Embed Size (px)
Citation preview
Systematic Review
Grocery Store (or Supermarket) Tours asan Effective Nutrition Education Medium:A Systematic Review Cassandra J. Nikolaus, BS1; Henna Muzaffar, PhD, RD2;Sharon M. Nickols-Richardson, PhD, RD21Divisiobana, IL2DepartmChampaConflict owith thiAddressFood ScIL, 905 S4498; Fa�2016 Sreservedhttp://dx
544
ABSTRACT
Objective: To evaluate evidence regarding grocery store tours as an effective nutrition education me-dium for improving nutrition knowledge and food-related behaviors.Design: A systematic literature review of studies published from 1984 to 2015 concerning grocery store(or supermarket) tours and impact on nutrition knowledge and behaviors. Three investigators indepen-dently reviewed articles, extracted details, and assessed the quality of each study.Results: Of 307 citations identified, 8 were reviewed and 6 were of neutral quality. Increases in nutritionknowledge were reported in 4 studies, as evaluated by investigator-designed quizzes, with short intervalsbetween tours and assessments. Six programs assessed behavior change using subjective reports or objectivepurchasing behavior measures; 2 studies did not perform statistical analyses. The 6 studies that reported pos-itive health-related outcomes had varying topics, tour lengths, and target audiences.Conclusions and Implications: Grocery store tours are increasingly used as an avenue for nutrition ed-ucation to improve knowledge and/or alter food selection behaviors and may result in positive outcomes,but it is unknownwhether these outcomes persist for longer than 3months after the tour and whether thereare common attributes of effective grocery store tours. More rigorous studies with uniform methodologyin study design and outcome measures are needed to confirm the effectiveness of supermarket tours.Key Words: grocery store, health behaviors, nutrition education, supermarket, tour (J Nutr Educ Behav.2016;48:544-554.)
Accepted May 28, 2016. Published online July 14, 2016.
INTRODUCTION
Despite a well-established link betweenhealthful dietary patterns and lowerrisk of chronic disease, many adultsand children fall short of dietary recom-mendations that promote health.1 Oneactivity that may facilitate the achieve-mentofhealthfuldietarypatterns is gro-cery shopping for healthy foods.Interestingly, the average number oftrips to a supermarket in the US has
n of Nutritional Sciences at the Unive
ent of Food Science and Human Nutrign, Urbana, ILf Interest Disclosure: The authors’ conflis article on www.jneb.org.for correspondence: Sharon M. Nickoience and Human Nutrition, University. Goodwin Ave, 260 Bevier Hall, MCx: (217) 265-0925; E-mail: nickrich@ilociety for Nutrition Education and Beh..doi.org/10.1016/j.jneb.2016.05.016
declined from 2.2 trips/wk in 2010 to1.5 in2015,2 likelyowing toacontinuedincrease ineatingaway fromhome3andgreater use of convenience foods.4-6
Participants in a recent focus groupindicated that price and knowledgeabout the risks and benefits of foodchoices were important motivators forhealthful grocery purchasing choices.7
In addition, 53% to 62% of US con-sumers report at least sometimes usingnutrition facts labels,8,9 but many
rsity of Illinois at Urbana-Champaign, Ur-
ition at the University of Illinois at Urbana-
ct of interest disclosures can be found online
ls-Richardson, PhD, RD, Department ofof Illinois at Urbana-Champaign, Urbana,-182, Urbana, IL 61801; Phone: (217) 244-linois.eduavior. Published by Elsevier, Inc. All rights
Journal of Nutrition Education and Beh
perceive actual label comprehensionand use for product comparisons to beconfusing and difficult skills.10 Re-searchers and policymakers have calledfor strategies to increase nutrition labeluse and comprehension.1,9,11,12
One method to address consumers'grocery shopping practices and increasenutrition label comprehension isfacilitator-led grocery store (or super-market) tours. Such structured tourstake small groups throughagrocery storeto teach participants strategies andskills that enable healthful foodpurchasing choices. Dietitian LeniReed is credited with beginning super-market tours in the early 1980s, asshe took individuals out of traditionalclassrooms and brought them into storesforexperiential learning.13,14Basedonthepracticality of grocery stores containingtangible food items and providing con-textual learning environments, super-market tours have flourished as anutrition education medium. However,the effectiveness of grocery store tourshas not been evaluated systematically.
avior � Volume 48, Number 8, 2016
Additional records identified with reference lists and a cited reference search,
n = 10
Total records identified,
Records identified through database searching,
n = 297
Journal of Nutrition Education and Behavior � Volume 48, Number 8, 2016 Nikolaus et al 545
A recent systematic review summa-rized many supermarket-based inter-ventions.15 Only 2 grocery store tourswere included,andfindingswerepooledwith other assorted point-of-purchaseinterventions that used shelf markersor printed brochures as the primaryintervention.15 In contrast, the currentreview focuses only on outcomes fromin-persongrocery store tours to examinethe effectiveness of contextual learningon consumer knowledge and behaviors.
