FoU i Västra Götalandsregionen
Psykisk och kognitiv hälsa i relation till funktionsförmåga och somatisk hälsa hos äldre. Påverkan av ålder, kön, födelsekohort, genetiska, sociala och neurobiologiska faktorer
Psykisk och kognitiv hälsa i relation till funktionsförmåga och somatisk hälsa hos äldre. Påverkan av ålder, kön, födelsekohort, genetiska, sociala och neurobiologiska faktorer
Project number : 240461
Created by: Ingmar Skoog, 2017-12-28
Last revised by: Ingmar Skoog, 2017-12-28
Project created in: FoU i Västra Götalandsregionen

PublishedPublished

1. Översiktlig projektbeskrivning

Populärvetenskaplig sammanfattning av projektet

Målsättning: Att studera psykisk, kognitiv och somatisk hälsa hos äldre, dess riskfaktorer och dess konsekvenser för funktionsförmåga och välbefinnande, och hur de påverkas av ålder, kön, socioekonomiska skillnader, psykosociala, neurobiologiska och genetiska faktorer under livsloppet och mellan olika födelsekohorter. Metoder: Studierna omfattar representativa födelsekohorter av äldre som följs longitudinellt; a) H70-studien, som omfattar kohorter födda 1901-02 (följda från 70-102 års ålder), 1905-6, 1922 and 1930 (följda från 70-79 år, ny uppföljning vid 85 års ålder 2015) och 1944 (undersökning pågår vid 70 års ålder); b) H85-studien, som omfattar kohorter födda 1901-02 (undersökta från 85-102 års ålder), 1923-24 (undersökta från 85-90 år, fortsatt uppföljning planeras), och 1930 (undersöks vid 85 år 2015); c) 95+ studien (population 95 år och äldre; N=950); d) Kvinnoundersökningen (1462 kvinnor följda under 47 år från 1968-2015). Studierna innehåller psykiatriska, somatiska, psykosociala, oftalmologiska, audiologiska, psykometriska, genetiska, likvor och dietistundersökningar, kroppsimpedans, laboratorietester, datortomografi och magnetkameraundersökning av hjärna, och bedömning av funktionsförmåga. Betydelse: Kunskapen om psykisk hälsa hos äldre är otillräcklig. Projektet är unikt med hänsyn till de omfattande undersökningarna, de långa uppföljningarna, den höga åldern, och att födelsekohorter undersöks över 40 år. Projektet har klinisk relevans för sjukvården, i förhållande till prevention, tidig diagnostik, patogenes och prognos, och kan förbättra omhändertagandet av äldre personer i vården.

Vetenskaplig sammanfattning av projektet

Aims: To examine mental, cognitive and somatic health in old age, their risk factors and impact on functional ability and well-being, taking into account the complex interactions with age, sex, gender, socio-economic gradients, secular changes, psychosocial, neurobiological, and genetic factors occurring across the life course and in different birth cohorts. Methods: The studies include representative birth cohorts followed longitudinally in Gothenburg; a) H70-study comprising cohorts born 1901-02 (studied from age 70 to 102 years), 1905-6 (examined at age 70 to 79), 1922 (examined at age 70, 78, 83 and 94 years), 1930 (examined from age 70 to 85, new exams planned at age 88 and 90) and 1944 (studied at age 70, new follow-ups planned at age 75 and 79); b) H85-study comprising cohorts born 1901-02 (examined from 85 to 102 years), 1923-24 (studied from age 85 to 90, new exam planned at age 95), and 1930 (examined at age 85, new follow-up planned at age 88 and 90); c) the 95+ study (populations aged 95 and over; N=1020); d) Prospective Population Study of Women (1462 women followed 48 years 1968-2015, new follow up 2018). The studies include psychiatric, somatic, audiological, opthalmological, psychosocial, genetic, dietary, and psychometric examinations, collection of blood, plasma, serum, and cerebrospinal fluid, examinations with CT-scan, MRI and PET of the head, body composition DEXA, and assessment of functional abilities. Significance: The project is unique in relation to the long follow-ups of birth cohorts examined four decades apart, and the comprehensive multidisciplinary examinations. It has clinical relevance in relation to prevention, early diagnosis, understanding of pathogenesis, clinical picture and prognosis, and may enhance the quality of care of older persons

Typ av projekt

Forskningsprojekt

MeSH-termer för att beskriva typ av studier

checked Fokusgrupper (Focus Groups)
checked Hälso- och sjukvårdsundersökningar (Health Care Surveys)
checked Longitudinella studier (Longitudinal Studies)
checked Prospektiva studier (Prospective Studies)
checked Retrospektiva studier (Retrospective Studies)
checked Kohortstudier (Cohort Studies)
checked Tvärsnittsstudier (Cross-Sectional Studies)
checked Befolkningsstudier (Population Surveillance)


(Only selected options are displayed. Click here to display all options)

MeSH-termer för att beskriva ämnesområdet

information Added MeSH terms
Mental Disorders
Psychiatric illness or diseases manifested by breakdowns in the adaptational process expressed primarily as abnormalities of thought, feeling, and behavior producing either distress or impairment of function.
Nervous System Diseases
Diseases of the central and peripheral nervous system. This includes disorders of the brain, spinal cord, cranial nerves, peripheral nerves, nerve roots, autonomic nervous system, neuromuscular junction, and muscle.

