UBC Faculty Research and Publications

Predictors of trips to food destinations Kerr, Jacqueline; Frank, Lawrence; Sallis, James F; Saelens, Brian; Glanz, Karen; Chapman, Jim May 20, 2012

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-12966_2011_Article_582.pdf [ 216.19kB ]
JSON: 52383-1.0221581.json
JSON-LD: 52383-1.0221581-ld.json
RDF/XML (Pretty): 52383-1.0221581-rdf.xml
RDF/JSON: 52383-1.0221581-rdf.json
Turtle: 52383-1.0221581-turtle.txt
N-Triples: 52383-1.0221581-rdf-ntriples.txt
Original Record: 52383-1.0221581-source.json
Full Text

Full Text

RESEARCH Open AccessPredictors of trips to food destinationsJacqueline Kerr1*, Lawrence Frank2, James F Sallis3, Brian Saelens4, Karen Glanz5 and Jim Chapman6AbstractBackground: Food environment studies have focused on ethnic and income disparities in food access. Few studieshave investigated distance travelled for food and did not aim to inform the geographic scales at which to study therelationship between food environments and obesity. Further, studies have not considered neighborhood design asa predictor of food purchasing behavior.Methods: Atlanta residents (N = 4800) who completed a travel diary and reported purchasing or consuming food atone of five food locations were included in the analyses. A total of 11,995 food-related trips were reported.Using mixed modeling to adjust for clustering of trips by participants and households, person-level variables(e.g. demographics), neighborhood-level urban form measures, created in GIS, and trip characteristics (e.g. time ofday, origin and destination) were investigated as correlates of distance travelled for food and frequency of grocerystore and fast food outlet trips.Results: Mean travel distance for food ranged from 4.5 miles for coffee shops to 6.3 miles for superstores. Type ofstore, urban form, type of tour, day of the week and ethnicity were all significantly related to distance travelled forfood. Origin and destination environment, type of tour, day of week, age, gender, income, ethnicity, vehicle accessand obesity status were all significantly related to visiting a grocery store. Home neighborhood environment,day of week, type of tour, gender, income, education level, age, and obesity status were all significantly related tolikelihood of visiting a fastfood outlet.Conclusions: The present study demonstrated that people travel sizeable distances for food and this distance isrelated to urban. Results suggest that researchers need to employ different methods to characterize foodenvironments than have been used to assess urban form in studies of physical activity. Food is most oftenpurchased while traveling from locations other than home, so future studies should assess the food environmentaround work, school or other frequently visited destinations, as well as along frequently traveled routes.Keywords: Built environment, Food environment, Urban form, Travel, Nutrition, ObesityIntroductionSome studies document built environment-obesity asso-ciations [1-5]. Both physical activity environments andfood environments could contribute to the relationshipbetween obesity and urban form. It is well documentedthat neighborhood form (e.g., land use patterns) isrelated to physical activity [6-9], but evidence regardingthe relation of food environments to food purchasingpatterns and eating behaviors is limited [10,11]. In theUS, the number and distance to healthful food stores andrestaurants varies by neighborhood income and ethniccomposition [12-15]. Indeed, most food environmentstudies have focused on income and ethnic disparitiesin obesity rates and diet quality and their relationshipto availability of fast food restaurants and grocery stores[16-18]. However, most studies of food environments inthe U.S. have not considered urban form or other factorsthat impact access to food stores.Definitions of food environment access and availabilityhave included the number or density of food outlets in agiven area and/or home-to-food outlet distances [4]. Theoften-applied gravity model asserts that closer destina-tions are exponentially more attractive, saving time andmoney on travel [19]. Trip tour data, however, also indi-cate that people piece together trips and stops to be con-venient. Food outlet proximity is impacted by land usemix and street network patterns, with more griddedstreets and a mixture of retail and residential land usessupporting shorter trips and more travel by walking and* Correspondence: jkerr@ucsd.edu1University of California, San Diego, USAFull list of author information is available at the end of the article© 2012 Kerr et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the CreativeCommons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly cited.Kerr et al. International Journal of Behavioral Nutrition and Physical Activity 2012, 9:58http://www.ijbnpa.org/content/9/1/58cycling [6-8,20]. It is plausible that increased access tofood (and other destinations) near where someone livesmay result in less food purchasing to and from work andother destinations. One study showed that increased landuse mix where people live resulted in simpler tours (lessstops) to and from work [21]. Research has not yetexamined if access to and use of healthy food storesis greater in these more ‘walkable’ neighborhoods [10],although one study found distance from home to foodstores decreased with increasing population density, amarker of greater walkability [22]. The actual distancepeople go for food purchasing and the trip characteristicsare understudied.To date, environmental correlates of physical activityand obesity have been examined at scales up to a onemile buffer around residents’ homes [23,24], or evenshorter distances for children [25,26]. These distancesare reasonable estimates of how far individuals will walk,with the focus almost exclusively on environmentsaround the home, where it is assumed most physicalactivity occurs. Walkability of neighborhoods measuredat these scales is consistently related to walking fortransportation in adults [6-9]. In