Life course aetiology of dementia and cognitive decline: improving causal inference
Lead Research Organisation:
University of Bristol
Department Name: Social Medicine
Abstract
Dementia describes a set of symptoms including memory loss and difficulties with thinking, problem-solving or language. The term 'dementia' encompasses several different types of disease, with Alzheimer's being the most common. Dementia is now the second leading cause of death in England and Wales, thus, finding ways to prevent it is a public health priority. Dementia is not only a great burden on our economy and health care system, but also on the families and friends of those suffering with the disease. Furthermore, treatments that are currently available are unable to delay the onset of, or cure, dementia. Therefore, it is essential that we identify factors that increase a person's risk of dementia, especially things that we are able to change (such as smoking or diet), so that we can intervene and prevent it occurring in the first place.
Findings from existing studies that have tried to identify risk factors for dementia are conflicting. The only risk factors for which we have good, consistent evidence so far are older age, being female, low education, head trauma and some cardiovascular risk factors such as diabetes. Evidence for other risk factors such as smoking, alcohol consumption and physical activity, is less clear, in that they have reported positive (i.e. increasing risk), negative (i.e. decreasing risk) and no effects on dementia risk. One key problem is that there are many complications when studying causes of dementia. It is a complex disease that has many possible causes and can begin up to 20 years before people start showing noticeable symptoms. This makes it difficult to assess whether the risk factors being studied are a cause or a consequence of dementia. For example, it is still unclear whether depression in adulthood is a risk factor for dementia, or whether the early brain changes that occur in dementia result in depressive symptoms. Moreover, people with dementia who are still actively participating in studies are often more 'healthy' than those who leave studies due to illness or death, which can bias study results (i.e. give us the wrong answer).
My project aims to characterise if and how people at increased genetic risk of dementia differ in terms of their cognitive capability across the whole life course. For example, do people with increased genetic risk of dementia achieve lower grades at school and achieve a lower cognitive peak in early life? Do they start to decline cognitively at an earlier age, or at a faster rate than people who are not at increased genetic risk of dementia? My project also aims to identify causal risk factors for dementia by employing state of the art analytical approaches and by studying many different groups of people (cohorts). Consistent research findings across multiple cohorts of people will not only provide confidence in my findings and yield important insights into true, causal risk factors for dementia, but it will also help to inform dementia prevention strategies.
In summary, the proposed research will help to bring clarity to some of the conflicting literature on dementia risk factors, which in turn will improve translation into public health policies for interventions, with the overriding aim of preventing dementia.
Findings from existing studies that have tried to identify risk factors for dementia are conflicting. The only risk factors for which we have good, consistent evidence so far are older age, being female, low education, head trauma and some cardiovascular risk factors such as diabetes. Evidence for other risk factors such as smoking, alcohol consumption and physical activity, is less clear, in that they have reported positive (i.e. increasing risk), negative (i.e. decreasing risk) and no effects on dementia risk. One key problem is that there are many complications when studying causes of dementia. It is a complex disease that has many possible causes and can begin up to 20 years before people start showing noticeable symptoms. This makes it difficult to assess whether the risk factors being studied are a cause or a consequence of dementia. For example, it is still unclear whether depression in adulthood is a risk factor for dementia, or whether the early brain changes that occur in dementia result in depressive symptoms. Moreover, people with dementia who are still actively participating in studies are often more 'healthy' than those who leave studies due to illness or death, which can bias study results (i.e. give us the wrong answer).
My project aims to characterise if and how people at increased genetic risk of dementia differ in terms of their cognitive capability across the whole life course. For example, do people with increased genetic risk of dementia achieve lower grades at school and achieve a lower cognitive peak in early life? Do they start to decline cognitively at an earlier age, or at a faster rate than people who are not at increased genetic risk of dementia? My project also aims to identify causal risk factors for dementia by employing state of the art analytical approaches and by studying many different groups of people (cohorts). Consistent research findings across multiple cohorts of people will not only provide confidence in my findings and yield important insights into true, causal risk factors for dementia, but it will also help to inform dementia prevention strategies.
In summary, the proposed research will help to bring clarity to some of the conflicting literature on dementia risk factors, which in turn will improve translation into public health policies for interventions, with the overriding aim of preventing dementia.
