Estimating the number of people who have dementia is important for both local planning and national guidance.
On the face of it, this should be a straightforward metric but epidemiology is an intricate science.
Most current estimates of dementia prevalence (the number of people affected by the disorder) and incidence (the number of people developing it over a defined period, usually one year) are based on studies dating back to the 1980s.
The figures are sometimes modified by a Delphi approach whereby expert opinion fine-tunes the results of studies for local information. Current estimates of the national totals are based on the Alzheimer’s Society figures from their own report which are broadly in line with those of Alzheimer Research UK.
The headline figures rarely subtype dementia into the causes – Alzheimer’s disease, vascular dementia.
A recent report in the in the Medical Journal, the Lancet on the Cognitive Functioning and Ageing Study (CFAS) gives a very important update on the numbers of people with dementia in England.
The study is uniquely placed to inform the discussion about the prevalence of dementia.
It was designed specifically to look at changes over time and has recorded the difference in prevalence between 1991 (CFAS 1) and 2011 (CFAS 2) by using the same methods.
Diagnostic methods for dementia and research approaches have undergone a subtle change over that time and the consistency of the CFAS design is essential to allow such a direct comparison.
The study deals specifically with dementia in England, and looked at three areas – Cambridgeshire, Nottingham and Newcastle.
In 1991 there were an estimated 664,000 people with dementia. Increasing that in-line with the increase in the older population suggested that in 2011 the figure would rise to around 884,000. But the current figures actually suggest the number is 670,000 – 214,000 fewer than expected.
The number of people with dementia in care homes increased from 56 per cent to 70 per cent and the challenges of carrying out longitudinal research in this population is suggested by the response rate declining from 80 per cent in 1991 to 56 per cent in 2011.
The authors should be congratulated on such an important study which has significant implications for dementia.
There are three things which are important to take away:
First, there should not be a knee jerk reaction to change the prevalence data on which current rates of dementia are predicated. The study is an incredibly important one but the authors themselves describe some of the limitations – perceptions and attitudes to research, the accuracy of population registers on which some of the data are based, and the higher rate of people refusing to take part in the study (people who decline to research may be more likely to have dementia). The authors carried out some sophisticated modelling as to the effects of these limitations but they still did not suggest that the prevalence of dementia had anything other than decreased. NHS England will be working with colleagues to develop support for CCGs for the diagnosis of dementia which will include a consideration of all the current evidence of the prevalence of dementia. This will involve gathering experts from a variety of backgrounds to look at the new information and to make a balanced judgement as to what changes, if any, should be made to the prevalence figures. This will take weeks rather than days or months.
Second, although the overall prevalence of dementia may well have decreased, the variability seen between regions in the UK is unaffected by the absolute numbers.
Third, the results do suggest that the numbers of people with dementia can be influenced and raises the real possibility of the reality of prevention. Changes in the habits and the illness profile of the population could have either an upward or downward effect on the numbers of people with dementia. For example, better survival after a stroke may increase the number of people at risk of dementia but primary prevention (that is, reducing the numbers of people having strokes) may reduce the number. Similarly, better management of diabetes could lead to reduced vascular dementia. Improved educational levels may also protect against the development of dementia.
All in all, this is great news for dementia raising the possibility that prevention can become a reality while not diminishing the importance of the disorder in terms of individual impact and the benefits of empowering people through awareness and diagnosis.
Acknowledgement: Thanks to Piers Kotting of DeNDRoN for the concept of turning the tide.