Data project

95+

95+

Summary

Strengths 1 One of the strengths of research based on these data sets is its interdisciplinary character. The EpiLife Center have principal investigators and co-investigators from the following public health-related disciplines: Epidemiology, Psychiatry, Cardiovascular Medicine, Social medicine, Primary health care, Geriatrics & gerontology, Statistics, Nutrition & metabolism, Genetics & molecular epidemiology, Psychology, Education, Sociology, Cancer epidemiology, and Caring sciences. There are also extended links with other research groups throughout Sweden, and through providing a basis to extend our international collaborations. In addition to the EpiLife team, visiting researchers have been collaborating on several of the population studies. ‘The research teams are involved in international research collaborations with studies involving all age groups, from the very young to the very old.’ All samples are systematically obtained, based on birth dates, from the Swedish Population Register, which covers names and addresses of all people living in Sweden. The studies include both persons living in private households as well as in institutions. Response rates 60-85%. Strengths 2 (See http://ageing.oxfordjournals.org/content/41/4/529.full.pdf+html) Strengths of the data set are the population-based design, the large sample, the comprehensive examinations, the use of several information sources, and that all diagnoses were made by geriatric neuropsychiatrists. Limitations Some possible limitations should also be considered. First, the validity of the information sources could be questioned. However, information from self-reports and key informants was evaluated by neuropsychiatrists. Furthermore, criteria for dementia and stroke were strict, allowing only cases with a clear history of focal symptoms for a stroke diagnosis and requiring information from both neuropsychiatric examinations and key informant interviews for a dementia diagnosis. Second, it is possible that we underestimated the prevalence of stroke/TIA, as only 450 out of 591 individuals had key informant interviews. Third, information from the hospital discharge register may be questioned since diagnoses were made by many different physicians working under different circumstances. However, most cases were diagnosed by neurologists or internists at the university hospital in Gothenburg. Fourth, the response rate was 65%, a fairly satisfactory response rate in this age group.

Type of data

Data Source
Registry + Survey

Type of Study
Survey same
Survey different
Crosssection regular
Cohort study

Data gathering method
Telephone
Face-to-face
Registries
Self administered questionnaire
Other: Physical examinations, cognitive testing

Access to data

Conditions of access
After contact with responsible research group, data can be analysed at site and at times as unidentifiable files.

Type of available data (e.g. anonymised microdata, aggregated tables, etc.)
anonymised

Formats available
Primarily SAS and SPSS, but the data set is compatible and other format can be used.

Coverage

Coverage Years of collection, reference years, and sample sizes
Cohort 1 1901-1903: We examined 95-year-olds born in 1901-1903 (N=338, response rate 65%) beginning in 1996. Follow-ups were conducted at ages 97 and 99 years, when the examination was enlarged to also include individuals born between 1904 and 1909. Yearly follow-ups were conducted from age 99 and onwards. The examination of 97-years-olds born from 1901 to 1909 was finished in 2007 (N=591; 107 men, 484 women), of 99-year-olds in 2010 (N=348; 295 women, 53 men) and of 100-year-olds in 2011 (N=245; 209 women, 36 men; including also individuals born 1910-11). In total, we will also examine 150 101-year-olds (so far 135 individuals), 120 102-year-olds (so far 80 individuals), 60 103-year-olds (so far 39 individuals), 30 104-year-olds (so far 18 individuals), 20 105-year-olds (so far 12 individuals) and 11 106-year-olds (so far 5 individuals). One person has follow-up from age 95 to 109 years. The response rate is around 70%. In total, we have examined 950 persons in the study.

First year of collection
1996

Stratification if applicable
Yes – in relation to socio-economic status, education and living conditions and general medical conditions, age, gender and region

