Two of our articles name medication as a cause of cognitive fog and move on. What actually causes brain fog lists it among several causes; the blood-test piece calls it something bloods will not catch. This is the follow-up neither gave it.
What anticholinergic burden actually is
Acetylcholine is used throughout the body — gut, bladder, airways, salivary glands — and in the brain circuits supporting attention and memory. Many drugs block its muscarinic receptors. For a few that blockade is the point; for most it is incidental. So someone can take four or five drugs, none described as an anticholinergic, and carry a substantial cumulative load. That load is anticholinergic burden, and it is additive.
The immediate effects are recognisable: dry mouth, constipation, blurred near vision, urinary hesitancy, and centrally, sedation, slowed thinking and poor recall. In older people the same load raises the risk of delirium and falls.
The classes that carry it
The usual suspects, all ordinary drugs:
- First-generation antihistamines — diphenhydramine, chlorphenamine, promethazine, hydroxyzine. Mostly available without prescription, and present in many combination cold remedies and over-the-counter sleep aids.
- Tricyclic antidepressants, particularly tertiary amines: amitriptyline, imipramine, clomipramine. Often prescribed at low doses for pain, migraine prevention or sleep rather than depression, which is why people forget they are on one.
- Bladder antimuscarinics — oxybutynin, tolterodine, solifenacin.
- Some antipsychotics, and some antiemetic and antivertigo drugs such as prochlorperazine and cyclizine.
- Anticholinergic antiparkinson drugs such as procyclidine, and several antispasmodics and muscle relaxants.
The 2023 American Geriatrics Society Beers Criteria advise avoiding first-generation antihistamines in older adults — a strong recommendation on moderate-quality evidence — and say the same of strongly anticholinergic tricyclics on high-quality evidence.
The scales, and why they disagree with each other
Burden is scored with published scales. The Anticholinergic Cognitive Burden scale is the most used, rating each drug 1, 2 or 3 and summing them; others are the Anticholinergic Risk Scale, the Anticholinergic Drug Scale and the Drug Burden Index.
They are not interchangeable, and the gap is larger than most articles admit. A study of 330,684 Western Australian adults aged 65 and over compared six instruments on the same data: the proportion classified as exposed ranged from 13.1% to 53.6% and high exposure from 0.7% to 6.9%. In a memory clinic of 657 people, the ACB scale called 14.8% high burden where a newer Swedish scale called 27.2%. A burden score is a prompt for a conversation, not a measurement of your risk.
The cohort evidence, and the caveat that has to travel with it
The prospective Adult Changes in Thought study followed 3,434 people aged 65 and over for a mean of 7.3 years, during which 797 (23.2%) developed dementia. Cumulative exposure was measured in total standardised daily doses over ten years, excluding the most recent twelve months to avoid catching drugs prescribed for early symptoms.
The dose-response trend was significant (P < .001). The individual hazard ratios are worth reading in full: 0.92 (95% CI 0.74–1.16) for 1–90 doses, 1.19 (0.94–1.51) for 91–365, 1.23 (0.94–1.62) for 366–1,095, 1.54 (1.21–1.96) above 1,095. Only the highest category reached significance; the three below it all cross 1.00.
A much larger nested case-control study in JAMA Internal Medicine used UK primary-care records: 58,769 people with dementia and 225,574 matched controls aged 55 and over. Adjusted odds ratios rose from 1.06 (1.03–1.09) in the lowest exposure band to 1.49 (1.44–1.54) in the highest. By class at the highest exposure: antipsychotics 1.70, bladder antimuscarinics 1.65, antiparkinson drugs 1.52, antiepileptics 1.39, antidepressants 1.29. Lag windows of 3–13 and 5–20 years gave similar results.
Why none of this proves causation
Confounding by indication is the central problem, and it is not a technicality.
The conditions these drugs treat — depression, sleep disturbance, urinary urgency, constipation, dizziness — are themselves early features of neurodegenerative disease, sometimes years before diagnosis. People in the prodromal phase get prescribed more of exactly these drugs. A study finding they later developed dementia may be watching the disease choose the prescriptions.
A 2018 BMJ case-control study of 40,770 dementia cases and 283,933 controls aged 65 to 99 contains the sharpest test of this. Overall, any drug scored 3 on the ACB scale carried an odds ratio of 1.11 (1.08–1.14). Risk rose with exposure for antidepressant, urological and antiparkinson drugs, and that held even for exposure 15 to 20 years before diagnosis. But gastrointestinal anticholinergics scored 3 were not distinctively linked to dementia at all.
That is the finding to sit with. If the anticholinergic property itself were doing the damage, gut anticholinergics should carry the same risk as bladder ones. They did not. The authors said the association could reflect a class-specific effect or drugs used for very early symptoms, and recommended studying classes rather than summed scores.
