Background

Noncommunicable diseases (NCDs) are the leading causes of death and disability worldwide [1]. Reports from the World Health Organization estimate that in 2021, 71% of deaths globally were attributable to NCDs, rising from 63% in 2013 [2, 3]. Lifestyle diseases like cardiovascular disease (CVD), type 2 diabetes and obesity continue to be a treatment conundrum for public health policy makers. The ever-ubiquitous nature of overweight and obesity [4], alongside the rising prevalence of type 2 diabetes compounds treatment approaches that aim to reduce the debilitating effects of CVD [1, 5]. Evidently, disability-adjusted life years for CVD have risen by 16.4% between the years 2007 and 2017 [6], further underscoring the need for action.

Considering the pervasive nature of these diseases, novel treatment approaches have been proposed, including using the workplace as a front for the dissemination and promotion of positive health behaviors [7]. Modifiable health behaviors such as physical inactivity, poor diet, excessive alcohol consumption and tobacco smoking, can all negatively contribute to NCDs [8]. Hence, health programs which look to target these behaviors in the workplace could prove valuable in the ongoing fight to curb the morbidity and mortality of these diseases.

Workplace health promotion programs offer a coordinated and strategic approach to health promotion, empowering employees to live healthier lives through policies, environmental support, and health promotion activities which can be conducted in the workplace [9]. There is a growing evidence base to support these programs [7], especially with regards to modifying dietary and obesogenic habits, as well as improving cardiometabolic health outcomes [7]. Specifically, a meta-analysis of 121 studies by Peñalvo et al., [7], found that workplace wellness programs can statistically improve fruit and vegetable consumption (0.27 servings per day [95% CI 0.16 to 0.37]), as well as systolic blood pressure (–2.03 mm Hg [–3.16 to –0.89]) and bodyweight (–0.92 kg [–1.11 to –0.72]) [7]. Other studies have investigated effects of workplace physical activity interventions on men, and have found mixed success [10].

As a front for health promotion, workplaces offer the necessary infrastructure and sustained reach to a large number of individuals for prolonged periods each day [11]. This larger reach also means workplaces have access to demographics who would otherwise be difficult to engage in health promotion. Blue-collar male dominated industries such as mining [12], construction [13], and manufacturing [14], could all potentially benefit greatly from workplace health programs as this population is usually underrepresented in health research [15], yet may benefit the most from such interventions [16].

Globally, Australia is one of the biggest mining nations [17], and is just one example of a blue-collar male dominated industry that could benefit from workplace health promotion. The Minerals Council of Australia, the country’s governing mining industry body reaffirms that core to the industry, there is a strong emphasis on, and commitment to the safety, health, and psychological wellbeing of its workforce [18]. Despite this, the physical health and wellbeing of employees in the industry is notably falling behind [12, 19, 20]. The industry comprises of approximately 90% males [12], with roughly 85% of males in the industry classified as overweight or obese [21]. Employees in this industry have also reported lower consumption of fruit and vegetables compared to gender and aged match general population samples [12, 22]. In a cross section of 949 coal miners across New South Wales (NSW), Australia, it has been reported that only 3.5% of employees meet current recommendations for vegetable intake, as well as 42.2% for fruit (compared to 9% and 51.4% nationally) [12]. This is in conjunction with an episodic hazardous drinking culture [23], capped off by an obesogenic work environment whereby employees spend large parts of the day sedentary.

Knowledge gaps remain about the success of diet and physical activity programs within blue-collar workers, particularly miners. This paper presents baseline and follow-up findings from a pilot study that implemented a 12-week multicomponent workplace health promotion program at a mine site in rural NSW, Australia. The study aimed to assess the efficacy of a workplace health program that targeted nutrition, physical activity, and obesity outcomes, implemented under the broader RESHAPE framework. Secondary outcomes of interest were whether workplace factors contributed to poorer health behaviors.

Materials and methods

RESHAPE is an eight-step framework based on the WHO ‘Health Workplace Framework and Model’ [24]. The framework was developed in response to an industry identified need that was highlighted from the ‘Blueprint for the management of overweight and obesity in the NSW mining industry’ [25]. Coal mining, and mining more generally, is an incredibly dynamic and variable environment. As such, RESHAPE does not adopt the mantra of one size fits all, rather it looks at problems pragmatically, and applies appropriate mine site specific solutions. Whilst this study is focusing on the issue of overweight and obesity within the workplace, the RESHAPE process is designed to be used for any modifiable health risk factor including smoking, nutrition, alcohol, physical activity, sleep or mental health.