contrast, food environ-ment studies mostly assess communities [27] or neigh-borhoods, defined by census blocks or tracts [28].However, it is not clear what are meaningful scales ordistances for defining food environments and/or whetherfood outlet type is important to scale/distance considera-tions. Obesity rates may also be impacted by the type offood outlet (store type). For example, grocery stores tendto sell higher quality and cheaper fresh fruits and vegeta-bles [29,30], more low fat products and fresh productsthan fast food restaurants or convenience stores whichtend to sell processed foods commonly high in fat andsodium [29,31].The current study seeks to inform the understandingof the scale at which food purchasing from stores andrestaurants should be evaluated by documenting, thenexamining correlates of how far people actually travel forfood. The geographic scale at which food is obtained islikely a function of many factors including daily traveland commute patterns, presence of food options thatmatch individual preferences and other individual factors(e.g., age), land use around the residence, income, andprice. There is limited evidence on how far individualsactually travel for food [32] but getting food is the sec-ond most common travel purpose. Travel patterns forfood are not well understood and it is not clear whatproportion of food purchases are performed with homeas the starting point. Associations of the food environ-ment with diet and obesity could be obscured and resultsmay be misleading if we continue to assume that foodpurchasing occurs only near one’s residence, and/or touse the same buffer sizes to measure the foodenvironment as employed for physical activity environ-ment studies. The association between residential neigh-borhood and obesity may be particularly misleading forlow income ethnic groups most at risk for obesity,because these individuals spend large amounts of timeaway from their home neighborhood attending to familyand work responsibilities [33]. Previous studies of foodenvironments have not provided data on actual food pur-chasing behavior, where and when people buy food, andhow far they travel to buy food in a large sample withtrips extending beyond the local neighborhood.MethodsData collected for the cross-sectional SMARTRAQ (Strat-egies for Metropolitan Atlanta’s Regional Transportationand Air Quality; see www.act-trans.ubc.ca/smartraq)household travel survey in the 13-county Atlanta regionin 2001–2002 were analyzed. Data collection was stratifiedacross 4 ranges of income and household size and 5 levelsof residential density, meaning some population groupswere oversampled to ensure variation in socio-demo-graphics and urban form. Study method details are pub-lished elsewhere [5,23]. The overall response ratewas typical for travel surveys at 30.4%; partly reflectingsubstantial study demands on participants. Verbal con-sent was acquired from participants and the study wasapproved by the local ethical review board. While the pri-mary aim of the study was to study travel behaviors toinform transportation and air quality research and plan-ning, the data included trips for the recorded purposeof eating or purchasing food which allowed the currentanalyses to be performed.MeasuresThe aim of this study was to explore environmental, in-dividual and trip level factors related to shopping forfood. The analyses were framed by the ecological modelof behavior change that includes multiple levels of influ-ence and would include factors from multiple types ofenvironments e.g. home, work, school etc. Unfortunatelyin practice, most ecological studies to date focus only onthe home environment. This study collected informa-tion from each location that was visited to allow a morecomplete analysis of travel for food predictors.Participants completed a paper travel diary, recordingdestinations visited, travel mode and purpose, and timeof day across two days assigned by the research team toensure an even distribution of all weekday and weekenddays across the sample. Socio-demographic informationwas provided by a head of household in a recruitmentcall through the use of a computer aided telephone inter-view (CATI) protocol. Height and weight were reportedindividually by household members. BMI was computedas kg/m2.Kerr et al. International Journal of Behavioral Nutrition and Physical Activity 2012, 9:58 Page 2 of 10http://www.ijbnpa.org/content/9/1/58The built environment variables were created usingNetwork Analyst, which is an extension to the GIS soft-ware product developed by the ESRI corporation knownas ARCVIEW (ArcView GIS 3.2; ESRI Inc., RedlandsCA, 2000). GIS was used to assess the distance partici-pants traveled to food sources provided on the traveldiary and to define the urban form around each address.This study aimed to assess whether urban form vari-ables related to physical activity and obesity were alsorelated to food purchasing. A one kilometer road net-work buffer was developed around each trip origin,destination and home address to create urban formmeasures. A combination of county level Tax Assessorsparcel data and census data were used to measure resi-dential density and mixing of land uses, and street net-work files were used to measure street connectivitywithin the 1 kilometer buffer. These values were normal-ized for the sample and summed to create a measure ofdestination accessibility that has been related to bothwalking and vehicle miles travelled [21]. An index wasemployed to simplify the analyses because multiple loca-tions were being investigated (origins, destinations andhome) and comparisons across them could be made.These are the same methods published in two paperslinking built environment measures with physical activityand obesity in adults from SMARTRAQ [5,23].Car travel time and distance walked were calculatedusing GIS, with each trip, origin and destination fromthe travel diary placed on the street network. The short-est time (for car travel) and distance (for walking travel)between the origin and