About the study
For many years, Sweden has been one of the leading nations in work environment research. This is not largely due to the fact that we have had careful surveys of the conditions at the workplaces, e.g. through the Swedish Work Environment Surveys (SWES) that have been carried out every second year since 1989. The surveys have given a very good picture of the development of a wide range of conditions in working life. However, society is constantly changing, which means that new questions continually arise.
Associations between work environment and health have been shown in many studies. Most often, however, these are so-called cross-sectional studies, where you ask about work environment and health at the same time. You don't know which came first. Is it a bad working environment that impacts on health or is it possible that people with poor health tend to get jobs with worse working environments, or alternatively experience the working environment as worse? These questions can only be answered by following responses over time. You can then see if a certain work environment is related to how people’s health and wellbeing develop over time, or if improvements or deteriorations in the work environment predict changes in health.
Other aspects highlighted in the SLOSH are conditions outside of the workplace, for example how retirement affects health. By asking a series of questions that have not previously been asked in a nationally representative sample, we can also shed light on the existence of new or relatively little- studied conditions in the workplace, for example technostress, organizational changes and telework. SLOSH thus creates opportunities for research into the health consequences of work that are unique both in Sweden and internationally.
Since SLOSH follows participants whether they continue to work or not, SLOSH can also be used to study e.g. unemployment and ageing. In the future, we will also include younger people in order to investigate health in relation to studies, entry into the labour market and the consequences of not working or studying.
Method
The SLOSH sample consists of all people who answered a work environment survey (SWES) between 2003 and 2019. The SWES is run by the Swedish Work Environment Agency and Statistics Sweden (SCB) and is in turn based on the Labor Force Surveys (LFS), which are carried out every year. In these, randomly selected individuals of working age are asked about employment. In the sample drawn for the LFS, stratification is done by to county, sex, citizenship and employment according to the employment register. The sample for the Swedish Work Environment Survey is a sub-sample of the people who participated in the LFS and who were employed at the time of the interview.
The SLOSH questionnaires consist of detailed questions about working life, general life situation, health and well-being. Data from administrative registers on, among other things, sick leave and hospital admission are linked to the questionnaires both retrospectively (backwards in time) and prospectively (forwards in time). A list of which questions are included in the different years is provided on request to data@slosh.se.
The Research Council for Working Life and Social Sciences (FAS) was the original funder, followed by Forte, and from 2010 the Swedish Research Council (VR) is the main financier. Since 2018, SLOSH has been part of the national infrastructure REWHARD, which is financed by VR together with Stockholm University and Karolinska Institutet. Funding is currently available until the year 2028. The study is approved by the Regional Ethics Review Board in Stockholm and the Erikprövingsmyndigheten. For a more detailed description, refer to the published cohort profile.
Advantages and limitations
One of the major advantages of SLOSH is that it is based on a large, nationally representative sample of the working population. The large sample makes it possible to draw conclusions that apply to the general working population, but also to study certain groups, e.g. people who work in human services professions. However, the greatest advantage of the SLOSH study lies in its longitudinal design. Following the same people with similar questions over time provides the opportunity to analyze not only correlations, but also causality. Something that cannot be done in a cross-sectional study like SWES.
The longitudinal design has another very important advantage – it becomes possible to study the health consequences of changes as well. In today's mobile working life, where few people can expect to have the same job for the rest of their lives, these changes can have at least as large consequences as the working environment in an individual workplace. Considering the aging population, questions about retirement and possibilities for an extended working life become particularly interesting.
The opportunities to study changes also have a technical side. If you can show that a change in a certain measure has a prospective relationship with how health changes, then that is an even stronger sign of a causal relationship than if you just show that different levels at one time are related to health at a later time.
SLOSH builds on a tradition of high-quality cohort studies – notably Whitehall II, GAZEL and the Ten Town/Finnish Public Sector Study – but is unique in that it is based on a nationally representative sample followed over a long time period. SLOSH is a valuable base for a large number of ongoing and planned studies. Its value for research will increase even more over time through new data collections, refreshments with new individuals and an increased follow-up time.
Table. Overview of the development of the SLOSH cohort including response rates.
Wave | SWES year(s) | Culmulative cohort size | No. of contacted persons* | No. of responses registered | Response frequency |
2006 | 2003 | 9 214 | 9 149 | 5 985 | 65.4% |
2008 | 2003-2005 | 18 917 | 18 734 | 11 441 | 61.1% |
2010 | 2003-2005** | 21 489 | 20 291 | 11 525 | 56.8% |
2012 | 2003-2005 | 18 917 | 17 409 | 9 880 | 56.8% |
2014 | 2003-2011 | 40 877 | 38 657 | 20 316 | 52.5% |
2016 | 2003-2011 | 40 877 | 38 012 | 19 360 | 50.9% |
2018 | 2003-2011 | 40 877 | 37 043 | 17 841 | 48.2% |
2020 | 2003-2011 | 40 877 | 35 700 | 17 489 | 49.0% |
2022 | 2003-2019 | 57 105 | 51 412 | 21 867 | 42.5% |
2023 | 2003-2019 | 57 104*** | 43.7% |
* After exclusion of dead and emigrated persons as well as those who have actively opted out.
** Also including SWES 2007 respondents from Stockholm and Gothenburg.
***After exclusion of former participants who have demanded deletion of all their data.