select search filters
roundups & rapid reactions
before the headlines
Fiona fox's blog

expert reaction to two studies on word and number puzzles improving adult cognitive function

Research published in the International Journal of Geriatric Psychiatry demonstrates that the frequency of word puzzle use is directly related to cognitive function in adults aged 50 and over.


Prof Til Wykes, Professor of Clinical Psychology and Rehabilitation, Institute of Psychiatry Psychology & Neuroscience, King’s College London (IoPPN), said:

“The two papers describe cross sectional analyses of people over 50 (who probably wouldn’t at this age think of themselves as old). The average age is 62. They find that those who use either number or word puzzles often have better cognition on their tests of memory, attention and reasoning after controlling for age, education and gender. But, and this is a big but, there are several different explanations of the results. First, the analysis is not without its problems. Education is controlled and is usually considered as a proxy for IQ but we know that expectations of education have changed radically over time and therefore older people may not have had the opportunity for higher education thus the results may be biased, especially for women. Those with better cognition may have enjoyed puzzles that they were good at because they were brighter and this has not been excluded as an explanation. As the authors spell out in one of the papers – we do not know the causal and effect.

“Secondly, there is the cohort itself which is unlikely to contain poorer individuals as you need the internet and time to complete the study. I admit that I tried to enrol and gave up as it was so time consuming. Data quantity does not trump data quality. Biased data can only produce biased results. Researchers should be trying to recruit the older participants of Eastenders (and make it easy) as well as the time rich middle classes if we are going to generalise study results.”


Prof Oliver Braddick, Emeritus Professor of Experimental Psychology, University of Oxford, said:

“It may be the case that regular puzzle solving is protective against the onset of dementia, but these studies certainly don’t prove it.  The major point is that the study demonstrated a correlation between puzzle use and cognitive performance at a particular point of time.  This could just as well result from those with “sharper brains” being more likely to want to do puzzles, as the puzzles having any causative role.  It’s likely that you won’t do puzzles so often if you find you aren’t very good at solving them.  Stronger evidence, though not proof, would be provided if any decline in cognitive performance, over say a 5 year period, was less in those who did more puzzles.

“This point is briefly acknowledged in the discussion section of the papers, but is played down, I feel, in the press release, especially since the quotation from Dr Corbett refers repeatedly to “improvement” – there is no measure here of whether individuals improved – the findings are of differences across groups measured on a single occasion.  I think it would be seriously misleading to use words like “improve cognitive function” in a headline.

“Any study such as this has to be careful that the two measures being associated don’t do so because they share a common factor, such as age or educational level.  The studies here did control for these two as covariates, which is satisfactory although the analysis is not reported in great detail, so we don’t know whether these effects might interact or be “non-linear” – i.e. not act in a uniform way across the range.  General health might also be a factor affecting both measures. We also aren’t told how large were the effects of age and education, compared to the association with puzzle use.

“In evaluating this work, you should be clear that it doesn’t say anything directly about people with dementia – this was a volunteer group who have access to, and use, the internet, so they are likely to reflect the more active and capable range of older people.  They are described as “healthy” although it is not clear whether this is based on anything beyond not having a diagnosis of dementia.”

“Finally, this was a very large group of nearly 20,000 people.  Such large studies provide very sensitive tests, but it must be realised that this means they can highlight very small effects in terms of the individual.  Dr Corbett is quoted in the press release in terms of the equivalent benefit in terms of younger age (“8 years younger” etc).  This way of presenting the data does not appear in the paper.  How much it means depends on how much performance changes in the age range – if decline is very shallow with age, a large age difference could correspond to a small difference in performance.  (Given the age range in the study, it’s also not clear whether 8 years younger refers to 60 vs 68, or 84 vs 92 – these are unlikely to show the same amount of change.)”


Prof Robert Howard, Professor of Old Age Psychiatry, University College London (UCL), said:

“These papers report an association between cognitive performance and engagement in word and number puzzles. It’s important to caution that this doesn’t necessarily imply that the puzzles have improved cognitive ability, because the data can’t show this and also can’t exclude the obvious explanation that people who are more highly educated or intellectually able are more likely to enjoy word and number puzzles.

What we do know from controlled clinical trials is that cognitive training has proved consistently disappointing in preventing cognitive decline in healthy people or improving cognitive function in people who have dementia. If people enjoy doing crosswords or Sudoku, that’s fine. But, this is the only reason to engage with these hobbies as there is no good evidence that they will protect you from cognitive decline or dementia.”


Dr James Pickett, Head of Research at Alzheimer’s Society, said:

“This research, much like research we’ve seen in the past, suggests that regularly enjoying word and number puzzles has a positive impact on thinking skills.

 “Unfortunately it doesn’t yet mean that regular games of Sudoku or jigsaw puzzles will definitely prevent dementia. It’s an important first step – and we are proud to have helped fund the study as it lays the foundations for more research into the relationship between a love of ‘puzzling’ and reducing dementia risk.  This looks to be a well conducted study, although it can only show that puzzling and thinking skills are linked, not that puzzling will improve thinking skills.  It also didn’t look for a link with dementia risk.

“With nothing yet to slow or stop dementia, prevention is key. If you enjoy puzzling, then continue, although don’t worry if you don’t. There are other ways we can reduce our risk of developing dementia by taking steps towards a healthy lifestyle, eating a balanced diet, avoiding smoking and heavy drinking, and exercising regularly.


Prof Martin Rossor, NIHR National Director for Dementia Research, University College London Hospitals (UCLH), said:

“The PROTECT study is beginning to provide valuable information. Those who engage in word and number puzzles do better on memory and other cognitive tests. The challenge is to know what is cause and effect; people with better memory and other cognitive skills may enjoy doing puzzles. The challenge is to determine if starting to do puzzles improves our brain function”


The relationship between the frequency of number‐puzzle use and baseline cognitive function in a large online sample of adults aged 50 and over’ by Brooker et al. and ‘An online investigation of the relationship between the frequency of word puzzle use and cognitive function in a large sample of older adults’ by Brooker el al. was published in International Journal of Geriatric Psychiatry at 04:00 UK time on Thursday 16th May.  


Declared interests

Prof Til Wykes: “I have co-created a software programme CIRCuiTS for cognitive remediation therapy for schizophrenia and have had grants from MRC and NIHR to develop and evaluate it”

Prof Robert Howard: “No competing interests.”

Prof Martin Rossor: “I am the NIHR director for dementia research but have no involvement in funding decisions and so do not see a conflict of interest”

in this section

filter RoundUps by year

search by tag