Swedish Panel Study of Living Conditions of the Oldest Old (SWEOLD)

Summary

Strengths: SWEOLD is based on a nationally representative study population (sample includes all despite living situation and health condition, even elderly persons living in institutions). The data contains high quality data and has a high response rate. We have a broad multidimensional approach of special importance to elderly people, with data covering individuals’ command over resources in terms of family and social relations, material living conditions, health, political life, leisure time activities, housing conditions, etc. The follow-up allows investigating long-term trajectories of change in health and other living conditions, as well as the impact of different exposures, during different life stages, on health and other living conditions in later life. Hence, SWEOLD provides a rich and multi-purpose longitudinal database for the scientific community. Weaknesses: A small sample size, (although complemented with an additional representative sample in SWEOLD 2011). Intervals of six to ten years between measurement waves are too long for trajectories of health decline, social and behavioral changes to be identified, especially end-of-life changes. SWEOLD has a tradition of a broad multi-dimensional approach with a rich amount of sociological data, but the drawback is the low ability to study specific health problems in detail, such as dementia diagnosis, depression, etc.

Type of data

Data Source
Registry + Survey

Type of Study
Survey same
Survey different

Data gathering method
Face-to-face
Registries

Access to data

Conditions of access
The data are available on request for social and epidemiological research. Data can only be provided after ethical testing and approval by the Stockholm Ethical Review Board. Information of the study is available at the website www.sweold.se. There are no costs associated with data providing. DVD/CD after agreement.

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

Formats available
All common applications are used and the data set is compatible to ASCII, CSV, Excel, SAS, SPSS, STATA.

Coverage

Coverage Years of collection, reference years, and sample sizes
Collection year: 1992, Reference years: 1977-98, Sample size: 537 individuals Collection year: 2002, Reference years: 1977-99, Sample size: 621 individuals Collection year: 2004, Reference years: 1969-100, Sample size: 1,180 individuals Collection year: 2011, Reference years: 1976-101, Sample size: 931 individuals

First year of collection
The first Panel Study of Living Conditions of the Oldest Old (SWEOLD) was carried out in 1992. The sample consisted of people born between February 1892 and February 1915 who had been interviewed in the Swedish Level of Living Surveys (LNU) 1968 - 1981. Of the 1,936 people meeting these criteria, 563 were still alive in January 1992. The sample was nationally representative for the Swedish population. All but two of the individuals could be located through intensive preparatory fieldwork. A total of 95.4 per cent was interviewed. The interviews were conducted by an external provider. The purpose of the 1992 study was to follow up on people who had participated in earlier surveys in order to create a longitudinal database including the oldest old.

Stratification if applicable
Data includes variables such as age, sex, region, country of birth, etc.

Base used for sampling

Geographical coverage and breakdowns
Nationally representative for Sweden.

Age range
76 and older in general, except for 2004 when age range started at 69 and older (also see Coverage).

Statistical representativeness
Population representative

Coverage of main and cross-cutting topics
Elderly people’s living conditions The SWEOLD study was established in order to increase the knowledge about older people’s living conditions. The questionnaire covers a wide variety of areas of concern to elderly people, such as mobility and activities of daily living (ADL and IADL), as well as health, housing, economy, family, political resources and leisure activities. Various tests are included to measure the physical and cognitive ability of these individuals. The longitudinal design (containing information about earlier living conditions from previous LNU surveys) makes it possible to do a follow up later in life.

Linkage

Standardisation
Internationally harmonised standards (e.g. ADL, ISCO, NACE, ISCED).

Possibility of linkage among databases
The SWEOLD database can be linked to available registers and to The Swedish Level of Living survey (LNU).

Data quality

Entry errors if applicable
The SWEOLD survey is nationally representative and holds a very high quality of data due to the fact that it includes all people, despite living situation and health condition. All individuals were located and contacted. The field work was planned in detail and held to a high quality through the whole process. We also reached a high response rate in all surveys. The data was cleaned properly and the study was well documented.

Breaks
Even though the SWEOLD studies are based on face-to-face interviews (except for year 2004 when all interviews were conducted as telephone interviews), we use different methods for people who are not able to answer questions or refuse to meet an interviewer face-to-face. In those cases, we have used questionnaires, telephone interviews or proxy interviews. Proxy interviews are used only when a respondent is unable to participate in the interview directly due to physical or cognitive disabilities. Next of kin or professional caregivers are used as proxies. Furthermore, the data collection of the SWEOLD 2011 survey was complemented with an additional sample of 335 women and men aged 85-99. This was done to enhance the original sample and create the possibility of making detailed cross-sectional analyses among the oldest old.

Consistency of terminology or coding used during collection
In general, the data has been labelled and coded in the same way in all survey years. It is easy to compare data from different survey years due to well documented descriptions (code books). Some questions have been improved and are therefore not fully comparable over time. But we have always been careful to maintain the comparability of the parts of the questions to the extent it has been possible.

Governance

Contact information
Carin Lennartsson
Aging Research Center (ARC/KI/SU)
Gävlegatan 16
113 30 Stockholm Sweden Phone: +46 8 690 58 53
Email: carin.lennartsson(at)ki.se
Url: http://www.sweold.se/aboutsweold.htm

Timeliness, transparency
There is about one year between data collection and publication.