Presented at the Dementias 2017 conference which was held in the Royal College of General Practitioners in London on 9th and 10th February 2017.

1. Prevention of Dementia and the
Lancet Commission
Gill Livingston
February 2017

2. Email :An invitation from The Lancet
I am one of the editors at The Lancet and I am writing to
invite you to participate in an exciting project that we are
planning…
We would like to invite you to be the Chair of the Commission
and take the lead on this project.
The content is entirely up to you but our initial thoughts are…

3. Our sponsors
• UCL
• Alzheimer's Society
• ESRC
• ARUK
And…..
The Lancet editors
Helen Frankish and Sabine Best

4. The commissioners
Gill Livingston Andrew Sommerlad Vasiliki Orgeta Sergi Costafreda Jonathan
Huntley David Ames Clive Ballard Sube Banerjee Alistair Burns Jiska Cohen-
Mansfield Claudia Cooper Nick Fox Laura Gitlin Robert Howard Helen Kales
Eric Larson Karen Ritchie Kenneth Rockwood Elizabeth Sampson Quincy
Samus Lon Schneider Geir Selbӕk Linda Teri, Naaheed Mukadam
https://www.ucl.ac.uk/psychiatry/research/olderpeople/lancet-international-commission

6. And then from the Lancet
A new systematic review or model would be great
and so we became
The Lancet international Commission on Dementia
Prevention and Care

7. Why do some people not develop
dementia?
• Dementia is by no means an inevitable consequence of
reaching retirement age, or even of entering the ninth decade.
• There are lifestyle factors that may reduce, or increase, an
individual’s risk of developing dementia.
• In some populations dementia is already being delayed for
years; while in others the numbers of people living with it has
increased

8. Prevention is better than cure
Two ways forward
• Make people more resilient
• Prevent damage

9. Population Attributable Fraction for
modifiable risk factors
The percentage reduction in new cases
over a given time if a particular risk
factor were completely eliminated.

10. Ritchie et al BMJ 2010

11. Barnes and Yaffe
Lancet Neurology
2011

12. Lamplight research?
65+ cohorts = 65+ risk factors
65+ biobanks = 65+ biomarkers

13. Lancet analysis
Lifecourse analysis- when should intervene
Taking into account new risk factors with evidence
Split into
Early life
Midlife aged 45-65 years
Late life as aged > 65.

14. Risk factors from
UK National Institute of Health and Care Excellence (NICE)
US National Institute of Health (NIH)
Used systematic reviews and meta-analyses and when
there was not one we calculated.
PAF depends on relative risk and prevalence

15. BUT
People have lots of risks factors together
So we had to adjust for communality

16. Formula for individual Population Attributable Fraction (PAF)
PAF = Pe (RRe-1) / [1 + Pe (RRe-1)]
Pe = prevalence of the exposure
RRe = relative risk of disease due to that exposure
Calculation of communality
Input data on all nine risk factors in our model - Calculate tetrachoric correlation to generate correlation coefficients and a correlation matrix
Conduct a principal-component analysis on the correlation matrix to generate eigenvectors, which are directions mapped onto the data
points and from which variance to the data is measured. These represent unobserved factors underlying all the variables that explain the
variance observed.
Components with eigenvalues ≥1 were retained in the model
Communality was calculated as the sum of the square of all factor loadings (i.e. how much each unobserved component explained each
measured variable).
Calculation of overall Population Attributable Fraction (PAF)
We then calculated overall PAF: PAF = 1-[(1-PAF1)(1-PAF2)(1-PAF3)…]
Each individual risk factor’s PAF was weighted according to its communality using the formula:
Weight (w) = 1-communality
Weighting was included in the calculation of overall PAF using the formula:
PAF = 1-[(1-w*PAF1)(1-w*PAF2)(1-w*PAF3)...]

17. So we calculated life course
potentially modifiable risk factors
through the life coursr
Sorry can’t give you new meta-analysis and life course model today
Thanks
Naaheed Mukadam
Andrew Sommerlad,
Sergi Costafreda

18. Just because it flies it doesn’t make it a
bird

19. Limitations
• The PAF model assumes a causal association between a risk
factor and dementia,
• The most convincing evidence of causality would be randomised
controlled trials in humans.
• This is not possible for many proposed dementia risk factors such
as education; but we know that falling age-specific incidence is
associated with more education
• Without experimental human evidence, causality criteria are:
strength, temporality, plausibility, biological gradient, consistency

20. Antihypertensives
RCT in non-demented but hypertensive aged >80 (160-200/<110mmHg)
stopped as CVAs in TAU
Underpowered (as stopped) but dementia risk= HR 0.87, 95% CI [0.76-
1.00]
• Peters et al 2008
• Meta-analysis of antihypertensive treatment groups (weighted mean
difference = 0.42; 95% CI [0.30-0.53])
• Cochrane 2009
• Pre- Diva trial found significant difference only in those with hypertension
• Moll van Charante 2016

21. Exercise
• Longitudinal studies show a strong relationship between taking exercise
and not developing dementia: meta-analysis hazard ratio 0.62 (95% CI
0.54-0.70). Dose dependent protection
• Sofi et al 2011
• Postulated to have a neuroprotective effect, potentially through
promoting release of Brain Derived Neurotrophic Factor, reducing
cortisol and reducing vascular risk
• One RCT of 40 minutes walking three times weekly for a year (versus
stretching and conditioning) showed exercise training increased
hippocampal size and improved memory in healthy adults aged 55-80
• Conflicting RCT findings about exercise- ? Too short or not adaptive or
other differences

22. BUT trials of
• non-steroidal anti-inflammatory drugs (NSAID)
• a 24 week RCT of an oral hypoglycaemic drug,
rosiglitazone
• oestrogen hormone replacement therapy, statins
• vitamins
• statins
• and ginkgo biloba extract
Have all been negative

23. Studies in
Low risk volunteers (PRE-DIVA) have
been negative or weakly positive
(FINGER)

24. Dementia intervention: what,
when, for how long and for whom?
• Not feasible to completely eliminate risk factors
• Future strategies either to target the whole
population over a long period
• Or those at higher risk.

25. Purpose of Lancet commission
• “Philosophers have only
interpreted the world in
various ways; the point,
however, is to change it.”
Karl Marx
Theses on Feuerbach
11th thesis
(And Sube Banerjee)

26. Be ambitious about prevention
• Thus while trials, which by their nature are relatively short and
include a smaller number of people, are disappointing,
• Results from risk factor modification for whole populations or
high risk populations have been more hopeful.
• Delaying dementia for some years for even a small
percentage of people would be an enormous achievement
• It looks like it may be within our reach

27. Thanks
To all commissioners
To sponsors
To those whose slides I have used
July 2017 Launch of commission