Lifestyle in the West project
The Lifestyle in the West (LIW) project was initiated as a joint venture between two of the largest employers in western Sweden: a manufacturing corporation representing the private sector and the regional public healthcare organization. The aim of the project was to promote healthy lifestyle behaviour among the employees. The target population for the project was in total the 65,000 employees of these two organizations. This study, however, considers only employees working in the healthcare sector. This 3.5-year WHP project, conducted between 2005 and 2008, used a broad all-inclusive approach, relying on individuals' decisions to participate instead of targeting specific at-risk individuals. The key phrases of the LIW project were: focus on the healthy (not on disease), achievement of balance (between work and leisure, and between activity and recovery), no pointing fingers (persons at risk were not specifically targeted), and individual will (all initiatives of the project were available for the individual, but not mandatory). Four lifestyle-related themes were decided on: physical activity, nutrition, sleep, and happiness/enjoyment. During a period of six to nine months, special attention was given to each theme, starting with the physical activity theme in 2005. A group of experts were appointed for each theme, responsible for initial training of 150-200 employees, of both sexes with different ages and occupations, to become local 'health coaches'. During the duration of the project, the health coaches were continuously updated, and provided with educational material (a website, PowerPoint presentations, and books) for each theme. Chief executives and union representatives also received training with the clear aim of increasing awareness on the importance of promoting a healthy lifestyle and lifestyle changes among the employees and making it possible for the employees to participate in different activities, even during working hours.
During the duration of the project, the health coaches functioned as local practical ambassadors of the project, organizing seminars, competitions, and theme-related events. Ideas for activities were provided in a project book available for the health coaches, but the activities could differ considerably between different workplaces. Educational books for each theme were produced and distributed to all employees. Examples of general activities were a pedometers competition between different departments, a cookbook with healthy recipes, environmental certification of hospital food and restaurants, and production of an educational TV show on healthy lifestyles. Furthermore, thousands of local activities were initiated during the project period.
At the time of the study, employees of the public healthcare organization numbered around 48,600 (80% women). Approximately one year before the onset of the LIW project, a questionnaire survey was posted to a random sample of 5300 employees. The sample included mainly HCWs (around 80%), others being administrative personnel working in hospitals, primary care, dental care, or central administration. An inclusion criterion was having been employed for at least one year and working 50% or more of full-time. A total of 3207 individuals (response rate 61%) consented to participate and completed the questionnaire (Table ). Seventy-three percent were working in hospitals (one large university hospital, four medium sized and eight small hospitals), 13% in primary care, 8% in dental care, and 5% in central administration offices. When comparing responders with the non-responders, women tended to respond more frequently than men and middle aged persons were slightly overrepresented (the proportion of women was 1.7% higher and the age group 45-54 years 1.3% higher compared to the random sample population). The responders of the baseline survey in 2004 were asked to participate in follow-ups to be conducted every second year. The response rate at the first follow-up in 2006 was 85%. At the second follow-up in 2008, 1971 persons responded (83%), of whom 1859 (94%) answered the question of whether they had participated in any LIW activity during the duration of the project. The study was approved by The Regional Ethical Review Board in Gothenburg and conduced in compliance with the Helsinki declaration.
Characteristics of all the healthcare workers responding at baseline (n = 3207).
The general aim of the survey was to investigate different aspects of psychosocial work environment, stress, and stress-related health in the target population, including a range of demographic factors and lifestyle parameters. At the four-year follow-up in 2008, questions about the LIW project were included, and the participants were asked to respond 'yes', 'no', or 'do not know' to the question of whether they had participated in any LIW activity. A following question was 'If yes, which theme(s) (nutrition, sleep, physical activity, and/or happiness/enjoyment) did you participate in?' Among the variables available from the baseline data, four indicators of lifestyle-related health/behaviour were identified that seemed best related to the respective theme of the LIW project. Thus, body mass index (BMI) was considered to best relate to the nutrition theme, self-reported PA to the physical activity theme, sleep disturbances to the sleep theme, and general mood (depressive or not) and feelings about work (positive or not) to the happiness/enjoyment theme. A single item on self-rated general health was also used as an indicator of a variety of experienced health problems.
