Participants and study groups
Flow of participants throughout the study is depicted in Figure . In collaboration with a White-Collar Union (Sif) and a Swedish Employers' Association (Almega), ten companies insured by the study's source of funding Alecta (an occupational pension plan company) were asked as to their interest in participating. The asked companies were selected and contacted by employees at Alecta, by mail and phone. The management departments of six out of the ten asked companies were interested. Informed of the basic inclusion criteria, i.e. minimum group of ten individuals and access to economic production data, 2–4 departments within each company were chosen and asked by the company management as to their interest in participating. The managers of the selected departments in turn asked their employees whether they were interested in participating. No incentives were offered to the participants, with exception of the extensive blood sampling including feedback of the results, which seemed to be a motivator for many participants.
Flow of participants. The figure illustrates the flow of participants through each stage of the randomized trial. Additionally, the numbers of excluded participants and drop-out rates, including generalized reasons for these actions are depicted.
For some companies the departments were located in different cities and for some in the same buildings or city. With the exception for one of the media companies where a whole department with five units enrolled, there was no "natural" connection between the participating departments. Consequently, there was only occasional risk of contamination of the extended intervention to the reference group. Moreover, the design of the websites for the intervention and reference group respectively was similar in appearance (see more information on the interventions below). The intervention group website only had two additional buttons, which makes it hard to notice any difference if the site would be exposed to a participant of the reference group.
The participating departments were, within each company, randomized by lottery to either the intervention or reference group. Thus each company had at least one intervention and one reference group. All departments received a 30-minute information session including 10 minutes for questions and answers. These information sessions included the aim of and general information about the study as well as general information on stress and health. Finally, after oral information each participant received written information about the project and consent forms. All the participants were informed, orally as well as in writing, that participation was voluntary and withdrawal was possible at any time.
There is no information on the exact number of employees that were asked to participate in the study. An exception was one of the media companies where 95 out of 100 possible participants chose to participate. In general there was also a great interest from the other departments and similar participation rates is therefore estimated. Altogether, 317 participants from 22 departments/units in four information technology and two media companies enrolled in the study. Fourteen participants were excluded because of communication-related problems (n = 7), change of mind to participate or quit their job before initiation of the study (n = 7). Thus 303 persons finally participated in the study, out of which 26 participants (8.6%) dropped out. The reasons for dropping out were job termination (n = 7), change of workplace (n = 2), foreign service or moving abroad (n = 6) or other reasons (n = 11). There were no significant differences in dropout rates between the groups (6.9% in the intervention group vs. 9.8% in the reference group, p between groups = n.s.). Nor were there any significant differences between the intervention and reference groups in socioeconomic background or psychophysiological measures at baseline. Regarding the participants, there was no information about possible mental or somatic disorders or medication.
The participants had professions such as IT technicians, programmers, system developers as well as journalists/reporters, news presenters, sound technicians and photographers. The main type of work-site was open plan offices. Many participants from the IT-companies were partly located in the work sites of their customers for longer or shorter periods. For the media companies, some participants, such as photographers and reporters, were partially ambulatory and worked in different locations. The common feature for all participants was regular and daily computer usage at work.
The web-based tool
Table provides a detailed description of the web-based tool and illustrates similarities and differences in the features that were offered to the intervention and reference group respectively. A web-based tool for health promotion and stress management was developed and offered all participants real-time monitoring of perceived current health and stress status, a diary and information about stress and health (Table ). In addition, participants in the intervention group were offered web-based cognitive exercises, aimed at decreasing unwanted stress and promoting health and recovery through health promotion initiatives. The exercises included techniques for relaxation, time management, cognitive reframing and a chat. Thus, the only things that distinguished the groups were the addition of the cognitive exercises and the chat in the intervention group. The web-based tool was developed by the researchers and most techniques are commonly utilized techniques in cognitive and behavioral psychology and stress management. These techniques were modified so that they could become more or less self-instructing to be used for self-help purposes.
Table 1 The web-based tools. The table depicts the different features included in the web-based tool for the study groups respectively. The only things that distinguished the groups were the addition of the cognitive exercises and the chat in the intervention (more ...)
The web-based tool and the exercises were not pilot tested before the study. However, the tool as well as exercises were chosen and adjusted on the basis that they had to be time efficient in order to be utilized. It was hypothesized that basic demands for regular usage were instant feedback on the questionnaire and that the measurement and exercises could be used rapidly. Consequently, it was decided that regular or daily monitoring should not take more than 20–40 seconds. Moreover, every exercise was labeled with information of time for accomplishment (time span 1–60 minutes). Some of the cognitive exercises, e.g. improving self-confidence, were designed such that they consumed 5–10 minutes when learning and then could be conducted in a matter of seconds when utilizing.
Most exercises were presented in three different modes; on the web-site as plain text, as a downloadable PDF-file (sometimes including descriptive images), and as a flash animation, guiding the participant with image and sound through the exercise. Since the intervention for both groups was completely web-based it could only be accessed online. All information was however printable, which made it possible for the participants to print material of interest and thus intervene elsewhere. Exposure to the intervention for both groups could only be logged via the number of logins to the website.