The aim of RESHAPE is to provide a sustained approach to fostering healthy, happy, safe, and productive workplaces. Through shared responsibility, the workplace can become an environment where healthy choices are valued. This produces organizational change through ongoing investment into workplace health and cultivates a workplace culture which is conducive to positive change. RESHAPE is a systematic approach to workplace wellness. Encompassed within this framework are wellness programs which this paper looks to discuss.

Intervention

The wellness intervention took a whole site approach whereby participation was open to all employees on site. All employees were made aware of the intervention and were encouraged to participate; however, the final level of involvement is unknown. The intervention was a commercially available 12-week multicomponent wellness program which was paid for by grant funding and made freely available to all employees. The program comprised of twelve weekly health focuses, with an emphasis on nutrition and physical activity to drive changes to obesity outcomes. The intervention used several aspects of gender targeting to increase intervention uptake across the site, including: [1] healthy weight challenge; [2] step challenge (who could do the most steps); [3] Core stability challenge (time plank adjusted for age) and [4] Healthy recipe mini challenge (submission of best home cooked recipes that was scored against a criterion for originality, presentation, and nutrition value).

The program included aspects of behavior change techniques including goal setting, self-monitoring (through an online website), modification of eating habits and problem-solving, and a large emphasis on knowledge and education. Goal setting was conducted via a generalized template asking several questions including: “Why is your health important to you?”, “What aspects of your lifestyle could you improve” and, “How could you improve these lifestyle behaviors?”. Participants were encouraged to set three action orientated goals. In conjunction with the nutrition aspects of the program, employees were educated about the Australian physical activity and exercise guidelines [22], and encouraged to increase physical activity to these levels. Handouts and communications were delivered to participating employees’ email, and participants logged into an online portal to record information about their diet and exercise. Table 1 outlines additional intervention aspects and flow.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. 12-week intervention breakdown.

https://doi.org/10.1371/journal.pone.0288244.t001

Sensitivity analysis

Due to missing data on the outcomes, sensitivity analyses were also performed on each outcome, imputing missing data via multiple imputation by chained equations (MICE) under the missing at random (MAR) assumption. The association between missing data on each outcome and demographic/work-related characteristics (age, gender, highest civilian qualification, relationship status, occupation, employment status, usual hours worked per week, shift work status) were utilized to inform the imputation models. For both the primary analysis and secondary analysis, effect estimates were stable after n = 50 imputed data sets (Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Sensitivity results health related outcome (m = 50).

https://doi.org/10.1371/journal.pone.0288244.t002

Results

There were n = 389 records at baseline and n = 420 recorded at follow-up, as well as 61 participants from both timepoints (8.2% repeated measures) (Table 3). Surveys were distributed to all employees onsite with a total of 556 surveys distributed at baseline and 583 at follow-up. This represents a response rate of 70% and 72% respectively at each timepoint. At baseline and follow-up, majority of participants belonged to the 25–34 age bracket or 35–44 age bracket (32% and 31% respectively). Most of the participants were male (89%) and were born in Australia (93%).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Participant demographic characteristics.

https://doi.org/10.1371/journal.pone.0288244.t003

The highest qualification obtained varied amongst participants with a trade / apprenticeship (36%) being the most reported, followed by school certificate / higher school certificate (34%). Most participants were married or in a de facto (living together but not married) relationship (76%) and permanent full-time employees (78%). Hours worked per week varied with 39–45 hours being the most reported (53%), followed by 46–56 hours (27%), as well as most employees participating in shift work (82%).