Projektets delaktighet i utbildning

checked Avhandling
checked Master
checked D-uppsats / Magisterexamen
checked ST-läkarutbildning
checked Annan utbildning


(Only selected options are displayed. Click here to display all options)

2. Projektorganisation och finansiering

Arbetsplatser involverade i projektet

information Added workplaces
Landsting - Västra Götalandsregionen - Specialiserad vård - Sahlgrenska Universitetssjukhuset - Område 2 - Neuropsykiatri workplace verified by Västra Götalandsregionen on 2018-02-27
Statligt - Universitet - Göteborgs universitet - Sahlgrenska akademin - Institutionen för neurovetenskap och fysiologi - Sektionen för psykiatri och neurokemi workplace verified by Sahlgrenska Akademin on 2017-04-05

3. Processen och projektets redovisning

Hur långt har projektet framskridit?

Rekrytering/datainsamling pågår

Projektstart (när planeringen påbörjas och börjar dokumenteras skriftligt)

1971-07-01

Datum då projektet är slutrapporterat

2030-12-31

Länk till webbplats / webbsida

www.epinep.gu.se

Detaljerad projektbeskrivning

PURPOSE AND AIMS

The majority of patients in the health care system are above age 65 years. Between 2009 and 2050, this age group will increase from 1.7 to 2.2 million in Sweden, and from 473 to 1457 million worldwide (UN). Mental, cognitive and somatic health are major determinants for well-being in old age. To improve health care for older persons, we need to to learn more about normal ageing and how it relates to disease, and to identify preventive factors and early markers for diseases.

This project examines mental and cognitive health, and its impact on functional ability and well-being, taking into account the complex interactions with age, sex, gender, somatic health, socio-economic gradients, secular changes, psychosocial, neurobiological, and genetic factors occurring across the life course. The influence of different factors may vary depending on when they occur during the life span, e.g. factors occurring before birth may interact with midlife factors to influence mental and somatic health in old age. Figure 1 illustrates the theoretical model underlying our proposal. Many associations are bidirectional, and factors outside the circle influence each other.

We aim to study

1) the association between mental and somatic health in older people and how it is influenced by birth cohort, ageing, sex, gender, genetic and socioeconomic factors

2) how neurobiological processes, measured in cerebrospinal fluid (CSF) and with brain imaging, influence mental, cognitive and somatic health, and especially determinants and prognosis in relation to preclinical Alzheimer’s disease

3) predictors and risk factors for mental, cognitive and somatic health in older persons using a life course perspective (from birth to old age)

4) the consequences of mental, cognitive and somatic health in older persons in relation to functional ability, well-being and mortality

BACKGROUND AND PRELIMINARY RESULTS

Mental and cognitive health

Mental and cognitive health in older populations are positioned on a continuum, from complete wellbeing to defined disorders, such as dementia (e.g. Alzheimer’s disease) and other mental disorders (e.g. depression). Even mild mental conditions have a profound impact on health, and how a person adapts to ageing and its related disorders.

The prevalence of dementia increases from 3% at age 70 to 52% at age 95 years. Mild cognitive impairment is even more common. Several risk factors for somatic disorders are related to dementia and Alzheimer’s disease (AD), including the APOE e4 allele, low education, psychological stress, poor lung function, and cardiovascular risk factors, such as hypertension, overweight, diabetes mellitus, and hypercholesterolemia (1, 2, 3). The importance of risk factors may differ depending on the age when they are measured. For example, stroke and Alzheimer encephalopathy are related to dementia, but the relative risk is lower in nonagenarians compared to septuagenarians and octagenarians (4).

The prevalence of depression is about 10% at all ages (5). Depression, including mild subsyndromal symptoms, has several consequences, such as poor quality of life, disability, increased use of health and home care services, cognitive decline, suicide, increased mortality rate, and increased risk for cardiovascular disease (CaD) (5, 6). Risk factors for depression include somatic health, family history, negative life events, personality, socio-economic gradient, loneliness, smoking, low education, and CaD (5, 6). Thus, depression may be both a consequence and a cause of somatic disorders, e.g. in relation to stroke (7). Risk factors for depression may present already during fetal development. Lower birth weight is related to depression (8) and several somatic disorders, maybe due to early programming as a response to intrauterine environment, indicating a role of epigenetics. Few studies have examined other mental disorders than depression in older people (5). In this project, we examine the whole range of mental disorders, including social phobia, generalized anxiety disorder (GAD) (PhD-student Johan Nilsson), specific phobia, obsessive-compulsive disorder (PhD-student Isak Freden-Klenfelt), psychotic syndromes, and alcohol and drug abuse (PhD-student Felicia Ahlner). Anxiety disorders, such as GAD, have been associated with increased risk for CaD (9). We found that paranoid ideations were related to myocardial infarction in non-demented 85-year-olds (5).

In our analyses, we will take into account socioeconomic gradient (10) and socioeconomic differences in cumulative dis/advantages (collaboration with Lotta Dellve, Björn Halleröd and PhD-student Caroline Hasselgren, Dept of Sociology), which are related to a number of health outcomes, and gender, which is related to depression and anxiety (5) (PhD-students Therese Rydberg, Hanna Olsson). We will also study multimorbidity, which is associated with death and functional disability, and probably also to depression.

