Raising awareness for dementia risk reduction in the general population Towards Primary Prevention of Dementia Irene Sophia Heger i i r f r ti ri r ti i t r l l ti Ir i r

Irene Sophia Heger Towards Primary Prevention of Dementia Raising awareness for dementia risk reduction in the general population

© Irene Sophia Heger, Maastricht, 2022 All rights reserved. No part of this book may be reproduced or transmitted in any form or by any means without permission from the first author, or where appropriate, from the publishers of the publication. Cover design and layout: Zuiderlicht Printed by: Gildeprint – ISBN: 978-94-6423-683-5 ISBN e-book: 978-94-6423-690-3

Ter verkrijging van de graad van doctor aan de Universiteit Maastricht, op gezag van de Rector Magnificus, Prof. dr. Pamela Habibovic, volgens het besluit van het College van Decanen, in het openbaar te verdedigen op donderdag 24 maart 2022 om 16.00 uur door Irene Sophia Heger Geboren op 31 oktober 1987 te Voorburg Towards Primary Prevention of Dementia Raising awareness for dementia risk reduction in the general population PROEFSCHRIFT

Promotores: Dr. S. Köhler Prof. dr. F. R. J. Verhey Co-promotor: Dr. M.P.J. van Boxtel Beoordelingscommissie: Prof. dr. R. Crutzen (voorzitter) Dr. J.A.H.R Claassen (Radboudumc, Nijmegen) Prof. dr. C.M. van Heugten Prof. dr. J.W.M. Muris Dr. C.H.M Smits (Pharos Expertisecentrum gezondheidsverschillen, Utrecht) The research described in this thesis was performed at the Department of Psychiatry and Neuropsychology, School for Mental Health and Neuroscience, Maastricht University, Alzheimer Centre Limburg, Maastricht, the Netherlands. The research presented in this thesis was supported by the Ministry of Economic Affairs by means of the PPP Allowance made available by the Top Sector Life Sciences & Health to stimulate public-private partnerships (LSHM17080-SGF); the Province of Limburg (SAS-2015-04931); and Health Foundation Limburg. Printing of this thesis was kindly supported by Alzheimer Nederland and Maastricht University.

Contents Chapter 1 General introduction, thesis aim and outline 11 Part 1 Epidemiological perspective 25 Chapter 2 Associations of the LIfestyle for BRAin Health (LIBRA) index with structural brain changes and cognition: results from The Maastricht Study Neurology, 2021 27 Chapter 3 Socioeconomic position, modifiable dementia risk and cognitive decline: results of 12-year the Maastricht Aging Study International Journal of Epidemiology, submitted 65 Part 2 Public health perspective 97 Chapter 4 Dementia awareness and risk perception in middle-aged and older individuals: baseline results of the MijnBreincoach survey on the association between lifestyle and brain health BMC Public Health, 2019 99 Chapter 5 Raising awareness for dementia risk reduction through a public health campaign: a pre-post study BMJ Open, 2020 131 Chapter 6 Appreciation of an mHealth tool to increase knowledge and beliefs and attitudes for dementia risk reduction: a pre-post proof-ofconcept study Journal of Medical Internet Research mHealth and uHealth, pending revisions 155 Chapter 7 General discussion 195 Addendum 211 Summary 213 Nederlandse samenvatting 219 Impact paragraph 225 List of publications 233 Dankwoord 239 Thesis defences from MHeNs 247 Author information 255

Chapter 1 General introduction, thesis aim and outline 11

General Introduction 13 Introduction Like in most other non-communicable diseases such as cancer and chronic lung disease1, a part of the dementia risk is modifiable. Latest estimates suggest that around 40% of all dementia cases worldwide are potentially preventable through lifestyle change and proper cardiovascular risk management2. While these insights are promising, awareness of dementia risk reduction in the general population is still low and most people still perceive dementia as an inevitable part of aging3. This General Introduction will place this thesis in the context of national and international dementia research. It defines what dementia is, including the current state of research on dementia risk reduction, awareness in the general population, and how this thesis aims to contribute to the field. Dementia Dementia is a syndrome with a wide variety of clinical manifestations and aetiologies. While mild decline in cognitive performance can be part of normal aging, the cognitive deficits in dementia are more substantial, most often progressive, interfere with independence in everyday activities and are caused by underlying neuropathology such as in Alzheimer’s disease and vascular dementia4, 5. The number of people with dementia is rising worldwide due to our aging population, as age is the most important risk factor. In the Netherlands, dementia prevalence has increased six-fold from 1950 to 2021, with estimates showing that the current 290.000 people with dementia in the Netherlands will increase to 620.000 in 20506. Dementia has a substantial impact on the person with the diagnosis, relatives and carers, as well as on the society and economy as a whole4-6. To date, there is no curative treatment available for dementia with clinically relevant health benefits4, 5. The potential of prevention Recent trends in dementia rates provide better understanding into the topic of dementia risk reduction and the potential of prevention. Epidemiological studies have recently shown that the age-specific prevalence and incidence of dementia are stable or even declining in high-income countries, probably due to improvements in vascular healthmanagement, risk factor control, and improved provision of and access to education and health care. In contrast, an increasing number of people in low-to-middle income countries are at increased risk for developing dementia, reflected by the increased presence of risk factors (e.g., diabetes, hypertension, and smoking), less health care opportunities, lower

