SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Objective: Based on previous worksite-wide intervention studies and an ecological framework, we created a behavioral intervention program to maintain or reduce weight through healthy eating and physical activity. The design and evaluation plan of the group-randomized trial and the recruitment of worksites are described. Preliminary results regarding the dietary and physical activity behaviors associated with BMI are discussed.

Research Methods and Procedures: The intervention used an ecological framework modified by qualitative methods that identified salient barriers and facilitators of behavioral change. Approximately 30 transportation, manufacturing, utilities, personal, household, and miscellaneous service companies in the greater Seattle area are being recruited to the trial. The study population for the present analysis consists of 18 worksites from the first two randomization waves. Dietary behavior was assessed, not by calories, but by behavioral measures related to BMI. Physical activity behaviors were surveyed. BMI is derived from reported height and weight at baseline.

Results: The intervention has been developed with a specified minimum suite of strategies within the defined framework. Response rates to the baseline survey among the 18 worksites are 81% on average. After adjusting for age, gender, race, and education, BMI was associated with frequency of intensity-adjusted physical activity, sweat-inducing exercise, fast food meals, soft drinks, eating while doing another activity, and fruit and vegetable intake.

Discussion: Worksite-wide intervention strategies can be adapted to target obesity prevention. Employees are willing to participate in surveys at high rates. Several measures of physical activity and eating choices are associated with baseline BMI.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

The increasing rate of overweight and obese Americans is concerning given the health risks associated with obesity. Risk of heart disease, diabetes, certain types of cancer, breathing problems, arthritis, and reproductive complications are all increased in overweight and obese persons (1). Moreover, overweight and obesity are associated with an increased risk of premature death; an estimated 300,000 deaths each year are attributed to obesity (2).

Given these risks, reducing the prevalence of obesity has become the focus of much research. The Healthy People 2010 objectives recommend increasing the portion of people who are at a healthy weight (BMI between 18.5 and 24.9 kg/m2). Healthy People 2010 supports decreasing the proportion of the U.S. population who is obese to 15% (3) from the current 30% (4).

In weight-stable individuals, energy intake balances energy expenditure. Bringing energy intake and expenditure back into balance is the best way to maintain or lose body weight, but we need new ways to get the message out and ensure the behaviors are incorporated into new social norms. Worksites may provide a unique opportunity to deliver intervention messages that encourage the balance between energy intake and expenditure. Worksites are like small communities, in that each worksite has its own social network and often has an infrastructure for disseminating information or messages to its employees. Employees spend about one-third of their waking hours at the worksite, and >62% of the U.S. adult population are employed (5). Therefore, worksite level interventions have the potential to reach a wide portion of the adult population. Indeed, several worksite studies have evaluated behavior change in relation to smoking cessation (678), increased fruit and vegetable intake (9101112), and decreased fat consumption (131415).

Worksite interventions that focused on weight control have used various strategies, including behavior modification (16). Examples of these techniques include virtual (computer-based) education programs to receive physical activity and wellness information (17); posting key health messages in workplace common areas (18); keeping journals of food intake, weight, and exercise to be used in counseling sessions (19); and self-contracting (19,20). In the Working Healthy Project, Emmons et al. (21) demonstrated efficacy in changing physical activity and healthy eating practices at worksites by targeting interventions around social norms and health-related policies at the worksites. Despite the ability of worksite interventions to foster changes in behavior that support energy balance and may lead to reductions in weight gain, the magnitude of the individual change in weight tends to be less than that observed in clinical trials (22). We build on this literature with a combination of proven worksite level and individual level strategies in the Promoting Activity and Changes in Eating (PACE)1 project, a randomized controlled trial of worksites in which the worksite is the unit of randomization.

Intervention Development

Intervention Design

The intervention was based on a modified ecological framework (23). At the social environmental level, worksites are systems in which the environment can be changed to encourage health behavior change (24). There are several worksite factors that have relevance for facilitating behavior changes that relate to eating and physical activity, including current finances in a company, unionization, insurance costs, company morale, and social norms. At the individual level, the intervention is guided by Social Learning Theory, which emphasizes self-efficacy and the role of support in behavior change through changing social norms, as shown in Figure 1.

image

Figure 1. Conceptual framework of intervention.

Download figure to PowerPoint

Formative Research

To understand attitudes, beliefs, and behaviors of employees that are relevant to weight maintenance and weight loss, PACE project staff sought 20 volunteers to complete a confidential interview. The goal of these elicitation interviews was to identify items that underlie eating and physical activity decisions. The interview questions were open-ended and semistructured with extensive probes about the behavior being discussed. The interviews were completed at a pilot worksite that fit the eligibility criteria for the study. The goal of the qualitative study was to refine the framework for the intervention. The interviews were tape-recorded, and common themes were analyzed.