Base used for sampling

Geographical coverage and breakdowns
Gothenburg

Age range
Individuals aged 95-109 years

Statistical representativeness
Population representative

Coverage of main and cross-cutting topics
The 95 study is the largest population study in the world on individuals above age 95. More than 800 individuals have been included so far. The aim of this longitudinal study is to examine the frequency of mental disorders in very old age (the fastest growing segment of Western populations) and to examine determinants of mentally healthy survival. All samples are systematically obtained, based on birth dates, from the Swedish Population Register, which covers names and addresses of all people living in Sweden. The studies include both persons living in private households as well as in institutions. Response rates are between 60-85%. Examinations have been virtually identical at each examination year to enhance possibilities of comparisons, and include psychiatric examinations, key informant interviews, physical examinations, nurse examinations, and examinations by physiotherapists, and psychometric testings. We also examine ADL (activities of daily living) and IADL (instrumental activities of daily living), cognitive and emotional function, physical activity, sensory functions, social function (social network) and psychiatric function, as well as chronic diseases. Data on socioeconomic status, marital status, living conditions, housing, social network and other social circumstances, education, religious activity, hobbies and life events are collected. Neurobiological examinations include CT scans of the head, and CSF examinations. Blood sampling is also included, and includes genetic testings of frozen blood samples. This is led by Prof. Kaj Blennow. We process whole frozen blood for DNA using standard procedures. We collaborate with Professor John Hardy, London, in a genome-wide association study (GWAS), which focuses on associations between single-nucleotide polymorphisms (SNPs) and traits like major diseases. The psychiatric examinations were performed by trained psychiatrists at an outpatient department, in the subject's home or at an institution until 2000. Since then, the examinations are performed by experienced psychiatric research nurses. To minimize the methodological problems with this approach, we perform extensive training programs. The inter-rater reliability tests between nurses and specialists in psychiatry is high, with Kappa values for the presence versus absence of signs and symptoms of depression and dementia between 0.62 and 1.00 indicating “good or excellent” agreement. The psychiatric examinations are semi-structured and include ratings of psychiatric symptoms and signs with the Comprehensive Psychopathological Rating Scale, the Mini-D, and questions about previous mental disorders, sexuality, sleep, and use of psychotropic drugs. Suicidal feelings are rated according to Paykel. The examination also includes tests of mental functioning (e.g. memory, proverbs, language, apraxia, construction, finger agnosia, agraphia, alexia, acalculia, and right-left disorientation), the Mini Mental State Examination, and ratings of other signs common in dementia (e.g. personality changes and motor symptoms). Personality inventories (MNT, CMPS och EPI) are also included. Somatic examinations include history of medical conditions (e.g. hypertension, myocardial infarction, diabetes mellitus, stroke, cardiac failure, atrial fibrillation, pulmonary diseases, fractures, thyroid diseases, head trauma, neurological disorders, cancer), use of health services and questions about urinary function, falls, hearing and vision, dizziness, alcohol and tobacco use. Family history of several psychiatric and somatic disorders is also included. The prescribed and actually taken medication is recorded and classified according to the Anatomical Therapeutic Chemical (ATC) classification system recommended by the World Health Organization. The examinations also include anthropometric measurements, blood pressure, electrocardiogram (ECG), lung function, measures of atherosclerosis, a battery of blood tests, grip strengths and walking capacity. Blood samples are taken after overnight fasting, and include a variety of measurements such as hemoglobin, HbA1c, thyroid function, cholesterol (total, HDL, LDL, triglycerides), homocystein, B12 and folate. Blood, serum and plasma are frozen for future analyses. Psychological examinations regarding memory and intelligence are performed in all examinations. Responsible for these examinations is Professor Boo Johanson, Dept. of Psychology, Göteborg University. Measures include SRB-1, SRB-2, SRB-3, Thurstone Picture Memory, Ten Word Memory List, Clock Test, Prose Recall, Digit Span Forward and Backward. Functional ability: Data is collected on basic abilities in activities of daily living (ADL), instrumental ADL and other abilities, and physical activities, such as walking and exercising. Social factors: Data on socioeconomic status, marital status, living conditions, housing, social network and other social circumstances, education, religious activity, hobbies and life events are collected. Close informant interview. After the examination, the participant is asked for permission to interview a close informant. The interview with a close informant is semi-structured and comprises IQCODE (Jorm), other questions about changes in behavior and intellectual function (e.g. changes in personality, memory, difficulties in finding one’s way in familiar surroundings, intellectual ability, language, psychiatric symptoms, activities of daily living, incontinence, neurological symptoms), family history, medical history (e.g. hypertension, myocardial infaction, diabetes mellitus, stroke, cardiac failure, atrial fibrillation, fractures, thyroid diseases, head trauma, neurological disorders, cancer) and, in case of dementia, questions about age at onset and course. Psychiatric symptoms include e.g. hallucinations, delusions and depression.

Linkage

Standardisation
The Diagnostic and Statistical Manual of Mental Disorders (DSM by American Psychiatric Association and other medical taxonomies. ADL-models also used.

Possibility of linkage among databases
Good technical opportunities to link to other Swedish datasets or registries due to the use of ID – personal identification number, but policies of data protection and ethical considerations have to be regarded.

Data quality

Entry errors if applicable
Varying response rate from 65 to 85%.

Breaks
Over the years the procedure has been retained as identical as possible. Participants have been invited to study and offered a health control. The investigation comprised a physical examination, blood and urine tests, ECG, impedance measurement and the testing of visual accuracy and hearing. Additionally, in subgroups psychological and psychiatric examinations, audiometric tests, analyses of body composition, oral health, dietary interviews, movement analyses and muscular ability tests were carried out.

Consistency of terminology or coding used during collection
High ambition to be consistent in terminology and coding, but there might be some variation due to person factors in coding.

Governance

Contact information
Ingmar Skoog
Dept of Neuroscience and Physiology
Wallinsgatan 6
431 41 Mölndal Sweden Phone: 00709-433 681
Email: ingmar.skoog(at)neuro.gu.se
Url: http://www.epilife.sahlgrenska.gu.se/cooperation/International/

Timeliness, transparency
It takes about 3-5 years until the data is released following collection.