"Dementia rate falls as men behave themselves," The Times reports. A UK study of dementia trends over the last 20 years suggests that the number of men developing the condition has dropped significantly, possibly as a result of lifestyle changes.
The study showed an unexpectedly large fall in the number of people in the UK aged over 65 with signs of dementia between two time periods – 1989-94 and 2008-11.
There was a dramatic decrease in rates of dementia among older men, which almost halved for those aged 80 and over. Although rates for women also fell, the changes were much smaller. It is unclear why a similar strong trend was not seen in women.
Both the authors and the media speculate that positive trends in men's health – such as reduced smoking levels, improved diets, and more men taking regular exercise – could be responsible for the falling rates. While these are certainly plausible suggestions, they are unproven.
However, there is a strong body of evidence that healthy living – such as not smoking, keeping to a healthy weight, and taking regular exercise – does reduce the chances of getting dementia, although it is still no guarantee.
Where did the story come from?
The study was carried out by researchers from Newcastle University and Cambridge University, and was funded by the Medical Research Council and the National Institute for Health Research.
The Daily Mail, The Daily Telegraph and the Sun all go with the angle that "new men", as the Telegraph quaintly calls them, are healthier, so less likely to get dementia.
The Times, somewhat patronisingly, echoes this with the claim that these days, "men behave themselves".
The Guardian and BBC News are more cautious, saying that the "most likely explanation" is improvements in male health.
What kind of research was this?
This was a combination of two cohort studies. Both studies had two phases: a baseline, when people were interviewed and their mental health assessed, and another two years later, when the interviews were repeated.
The studies aimed to discover the proportion of people who got dementia during the two-year time period between interviews.
Researchers wanted to see if this number – called incidence – had changed. Cohort studies can find information like this, but they can't tell us much about the reasons behind the results.
What did the research involve?
Researchers replicated a study first carried out with 7,635 people aged over 65 between 1989 and 1994 from sites around the UK.
They then used the same questions to assess the mental health of a group of 7,762 people between 2008 and 2011 from three of the areas originally studied.
In both studies, people were assessed once, then again two years later, to see if they had developed dementia. This allowed researchers to calculate the incidence of dementia, or the number of new cases per 1,000 people. They looked to see whether the incidence had changed in the two decades since the early 90s.
The researchers checked their figures for factors that could have affected the results – for example, whether people who didn't respond to the original request to be interviewed were more likely to have dementia already – and also assessed the impact of where people lived.
The original study was done in several stages, meaning that more people dropped out between interviews, so the researchers tried to account for any effect of that.
They decided to use the same criteria for diagnosing someone with dementia that were used in the first study, even though the criteria for diagnosing dementia had changed since then. They said this was important to keep the results consistent.
Finally, they calculated incidence rates for people by age range and sex.
What were the basic results?
Incidence rates overall dropped from 20 cases of dementia for every 1,000 people in the early 1990s (95% confidence interval [CI] 16.9 to 23.8), to 17.7 cases per 1,000 (95% CI 15.2 to 20.9) in the more recent study.
However, looking at figures for men and women separately, the most dramatic drops in incidence were among older men. Rates almost halved for men aged 85 or over, from 71 in 1,000 (95% CI 36.5 to 140.2) to 38 in 1,000 (95% CI 22.5 to 64.2).
Rates for women declined a little in each age range, except among those aged 80 to 84, where they rose a little.
The researchers calculated how many people you would expect to get dementia each year in the UK, based on the 1991 rates but with an increased elderly population, and came up with a figure of 251,000 new cases a year. Based on the newer incidence figures, however, that dropped to 209,600 new cases of dementia a year.
How did the researchers interpret the results?
The researchers said their findings suggest that fears of "huge increases of people with dementia" in the future may be wrong. However, they warn this might only apply to parts of the world where health has improved.
They say future investment should be directed at improving health across the whole of the life course, so that people have good circulation, plenty of opportunities to engage in society, and good education. They say this may be more cost effective than strategies to diagnose dementia early.
They questioned whether "earlier and earlier identification of at-risk states" is helpful, saying that their findings of reductions in dementia "will be offset within services by the concept of 'early' detection" and changes in diagnostic criteria.
"Individuals who were previously not diagnosed with dementia or cognitive impairment are now being tested and referred for specialist assessment of ever milder stages with unknown prognostic significance," they said.
The figures from this study are striking, particularly the drop in the incidence of dementia in older men. However, we don't know what is behind this dramatic drop.
While it would be great to think that it's because men in their 80s are smoking less, exercising more and generally living healthier lives, we don't know whether this is true or if it can completely account for the big drop in dementia rates.
It's possible that the figures for men aged 80 and over are less reliable than those for younger age groups, as there were fewer men of this age interviewed.
For example, only 205 men aged over 85 were interviewed at baseline in 1991, with 110 interviewed at follow-up. The numbers for the second cohort were 364 men interviewed in 2008, with 193 interviewed at follow-up.
These small numbers are reflected in the large confidence intervals for these results. The smaller the numbers in a specific group, the greater the chance that any perceived effect is, in fact, the result of chance.
The researchers' decision to use the 1991 study criteria for deciding whether someone had dementia has been criticised by one expert, Dr Sujoy Mukherjee, consultant psychiatrist at West London Mental Health Trust and a member of the Dementia Strategic Clinical Network.
Diagnostic criteria have changed, and people who were not diagnosed with dementia in 1991 might be seen as having dementia today. Dr Mukherjee says this could undermine the findings. But using modern criteria would have made it difficult to do a direct comparison between the two time periods.
Although it's right to be cautious about the study results and their interpretation, that doesn't change what we already know about how to reduce the risk of dementia. Keeping active, having a healthy social life, and being a healthy weight are all good ways to protect the brain in later life.
Analysis by Bazian
Edited by NHS Website
Links to the headlines
The Times, 20 April 2016
The Guardian, 19 April 2016
Daily Mail, 20 April 2016
The Daily Telegraph, 19 April 2016
BBC News, 19 April 2016
The Sun, 20 April 2016
ITV News, 19 April 2016
Links to the science
Nature Communications. Published online April 19 2016