Four major themes were extracted from the interview data, namely participants perceived physical activity to be rigorous exercise, participants found it difficult to track their dietary intake due to a lack of time or sense of personal restraint, participants preferred to receive healthy eating information from the Internet and through tangible items such as flyers and brochures that they could refer to at their leisure rather than face-to-face, and participants perceived work-sponsored health promotion programs to be beneficial because they are free and convenient. The content from the interviews suggested the importance of changing the connotation of the term “physical activity” from being that of rigorous exercise to that of moving more to achieve personal goals. Secondly, we developed a simplified tracking system called counting hundreds impact points (CHIPs) that would be an easy way to track how much food a participant could consume and the amount of activity a participant should complete each day to maintain and/or decrease energy balance. Thirdly, we developed a web site that would be easy to navigate and would provide participants with healthy eating and physical activity tips and messages. These messages were also made available in tip sheets and brochures. Finally, we developed a number of worksite-wide social activities that a worksite employee advisory board (EAB) (see below) can use to motivate and educate the employees about the importance of healthy eating and moderate physical activity and support them in their behavior change attempts.

Once the implementation framework of the intervention was designed, two focus groups were held to further refine the intervention strategies. The first focus group was made up of 11 volunteers from a worksite that had participated in an earlier worksite health project (25). The volunteers responded to a series of questions and statements concerning the pros and cons of reducing energy intake and increasing energy expenditure, with the intent of identifying key barriers and facilitators of physical activity and healthy eating behaviors. The goals of the first focus group were to gather insight regarding barriers and aids to improve dietary intake and increase physical activity to validate themes from the elicitation interviews, to seek ideas to encourage the use of tracking dietary intake and/or physical activity, to determine initial impressions of the CHIPs concept and understanding of the term energy balance, and, finally, to determine feasibility of implementing the intervention at worksites and gather suggestions regarding intervention activities and giveaway items. Data from transcripts and notes from the first focus group were then analyzed using content analysis. Messages were designed to overcome the barriers and enhance the facilitators of lifestyle behaviors.

The second focus group consisted of an additional 11 volunteers from the same pilot worksite. The volunteers were asked to brainstorm ideas on the best ways to present the messages developed from the first focus group. The second focus group was also asked to provide feedback on proposed poster titles, to give their impressions of a variety of food and physical activity images, to share their opinions regarding an assortment of promotional materials and to offer suggestions for other promotional methods, to select formatting preferences of informational materials, and to give feedback about proposed intervention activities and coinciding content material. A final analysis was completed on data from the second focus group, and the results were used to prepare materials that combined the messages and the mode of delivery.

Implementation Framework

The PACE Program intervention is divided into five phases. Each phase has a number of strategies that build and enhance a particular behavior change focus. The worksites initially focus on one behavior at a time, moving from simpler strategies to those that require broader social support for long-term maintenance, for a total intervention period of 15 to 18 months. Each worksite is required to complete a defined minimum number of components in each phase before moving to the next phase.

The purpose of Phase 1 of the PACE intervention is to foster awareness of physical activity and healthy eating among employees. To introduce the PACE program to the employees, a kick-off event is held at the worksite. It is highly publicized, and all employees are encouraged to participate. There are several booths that feature a variety of health topics. The publicity uses a variety of posters announcing that “something is coming!”. These are displayed around the worksite to entice employees and create interest in the PACE Program. The campaign promoting the kick-off, the kick-off event with at least five booths, and constant inescapable messages displayed throughout the worksite using at least two communication channels (e.g., paycheck stuffer, poster) are required.

Phase 2 focuses on providing motivation and support for increasing physical activity behavior. The activities in Phase 2 are designed to motivate physical activity by teaching individuals how to increase their routine daily movement. The activities may be completed individually, in small groups or among the worksite as a whole. In addition to these worksite-wide activities, an indoor/outdoor walking trail or path is established. Participants are also given the PACE Self-Help manual in this phase and are encouraged to use the tracking tools and self-assessment materials to evaluate their current level of physical activity and to set goals to increase their amount of routine, daily movement. The concept of balancing energy intake with energy expenditure is introduced, using approximate units of calorie hundreds (CHIPs). A walking loop, promoted using at least two communication channels to promote it, a copy of the self-help manual to each employee, and one additional activity to reward behavior change is required.

The goal of Phase 3 is to provide motivation and support to improve dietary intake among employees. The activities in this phase are designed to involve all members of the worksite to establish an environment that supports positive health behaviors. Examples of such activities include food demonstrations and taste testing. Participants are encouraged to use the tracking tools in the PACE Self-Help manual and additional self-assessment tools to evaluate current dietary intake and identify strategies to improve the quality of their diet. At least one activity, promoted using at least two communication channels and constant inescapable messages around tracking dietary intake with reference to the self-help manual, are required.

Phase 4 concentrates on establishing a support system to encourage increased physical activity and healthy eating intake. To accomplish this, the worksites are encouraged to adopt new avenues for supporting these behaviors. Examples include participating in a sponsored fun run/walk event, offering exercise classes at the worksite or an employee potluck featuring healthy foods. The worksites are also encouraged to offer discounts to health clubs for their employees and to provide healthy foods at company meetings and in worksite vending machines. At least one activity promoted using at least two communication channels is required.

