“Pizza Is Cheaper Than Salad”: Assessing Workers' Views for an Environmental Food Intervention


Division of Nutritional Sciences, Cornell University, 377 MVR, Ithaca, NY 14853-4401. E-mail: cmd10@cornell.edu


Objective: “Images of a Healthy Worksite” aims to provide easy access to healthful foods and to reduce sedentarism at the worksite—to prevent weight gain. Formative research for the nutrition intervention component was aimed at gaining a broad understanding of the sociocultural role of food and eating among workers and worker perspectives on socially feasible and culturally acceptable environmental intervention strategies.

Research Methods and Procedures: Using an adapted PRECEDE health planning model, we conducted ecological, educational, environmental, and administrative assessments at the worksite. Through 15 in-depth interviews, five focus groups, and community mapping at two sites with 79 administrators, managers, workers, and food service personnel (51% men, 82% white), we assessed workers' perspectives on physical, sociocultural, economic, and policy environments. Data were coded for predisposing, enabling, and reinforcing factors related to intervention strategies in vending, cafeteria, catering, and informal food environments. After classification for reach, intensity, and sustainability, objectives and evaluation plans were developed for each highly ranked strategy.

Results: Key sociocultural factors affecting food and eating included: stress-related eating in a downsizing workplace, enthusiasm for employer-sponsored weight gain prevention efforts that respect personal privacy, and the consequences of organizational culture on worker access to the food and eating environment. Workers supported healthier cafeteria and catering options, bringing healthful foods closer, and labeling of healthful options.

Discussion: We provide a practical and systematic approach to formative research and assess the interrelatedness of the physical, policy, economic, and sociocultural factors that affect environmental worksite interventions to prevent weight gain among employees.


High rates of overweight and obesity among U.S. adults (1) demand the design and implement action of environmental approaches to promote behavioral change that reach a large number of working adults. The goal of this analysis was to conduct formative research to design a socially feasible, culturally acceptable, and methodologically sound intervention trial for weight gain prevention at a large manufacturing company.

The intervention “Images of a Healthy Worksite” is a group-randomized trial to test weight gain prevention strategies that create synergy between a worksite environment and workers' food and physical activity choices. This paper reports on the use of formative educational, ecological, environmental, and administrative assessments of food and eating at the worksite to develop a nutrition intervention strategy, one part of a larger project.

Obesity has a multi-factorial etiology. Although overweight and obesity result from an energy imbalance, body weight is the result of genetics, behavior, and environments. Food choices and physical activity patterns are influenced by individual, environmental, and societal factors (2). There is general agreement that a contributor to the obesity epidemic is an obesogenic environment that encourages overeating and discourages physical activity (2,3,4). The worksite is a feasible and potentially effective environment to implement interventions reaching a large number of working adults. Relevant worksite characteristics include access to energy-dense foods, large portion sizes, policies and practices related to food access and worker movement, social norms related to times and places for eating, access to recreational facilities, and sedentary occupations.

Worksite interventions commonly target individual behavior (5). Recently, investigations to improve food and eating at the worksite have targeted one or two aspects of the food environment (2,6,7). A recent review of clinical trials of worksite interventions found strong evidence of the effectiveness of environmental modifications on fruit, vegetable, and fat intake (8). Environmental approaches offer promising opportunities for intervention but need rigorous testing (91011), and they benefit from active worker involvement in formative research that informs study design and implementation (12,13).

Formative research is conducted before an intervention to understand a target population or setting and inform an intervention appropriate to that context (13,14). In workplaces, formative research has been useful in understanding complex worksites with multiple settings and worker types (15,16). Lacking in the formative research literature is information about appropriate processes for collecting and applying formative research to intervention design (13). This study makes a contribution to filling that gap.

This formative research was guided by the PRECEDE-PROCEED model for health program planning and promotion, adapted for this project (17). The work described here focused on the planning or PRECEDE portion of the model, specifically on the educational, ecological, environmental, and administrative assessments at the worksite and the use of assessment data to design the nutrition intervention strategies. This research was part of a larger project that also collected individual data about workers’ weight, health, dietary behavior, knowledge, and attitudes about diet and health (epidemiological assessment) and assessment of the physical activity environment (data not shown).