Table 4 shows summary statistics for the outcomes. Numeric outcomes are summarized through mean (SD) and median (Q1, Q3), and categorical outcomes are summarized through frequencies and percentages (n (%)). The mean BMI at baseline was 30.01 kg/m², and at follow-up 29.79 kg/m². The percentage of participants not meeting the Australian physical activity and exercise guidelines decreased between baseline and follow-up (61% to 52% respectfully), in addition to a mean increase in the number of moderate physical activity minutes participated in per week between baseline and follow-up (153 minutes to 201 minutes). Dietary consumption patterns are reported across the varying cut offs. At baseline and follow-up, the most reported fruit intake was 1 serve per day, and similarly for vegetables, 1 serve per day.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Summary statistics of assessed outcome variables.

https://doi.org/10.1371/journal.pone.0288244.t004

Table 5 shows the crude and adjusted estimates for each outcome. Odds ratios and rate ratios are follow-up when compared to baseline. For BMI, in the crude models, the mean BMI at baseline was 30.02 and the mean BMI at follow-up was 29.76. The difference between baseline and follow-up is -0.27, but this was not statistically significant (p = 0.39) and similarly for the adjusted model -0.22 (p = 0.52). From baseline to follow-up, the crude model demonstrated that there were no statistically significant changes regarding fruit intake (OR = 1.13, p = 0.58), vegetable intake (OR = 0.93, p = 0.73), sugar sweetened beverages (OR = 1.29, p = 0.26), and fast-food intake (OR = 1.03, p = 0.86).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Crude and adjusted estimates for each outcome comparing follow-up against baseline.

https://doi.org/10.1371/journal.pone.0288244.t005

For physical activity and exercise guidelines in the crude models, people at follow-up had 111% higher odds of meeting the physical activity and exercise guidelines compared to people at baseline, but this was not statistically significant (OR = 2.11, p = 0.057). For moderate physical exercise in the crude models, people at follow-up had 81% lower odds of doing no moderate physical exercise compared to people at baseline, and this was statistically significant (OR = 0.09, p < 0.001). Furthermore, people at follow-up had 10% increase in the number of minutes spent doing moderate physical exercise when compared to people at baseline, but this was not statistically significant (RR = 1.10, p = 0.236).

Table 6 examined workplace factors as predictors of exercise outcomes. Physical activity and exercise guidelines, Minutes of moderate physical exercise per week, and Minutes of vigorous physical exercise per week were examined against the factors of interest. For Physical activity and exercise guidelines, people who were employed part-time had 5% higher odds of meeting the guidelines compared to people employed full-time, but this was not statistically significant (OR = 1.05, p = 0.96).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 6. Modelling of exercise related outcomes.

OR = Odds Ratio, RR = Rate Ratio.

https://doi.org/10.1371/journal.pone.0288244.t006

For moderate activity (binary), people at follow-up had 97% lower odds of not doing any moderate activity compared to people at baseline, and this was statistically significant (OR = 0.03, p < 0.001). For vigorous activity (gamma), managers had 10% lower number of minutes spent performing vigorous activity compared to machinery operators, but this was not statistically significant (RR = 0.9, p = 0.690).

Discussion

This study underscores the potential benefits of workplace health programs in improving employee’s modifiable health behaviors within the coal mining industry. Noting that people at follow-up compared to baseline had 111% (p = 0.057) higher odds of meeting physical activity and exercise guidelines, and 81% lower odds of doing no moderate physical exercise (OR = 0.09, p < 0.001). When examining predictors of exercise outcomes against workplace factors (shift work status, employment type, hours worked per week), there were not significant associations. Notably, there was no differences in fruit (p = 0.39) and vegetable (p = 0.39) intake between timepoints.

Physical inactivity has established links with a raft of non-communicable diseases and premature mortality [34, 35]. So much so, that globally, 7.6% of cardiovascular disease deaths can be attributed to physical inactivity, and similarly for all-causes of death (7.2%) [36]. Combined with the notion that one-in-four adults globally do not meet current physical activity and exercise guidelines [37], there is significant potential for workplace health promotion programs that target these behaviors. Considering the efficacious results of this paper, as well as others [7, 38], the authors posit that the workplace should be more heavily utilized in the promotion of physical activity and associated health behaviors. One particular study which found similar success in increasing physical activity outcomes in a similar male dominated industry was the Preventing Obesity Without Eating like a Rabbit (POWER) study by Morgan et al., [39]. Amongst 110 workers with overweight or obesity in a manufacturing plant in Australia, during a 14 week wait-list randomized control trial, workers in the intervention group significantly increased current physical activity levels (p<0.001). Similarities between the POWER study and this study was the utilization of behavior change techniques (BCTs) in intervention materials and design which has been highlighted as a powerful tool for workplace health programs that target modifiable health behaviors in men and male dominated industries [38].