In the current review, a supermarkettour or grocery store tour is operation-ally defined as the dissemination ofnutrition information and/or foodshopping strategies by an educator toa small group of individuals while mov-ing from aisle to aisle within a marketthat sells a wide variety of food prod-ucts. Using this definition, 3 primaryresearch questions guided this review:(1) Do grocery store tours lead to knowl-edge gains? If so, are increases in knowl-edge retained $ 3 months after theintervention? (2) Do grocery store tourslead to behavior change? If so, do thesechanges remain $ 3 months after theintervention? (3) What attributes ofgrocery store tours are associated withhealth-related outcomes being posi-tively met? Three months was consid-ered a desirable follow-up intervalbased on the Transtheoretical Modelof behavior change, which considersthis to be an appropriate estimate forthe transition from an action to amain-tenance stage of change.16
n = 307
Records after duplicates removed,n = 274
Records screened,n = 274
Full-text records assessed for eligibility,
n = 32
Records excluded based on title and/or abstract, n = 241
Studied included in qualitative synthesis,
n = 8
Full-text articles excluded, n = 24› Grocery tour played minor role in
intervention, n =13› No health-related outcome data, n = 7› Only abstract available, n = 3› Grocery tour held virtually, n = 1
Unable to obtain full-text of potentially relevant article, n = 1
Figure 1. Flow diagram depicting study selection for the systematic literature reviewon grocery store (supermarket) tours.
METHODSSearch Strategy
One graduate student of nutritional sci-ences (CJN) and2 registereddietitian in-vestigators (HM and SNR) conducted asystematic searchand reviewof the liter-ature published between December,2014 and June, 2015, using guidelinespublished by the Centre for Reviewsand Dissemination.17 Two Internet da-tabases were used to identify resources(EBSCO Host Academic Search Com-plete and the Springer Standard Collec-tion, which index over 14,000 journalsand abstracts). Search terms, used invarying combinations, included ‘‘tour,’’‘‘nutrition,’’ ‘‘grocery,’’ ‘‘education,’’ ‘‘su-permarket,’’and ‘‘grocerystore.’’ Inaddi-tion, a manual search of each relevantarticle's references was conducted, anda cited reference searchbasedon all rele-
vant articles was performed to expandthe scope to the latest publications viaWeb of Science.
For inclusion, resources had to havebeen published in English as aresearch-based article or abstract be-tween January, 1984 and April, 2015,and had to have used a supermarkettour as a nutrition education methodand reported outcomes directly attribut-able to the tour. All types of research de-signs were included. Contact withauthors of relevant published paperswas prohibited after the review wasinitiated, to avoid potential bias.18 Toprovide a comprehensive review of theliterature, studies were not excludedbased on the type of grocer, size of thesample, target audience, specific focusof the tour's subject matter, or studydesign. Abstracts were not includedowing to inadequate details for fulldata extraction and the inability toassess quality. A total of 307 citationswere identified. After duplicates wereremoved, 241 records were excludedbased on their title or abstract. One rele-vant article was not included because ofthe inability to locate research detailsbeyond the title. The lead investigator
examined 32 full-text publications. Ar-ticles were excluded if no knowledgeor behavior outcome data were re-ported, if the tour did not match theoperational definition provided, or ifthe supermarket tour had a minor rolein the intervention and the outcomescould not be linked solely to the tour.Eight studies were included after exclu-sion criteria were applied (Figure 1) andwere critically reviewed by each of 3 in-vestigators, during which details wereextracted.
Data Extraction and QualityAssessment
Data related to study characteristics,participant characteristics, interven-tion, and setting, as well as outcomedata and results were extracted fromeach study in the final critical review.17
Although the focus was on nutritionknowledge and behavior, all outcomeswere examined and evaluated. Eachinvestigator independently performeddata extractions, and component qual-ities were agreed upon after the au-thors discussed any discrepancies.
546 Nikolaus et al Journal of Nutrition Education and Behavior � Volume 48, Number 8, 2016
Each study was classified as A, B, C,or D, according to the strength of theresearch design, based on the EvidenceAnalysis Manual's19 hierarchy andclassification system, to provide aninitial snapshot of its level of evi-dence. A quality criteria checklist19
was used to rate each study as nega-tive, neutral, or positive. These ratingswere based on answers to 4 relevancequestions and 10 validity questionsas determined by details reported ineach article. If all relevance and mostvalidity questions were positivelyanswered as yes, an article was ratedpositive. In contrast, negative ratingswere assigned when a minority of val-idity questions was answered posi-tively, and neutral ratings wereassigned when there was a mixtureof positive and negative answers.Three investigators independentlyconducted quality assessments foreach full article, and final quality rat-ings were agreed upon after any dis-crepancies were discussed. Researchcharacteristics and quality ratings foreach article are displayed in the Table.
