A systematic review of trucking food, physical activity, and tobacco environments and tractor-trailer drivers’ related patterns and practices in the United States and Canada, 1993–2021

Highlights • The built environment is recognized to influence health patterns and practices.• No review has explored trucking food, physical activity, and tobacco environments.• Trucking built environment science is underdeveloped and requires validated tools.• Future research should explore truckers’ views on built environment interventions.• More emphasis on the trucking built environment and health equity is warranted.


Review outcomes
Microsoft® Excel® (2018 Microsoft Corporation, Redmond, WA, USA) was used to guide data extraction for article characteristics (study objective, design, location), environmental data (setting, measurement, and food, PA, and/or tobacco environment results), and data about related perceptions and patterns/practices (recruitment strategy, participant characteristics, qualitative or quantitative measures, and results). Categories for data extraction are presented in the below Tables. A graduate research assistant with a background in community health extracted data from all sources (KK) and co-authors (BH, LB, LM, and CBS) each reviewed extractions for suggested changes. A consensus was reached between two co-authors for all data points. Three topic experts (BH, LB, and LM) constructed a narration of results by topic area for food, PA, and tobacco, respectively, using evidence tables.

Study quality
Research quality was assessed independently between two authors (BH and LB) using the 2018 Mixed-Method Appraisal Tool (Hong et al., 2019;Souto et al., 2015). This tool allows for the assessment of quality among quantitative, qualitative, and mixed-method studies (Pluye and Hong, 2014), which were all captured in the review scope. For each study type, the 2018 Mixed-Method Appraisal Tool prompts seven "Yes", "No", or "Can't Tell" responses to questions about study rationale, design, methods, and conclusions (Hong et al., 2019;Souto et al., 2015). Discrepancies in ratings were discussed until a consensus was reached. Quality ratings are communicated below based on the number of "Yes" determinations regarding study quality indices.

Food environments
Three studies reported data from the measurement of trucking food environments including truck terminals, warehouses, truck stop restaurants, convenience stores, and travel stop vending machines (Apostolopoulos et al., 2011;Lincoln et al., 2018;McGuirt et al., 2019). Lincoln et al. (Lincoln et al., 2018) found most of the sampled restaurant and convenience store settings at truck stops to have fruits or vegetables available for purchase (Lincoln et al., 2018). Overall, however, trucking environments were characterized as 'unhealthy' given products with high amounts of saturated fat, added sugars, and sodium were found widely available (Apostolopoulos et al., 2011;Lincoln et al., 2018;McGuirt et al., 2019).

Physical activity environments
Two studies measured PA environments at truck stops Lincoln et al., 2018). Both found most trucking environments to have showers available Lincoln et al., 2018) and to lack resources and facilities that encourage PA, including walking paths and exercise centers. Apostolopoulos et al.  also assessed communities near truck stop locations and found most were without PA amenities . See Table 2.

Physical activity environment perceptions
Twelve studies assessed truckers' perceptions of PA environments (Apostolopoulos et al., 2013;Holmes and Power, 1996;Layne et al., 2009;Lemke et al., 2016;McDonough et al., 2014;Passey et al., 2014;Staško and Neale, 2007;Turner and Reed, 2011;Versteeg et al., 2018;Wenger, 2008). Although truckers believed a lack of exercise contributed to weight gain and poor health (Holmes and Power, 1996;Whitfield Jacobson et al., 2007) and recognized the importance of PA Versteeg et al., 2018), there were many reported barriers to PA including lack of time, long and inflexible work hours, and fatigue Passey et al., 2014;Turner and Reed, 2011;Wenger, 2008). Poor or no access to PA facilities and equipment was another primary barrier (Apostolopoulos et al., 2013;Passey et al., 2014;Turner and Reed, 2011;Wenger, 2008). In one study, limited parking resulted in needing to stop in unsafe locations and a lack of showering facilities at available rest stops was described , which may influence the likelihood of PA. Finally, some truckers expressed interest in PA and explained they would use exercise equipment, gym memberships, and walking paths if available in the environment Staško and Neale, 2007;Turner and Reed, 2011;Versteeg et al., 2018).