The strategy in Phase 5 is to support the maintenance of the physical activity and dietary changes that were made in the earlier phases by providing the resources and social support to maintain these behavior changes. Topics include eating out, holiday events, using tracking tools, and socializing smart. At least one activity promoted using at least two communication channels is required.

EAB

Each worksite randomized to the intervention is required to establish an EAB. The EAB consists of four to seven employees who volunteer or are recruited or nominated by a worksite primary contact person (usually the person at the worksite with whom the team communicates for arranging assessment visits). The EAB is made up of employees from all occupational sectors in the worksite and works closely with the project liaisons (from the research team) to design, plan, and implement intervention activities that are best suited for the worksite. By empowering employees, it is thought to be more likely that intervention activities will continue beyond the life of the research (26, 27).

Each EAB member is given a handbook that describes the study, explains their role as an EAB member, and provides the intervention framework necessary to carry out the intervention in their worksite. The handbook is similar in approach to that used in the Seattle 5-A-Day study (10). The intervention framework is outlined in the handbook providing a number of activities and messages that the EAB can tailor for their worksite. The EAB, although encouraged to do more, is required to implement the minimum framework outlined in the handbook and is responsible for tailoring the intervention to best suit their specific worksite needs. The EAB meets a minimum of once every 4 months to identify and plan activities. The EAB meeting is scheduled on company time and the length of time of the EAB meeting is dependent on the worksite but, on average, is no longer than 30 to 45 minutes.

Web Site

Employees from the PACE Program intervention worksites are encouraged to access the PACE web site for information on physical activity and dietary change. The web site includes a number of interactive features, such as quizzes and self-assessment tools that are designed to attract and assist the employee making changes. The web site also features a CHIP plan that calculates an employee's Base CHIPs (calorie hundreds for fueling basic daily activities based on age and gender), food CHIPs (calorie hundreds consumed), and activity CHIPs (expenditure on activities beyond the basic) for the purpose of tracking one's food and physical activity. The CHIP allowance (base CHIPs plus activity CHIPs) should equal food CHIPs, on average, for weight maintenance.

Research Methods and Procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

This manuscript reports the baseline results from the first two waves of worksites taking part in PACE, a group-randomized trial of an intervention to reduce weight gain, evaluated by baseline and 2-year follow-up assessments of BMI. Worksites in the Seattle Metropolitan area and along the I-5 corridor between Everett and Tacoma, WA were identified using U.S. Standard Industrial Classification (SIC) two-digit codes (28). These included manufacturing (SIC 20 to 39), transportation or utilities (SIC 40 to 49), personal services (SIC 70 to 79), household and miscellaneous services (SIC 88 to 89), and non-classifiable establishments (SIC 99). The identified worksites that had between 40 and 450 employees and were eligible on SIC code were called to initiate the first screening encounters. The first screen verified a worksite primary contact and the location and size of the worksites and obtained information on proportion of sedentary workers and Internet access. Priority for second screening was given to worksites where employees had at least 10 minutes access to a computer each week and whose location was within an hour of the city center. Initially, priority was also given according to proportion of sedentary employees. Later priority was given to blue collar or service industry worksites. For priority worksites, a letter was sent to the primary contact explaining the PACE study and then followed by a telephone call 1 week later. This second detailed eligibility screening also asked about interest in participation. Eligible worksites were recontacted for a recruitment meeting at which more details of the study were provided, and interest in participation and eligibility were checked.

Eligibility criteria include worksites with: a large portion of sedentary workers (>25%), a low turnover rate (<30%) over the past 2 years, a low proportion of non-English-speaking employees (<30%), a workforce between 40 and 350 employees; no more than two locations participating, at least a 3-year history of being in business, and willingness to be randomized to either the intervention or comparison (delayed intervention) arm of the study. Worksites with a wellness program that had an on-site, active physical activity or nutritional component were excluded.

At the formal recruitment contact, we explained the study in greater depth and requested the worksite provide all of the necessary documentation to participate in the PACE study (letter of intent to participate, employee list on date of baseline). A rolling recruitment process was established, such that worksites will be recruited in four waves. The recruitment goal is to have at least 34 worksites start the run-in process of documentation and baseline assessment to randomize 32 companies (16 pairs) to either the intervention or delayed intervention arm of the study. The intervention will last ∼18 months, with follow-up at the 2-year anniversary of baseline. Comparison worksites will get the intervention materials, assistance with establishing an EAB, the EAB handbook for members of that Board, and assistance with the first worksite wide event after completion of all follow-up surveys. Only baseline data are included in the present manuscript.