Self-reported physical activity
The participants rated their PA level according to an adapted version of the widely used four-level scale originally developed by Saltin and Grimby [19
]. This simple instrument has been shown to discriminate between sedentary and active counterparts regarding maximal oxygen uptake [20
] and has been validated against biological measures [21
]. Furthermore, it has been widely used in several large epidemiological studies that show the relation between self-reported PA and morbidity as well as to mortality [22
The participants reported the level that best corresponded to their PA during the last three months: mostly sedentary (group 1), light PA (such as gardening, or walking or bicycling to work) at least two hours a week (group 2), moderate PA (such as doing aerobics, dancing, swimming, playing football, or doing heavy gardening) at least two hours a week (group 3), or vigorous PA several times a week, for at least five hours with high intensity (group 4).
Due to the fact that only 2.4% of the participants reported vigorous PA, we reduced the four categories to three distinctive groups: sedentary (group 1), light PA (group 2), and moderate-to-vigorous PA (groups 3 and 4). When dichotomized in two categories (sedentary and physically active), both the light PA and moderate-to-vigorous PA groups were placed together as the physically active group.
Two items were used to construct the sleep-disturbance variable. The participants were asked to answer how often during the past three months they had experienced: a) difficulty falling asleep and b) repeated awakenings with difficulty falling back asleep. Five different response alternatives were used for both items as follows: 1) never, 2) seldom, 3) sometimes, 4) several times a week, or 5) every day. Participants who selected alternative four or five for at least one of the two items were defined as suffering from a sleep disturbance.
The Hospital Anxiety and Depression (HAD) scale was included in the questionnaire [25
] and the subscale for depression (HAD-D) was used as an indication of presence or absence of joy/happiness at baseline. This subscale has seven items, each with four response categories. A total sum score of seven or more was defined as a relative lack of joy/happiness (depressive mood).
Feelings about work
A single question--'What feelings do you usually have about your work when you are on your way there?'--was used to gauge how respondents felt about their jobs. The five different response alternatives are as follows: 1) I feel happy and satisfied about the interesting workday ahead, 2) I have a fairly positive feeling about work, 3) I have neither positive nor negative feelings about work, 4) I feel rather uneasy about work, and 5) I have a strong uneasy feeling about work. These five alternatives were reduced to three categories as alternatives 1 and 2 were combined into the category 'Happy or fairly happy and satisfied about the workday ahead' and similarly alternatives 4 and 5 were put into the same category of 'Rather or strong uneasy feeling about going to work', leaving alternative 3 as 'Neither positive nor negative feelings about going to work'.
Self-rated general health
The single item taken from the SF-36 short-form health survey was used: 'In general, how would you describe your health?' [26
]. From the five response alternatives, three categories of self-rated health were defined: very good (excellent/very good), neither good nor bad (neutral), and not good (fair/poor).
Procedures and analysis of dropouts
The LIW project started in May-June 2005 and continued until May 2008. Characteristics of the responders at baseline in May-June 2004 were analysed to explore indicators of lifestyle-related health/behaviour in the total population before the onset of the healthy-lifestyle-promotion project. Data from the four-year follow-up, conducted in May 2008, was then used to analyse how participation in LIW was related to these indicators, among the 1859 HCWs available for analysis. When comparing dropouts with the individuals still responding at follow-up, the latter population was slightly younger on average, 46.4 years (SD 9.3) compared to 47.4 years (SD 11.1) for the dropouts (p = .001). This was expected since individuals who retired from work between 2004 and 2006 were not invited to participate in 2008. The proportion of women was greater in the responding population 2008 (62.4% versus 55.2%; p = 0.004). The distribution according to educational level, based on profession (whether the occupation requires college/university studies), was slightly different between the groups as 61% of the responders were classified as having higher education compared to 58% among the dropouts (p = .049). BMI was similar in the two groups, but the responders seem to constitute a slightly healthier population as only 4.5% reported their general health to be 'not good', compared to 8% among the dropouts (p = .0005). Furthermore, depressive mood (HAD-D score ≥ 7) was found in 16% and 21%, respectively (p = .001). Also, the proportion experiencing sleep disturbances at baseline was lower in the former group (37.3% versus 41.6%; p = 0.031), and the proportion reporting a sedentary lifestyle was 13.6% compared to 16.6% in the dropout group (p = .024).
Data from the survey in 2004 was used to analyse and describe the target population for the LIW project. No data imputation was done, and the number of observations can thus vary somewhat for the respective analysis. Descriptive statistics are given in terms of counts and percentages. Pearson's chi-square test was used to analyse group differences except for BMI and age where Mann-Whitney U test was used. The level of significance was set at p < 0.05. The association between the respective indicator of lifestyle-related health/behaviour and participation was assessed by Cox regression and expressed as participation ratios (PR) (which equals prevalence ratios at follow-up) with 95% confidence intervals. All analyses were performed by using SPSS version 15.0.