The Dementia Tsunami?

Summary of the Bristol Festival of Ideas/MRC ICU talk by Carol Brayne, Director of the Cambridge Institute for Public Health titled “Challenging the ‘Dementia Tsunami’ “. 

It’s pretty much common knowledge that life expectancy is increasing. This is in large part due to improvements in child health which means that more and more children survive into adulthood. But even for those who reached adulthood, living into dotage was an uncommon occurrence until relatively recently. Unfortunately, by fixing the health problems of the young we are finding that there are a myriad of health problems in the old - you win some, you lose some.

One of the most common problems associated with ageing is that of dementia. Dementia is not a specific disease but is a syndrome – a cluster of phenomena that combines a decline in mental function with a decline in the ability to function independently but without an associated decline in consciousness. It has multiple causes, the most common being Alzheimer’s disease and vascular dementia. Recently it has been recognised that mixed dementia, a combination of Alzheimer’s and vascular, is also common.

Diagnosing dementia is not an easy task. It is principally recognised by a decline in cognitive ability but this is a complex character to measure and consistent decline can be hard to recognise – after all, we all forget where we put our keys once in a while so how often is too often?

In response to the recognition of this continuum, a new diagnosis has been developed – that of mild cognitive impairment (MCI). Unfortunately, the continuum that led to its creation is also its downfall. As in all continuums, the extremes are easy to identify but drawing a clear line in the middle can be extremely difficult. Borderline cases of MCI and dementia are difficult to consistently and accurately diagnose.

The Medical Research Council’s Cognitive Function and Ageing Study (CFAS) investigated the problems with diagnosing MCI. This is what they found:

Summary of the pathways to different definitions of mild cognitive impairment currently in use (figure 1 from Matthews et al. (2007) PLoS Medicine, 4: e304).

It is not necessary to look at the details, just the larger picture. The most obvious aspect is that it is a very large picture! A single patient displaying a set of symptoms could, depending on the thoroughness of the doctor or their knowledge of particular diagnoses, be diagnosed with one of any of the eighteen diagnoses. Additionally, some diagnoses require only a single symptom while others require many more, so time-constrained clinicians may cling to the most straightforward diagnoses. Most damningly, the diagnoses do not appear to have any predictive ability. Telling a patient that they have a Mild Cognitive Disorder does not come with any information about how likely they are to develop dementia or what the timescale might be. The research concluded that there was a “lack of consistency [which] calls into question what exactly is being captured in each classification” (Matthews et al. 2007, p1618).

Despite these problems, MCI has received a lot of attention. In part this is due to the popularity of screening programmes designed to diagnose people at risk of diseases so that they may be more closely monitored and their treatment began earlier than might otherwise be the case. However, as described above, there is a major failing in screening people and offering the MCI diagnosis as it does not tell people how likely they are to develop dementia. Additionally, different diagnoses recognise different people, so depending on which diagnostic path a doctor is using, different patients could be diagnosed when compared with another doctor using another diagnostic path.

Another, more cynical, reason for the popularity of diagnosing MCI is that companies can offer treatments that promise to delay the onset of dementia in MCI patients. And as the diagnosis does not necessarily mean a person will get dementia, they can have high success rates without necessarily being effective.

For a patient who has been diagnosed with MCI or dementia, how they are treated will depend very much on the clinical path they follow. An illustration of this was provided which I have attempted to recreate here:

A patient suffering from dementia will go to their doctor where they will be diagnosed, or they will be sent to a specialist for diagnosis. Which type of specialist will depend on many factors, including the age and general health of the patient. This splitting of patients according to factors other than their dementia leads to ‘silos’ of specialists, focusing on their particular area of dementia in isolation from other researchers. For example, patients referred to the neurology service are typically atypical dementia patients – they are generally younger and otherwise healthy. Conversely, those in institutional care are typically typical dementia patients – old, infirm and often with additional health problems. As dementia is a neurological disease this leads to the unfortunate result that neurologists most often see atypical patients, presenting them with atypical models of the disease. Equally, drug companies looking to recruit patients for trials go to people who are otherwise healthy as they are more likely to be able to complete trials successfully. Again, this leads to atypical study groups. The effects of a particular drug may be promising on this subset of patients but there’s no particular reason why this should generalise to other groups of patients. This focus on young patients ultimately draws attention and resources away from the majority of sufferers who are older. And this is unfortunate because it has been found that the factor most associated with dementia is age.

CFAS has been running for several decades and has performed a massive amount of research. Beginning in the late 1980s it recruited over 18,000 participants in a national study to examine dementia in the population. It found that most people with dementia are over 85 years of age. The median age of onset was 84 years and the median survival time from the onset of dementia was 4.1 years for men and 4.6 years for women (Xie et al. 2008), meaning that the median age at death from dementia was 87 for men and 90 for women.