Neurobiology

Neurobiological factors play a role in both mental and somatic disorders. The brain regulates a number of body systems, for example blood pressure and vessel tone. Many biological systems are disturbed in depression and GAD, leading to increased platelet aggregation and coagulation, activation of the sympathetic nervous system, immune activation, increased inflammatory response, disruption of the hypothalamic-pituitary-adrenal axis, catecholamine activation, increased insuline resistance, increased blood serotonin concentrations, decreased cardiac variability, vasoconstriction and vascular reactivity, smooth muscle cell movements and other vascular changes (6). This may partly explain the association between depression and CaD. AD may also influence body systems, e.g. blood pressure and BMI decreases several years before clinical onset of AD (1, 2). We now have CSF biomarkers and brain imaging modalities to measure the major biological processes related to ageing, AD and psychiatric disorders. One central mechanism in AD is the accumulation of beta-amyloid and tau proteins in the brain, reflected in CSF by increased levels of T-tau and reduction in beta-amyloid-42, which we will study also in relation to somatic markers. According to the theory of Jack et al (11), CSF-beta-amyloid levels decrease two-three decades before onset of clinical AD. Later in the disease process, CSF-tau increases and early changes in brain volume on MRI develop in the enthorinal cortex and medial temporal lobe structures one and a half decade before the first clinical symptoms. This theory has not been confirmed in longitudial studies. Preliminary data from our longitudinal studies shows that lower levels of CSF-beta-amyloid are present up to 18 years before dementia onset, and that none of those who developed AD during 18 years follow-up (from mean ages 70 to 90 years) had CSF beta-amyloid levels above the median. We have an unique opportunity to examine the long preclinical phase of AD in more detail during the period of the present grant (Silke Kern, Simona Sacuiu, PhD-student Mats Ribbe) and also to further examine the finding that high CSF beta-amylod levels may prevent AD. In addition, we will be able to examine longitudinal change of CSF markers in initially normal populations, and relate it to onset of dementia and cognitive decline. This has previously not been done in population studies. It is also of interest to examine biological markers of AD in relation to assessments of body composition, linking body and brain with ageing and disease, and in relation to the decline in blood pressure and BMI observed in the years preceeding dementia.

Brain imaging, such as CT and MRI, shows that cerebrovascular diseases are common in the elderly. We have shown that ischemic white matter lesions (WMLs) on CT are related to dementia, cognitive dysfunction and depression (12). WMLs are also related to increased risk for hip fracture and stroke. WMLs are even more common on MRI, but the relevance of mild WMLs on MRI is not clear. With MRI we are also able to examine microbleeds and silent infarcts in relation to dementia and depression. Brain atrophy is visualized on CT and MRI and may reflect neurodegeneration. These changes are common in AD and vascular dementia. We recently showed that temporal lobe atrophy is related to an increased risk for development of dementia and depression during 9 year follow-up (12). Changes in brain volume over time on MRI, and changes in certain areas of the brain, such as the hippocampus and medial temporal lobe, has also been shown to predict dementia. This has rarely been examined in population studies. In addition, we aim to use Positron Emission Tomography (PET), using tracers for tau (in collaboration with Michael Schöll, Wallenberg laboratory), to further examine early processes of AD in the population.

Genetic diversity may influence mental and somatic disorders, and their interactions. For example, the Apoe4 allele is a vulnerability factor for AD, but also a vulnerability factor for other conditions, such as atherosclerosis (13), and geriatric depression (14). However, traditional genetics alone is not sufficient to explain the complete picture on how genetic variability influences the phenotype. Epigenetic control mechanisms, which are modified by behavioural and environmental circumstances, may also play a role (15). This part of the study is a collaboration with Anna Zettergren, and John Hardy’s lab in London.

Secular changes

Large societal changes occurred during the 20th century in areas such as perinatal care, urbanisation, education, work related and retirement conditions, housing and hygiene, dietary habits, and health care. In our own studies, the prevalence of dementia and major depression was unchanged between 70- and 75-year-olds born 1901-02 and 1930 (16). However, the prevalence of mild depression increased from the 1970s to the 2000s. Several studies report that the prevalence of CaD risk factors, such as smoking, hypertension, and hypercholesterolemia have decreased, and the proportion with overweight and diabetes increased among middle-aged populations between the 1960s and 2000s. Similar findings were seen in 75-year-olds examined 1976-77 and 2005-6 (17). In that study, the prevalence of CaD decreased. However, in the 1970s, CaD was more common among women, while it was more common among men in the 2000s. In our study, blood pressure decreased from the 1970s to the 2000s, with a larger decrease observed among women (17), suggesting that factors which decreased blood pressure had their largest effect in women. It needs to be elucidated how the changing pattern in CaD influences the relation between mental health and somatic disorders.

