Please can you outline your recent analysis that suggests the prevalence and incidence of dementia may be falling?
In this analysis we brought together the few European studies that have looked, within geographical areas, at the prevalence or incidence of dementia in the population aged 65 and over.
The studies had also been repeated with a range of gaps, up to 20 years afterwards. Those studies kept their methods as steady and stable as possible and were therefore able to compare the prevalence or incidence of dementia in contemporary populations over 65 with those of the same age in populations aged 65 and over in the past.
There were only a small number of studies; two in Sweden, one in the Netherlands, one in Spain and we have three sites in the UK, in England. Not all of them showed the same findings, but taken together, they provide an indication that dementia in populations might be changing.
Were you surprised by these findings given population ageing and the “dementia epidemic” reported in some recent studies?
I wasn't surprised because we published our data from England two years ago, which showed a more than 20% drop in the age-specific prevalence of dementia.
At that time, my peer researchers from around Europe had been beginning to publish or had already published their findings. I suggested to them that we work together and bring in any other studies that might meet the eligibility criteria and put them together, because they would be more powerful together than individually.
The other studies do not share the same design, so the paper goes into quite a lot of detail about the designs of the studies to try and provide an indication of the strengths and limitations of each one.
Two of the other studies suggest that dementia incidence (the occurrence of new dementia) has decreased and another suggests the prevalence has decreased.
The Spanish one shows a reduction in prevalence, a reduction that was driven by the male subjects. The Dutch study and one of the Swedish studies showed a reduction in incidence.
This wasn't a surprise because this was bringing together data that had already been published.
What are the main reasons thought to be contributing to the stabilizing of the number of people with dementia in some Western European countries?
These studies don't provide the answers to that question; it's speculation. There is a lot of other research evidence, which we partly used to justify the funding for the questions raised for the new generation studies.
The other evidence that we have arises from modeling of the risk factors for dementia, the modifiable ones such as those that contribute to vascular risk.
In the presence of neuropathology, there are some things that appear to compensate. For example, at a given level of neuropathology, a higher level of education is associated with a lower risk of dementia. The changes in the brain might be the same, but if you've got a higher level of education, then on a group level, you’re less likely to express dementia.
All these things have changed dramatically in society as a whole. The justification for raising money and being funded, was that we should be looking for changes because we would expect them if these factors really are risk factors for dementia. We should see changes, and indeed even changes in the nature of dementia, in our older population.
The explanation for reduced incidence and/or prevalence is that there may have been things affecting those particular cohorts at particular life stages, including their later life with dementia.
Why are accurate estimates of the proportion of dementia cases within countries so important?
You can see how interested the politicians, charities, industries and commercial developments involved in these societal concerns are. People on the ground are also very interested. Individuals are very interested, as are their families. I think people generally would like to know whether dementia is a fixed thing associated with fixed risks at different ages and only something that shifts if the neuropathology changes it, or whether it's a more fluid thing that is related to all sorts of things such as brain health across the life course.
I think lots of people are interested in it for very specific reasons;
- Governments because they have come to understand that dementia is important for aging populations;
- Charities because they need to know the numbers in order to pitch their messages and communicate with their members, as well as the external world;
- The UN because it wants to understand what might be changing across different populations at different times and how things might unfold in the future;
- Local authorities are interested in terms of adaptations of the environment and support for families and carers.
The results that we have presented in the paper are fairly crude and based simply on whether a person has dementia or not. There is much more nuance than that.
People with dementia are every bit as variable as people without dementia. There's a very grey area in terms of the transition into dementia in somebody with cognitive impairment, but who is functional and maybe has difficulties with day-to-day living.
There will be very different groups within those frail older people and robust older people. There are people with dementia who are incredibly physically fit, doing very well and are sustained in their community, whereas there are people with mild cognitive impairment who may have five or six medical conditions, be on lots of medications and who are very frail and need a lot of care.
There's every permutation that you can imagine in the older population. We need to measure that and try to help society work out how to best support people, what might be prevented and how we might reduce our risks for dementia.
What impact do you think your latest findings will have on planning for the provision of care?
The results within countries can immediately be used to map expectations regarding the amount of health and social care services that might be needed in any particular area.
In the UK, the age, sex and deprivation profiles can be used to estimate how many people with dementia there may be. It is not highly accurate and once you are dealing with the small numbers, it becomes much more unstable and I am absolutely not saying that screening is indicated.
However, it gives a ballpark for people who are providing services at the population level and enables them to think about services appropriate to need rather than services just being based on the fact that someone has been diagnosed through the system.
In terms of larger policy making, I think it means that we can use the data to model long-term care needs into the future and do more general population modeling.
If you bring in the risk estimations and other research that we and others have done, then you can start to say how much risk reduction we may have seen already and how much we may still be able to achieve if we look at the whole population health across the life course. We can start to look at what we can do at different life stages to implement that.