14 Chapter 1 risk awareness and less educational attainment7-12. An important underlying concept in understanding the protective effect of educational attainment and other cognitively stimulating activities is cognitive reserve13. Cognitive reserve explains the differences between individuals in level of tolerance to age-related brain changes. Some people seem less susceptible to these brain changes and can still maintain their cognitive function, compared to others. It appears that this level of tolerance -the cognitive reserve- can be increased by engaging in cognitively stimulating activities, leading to a delay in cognitive decline and dementia onset13. Above all, the diverging trend in dementia rates between high- and low-tomiddle income countries shows the worldwide health inequalities that play a significant role in health status and incidence of diseases. It also shows the potential for primary prevention of dementia. Despite unmodifiable factors, such as genetics and age, there seems to be room to improve brain health and delay or prevent dementia onset via improved general health, lifestyle and wellbeing. Recent developments in dementia research and policy The life-course model of the Lancet Commission on Dementia Prevention, Intervention and Care14 has been updated in 2020 and its authors conclude that around 40% of dementia is potentially attributable to 12 modifiable risk factors. Risk factors vary across the life span, from less education in early life, hypertension and excessive alcohol consumption in midlife, and smoking, depression and physical inactivity in later life. One of the key messages of the Commission is to “be ambitious about prevention”2. To scaffold these associations between modifiable risk factors and dementia as being causal in nature, randomized controlled trials are needed. The 2-year Finnish Geriatric Intervention Study to Prevent Cognitive Impairment and Disability (FINGER) was the first large-scale multi-domain lifestyle trial to find beneficial effects on cognition15. A recent initiative called World Wide FINGERS was set up to harmonize worldwide intervention trials aimed at delaying or preventing dementia onset16. The Dutch branch, the FINGER-NL study, was launched in 2021. From a policy perspective, an important mark was the identification of risk reduction of dementia as one of the action areas for 2017-2025 by the World Health Organization17, followed by publishing guidelines for risk reduction of

General Introduction 15 cognitive decline and dementia in 201918. In the same year, 67 Dutch health professionals and scientists published a letter in a national newspaper in which they summoned the Minister of Health, Welfare and Sports to invest in primary prevention of dementia19. The National Dementia Strategy 2021-2030 was published in the following year. Overall, the publication of this Strategy seems a promising development for the Dutch dementia policy. However, dementia risk reduction is not incorporated as one of the major pillars of the strategy yet20. Estimating the risk of dementia Dementia risk indices aim to quantify the dementia risk to identify at-risk individuals that can be targeted for prevention initiatives. However, most existing indices are developed within a single cohort study and have not been externally validated. This is problematic since such indices are prone to capitalizing on chance or on associations that only exist in this specific cohort21. In addition, they usually incorporate factors that are not amenable to change, most notably age22-27. Age remains the best predictor of dementia risk and has on its own almost the same predictive validity as complete indices28. Age essentially remains a container variable, which presumably makes up a proxy measure of yet unidentified age-related disease mechanisms with an impact on cognitive outcome. In addition, age does not inform about the room for improvement. In 2015, the LIfestyle for BRAin health (LIBRA) index was developed by researchers fromMaastricht University, based on a comprehensive literature review and international Delphi consensus rounds29. As shown in Figure 1, LIBRA incorporates twelve potentially modifiable factors and aims to identify at-risk individuals who will benefit most from lifestyle interventions and cardiovascular risk management. Therefore, LIBRA is considered to be useful as a participant selection tool and as a surrogate outcome measure in lifestyle intervention trials30, and to inform people about target behaviours in public health initiatives. Indeed, to date LIBRA was shown to predict cognitive decline and dementia risk in several prospective cohort studies31-36.

16 Chapter 1 9% 11% 17% 19% 5% 6% 7% 8% 9% 9% Healthy diet/ Mediterranean diet Depression High cognitive activity Remember to manage: - Chronic kidney disease 6% - Diabetes 7% - Coronary heart disease 6% Low/moderate alcohol intake Physical inactivity Cholesterol Smoking Midlife obesity Midlife hypertension Figure 1. The individual contribution of each risk or protective factor included in the total score of the LIfestyle for BRAin health (LIBRA) index (based on weight from existing meta-analyses). Risk awareness From a public health perspective, these results are promising. However, most people in the general population are still unaware of the potential of dementia prevention, let alone of specific actions to reduce dementia risk. The British Social Attitudes survey from 2015 concluded that the public knowledge of dementia risk factors is considerably low. The majority of people did select dementia in their top three from a list of health conditions for clinicians and scientists to prevent37. According to a systematic review from 2018, almost half of all respondents perceived dementia as a condition which is “nonpreventable”3. Gaps of knowledge exist in particular for the cardiovascular risk factors of dementia (e.g., hypertension or obesity) which means that most people are not aware of the insight that “what is healthy for the heart, is healthy for the brain”.

General Introduction 17 Thesis aims and outline This thesis used two different perspectives to contribute to the field of dementia risk reduction. In the first part, an epidemiological perspective was taken to further validate LIBRA. To explain underlying mechanisms of the well-established association between LIBRA and cognitive performance and dementia risk, we explored plausible biological pathways of this association, using volumetric markers in brain Magnetic Resonance Imaging (MRI). In another study, we investigated whether the association between LIBRA and rate of cognitive decline differs across different socioeconomic strata. With this study, we aimed to identify at-risk groups that are important to target and design interventions for. In the second part of this thesis, a public health perspective was chosen, in which we assessed the level of awareness of dementia risk reduction in the general population and conducted proof-ofconcept studies aimed at raising awareness of dementia risk reduction in the general population. For this purpose, a public health campaign (“ We zijn zelf het medicijn”) and mHealth intervention (“ MijnBreincoach” or “MyBraincoach”) were developed and evaluated. LIBRA was used to identify individual roomfor-lifestyle-improvement and to motivate people to make brain-healthy lifestyle choices. The underlying theoretical behavioural change model for these proof-of-concept studies was the Theory of Planned Behaviour (see Figure 2). In short, this theory assumes that behavioural beliefs (i.e., the likely consequences of the behaviour) determine the attitudes toward the behaviour, normative beliefs (i.e., expectations from others) determine the subjective norm (i.e., the perceived social pressure) and control beliefs (i.e., facilitators and barriers to engage in the behaviour) determine the perceived behavioural control38, 39. These constructs, subsequently, influence the intention to perform the behaviour. The intention and the behavioural control, in turn, predict the actual behaviour38, 39. We aimed to identify both the barriers and difficulties in developing and executing public health initiatives for raising awareness of dementia risk reduction, as well as the advantages and gains, to improve our understanding of key factors that need to be considered in future studies. Also, we aimed to facilitate steps towards joining (inter)national forces towards primary prevention of dementia.