Technical Summary
Existing evidence for modifiable risk factors for dementia is inconsistent. This is perhaps unsurprising given the , such as the potential for reverse causation and selection bias. The overall aim of the proposed work is to improve understanding of the causal determinants of, and mechanistic pathways to, dementia and cognitive decline, using novel analytical approaches and cross-cohort comparisons. I will improve causal inference in Mendelian randomisation studies of dementia by using simulation analyses to investigate and apply methods for dealing with selection bias. I will assess whether people at increased genetic risk of Alzheimer's (based on a polygenic risk score) have different life course trajectories of cognitive capability than people not at increased genetic risk. I will use novel statistical methods (such as Mendelian randomisation, offspring instrumental variable analysis and bivariate multilevel modelling) to identify causal risk factors for dementia and cognitive decline across the life course and use novel mediation methods (such as network Mendelian randomisation, counterfactual, decomposition and G computation) to assess possible mediation by circulating metabolites. The triangulation of research findings across different analytical methods and across cohorts in different settings is novel and will increase confidence in my findings. I will use data from cohorts within Dementias Platform UK, the Avon Longitudinal Study of Parents And Children (UK), EPIPorto (Portugal) and The HUNT Study. This work has the potential to inform preventive strategies, which is important given the rising economic burden of the ageing population and because current treatments are unable to delay the onset of, or cure, dementia.
Planned Impact
Academia
The primary benefit to academia will be in the provision of new, in-depth understanding of modifiable life course risk factors for dementia and cognitive decline. The outputs of my research will have a wide and diverse academic appeal, since they may be used to generate new hypotheses about the identified risk factors and to develop public health interventions. I will develop trajectories of repeated measures of risk factors within multiple cohorts and will make these available to other researchers wishing to examine them in relation to other exposures and outcomes. I will develop an advanced epidemiological skill set by learning novel analytical approaches and working with experts in both epidemiology and dementia. I will directly reinvest these new skills back into the UK and USA research base through dissemination in methodological journals, conferences, teaching, seminars and talks.
Health policy makers
Authorities involved in setting clinical guidance will benefit from this research by receiving new knowledge about modifiable causes of dementia and cognitive decline. Given that my research will investigate modifiable causes of dementia across the life course, it has the potential to shed light on early life prevention opportunities, which may be more effective than prevention efforts later in life.
Health service planners and health economists
This research could be included in statistical models to predict future burden of disease due to the risk factors studied in this fellowship and also be used to predict the future likely benefit of prevention by intervening on each of the risk factors identified.
Charities such as Alzheimer's Association UK
One of the key priorities of Alzheimer's Association UK is the promotion of brain health. Thus, through generating evidence for risk factors for dementia, my work is directly relevant to informing the research agendas and prevention work of these charities.
The wider public
Improvements in policy, services or guidance as a result of this research will have a direct impact on improving the cognitive health of the general public. Improving the effectiveness of public services and policy may also result in reduced direct and indirect healthcare costs and lower economic impacts of dementia, which will benefit society as a whole. This research will also help to reduce the burden that ageing populations place on health and care provision services.
Benefits of this research could be both within the lifetime of the fellowship, e.g. providing new aetiological insights, and beyond the lifetime of the project, e.g. changes to guidance and policy for prevention and the potential health care savings following this.
Details of the actions I will take to achieve these benefits are detailed in the 'Pathways to Impact' attachment.
The primary benefit to academia will be in the provision of new, in-depth understanding of modifiable life course risk factors for dementia and cognitive decline. The outputs of my research will have a wide and diverse academic appeal, since they may be used to generate new hypotheses about the identified risk factors and to develop public health interventions. I will develop trajectories of repeated measures of risk factors within multiple cohorts and will make these available to other researchers wishing to examine them in relation to other exposures and outcomes. I will develop an advanced epidemiological skill set by learning novel analytical approaches and working with experts in both epidemiology and dementia. I will directly reinvest these new skills back into the UK and USA research base through dissemination in methodological journals, conferences, teaching, seminars and talks.
Health policy makers
Authorities involved in setting clinical guidance will benefit from this research by receiving new knowledge about modifiable causes of dementia and cognitive decline. Given that my research will investigate modifiable causes of dementia across the life course, it has the potential to shed light on early life prevention opportunities, which may be more effective than prevention efforts later in life.
Health service planners and health economists
This research could be included in statistical models to predict future burden of disease due to the risk factors studied in this fellowship and also be used to predict the future likely benefit of prevention by intervening on each of the risk factors identified.
Charities such as Alzheimer's Association UK
One of the key priorities of Alzheimer's Association UK is the promotion of brain health. Thus, through generating evidence for risk factors for dementia, my work is directly relevant to informing the research agendas and prevention work of these charities.