Behavior change techniques can be thought of as the active ingredients of behavior change interventions [40]. Meta-analytic models by Sharp et al., [41], have shown in males, the utilization of BCTs is a key driver in increasing physical activity in intervention studies. Overall, behavior change interventions can have a small but significant effect on increasing physical activity levels (d = 0.35) in men [41]. This effect size is consistent with an increase of approximately 97 minutes of total physical activity per week or 980 steps per day [41]. Interestingly, studies that employed four or more BCTs were found to have a larger effect size (d = 0.51; 95% CI 0.38,0.63) than studies that used three or less (d = 0.24; 95% CI 0.15,0.34) [41]. Utilizing appropriate BCTs to increase participation and uptake of physical activity intervention could prove an advantageous design decision for workplace health programs to targeting modifiable health behaviors [34, 35].

Workplace health programs have shown effectiveness in statistically reducing overall body weight. Peñalvo et al., [7], recently conducted a systematic review and meta-analysis assessing the effectiveness of workplace wellness programs for dietary habits, overweight, and cardiometabolic health. The review included 121 studies and found that workplace wellness programs significantly reduced BMI (–0.22 kg/m2 [95% CI –0.28 to –0.17]), bodyweight (–0.92 kg [–1.11 to –0.72]), and waist circumference (–1.47 cm [–1.96 to –0.98]. Whilst the findings from this study could not replicate the statistical significance produced by the aforementioned review, differences of -0.88kg body weight (p = 0.32) draws attention to the potential of these programs in improving health outcomes [7].

Considering this, from a health and safety perspective, even modest changes to weight can have substantial impacts on safety outcomes within the mining industry and the workplace more broadly. Wilson [42], reinforces this notion in a study investigating safety incident risk of 868 underground and opencut miners over a 5-year period. The study found that for every one-unit increase BMI over a reference 25 (upper threshold of healthy), there was an 8% (p < 0.05) increase in incident risk. Additionally, Maisey et al., [43], when studying workplace fatigue in fly-in-fly-out mining workers found that for every 1 unit increase in BMI, the odds of risk for obstructive sleep apnea increased by 19% (p < 0.05). Obstructive sleep apnea is one of the most prevalent sleep disorders the mining population (31%), second only to shift work disorder (44%), and a significant contributor to fatigue [43]. Safety issues such as fatigue pose a threat to worker health and safety with an estimated 65% of truck driving accidents in open cut mines being fatigue related [44]. The benefits of workplace health programs can extend beyond the individual behaviors they are targeting and double down as a safety incident risk management measure.

Strengths of this study include a large sample size and an intervention that took a whole of site approach. This approach has several benefits as implementing a health program in a non-controlled environment allows findings to be more applicable to industry, thus increasing the translation of research into practice. Limitations include not knowing the exact level of participation in the wellness intervention. Whilst all employees were made aware of the intervention and could participate, the level of involvement is unknown. Additionally, due to COVID-19 extending the follow-up timepoint (due to site access constraints), potential contamination regarding wellness outcomes ought to be acknowledged. Furthermore, this study acknowledges the limitation of not including data on pre-existing medical conditions, which could have provided additional context regarding the importance of workplace health promotion programs. Lastly, reporting bias due to the self-reported nature of measures and the small sample of paired data should be recognized. Due to the small sample of paired data, caution is prescribed in interpreting results on an individual level. Despite this, due to the high response rate at baseline and follow-up (70% and 72% respectively), the authors believe these results still offer a snapshot of the workplace as a whole.

Conclusions

This study assessed the efficacy of a multicomponent workplace health program targeting modifiable health behaviors in NSW coal miners. Notably, improvements to physical activity outcomes and to a lesser degree, marginal weight loss does indicate that the workplace could be a viable option in health promotion and improving employee’s health outcomes. Further research is needed to determine the true effectiveness of these programs, particularly in an environment as challenging and dynamic as the mining industry.

Acknowledgments

The Authors wish to acknowledge and thank all staff and participating employees across both timepoints and sites.