destination along the road net-work was computed. For car travel, expected travel timeswere developed based on time of day and direction oftravel to adjust for congestion level, using data from theAtlanta Regional Commission’s Regional Travel Model.Corresponding morning, afternoon, and off peak zone-to-zone link-based travel times for reported trips weredrawn from the regional travel model. The distances alongthe shortest route were then measured and employed inthe current analyses.Identifying food location typeFrom a list, participants indicated the primary and up tofour secondary activities they did at each destination andthe destination name and address. Destinations werecoded as food-related if one of the two food-relatedactivities (purchasing food or eating) was recordedamong the primary or secondary activities. Destinationsonly in which individuals indicated engaging in a food-related activity were categorized into food establishmenttypes based on location name.Food related activity destinations with food-relatedactivities included convenience stores, bars, schools, churches,hospitals, entertainment centers and malls. Most of thesewere excluded from analyses as there were few trips tothese destinations. Convenience stores were often visitedbut it was not clear that food shopping, as opposed toshopping for gas, had occurred. Many convenience storesare attached to gas stations, and there were few instancesof eating reported in these locations. Trips to five foodoutlets were evaluated including to fast food restaurants,sit down restaurants, grocery stores, coffee shops andlarge superstores.Destinations were assigned to the “fast food” categoryif they contained any of the following words: “burger,burrito, cafeteria, chicken, deli, food court, hot dog,pizza, sub, taco, wings”. Regional and national chainnames, e.g. Burger King, Kentucky Fried Chicken,Krystal, McDonald’s and Mrs. Winner’s, were alsoincluded in the “fast food” category. Locations withthe word “restaurant” in the name were included in thesit down restaurant group. After categorizing recordsinto the above categories the remaining set of locationswere reviewed for possible inclusion in the “restaurant”category. Locations were labeled sit-down restaurantsbased on a record-by-record review using known local/regional/nation restaurant chain names (e.g. Flying Biscuit,Fuddruckers, and Hard Rock Café), investigators’ know-ledge of the region, and internet searches for locationdescriptions. Locations with the word “grocery” in thename were reviewed and included in the grocery storescategory. Regional and national supermarket chains (e.g.Kroger, Publix), were also identified and included as‘grocery stores’. The coffee shop category was developedfrom locations categorized by the research team as“bakery, doughnut shop, coffee shop, etc.” or if the loca-tion name included any of these words; “bakery”,“doughnut,” “bagel,” “bread” or “coffee.” Locations werecategorized as “large superstores” if they were named oneof the following—Belk Department Store, Costco Ware-house, Home Depot, Goody’s Family Clothing, HomeDepot Expo, JC Penney, K-Mart, Kohl’s, Lowes, Rich’s,Sam’s Club, Sears, Target or Walmart. Although some ofthese stores are not principally food outlets, they wereincluded if food eating/purchasing was reportedby participants in these stores.Participants who completed the travel diary and indi-cated at least one food related activity in either day atone of the 5 food establishment categories describedabove were included in analyses.Variables included in the analysesFor these analyses only food destinations were consid-ered, and the immediate location before the food destin-ation was considered the trip origin. The distance fromthe origin location to the food outlet was examined. Thereturn trip was not included. For each trip origin anddestination location, destination accessiblity scores wereKerr et al. International Journal of Behavioral Nutrition and Physical Activity 2012, 9:58 Page 3 of 10http://www.ijbnpa.org/content/9/1/58calculated from the index combining land use mix, inter-section density and residential density. The urban formscores were split into tertiles based on these analyses.For these exploratory analyses simple trips wereconsidered in a set of predictive models. Travel be-havior research generally considers more complextours and trip chains [34,35]; often involving stopsbetween home, work or other major non-work loca-tions. To investigate the relationship between homeresidential environment and food purchasing a sim-ple home-food-home tour category was created. Thiscategory included only trips from home to a fooddestination followed by a return trip to home, with-out other stops. Similarly, a simple work-food-worktour category was created. This category includesonly trips from work to a food destination followedby a return trip to work, without other stops. Twoother simple tours were created: a home-food-worktour and a work-food-home tour. The analyses inves-tigated whether the travel behavior varieddepending on whether these types of tours occurred.Trip, personal and household variables were includedas correlates of distance traveled for food purchasing.Distance and frequency of visits to the five food destina-tions were compared by day of the week, origin of thetrip (home, not home) and urban form of the trip originand destination. Household level income (<$50,000,$50-74,000, $75,000+) and number of vehicles ownedwere compared for distance and frequency of travel.At the person level, gender, race (white/non white), edu-cation (college degree or not), work status and obesitystatus (BMI greater than or equal to 30 or less than 30)were related to the dependent variables distance andfrequency of travel.AnalysesThree dependent variables were analyzed; 1) Distancetravelled for food (in miles, continuous), 2) Visit toa fastfood restaurant (vs visit to any other food loca-tion) and 3) Visit to a grocery store (vs visit toany other food location). Mixed model analyses wereemployed