Those born 1901-02 experienced more factors which weared them down than later-born cohorts. Later-born cohorts may therefore have a larger cognitive and physical reserve, making them less vulnerable to the consequences of mental and physical disorders. For example, cognitive function (18) and poor lung function (19) predicted 5-year mortality in the 1970s, but not in the 2000s. Furthermore, a fraility index including 70 negative social, somatic, psychological, and functional factors predicted survival between ages 70 and 80 in the cohort born 1901-02, but less so in those born 1930 (20). Another finding supporting the reserve model is that the prevalence of dementia in 85-year-olds decreased from 30% in 1986-87 to 22% in 2008-10, which was related to higher educational level in 2008-10, and to a decreased association between stroke and dementia. A better reserve may postpone the clinical manifestations of diseases. Our data makes it possible to study changes in mental and somatic health, their interactions, as well as their determinants, over four decades. Secular changes in risk factors may influence health in different ways. The association between negative factors and health may be constant, or the importance of risk factors may change when the prevalence of modifying factors changes. We have recently started a collaboration with the Department of Historical Studies (Helene Castenbrandt, Ulrika Lagerlöf Nilsson) to put our studies into a historical context. The study on how the ageing process changes between generations may give clues on how much of the ageing process is determined by life style factors and how much is genetically determined, as it is not likely that large changes would occur in the gene pool over 40 years. However, recent epigenetic research suggest that environmental factors may also change gene expression (15).

Increased survival in later-born cohorts may reflect a healthier population, but more people will also survive with chronic disorders, and more will develop age-related disorders. We will examine if individuals with poor mental health and somatic disorders live longer with their disorders today, and if the consequences of these disorders have changed. Among 75-year-olds, the prevalence of stroke increased between 1976-77 and 2005-6, maybe because more people survive with stroke (17). The complexity is further emphasized by studies reporting both better and worse health in later-born birth cohorts of elderly. The question on whether increased survival leads to compression or expansion of morbidity remains unanswered. It is necessary to use longitudinal studies of different birth cohorts who are followed up to extreme old age to answer these questions.

DESIGN, METHODS, WORK PLAN

The main part of the project centers around population studies of elderly in Gothenburg. These include longitudinally followed representative populations examined over more than 40 years; The Prospective Population Study of Women (PPSW), H70, H85 and 95+. All samples are systematically obtained from the Swedish Population Register based on birth dates. Response rates are 60-85%. Part of examinations have been virtually identical between studies to enhance possibilities of comparisons, and are done at the Neuropsychiatric outpatient clinic at Wallinsgatan, or in the participant’s home. New and modern examinations have also been added over time. The populations are described below, and in figure 2. We are part of NEAR (National e-infrastructur for Ageing Research in Sweden), to make databases more accessible to outside users and for adding new data. We are also part of a Swedish genomics consortium (GAPS: Genomic Aggregation Project in Sweden), and the Harvard Alzheimer Consortium (12 North-American and European studies).

The H70-study

Cohort 1901-02: In 1971-72, 70-year-olds born 1901-1902 (N=973, 85% response rate) were examined, with re-examinations at ages 75, 79, 81, 83, 85, 88, 90, 92, 95, 97, 99 and 100.

Cohort 1906-07: In 1976-77, 70-year-olds born 1906-1907 (N=1036, 81% response rate) were examined, with re-examinations at ages 75 and 79 years.

Cohort 1911-12: In 1981-82, 70-year-olds born 1911-1912 (N=619, 77% response rate) were examined, with re-examination at age 75 years.

Cohort 1922: In 1992, 70-year-olds born 1922 (N=500, 66% response rate) were examined, with re-examinations at ages 78, 83 87 and 94 years.

Cohort 1930: In 2000-2001, 70-year-olds born 1930 (N=522) were examined. At age 75, the sample was extended to include 753 persons. Re-examinations were done at ages 79 and 85 years (in 2015-16; N=491). Further follow-ups are planned at ages 88 (2018), 90 (2020), 92, 95, 97, 99 and 100.

Cohort 1944: In 2014-16, we examined 70-year-olds born 1944 (N=1202, response rate 72%). Re-examinations are planned at ages 75 (in 2019), 79 and 85 years.

The H85-study

Cohort 1901-02: In 1986-87, 85-year-olds, born 1901-02 were examined (N=962; response rate 64%), with re-examinations at ages 88, 90, 92, 95, 97, 99, 100, 102, and 103 years.

Cohort 1923-24: In 2008-10, 85-year-olds born 1923-24 were examined (N=571, response rate 61%) with re-examinations at ages 88 and 90 years, and planned at age 95 years.

Cohort 1930: In 2015-16, we examined 85-year-olds born 1930 (N=491), with planned re-examinations at ages 88 (2018), 90 (2020), 92 and 95 years etc.

The 95+ StudySTRONG·

Longitudinal examinations of 95-year-olds and older born 1901-1911 (total N=1020, response rate 65%). The examination of 97-years-olds was finished in 2007 (N=591), 99-year-olds in 2010 (N=348), 100-year olds in 2011 (N=245), 101-year-olds in 2012 (N=144), 102-year-old in 2013 (N=85), and 103-year-olds in 2014 (N=47). We have so far examined 25 104-year-olds, 15 105-year-olds, 6 106-year-olds, 4 at age 107, one at age 108 and two at age 109.

PPSW

This is a study on women born 1908, 1914, 1918, 1922 and 1930. They were examined in 1968-69 (N=1467), and re-examined in 1974-75, 1981-82, 1992-93, 2000-2002, 2005-2006, 2009-11, and 2015-16. Next follow-up will be in 2018, 50 years after the baseline examination (N=cirka 350).