Baseline Measurements

Baseline measurements were collected from all employees if the worksite had between 40 and 150 employees. In worksites with >150 employees, baseline measurements were collected on a random selection of 125 employees. For the worksite to qualify to be randomized, at least 65% of employees sampled had to have completed the self-administered survey. Once the survey response rate reached 65%, a subgroup of employees was randomly selected to participate in additional measurements, which included a blood draw for determining total cholesterol and biomarkers of obesity sequelae, 24-hour dietary recall, and completion of a 7-day pedometer log to estimate steps taken in 1 week. Information on the worksites was assembled on baseline response rate, worksite size, portion of women, portion sedentary, and whether or not the SIC code was non-classifiable. The shortest distance between worksites was calculated for all possible pairs using these five variables. The best matching pairs were chosen based on the shortest combined distance, and then one worksite per pair was randomized into the immediate intervention arm or delayed intervention (comparison arm).

Assessment of BMI

Our main outcome measure for this manuscript is BMI, calculated from self-reported weight and height from the first of two independent cross-sectional samples of employees at each site. Measured height and weight were also obtained from employees at baseline using one of three methods: at the time of a proctored survey administration by providing the machines in a separate space convenient to the employees, at a separate height and weight clinic held on a different day at the worksite, or at the time of the intensive assessment for those employees not previously measured. These data are not yet available for analysis.

Assessment of Dietary Behavior

Several studies have demonstrated that certain dietary choices and eating behaviors are strongly related to healthy eating and body weight, including fast food meals (293031), TV viewing (30,32), eating while doing other activities (32), fruit and vegetable intake (33), and sweetened beverage consumption (32). Rather than assess caloric intake directly, we have included assessment of four of these index behaviors: usual daily servings of fruits and vegetables (34), frequency of eating at fast food restaurants (31), frequency of drinking sodas (29), and eating while doing another activity such as watching television (32). The utility of these behavioral indices as markers of energy intake or obesigenic behavior is suggested by Liebman et al. (32), who showed that sweetened beverage consumption and watching TV were independently associated with overweight and obesity, after adjusting for demographic factors and other behaviors. The PACE project will be able to evaluate their use in relation to energy intake from the 24-hour recall and measured weight for height. These data are not yet available.

Assessment of Physical Activity

Free-time physical activity of at least 10 minutes was assessed using a modification of the Godin questionnaire (35). The Godin leisure-time exercise questionnaire has been widely used in clinical and epidemiological studies, including in assessment of physical activity (363738394041). Godin and Shephard (35) have shown that it is both reliable (the test-retest correlation coefficient ranged from 0.48 for light activity to 0.94 for strenuous activity) and valid in relation to maximal oxygen consumption. More recent studies have confirmed both reliability (42,43) and validity (424344). The questionnaire provides direct estimates of frequency of vigorous, moderate, and light exercise, and we also combined the frequencies into an intensity-weighted frequency score (metabolic equivalent frequency per week). Frequency of sweat-induced exercise was also assessed (35).

Statistical Methods

Employee characteristics were examined for differences between intervention and control groups. For these descriptive analyses, there was no adjustment for random worksite characteristics. The dataset was then restricted to employees with non-missing values for age, race, education, physical activity, and dietary behaviors (86% of the dataset). Using log (BMI) as the dependent variable, linear mixed models were conducted separately by gender to explore the associations of BMI with physical activity and dietary behaviors, for fixed age, race, and education effects and random worksite effects. Predicted mean values were obtained using the adjustment variables as distributed in the dataset. The overall mean and 95% confidence limits were calculated adjusting only for random worksite effects.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

The PACE study used a standard recruitment procedure that was originally established for the Seattle 5-A-Day project (45). An SIC code list of 1109 worksites was purchased. Of the 1109 worksites screened for eligibility by the PACE study, only 378 worksites passed the first telephone screen, as illustrated in Table 1. Priority worksites were identified and were sent a letter of interest and study brochure. These 196 worksites have completed the second screening, and only 81 worksites remained eligible. Of these, 34 worksites refused to have an on-site recruitment meeting, nine refused participation, and three others were found to be ineligible after the recruitment meeting, leaving 35 worksites. Therefore, the eligibility adjusted agreement rate is 36%, after the screenings, letter, and recruitment meeting.

Table 1.  Worksite recruitment
  First screenSecond screenMeeting set-upMeetingAgreement (%)
  1. SIC, Standard Industrial Classification.

Worksites from SIC code list1109     
 Ineligible 619    
 Refused 112    
 Eligible 378   77
  Priority location 196    
   Ineligible  115   
   Eligible  81  100
    Refused meeting   34  
    Meeting at worksite or by telephone   47 58
     Ineligible    3 
     Refused    9 
     Eligible to start run-in    3580
Overall agreement rate     36

The first two waves of worksites consisted of nine worksites randomized into the intervention and nine comparison worksites. The average intervention worksite was slightly larger compared with the average comparison worksite (149 and 142 employees, respectively). A greater portion of employees in the intervention were younger (Table 2), with the average employee age in the intervention worksites being 41 ± 12 years (range, 17 to 75 years), and a greater portion of employees in the <25 years old (6.4%) and the >40 years old (42.1%) age categories than in the comparison worksites. The majority of employees at both intervention and comparison worksites were white (74.1% and 80.1%, respectively), which reflects the local population of the Pacific Northwest. The intervention arm had a greater portion of employees who identified as Asian (20.2% vs. 11.6%). One worksite randomized to the intervention had a workforce that was >40% Asian. There were also fewer employees in the intervention worksites who reported an annual household income greater than $50,000 per year.