Aspects of the PRECEDE model included in this analysis are shown in Figure 1 (17). The social assessment includes participants’ views of the relationships among work, health, and weight as they relate to quality of life. The PRE educational and ecological assessment phase of the PRECEDE model investigates: predisposing (P),1 reinforcing (R), and enabling (E) factors. P factors include circumstances and situations existing before a behavior occurs and the motivation for that behavior. R factors include positive or negative feedback that follows a behavior. E factors include skills, resources, and barriers that help or hinder a behavioral or environmental change. The environmental assessment phase of the PRECEDE model includes aspects of the environment that support health behaviors. We were guided in this phase by Swinburn's (18) model of four key environments (physical, sociocultural, economic, and policy) that should be considered in relation to obesity. In the administrative and policy assessment phase of the PRECEDE model, we assessed organizational policies, resources, and settings that could support or interfere with the intervention (17).

Figure 1.

Study design: adapted from the PRECEDE-PROCEED Model for Health Program Planning and Evaluation (17). Aspects of the model described in this paper are in bold. Aspects of the model included in this project, but not presented here, are shown in italics. Aspects not included are regular text. (Used with permission.)

The objectives of this formative research were to investigate multiple perspectives on the barriers and enablers of healthy eating from various worksite stakeholders, develop an understanding of the sociocultural role of food and eating in the lives of the workers, and examine the influence of physical, economic, policy, and sociocultural environments on the consumption of food at work for the purpose of developing intervention strategies.

Research Methods and Procedures

The intervention and control worksites are multiple workplaces within a large manufacturing company in a metropolitan area in the Northeastern U.S. The worksites include light and heavy manufacturing facilities and research and development settings. The company employs ∼8000 people in the area. At the time of the assessment, the company was undergoing significant changes in direction accompanied by a series of layoffs.

We used three methods in a rapid anthropological assessment to meet the research objectives: individual interviews, focus groups, and community mapping with key informants (19). Rapid assessment procedures were originally developed to provide practical anthropological methods for health workers assessing health behaviors in community settings (20). Respondents, who were 18 years of age or older and employed full time at the site, were recruited for all three methods through worksite flyers and snowball sampling. Respondents were recruited from a variety of job types to assess the opinions of workers from different layers of the company.

Two experienced qualitative interviewers conducted 15 open-ended, in-depth, interviews with company executives, managers, workers, worksite nurses, and food service personnel. The sampling plan for the interviews was emergent; themes and issues from earlier interviews informed the choice of subsequent participants. Interviewers queried workers’ perspectives on work demands, health and weight, food, eating and physical activity, possible solutions to overcoming workplace barriers to healthy eating and physical activity, and previous attempts to introduce healthy foods in the worksite. Interviewers allowed the respondents maximum latitude in introducing and expanding on ideas that were important to them. The interviews took place in private offices or worksite conference rooms over about a month. Some interviews were conducted during the workday, others at the end of a shift. Interviews were audio-recorded, transcribed verbatim, and verified by interviewers.

We subsequently convened five focus groups with groups of workers from study sites. Focus group composition was homogeneous with respect to social factors that may influence attitudes, beliefs, and daily practices (gender, race, ethnicity, and work place occupation). Separate focus groups were conducted for male and female workers, managers, supervisors, and workers with sedentary and active jobs (57 individuals). The findings of the individual interviews framed the focus group discussion topics: perceptions about overweight and obesity, weight and productivity, changes needed in the worksite to promote healthy weight, and satisfaction with current food options. A trained researcher/moderator conducted each of the focus groups assisted by an analyst who took notes, made observations, tape-recorded the sessions, and wrote up the participant remarks and interactions. Moderators were women, except for a male moderator for the male worker group. Focus groups lasted 95 minutes, included a light meal, and took place at the end of the workday in a nearby hotel. Although designed as a data collection tool, the focus group interactions also increased participants’ awareness of the issues at hand.

We implemented the third method through mapping of community environments within the worksites with the assistance of key informants (19). The mapping process included appraisal of the physical space and the social interactions within that space relative to food and eating. In each setting, the investigators mapped physical, economic, policy, and sociocultural environments for food and eating. This included who ate where and with whom; what, when, and how they ate; and the cost of food choices in terms of money and time. Two investigators independently mapped the facilities, probing for barriers to and opportunities for healthy eating, with the help in each case of a long-time worker. We carried out mapping in two sites within the company. In each case, mapping included cafeteria, vending, catering, neighborhood, and work settings. All study participants provided written consent according to a protocol approved by the institutional review boards of both of the academic institutions involved and the company. All study participants received a modest monetary incentive for their participation.