adjusting for clustering of trips by participantand household.Environment factorsTertiles of destination accessiblity in the one kilometernetwork buffer around the origin and destination of thefood trip, and participants’ home was used.Trip factorsThese included day of the week (working day or not),whether the food trip started and ended at home,whether the food trip started and ended at work, the fivecategories of food locations (in the distance model only).Participant demographic factorsVehicles in the home, annual household income categor-ies, educational status, employment status, obesity status,race, and gender.ResultsA total of 116,541 trips were made by 7665 participants.Of these, 4800 participants made 11,995 trips thatincluded a food activity (e.g., purchasing or eating food)during a visit to one of the five types of stores identifiedfor these analyses. Across the two day diary period 31.1%of food trips were made to a grocery store, 29.9% to a sitdown restaurant, 19.2% to a fastfood outlet, 13.1% to asuperstore and 6.7% to a coffee shop which included afood purchase. Only 7% of all trips to a food outlet weremade on foot.Distance traveled to any food storeThe unadjusted mean distance travelled to each of thefive food locations (and standard deviations) and for eachindependent variable can be found in Table 1. Table 1also presents the results of the mixed methods modelingadjusting for person sampled and number of participantsin the household. The data represent the trips made overthe two day travel diary period.Participants travelled furthest for superstore foodshopping and the least distance to grocery stores andcoffee shops. Those living in less accessible environmentsor making trips to and from less accessible environmentstraveled farther.Participants travelled further to food stores when thetrip was part of a larger tour with differing origins anddestinations before and after the trip to the food loca-tion; i.e., work food home tour or home food work tour.When the tour was work food work, distances travelledwere shorter. Participants travelled farther on non-workdays for food.In the adjusted analyses, lowest income participants,non whites, and those without a degree travelled furtherfor food.Grocery storesTable 2 presents the percentage of trips made to the gro-cery store by environment, trip and person levelvariables and the results of the adjusted analyses. Thosestarting a trip from the least accessible neighborhoodwere less likely to visit a grocery store than those start-ing from the most accessible. Those travelling to ahighly accessible destination were less likely to visit agrocery store than those traveling to a medium access-ible community. The destination accessibility of thehome environment was not significant.People were more likely to travel to a grocery storewith a tour starting and ending at home and less likelyKerr et al. International Journal of Behavioral Nutrition and Physical Activity 2012, 9:58 Page 4 of 10http://www.ijbnpa.org/content/9/1/58to travel to a grocery store during a break from work (i.e.a work food work tour). Participants were less likely tovisit a grocery store on the way to work and more likelyto visit a grocery store on the way home from work.Participants, however, were most likely to visit a grocerystore on a non work day.Those in high income households were least likely tovisit a grocery store compared to low and mid incomegroups over the two day survey period. Those without acar were more likely to visit a grocery store than othertypes of food store. Men were less likely to visit a grocerystore than women. Relative to their counterparts, nonwhites, non-obese participants and older adults weremore likely to visit a grocery store.Fast food outletsTable 3 presents the results of the unadjusted andadjusted analyses for the dependent variable, visit to afastfood outlet.The accessibility level of the trip origin and destinationwas not related to visiting a fast food outlet. Trips tofastfood were more likely when the home neighborhoodenvironment was least accessible vs most accessible.Trips to fastfood were more likely to made on workdays than non work days. Tours to and from home wereless likely to result in a fastfood stop. Tours that startedand ended at work were more likely to include a visit toa fastfood restaurant. Tours from home on the way towork were more likely to include a fastfood stop. Toursfrom work on the way home were less likely to result ina fastfood stop.Men were more likely to visit a fastfood outlet thatwomen. Those earning less than $50K were more likelyto eat fastfood than those earning $75K or more. Thosewithout a college degree were more likely to visit a fas-tfood outlet. Older participants were less likely to visit afastfood outlet. Obese adults were more likely to visit afastfood restaurant.DiscussionIn this study of Atlantans, the lowest average distance trav-eled to a food establishment was greater than 4.5 miles. ThisTable 1 Distance to food storesUnadjustedmean distance(SD) in miles*AdjustedT statisticPvalueOutlet type visited on tripGrocery store 4.67 (4.35) −9.86 .001Sit down 6.10 (4.87) -.31 .75Fastfood 4.96 (4.41) −6.44 .001Coffee shop 4.50 (4.40) −7.02 .001Superstore 6.32 (4.77) RefDestination accessibility of trip origin locationLow 6.65 (4.51) 9.22 .001Medium 4.99 (4.59) 1.39 .16High 4.46 (4.57) RefDestination accessibility of trip destination locationLow 6.92 (4.85) 9.23 .001Medium 5.41 (4.59) 3.50 .001High 4.56 (4.44) RefDestination accessibility of home locationLow 6.40 (4.73) 5.54 .001Medium 5.32 (4.47) 3.19 .001High 4.47 (4.40) RefHome food home tourTrip not part of tour 5.34 (4.70) 1.32 .19Home food home tour 5.37 (4.44) RefWork food work tourTrip not part of tour 5.42 (4.66) 3.67 .001Work food work tour 4.38 (4.03) RefHome food work tourTrip not part of tour 5.32 (4.59) −5.67 .001Home food work tour 6.07 (5.38) RefWork food home tourTrip not part of tour 5.23 (4.54) −9.91 .001Work food home tour 6.51 (5.30) RefDay of the week of tripNon work day 5.36 (4.51) 2.33 .02Workday 5.34 (4.73) RefHousehold income level of trip participantIncome< $50,000 5.31 (4.73) 2.11 .04Income $50,000–74,000 5.39 (4.50) .54 .59Income $75,000 + 5.34 (4.65) RefVehicle ownership in household of trip participantNo vehicles 4.76 (5.56) −1.11 .27At least 1 vehicle 5.36 (4.62) RefGender of trip participantMale 5.36 (4.63) .82 .42Female 5.34 (4.63) RefEducation of trip participantNo degree 5.70 2.33 .02Degree 5.12 RefTable 1 Distance to food stores (Continued)Ethnicity of trip participantNon white 5.57 (4.78) 3.50 .001White 5.30 (4.60) RefEmployment status of trip participantDoes not work 5.31 (4.48) −1.61 .11Works 5.34 (4.73) RefAge (continuous) - −1.49 .14* adjusted for all independent variables as well as person taking the trip andnumber of people in household taking trips.Kerr et al. International Journal of Behavioral Nutrition and Physical Activity 2012, 9:58 Page 5 of 10http://www.ijbnpa.org/content/9/1/58is a much longer distance than the 1-km to 1-mile buffersaround homes that are often constructed in studies of phys-ical activity environments [23]. Therefore we have to becautious about interpreting the relationship between obesityif it is assumed that an individual’s food environment is con-stituted only or mostly within these buffers.Although many food trips are currently outside of theresidential neighborhood, a 1-km buffer is still a good in-dicator of food access that may be related to shorterjourneys, more walking for transportation and greatersupport of local food sources. Participants travelled fur-thest to larger superstores for food, as might be expectedbased on variety and cost considerations. The implica-tions for greenhouse gas emissions and air pollution forlonger trips to superstores should be evaluated relativeto potential efficiencies if fewer trips to obtain food over-all are made. Travel for food or food miles is becoming acentral focus of strategies to reduce greenhouse gasemissions [36].Food trips were longer when people started from anon-home location or lived in suburban type locationwithout connected streets, mixed land uses or high resi-dential density. Non-white, those without a degree, andlower-income subgroups travelled longer distances forfood, suggesting local food sources may be unavailableand travel to food may be an additional hardship for theunderserved. Increased time spent in cars may also berelated to obesity [23]. In one study, supermarkets wereon average 1.15 miles further away for residents of blackcompared with white neighborhoods [37]. In anotherstudy, researchers estimated that residents of low-incomeneighborhoods would have to travel more than 2 milesto have access to the same number of supermarkets asresidents of higher-income areas [32]. Our findings foractual travel distances for food indicated that local accessto foods around the home is less frequent than access tofoods along routes that occur as part of people’s everydaylives. This may be the result of home environments de-void of local food outlets. Examining accessibility alongcommon routes requires a completely different measure-ment strategy than focusing only on residential neigh-borhoods and could include use of GPS devices whichtrack individuals across multiple locations and routes[38]. This is consistent with current tour basedapproaches to modeling and predicting travelTable 2 Grocery store visitUnadjusted% *AdjustedT statisticPvalueDestination accessibility of trip origin locationLow 28.1 −3.02 .003Medium 34.3 .79 .43High 31.9 RefDestination accessibility of trip destination locationLow 29.4 1.65 .10Medium 34.4 4.28 .001High 29.5 RefDestination accessibility of home locationLow 28.0 −1.60 .11Medium 32.4 .39 .70High 32.7 RefHome food home tourTrip not part of tour 28.9 −7.47 .001Home food home tour 37.7 RefWork food work tourTrip not part of tour 32.9 9.73 .001Work food work tour 5.9 RefHome food work tourTrip not part of tour 31.8 4.78 .001Home food work tour 15.9 RefWork food home tourTrip not part of tour 29.5 −9.57 .001Work food home tour 44.9 RefDay of the week of tripNon work day 36.1 2.01 .05Workday 27.0 RefHousehold income levelof trip participantIncome< $50,000 35.6 3.53 .001Income $50,000-74,000 30.9 2.06 .04Income $75,000 + 27.6 RefVehicle ownership in household of trip participantNo vehicles 43.3 2.02 .04At least 1 vehicle 30.9 RefGender of trip participantMale 26.6 −5.70 .001Female 34.8 RefEducation of trip participantNo degree 31.9 -.46 .65Degree 30.5 RefEthnicity of trip participantNon white 36.8 5.79 .001White 29.7 RefEmployment status of trip participantDoes not work 37.6 .35 .73Works 28.7 RefTable 2 Grocery store visit (Continued)Obesity status (from BMI) of trip participantNot obese 32.1 −2.63 .009Obese (BMI 30+) 29.7 RefAge (continuous) 6.74 .001* adjusted for all independent variables as well as person taking the trip andnumber of people in household taking trips.Kerr et al. International Journal of Behavioral Nutrition and Physical Activity 2012, 9:58 Page 6 of 10http://www.ijbnpa.org/content/9/1/58behavior where food outlet visits are often a mid tourstop [20].Origin and destination environment, type of tour, dayof week, age, gender, income, ethnicity, vehicle accessand obesity status were all related to visiting a grocerystore. Home environment, day of week, type of tour, gen-der, income, education level, age, and obesity status wereall related to likelihood of visiting a fastfood outlet.Many of the same results were seen for the grocery storebut operating in the opposite direction. For examplemen were more likely to visit a fastfood outlet and lesslikely to visit a grocery store, and older adults were morelikely to visit a grocery store than a fastfood location.Fastfood outlet visitation (and dining out overall) mayrepresent a less healthy lifestyle with mostly high fat pro-cessed foods available, and grocery store visits may rep-resent a healthier lifestyle because fresh fruits andvegetables are available.More trips were taken to fastfood on working daysthan non working days (and