Figure 2. The populations in relation to birth year, examination year and age (in the squares)

During the period 2018-2020, we will thus do follow-ups at age 88 and 90 of the cohort born 1930 (N=cirka 420), at age 75 of the cohort born 1944 (N=cirka 1000), at age 95 and 97 of the cohort born 1923-24 (N=cirka 150), and a 50 year follow-up of PPSW (N=cirka 350).

Examinations

The psychiatric examinations were performed by psychiatrists until 2000, and by MDs and psychiatric research nurses thereafter. Inter-rater reliability between nurses and specialists in psychiatry is high (kappa values for signs and symptoms of depression and dementia between 0.62 and 1.00). The psychiatric examinations include ratings of psychiatric symptoms and signs with Comprehensive Psychopathological Rating Scale, Mini-D, tests of mental functioning (e.g. memory), Mini Mental State Examination, other signs common in dementia, personality inventories (Five Factor Model, EPI) and questions about previous mental disorders, sexuality, sleep, and suicidal feelings.

Somatic examinations include assessment of medical conditions (e.g. hypertension, myocardial infarction, angina pectoris, stroke, cardiac failure, atrial fibrillation, diabetes mellitus, pulmonary diseases, fractures, thyroid diseases, head trauma, cancer), the Cumulative Illness Rating Scale for Geriatrics, anthropometric measurements, blood pressure, ECG, lung function (PI Kjell Torén, Occupational Medicine), measures of atherosclerosis, and body composition (dexa). Questions are asked about urinary function, falls, dizziness, alcohol and tobacco use, family history of psychiatric and somatic disorders, and use of health services. Medication is classified according to the Anatomical Therapeutic Chemical classification.STRONG Blood samples are taken and include e.g. hemoglobin, HbA1c, cholesterol (HDL, LDL), and homocystein. Blood, serum and plasma are frozen for future analyses. Dietary examinations are conducted by dietists (PI Elisabeth Rothenberg), ophtalmologic examinations by ophtalmological nurses (PI Madeleine Zetterberg) and audiological examinations by audiologists (PI Ulf Rosenhall).

Psychological examinations assess memory and intelligence using SRB-1, SRB-2, SRB-3, Thurstone Picture Memory, Ten Word Memory List, Clock Test, Prose Recall, and Digit Span (PI Boo Johanson, Dept of Psychology).EM 

Functional ability data is collected on basic abilities in ADL, and iADL (Katz, Lawton). Functional status is examined by a physiotherapist (PI Kerstin Frändin, Helena Hörder), including grip strength, walking speed, balance, and chair stand.

Social factors are collected on socioeconomic status, marital status, living conditions, social network, education, hobbies, previous and current work life, cultural activities and life events. This is a collaboration with Björn Halleröd and Lotta Dellve at the Department of Sociology.

Close informant interviews comprise IQCODE, questions on changes in behavior and intellectual function, psychiatric symptoms, ADL/iADL, family history, medical history and onset and course of dementia.

Genetic epidemiology is done in collaboration with Kaj Blennow, Henrik Zetterberg and Anna Zettergren at our department, and with John Hardy, University College London. DNA is extracted from whole blood using a standard procedure. We currently have frozen blood samples from over 3,600 individuals. All the DNA samples are analysed using the NeuroX genotyping array from Illumina, covering all mutations in genes known to cause neurodegenerative diseases, key tagging SNPs for all genome‐wide significant loci published in human genome wide association studies (GWAS) related to neurological disorders, the top ~1,000 SNPs just below genome‐wide significance, ~240,000 coding sequence variants covering the exome, and novel coding variants detected by whole exome sequencing of familial cases. We will also run additional genotyping arrays (e.g the PsychChip covering psychiatric disorders), and single SNPs of specific interest. Approximately 550 individuals in 95+ , who have frozen blood, will be analysed by whole exome sequencing.

Brain imaging

Since 2014, we have done 800 3 teslas MRI on 70-year-olds born 1944 and 200 on 85-year-olds born 1930. The examinations include common MRI-sequences (T1, T2, FLAIR) and newer techniques, such as susceptibility weighted imaging (SWI), diffusion tension imaging (DTI), resting state fMRI, and default mode network (DMN) imaging. MRI gives a more detailed picture of the brain and its function than CT, and can e.g.visualize microbleedings, medial temporal lobe atrophy, and hippocampal atrophy, and gives more details on white matter tracts and neural networks. Our study is currently the only population study, which examines risk factors and consequences of brain changes detected by CT-scans. In total, 4000 CT-scans have been done since 1986. CT is analysed for brain atrophy, white matter lesions, infarcts, basal ganglia calcifications, and brain volume New CTs and MRI will be done at the follow-up of those born 1930 at age 88 (N=250), and 90 years and in the H70 cohort born 1944 (N= about 700) at the follow-up at age 75.

We also plan to include PET studies using tau tracers in the study on 88-year-olds, and in the follow-up of the cohort born 1944 at age 75 (PI Michael Schöll, Wallenberg Center), to give a more detailed picture of preclinical Alzheimer’s disease.