Data Analysis

Data analytic procedures included social analysis, environmental analysis, educational and ecological analysis, and administrative analysis (17). We used the data from interviews, focus groups, and community mapping for these analyses. In social analysis, data were open coded (21) for participant's perceptions about work, health, diet, and body weight (data not shown). Widespread concerns about diet, health, and weight and their links to workplace food and eating environments included food availability, work pressures, and eating norms, confirming that the proposed intervention addressed a felt need. Environmental analysis included coding for the physical, sociocultural, economic, and policy elements of the worksite environment that were described by workers as influences on their food and eating.

Steps in the educational and ecological analysis included coding, classification, matching, and writing of outcome objectives for P, R, and E factors. All investigators read all of the interview and focus group transcripts and the written summaries of the mapping activities. One investigator coded data from interviews, focus groups, and mapping for PRE factors related to strategies for environmental interventions. After coding, emergent strategies were separated into categories (cafeteria, vending, catering, informal food brought by workers). Each emergent strategy was then independently classified as high, medium, or low on three criteria: reach (ability to reach large numbers of workers), intensity (frequency of exposure), and sustainability (cost and support in the worksite). Classifications were assessed by one researcher and confirmed by another, working together until consensus was achieved. Outcome objectives with supporting PRE factors were developed for each strategy that was classified as having high reach, intensity, and sustainability. Peer debriefing (22) was conducted through regular research meetings to review proposed coding strategies and emerging themes and through presentation of findings to the whole research team. Peer review of strategies was informed by evidence-based determinants and strategies for environmental interventions from the published literature (data not shown).

We used the same formative data for the environmental and administrative assessment to identify all major routes of access to foods in the workplace environment and to develop a company food access organizational chart. Using an ecological approach (17) we matched our behavioral objectives from the PRE analysis with the administrative structures on the food access organization chart, selecting the channels and mediators through which each intervention strategy could be implemented.


Seventy-nine workers (40 men and 39 women) participated in the study. Similar to worksite demographics, 82% of participants were white, 4% were African American, 1% was Asian, and 1% was mixed race; 12% did not identify a racial group. Ninety-one percent were non-Hispanic, 2.5% were Hispanic, and 6.5% identified no ethnic group. The average age of participants was 48.6 years (worksite average = 48 years). Approximately 49% of the participants were women. Although data on gender distribution are not available for the whole worksite, 37% of the employees who participated in baseline data collection (n = 2885) were women. Participants included administrators, managers, workers, and food service personnel. Only two (2.5%) of the participants reported working a production or manufacturing job (43% did not report job type). With this exception, participants in the formative research were similar in demographic characteristics to the workers who participated in the baseline data collection (data not shown).

Participants described an aging workforce with little turnover that was “wearing down,” with more than one-half having “major health issues” including overweight and obesity. Major barriers to health were described as work schedules, on-going layoffs that left those remaining “doing the work of 5 people,” stress-related eating, and an organizational culture ‘shared values and expectations by members of the organization (23)’ that rewarded working through breaks. One participant described significant layoffs during the preceding week at her worksite. “This last week was a tough week, they were so stressed; a lot of people weren't eating right.” Another used food to relieve stress at work. “It's like a quick pat on the shoulder or something like that. You've kept up with your day; you've done a great job, here have a cookie ‘laughs’.”

Participants expressed support for worksite opportunities to stay healthy and prevent obesity and concerns about discrimination based on weight status. Focus group participants did not want the company to have a “record of my weight.” They said the word obesity would embarrass people and turn them away. Many longed for past employer-sponsored wellness initiatives.

Study participants described physical, economic, sociocultural, and policy environments during interviews, focus groups, and community mapping. Participants described the physical characteristics of food available in their work environments, e.g., “I don't eat at the … cafeterias because … I'd say 80% of the choices available are unhealthy” and “The fresh fruit is not in the ‘cafeteria’ flow.” Workers described the economic aspects of the foods offered, e.g., “Pizza is cheaper than salad.” They commented on ways that the organizational culture provided an environment for eating: e.g., “We work through lunch.” These workers described how the work-related strain due to recent layoffs had an impact on food choices and eating as they expressed reluctance to leave their work stations to go to the cafeteria in case a new order came in or when they talked about using brownies as stress relievers or skipping lunch. “You're supposed to take one ‘a lunch break’ but it's not worth it.” The policy environment was evoked when workers described changes they would like to see in the foods available at work, including changes related to nutrition and health (e.g., spinach on the salad bar, vegetables cooked without fat, more low-fat salad dressings), cost (e.g., less expensive fruits), convenience (e.g., more grab-and-go items), and social events (e.g., replacing cake at celebrations with fruit).

