Does It Keep You Sharp?

Does It Keep You Sharp?

You'll have seen the headline. Crosswords keep the mind young; crosswords hold off dementia; do the puzzle and buy yourself years.

We sell crosswords. If any of that were solidly established we'd be the first to say so. It is not - and nearly every study in this field measures one of two very different things. Level is how good your cognition is now. Rate of decline is how fast you lose it. Puzzle-doers reliably score better on the first. The best-designed studies find they do no better on the second.

The big one, read properly

The largest study is PROTECT, run from Exeter and King's College London: 19,078 adults aged 50 to 93. All fourteen cognitive measures showed a significant effect of word-puzzle frequency, and those who never did puzzles scored worst on every one. The framing that travelled the world: brain function ten years younger.

It was cross-sectional. A snapshot, with no follow-up and no dementia outcome. Its lead author, Dr Anne Corbett, was plain about it:

We can't say that playing these puzzles necessarily reduces the risk of dementia in later life.

Higher start, same fall

The Aberdeen Birth Cohort followed 498 people over fifteen years, all born in 1936 and all of whom had sat an intelligence test at age eleven - so their childhood ability was known. Engagement predicted a higher score. It did not predict the slope. Dr Roger Staff summarised his own findings as showing that engagement "does not protect an individual from decline but imparts a higher starting point from which decline is observed". Four other cohorts, running up to twenty-one years, agree.

Then the sharpest test. In the Lothian Birth Cohort, 1,091 people tested at seventy also had IQ scores from age eleven. Once childhood intelligence and social class were accounted for, the link between intellectual activity and cognition disappeared. Only physical activity survived.

Which way the arrow points

The researchers put the problem better than we can:

Put simply, this study design does not answer whether completing crossword puzzles reduces one's risk of cognitive decline or if cognitive decline reduces the likelihood that one will complete crossword puzzles.

People in the earliest, undiagnosed stage of dementia quietly give up the crossword, because it's stopped being fun. That withdrawal begins years before diagnosis, and produces exactly the pattern these studies find.

About those 2.54 years

The most-quoted crossword statistic deserves unpicking. A 2011 study found accelerated memory decline began 2.54 years later in crossword players than non-players.

It was observational, and it rested on seventeen crossword players. Every one of the 101 people in that analysis developed dementia regardless: it was a delay in the onset of decline, not a delay in dementia.

The half never repeated is that once decline began, the players fell faster. A shorter, steeper descent - which is what cognitive reserve predicts, reserve being a theory about hiding damage rather than preventing it.

The one real trial

There's one properly randomised trial, and crosswords did well in it: 107 people with mild cognitive impairment, seventy-eight weeks, crosswords against computerised brain-training games. Crosswords won on cognition and daily function, with slightly less shrinkage on MRI.

The caveats matter. There was no do-nothing arm, so it shows crosswords beat brain-training apps, not that they beat nothing. The margin barely cleared significance. And on progression to dementia - six of 56 against eight of 51 - the difference vanished.

What we won't tell you

The Alzheimer's Society is unambiguous: "There is no strong evidence that brain training activities will reduce a person's risk of developing dementia." It adds that people should be cautious of commercial packages claiming otherwise. That warning is aimed at sites like this one, and it's correct.

The Lancet Commission lists fourteen modifiable risk factors for dementia. Doing puzzles is not among them. Education is.

So: crosswords are absorbing, verbal, sociable, difficult in a way that rewards you, and cost almost nothing. Nobody has shown they do the slightest harm. Whether they defend your brain is unsettled, and may stay that way - the trial that would settle it would have to run for decades.

Do them because you like them. That was always enough.