Does quitting smoking reduce dementia risk—and how certain is that?

01

What changed from the 2019 guideline?

The current edition is from 2026, not merely the 2019 guideline with a new review date. Its tobacco section discusses newer large cohorts and validates the existing recommendation. It retains strong recommendation and low certainty rather than upgrading cessation to proven dementia prevention for each individual.

The question concerns risk reduction in adults with normal cognition or mild cognitive impairment. It is not an evaluation of cessation as a cure for diagnosed dementia. WHO also considers tobacco's wider harms and the case for accessible support; low certainty here is not a recommendation to keep smoking.

[1]

02

Where the ‘nine years’ category came from

One cohort cited by WHO, the US ARIC study, examined dementia in 13,002 participants and compared current and former smokers with never smokers. Its nine-year cut point came from the average interval between two study visits: people who already reported former smoking at the earlier visit were classified as having quit at least nine years earlier. This was a study grouping, not a discovered biological reset day.

Long-term quitters did not have a statistically significant excess dementia rate relative to never smokers. That does not prove that everyone's risk becomes identical at nine years. In the same paper, smoking groups did not differ significantly in measured cognitive decline. Incident dementia and change in cognitive-test scores are different outcomes; one result cannot promise immediate improvement in memory.

[2]

03

Why the uncertainty matters without cancelling the action

Smoking history may be self-reported, people may quit because they are already unwell, and quitters may differ in other behaviours. Death and loss to follow-up also matter: people who die earlier cannot later receive a dementia diagnosis. ARIC examined these problems and adjusted its analyses, but that does not make quitting a randomly assigned intervention.

WHO identifies these as continuing research gaps, including better lifetime smoking measurement and studies of cessation timing. A small dementia count among older surviving smokers would not show that smoking protects the brain. Nor should a risk estimate from a cohort be converted into your personal dementia probability or a timer counting down until safety.

[1][2]

04

Long-term risk and a current memory concern need separate answers

Dementia can affect memory, thinking and everyday activities; it is not another name for every lapse of concentration. New, persistent or worsening cognitive concerns belong with qualified local care, not a self-diagnosis based on years since quitting. For someone already living with dementia, care and support still matter, but this risk-reduction evidence cannot promise reversal.

You can ask separately for an assessment of the concern and for help stopping tobacco. In England, NHS local stop-smoking services are a help route. This page needs no cognitive score, symptom diary or smoking history. There is no need to calculate a personal dementia risk before seeking cessation support.

[3][4]

05

What to keep in mind

• Use the current 2026 recommendation without upgrading its certainty.
• A cohort's duration category is not a personal reset deadline.
• Memory assessment and cessation support are separate, compatible needs.