The administrative and policy assessment revealed that there were four major routes of exposure to food at the worksite: cafeterias, vending machines, catered food at meetings, and informal food. Informal food was food brought in by individuals and kept in their personal pantries at work or occasional treats or meals purchased by supervisors during meetings at neighborhood restaurants. All food in the first three categories was controlled by the staff of the external food service management contractor (Figure 2); these were the main targets of the environmental intervention. The informal food category was not a main intervention target because changes in these foods would not reach large numbers of workers.

Figure 2.

Organization chart for worksite food access. Units that are not part of the company are shown in italics. Contractual relationships are shown as dotted lines.

A company-specific table of objectives, PRE factors, strategies, and evaluation plans was developed for each of three main food access sources: cafeterias, vending machines, and catered food at meetings (Tables 1 2 to 3). An example from worksite catering illustrates how intervention objectives were linked to PRE factors and participant inputs. Interview participants said they would like to have healthier menu options for the frequent meetings at work. This request was related to their own P motivations including concerns about health, a desire to lose weight, awareness of nutrition and health guidelines, general health consciousness, and an expressed desire to make food choices informed by nutrition and health information. An R factor was acknowledgment of their own aging and potential for health problems. Potential E factors supporting healthy food at meetings included willingness to eat healthy foods if they were tasty, reasonably priced, and of good quality and the control of all food at the worksite by a single source that could institutionalize such change. An objective was developed that workers would choose healthier foods at worksite meetings if healthy menu options were available and met their criteria for taste, cost, and quality. Intervention strategies included addition of new nutrient-dense and lower calorie catering menu options, marketing of the healthy options to meeting planners and employees, and point-of-sale icons denoting healthy options. This strategy would be evaluated by tracking orders for the healthy menu options and through a user satisfaction survey of planners and consumers of food at these meetings. In Tables 4 and 5, we show administrative and policy-relevant and social marketing-objectives that emerged from the formative research.

Table 1.  Cafeteria environment
ObjectivesIntervention strategiesPRECEDE factors
  1. P, predisposing; R, reinforcing; E, enabling.