The studies are the first world-wide to compare CT and MRI in a general population. Few studies have compared these examinations in their ability to diagnose brain diseases, such as WMLs and atrophy. This study gives a unique possibility to compare MRI and CT, and to validate these modalities with each other and with clinical data. This will have importance for clinical and research centers without access to MRI, and for patients with contraindications for MRI. CT remains the most widely used brain imaging tool worldwide, especially in the developing world but also in the western world, due to its availability and its relatively low cost. Brain imaging is done in collaboration with Lars-Olof Wahlund and Eric Westman, Karolinska Institute.

The CT-scans done until 2000 were stored as X-ray films. We have now digitised our old films (N=1576). Lars-Olof Wahlund, Karolinska Institute, apply automated brain image segmentation algorithms on all CT images, compare the resulting brain volume against manual delineation, and assess the performance of manual and automated segmentation from CT to MRI. Examinations for Normal Pressure Hydrocephalus (NPH) on CT is done by prof Carsten Wikkelsö and Daniel Jaraj, Dept of Neurology, which resulted in the first publication on the prevalence of NPH in elderly populations.

Cerebrospinal fluid (CSF) is measured in the population studies since 1986 (total N=750). New LPs were done in the cohort born 1944 at age 70 (N=323), and in those born 1930 at age 85 (N=50). LPs will be done at the follow-up at age 75 (N=200) ), 88 (N=50) and 90 years (N=25). This will enable us to examine age-related changes in CSF-levels in population-based samples followed longitudinally, to detect preclinical Alzheimer’s disease, and to test early markers for dementia in relation to CSF markers, such as sAPPalpha, sAPPbeta, Aβ38, Aβ40, Aβ42, P-tau181, Ptau231, total tau, neurofilament light, H-FABP, ViLiP-1, SNAP-25, dendritic protein neurogranin, CSF/serum albumin ratio, MBP, cytokines, YKL-40 and chitotriosidase, and to study risk factors for pathological CSF markers. We will also use a novel ultra-sensitive analytical technique, called single molecule digital ELISA for serum samples (responsible for CSF are Kaj Blennow, Henrik Zetterberg, Silke Kern). In relation to early markers for Alzheimer’s disease, we have also initiated a collaboration with Jansen Prevention in Netherlands.

Diagnoses. Dementia, depression, psychotic disorders, sleep disorders, obsessive compulsive disorder, anxiety disorders (social phobia, specific phobia, panic disorder, generalized anxiety disorder) and post-traumatic stress disorder are diagnosed according to DSM-III-R, DSM-IV or DSM-5. Alzheimer's disease is diagnosed according to NINCDS-ADRDA-criteria. Vascular dementia is diagnosed similar to NINDS-AIREN-criteria. Somatic disorders are classified according to established criteria, using information from examinations, close informants, and hospital register data.

Medical records from all major hospitals and geriatric and psychiatric institutions and outpatient services in Gothenburg are examined, and the Swedish Hospital Discharge Register is used to find cases of mental and somatic disorders.

Register data

We will take advantage of databases in Sweden, including Statistics Sweden (socioeconomic and demografic data, the LISA register), the National Board of Health and Wellfare (the Cancer register, the Hospital Discharge Register, the Causes of Death Register, the pharmaceutical drug register, Care and social services), the Swedish Tax Agency’s register, Region Västra Götalands´ health care register (VEGA), the Primary Care Register (QRegPV), National Swedish Quality Registers (Stroke, Dementia, Diabetes, Fracture, Hip fracture, Cardiovascular, Heart failure, Blood pressure) and for men, the compulsory military service register, which includes data from the drafting of 18-yar-old Swedish men.

Statistical analyses

We have four statisticians (Erik Joas, Kristoffer Bäckman, Yadi Nejad, Valter Sundh). Cross sectional analyses are performed using standard regression techniques, such as logistic regression for dichotomous outcomes or linear regression models for continuous outcomes. In cohort analyses, we include birth cohort as a covariate to analyze whether cohort has an association with the outcome. In analyses of associations, birth cohort is used as an interaction term to analyze if the effect of an independent variable differ by cohort. Longitudinal data of different birth cohorts will be analyzed with linear mixed models if the outcome is a continuous variable or generalized linear mixed models if dichotomous. These models will account for the non-independence of repeated measurements and the heterogeneity of the sample. In these models, cohort can be included both as a covariate and as an interaction factor. Furthermore, time dependent covariates can be entered as explanatory variables which are allowed to change over time. As an example, early life trauma does not change over time, but mental health and cognitive function might. Interaction covariates could also be specified, e.g. the association between cognitive function and functional ability could differ by age. These models can be employed to cover many potential research questions such as accumulation of disability over time and how other factors are changing simultaneously within the individual. For time to event analysis, we use the cox regression model. Potential effect modifiers can be included as interaction terms, e.g. to answer the question if several affected health areas have only additive effects on the hazard of death or if it also leads to multiplicative effects. For analyses of causal pathways, crossed lagged panel regression models (belonging to the structural equation modelling-SEM family) can be used. For example, it is of interest to see if multimorbidity precedes mental disorder or if mental disorder precedes multimorbidity. These models provide a mechanism to answer the question if a change in a given variable precedes change in another variable and give quantifications of these effects. These models may provide answers to the question of casual ordering. We will adjust for non-response using inverse probability weighting.