What policies would you suggest putting in place to reduce dementia risk moving forwards?
I'd use a whole population approach. I'd think about maternal health, about the environment in which children grow up and their education. I'd think about health interests and the way we educate children to understand their own risks and the things that they can do themselves.
I would want to look at the whole system; not only the health aspects across the life course, but how we can encourage localities and governments to implement strategies, whether it be laws, transport systems, or ways to address local communities that are sensitive to the risks.
For example, to reduce our risks of obesity, diabetes and lack of physical activity and provide people with a balanced understanding of the risks and benefits of various policies.
I'd love to see policies that encompass health impact, including brain health which would therefore include your risk of dementia in later life. You could look at every life stage as a raft of packages which are about general health, but usually the dominant feature is vascular health and risks such as diabetes.
A lot of the other risks such as alcohol, smoking and drugs also have a bearing on brain health. It would be good to look at those for each life stage and ask how we can structure our society to give people good health at that age but also good health at that age that would promote healthy aging and, therefore, healthy aging of the brain.
Where do you think research should now focus in order to aid dementia prevention?
I think it should be balanced. I think we invest a very large amount in drug discovery, which may bear fruit in the future for specific neuropathologies. It also has major implications in that if one or more of these medications turns out to be promising, then at what life stage are we going to ask people to take such a drug?
What are the potential downsides of that and how are we going to track it if people are willing?
People who are at very high risk might be willing to be the front-runners in this. There are lots of research programs which are predicated on that model so they would have regular imaging and we'd have to look at those images.
We already know that a significant portion of people who take some of the medications being used in these cutting-edge trials have imaging changes in their brain. We need to understand all of those things.
I would say that we need to look at the investments that we're making currently. A lot of this is really interesting and important work, particularly for people who are at a very high risk of the early onset form of dementia.
There are purer types of dementia, the pure Alzheimer's disease that begins earlier in life. We don't know what benefit such research will bring for the later onset form. That is by far the most common in the ninth decade and affects people who are in their eighties.
In the high-income countries, most people with dementia are aged 80 and older. In that age group, we know that you see a mixture of different types of pathology; vascular pathology and Alzheimer’s pathology. We just don't know how effective a single drug is going to be and until now, we've really only had symptomatic drugs.
I'd want to look at that and ask what we can understand about risks at different life stages and what things there are that we might be able to look at in terms of shifting risk patterns across the life course, which we understand might then lead to healthy brain aging.
It would be good to have a raft of research programs that have that sort of approach. Some of it would be individually-based and lot would be community-based. There might be legislation, for which we need to have a lot of evidence.
We could work out how things like the pricing of alcohol might affect dementia risk. Anecdotally, we are seeing an increase in women being admitted with serious alcohol problems in earlier life than would have been expected from the earlier patterns observed. We need to look at the population and say what are we doing with our risks and with our protective factors? Let's have a research program that looks at the future of all these people.
I'd like to see balanced investment. I completely understand why it is that the UK government is very interested in drug discovery because it's very important to the UK as a country. Our pharmaceutical companies are very important. I think that a balance and understanding of the whole context is what's important here and making investment across a range of areas rather than just certain areas.
Where can readers find more information?
The paper is published in The Lancet Neurology: http://www.thelancet.com/journals/lanneurol/article/PIIS1474-4422%2815%2900092-7/abstract
There’s also a paper on the two-decade comparison published in The Lancet by Matthews et al. It was published two years ago and is open access: http://www.thelancet.com/journals/lancet/article/PIIS0140-6736%2813%2961570-6/abstract
There is also a paper by Norton et al. that was published last year and is about the preventable fraction of dementia. That was a re-run of a previous paper but it was more accurate because it took risk factors and the correlation of risk factors into account. http://www.thelancet.com/journals/laneur/article/PIIS1474-4422%2814%2970136-X/abstract
About Professor Carol Brayne
Carol Brayne is a Professor of Public Health Medicine in the Department of Public Health and Primary Care at the University of Cambridge. She is a medically qualified epidemiologist and public health academic.
She graduated in medicine from the Royal Free Hospital School of Medicine, University of London and went on to train in general medicine. After gaining membership she moved on to training in epidemiology with a Training Fellowship with the Medical Research Council. The research area for this Fellowship was ageing and dementia.
Since the mid-eighties her main research area has been longitudinal studies of older people following changes over time in cognition, dementia natural history and associated features with a public health perspective.
She is lead principal investigator in the group of MRC CFA Studies which have informed and will continue to inform national policy and scientific understanding of dementia in whole populations.
She is Director of the Cambridge Institute of Public Health at the University of Cambridge and has played a lead role for teaching and training programs in epidemiology and public health for under and postgraduates at the University of Cambridge. She is a Fellow of the UK’s Academy of Medical Sciences.
Frailty may predict dementia risk years before onset, study finds