The honest position: the association is consistent across very large datasets and survives long lag windows, and it is still not established as causal. Reducing unnecessary exposure is worth doing on the short-term effects alone.
The short-term effects, which are not in doubt
This is where the evidence is cleanest and the advice firmest. A randomised crossover trial in the Iowa Driving Simulator gave 40 drivers with seasonal allergic rhinitis, on separate weeks, fexofenadine 60 mg, diphenhydramine 50 mg, alcohol to roughly 0.1% blood alcohol concentration, or placebo, then an hour of driving. Driving performance was poorest after diphenhydramine — worse than after the alcohol. Lane keeping was impaired by both alcohol and diphenhydramine relative to fexofenadine; performance on fexofenadine resembled placebo.
The detail that matters most: self-reported drowsiness did not predict impairment. People could not tell. A later simulator study reproduced the pattern — diphenhydramine impaired psychomotor performance, levocetirizine and fexofenadine did not. If fog arrived with a new antihistamine, or you take a sedating one nightly for sleep, that is a specific and testable explanation.
Benzodiazepines and z-drugs: a different shape of evidence
These get bundled in with anticholinergics, and the data behave differently. The same cohort examined cumulative benzodiazepine use over ten years: hazard ratios for dementia were 1.25 (1.03–1.51) for 1–30 doses, 1.31 (1.00–1.71) for 31–120, and 1.07 (0.82–1.39) for 121 or more. Higher use was not associated with faster cognitive decline, and the authors concluded the results do not support a causal association.
A risk that fails to rise with dose, and is highest in those who took least, is the signature of prescribing responding to early symptoms rather than causing them.
That is not a clean bill of health. The Beers Criteria advise avoiding benzodiazepines in older adults — a strong recommendation — because sensitivity increases with age and they raise the risk of cognitive impairment, delirium, falls, fractures and crashes. Z-drugs such as zolpidem carry the same advice. The case against them rests on falls, next-day impairment and dependence, all well established, not on a dementia argument that has not held up.
For long-term insomnia the best-evidenced treatment is not a drug: CBT for insomnia outperforms everything we sell, and we have said so. It is also the route out of a hypnotic that has stopped working, alongside the non-melatonin options.
What to actually do
Ask for a structured medication review — a deprescribing review — with your GP or a community pharmacist. Not a repeat prescription: a review.
Take a genuinely complete list: prescriptions, over-the-counter purchases, antihistamines, sleep aids, cold and flu remedies, travel-sickness tablets, supplements including ours. Over-the-counter agents are missed most often — a study of 45,183 US long-term care residents found 33.4% using anticholinergics, first-generation antihistamine use reaching 29.5% among residents with dementia, and noted claims data underestimates burden because it misses them.
Say when the fog started and what changed then. Ask whether anything on the list is strongly anticholinergic, whether a less anticholinergic alternative exists, and whether anything is continuing out of habit.
Expect a modest result rather than a transformation. A 2025 hospital deprescribing study shows how hard this is: of 301 patients with an ACB score of 3 or more, 21 took part, and mean scores moved from 3.85 on admission to 3.44 at three months, with slight changes in cognition and delirium and none in mood or quality of life. Deprescribing is the right direction, not a proven cognitive intervention.
Where we sit
Medication burden is one of the more actionable items on the fog list, and the one most often skipped in favour of buying something. It sits alongside the modifiable factors behind longer-term risk, covered in what the 45% figure actually means.
Our own products belong on that list and are not exempt from the interaction conversation — we have set out where they and mental-health medications intersect. No supplement reduces anticholinergic burden, and adding one to a long medication list is the opposite of what this article argues for. If symptoms are persistent, worsening, or came with anything neurological, that needs a doctor, not a supplement; if the worry is memory, this is the more useful article.
This article is not medical advice and is not a basis for changing any medication. Speak to your prescriber or a pharmacist.
Sources: Gray SL. Cumulative use of strong anticholinergics and incident dementia. JAMA Intern Med. 2015 · Coupland CAC. Anticholinergic drug exposure and the risk of dementia. JAMA Intern Med. 2019 · Richardson K. Anticholinergic drugs and risk of dementia. BMJ. 2018 · Gray SL. Benzodiazepine use and risk of incident dementia or cognitive decline. BMJ. 2016 · Weiler JM. Effects of fexofenadine, diphenhydramine and alcohol on driving. Ann Intern Med. 2000 · Simulated car-driving study of levocetirizine, fexofenadine and diphenhydramine. Hum Psychopharmacol. 2016 · Differences in anticholinergic exposure assessment across burden scales. Drugs Aging. 2026 · Anticholinergic medication use in long-term care residents. J Am Geriatr Soc. 2026 · Feasibility of collaborative deprescribing to reduce anticholinergic burden. Basic Clin Pharmacol Toxicol. 2026 · AGS 2023 updated Beers Criteria. J Am Geriatr Soc. 2023