Food, PA, and tobacco patterns or practices
Twenty-seven of 38 studies (71%) had data about truckers' food, PA, and/or tobacco patterns or practices (Table 4).

Food patterns/practices
Twenty-one studies had data about truckers' dietary patterns/   practices Versteeg et al., 2018;Van Hemel and Rogers, 1998;Heaton and Griffin, 2015;Hege et al., 2019;Holmes and Power, 1996;Korelitz et al., 1993;Mullins et al., 2013;Olson et al., 2009;Ronna et al., 2016;Sieber et al., 2014;Staško and Neale, 2007;Thiese et al., 2015;Whitfield Jacobson et al., 2007;Gay Anderson and Riley, 2008;Solomon et al., 2004). Truckers' diets were overall characterized as poor Versteeg et al., 2018) aside from truckers surveyed in two studies Solomon et al., 2004) -with high amounts of both sodium  and %kcal from fat (Olson et al., 2009;Thiese et al., 2015) in relation to dietary guidance recommendations (Health Canada, 2019; U.S. Department of Agriculture, 2020). Few drivers met recommendations for daily fruit and vegetable consumption (USDA & HHS, 2020; Holmes and Power, 1996;Whitfield Jacobson et al., 2007;Olson et al., 2009;. Truckers reported consuming around 1-2 meals per day in one study (Korelitz et al., 1993), and another found dinner was the most common meal consumed during weekdays on the road (breakfast was less common) (Holmes and Power, 1996). Snacking was frequently reported among drivers (Holmes and Power, 1996;Korelitz et al., 1993). Truckers reported frequent consumption of meats, salty snacks, fruit, and sweets (Holmes and Power, 1996). Truckers reported consuming sugary snacks, sugar-sweetened beverages, and fast-food at least 2 times per week on average, with sugary drinks reported most frequently (Olson et al., 2009). In another study truckers reported consuming these products about 4 times/week . Caffeinated beverage intake, including coffee, was frequent (Heaton and Griffin, 2015;Staško and Neale, 2007). For example, the majority of truckers interviewed by Staško et al (2007) (Staško and Neale, 2007) reported drinking up to or more than 5 cups of coffee each day. Caffeinated beverages and snacking were reported to help fight fatigue (Van Hemel and Rogers, 1998). Eleven studies assessed alcohol intake Hege et al., 2019;Korelitz et al., 1993;Wenger, 2008;Gay Anderson and Riley, 2008;Ronna et al., 2016;Sieber et al., 2014;Staško and Neale, 2007). The majority of truckers among six studies did not report high alcohol use in general Hege et al., 2019;Staško and Neale, 2007;. In two studies, over half reported drinking alcohol within the past year (Ronna et al., 2016;Gay Anderson and Riley, 2008). A minority reported consuming alcohol weekly (Staško and Neale, 2007; or daily (Gay Anderson and Riley, 2008). Other studies found most truckers reported being current drinkers (Korelitz et al., 1993;Sieber et al., 2014).
One study compared smoking rates between independent and company drivers and found no significant differences between groups (though both had high rates of smoking: 47% of company and 56% of independent truckers reported current smoking) (Bachmann et al., 2018). One study assessing sleep behaviors and fatigue among truckers reported that less than 1% of truckers report smoking or using other tobacco products to stay awake on the road (Van Hemel and Rogers, 1998), and another found that roughly a third of truckers in their sample (31%) reported quitting smoking (Solomon et al., 2004). A final study assessing trucker health knowledge using an online forum found that while online posts related to smoking comprised only 3% of posts, the subject of smoking and e-cigarette use resulted in the highest number of reply posts, indicating that truckers were engaged with these topics (Versteeg et al., 2018). Some reported smoking 'more than ever' (Wenger, 2008).