1. Workers will purchase healthier entrees if they are available, appealing, satisfy preferences and health concerns, and are priced at or below other optionsEducate chefs and managersP, food service staff want satisfied customers
 Establish healthy menu guidelines 
 Price healthy selections ≤ othersP, active chronic disease among workers
 Monitor food preferences of all workersP, worker interest in weight loss
  P, some health conscious workers
Evaluate: per unit sales monitor menu monitor food satisfactionProvide equipment to prepare healthy foodsR, colleagues share unhealthy foods
  R, worker age is increasing
 Taste testing in/outside of cafeteriaE, cost, taste, and quality influence choices
 Incentives for healthy options 
  E, cost influences food procurement
  E, food access governed by contractor
  E, menu decisions driven by infrastructure and cost
  E, lack of fresh fruits/vegetables in cafeteria
  E, menu decisions by cafeteria managers
2. Chefs and managers will increase self-efficacy for preparing and serving healthy foodsEducate chefs and managers 
 Recipe competitions for chefs 
 Recognition for chefs 
Evaluate: chef surveys  
3. Workers will purchase healthier foods if they can identify the nutritional value/ingredientsPoint-of-sale icons denoting healthy optionsP, workers want to make choices based on ingredient/nutrition information
 Educate cafeteria staff on nutrition, weight, and healthP, food service staff want satisfied customers
Evaluate: sales  
 Taste testing in/outside of cafeteriaP, active chronic disease among workers
 Incentives for healthy optionsP, worker interest in weight loss
  P, some health-conscious staff
  R, worker age is increasing
4. Workers will select low-calorie salad dressings if they have choicesAdd more low-calorie salad dressingsE, menu decisions by cafeteria managers
  P, worker interest in weight loss
 Point-of-sale icons denoting healthy optionsP, active chronic disease among workers
Evaluate: use of low-calorie dressings P, some health-conscious workers
5. Workers will purchase more fresh fruits and vegetables if the quality is good and the price is ≤ othersPrice healthy foods ≤ othersE, lack of fresh fruits/vegetables in cafeteria
 Offer more fresh fruits and vegetables in cafeteria 
  E, menu decisions by cafeteria managers
 Point-of-sale icons denoting healthy options and nutrient informationP, worker awareness of strategies to attain healthy weight
Evaluate: sales  
  P, some health-conscious workers
 Host farmers' marketE, cost, taste, quality influence decisions
 Offer healthier salad bar choices 
6. Workers will purchase smaller entrées and sandwiches if offered and proportionally pricedOffer menu items in smaller sizesE, cost, taste, and quality influence decisions
 Point-of-sale icons denoting healthy menu optionsP, worker interest in weight loss
  P, worker awareness of strategies to attain healthy weight
Evaluate: sales  
  P, participation in weight management programs
7. Workers on late shifts or working far from or lacking cafeterias will increase healthy food purchases if healthy options are accessibleTake orders for meals-to-go picked up after hours.E, worker schedules are constrained by out-of-work commitments
 Provide chilled storage for meals-to-go at work Institute “lunch wagons”/farmers’ market wagon in areas lacking a cafeteriaE, cafeteria hours are limited
  E, food access governed by contractor
  R, pressure to work beyond the limits (skip breaks, eat at desk)
Evaluate: sales or repeat customers  
  E, unhealthy vending options
  E, not all buildings have refrigerated vending
  E, long distances to cafeteria for some workers
Table 2.  Vending environment
ObjectivesIntervention strategiesPRECEDE factors
  1. P, predisposing; R, reinforcing; E, enabling.

1. Workers will purchase healthier vending machine foods if available and priced ≤ othersEstablish policy for healthy vending choices and monitor for complianceP, active chronic disease among workers
  P, interest in weight loss among workers
  P, some building staff are health conscious
Evaluate: sales customer satisfactionIncrease healthy vending choicesP, worker awareness of healthy weight strategies
 Price healthy vending options ≤ others 
 Make available vending machines for sale of fruit and low-fat dairy products 
2. Workers will purchase fresh fruit, low-fat dairy, and healthy entrée selections if the quality is good and the price is at or below similar productsPoint-of-sale icons denoting healthy optionsP, workers want to make informed choices based on ingredient/nutrition information
 Incentives for trying healthy optionsP, participation in weight management programs
  P, some building staff are health conscious
Evaluate: sales R, worker age is increasing
  R, pressure to work beyond the limits (skip breaks, eat at desk)
  E, cost, taste, quality influence food choices
  E, cost influences food procurement
  E, food access decisions by contractor
  E, lack of fresh fruits and vegetables in vending
  E, cafeteria hours limited
  E, worker schedules are constrained by out-of-work time commitments
  E, unhealthy vending options
  E, not all sites have refrigerated vending
  E, distances to cafeteria for some workers
Table 3.  Catering environment
ObjectivesIntervention strategiesPRECEDE factors
  1. P, predisposing; R, reinforcing; E, enabling.

1. Workers will eat healthier at meetings and celebrationsSocial marketing to promote healthy meeting optionsP-Active worker chronic disease
  P, interest in weight loss among workers
Evaluate: track orders for catering options user satisfactionPoint-of-sale icons denoting healthy choices on catered menus and food plattersP, some building staff are health conscious
  P, worker awareness of healthy weight strategies
  P, workers desire to make informed choices
  R, increasing worker age
  E, cost, taste, quality influence food choices
  E, food access control by food service
2. Meeting planners will order healthy selectionsAdd healthy options to catering menusP, some building staff are health conscious
Evaluate: track ordersPrice healthy selections ≤ othersR, perceived pressure to work beyond the limits (skip breaks, work through lunch)
 Incentives for those ordering healthy options 
  E, cost, taste, quality influence food choices
Table 4.  Administration and policy environment
ObjectivesIntervention strategiesPRECEDE factors
  1. P, predisposing; R, reinforcing; E, enabling.