18 Chapter 1 Figure 2. Diagram of the different constructs of the Theory of Planned Behaviour. Copyright © 2019 Icek Ajzen Behavioural beliefs Attitude towards the behaviour Normative beliefs Subjective norm Control beliefs Percieved behavioural control Intention Behaviour Actual behavioural control

General Introduction 19 The outline of this thesis is as follows: Part 1: Epidemiological perspective Chapter 2 addressed the following research question: To what extent can volumetric brain markers explain the association between LIBRA and cognitive performance? This question was addressed using data of The Maastricht Study, a population-based cohort study in the south of Limburg, the Netherlands. A cross-sectional analysis was conducted to analyse the indirect, mediating effect of volumetric brain markers in the association between LIBRA and cognitive performance (including memory function, information processing speed and executive functioning and attention). The role of socioeconomic status was the focus of Chapter 3, with the following research question: Are there socioeconomic differences in the relationship between LIBRA and cognitive functioning over time? We analysed this question using the 12-year follow-up of the Maastricht Aging Study (MAAS), an ongoing prospective cohort study in the province of Limburg, the Netherlands. Part 2: Public health perspective Chapter 4 focussed on assessing literacy of dementia risk reduction in a random sample of middle-aged community-dwelling individuals in the province of Limburg. We aimed to answer the following research question: What does the general public know about modifiable dementia risk and protective factors and what are their needs, wishes and barriers to engage in a brain-healthy lifestyle? This study was used as the rationale for the developed public health initiatives that are described in Chapter 5 and Chapter 6. Chapter 5 described and evaluated a 10-month public health campaign that was launched in the Province of Limburg, the Netherlands, aimed at raising awareness of dementia risk reduction. This chapter focussed on the research question: What is the effect of a public health campaign on level of awareness of dementia risk reduction and what are the lessons learned for future studies and public health initiatives? The assessment that was described in Chapter 4 was repeated in a new random sample after the campaign to investigate pre/post-campaign differences in level of awareness.

20 Chapter 1 Chapter 6 focussed on the MijnBreincoach mHealth intervention with the research question: To what extent is the MijnBreincoach app used and appreciated, and what is the effect of the app on perceived knowledge and beliefs and attitudes towards dementia risk reduction? A proof-of-concept study was carried out in which two versions of the tool were used, which differed on the level of personalization (“tailoring”). The last Chapter 7 provided a general discussion of the main finding of this thesis, including the methodological considerations, implications, and future directions.

General Introduction 21 9. Prince M, Ali G-C, Guerchet M, Prina AM, Albanese E, Wu Y-T. Recent global trends in the prevalence and incidence of dementia, and survival with dementia. Alzheimers Res Ther 2016;8:23-23. 10. Satizabal CL, Beiser AS, Chouraki V, Chêne G, Dufouil C, Seshadri S. Incidence of Dementia over Three Decades in the Framingham Heart Study. NEJM 2016;374:523-32. 11. Wolters FJ, Chibnik LB, Waziry R, et al. Twenty-seven-year time trends in dementia incidence in Europe and the United States: The Alzheimer Cohorts Consortium. Neurology 2020;95:e519-e31. 12. Prince M, Albanese E, Guerchet M, Prina AM, World Alzheimer Report 2014. Dementia and risk reduction. An analysis of protective and modifiable factors. 2014, Alzheimer’s Disease International (ADI): London. 13. Stern Y. Cognitive reserve in ageing and Alzheimer’s disease. Lancet Neurol 2012;11:1006-12. 14. Livingston G, Sommerlad A, Orgeta V, et al. Dementia prevention, intervention, and care. The Lancet 2017;390:2673734. 15. Ngandu T, Lehtisalo J, Solomon A, et al. A 2 year multidomain intervention of diet, exercise, cognitive training, and vascular risk monitoring versus control to prevent cognitive decline in at-risk elderly people (FINGER): a randomised controlled trial. Lancet 2015;385:2255-63. 16. Kivipelto M, Mangialasche F, Snyder HM, References 1. World Health Organization. Preventing noncommunicable diseases. https:// noncommunicable-diseases (14-04-2021, date last accessed). 2. Livingston G, Huntley J, Sommerlad A, et al. Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Lancet 2020;396:413-46. 3. Cations M, Radisic G, Crotty M, Laver KE. What does the general public understand about prevention and treatment of dementia? A systematic review of population-based surveys. PLoS One 2018;13. 4. Patterson C, World Alzheimer Report 2018, in The state of the art of dementia research: New frontiers . 2018, Alzheimer’s Disease International: London. 5. World Health Organization. Dementia. 2020. fact-sheets/detail/dementia (14-04-2021, date last accessed). 6. Alzheimer Nederland. Feiten over dementie. 2017. (19 Jan, date last accessed). 7. Hachinski V, Einhäupl K, Ganten D, et al. Preventing dementia by preventing stroke: The Berlin Manifesto. Alzheimers Dement 2019;15:961-84. 8. Roehr S, Pabst A, Luck T, Riedel-Heller SG. Is dementia incidence declining in high-income countries? A systematic review and meta-analysis. Clin Epidemiol 2018;10:1233-47.