Main findings and importance
Available data about trucking food, PA, tobacco environments, and truckers' related patterns and practices were synthesized in this review to inform needed opportunities for high-impact interventions that target factors beyond individual health knowledge and skills (Ng et al., 2015;Frieden, 2010;Crizzle et al., 2017). Only four reviewed studies used environmental measures Apostolopoulos et al., 2011;Lincoln et al., 2018;McGuirt et al., 2019) to examine trucking settings and only one study used validated environmental tools (McGuirt et al., 2019). A higher number of studies explored perceptions about environmental factors that influence health patterns and practices (Apostolopoulos et al., 2013;Hege et al., 2019;Holmes and Power, 1996;Layne et al., 2009;Lemke et al., 2016;McDonough et al., 2014;Passey et al., 2014;Staško and Neale, 2007;Turner and Reed, 2011;Versteeg et al., 2018;Wenger, 2008;Whitfield Jacobson et al., 2007); however, these perceptions did not consider truckers' views of possible environmental change strategies to improve health. Last, there was no evidence about trucking tobacco environments. Thus, a main contribution of this review is a highlighted dearth of empirical information about attributes of trucking environments requiring intervention that negatively influence truckers' reported food, PA, and tobacco patterns and practices.
While not directly related to the specific focus of our review, several reviewed research studies described an unsupportive trucking workplace culture that was at times implicated for food, PA, and/or tobacco practices (Apostolopoulos et al., 2013;Apostolopoulos et al., 2016;Apostolopoulos et al., 2011;Lemke et al., 2016;McDonough et al., 2014;Wenger, 2008;Johnson et al., 2021;Williams et al., 2017). In general, the structure and nature (sedentary, stressful, remote) of the job as well as characteristics of trucking settings (limited health care) were reported barriers to positive health patterns and practices. This indicates future steps to build the research base and inform policy, systems, and environmental change efforts to improve truckers' workplace environments also need to account for truckers' views and sociocultural contexts and "begin with the end in mind" by understanding the acceptability, appropriateness, and feasibility of potential built environment interventions (Weiner et al., 2017;Klesges et al., 2005;Proctor et al., 2011). Specific recommendations to move the state of the science forward are described in more detail below.

Trucking food environments
Both objective and perceived assessments of the food environment are widely applied in the public health nutrition literature (Neve et al., 2021; National Collaborative on Childhood Obesity Research, 2019) and are promising approaches to improve the state of the science regarding trucking food environments. For example, details about the development and reliability and validity of the HEATWAI tool used to measure food and PA environments in two studies Apostolopoulos et al., 2011) and the National Institute for Occupational Safety and Health checklist used in one study (Lincoln et al., 2018) were not able to be identified. To build a coherent evidence base to inform future trucking food environment interventions, the use of standard tools assessed for reliability and validity is critical. As one example, the Nutrition Environments Measures Survey (NEMS) variants (e.g., for convenience stores, restaurants, vending machines) (Center for Health Behavior Research, 2021) are publicly available, accompanied by training, and have standardized scoring protocols for analysis. Additionally, standardized survey metrics for assessing dietary patterns and practices (e.g., Behavioral Risk Factor Surveillance System (diet and alcohol) (Centers for Disease Control and Prevention, 2019), National Cancer Institute Fruit and Vegetable screener (National Cancer Institute, 2021), or the Beverage Intake Questionnaire for Habitual Beverage Intake (Fausnacht et al., 2020) were not consistently applied and are also recommended for trucking research moving forward. Improving the use of standard measures will be especially useful for correlational research linking food environment properties to truckers' dietary purchases.

Physical activity environments
Measures of the PA environment were also limited. This is unsurprising, as most of the evidence on the association between environment and PA behavior comes from data on individuals' perceptions of the environment rather than objective assessments (Brownson et al., 2009). However, more information about characteristics of existing  Food: 48% of truckers were characterized as having a poor diet. 82% reported very high salt intake (above 2300 mg/day) and over 96% consumed above 1,500 mg/day. 42% reported high alcohol intake.
Tobacco: 31.5% reported cigarette smoking. Apostolopoulos, 2013 X Healthy Trucker Survey (HEATS), based on Long-Haul Trucker Interview Guide, the Health Appraisal Survey, and the Health Survey of the NSW Transport Industry, included questions on work history, workplace conditions, physical health, wellness, mental health, healthcare access and medical treatment history.
PA: 70% of drivers reported not participating in some form of regular exercise.
Bachmann, 2018 X X Self-reported health including PA and smoking status. PA: Differences between company and independent drivers were not statistically significant. 62.8% of company and 50% of independent drivers reported exercising in the past 30 days (p = 0.10). 64.5% of all truckers reported feeling out of shape.