One of the most important things to note is that cognitive decline is normal. It is very rare for someone to reach old age without having experienced any cognitive decline. The problem, if it can be described as such, is that while cognitive scores are generally pretty uniform across younger populations, older populations have much more variation.

CAMCOG (a score of cognitive ability) across age groups from a population sample.

(Figure 3 from Brayne. 2007. Nature Reviews Neuroscience 8: 233-239.)

So if cognitive decline is normal does that mean that dementia is normal too? Yes, in that by the age of 90 it is ‘normal’ to have severe cognitive impairment. But it is extremely difficult to lead a ‘normal’ life when you can no longer remember where you live or that your husband is dead and your children are now grown. So just because it is normal does not mean that it is not debilitating and requires help to counter the worst of the effects.

The recognition that cognitive decline is normal has come at the same time as one definitive test for dementia has been recognised as deeply flawed. One of most diagnostic physical manifestations of dementia has been plaques and tangles on the brain.

Plaques and tangles build up on neuron endings in patients with Alzheimer's Disease.
Credit http://ksj.mit.edu/tracker/2010/08/nytimes-default-hypothesis-alzheimers-ma
Source: http://bit.ly/1KjaBhp

Autopsies of patients with dementia, particularly Alzheimer’s disease, found that neurons in their brains had a lot of abnormal proteins called plaques and tangles. It was thought that these were causing cell death and tissue loss which caused the cognitive decline. However, a fantastic study by CFAS has found that the presence of plaques and tangles is not the diagnostic certainty it once was. The study examined 456 brains donated for research. They were examined for the typical signs dementia by people who did not know anything about the person so that there was no risk of bias. Their results were surprising.

Prevalence of moderate or severe pathological lesions in people with dementia (yellow) and without (blue) by age (figure 1 from Savva et al. (2009) New England Journal of Medicine, 360: 2302-2309).

Their results showed that these neurological signs of dementia were not great predictors of dementia. In young people plaques and tangles were good indicators of dementia but in older people as many people without dementia had plaques and tangles as those with dementia. In other words, plaques and tangles may be necessary but are not sufficient to cause dementia in those age groups most prone.

So if the plaques and tangles aren’t causing dementia then what is? Research by CFAS and others has found two potential influences. The first is education. Those with higher levels of education are found to have a lower risk of dementia even though they have the same level of plaques and tangles in their brains (Brayne et al. 2010). This suggests that education is not a protective measure but allows for compensation so that the effects of the plaques and tangles are minimised.

The second cause is related to vascular dementia. A follow-up to CFAS, CFAS II, took place in the 2000s. The follow-up study followed exactly the same protocols as the original study, enabling direct comparison of data. It found that dementia rates had fallen in almost all age groups.

Comparison of dementia rates for people in 1989-1994 (CFAS I - blue) and 2008-2011 (CFAS II - orange) (figure 1 from Matthews et al. 2013. The Lancet, 382: 1405-1412)

Other studies across Europe have found similar results and it is thought that the fall in the prevalence of strokes is to ‘blame’. What is not clear is whether this fall is a one-off or is the beginning of a longer-term trend. Only time will tell.

It seems that dementia is an unfortunate but normal part of ageing. While it is possible to minimise the effects of dementia this seems to be achievable through a lifetime of mental stimulation and healthy living. In other words those ‘brain training’ games may be fun but they aren’t the preventative they claim to be.

What is most important from a societal point of view is that we recognise that people suffering from dementia are not ‘ill’ in the normal sense of the word. They are merely experiencing the effects of normal cognitive decline more severely than others. But everyone will experience cognitive decline and it is important that we recognise this and create services that allow people to live rewarding lives in spite of this. This is starting to happen in places like the Netherlands. These services are expensive but I believe that as more and more people see their grandparents and parents experience age-related cognitive decline we will feel the expense is well worth it.

After all, we’re all going to get old.

References 

Brayne (2007) The elephant in the room—healthy brains in later life, epidemiology and public health. Nature Reviews Neuroscience, 8: 233-239.

Brayne et al. (2010) Education, the brain and dementia: neuroprotection or compensation? Brain, 133: 2210-2216.

Matthews et al. (2007) Operationalisation of mild cognitive impairment:
A graphical approach. PLoS Medicine, 4: e304.

Matthews et al. (2013) A two-decade comparison of prevalence of dementia in individuals aged 65 years and older from three geographical areas of England: results of the Cognitive Function and Ageing Study I and II. The Lancet, 382: 1405-1412

Savva et al. (2009) Age, Neuropathology, and Dementia. New England Journal of Medicine, 360: 2302-2309.

Xie et al. (2008) Survival times in people with dementia: analysis from population based cohort study with 14 year follow-up. British Medical Journal, 336: 258-262.