22 Chapter 1 et al. World-Wide FINGERS Network: A global approach to risk reduction and prevention of dementia. Alzheimers Dement 2020;16:1078-94. 17. Global action plan on the public health response to dementia 2017-2025. 2017, World Health Organization: Geneva. 18. Risk reduction of cognitive decline and dementia: WHO guidelines. 2019, World Health Organization: Geneva. 19. Laten we de duurste ziekte aanpakken - dementie, in NRC Handelsblad. 2019: Amsterdam. 20. Nationale dementiestrategie. 2020, Ministerie van Volksgezondheid, Welzijn en Sport. 21. Tang EY, Harrison SL, Errington L, et al. Current Developments in Dementia Risk Prediction Modelling: An Updated Systematic Review. PLoS One 2015;10:e0136181. 22. Stephen R, Liu Y, Ngandu T, et al. Associations of CAIDE Dementia Risk Score with MRI, PIB-PET measures, and cognition. J Alzheimers Dis 2017;59:695705. 23. Kivipelto M, Ngandu T, Laatikainen T, Winblad B, Soininen H, Tuomilehto J. Risk score for the prediction of dementia risk in 20 years among middle aged people: a longitudinal, population-based study. Lancet Neurol 2006;5:735-41. 24. Reitz C, Tang MX, Schupf N, Manly JJ, Mayeux R, Luchsinger JA. A summary risk score for the prediction of Alzheimer disease in elderly persons. Arch Neurol 2010;67:835-41. 25. Exalto LG, Quesenberry CP, Barnes D, Kivipelto M, Biessels GJ, Whitmer RA. Midlife risk score for the prediction of dementia four decades later. Alzheimers Dement 2014;10:562-70. 26. Cherbuin N, Shaw ME, Walsh E, Sachdev P, Anstey KJ. Validated Alzheimer’s Disease Risk Index (ANU-ADRI) is associated with smaller volumes in the default mode network in the early 60s. Brain Imaging Behav 2019;13:65-74. 27. Barnes DE, Yaffe K. Predicting dementia: role of dementia risk indices. Future Neurol 2009;4:555-60. 28. Licher S, Yilmaz P, Leening MJG, et al. External validation of four dementia prediction models for use in the general community-dwelling population: a comparative analysis from the Rotterdam Study. Eur J Epidemiol 2018;33:645-55. 29. Deckers K, van Boxtel MP, Schiepers OJ, et al. Target risk factors for dementia prevention: a systematic review and Delphi consensus study on the evidence from observational studies. Int J Geriatr Psychiatry 2015;30:234-46. 30. Coley N, Hoevenaar-Blom MP, van Dalen JW, et al. Dementia risk scores as surrogate outcomes for lifestyle-based multidomain prevention trials-rationale, preliminary evidence and chal lenges. Alzheimers Dement 2020;16:1674-85. 31. Vos SJB, van Boxtel MPJ, Schiepers OJG, et al. Modifiable Risk Factors for Prevention of Dementia in Midlife, Late Life and the Oldest-Old: Validation of the LIBRA Index. J Alzheimers Dis 2017;58:537-47. 32. Schiepers OJG, Kohler S, Deckers K, et al. Lifestyle for Brain Health (LIBRA): a

General Introduction 23 new model for dementia prevention. Int J Geriatr Psychiatry 2018;33:167-75. 33. Pons A, LaMonica HM, Mowszowski L, Kohler S, Deckers K, Naismith SL. Utility of the LIBRA Index in Relation to Cognitive Functioning in a Clinical Health Seeking Sample. J Alzheimers Dis 2018;62:373-84. 34. Deckers K, Nooyens A, van Boxtel M, Verhey F, Verschuren M, Kohler S. Gender and Educational Differences in the Association between Lifestyle and Cognitive Decline over 10 Years: The Doetinchem Cohort Study. J Alzheimers Dis 2018;70:S31-S41. 35. Deckers K, Cadar D, van Boxtel MPJ, Verhey FRJ, Steptoe A, Kohler S. Modifiable Risk Factors Explain Socioeconomic Inequalities in Dementia Risk: Evidence from a Population-Based Prospective Cohort Study. J Alzheimers Dis 2019;71:549-57. 36. Deckers K, Köhler S, Ngandu T, et al. Quantifying dementia prevention potential in the FINGER randomized controlled trial using the LIBRA prevention index. Alzheimers Dement 2021. 37. Marcinkiewicz A, Reid S, Attitudes to dementia: Findings from the 2016 British Social Attitudes survey 2016, NatCen Social Research: London. 38. Ajzen I. Theory of Planned Behavior. html (May 11, 2021, date last accessed). 39. Godin G, Kok G. The theory of planned behavior: a review of its applications to health-related behaviors. AJHP 1996;11:87-98.

24 Chapter 2

Associations of LIBRA with structural brain changes and cognition 25 Part 1 Epidemiological perspective

Associations of LIBRA with structural brain changes and cognition 27 Chapter 2 Associations of the LIfestyle for BRAin Health (LIBRA) index with structural brain changes and cognition: results from The Maastricht Study Irene Heger, Kay Deckers, Miranda Schram, Coen Stehouwer, Nicolaas Schaper, Pieter Dagnelie, Carla van der Kallen, Annemarie Koster, Simone Eussen, Jacobus Jansen, Frans Verhey, Martin van Boxtel, Sebastian Köhler Neurology, 2021

Associations of LIBRA with structural brain changes and cognition 29 Abstract Background and Objectives: Observational research has shown that a substantial proportion of all dementia cases worldwide is attributable to modifiable risk factors. Dementia risk scores might be useful to identify highrisk individuals and monitor treatment adherence. The objective of this study was to investigate whether a dementia risk score, the LIfestyle for BRAin health (LIBRA) index, is associated with MRI markers and cognitive functioning/ impairment in the general population. Methods: Cross-sectional data was used from the observational populationbased cohort of The Maastricht Study. The weighted compound score of LIBRA (including twelve dementia risk and protective factors, e.g., hypertension, physical inactivity) was calculated, with higher scores indicating higher dementia risk. Standardized volumes of white matter, grey matter, CSF (as proxy for general brain atrophy), white matter hyperintensities, and presence of cerebral small vessel disease were derived from 3T MRI. Cognitive functioning was tested in three domains: memory, information processing speed, and executive function and attention. Values ≤1.5 SD below the average were defined as cognitive impairment. Multiple regression analyses and structural equation modelling were used, adjusted for age, sex, education, intracranial volume and type-2 diabetes. Results: Participants (n=4,164; mean age 59y; 49.7% men) with higher LIBRA scores (mean=1.19, range=-2.7 to +9.2), denoting higher dementia risk, had higher volumes of white matter hyperintensities (β=0.051, p=.002), and lower scores on information processing speed (β=-0.067, p=.001) and executive function and attention (β=-0.065, p=.004). Only in men, associations between LIBRA and volumes of grey matter (β=-0.093, p<.001), CSF (β=0.104, p<.001) and memory (β=-0.054, p=.026) were found. White matter hyperintensities and CSF volume partly mediated the association between LIBRA and cognition. Discussion: Higher health- and lifestyle-based dementia risk is associated with markers of general brain atrophy, cerebrovascular pathology and worse cognition, suggesting that LIBRA meaningfully summarizes individual lifestylerelated brain health. Improving LIBRA factors on an individual level might improve population brain health. Sex differences in lifestyle-related pathology and cognition need to be further explored.