Qualitative studies

To further deepening our understanding of psychiatric disorders in older people, we will perform qualitative studies embedded in the population-studies, e.g. selecting persons with a history of depression, or with multimorbidity. Self-rated health assesses unknown perceptions, and weights them according to beliefs and preferences. There is an independent association between self-rated health and mortality that persists when other health indicators and relevant covariates are included. Few epidemiological studies combine the insider and outsider perspective in their estimates of health in old age. We will combine quantitative methods with qualitative descriptions of meaning and understanding using mixed designs (PI Hanna Falk, Synneve Dahlin-Ivanoff). The ranking of health outcomes according to personal priorities are based on individual explanatory models of illness and health. Purposeful samples will be identified in the 1923-1924, 1930, and 1944 cohorts. Qualitative data will be obtained through inductive individual interviews as well as focus group interviews organized according to the Explanatory Model Interview Catalogue (EMIC). Using two different modes of interviewing takes advantage of both small group discussions and processes and individual illness narratives, to enhance understanding of the relationships between biomedical models and personal experiences. Qualitative interviews are currently performed with participants in the H70-study to increase our knowledge regarding the experience of taking part in the studies, and to deepen our understanding of depression.

Time plan

2018: Follow-up of the birth cohort born 1930 at age 88 years, the cohort born 1923-24 at age 95 years, and the PPSW after 50 years.

2019: Follow-up of the birth cohort born 1944 at age 75 years.

2020: Continued examinations of the cohort born 1944 at age 75 years, and of those born 1930 at age 90

THE RESEARCH GROUP

The Neuropsychiatric Epidemiology Research GroupSTRONG (Epinep; home page www.epinep.gu.se) is led by Ingmar Skoog and includes 44 person: 22 researchers with a PhD (3 professors, 5 associate professors, 14 post-docs), 11 PhD students, 4 statisticians and database managers, 5 research nurses, and two research administrators. We also have master students at a regular basis from different programs at the University of Gothenburg, as well as from other national universities. Five researchers are specialists in psychiatry (Ingmar Skoog, Silke Kern, Simona Sacuiu, Svante Östling, Anne Börjesson-Hanson); four are MD at ST-level (Robert Sigström, Xinxin Guo, Mats Ribbe, Isak Fréden Klenfelt) and two are AT (Johan Nilsson, Jenna Al-Najjar). The group also provides an infrastructure with locations and computers.

INTERNATIONAL AND NATIONAL COLLABORATIONS

International collaborations

The European Union’s concerted action on depressive disorders in the elderly

(EURODEP) merges data from 13 population studies in Europe to study mental disorders in elderly populations.

The 21 st Century EURODEM, coordinated by Carol Brayne, University of Cambridge, is a collaboration between the 12 leading population studies on dementia in Europe. Our study was one of eight included in an application to EU:s JPND call.

International Consortium of Centenarian-Dementia studies is a worldwide collaboration among 13 studies on dementia in centenarians coordinated by Perminder Sachdev, University of Melbourn Australia.

MELODEM is an international consortium led from Bordeaux, with the aim to improve statistical methods in epidemiological studies.

The Harvard Alzheimer Consortium includes 12 North-American and European studies, and is led by prof Albert Hoffman, Harvard School of Medicine.

ROADMAP is an EU-financed IMI consortium led by John Gallacher, Oxford University, including university and commercial participants from EU with the aim to study real world outcomes across the Alzheimer spectrum.

John Hardy, University College of London, is processing the genetic analyses.

Grants from Forte and VR support collaboration withEM[ the]10/66 Study (led by Martin Prince, Institute of Social Psychiatry and Epidemiology, London), and the ESPRIT study (led by Karen Ritchie, Montpellier University).

We also have collaborations with researchers from Johns Hopkins University in Baltimore; Ludwig-Maximilians-University, Munich, Germany; University of Florence, Italy; University of Vienna, Austria; Rochester University, USA; Brigham Young University, Provo, Utah, USA; Dalhuise University, Halifax, Canada; and University of Leipzig.

National collaborations

The PI is director of the Forte-Centre for Ageing and Health (AgeCap; www.agecap.gu.se) at the University of Gothenburg, in which the Epinep Research Group, and this project, is embedded. The center provides multidisciplinary collaborations, and includes 5 faculties, 15 institutions and more than 100 researchers. Of special relevance for the present project are Neurochemistry (Kaj Blennow, Henrik Zetterberg), Dept of Psychology (Boo Johansson), Dept of Sociology (Lotta Dellve, Björn Halleröd), Rehabilitation (Synneve Dahlin-Ivanoff), Historical science (Helene Castenbrandt, Ulrika Lagerlöf Nilsson), and Journalism (Maria Edström, Annika Bergström). We also collaborate with Lars-Olof Wahlund and Eric Westman, Karolinska Institute, on brain imaging studies. Joint projects are done with Laura Fratiglioni at Karolinska Institute. We also collaborate with ophtalmology (Madeleine Zetterberg), audiology (Ulf Rosenhall), nutrition (Elisabeth Rothenberg), neurology (Carsten Wikkelsö), rheumatology (Mats Dehlin), medicine (Kjell Torén).