PA:
Of the 47 drivers with depressive symptoms, PA was reported 3.0 ± 2.3 days/week. Of the 60 drivers without depressive symptoms, PA was reported 2.8 ± 2. days/week (p = 0.638).
Food: Drivers reported drinking an average of 16.4 oz of caffeine drinks when on duty and 15.1 oz when off duty. Hege, 2019 X X X Self-reported health behaviors regarding the number of alcohol drinks on non-workdays, caffeine intake, smoking status, and daily exercise habits.
Food: 51.4% reported not consuming alcohol on non-workdays and 48.6% reported having one or more alcoholic drinks on nonworkdays.
PA: 39.4% of truckers were sedentary and 60.6% were moderately active.
(continued on next page) Food: 55% of drivers ate dinner while on the road Mondays-Fridays (lowest frequencies of all meals on Saturdays/Sundays). 48% of drivers stopped for breakfast on Mondays-Fridays. 44% of drivers ate lunch on the road during weekdays. Snacks were also high during weekdays (range 45-60% of drivers). Drivers reported their favorite meals on the road were steak (29%), burgers (25%), chicken (17%), and a buffet (13%). Drivers most frequently reported snacks were chips, then fruit, followed closely by candy, donuts, and cookies. Less frequently reported snacks in comparison were popcorn, crackers, pop, and ice cream. Most drivers were found to meet daily protein and dairy requirements and only a minority met servings for fruits, vegetables, and cereals and breads. Korelitz, 1993 X X X Self-administered, close-ended questionnaire that measured participants' personal characteristics, health status, and health interests.
Food: More than 80% reported eating only one or two meals per day and 36% reported consuming three or more snacks each day. 59.2% were current drinkers. Food: Scaled responses for eating 6-11 servings of bread, cereal, rice, and pasta/day was 3.28 ± 0.72 (Spanish version) and 1.77 ± 0.69 (English version) (P = 0.02). The study concluded the HPLP II nutrition subscale may not adequately assess behavior among the study population.
PA: Scaled responses for exercising vigorously for 20 or more minutes at least three times per week (brisk walking, bicycling, aerobic dancing, using a stair climber) were on average 2.64 ± 1.00 (Spanish version) and 2.00 ± 0.91 (English version) (P = 0.04). Responses for taking part in light-to-moderate PA (sustained walking for 30-40 min 5 or more times/week) were on average 3.14 ± 0.89 (Spanish version) and 2.00 ± 1.00 (English version) (P = 0.012). Responses for doing stretching exercises at least 3 times a week were on average 2.57 ± 1.01 (Spanish version) and 1.81 ± 0.87 (English version) (P = 0.02 less than 0.05). The study concluded the HPLP II physical activity subscale may not adequately assess behavior among the study population. Olson, 2009 X X Dietary behaviors measured using validated measures of daily fruit and vegetable consumption and using "Gear Up for Health" to assess dietary fat and sugar consumption. Exercise behaviors measured using the 7-day PA Recall interview, which was used to convert to active kcals/kg/week of moderate to vigorous PA. Other measures were used to assess fitness: strength (maximum pushups and timed curl-ups; grip strength measured with a hand dynamometer); flexibility (sitand-reach test); and the 6-minute walk test.
Food: 75% of drivers reported consuming a mean of 36.6% kcals from fat over the past month. Mean sugary snack consumption was reportedly 2.55 times/week, mean sugary drink consumption was reportedly 4.22 times/week, and mean fastfood consumption was reportedly 2.32 times/week. 69% reported consuming an average of 3.05 fruit and vegetables servings per day and 76% consumed an average of 2.31 meals from home per week. Among most participants (75%), the frequency of sugary drinks was "5 or 6 times a week, the frequency of sugary snacks was "1 or 2 times a week", and frequency of fast-food was reportedly "1 or 2 times a week."