30 Chapter 2 Classification of Evidence: This study provides Class II evidence that higher LIBRA scores are significantly associated with lower scores on some cognitive domains and a higher risk of cognitive impairment.

Associations of LIBRA with structural brain changes and cognition 31 Introduction A substantial proportion of dementia cases might be attributable to modifiable risk factors1,2. Early detection of individuals at risk, allowing for timely management, has great public health implications1, as echoed by recent reports of the Lancet Commission on Dementia Prevention, Intervention and Care2 and the World Health Organization (WHO)3. Dementia risk scores, summarizing individual risks, might be useful for selection of high-risk individuals and could serve as intermediate outcomes to monitor treatment adherence. Some risk scores have been associated with structural brain changes and cognitive functioning4-7, but most are based on single cohort studies and/or include factors that are not amenable to change, e.g., age4-8, known to be highly correlated with brain markers. The LIfestyle for BRAin health (LIBRA) index is based on a systematic literature review and Delphi consensus on factors amendable to change9, thereby summarizing one’s potential for brain health improvement9. Criterion validity has been established by several prospective studies relating higher LIBRA scores with steeper cognitive decline, incident cognitive impairment and dementia in mid- and late life9-14, and intervention effects in multifactorial randomized controlled trials15. Whether LIBRA is also related to brain markers, reflecting more direct neurobiological markers of ‘brain health’, remains to be elucidated. Therefore, this study aimed to examine the association of LIBRA with cognitive performance and impairment, and evidence of neuroimaging abnormalities in the general adult population (aged 40-75 years). In addition, we investigated biological plausible pathways by testing whether MRI markers mediated the association of LIBRA with cognition.

32 Chapter 2 Methods Participants Data were used from The Maastricht Study, an observational population-based cohort study, of which the rationale and methodology has been described previously16. In brief, the study focusses on the etiology, pathophysiology, complications and comorbidities of type 2 diabetes mellitus (T2DM) and is characterized by an extensive phenotyping approach. Eligible for participation were individuals aged between 40 and 75 years and living in the southern part of theNetherlands. Participantswere recruited throughmassmedia campaigns and from the municipal registries and the regional Diabetes Patient Registry (which includes virtually all T2DM individuals in primary, secondary or tertiary care in the targeted population) via mailings. Recruitment was stratified according to known T2DM status, with an oversampling of individuals with T2DM, for reasons of efficiency, while at the same time monitoring the representation of the source population continuously (reprinted with permission)16,17. The present report addresses the following primary research questions: Are higher (i.e., more unhealthy) LIBRA scores associated with lower scores on cognitive functioning and a higher odds of cognitive impairment (Class II Evidence)? Are higher LIBRA scores associated with lower volumes of MRI markers and a higher odds of cerebral small vessel disease (Class II Evidence)? To what extent can volumetric MRI markers explain the association between LIBRA and cognitive functioning (Class II Evidence)? Cross-sectional data was used from participants who completed the baseline survey between November 2010 and January 2018. The examinations of each participant were performed within a time window of three months. MRI measurements were implemented from December 2013 onwards. Participants were included in the analyses if data on MRI outcomes, at least 11 LIBRA factors (see Table 1 and Operationalization of the LIBRA score) and cognition were available. Operationalization of the LIBRA score The individual LIBRA factors were created based on clinical data from physical examination and/or self-reported questionnaires from the baseline measurement of The Maastricht Study and then dichotomized (presence of LIBRA factor yes/no) according to established cut-offs. The LIBRA total score is computed by assigning a weight (positive for presence of risk factors; negative for presence of protective factors) to each factor, based on the relative risks from published meta-analyses9,18. Weights are then standardized and summed

Associations of LIBRA with structural brain changes and cognition 33 up to a total score. A higher LIBRA score reflects higher dementia risk, with scores ranging from -5.9 to +12.79. All LIBRA factors could be operationalized in The Maastricht Study, except for the LIBRA factor high cognitive activity. Engagement in cognitively stimulating activities was not available in the dataset and, therefore, this LIBRA factor could not be included in the risk calculation. Available protective factors were adherence to a Mediterranean diet and low to moderate alcohol use. Risk factors were physical inactivity, smoking, obesity, depression, T2DM, hypertension, hypercholesterolemia, heart disease and chronic kidney disease. See Table 1 for an overview of all individual LIBRA factors, assigned weights and operationalization in this dataset. Adherence to a Mediterranean diet was based on the Greek Mediterranean diet score derived from a comprehensive 253-item self-administered Food Frequency Questionnaire (FFQ) on frequency (not used to 7 days/week) and consumed amounts (<1/day to >12/day), with a 1-year reference period19. The Mediterranean diet score consists of the reported intake of vegetables, fruit and nuts, fish, cereal intake, dairy, meat and alcohol, with scores ranging from 0 to 9. A score of ≥6 is used as a cut-off for adhering to the diet20. Nonadherence to this diet does not necessarily imply non-adherence to the Dutch food-based dietary guidelines, which provide a more general guideline for a healthy diet in relation to numerous chronic diseases than specifically for brain health and dementia21. Physical inactivity was based on self-reported moderate to vigorous physical activity in the past two weeks, calculated from a modified version of the Community Healthy Activities Model Program for Seniors (CHAMPS) questionnaire22. Less than 150 minutes per week of moderate to vigorous physical activity was categorized as physically inactive, based on the Dutch physical activity guidelines23. Smoking status was defined by selfreported data on smoking cigarettes, with response options ‘never smoked’, ‘ever smoked’ and ‘currently smoking’: current smokers were assigned to the risk group. Low to moderate alcohol use was based on self-reported alcohol use per day based on an item of the FFQ, converted into grams of ethanol per day. Low to moderate alcohol intake was defined as ≤ 70 grams per week, based on the Dutch guidelines recommending not to drink, or to drink no more than one glass of alcohol a day21. Obesity was based on the WHO categories24, in which a BMI (calculated from physical examination at the research centre) of ≥ 30 kg/m2 was defined as obese. The presence of depression was assessed using the Mini International Neuropsychiatric Interview (MINI; current major or minor depressive episode)25. In case of missing data on the MINI, the Patient Health