Table 2.  Baseline demographic characteristics of employees at PACE worksites (Waves 1 and 2)
 Intervention (N = 9) (n = 865)1Control (N = 9) (n = 768)
  • PACE, Promoting Activity and Changes in Eating; GED, general equivalency diploma.

  • *

    N, number of worksites; n, number of employees.

  • Other is self-specified.

 n(%)n(%)
Age (yrs)    
 <2554(6.4)18(2.4)
 25 to 39352(42.1)267(35.1)
 40 to 54304(36.3)346(45.5)
 55+127(15.2)129(17.0)
Gender    
 Women421(49.8)357(46.8)
 Men424(50.2)406(53.2)
Race    
 White615(74.1)602(80.1)
 Black or African American24(2.9)28(3.7)
 Asian168(20.2)87(11.6)
 Hawaii or Pacific Islander12(1.5)10(1.3)
 Native American, Alaskan Native2(0.2)9(1.2)
 Other9(1.1)16(2.1)
Hispanic ethnicity    
 Yes42(5.0)37(5.0)
 No791(95.0)699(95.0)
Education    
 Less than high school66(7.9)18(2.4)
 High school graduate or GED137(16.3)137(18.0)
 Technical college109(13.0)102(13.4)
 College429(51.1)376(49.4)
 Postgraduate or professional degree98(11.7)128(16.8)
Household income    
 <$25,00063(8.9)30(4.4)
 $25,000 to $49,999208(29.5)153(22.4)
 $50,000 to $74,999160(22.7)162(23.7)
 $75,000 to $100,000130(18.4)144(21.1)
 >$100,000145(20.5)194(28.4)

The average (self-reported) employee BMI was similar between intervention and comparison worksites at baseline (Table 3). Average employee free-time intensity-adjusted physical activity frequencies were not significantly different, nor were eating behaviors between intervention and comparison groups. Employees in worksites randomized to the intervention reported similar eating habits to participants at comparison worksites. Average fast food meals per month and servings of fruit and vegetables per day were comparable.

Table 3.  Baseline BMI and behaviors of employees at PACE worksites (Waves 1 and 2)
 Intervention (N = 9) (n = 865)*Control (N = 9) (n = 768)
  • PACE, Promoting Activity and Changes in Eating; SD, standard deviation.

  • *

    N, number of worksites; n, number of employees.

BMI (kg/m2)n(%)n(%)
 <18.512(1.5)7(1.0)
 18.5 to 24.99314(39.4)272(38.2)
 25 to 29.99285(35.8)239(33.6)
 30+186(23.3)194(27.2)
Servings fruits/vegetables per day    
 <119(2.2)18(2.4)
 1 to 2375(43.9)344(45.4)
 3 to 4317(37.1)271(35.8)
 5+143(16.7)124(16.4)
 Mean(SD)Mean(SD)
Fast food meals per month2.18(2.71)2.23(2.86)
Free time activity score29.01(24.00)29.92(34.07)
BMI (kg/m2)27.03(5.83)27.59(6.25)
Servings fruits/vegetables per day2.93(1.72)2.89(1.73)

Linear mixed models on log(BMI) showed that quintiles of intensity weighted free-time physical activity frequency are significantly associated with BMI (adjusted for age, race, and education) (Tables 4 and 5). Women in the least active quintile had an average BMI of 28.1 kg/m2 (95% confidence interval, 27.2 to 29.2) compared with 25.1 kg/m2 (95% confidence interval, 24.0 to 26.2) for women reporting a free-time exercise score in the uppermost quintile (p value for trend < 0.001, measuring activity score as a continuous variable). Average BMI among men demonstrated a similar trend (p = 0.04).

Table 4.  Mean BMI by baseline reported physical activity and dietary behaviors (females)
inline image
Table 5.  Mean BMI by baseline reported physical activity and dietary behaviors (males)
inline image

Increasing fast food meals (p < 0.001, p = 0.003) and soft drinks (p < 0.001, p < 0.001) consumed were both significantly associated with higher BMIs among men and women, respectively. Greater daily servings of fruits and vegetables were associated with lower BMI among both genders, although only among men was this trend statistically significant (p < 0.001). Eating while doing other activities was associated with greater BMI among women (p < 0.001).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

The methods of group-randomized trials of fruit and vegetable behavior change can be adapted to trials of eating and physical activity behavior to improve weight maintenance. Theoretical frameworks of demonstrated applicability can be used again in obesity prevention, with details being informed by formative research using elicitation interviews and focus groups during the development phase. Several studies using these methods have been summarized recently (46). Formative research has been used in intervention development to identify salient themes, key behaviors, and gaps in knowledge and to test messages and approaches, similar to what we have done. The importance of tailoring an intervention to the target audience has been stressed, and the use of formative research to identify the pertinent barriers and motivators has been reported for vegetables, fruit, and healthy eating behavior (47). In this report, losing weight was viewed as a strong motivator for eating healthier.