PA:
The mean reported active kcals/kg/week among all drivers was 23.32. Mean active kcal/ kg/week for drivers who reported a typical week before interviews (n = 7) was 13.14. Around a quarter of drivers (27%) reported engaging in 40 min of moderate exercise most days each week. On average, the 6minute walking test results were 525.05 m, the push-up results were 3.36, the valid curl ups were 6.41, and the flexibility results were 12.59 in..  X X X Self-reported daily fruit and vegetable intake, percent of calories from fat, and frequency of sugary food, sugary drink, and fast-food consumption.  (continued on next page) environments is needed to guide the selection, adaptation, and implementation of built environment interventions to promote PA. For example, assessing the presence or absence and quality of features that affect PA (e.g., streets, sidewalks, public spaces, signage, litter, lighting, incivilities, shade) (Brownson et al., 2009) can identify barriers and aid in the selection of built environment interventions to improve PA patterns and practices among trucking settings. Moving forward, the Workplace Walkability Audit could be adapted and pilot tested for various trucking settings to assess PA attributes (Dannenberg et al., 2005).
Comparing PA patterns and practices across studies to inform PA interventions was also difficult due to the different measures used in each study. Only two studies (Ronna et al., 2016;Turner and Reed, 2011) that assessed PA used measures aligned with meeting the national PA guidelines (i.e., 150 min per week of moderate-intensity aerobic activity or 75 min of vigorous intensity aerobic activity or an equivalent combination of each, and two sessions of full-body strength training per week) (HHS, 2018). For example, one study assessed whether truckers engaged in 30 min of "sustained" PA three times per week (Whitfield Jacobson et al., 2007). This is difficult to align with national PA recommendations, as aerobic activity does not need to be done in 30-minute periods. The most recent version of the PA guidelines specifies that any duration counts, while the 2008 version specified that aerobic activity needed to be done 10 minutes or more at a time (HHS, 2018). Additionally, only two studies fully assessed compliance with PA guidelines, including the strength training portion of the recommendations. Two studies assessed frequency of stretching (Mullins et al., 2013;Turner and Reed, 2011), which is not included in the guidelines and is not associated with chronic disease prevention (HHS, 2018).
Last, in 1998 a reporter for Overdrive noted truck stop gyms were "popping up around the country" to address truck drivers' health (Cox, 1998). Truckers in some of the reviewed research described truck stop gyms (when available) as "ratty" or time a barrier to their use . Evidence-based interventions, such as built environment PA: 34% reported exercising (extent ranged from stretching in the truck, to getting out and kicking tires, to running) and 5.5% reported maintaining fitness (e.g., watch weight, get enough exercise).
Tobacco: Less than 1% of drivers reported smoking or using other tobacco products (e.g., chewing tobacco, snuff, dip) to fight fatigue. Versteeg, 2018 X X Driver Health Forum posts from the Trucker Report Website were coded and analyzed.
Food: There were 289 (16%) food-related posts, which was the highest number of posts by category. Typical foods consumed were categorized as energy dense and ultra-processed foods.
PA: There were 85 (4.8%) exercise-related posts. Drivers wanted more opportunities to exercise in trucks and suggested building walking or running trails around truck stops. Offering gym memberships was also a recommendation.  (Kahn et al., 2002;McCormack and Shiell, 2011;Sallis et al., 2012), may fail to be effective initially and over time if endusers' perceptions are not considered (Weiner et al., 2017;Klesges et al., 2005;Proctor et al., 2011). Truckers' buy-in will be critical for future interventions.

Tobacco
No studies to date have examined tobacco environments (either objective or perceived) for truckers (e.g., tobacco availability and marketing in truck stops). The lack of objective and perceived environmental data on retail tobacco environments for truckers represents a considerable gap in the literature given the high prevalence of smoking reported among truckers Hege et al., 2019;Korelitz et al., 1993;Ronna et al., 2016;Sieber et al., 2014;Staško and Neale, 2007). Methods to assess the retail tobacco marketing environment have been developed (Lee et al., 2014;Henriksen et al., 2016). Given the high prevalence of smoking and the frequency with which truckers patronize truck stops, these settings may be an ideal location to identify opportunities to support policy and environmental changes conducive to smoking cessation. Globally, point-of-sale tobacco display bans have been found to reduce adult daily smoking by approximately 7% (Ribisl et al., 2017) and a menthol sales ban was found to reduce retail sales by 93% in Ontario (Page et al., 2020). Additionally, a 2019 study by Rust et al. found stores authorized by the Supplemental Nutrition Assistance Program (SNAP) -a federal nutrition assistance program to promote food security among Americans with lower incomes (Tiehen, 2020)had almost 3 times greater odds of interior tobacco marketing than comparison stores (Rust et al., 2019). Of note, many truck stop locations are SNAP-authorized (USDA, 2021), indicating these sites may have more tobacco advertisements than other locations and thus may be prime sites for feasible policy solutions. Similar environmental approaches to identify smoking cessation intervention opportunities may also increase gender equality, given that women truckers have substantially higher rates of smoking relative to their male counterparts (Birdsey et al., 2015;Layne et al., 2009).