34 Chapter 2 Questionnaire (PHQ-9) was used to determine presence of moderate to severe depressive symptoms (range 0-27; cut-off ≥10)26. T2DM was defined based on glucose tolerance status based on fasting glucose (≥7.0), oral glucose tolerance test (≥ 11.1), according to WHO definition, or information on current diabetes medication use27. For sensitivity analyses, a second variable was computed based on impaired glucose metabolism, which includes both prediabetes and T2DM. Hypertension was based on average office blood pressure measurement (systolic blood pressure ≥ 140 mmHg or diastolic blood pressure ≥ 90 mmHg) and/or current antihypertensive medication use. Hypercholesterolemia was calculated from serum total cholesterol using a cut-off of ≥6.5 mmol/l. The LIBRA factor heart disease was based on self-reported history of cardiovascular disease from the Rose Questionnaire28 (i.e., myocardial infarction, and/or percutaneous artery angioplasty of the coronary arteries, abdominal arteries peripheral arteries or carotid artery, and/or vascular surgery on coronary arteries, abdominal arteries peripheral arteries or carotid artery). Presence of cerebrovascular infarction and haemorrhage were not included in the risk calculation of the LIBRA factor heart disease. For sensitivity analyses, a second variable was computed based only on self-reported history of myocardial infarction29, thereby including only coronary heart disease. Chronic kidney disease was derived from CKD-EPI equation estimated glomerular filtration rate using serum cystatin C (serum cystatin C of < 60) and/or average urinary albumin excretion (both microalbuminuria and macro-albuminuria defined as risk)30.

Associations of LIBRA with structural brain changes and cognition 35 LIBRA factor Weighta Operationalized in The Maastricht Study Adherence to a mediterranean diet -1.7 Greek Mediterranean diet score (range 0-9) based on a 253-item Food Frequency Questionnaire (FFQ; 1 year reference period). Scores of ≥ 6 are categorized as “adherence to the diet”. Physical inactivity +1.1 Less than 150 minutes per week of (self-reported on CHAMPS questionnaire) moderate to vigorous physical activity in the past 2 weeks was categorized as “physically inactive”. Smoking +1.5 Self-reported data on smoking cigarettes based on an item of the FFQ. Current smokers were included in the risk score. Low-to-moderate alchohol intake -1.0 Self-reported alcohol intake based on the FFQ. Low to moderate alcohol use was defined as <70 grams per week. Obesity +1.6 BMI ≥ 30 kg/m2 calculated from physical examination at the research centre. Depression +2.1 Current major or minor depressive episode based on the MINI or presence of moderate to severe depressive symp toms based on the PHQ9 (range 0-27; cut-off >=10). Table 1. Operationalization of LIBRA factors. Table continues on next page.

36 Chapter 2 Type-2 diabetes +1.3 Glucose tolerance status based on fasting glucose (≥7.0) or oral glucose tolerance test (≥11.1), or information on current diabetes medications. Hypertension +1.6 Average systolic blood pressure ≥ 140 mmHg, or diastolic blood pressure ≥ 90, and/or current antihypertensive medication use. High cholesterol +1.4 Serum total cholesterol ≥6.5 mmol/l. Heart disease +1.0 Self-reported history of cardiovascular disease (cerebrovascular accidents excluded). Chronic kidney disease +1.1 Levels of serum cystatin C of < 60 and/or average albuminuria categories, based on average urinary albumin excretion. Microalbuminuria and macro-albuminuria were defined as risk. Cognitive activity -3.2 Data not available in dataset. Abbreviations: LIBRA – LIfestyle for BRAin health index; BMI – Body Mass Index; MINI – Mini International Neuropsychiatric Interview; PHQ9 – Patient Health Questionnaire. a Positive weights are assigned to risk factors and negative weights are assigned to protective factors. Total range: -5.9 to +12.7; range adjusted to this study -2.7 to +12.7.

