The majority of the participants in this study were female which is representative of the gender profile of both local and global clothing manufacturing employees [28, 29]. The age distribution of the sample in this study is similar to the South African economically active population range of 25-55years that allows the results to be generalised to the South African working age population. It may be argued that although the subjects were randomly assigned to the experimental and control groups, the sample comprised of volunteers who were interested in improving their own health and thereby may not be representative of all the employees within the clothing and textile manufacturing industry. Participants in this study were particularly unhealthy with regards to baseline BMI measurements. The mean BMI of the sample was 29.36 at baseline, which is indicative of overweight but closely approaching obesity. This is representative of adults in the Western Cape Province .
There was little or no attrition in the number of participants that would have threatened the internal validity of this study. Attendance at the weekly intervention sessions was generally high which further strengthened the study’s internal validity. Similarly, other worksite intervention studies have also reported a high attendance and participation rate but these studies were not randomised controlled trials [32, 33]. The high attendance and compliance of participants in this study could partly be attributed to excitement at the companies associated with the EWP as it was the first time an EWP was implemented within the South African Clothing Manufacturing Industry. These findings suggest that clothing industry workers welcome the idea of worksite wellness programmes and are motivated to participate in such programmes.
The outcome of this study demonstrated that the CBT-based EWP implemented over a period of six weeks was no more effective in improving self-reported difficulties on the EQ-5D than the once-off education session. The lack of significant difference between the two groups EQ-5D health domains (mobility, self-care, usual activities, pain, and depression) was not completely unexpected as the short-term EWP was not focused on symptomatic relief, but rather designed to equip employees with the knowledge and skill of improving health and promoting positive health behaviour change. This approach was utilised to afford the participants an opportunity to set realistic short-term goals for themselves and to gradually progress their behaviour changes. Similarly, a study by Baker et al. (1998)  showed that participating in a short-term combined walking and education programme did not produce significant changes in the EQ-5D utility scores amongst the participants.
Although we may not have expected large symptomatic improvements in this study, the lack of significant change could also be attributed to the choice of outcome measure. The EQ-5D is a generic instrument that measures health-related quality of life and is not disease specific . Although the EQ-5D has been validated for use among the South African population , a few studies have reported it to be insensitive to detect a change when baseline scores are high [36, 37]. In this study, the majority of the subjects had high baseline scores for the EQ-5D health state descriptors indicating no problems with most of the functional domains. Therefore, the lack of significance in the change of scores could be attributed to insensitivity of the EQ-5D.
The experimental group’s improvement in EQ-5D VAS scores at six weeks post intervention could be attributed to their participation in weekly exercise and the perceived knowledge gained from the workshops. Individuals with improved levels of self-esteem and good psychological wellbeing are more likely to report better perceived health states . Consistent with this study’s findings, other studies have also reported that a combination of behaviour therapy and supervised exercise programs is effective in improving the perceived health state of patients with musculoskeletal type pain [15, 39, 40]. In the study by Dahl and Nillson (2001) the cognitive behaviour therapy sessions were administered by a physiotherapist and registered nurse on an individual basis, twice a week for four weeks . The study by Wigers and Finset (2007) showed improved overall HRQoL at six months amongst patients with chronic musculoskeletal pain after participating in a four week CBT education and exercise programme. However, their study did not have a control group and in conjunction with the behavioural intervention the participants also attended individual physiotherapy sessions four times per week for myofascial release.
However, since a ten point improvement was noted in both the control and the experimental group’s EQ-5D VAS scores at six weeks post-intervention, the improvements in perceived HRQoL cannot only be attributed to the EWP. The overall improvement may be due to the Hawthorne or other non-specific effects as a result of being part of the study. The subjects may have perceived their health as improved due to being a participant in a health and wellness study. Long-term follow-up studies are recommended to investigate whether the improvements of perceived HRQoL is maintained after the intervention ends and whether the difference between the two groups becomes more apparent in the long-term.
The EWP encouraged participants to establish weekly goals for themselves, exercise at their own pace and to engage in leisure time physical activities that they enjoyed. This could have contributed to the increase of physical activity behaviours amongst the experimental group. Literature suggests that individual goal setting abilities have a positive effect on short-term behaviour change and exercise maintenance [41, 42].
Educating employees on nutrition and cooking methods coupled with supervised exercise programmes appear to have an impact on reducing BMI. The EWP in this study encouraged participants to set weekly nutrition goals, increase their daily fruit and vegetable consumption and advised on healthy cooking methods. This, in conjunction with increased exercise, could explain the reduction in BMI over the short time-frame. Similarly, a study by Aldana et al. (2005) proved that a worksite intervention programme addressing nutritional behaviours and encouraging self-monitoring of nutrition was effective in reducing the BMI of the participants at six weeks post intervention. Although their study did not include participation in supervised aerobic exercise classes , their educational workshops were much longer in duration and were more frequently held in comparison to this study.
However, this study also reported a reduction in the control group’s BMI. The control group did receive a once-off health promotion talk and education pamphlets that guided on healthy eating and cooking methods. In addition, all subjects entered the study highly motivated to improve their health. Self-motivation and intention to change has been associated with successful health behaviour change . Overall, these findings suggest that worksite health promotion intervention has an effect on reducing workers BMI.
Similar to the decrease in BMI across the two groups, this study found a statistically significant reduction in sickness related absenteeism of the entire sample group post completion of the study. The overall reduction in sickness related absenteeism could be attributed to improved morale amongst the study participants due to the excitement of the wellness study. The fact that the employers allowed the wellness study to occur during paid working hours may have resulted in a perception that management was interested in their health and wellbeing. According to Nawaz (2006) acknowledging and appreciating employees promotes improved morale at the workplace . The overall reduction in absenteeism in this study indicates that worksite wellness programmes can potentially offer positive economic benefits to the company. However, in order to determine the financial effects a larger population would be required and absenteeism rates would need to be monitored over a much longer period than was done in this study.
Strengths and limitations
The major strength of the study was the use of a randomized controlled experimental design. The sample size was sufficient to detect a predicted difference. The same person conducted the wellness programme at the three companies and thereby standardization of the intervention programme was maintained.
A major short-coming of the study was that it failed to record the detailed medical history of the participants. It could be argued that the groups were not matched for co-morbidities which could have impacted on results and adherence to the programme. Therefore it is unknown whether the presence of co-morbidities in either group impacted on the results. Apart from the BMI measurements and absenteeism records assessed, the study relied significantly on self-report outcome measures which could have introduced bias. A possible Hawthorne effect may have caused information bias as the participants may have tried to please the researcher by giving favorable and not honest answers. However, considering that this study was primarily concerned with the effects of the intervention on perceived HRQoL and exercise behaviours, self-report measures are the only possible indicators of change. Double-blinding is impossible with this type of intervention as participants can easily identify what is the intervention being tested. A further limitation of the study was that although subjects were randomly selected, the subjects that consented to being part of the study were all motivated and interested in improving their health. Therefore it remains unknown whether the EWP would have the same effect on subjects who are not motivated to change their behaviours.