Health disparities, health equity, and the COVID-19 pandemic
More information about trucking population differences in access to and experiences with trucking food, PA, and tobacco environments is needed by sociodemographic characteristics, given this was a limited focus among reviewed studies. The trucking workforce is diversifying (Cheesman Day and Hait, 2019) and health inequities experienced by truckers in general (Sieber et al., 2014) are likely more severe for women and racial and ethnic minority groups in this occupation (Singh et al., 2017). Understanding how trucking populations navigate their environments is critical, including experiences such as safety concerns (Gray and Lindsay, 2019) which may disproportionately affect certain populations and influence how truckers interact with truck stop food, PA, or tobacco amenities. Also, review results may not be reflective of the current trucking food, PA, or tobacco environment. Since the COVID-19 pandemic truckers, who are essential workers, have reported lower pay and closed travel stops and restaurants (or, in some cases, a shift to drive-through ordering) (Hennessy-Fiske, 2020; Cole, 2021;Dills, 2021;Maiden, 2020), which suggests opportunities for healthy eating and PA while on the road may be further reduced.

Limitations
This review has several important contributions: 1) the review scope is a novel characterization of trucking environments and truckers' patterns and practices related to food, PA, and tobacco that have not been investigated previously; 2) review results demonstrate the state of the science regarding trucking food, PA, and tobacco environments is underdeveloped and requires more focus using better measures; and 3) several future directions are highlighted that introduce public health concepts that could be applied to complement and build on the work of occupational health and safety scientists to improve the state of the science in this regard. However, review results should be interpreted with respect to limitations. Results are limited by the search databases used, which may have been inadequate to identify all articles meeting the inclusion criteria. Five research databases, Google searches, and reference searching were used to minimize this limitation. Some articles identified for review inclusion had a specific field focus (i.e., mental health), so relevant articles may have been overlooked during the title/ abstract or reference searching processes due to a lack of perceived focus on food, PA, or tobacco. However, researchers were liberal in flagging sources for full-text review. The exclusion of gray literature is also a limitation given industry reports may have included important information but were not included. Given a source quality assessment was a priority, gray literature was outside the scope of this work. Last, results in majority are reflective of U.S. trucking settings and more investigations are warranted in Canada.

Conclusion
The science of trucking food, PA, and tobacco environments is underdeveloped and requires much more focus using validated measures. Future steps to build the research base and inform policy, systems, and environmental change efforts to improve truckers' workplace environments should account for truckers' views and socio-cultural contexts and "begin with the end in mind" by understanding the acceptability, appropriateness, and feasibility of potential built environment interventions aimed at improving truckers' food, PA, and tobacco patterns and practices.

Ethics approval and consent to participate
Not applicable; Review did not engage human subjects.

Consent for publication
Not applicable.

Availability of data and materials
Review data is available in reported Tables. The search strategy and quality review indices are available in supplementary files.

Authors contributions
BH led all aspects of this review. BH, LB, LM, RM, and CBS were responsible for research conception and review design. BH and KK implemented the search strategy and extracted/reviewed data in collaboration with LB, LM, and CBS. BH and LB completed the quality assessment. BH, LB, and LM were responsible for writing the manuscript and KK assisted with Tables and formatted the manuscript to Journal requirements. All authors critically reviewed the draft manuscript and approved the final version.

Funding
This research was supported by the Louisiana State University Agricultural Center and the U.S. Department of Agriculture National Institute of Food and Agriculture, Hatch project 1024670.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.