Associations of LIBRA with structural brain changes and cognition 37 Cognitive performance Cognitive performance was assessed by a concise (30-minute) neuropsychological test battery16. For conceptual clarity, individual neuropsychological test scores were standardized and divided into three cognitive domains (memory function, information processing speed, and executive function and attention (reprinted with permission)16,17. Briefly, memory function was evaluated using the Verbal Leaning Test31 and a memory domain score was derived by calculating the average of total immediate and delayed recall standardized scores. An information processing speed domain score was derived from standardized scores of the Stroop Colour-Word Test (SCWT) part I and II32, the Concept Shifting Test (CST) part A and B33 and the Letter-Digit Substitution Test34. The executive function and attention domain score was calculated from the average score of the SCWT part III and the CST part C. If necessary, individual test scores were log-transformed to reduce skewness of distributions and/or inverted so that higher scores indicated better cognitive performance. In addition, participants were categorized as cognitively impaired (yes/no) based on a regression-based normalization procedure per test that predicted expected scores for each individual given their age, sex and level of education from a published normative sample31-34. The difference between observed and expected scores and their standard deviation were used to calculate z-scores, which were then averaged per domain and re-standardized. Individuals performing ≤1.5 SD below their norm-based expected score in any of the three cognitive domains were categorized as having cognitive impairment. Brain MRI Brain MRI was performed on a 3 Tesla (3T) MRI scanner (MAGNETOM Prismafit Syngo MR D13D; Siemens Healthcare, Erlangen, Germany) by use of a 64-element head coil for parallel imaging, as previously described16. Measurement of brain volumes and cerebral small vessel disease T1 images and T2-weighted FLAIR images were analysed by use of an ISO13485:2012 certified automatedmethod (which included visual inspection)35,36. T1 images were segmented into grey matter (GM), white matter (WM) and (as an inverted measure of brain atrophy) CSF (1 voxel=1.00mm3=0.001 ml)35. Intracranial volume (ICV) was calculated as the sum of GM, WM and CSF. T2-weighted FLAIR and T1 images were used to calculate white matter hyperintensity (WMH) volume36. Identified WMHs were summed to assess total

38 Chapter 2 WMH burden in ml. Additionally, WMHs were visually rated with the Fazekas scale37. Lacunar infarcts and cerebral microbleeds were counted manually by three neuro-radiologists in accordance with the Microbleed Anatomical Rating Scale (MARS)38,39. Presence of cerebral small vessel disease (CSVD) was defined as a Fazekas score of ≥2, presence of lacunar infarcts, and/or presence of cerebral microbleeds. Statistical analysis Independent samples t-tests and χ2-tests were used to investigate differences in demographic variables and LIBRA scores between the actual study sample used in the present study and the excluded group, and between three LIBRA groups (low risk: ≤1 SD below sample mean, middle risk: between -1 and +1 SD, and high risk: ≥1 SD above sample mean). The association between LIBRA and the structural MRI markers and between LIBRA and the three cognitive domains were analysed in separate multiple linear regression analyses. A quadratic term of LIBRA was added to the linear function in the analyses of the cognitive domains information processing speed and executive function and attention as this improved model fit. For direct comparison of strength of associations, we report the standardized regression coefficient beta and 95% CI. Logistic regression analyses were used to examine the association between LIBRA and CSVD, and between LIBRA and cognitive impairment, yielding odds ratios (OR) and 95% CIs. Structural equation modelling was used to study mediation of LIBRA on cognition by MRI markers by decomposing the total association into direct and indirect associations. Because the regression analysis suggested a curvilinear association between LIBRA and two cognitive domains, we used a technique that allows for estimating non-linear mediation effects, which is not taken into account in traditional linear or log-linear mediation models (see Figure 1)40. For this, we estimated the instantaneous indirect effect θ, which tests the mediation effect at different levels of the independent predictor variable (LIBRA), showing how the mediation effects changes as the level of the independent variable changes. Following this approach, we estimated the instantaneous indirect effects θ at three levels of LIBRA: 1 SD below the LIBRA sample mean (LIBRA score = -0.87), at the LIBRA sample mean (LIBRA score = 1.19) and 1 SD above the LIBRA sample mean (LIBRA score = 3.25), following previous recommendations40. To estimate robust 95% CIs, we used bootstrapping with 10,000 repetitions.

Associations of LIBRA with structural brain changes and cognition 39 Associations with cognition were adjusted for age, sex and level of education. Associations with structural brain markers were additionally adjusted for ICV to correct for head size, and the variable MRI lag time to adjust for the time (in years) between inclusion and MRI scan. The oversampling of participants with T2DM by design urged us to adjust for diabetes status in all the analyses in order to make sure that the overexpression of LIBRA risk factors in T2DM, such as obesity, hypercholesterolemia, hypertension or depression, did not confound the observed associations between LIBRA, MRI markers and cognition. Interaction terms were included in additional analyses to investigate whether the associations between LIBRA and brain markers or cognitive performance were moderated by sex and T2DM status. Finally, we did a series of sensitivity analyses to test the robustness of findings after assigning those with prediabetes the risk weight for T2DM, and after assigning a risk weight only to those with coronary heart disease. Statistical analyses were done with Stata 13.1 (StataCorp, TX) and Mplus8 (Muthen & Muthen) using two-sided hypothesis testing and an alpha-level of < .05. Standard Protocol Approvals, Registrations, and Patient Consents The Maastricht Study has been approved by the institutional medical ethical committee (NL31329.068.10) and the Ministry of Health, Welfare and Sports of the Netherlands (Permit 131088-105234-PG). All participants gave their written informed consent16. Data Availability Statement Data are unsuitable for public deposition due to ethical restrictions and privacy regulation of participant data. Data from The Maastricht Study are available to any interested researcher who meets the criteria for access to confidential data. Data requests may be submitted to The Maastricht Study Management Team (

40 Chapter 2 Covariates MRI-markers LIBRA LIBRA2 Cognitive domains i i Figure 1. Path model to quantify the instantaneous indirect effect of LIBRA on cognition. Abbreviations: LIBRA – LIfestyle for BRAin health (continuous); LIBRA2 – LIfestyle for BRAin health (squared); ICV – intracranial volume. Covariates: sex, age, level of education, time between assessment and MRI, ICV and diabetes status. i Standard error.