A questionnaire was compiled and included about 100 questions concerning socioeconomic status, consumption of caffeine drinks, expectations about the research project, self-rated health (SRH), stress and wellbeing at work as well as during leisure time, health economics and performance at work (Table ). Most of the questions were presented as Visual Analogue Scales (VAS) and some, concerning health economy, work time, basic daily functioning and symptoms of ill health, were presented as multiple-choice questions. Most of the newly constructed single VAS questions were based on previously validated Likert-based items or indices [20
]. Participants filled out the questionnaire online at baseline (before the initiation of the study) and at the end of the six-month intervention.
Questionnaire. The table illustrates theoretical models, items and topics covered by the questionnaire. Most items were presented as "straight forward" VAS, e.g. How is your overall sleep quality (Very poor – Very good).
The complete list of biological markers analyzed in the current study is presented in Table . More biological markers of general nature, such as blood status, were collected for overall health matters or all-purpose profiling. These markers were not analyzed in the present study. Furthermore, P-substance P, S-IL-1beta and P-endothelin were also collected. However, in the first measurement there was not enough blood collected to render the exact results needed for more sensitive analyses of these variables, resulting in the decision to not include them in the present study. Thus, the biological markers analyzed in the present study were only the ones that could be related to various stress-related hypotheses.
Blood sampling and physiological measures. The table illustrates the biological markers and physiological measures sampled at baseline and after the six months intervention.
Blood samples were collected from study participants between 7.00–11.30 am at each specific worksite (or nearby). Unfortunately, it was impossible for practical reasons to sample the blood within more narrow time limits. Questionnaires were filled out during the same time period (usually same day or week) in order for the outcome of the blood and questionnaire data to be as comparable as possible. The exact time for blood sampling was recorded for each participant at baseline and at the end of the study so that the blood could be collected at the same time (± 15 minutes). Participants were instructed not to eat or drink (except water), nor use nicotinic substances at least ten hours before blood sampling. The blood samples were analyzed by the Karolinska University Hospital laboratory that is qualified by SWEDAC (Swedish Board for Accreditation and Conformity Assessment) that accredits laboratories in the medical sector according to the standard ISO/IEC 17025. Intra-assay and inter-assay coefficients of variation can be obtained from the laboratory peter.matha/at/karolinska.se) or by e-mailing the authors dan.hasson/at/pubcare.uu.se.
The program SPSS 11.5 for windows was used for statistical analyses and an intention-to-treat approach was utilized. This means that all subjects in both groups were included in the follow-up regardless of how much they participated in the intervention programs. And the evaluation is based upon the assumption that everybody – even those who did not participate at all – in the intervention group were compared with everybody in the control group.
Initially, all variables were assessed for normality using Kolmogorov-Smirnov test. Changes over time (time, group and group × time) were assessed using two-way analysis of covariance (ANCOVA). ANCOVA adjusts for initial differences so that the results more precisely reflect possible intervention effects, and thus permits a more sensitive analysis compared to regular analysis of variance (ANOVA). The increase in sensitivity arises from the fact that the covariance reduces the error term (within-group variability) against which intervention effects are compared. Furthermore, ANCOVA is not very sensitive to small deviations from a normal distribution [26
]. In the present study, baseline values of the assessed variables were used as covariates. Analyses were in some cases, such as sex hormones stratified with a break up by gender. Finally, to adjust all results of the ANCOVA analyses for the possible effect of multiple comparisons (mass-significance), Bonferroni correction was utilized for each analysis.
Since VAS can be treated as an interval or ordinal scale [27
] and all variables were not normally distributed, both parametric and non-parametric tests were used where statistically significant differences between the groups were detected in the ANCOVA. Thus, it was decided that changes over time and differences between the groups would only be considered in cases where both parametric and non-parametric tests unanimously were statistically significant. For the non-parametric analyses, new variables (so called Δ variables) based on change between the first and second measurements were constructed. Differences between the groups were then assessed using a Mann-Whitney U test.
Logistic regression was used to model the probability of improvement in the significant Δ variables (dependent variables). Factors such as socioeconomic status, marital status and gender are known to be associated with outcomes in health and stress and were therefore included as covariates in the first step of the regression analysis. Also group was included as a factor in the first step to adjust for possible study group effects. The dependent and independent VAS variables used in the logistic regression were divided by quartile split into high (top quartile) and low (remaining quartiles) categories. Similarly the number of logins was dichotomized by quartile split. The independent VAS variables were selected for two subsequent steps in the regression analysis. The rationale for the second step was to adjust for work related factors that might disturb the relationships, i.e. working hours per week, working atmosphere, work intensity and number of breaks during a working day. The third step included all the remaining dependent variables. In order for the physiological markers to render comparable odds ratio they were dichotomized by quartile split. The fourth and final step included the number of logins to adjust for possible effects of high vs. low frequency of logins to the website.
The specific hypothesis tested in the present study was that the intervention group would improve compared to reference group on biological stress markers and health- and recovery-related ratings captured by the questionnaire. Decreased levels in indicators of catabolism and increased levels in indicators of anabolism were expected in the intervention group compared to the reference group. Since both groups received an intervention, some beneficial changes in the reference group, e.g. in SRH, might be expected as well.
The ethics committees of Uppsala University (Dnr 01–188) and Karolinska Institute (Dnr 01355) approved the research project. A modified version of the web-based tool used in the present study can be found at http://www.pql.se
Role of the funding source
The funding source had no involvement in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the paper for publication.