1. Workers will be supported in their efforts to make healthy food choices through “buy-in” by company executivesCommunicate with all levels of management of the relationship of weight to workplace safety, productivity, and corporate profitabilityE, perceived lack of corporate concern of worker needs
  R, diminished number of health care staff
Evaluate: manager support of recruiting and participationAlign interventions to existing worksite culture and institutional valuesR, discontinuation of health teams
  R, less staff to produce outputs, increases each person's workload
  R, average worker age is increasing
2. Food service workers will be supported in efforts to satisfy customer requests for healthy food choices through buy-in by food service managementCommunicate with all levels of food service management of the cost benefits associated with offering healthier food choicesE, food access decisions governed by food service contractor
  E, cost influences contractor's decisions
  E, cost, taste, and quality influence consumer decisions
Evaluate: manager support of interventions  
Table 5.  Social marketing environment
ObjectivesIntervention strategiesPRECEDE factors
  1. P, predisposing; R, reinforcing; E, enabling.

1. Workers will receive clear and consistent messages about the nutrition interventionManagers will communicate corporate supportP, active chronic disease among workers
 Market environmental changes broadly 
 Provide workers with nutrition and health information in a form they can useP, acceptance of overweight as normal
Evaluate: number and kind of messages  
  P, interest in weight loss among workers
 worker awareness  
  P, some buildings have health-conscious staff
  P, worker awareness of strategies to attain healthy weight
  P, workers want to make informed choices based on ingredient/ nutrition information
  R, colleagues share unhealthy foods
2. Workers will be recognized and rewarded for positive behavior changeConnect initiative to corporate winning and inclusive culture and the corporate value of recognition and celebrationP, some buildings have health-conscious staff
  P, food service staff want satisfied customers
Evaluate: corporate communications worker satisfactionProvide meaningful incentives 
 Support and recognize positive role modelsR, job tasks are often isolative
  R, some social support among peers (walking, dieting)
  P, participation in weight management programs

Participants in the formative research expressed enthusiasm for several high reach, high-intensity, and modest cost interventions that addressed the company environment to support healthier food choices. These included education of chefs and food service staff to develop their capacity to provide healthful options in cafeterias and at meetings, increases in healthy vending options, healthy menus for foods at meetings, and point-of-sale icons and nutrition labels to identify healthful foods. Workers also expressed support for higher cost intervention strategies that would bring healthful food closer to busy workers such as mini-lunch wagons, meals-to-go, and on-site farmers' markets.

Some responses differed by type of worker. For example, technical workers expressed interest in information about body weight and the nutritional value of food. Managers emphasized personal responsibility for weight and the need to avoid any strategy that would interfere with the workflow. Lower-wage workers were concerned about food costs and equity of food options across sites. All participants described chronic health problems that they connected to extreme work-related strain and poor workplace food choices. All strategies were subsequently presented to and discussed with worker advisory boards at each intervention site in the development of the final intervention design and strategies (data not shown). This last step made it possible to tailor strategies to the concerns of workers in each site.


Our findings provide a demonstration of the value of formative research based on the PRECEDE model to systematically and efficiently collect and assess formative data that was directly translated into intervention strategies at a particular worksite. Although formative research often precedes intervention studies, the contribution of this study to intervention research is the provision of an explicit process for translating formative findings into intervention design elements, helping to fill an identified gap in this area (13,24).

Respondents were aware of and able to articulate specific educational, ecological, administrative, policy, and environmental targets for weight gain prevention intervention within their work areas. These workers described promotion of healthy eating and weight gain prevention as an appropriate and valued worksite activity that was past due. Workers described the important influence of the worksite on their food choices, from the foods that were available and the support and help they received from their coworkers in meeting individual health and nutrition goals. Respondents highlighted the public-private nature of body weight and weight control. Although opportunities to have healthier food choice options were welcomed, respondents were also clear that they did not want employers knowing what they weighed or discriminating against them if they were overweight. Worksite nutrition or weight programs have become common, especially among larger employers (25). These findings suggest that environmental approaches that focus on healthy eating are likely to be more widely accepted by workers than those that target obese workers.