Associations of LIBRA with structural brain changes and cognition 41 Results Study design and sample characteristics Of all 7,689 participants (mean age 59.8 years; 50.4% men; 34.7% low educated; 24.6% T2DM), 45.8% was excluded from the present study, largely due to absence of MRI data. LIBRA factors that were most often missing were physical inactivity (9.8% missing) and adherence to a Mediterranean diet and low to moderate alcohol intake (from the same food questionnaire; 5.2% missing). All other LIBRA factors were below 3.7% missing. See Figure 2 for a flowchart. Compared to the study sample (n=4,164), excluded participants (n=3,525) had a higher mean age (59.2 years vs. 60.5 years; (t(7687)=6.5, p <0.001) and had lower education (sample low education 30.2%, excluded low education 40.2%; X2 (2)=86.6, p <0.001). Excluded participants had a more unfavourable LIBRA risk profile (1.19 vs. 1.95; t(7687)=15.4, p <0.001), with higher presence of T2DM (19.0% vs. 31.3%; X2 (1)=156.1, p <0.001), hypertension (49.0% vs. 59.7%; X2 (1)=87.0, p <0.001), heart disease (10.1% vs. 20.3%; X2 (1)=152.3, p <0.001), obesity (18.0% vs. 25.9%; X2 (1)=70.9, p <0.001), chronic kidney disease (5.2% vs. 7.6%; X2 (1)=19.3, p <0.001), depression (4.2% vs. 6.1%; X2 (1)=13.7, p <0.001), and were more often smokers (11.0% vs. 16.4%; X2 (1)=47.5, p <0.001), physically inactive (25.3% vs. 31.6%, X2 (1)=31.9, p <0.001) and less often adhered to the Mediterranean diet (28.5% vs. 26.2%; X2 (1)=4.6, p = 0.032). Low to moderate alcohol intake was more common in the excluded group (54.9% vs. 59.4%; X2 (1)=14.7, p <0.001) and hypercholesterolemia was more common in the study sample compared to excluded participants (15.4% vs. 12.3%; X2 (1)=14.4, p <0.001). Men had higher (more unhealthy) average LIBRA scores (1.5) compared to women (0.9; t(4162)=10.3, p <0.001), including higher presence of T2DM (25.6% vs. 12.5%, X2 (1)=116.1, p<0.001), hypertension (57.7% vs. 40.4%, X2 (1)=125.1, p<0.001), and physical inactivity (28.1% vs. 22.5%, X2 (1)=17.1, <0.001). The characteristics of the total study sample and those with a low (≤1 SD below sample mean), middle (between -1 and +1 SD) and a high (≥1 SD above sample mean) LIBRA score are summarized in Table 2.

42 Chapter 2 Total sample The Maastricht Study N = 7,689 Sample with MRI data available n = 5,183 Sample with availability of 11 LIBRA factors n = 4,300 Study sample n = 4,164 Missing MRI data (n=2,506) Less than 11 LIBRA factors available (n=883) Physical inactivity (n=507) Diet (n=270) Alcohol (n=269) Heart disease (n=193) Smoking (n=35) Hypertension (n=5) Depression (n=5) Cholesterol (n=4) Missing on all cognitive domains (n=136) Memory (n=153) Information processing speed (n=169) Executive functioning and attention (n=178) Individual LIBRA factors missing: Individual domains missing: Figure 2. Flowchart of the study sample selection. Abbreviation: LIBRA – LIfestyle for BRAin health index.

Associations of LIBRA with structural brain changes and cognition 43 Table 2. Characteristics of the total sample and of participants with low, middle and high risk based on LIBRA scores. Variablesa Total sample N=4,164 Low riskb n=848 Middle riskb n=2,665 High riskb n=651 Men; n (%) 2,070 (49.7%) 319 (37.6%) 1,354 (50.8%) 397 (61.0%) Age; mean (SD) 59.2 (8.6) 55.2 (8.4) 59.8 (8.3) 62.1 (8.0) Educationc; n (%) Low 1,252 (30.2%) 167 (19.7%) 795 (29.9%) 290 (45.4%) Middle 1,184 (28.6%) 255 (30.1%) 770 (29.0%) 159 (24.9%) High 1,706 (41.2%) 424 (50.1%) 1,092 (41.1%) 190 (29.7%) Marital status; n (%) Single 287 (6.9%) 64 (7.6%) 178 (6.7%) 45 (6.9%) Married or registered 3,417 (82.1%) 704 (83.1%) 2,199 (82.5%) 514 (79.0%) Widowed/divorced 452 (10.9%) 78 (9.2%) 283 (10.6%) 91 (14.0%) Other 7 (0.2%) 1 (0.1%) 5 (0.2%) 1 (0.2%) Table continues on next page.

44 Chapter 2 LIBRA total score; mean (SD) 1.19 (2.06) -1.47 (0.61) 1.20 (1.05) 4.6 (1.08) Individual LIBRA factors; n (%) Type-2 diabetes 790 (19.0%) 4 (0.5%) 390 (14.6%) 396 (60.8%) Hypertension 2,041 (49.0%) 61 (7.2%) 1,381 (51.8%) 599 (92.0%) High cholesterol 639 (15.4%) 27 (3.2%) 493 (18.5%) 119 (18.3%) Mediterranean diet 1,186 (28.5%) 457 (53.9%) 691 (25.9%) 38 (5.8%) Heart disease 419 (10.1) 10 (1.2%) 254 (9.5%) 155 (23.8%) Chronic kidney disease 216 (5.2%) 3 (0.4%) 117 (4.4%) 96 (14.8%) Low/moderate alcohol use 2,285 (54.9%) 677 (79.8%) 1,258 (47.2%) 350 (53.8%) Physical inactivity 1,054 (25.3%) 23 (2.7%) 677 (25.4%) 354 (54.4%) Depression 175 (4.2%) 0 (0%) 67 (2.5%) 108 (16.6%) Obesity 749 (18.0%) 6 (0.7%) 320 (12.0%) 423 (65.0%) Smoking 458 (11.0%) 12 (1.4%) 300 (11.3%) 146 (22.4%) Abbreviations: LIBRA – LIfestyle for BRAin Health score (higher is more risk). a Maximum values and percentages do not count up due to missing values and rounding issues; b Low (risk) score = ≤1SD below sample mean, middle (risk) score = between -1 and +1 SD, high (risk) score = ≥1SD above sample mean; c Educational level was divided from 9 ordinal levels to 3 categories (low: no education, primary education, lower vocational education; middle: intermediate vocational education, higher secondary education; high: higher professional education, university education).