vice versa for grocerystores). In particular, a break during the work day waslikely to include a fastfood trip, as was the journey towork at the start of the day. This suggests that when par-ticipants are pressed for time they resort to stopping forfastfood. Interventions could target these particular triphabits and suggest preparing a healthy breakfast andlunch at home and taking it to work to avoid such un-healthy stops or alternately the provision of healthfulfood options by employers.While it is now possible to order healthier options atfastfood outlets, the relationship between obesity andfastfood outlet visitation suggests that patrons are stillpurchasing high fat foods. Continuing to work on pro-grams to provide healthier options where stops are madeis important. One study found that visiting a fast foodrestaurant was a significant predictor of higher BMI andvisiting a grocery store a significant predictor of lowerBMI in women [39].Those starting a trip in a neighborhood with few desti-nations were less likely to travel to a grocery store, butthose traveling to a medium accessible environment weremore likely to visit a grocery store. This may reflect theavailability of parking at many grocery stores, which al-though surrounded by other destinations, would contrib-ute to the grocery store being in a moderately accessibleTable 3 Fastfood outlet visitUnadjusted% *AdjustedT statisticPvalueDestination accessibility of trip origin locationLow 18.3 -.60 .55Medium 18.4 −1.78 .08High 19.8 RefDestination accessibility of trip destination locationLow 18.5 −1.11 .27Medium 19.1 -.23 .82High 18.8 RefDestination accessibility of home locationLow 20.1 2.39 .02Medium 19.0 1.54 .12High 18.5 RefHome food home tourTrip not part of tour 21.2 7.13 .001Home food home tour 13.0 RefWork food work tourTrip not part of tour 18.0 −7.51 .001Work food work tour 34.9 RefHome food work tourTrip not part of tour 18.9 −1.93 .05Home food work tour 25.0 RefWork food home tourTrip not part of tour 19.8 5.68 .001Work food home tour 13.9 RefDay of the week of tripNon work day 16.1 −2.57 .01Workday 21.7 RefHousehold income level of trip participantIncome< $50,000 20.7 3.51 .001Income $50,000-74,000 18.8 1.53 .13Income $75,000 + 18.2 RefVehicle ownership in household of trip participantNo vehicles 19.8 .80 .42At least 1 vehicle 19.1 RefGender of trip participantMale 20.5 2.75 .006Female 18.0 RefEducation of trip participantNo degree 21.1 2.37 .02Degree 17.7 RefEthnicity of trip participantNon white 19.6 −1.28 .20White 18.9 RefEmployment status of trip participantDoes not work 16.2 1.79 .07Works 19.4 RefTable 3 Fastfood outlet visit (Continued)Obesity status (from BMI) of trip participantNot obese 17.9 −2.74 .006Obese (BMI 30+) 21.7 RefAge (continuous) - −5.50 .001* adjusted for all independent variables as well as person taking the trip andnumber of people in household taking trips.Kerr et al. International Journal of Behavioral Nutrition and Physical Activity 2012, 9:58 Page 7 of 10http://www.ijbnpa.org/content/9/1/58community. Some unexpected results were found forgrocery store visits, including low income participants,non whites and those without a vehicle visiting a grocerystore more often than other types of stores. Perhapsthose without a car cannot carry as much food withthem in a single journey and are more likely to makesmaller, frequent trips to the grocery store.The present study clearly shows that people get foodfrom a variety of locations, many of which are outside oftheir local community. Therefore, examining the residen-tial food environment alone is insufficient. Many trips topurchase food begin at locations other than home. Thissuggests studying locations that individuals frequent maybe more useful than estimating access to food onlyaround the home; many people spend as much or evenmore of their waking hours at work. One study foundthat fast foods restaurants tended to cluster aroundschools [40]. To date, few studies have investigatedwhether the availability of healthy or unhealthy foodsoutside of the residential environment, along the routesor ‘activity space’ of an individual, is related to diet orobesity [38]. One small study of migrants did collect atravel and food diary over a week period to assess accessto foods locally. [41]. A recent review also emphasizesthe need to study locations other than home neighbor-hoods and food environments associated with commut-ing behaviors [42]. Fruitful locations may be aroundworkplaces or along frequently traveled routes.Although the literature has documented some dispar-ities in obesity related to differences in healthy food ac-cess (6 out of 10 studies) [43], these studies have notdirectly assessed where food was purchased. Our studyindicated that income, ethnicity, and education wasrelated to distance travelled. Studies suggests thatfood deserts may be racially driven, not just incomerelated [44]. Another explanation comes from studiesthat have also shown that some population groupsspend large amounts of time traveling outside of theirhome environments due to work and care commitments[23,33]. Assessing food purchasing patterns outside ofhome neighborhoods may be particularly important forthis group.The strengths of this study were the large and diversesample as well as specific measures of places where indivi-duals purchased food. The limitations included use of anactivity-based travel survey that relied upon self-report oftrips and activities, and limiting the environment to anurban/suburban sample. One study indicated that dis-tances to food stores may be even greater in rural areas[22]. A two-day travel diary may not be representative orinclusive of the habitual food outlet visitation for an indi-vidual or a household. The response rate was lower thanfor a typical survey study, but the respondent burden forthe present study was higher, and there was no informationcollected about non-respondents. The study did not exam-ine the possible food stores that were available to travel to(and/or their quality), making it difficult