We have found that worksites satisfying eligibility criteria are generally willing to participate in the PACE program and that their employees are willing to participate in surveys and intensive assessments at high rates. This is important for internal validity of the findings from the trial when it is completed. In the current climate of suspicion associated with unsolicited telephone surveys [e.g., low response rates in random digit dialing surveys (48)], using the worksite as a partner in the research affords access for evaluation also.

As expected, we found that the measures of physical activity and eating behaviors we chose to include are associated with baseline BMI. In both men and women, strong associations were found with frequency of soft drink consumption and fast food restaurant use. In men, intake of fruits and vegetables was also strongly associated with BMI, and in women, physical activity was strongly related to BMI, as was frequency of eating while doing another activity. The strong gradient in frequency of consumption of soft drinks found in our study for both men and women is consistent with reports using a similar question in a study of rural communities in Wyoming that showed increased risk of overweight or obesity among regular users of soft drinks or sodas (32). Fast food consumption and TV viewing were also associated with energy intake and body weight in the Wyoming study (32) and also among men and women in Minnesota (29,30).

Limitations of the study include the preliminary nature of the findings reported. The results are based on the first two recruitment waves, and two additional waves are in progress. The choices of measures reflective of dietary and physical activity behaviors that could be expected to be responsive to the intervention and likely to lead to changes in BMI have been one of the challenges in the evaluation part of the design of this study. In a study based on the hypothesis that snacking pattern may be a key factor in obesity, the assessment measure used a grid to describe meal patterns on an ordinary day. Four meal types were available, main meal, light meal/breakfast, snack meal, and drink meal. The aim was to identify all intake occasions, including those consisting only of a drink. The frequency of meals and the number of snack meals was found to correlate strongly with energy intake (49). We have also included this grid, adapted for American snacking patterns in the baseline assessment, but results are not yet available. As is well known, the dietary assessment method used most frequently in large-scale population studies is the Food Frequency Questionnaire, but its estimate of caloric intake is biased (50). Therefore, this method was not included in our surveys. To obtain some estimate of caloric intake, we have included an intensive assessment subsample of willing employees from each participating worksite. For this group, employees are contacted by telephone, unannounced, during off-work hours at each evaluation point. They are asked whether they are willing to complete one 20-minute interview over the next week. If respondents agree, the 24-hour recall interview will be conducted, and data will be entered using the University of Minnesota's Diet Analysis System. Participants are also invited to a clinic at the worksite where blood is drawn for biomarkers of metabolic disturbances related to obesity. Participants who attend the blood draw are also asked to use a pedometer and record exercise in a log for 7 days to assess physical activity. It should be noted that a more precise estimate of BMI is obtained from those employees agreeing to objective measurement of height and weight using a calibrated portable stadiometer. Results from these additional measures are not yet available, but they will allow us to conduct correlation studies of the measures of dietary behavior, physical activity, and self-reported BMI and explore further the relationships between them.

The design of this study is one of its strengths. The intervention holds promise of success in obesity prevention because it invokes multiple levels of influence on individual behavior. Motivators and skills introduced in the intervention relate to the focus on fruit and vegetable consumption, to healthy food choices, and to increased frequency of physical activity, all of which are associated with BMI in either men or women or both at baseline. The intervention effect will be evaluated using a conservative, rigorous approach, namely a group-randomized trial, with repeat cross-sectional surveys and assessment of employees at baseline and 2-year follow-up.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

This study was supported by National Institutes of Health Grant HL079491. We thank Dale McLerran for advice on statistical analysis. We are grateful to the liaisons at collaborating worksites and to the employees participating in the surveys.

