New estimates suggest that prevalence of dementia is falling in UK


A new study, which was led by Carol Brayne, professor, at the Cambridge Institute of Public Health (CIPH), University of Cambridge, UK, is part of the Medical Research Council-funded Cognitive Function and Ageing Studies (CFAS), a large, collaborative set of studies looking at different aspects of health and cognitive function in older people across two decades in different regions of the UK. The study was published in The Lancet.

Study participants were randomly chosen from people aged 65 and over living in Cambridgeshire, Newcastle, and Nottingham.  The first part of the study, CFAS I, took place between 1989 and 1994, with over 7000 people interviewed about lifestyle, health, socioeconomic factors, medication, social care, and other factors, with around 1,500 participants going on to be diagnostically assessed to determine whether they had dementia. The second part of the study, CFAS II, used identical methods to survey 7,796 randomly selected people aged 65 and over in the same geographical regions, between 2008 and 2011. 

The data collected in both parts of the study allowed researchers to estimate the prevalence of dementia in people aged 65 and over (ie: the percentage of the population with dementia) two decades ago, in 1991 (based on CFAS I data) and in 2011 (based on CFAS II data).

There have been marked changes in the older population over the last 20 years, with more people living to increasingly old ages at which they are expected to be at the highest risk of dementia. Applying the prevalence from 20 years ago to the current population, the authors estimated that 884,000 people would be expected to have dementia (8.3% of the population aged 65 and over). However, the prevalence of dementia identified in CFAS II is lower, with under seven in every hundred (6.5%) people aged 65 and over now estimated to have dementia, 670 000 people in total. This is a reduction of nearly a quarter (24.1%) in the numbers expected to have dementia.

Prevalence of dementia remains considerably higher in women, with 7.7% of women over 65 thought to have dementia, compared to 4.9% of men. Although the overall prevalence of dementia seems to have fallen, the prevalence of dementia among people living in care homes has increased, from about 56% of care home residents 20 years ago, to 70% today.

The analysis also takes social deprivation into account, and wide regional variation was observed at both time points. Areas with a large older population such as Barking and Dagenham, and Torbay are estimated to have the highest prevalence of dementia, whereas those with relatively young and more affluent profiles, such as Central Bedfordshire and Staffordshire, have the lowest prevalence.

“Changing diagnostic criteria mean that comparing newer estimates of dementia with older estimates is usually not possible, but CFAS is one of very few studies that was designed to allow us to track changing rates of dementia prevalence over time,” says study author Fiona Matthews, of the MRC Biostatics Unit, Cambridge Institute of Public Health (CIPH), University of Cambridge, UK.

Contemporary estimates of the prevalence of dementia are needed to plan for the provision of care, yet much of the evidence relied upon by government and health policy makers to allocate resources is now out of date. 

The study allows meaningful comparisons of prevalence over time, as well as differences in prevalence between regions of the UK, because it uses the same method of diagnosing dementia – the algorithmic diagnosis method.

“This study provides compelling evidence of a reduction in the prevalence of dementia in the older population over two decades,” says Brayne.  “Whether or not these gains for the current older population will be borne out in future generations would seem to depend on whether further improvements in primary prevention and effective health care for conditions which increase dementia risk can be achieved.”