The wider public
Improvements in policy, services or guidance as a result of this research will have a direct impact on improving the cognitive health of the general public. Improving the effectiveness of public services and policy may also result in reduced direct and indirect healthcare costs and lower economic impacts of dementia, which will benefit society as a whole. This research will also help to reduce the burden that ageing populations place on health and care provision services.
Benefits of this research could be both within the lifetime of the fellowship, e.g. providing new aetiological insights, and beyond the lifetime of the project, e.g. changes to guidance and policy for prevention and the potential health care savings following this.
Details of the actions I will take to achieve these benefits are detailed in the 'Pathways to Impact' attachment.
Organisations
- University of Bristol (Lead Research Organisation)
- UNIVERSITY OF OXFORD (Collaboration)
- HARVARD UNIVERSITY (Collaboration)
- University College London (Collaboration)
- University of Bordeaux (Collaboration)
- Norwegian University of Science and Technology (NTNU) (Collaboration)
- UNIVERSITY OF EXETER (Collaboration)
- University College London (Fellow)
People |
ORCID iD |
Emma Anderson (Principal Investigator / Fellow) |
Publications
Anderson E
(2018)
Locus of control as a modifiable risk factor for cognitive function in midlife.
in Aging
Anderson EL
(2023)
Drug target Mendelian randomisation: are we really instrumenting drug use?
in Diabetologia
Anderson EL
(2017)
Adversity in childhood and measures of aging in midlife: Findings from a cohort of british women.
in Psychology and aging
Anderson EL
(2017)
Prospective associations of psychosocial adversity in childhood with risk factors for cardiovascular disease in adulthood: the MRC National Survey of Health and Development.
in International journal for equity in health
Anderson EL
(2018)
Associations of adversity in childhood and risk factors for cardiovascular disease in mid-adulthood.
in Child abuse & neglect
Anderson EL
(2022)
Little genomic support for Cyclophilin A-matrix metalloproteinase-9 pathway as a therapeutic target for cognitive impairment in APOE4 carriers.
in Scientific reports
Anderson EL
(2020)
Education, intelligence and Alzheimer's disease: evidence from a multivariable two-sample Mendelian randomization study.
in International journal of epidemiology
Anderson EL
(2021)
Is disrupted sleep a risk factor for Alzheimer's disease? Evidence from a two-sample Mendelian randomization analysis.
in International journal of epidemiology
Caleyachetty R
(2018)
Modeling Exposure to Multiple Childhood Social Risk Factors and Physical Capability and Common Affective Symptoms in Later Life.
in Journal of aging and health
Caleyachetty R
(2021)
Exposure to multiple childhood social risk factors and adult body mass index trajectories from ages 20 to 64 years.
in European journal of public health
Crick DCP
(2022)
Associations between Adverse Childhood Experiences and the novel inflammatory marker glycoprotein acetyls in two generations of the Avon Longitudinal Study of Parents and Children birth cohort.
in Brain, behavior, and immunity
Dardani C
(2022)
Is genetic liability to ADHD and ASD causally linked to educational attainment?
in International journal of epidemiology
Davies NM
(2021)
BMI is unlikely to be a plausible intervention target for reducing the incidence of dementia.
in International journal of epidemiology
Guyatt AL
(2017)
Early life adiposity and telomere length across the life course: a systematic review and meta-analysis.
in Wellcome open research
Hartley A
(2018)
Metabolic profiling of adolescent non-alcoholic fatty liver disease.
in Wellcome open research
Hartley A
(2018)
Metabolic profiling of adolescent non-alcoholic fatty liver disease
in Wellcome Open Research
Houtepen LC
(2018)
Childhood adversity and DNA methylation in two population-based cohorts.
in Translational psychiatry
Korologou-Linden R
(2019)
The causes and consequences of Alzheimer's disease: phenome-wide evidence from Mendelian randomization
Korologou-Linden R
(2022)
The causes and consequences of Alzheimer's disease: phenome-wide evidence from Mendelian randomization.
in Nature communications
Korologou-Linden R
(2019)
Polygenic risk score for Alzheimer's disease and trajectories of cardiometabolic risk factors in children
in Wellcome Open Research
Korologou-Linden R
(2019)
Polygenic risk scores for Alzheimer's disease, and academic achievement, cognitive and behavioural measures in children from the general population.
in International journal of epidemiology
Lawn RB
(2018)
Psychosocial adversity and socioeconomic position during childhood and epigenetic age: analysis of two prospective cohort studies.