Our findings were dominated by expressions of high levels of job insecurity among workers of this company, attributed to the many past and ongoing layoffs. Both job insecurity and job strain have been related to weight gain and obesity among workers (2627, 28). Several respondents expressed interest in health enhancement activities as one way of coping with the strain; some saw the health promotion project as a positive sign of management interest in their health. This is consistent with evidence that employer changes to reduce occupational risks may encourage workers to participate in health promotion programs (29)

An important feature of the formative research was the ability to concurrently assess features of both physical and sociocultural environments. Environmental interventions that focus solely on the physical environment may miss important opportunities and barriers from the social environment that affect access to resources in the built environment. For example, for workers who are too busy, tired, or worried to leave their work areas to go to the cafeteria, the food choices in the cafeteria may be irrelevant, and alternative solutions will be needed. Work fatigue and overtime work have been associated with weight gain (30).

The triangulation of individual interviews, focus groups, and community mapping provided richer findings than any one of these methods alone. We were able to check the extent of convergence in the focus groups and the mapping process with points made in individual interviews. Some themes such as stress-related eating and concerns about diet and health were common among all three sources, but other themes diverged. For example, past attempts to introduce healthier foods in cafeterias were presented as unappreciated in interviews with food service staff and as inadequate by focus group participants. The interviews allowed us to target high- and midlevel managers, specialty workers such as health workers, and food service workers who may not have been comfortable expressing their opinions in a group setting.

The focus groups allowed us to explore the social feasibility and worker acceptance of proposed intervention strategies and to incorporate social concerns into the intervention. For example, we were able to explore with workers how intervention strategies and timing could address the feelings of insecurity related to company downsizing.

The workplace mapping allowed us to see the food and eating environment firsthand with the assistance of an employee guide who could explain the social conventions of usage. For example, we were able to both observe and ask workers about the individual pantries that they assembled to limit their need to leave their work areas during demanding work periods.

Like other worksite interventions with environmental components that have used the PRECEDE model to develop health-related interventions (15, 16), this formative research strategy proved effective in understanding a complex worksite. One limitation of this research was the lack of information about the dietary intake of workers at the time of the formative phase of the study. This information would have aided judgments about which intervention strategies might have the most impact. We did have information about the relative sales and popularity of various foods elicited from the in-depth interviews with food service staff and other workers. Dietary intake was subsequently assessed as part of baseline data collection, after recruitment into the study (data not shown).

Another study limitation was our lack of success in recruiting more blue collar production workers into our formative research despite a monetary incentive. This may have been due to lack of comfort with research, lack of trust toward researchers, job insecurity, or job demands such as 12-hour shifts that made participation difficult. Unequal access to worksite health programs because of larger political, social, and economic influences on people's lives is also likely to play a role in participation, although participation is more dependent on availability than on worker characteristics (31, 32). Although the sample size and extensiveness was consistent with other studies of this sort (33), lack of participation by production workers raises concerns about the applicability of these findings to all worker groups. It is possible that the office, technical, and management workers who did participate had more positive attitudes toward diet and health. Our experience is not unique; blue collar workers have been difficult to recruit into worksite interventions (29). It is also possible that social desirability bias may have affected worker trust in and disclosure to outside researchers, especially in the environment of this workplace in transition. Subsequent to the analysis presented here, the results of this formative research were shared with worker advisory boards in each intervention site as a form of member check (22) and to obtain additional input into the specifics of the implementation of the intervention (24). Each worker advisory board developed an intervention strategy survey to distribute to the rest of employees in the worksite to minimize the lack of representation of production workers in formative research. Feedback from these groups is aiding in the on-going selection and form of the final intervention delivery, tailored to each site. The participation of production workers in the advisory groups may minimize their lack of representation in the formative research.

There have been calls for environmental interventions at the worksite to promote the health of workers (34). This paper documents a systematic process for using formative research to plan a worksite intervention that is targeted to worker-identified needs and strategies. We demonstrated a model for conducting rapid formative research in a manufacturing worksite to enhance the relevance of interventions among workers. The PRECEDE health planning model provided a practical basis for the collection and analysis of formative individual, group, and observational data to inform environmental intervention strategies. Our approach may be useful to other researchers. Ultimately, the effectiveness of the interventions aimed at improving diets among the workers must be demonstrated through rigorous process and outcome evaluation.


This work was supported by the National Heart, Lung, and Blood Institute (Grant 5R01HL079511 to I.D.F.). We thank the workers, managers, and food service workers who contributed to this study and Elizabeth Holtsclaw, the project coordinator at the time of this research.


  • 1

    Nonstandard abbreviations: P, predisposing; R, reinforcing; E, enabling.