Approximately 140 million U.S. adults (60% of the free-living population) are employed . Recent statistics indicate that 68% of U.S. adults are overweight, with 33% classified as obese . Adults are at continued risk for weight gain over time, with average increases of two pounds per year . Obesity is rising across all employment groups, with no differences in increases by race/ethnicity or sex . Obesity is associated with negative consequences in working populations, including more frequent absenteeism, sick leave [5, 6] or workplace injury and disability pension claims [7, 8], and greater health care costs .
As employed adults spend approximately half of waking time at work , worksites provide a logical setting in which the environment might be reshaped to promote healthier behaviors and improve weight control. Employers may be motivated to make changes due to concerns about quality or cost of employee health care , and worksites may be in a unique position to leverage resources (e.g. food service, communication networks, Human Resources departments) and promote social support among co-workers for obesity-preventive behaviors.
Environmental changes that make conceptual sense for obesity prevention include targeting food service (i.e. availability of energy-dense foods, portion sizes, cost), the physical environment (e.g. opportunities for exercise), and information distributed to increase knowledge of behaviors related to obesity risk [3, 11]. Previous studies designed to address obesity in the workplace typically have focused on a single factor from this list of options or had other limitations to the study design that preclude systematic examination of the effects of environmental changes on body weight over a prolonged period of time.
With regard to promoting changes to the food environment, one study focused on availability of healthy foods in cafeterias and was successful in reducing prices by 50% and increasing purchasing of fruit and salad by 300%, but was conducted in only a single worksite over a three-week period, using a pre-post observational design rather than randomizing to intervention or control conditions, and did not assess body weight as part of the study . Another study targeted pricing and promotion of healthy vending machine snacks in worksites and schools over a one-year period and achieved significant increases in purchases of healthy snacks by lowering prices by 50% and enhancing promotion; however, as with the study noted above, body weight was not monitored as part of this trial, so the effects on weight outcomes over time are not known .
Other studies have focused on aspects of the environment that lend themselves to physical activity. A 2004 study examined pedometer use among 177 sedentary employees using a 12-week pre-post design; 59% of employees completed the program, which resulted in an average increase of over 3,400 steps per day from baseline and a small but statistically significant change in BMI; this study is limited by a relatively small sample size, short study duration, and lack of randomization to conditions . Stair use at work has also been targeted for intervention. In one study, the Centers for Disease Control and Prevention in the United States used one of their office buildings to test the effects of stairwell enhancements (updated carpet and paint, artwork, music, and motivational signs throughout the building) on employee stair use assessed over a 3.5-year period . Stair use, measured objectively using infrared proximity sensors, increased by 4.7% overall; however, this study did not include a comparison condition without stair intervention, nor did it assess body weight or track changes in weight over time .
In addition to the limited focus or methodological limitations of the trials noted above, most studies in this area have focused on behavioral and informational changes in worksites rather than strictly environmental or policy-related changes . For example, an earlier trial by the present research team investigated the effects of a program offering weight control or smoking cessation classes four times in two years at 16 worksites randomized to intervention, versus 16 sites randomized to an assessment-only control condition . The intervention had no significant effects on weight over two years relative to the control condition, likely due to very low participation in weight loss classes during the study (16% of all employees during the first wave, approximately 5% in subsequent waves) and reliance on attracting employees to a separate educational program for weight loss rather than integrating healthy weight-related behaviors into the worksite environment as a whole .
To address obesity among working adults, the goal of the HealthWorks group-randomized trial was to implement a multi-component program to alter the worksite environment, with the intent of reducing weight gain [expressed in body mass index (BMI) units] over the study period, relative to control sites with no investigator-initiated alterations to the environment. The results of the first weight-gain prevention trial funded by NIH set the stage for the approach taken in the current trial. The Pound of Prevention (POP) study was conducted from 1995 to 1999 by the present investigators . Its purpose was to evaluate whether weight gain could be prevented in healthy young community-dwelling adults through education. Participants (228 men and 998 women aged 20 through 45 years) were recruited as individuals via telephone contacts, mailed or newspaper advertisements, and face-to-face contacts, and randomized to an intervention consisting principally of a monthly educational newsletter or to a no-treatment control condition. Five behavior changes were advocated: 1) increase frequency of self-weighing, 2) increase physical activity, 3) increase fruit intake, 4) increase vegetable intake, and 5) decrease intake of high-fat foods. After 3 years, the intervention was successful in increasing knowledge, self-weighing frequency, and healthy weight control behaviors . In addition, aggregated across all treatment groups, individuals who reported adopting these study recommendations were more successful in preventing weight gain over time than those who did not . However, the intervention did not significantly slow the rate of weight gain over time in the intervention group when compared to the group receiving no treatment.
Drawing on the literature described above in terms of representative studies, the two main goals of the HealthWorks trial were to a) conduct a simultaneous test of environmental approaches to healthy weight choices (including eating, physical activity, and weight monitoring behaviors), and b) extend the test of these approaches to a more sustained time period than typically observed. The study was conceptualized in terms of social cognitive theory (SCT), which argues that environment, behavior, and cognition are reciprocally related, such that changes to the environment should influence behavior . Intervention efforts primarily focused on the environmental change component due to a relative dearth of studies in that area, as well as being driven by a social ecological interest in taking a broader worksite-level approach to obesity prevention rather than an individualized behavioral approach to the issue , with the consideration that such a program has potential to be sustained in worksites at a lower cost than high intensity, individual behavior change efforts.
The four major environmental change components addressed food selection, promotion of walking/stair use, weight self-monitoring, and health information at work; we hypothesized that employees at sites with environmental changes would gain less weight over the two-year study period, relative to those employees at sites with no changes; based on weight gain expectations for an untreated adult population over a two-year period, we expected to observe weight gain of approximately 1 kilogram (kg) per year . These components have been tested separately in previous worksite or other obesity prevention studies and found to be effective in terms of impact on the environment, though most previous trials were of relatively short duration and not all assessed weight changes as an outcome [12–17, 23]. The addition of a weight self-monitoring component to the current project addresses an obesity prevention strategy with minimal evaluation in worksites. The current study represents an effort to test these components using an integrated worksite environmental approach, rather than by relying on individual or small-group weight education and health promotion activities  or attempts to alter workplace norms related to weight . Knowledge from this trial has the potential to promote a healthier workforce and improve population weight control.