Footnotes
  • 1

    Nonstandard abbreviations: PACE, Promoting Activity and Changes in Eating; CHIP, counting hundreds impact point; EAB, employee advisory board; SIC, Standard Industrial Classification.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  • 1
    U.S. Department of Health and Human Services. Overweight and Obesity: Health Consequences. http:www.surgeongeneral.govtopicsobesitycalltoactionfact_consequences.htm (Accessed March 1, 2004).
  • 2
    Allison, DB, Zannolli, R., Faith, MS, et al. Weight loss increases and fat loss decreases all-cause mortality rate: results from two independent cohort studies. Int J Obes Relat Metab Disord.. 23: 1999; 603611.
  • 3
    U.S. Department of Health and Human Services. Healthy People 2010: Understanding and Improving Health 2001; U.S. Department of Health and Human Services, Washington, DC.
  • 4
    Centers for Disease Control and Prevention. Overweight and Obesity: Home; http:www.cdc.govnccdphpdnpaobesityindex.htm (Accessed January 30, 2007).
  • 5
    U.S. Department of Labor Statistics. Current Employee Statistics (CES); http:www.bls.govces (Accessed January 23, 2007).
  • 6
    Sorensen, G., Stoddard, AM, LaMontagne, AD, et al. A comprehensive worksite cancer prevention intervention: behavior change results from a randomized controlled trial (United States). J Public Health Policy. 24: 2003; 525.
  • 7
    Glasgow, RE, Sorenson, G., Giffen, C., Shipley, RH, Corbett, K., Lynn, W. Promoting worksite smoking control policies and actions: the Community Intervention Trial for Smoking Cessation (COMMIT) experience: The COMMIT Research Group. Prev Med.. 25: 1996; 186194.
  • 8
    Biener, L., Glanz, K., McLerran, D., et al. Impact of the Working Well Trial on the worksite smoking and nutrition environment. Health Educ Behav.. 26: 1999; 478494.
  • 9
    Campbell, MK, Tessaro, I., DeVellis, B., et al. Effects of a tailored health promotion program for female blue-collar workers: health works for women. Prev Med.. 34: 2002; 313323.
  • 10
    Beresford, SA, Thompson, B., Feng, Z., Christianson, A., McLerran, D., Patrick, DL Seattle 5 a Day worksite program to increase fruit and vegetable consumption. Prev Med.. 32: 2001; 230238.
  • 11
    Buller, DB, Morrill, C., Taren, D., et al. Randomized trial testing the effect of peer education at increasing fruit and vegetable intake. J Nat Cancer Inst.. 91: 1999; 14911500.
  • 12
    Sorensen, G., Stoddard, A., Peterson, K., et al. Increasing fruit and vegetable consumption through worksites and families in the Treatwell 5-a-day study. Am J Public Health. 89: 1999; 5460.
  • 13
    Sorensen, G., Stoddard, A., Hunt, MK, et al. The effects of a health promotion-health protection intervention on behavior change: the WellWorks Study. Am J Public Health. 88: 1998; 16851690.
  • 14
    Elliot, DL, Goldberg, L., Duncan, TE, et al. The PHLAME firefighters’ study: feasibility and findings. Am J Health Behav.. 28: 2004; 1323.
  • 15
    Tilley, BC, Glanz, K., Kristal, AR, et al. Nutrition intervention for high-risk auto workers: results of the Next Step Trial. Prev Med.. 28: 1999; 284292.
  • 16
    Hennrikus, DJ, Jeffery, RW Worksite intervention for weight control: a review of the literature. Am J Health Promot.. 10: 1996; 471498.
  • 17
    Haines, DJ, Davis, L., Rancour, P., Robinson, M., Neel-Wilson, T., Wagner, S. A pilot intervention to promote walking and wellness and to improve the health of college faculty and staff. J Am Coll Health. 55: 2007; 219225.
  • 18
    Eves, FF, Webb, OJ, Mutrie, N. A workplace intervention to promote stair climbing: greater effects in the overweight. Obesity (Silver Spring). 14: 2006; 22102216.
  • 19
    Briley, ME, Montgomery, DH, Blewett, J. Worksite nutrition education can lower total cholesterol levels and promote weight loss among police department employees. J Am Diet Assoc.. 92: 1992; 13821384.
  • 20
    Sumner, SK A weight control and nutrition education program for insurance company employees. J Nutr Educ.. 18: 1986; S60S62.
  • 21
    Emmons, KM, Linnan, LA, Shadel, WG, Marcus, B., Abrams, DB The Working Healthy Project: a worksite health-promotion trial targeting physical activity, diet, and smoking. J Occup Environ Med.. 41: 1999; 545555.
  • 22
    Jeffery, RW Minnesota studies on community-based approaches to weight loss and control. Ann Intern Med.. 119: 1993; 719721.
  • 23
    Stokols, D. Establishing and maintaining healthy environments: toward a social ecology of health promotion. Am Psychol.. 47: 1992; 622.
  • 24
    Jacobson, MI, Yenney, SL, Bisgard, JC An organizational perspective on worksite health promotion. Occup Med.. 5: 1990; 653664.
  • 25
    Thompson, B., Hannon, PA, Bishop, SK, West, BE, Peterson, AK, Beresford, SA Factors related to participatory employee advisory boards in small, blue-collar worksites. Am J Health Promot.. 19: 2005; 430437.