to determinewhether respondents were making trade-offs regardingdistance to/from food outlets and quality; i.e., traveling fur-ther to get better quality or cheaper food, or simply choos-ing to purchase food while on a trip. More complex foodtours should be analyzed in future studies, and new techni-ques from transportation research may be enlightening,such as employing mental mapping of daily travel behavior[45]. Food trips were not compared to trips for otherpurposes.Enumerating food locations accurately [46] and con-ducting quality audits [29] are both labor-intensive pro-cesses that would be prohibitive for the 13-countyAtlanta region. This study would be strengthened by spe-cific food intake data or food purchasing receipts to con-firm what food locations were most impactingparticipants’ diet and weight. A recent study in theWaterloo Region of Ontario Canada known as “NEW-PATH” was patterned after the Atlanta SMARTRAQstudy but also included dietary data collection if a loca-tion was visited that involved a food purchase [47]. Ex-clusion of convenience stores due to the unclear activitycriteria was a study weakness as these are common loca-tions for food purchasing. Additional food locationsshould be considered beyond the five primary ones inthe current analyses and inclusion of food locationswhere food can be grown not just purchased may beinformative.The present study demonstrated that people travelledsizeable distances for food and this distance is relatedwith urban form. Results strongly suggest that research-ers need to employ different methods to characterizefood environments than have been used to operationalizebuilt environment in studies of physical activity. Further,food is most often purchased while traveling from loca-tions other than home, so future studies should assessthe food environment around work, school or other fre-quently visited destinations, as well as along frequentlytraveled routes. Increasing our understanding of travelpatterns to purchase food is important for improving ourhealth and the health of our environment.Competing interestThe authors have no competing interests to declare.Author details1University of California, San Diego, USA. 2University of British Columbia,Vancouver, Canada. 3San Diego State University and University of California,San Diego, USA. 4Seattle Children’s Research Institute and University ofWashington, Washington, USA. 5University of Pennsylvania, Pennsylvania,USA. 6Urban Design for Health, Washington, USA.Authors’ contributionJK analyzed the data, drafted and edited the manuscript. LF designed thestudy, collected the data, drafted and edited the manuscript. JS drafted andKerr et al. International Journal of Behavioral Nutrition and Physical Activity 2012, 9:58 Page 8 of 10http://www.ijbnpa.org/content/9/1/58edited the manuscript. BS drafted and edited the manuscript. KG drafted andedited the manuscript. JC collected the data, created the variables andcontributed sections to the manuscript. All authors read and approved thefinal manuscript.Received: 22 June 2011 Accepted: 20 May 2012Published: 20 May 2012References1. James WP: The fundamental drivers of the obesity epidemic. Obes Rev2008, 9(S1):S6–S13.2. Giskes K, Kamphuis CB, van Lenthe FJ, Kremers S, Droomers M, Brug JA:A systematic review of associations between environmental factorsenergy and fat intakes among adults: is there evidence for environmentsthat encourage obesogenic dietary intakes? Public Health Nutr 2007,10(10):1005–1017.3. Booth KM, Pinkston MM, Poston WSG: Obesity and the built environment.J Am Diet Assoc 2005, 105(S1):S110–S117.4. Ford PB, Dzewaltowski DA: Disparities in obesity prevalence due tovariation in the retail food environment: three testable hypotheses.Nutr Rev 2008, 66(4):216–228.5. Frank LD, Schmid TL, Sallis JF, Chapman JE, Saelens BE: Linking objectivephysical activity data with objective measures of urban form. Am J PrevMed 2005, 28(2):117–125.6. Gebel K, Bauman AE, Petticrew M: The physical environment and physicalactivity: a critical appraisal of review articles. Am J Prev Med 2007, 32:361–369.7. Heath GR, Brownson J, Krueger R, Miles KE, Powell LT, Ramset LT, The TaskForce on Community Preventive Services: The effectiveness of urbandesign and land use and transport policies and practices to increasephysical activity: a systematic review. J Phys Act Health 2006, 3(S1):S55–S76.8. Sallis JF, Kerr J: Physical activity and the built environment. President’sCouncil on Physical Fitness and Sports 2006, 7:1–8.9. Saelens BE, Handy SL: Built environment correlates of walking: a review.Medicine & Science in Sports & Exercise 2008, 40(7):S550–S566.10. Glanz K, Sallis JF, Saelens BE, Frank LD: Healthy nutrition environments:concepts and measures. Am J Health Promot 2005, 19:330–333.11. Sallis JF, Glanz K: The physical activity environment and foodenvironments: toward solutions to the obesity epidemic. Milbank Q 2009,87:123–154.12. Story M, Kaphingst KM, Robinson -O’Brien R, Glanz K: Creating healthyfood and eating environments: policy and environmental approaches.Annual Reviews of Public Health 2008, 29:253–272.13. Frank LD, Glanz K, Sallis J, Saelens B, Chapman J, McCarren M: The spatialdistribution of food outlet type and quality around schools in differingbuilt environment and demographic contexts. Berkeley Planning Journal2007, 19:31–47.14. Galvez MP, Morland K, Raines C, Kobil J, Siskind J, Godbold J, Brenner B:Race and food store availability in an inner-city neighbourhood.Public Health Nutr 2008, 11(6):624–631.15. Powell LM, Slater S, Mirtcheva D, Bao Y, Chaloupka FJ: Food storeavailability and neighborhood characteristics in the United States.Prev Med 2007, 44(3):189–195.16. Moore LV, Diez Roux AV: Associations of neighborhood characteristicswith the location and type of food stores. Am J Public Health 2006,92(2):325–331.17. Cummins S, Macintyre S: Food environments and obesity - neighborhoodor nation? Int J Epidemiol 2006, 25:100–104.18. Raja S, Ma C, Yadav P: Beyond food deserts: measuring and mappingracial