in Human molecular genetics
Magnus MC
(2018)
Childhood psychosocial adversity and female reproductive timing: a cohort study of the ALSPAC mothers.
in Journal of epidemiology and community health
Meddens SFW
(2021)
Genomic analysis of diet composition finds novel loci and associations with health and lifestyle.
in Molecular psychiatry
O'Keeffe LM
(2018)
Associations of Y chromosomal haplogroups with cardiometabolic risk factors and subclinical vascular measures in males during childhood and adolescence.
in Atherosclerosis
O'Keeffe LM
(2018)
Sex-specific trajectories of measures of cardiovascular health during childhood and adolescence: A prospective cohort study.
in Atherosclerosis
O'Keeffe LM
(2019)
Data on trajectories of measures of cardiovascular health in the Avon Longitudinal Study of Parents and Children (ALSPAC).
in Data in brief
Pagoni P
(2022)
Causal effects of circulating cytokine concentrations on risk of Alzheimer's disease and cognitive function.
in Brain, behavior, and immunity
Pagoni P
(2022)
Exploring the causal effects of genetic liability to ADHD and Autism on Alzheimer's disease.
in Translational psychiatry
Prince C
(2019)
Investigating the impact of cigarette smoking behaviours on DNA methylation patterns in adolescence.
in Human molecular genetics
Richmond RC
(2019)
Investigating causal relations between sleep traits and risk of breast cancer in women: mendelian randomisation study.
in BMJ (Clinical research ed.)
Rouanet A
(2022)
How Selection Over Time Contributes to the Inconsistency of the Association Between Sex/Gender and Cognitive Decline Across Cognitive Aging Cohorts.
in American journal of epidemiology
Walkden GJ
(2018)
Frailty in older-age European migrants: Cross-sectional and longitudinal analyses of the Survey of Health, Aging and Retirement in Europe (SHARE).
in Social science & medicine (1982)
Description | Presentation on Mendelian Randomization at University of Bordeaux |
Geographic Reach | Local/Municipal/Regional |
Policy Influence Type | Influenced training of practitioners or researchers |
Description | Presentation on Mendelian Randomization to University of Oxford |
Geographic Reach | National |
Policy Influence Type | Influenced training of practitioners or researchers |
Description | Identifying causal risk factors for cognitive decline |
Amount | £215,115 (GBP) |
Funding ID | TAKEDA 2020-2609 |
Organisation | Takeda Pharmaceutical Company |
Sector | Private |
Country | Japan |
Start | 11/2019 |
End | 11/2022 |
Description | Harvard University |
Organisation | Harvard University |
Country | United States |
Sector | Academic/University |
PI Contribution | Co-authorship of collaborators on my published paper |
Collaborator Contribution | Provided access to data, provided critical comments on the manuscript |
Impact | https://academic.oup.com/ije/article/50/3/817/5956327 The collaboration was multidisciplinary involving bioinformaticians, geneticists, epidemiologists, sleep scientists and statisticians. |
Start Year | 2018 |
Description | Norwegian University of Science and Technology (NTNU) |
Organisation | Norwegian University of Science and Technology (NTNU) |
Country | Norway |
Sector | Academic/University |
PI Contribution | I have worked closely with Ben Brumpton and Bjorn Olav Asvold from The HUNT Study based at NTNU. I used their cohort to validate some MR results in my bidirectional sleep-AD MR paper. I am now planning a visit to their department to complete some analyses for my offspring IV analysis of CVD risk factors on Dementia risk. |
Collaborator Contribution | They contributed data, advice on analyses, and comments on the resulting paper |
Impact | The paper resulting from my collaboration with NTNU is currently in press. I will add this once it has been published. |
Start Year | 2017 |
Description | University College London |
Organisation | University College London |
Country | United Kingdom |
Sector | Academic/University |
PI Contribution | I have formed collaborations with Mika Kivimaki from University College London as I have applied to complete a project examining associations between a polygenic risk score for Alzheimer's disease and cardiovascular risk factor trajectories. My application has been accepted and I am awaiting data to be uploaded onto the analysis portal. |
Collaborator Contribution | Mika has given me feedback on my project proposal and supervised my application for the data. He will also supervise the project and be a co-author on the paper |
Impact | The analyses have not yet been conducted but we have an analysis plan and an outline for a paper |
Start Year | 2020 |
Description | University of Bordeaux |
Organisation | University of Bordeaux |
Country | France |
Sector | Academic/University |
PI Contribution | I spent one month working Carole Dufouil, Cecile Proust-Lima an Stephanie Debette at INSERM at the University of Bordeaux. I helped postdocs with their Mendelian Randomization analyses (coding and interpretation) and also was invited to do a seminar on work I have carried out as part of my fellowship, looking at the causal effect of educational attainment on risk of Alzheimer's Disease. |