PRELIMINARY RESULTS

An on-going project like this has by default a large number of preliminary results, and the already published results support the feasability of the project. Some unpublished results could be highlighted as they point to some of the areas that we will persude in more details during the next project time. One is the finding that low CSF beta-amyloid-42 levels preceed dementia onset by up to 18 years, and that high CSF tau-levels preceed dementia by up to 8 years. Furthermore, CSF beta-amyloid levels above the median almost excluded a diagnosis of Alzheimer’s disease between ages 70 and 90 years. Thus, low levels of CSF beta-amyloid were necessary, but not sufficent, to develop AD. Furthermore, in 70-year-olds born 1944, 26% of those without dementia had pathological CSF-amyloid-42, and 44% had some AD pathology (pathological beta-amyloid-42, total-tau or phospho-tau) in CSF. We will elaborate further on these preliminary findings by examining biological markers of AD longitudinally in 70-year-olds free from dementia who will be followed up at age 75 2019, and by studying risk factors for dementia in those in the lower 50-percentile of CSF beta amyloid (and thus at increased risk for AD). The other important preliminary finding is that 44% of women in PPSW followed over 42 years had major depressive syndrome at least once, and 70% had some clinically relevant depression. The diagnoses were made by experienced psychiatrists. The high life-time prevalence may explain why it is difficult to find genetic causes for depression, as the control groups will include a large proportion with previous unreported depression. The large number of life-time depressions in our study makes it possible to study the natural course of the disorder over five decades, to examine true cases of late-onset depression and to examine social and biological origins of depression over the life-span.

SIGNIFICANCE (INCLUDING CLINICAL)

Older persons make up the fastest growing segment of the population, and constitute a majority of patients in the health care system. The project will increase our knowledge about mental and somatic disorders in the elderly by a unique combination of epidemiological and neurobiological methods. It will increase our knowledge on how biological and psychosocial factors interact in the development of dementia, depression and other mental disorders. The new examinations with MRI, PET, and the new types of neurochemical and genetic analyses, will further enhance the value of the studies, and deepen our understanding of late-life mental disorders. The follow-up at age 75 of the 1944 cohort will give a unique possibility to study change in biological markers, clinical picture, social and psychological circumstances in relation to cognitive change, and onset of dementia and other psychiatric disorders, as well as some somatic disorders. The studies will also give an unique opportunity to study secular trends in mental and somatic disorders over four decades. The number of individuals examined after age 95 is the largest published, and gives a possibility to examine genes related to dementia-free survival. Knowledge about psychiatric and somatic disorders is scarce in this age group. The detailed longitudinal examinations of Alzheimer processes that occur decades before clinical onset of the disease are unique, and have great potential for scientific break-throughs. The multidisciplinary approach examining biological, sociological, psychological, medical and genetic factors in longitudinally followed populations will deepen our understanding of ageing and mental disorders. The project has clinical relevance in relation to prevention, early diagnosis, clinical course, experience of illness, understanding pathogenesis and prognosis, and may enhance the quality of care of older persons. Results will easily translate into clinical practice due to the location of the research group at the memory clinic and the old age psychiatry clinic, and as several of the researchers work part-time as doctors or nurses at these clinics.

References

1) Skoog et al. Lancet 1996;347:1141-1145.

2) Gustafson D et al. Arch Intern Med. 2003;163:1524-8.

3) Johansson UP L, et al. Brain 2010;133(Pt 8):2217-24.

4) Andersson M, et al. Age Ageing 2012;41:529-33

5) Skoog I. Psychiatric disorders in the elderly. Can J Psychiatry 2011;56:387-97

6) Sobel RM, Markov D. Curr Psychiatry Rep 2005;7:206-12.

7) Linden L, et al. Stroke 2007;38:1860-3.

8) Gudmundsson P, et al. Eur J Epidemiol 2011;26(1):55-60

9) Martens et al. Arch Gen Psychiatry 2010;67:750-58

10) Marmot M. Status Syndrome: How Your Social Standing Directly Affects Your Health and Life Expectancy. London: Bloomsburry Publishing 2004

11) Jack CR, Jr., et al. Lancet Neurol. 2013;12(2):207-16.

12) Gudmundsson P, et al. Eur J Neurol. 2015;22(5):781.

13) Song et al. Annals of Internal Medicine. 2004;141(2):137-147.

14) Skoog I, et al. Biol Psychiatry 2015 Nov 15;78(10):730-6.

15) Hodes G. Biol Sex Diff 2013; Pubmed doi:10.1186/2042-6410-4-1

16) Wiberg P, et al. Psychol Med 2013; 43:2627-2634. doi:10.1017

17) Zhi X, et al. Aging Clin Exp Res 2013;25:377-83

18) Sacuiu et al. Neurology 2010;75:779-85.

19) Lak et al. Age Ageing. 2012;41:735-40.

20) Bäckman K,et al. J Gerontol A Biol Sci Med Sci. 2016 Aug 13.[Epub ahead


Psykisk och kognitiv hälsa i relation till funktionsförmåga och somatisk hälsa hos äldre. Påverkan av ålder, kön, födelsekohort, genetiska, sociala och neurobiologiska faktorer, from FoU i Västra Götalandsregionen
http://www.researchweb.org/is/vgr/project/240461