  • 26
    Minkler, M., Thompson, M., Bell, J., Rose, K. Contributions of community involvement to organizational-level empowerment: the federal Healthy Start experience. Health Educ Behav.. 28: 2001; 783807.
  • 27
    McFarlane, J., Fehir, J. De Madres a Madres: a community, primary health care program based on empowerment. Health Educ Q.. 21: 1994; 381394.
  • 28
    US Census Bureau 1987 SIC Matched to 1997 NAICS, Major Groups (2-Digit) http:www.census.govepcdnaicsnsic2ndx.htmS3 (Accessed May 29, 2007).
  • 29
    French, SA, Harnack, L., Jeffery, RW Fast food restaurant use among women in the Pound of Prevention study: dietary, behavioral and demographic correlates. Int J Obes Relat Metab Disord.. 24: 2000; 13531359.
  • 30
    Jeffery, RW, French, SA Epidemic obesity in the United States: are fast foods and television viewing contributing?. Am J Public Health. 88: 1998; 277280.
  • 31
    Pereira, MA, Kartashov, AI, Ebbeling, CB, et al. Fast-food habits, weight gain, and insulin resistance (the CARDIA study): 15-year prospective analysis. Lancet. 365: 2005; 3642.
  • 32
    Liebman, M., Pelican, S., Moore, SA, et al. Dietary intake, eating behavior, and physical activity-related determinants of high body mass index in rural communities in Wyoming, Montana, and Idaho. Int J Obes Relat Metab Disord.. 27: 2003; 684692.
  • 33
    He, K., Hu, FB, Colditz, GA, Manson, JE, Willett, WC, Liu, S. Changes in intake of fruits and vegetables in relation to risk of obesity and weight gain among middle-aged women. Int J Obes Relat Metab Disord.. 28: 2004; 15691574.
  • 34
    Thompson, B., Demark-Wahnefried, W., Taylor, G., et al. Baseline fruit and vegetable intake among adults in seven 5 a day study centers located in diverse geographic areas. J Am Diet Assoc.. 99: 1999; 12411248.
  • 35
    Godin, G., Shephard, RJ A simple method to assess exercise behavior in the community. Can J Appl Sport Sci.. 10: 1985; 141146.
  • 36
    Ainsworth, BE, Berry, CB, Schnyder, VN, Vickers, SR Leisure-time physical activity and aerobic fitness in African-American young adults. J Adolesc Health. 13: 1992; 606611.
  • 37
    Godin, GA, Colantonio, GM, Davis, GM, Shephard, RJ, Simard, C. Prediction of leisure time exercise behavior among a group of lower-limb disabled adults. J Clin Psychol. 42: 1986; 272279.
  • 38
    Godin, GA, Desharnais, R., Jobin, J., Cook, J. The impact of physical fitness and health-age appraisal upon exercise intentions and behavior. J Behav Med.. 10: 1987; 241250.
  • 39
    Godin, GA, Shephard, RJ, Colantonio, A. Children's perception of parental exercise: influence of sex and age. Percept Motor Skills. 62: 1986; 511516.
  • 40
    Godin, GA, Shephard, RJ Psycho-social predictors of exercise intentions among spouses. Can J Appl Sport Sci.. 10: 1985; 3643.
  • 41
    Noreau, LR, Shephard, RJ, Simard, C., Pare, G., Pomerleau, P. Relationship of impairment and functional ability to habitual activity and fitness following spinal cord injury. Int J Rehabil Res.. 16: 1993; 265275.
  • 42
    Jacobs, DR, Jr, Ainsworth, BE, Hartman, TJ, Leon, AS A simultaneous evaluation of 10 commonly used physical activity questionnaires. Med Sci Sports Exerc.. 25: 1993; 8191.
  • 43
    Sallis, JF, Buono, MJ, Roby, JJ, Micale, FG, Nelson, JA Seven-day recall and other physical activity self-reports in children and adolescents. Med Sci Sports Exerc.. 25: 1993; 99108.
  • 44
    Miller, DJ, Freedson, PS, Kline, GM Comparison of activity levels using Caltrac accelerometer and five questionnaires. Med Sci Sports Exerc.. 26: 1994; 376382.
  • 45
    Thompson, B., van Leynseele, J., Beresford, SA Recruiting worksites to participate in a health promotion research study. Am J Health Promot.. 11: 1997; 344351.
  • 46
    Gittelsohn, J., Steckler, A., Johnson, CC, et al. Formative research in school and community-based health programs and studies: “state of the art” and the TAAG approach. Health Educ Behav.. 33: 2006; 2539.
  • 47
    Strolla, LO, Gans, KM, Risica, PM Using qualitative and quantitative formative research to develop tailored nutrition intervention materials for a diverse low-income audience. Health Educ Res.. 21: 2006; 465476.
  • 48
    Kalton, G., Piesse, A. Survey research methods in evaluation and case-control studies. Stat Med.. 26: 2007; 16751687.
  • 49
    Forslund, HB, Lindroos, A., Sjöström, L., Lissner, L. Meal patterns and obesity in Swedish women: a simple instrument describing usual meal types, frequency and temporal distribution. Eur J Clin Nutr. 2002; 740747.
  • 50
    Subar, AF, Kipnis, V., Troiano, RP, et al. Using intake biomarkers to evaluate the extent of dietary misreporting in a large sample of adults: the OPEN study. Am J Epidemiol. 158: 2003; 113.