Collaborator Contribution | I learned about the Three City Dijon study and MEMENTO, which are housed at the university. I made applications to use the data and was able to get a condensed and anonymised version to do some preliminary analyses on. |
Impact | Invited seminar |
Start Year | 2017 |
Description | University of Exeter |
Organisation | University of Exeter |
Department | Medical School |
Country | United Kingdom |
Sector | Academic/University |
PI Contribution | I have assisted some of the group in their Mendelian Randomization analyses |
Collaborator Contribution | The group have spent the last year running genome-wide association studies on various sleep traits. I have collaborated with them to run bidirectional MR-analyses on various sleep traits on risk of Alzheimer's disease. I have also met with the dean and vice-dean of research to disucss potential collaborations on their PROTECT cohort (http://www.protectstudy.org.uk/) for some of my fellowship projects |
Impact | Currently writing up paper for publication |
Start Year | 2017 |
Description | University of Oxford |
Organisation | University of Oxford |
Department | Department of Psychiatry |
Country | United Kingdom |
Sector | Academic/University |
PI Contribution | I was invited to run a seminar on my fellowship work so far - disentangling the causal effects of education and IQ on risk of Alzheimer's disease.I also assisted some of the postdocs with their Mendelian randomization work, and gave advice on general epidemiology (as the department currently has no epidemiologists!) |
Collaborator Contribution | I met with Sarah Bauermeister who took me through the new DPUK portal, which I plan to use in the near future. I also completed the DPUK online training lecture. I worked with John Gallacher and some of his postdocs to come up with some exciting new project ideas using DPUK data. |
Impact | Invited seminar, invited to DPUK analysts conference |
Start Year | 2017 |
Description | Cheltenham Science Festival |
Form Of Engagement Activity | Participation in an activity, workshop or similar |
Part Of Official Scheme? | No |
Geographic Reach | Regional |
Primary Audience | Public/other audiences |
Results and Impact | I volunteer for the Alzheimer's Research UK stand at the Discover Zone of the Cheltenham Science Festival. I helped to run the following activities: A Walk Through Dementia- This is a virtual reality experience of dementia, that allows viewers insight into dementia. It is a great way to start conversations around the condition. We had four VR headsets on our stand and it was very popular. Sea Hero Quest-This is a game available on iPads and mobile devices. The navigation based game collects gameplay data which researchers are using to create a baseline of human spatial navigation and how it changes over ageing. Dementia explained - explore the brain games. Children's website (https://kids.alzheimersresearchuk.org/juniors/explore-the-brain/) has games that show how different parts of the brain work. We had two iPads that people could play these games on. Test your knowledge quiz - We had a blackboard with a short multiple choice quiz to test people's knowledge about the brain and dementia. Origami brain fact finders - This was a simple craft activity, and people couldtake these away to show off their knowledge. Brain hats - A simple craft activity to show people what different parts of the brain do, and giving them a hat to take away. Make a neuron - Using pipe cleaners, visitors learned the different parts of a neuron, what they are for and how they all link up together, and they could take their neuron away with them. |
Year(s) Of Engagement Activity | 2018 |
URL | https://www.cheltenhamfestivals.com/science/about/what-can-i-do-for-free/discover-zone/ |
Description | Invited Seminar at University of Oxford |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | Local |
Primary Audience | Professional Practitioners |
Results and Impact | I was invited to present a seminar on my fellowship work which uses multivariable Mendelian Randomization to disentangle the relative causal effects of educational attainment and intelligence on IQ. I had lots of interest with many questions after the talk and several people approaching me to ask advice. |
Year(s) Of Engagement Activity | 2017 |
Description | Invited seminar at University of Bordeaux |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Postgraduate students |
Results and Impact | I was invited to present on my recent work using Mendelian Randomization (MR) to exmaine the causal effect of educational attainment on risk of Alzheimer's Disease. The aim of the talk was to highlight a relatively novel causal inference method that can be used to identify risk factors for Alzheimer's and dementia, and provide a platform for people to engage with me to ask questions about the method, get help with analyses, or collaborate with me on future projects. Many people approached me after the talk to ask me questions and seemed very interested in and excited by this new method. They also noted my email address so they could contact me in future for assistance with/advice on MR, or to collaborate. |
Year(s) Of Engagement Activity | 2017 |