circumstances.

Poor media reporting also included: 

  • Failing to make it clear the antihistamines involved were only one, older class known to cause drowsiness (and avoided by many people because of this) – a mistake made by The Times, The Independent and the Mail.
  • Naming a brand (Benadryl) focused on by researchers that has a completely different drug in it in the UK – a mistake made by The Times, the Mail, The Independent and The Telegraph.
  • Having headlines that didn't make it clear the association was only seen in people over the age of 65 – a mistake made by most papers, except The Times.
  • Playing fast and loose with statistics – the Mail said up to 50% of elderly people could be taking an anticholinergic, a statement so vague it could mean half of them take them, or none take them.
  • Several media outlets also mistakenly reported a raised risk of dementia linked to taking 4 mg/day of diphenhydramine for three years, but that should have been 4 mg/day of chlorpheniramine (or 50 mg/day of diphenhydramine) for three years.

Today, the Mirror, with its front page "Shocking new report" headline, was perhaps the most overblown coverage, although it was among one of the most factually correct.

The Telegraph also did well to include suggestions for alternative antihistamines and antidepressants that could be used by over-65s.

What kind of research was this?

This was a prospective cohort study looking at whether the use of medicines that have anticholinergic effects are linked to dementia or Alzheimer's disease.

Medications with anticholinergic effects are commonly used for a wide variety of conditions affecting older adults, such as an overactive bladder. 

Some of these medications can be bought over the counter, such as antihistamines like chlorphenamine – which is mainly sold under the brand name Piriton and is not to be confused with other antihistamine products, such as Piriteze – and sleeping pills, such as diphenhydramine, sold under the Nytol brand.

The study authors stated the prevalence of anticholinergic use in older adults ranges from 8% to 37%.

A prospective cohort study cannot definitively prove this drug class causes Alzheimer's disease or dementia, but it can show they are linked in some way. Further research is needed to properly investigate and explain any links identified.

What did the research involve?

The research team analysed data on 3,434 US people aged over 65. These people had no dementia at the start of the study.

The study's participants were tracked for an average of 7.3 years to see who developed dementia or Alzheimer's disease.

The researchers also gathered information on what anticholinergic medications they were prescribed in the past as well as a partial record of past OTC use.

The researchers' main analysis looked for statistically significant links between these prescribed medications taken in the past 10 years and the likelihood of developing dementia or Alzheimer's disease.

Cases of dementia and Alzheimer's were first picked up using a test called the Cognitive Abilities Screening Instrument, which was given every two years.

This was followed up with investigations by a range of specialist doctors, and laboratory tests, to arrive at a consensus diagnosis.

Medication use was ascertained from a computerised pharmacy dispensing database that included the name, strength, route of administration (such as in tablets or in syrup), date dispensed, and amount dispensed for each drug. This was linked to each individual's electronic record at Group Health Cooperative, a US health care and insurance system, plan so it was personalised.

Usage in the most recent one-year period were excluded because of concerns about bias. This bias could occur when a medication is inadvertently prescribed for early signs of a disease that has not yet been diagnostically detected. For instance, medications may be prescribed for insomnia or depression, which can be early symptoms of dementia.

Drugs with a strong anticholinergic effect were defined as per an American Geriatrics Society consensus panel report. Data for the medicines was converted into an average daily dose, and this was added up over the number of years people were taking them to estimate their total cumulative exposure.

This cumulative exposure was defined as cumulative total standardised daily doses (TSDDs).

The statistical analysis adjusted for a range of potential confounders identified from past research, including:

  • demographic factors such as age, sex, and years of education
  • body mass index
  • whether or not they smoked
  • their exercise levels
  • self-rated health status
  • other medical problems, including hypertension, diabetes, stroke, and heart disease
  • whether they had a variant of the apolipoprotein E (APOE) gene
  • Parkinson's disease
  • high levels of depressive symptoms
  • cumulative use of benzodiazepine medicines – this could indicate a sleep or anxiety disorder

What were the basic results?

The most common anticholinergic classes used over the long term were antidepressants, antihistamines, and bladder control medicines.

During an average (mean) follow-up of 7.3 years, 797 participants (23.2%) developed dementia. Most people diagnosed with dementia (637 of the 797, 79.9%) had Alzheimer's disease.

Overall, as cumulative anticholinergic exposure over 10 years increased, so did the likelihood of developing dementia, including Alzheimer's disease. Results were reported to stand up to secondary analyses.

For dementia, cumulative anticholinergic use (compared with no use), was associated with:

  • for TSDDs of 1 to 90 days, a confounder adjusted hazard ratio (HR) of 0.92 (95% confidence interval [CI], 0.74-1.16) 
  • for TSDDs of 91 to 365 days 1.19 (95%CI, 0.94-1.51)
  • for TSDDs of 366 to 1,095 days 1.23 (95%CI, 0.94-1.62)
  • for TSDDs of more than 1,095 days 1.54 (95% CI, 1.21-1.96)

It is important to note the only statistically significant result was in the group with the highest long-term exposure level.

At standardised cumulative doses of between 1 and 1,095 days (three years), there was no statistically significant increase in incidence of dementia compared with those with no exposure.

However, those in the highest cumulative anticholinergic exposure group had an increased risk (hazard ratio of 1.54) for developing dementia compared with those with no anticholinergic exposure over the previous 10-year period .

How did the researchers interpret the results?

The researchers' conclusions were grounded, and warned of a potential risk if the results were true. They said that, "Higher cumulative anticholinergic use is associated with an increased risk for dementia.

"Efforts to increase awareness among healthcare professionals and older adults about this potential medication-related risk are important to minimise anticholinergic use over time."


This large US prospective cohort study suggests a link between those taking high levels of anticholinergic medicines for more than three years and developing dementia in adults over 65.

The main statistically significant finding was in a group taking the equivalent of any of the following medications daily for more than three years:

  • xybutynin chloride, 5mg
  • chlorpheniramine maleate, 4mg
  • olanzapine, 2.5mg
  • meclizine hydrochloride, 25mg
  • doxepin hydrochloride, 10mg

These are not unrealistic doses of medicine, so the results may be applicable to a significant proportion of older adults.

The main limitations of the research were recognised and openly discussed by the study authors. Although we don't expect them to have significantly biased the results, we cannot rule out the possibility.

These limitations include the potential misclassification of "exposure". This is possible because some anticholinergic medicines are available without a prescription – called "over-the-counter" medicines. These could have been partly missed in this study, which relied on a database of prescribed medicines and a partial record of over-the-counter medications.

It is therefore possible people who were reported to have no exposure may actually take, for example, regular doses of Piriton for hay fever without needing a prescription.

A related point is there is no guarantee the prescribed medications were actually taken – although it is likely they were, especially in the groups in the higher exposure categories.

Finally, we don't know whether these results can be generalised to other groups of people. The study sample was overwhelmingly white (91.5%) and university educated (66.4%). The findings will need replication in studies that recruit larger and more diverse participants to reflect wider society.

Studies are needed to better understand whether any increase in dementia risk is counteracted after people stop using anticholinergic medicines.

While there are biologically plausible theories, the mechanism by which anticholinergics might contribute to dementia risk is not well understood.

If you have been prescribed anticholinergic medicines, do not stop taking them without speaking to your GP first as everyone's circumstances are different. The harms of stopping might outweigh any potential benefits.

NHS Attribution