disparities in neighborhood food environments. J Planning Edu2008, 27:469–482.19. Zipf GK: Human Behaviour and the Principle of Least Effort. Massachusetts:Addison-Wesley; 1949.20. Frank LD, Sallis JF, Conway T, Chapman J, Saelens B, Bachman W: Manypathways from land use to health: walkability associations with activetransportation body mass index and air quality. J Am Plann Assoc 2006, 72(1):75–87.21. Frank LD, Bradley M, Kavage S, Chapman J, Lawton K: Urban form traveltime and cost relationships with work and non-work tour complexityand mode choice. Transportation 2007, 35(1):37–54.22. Sharkey JR, Horel S: Neighborhood socioeconomic deprivation andminority composition are associated with better potential spatial accessto the ground-truthed food environment in a large rural area. J Nutr2008, 138(3):620–627.23. Frank LD, Andresen MA, Schmid TL: Obesity relationships with communitydesign physical activity and time spent in cars. Am J Prev Med 2004,27(2):87–96.24. Jago R, Baranowski T, Baranowski JC: Observed GIS and self-reportedenvironmental features and adolescent physical activity. Am J HealthPromot 2006, 20(6):422–428.25. Roemmich JN, Epstein LH, Raja S, Yin L, Robinson J, Winiewicz D:Association of access to parks and recreational facilities with the physicalactivity of young children. Prev Med 2006, 43(6):437–441.26. Shoemaker Braza MW, Seely A: Neighborhood design and rates ofwalking and biking to elementary school in 34 California communities.Am J Health Promot 2004, 19(2):128–136.27. Block D, Kouba J: A comparison of the availability and affordability of amarket basket in two communities in the Chicago area. Public Health Nutr2007, 9(7):837–845.28. Morland K, Wing S, Diez Roux A, Poole C: Neighborhood characteristicsassociated with the location of food stores and food service places.Am J Prev Med 2002, 22(1):23–29.29. Glanz K, Sallis JF, Saelens BE, Frank LD: Nutrition Environment MeasuresSurvey in stores (NEMS-S): development and evaluation. Am J Prev Med2007, 32(4):282–289.30. Moore LV, Diez Roux AV, Nettleton JA, Jacobs DR Jr: Associations of thelocal food environment with diet quality—a comparison of assessmentsbased on surveys and geographic information systems: the multi-ethnicstudy of atherosclerosis. Am J Epidemiol 2008, 167(8):917–924.31. Liese AD, Weis KE, Pluto D, Smith E, Lawson A: Food store types availabilityand cost of foods in a rural environment. J Am Diet Assoc 2007, 107(11):1916–1923.32. Alwitt LA, Donley TD: Retail stores in poor urban neighborhoods.J Consum Aff 1997, 31(1):139–164.33. Matthews SA: The salience of neighborhood: some lessons fromsociology. Am J Prev Med 2008, 24(3):257–259.34. Bowman J, Ben-Akiva M: Activity-based disaggregate travel demandmodel system with activity schedules. Transportation Research Part A: Policyand Practice 2001, 35(1):1–28.35. Timmermans HJP, van der Waerden P, Alves M, Polak J, Ellis S, Harvey AS,et al: Spatial context and the complexity of daily travel patterns: aninternational comparison. J Transp Geogr 2003, 11:37–46.36. Weber CL, Matthews HS: Food-miles and the relative climate impacts of foodchoices in the United States. Environ Sci Technol 2008, 42(10):3508–3513.37. Zenk SN, Schulz AJ, Israel BA, James SA, Bao S, Wilson ML: Fruit andvegetable access differs by community racial composition andsocioeconomic position in Detroit Michigan. Ethn Dis 2006, 16(1):275–280.38. Zenk SN, Schulz AJ, Matthews SA, Odoms-Young A, Wilbur J, Wegrzyn L,Gibbs K, Braunschweig C, Stokes C: Activity space environment anddietary and physical activity behaviors: a pilot study. Health Place 2011, 17(5):1150–1161.39. Frank LD, Kerr J, Saelens B, Sallis J, Glanz K, Chapman J: Food outlet visitsphysical activity and body weight: variations by gender and race-ethnicity. Br J Sports Med 2009, 43:124–131.40. Austin SB, Melly SJ, Sanchez BN, Patel A, Buka S, Gortmaker SL: Clusteringof fast-food restaurants around schools: a novel application of spatialstatistics to the study of food environments. Am J Public Health 2005,95(9):1575–1581.41. Pereira CA, Larder N, Somerset S: Food acquisition habits in a group ofAfrican refugees recently settled in Australia. Health Place 2010,16(5):934–941.42. Kelly B, Flood VM, Yeatman H: Measuring local food environments:an overview of available methods and measures. Health Place 2011,17(6):1284–1293.43. Fleischhacker SE, Evenson KR, Rodriguez DA, Ammerman AS: A systematicreview of fast food access studies. Obes Rev 2011, 12(5):e460–e471.44. Walker RE, Keane CR, Burke JG: Disparities and access to healthy food inthe United States: a review of food deserts literature. Health Place 2010,16(5):876–884.45. Els Hannes, Davy Janssens, Geert Wets: Does Space Matter? Travel ModeScriptsin Daily Activity Travel. Environ Behav 2009, 41.46. Wang MC, Gonzalez AA, Ritchie LD, Winkleby MA: The neighborhood foodenvironment: sources of historical data on retail food stores. Int J BehavNutr Phys Act 2006, 3:1–5.Kerr et al. International Journal of Behavioral Nutrition and Physical Activity 2012, 9:58 Page 9 of 10http://www.ijbnpa.org/content/9/1/5847. Frank L, Van Loon J, Raine K, Mineker L, Fisher P, Thompson M: Underreview. Evaluating built environment relationships with physical activity,diet, and body weight: the NEWPATH study. Can J Public Health, .doi:10.1186/1479-5868-9-58Cite this article as: Kerr et al.: Predictors of trips to food destinations.International Journal of Behavioral Nutrition and Physical Activity 2012 9:58.Submit your next manuscript to BioMed Centraland take full advantage of: • Convenient online submission• Thorough peer review• No space constraints or color figure charges• Immediate publication on acceptance• Inclusion in PubMed, CAS, Scopus and Google Scholar• Research which is freely available for redistributionSubmit your manuscript at www.biomedcentral.com/submitKerr et al. International Journal of Behavioral Nutrition and Physical Activity 2012, 9:58 Page 10 of 10http://www.ijbnpa.org/content/9/1/58


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items