Unique measurement tools usedacross studies and insufficient report-ing of statistical analyses prevented ameta-analysis. Thus, a descriptive syn-thesis that explored themes and limita-tions of the current body of researchwas conducted. Summaries regardingknowledge and behavior outcomesalong with theoretical foundations ofstudies were prepared. The discussionof study characteristics associated withpositive health outcomes includesonly studies that reported statistical an-alyses, owing to limitationsof implyingsignificance without analyses.
RESULTS
Six studies were non-randomized non-controlled trials with a D classifica-tion20-25 and 2 were non-randomizedcontrolled trials with a C classifica-tion.26,27 Quality assessments of the 8articles revealed 1 positive rating,27 6neutral,20,22-26 and 1 negative.21
Some investigations did not explic-itly report participant gender.21-24 Ofstudies that did,20,25-27 most hostedtours for exclusively female partici-pants.25-27 No studies reported thesocioeconomic status of tour partici-pants, but low socioeconomic statuswas inferred in 1 study that conducted
tours with Special Supplemental NutritionProgram for Women, Infants, and Childrenparticipants.26 The 2 most commonfoci for grocery store tours were gen-eral healthful dietary behaviors21,24,27
and cardiac health–related dietary be-haviors.20,23
Study Designs
Six studies used a non-controlled trialresearch design in which an interven-tion group was the only source ofdata.20-25 Pre–post tests were used to as-sess participant outcomes in 4 of thesestudies22-25 whereas the other 2 includedonly posttests.20,21 Two programs usednaturally occurring control groups intheir recruited participant pool to com-pare with their tour group.26,27 Noneof the studies cited sample size goalsbased on power analyses. Sample sizesranged from 9 participants20 to 947.24
Of 7 studies that reported the size oftour groups,20-25,27 the majority hadgroups of # 15.20,21,23,25,27 In addition,statistical analyses were not presentedin 2 of the 8 articles.20,21
Knowledge
Five of the studies measured knowledgechange inparticipants20,22,24-26; 4 of thesefound that participants had an increasein knowledge after the tour.20,22,25,26
Investigator-designed, topic-specific,multiple-question quizzes were themost commonmethod for knowledge as-sessment.22,24-26 These quizzes weregiven to participants before and afterattending the tour, to assess change.One study assessed knowledge changebased on participants' written reportsof information they had gained afterthe tour in an open-answer format,and increased knowledgewas reportedbased on the number of individualswho opted to write in an answer.20
None of the articles reported the valid-ity or reliability of these instruments.
Each study that assessed knowl-edge20,22,24-26 chose to measure this atunique time intervals after the tour.One study assessed knowledge gainimmediately after the tour withoutadditional follow-up.22Other studies as-sessedknowledgewithadelayed follow-up ranging from 1week24 to 1month20
and even up to 2 months after thetour.25,26 None of the studies thatmeasured knowledge assessed reten-tion $ 3 months after the tour, the
amount of time that would placeindividuals in transition from an actionto a maintenance stage of behavioralchange.16
Behavior
Behavior assessments were collected in6 of the 8 studies.20,21,23,25-27 Measurementtools were widely heterogeneous amongstudies, butmost assessed behavior changewith subjective self-reports from par-ticipants. Self-report measures variedfrom general reports of ‘‘increasing va-riety indiets’’20 or ‘‘reducing fat intake’’25
to reported frequency of using strate-gies presented or purchasing productshighlighted in the tour.21,26,27 One studyassessed purchasing behavior changewith objective measurements, by usinggrocery store loyalty card purchasingdata for a 7-week period before the tourand a 7-week period after the tour.23
Neither the validity nor the reliability ofassessment instruments was addressed.
All 6 studies that assessed behaviorchange reported that participants had atleast 1 positive behavior change after at-tending the grocery store tours.20,21,23,25-27
However, few studies described dataanalyses.23,26,27 In addition, the intervalbetween the grocery tour and follow-upwithparticipants varied among studies.The shortest follow-up interval was 1month after the tour20,27 and the longestinterval was 3 months after the tour.21
The study using loyalty cards was un-clear regarding the interval betweenthe tour and the follow-up data collec-tion; it simply reported that data werecollected for a 7-week period.23
Attributes of Tours WithPositive Outcomes
The 8 studies measured additionalhealth-related outcomes, such as partic-ipant intentions,21,22,25,26 attitudes,24,26
dietarybehavior,20,21,25-27 andpurchasingbehaviors.23 The current researchersextracted characteristics from the 6 ar-ticles that reported statistical ana-lyses.22-27 Of these 6 studies, 5 foundpositive change in at least 1 health-related outcome.22,23,25-27 The topics ofthese tours varied, audiences differedby age and gender, and the length oftours ranged from11.5-hoursession23,25
to 3 1-hour sessions.26 Increasedknowledge and behavior change werereported after adult women of low
Table. Characteristics of Studies Assessing Grocery Store (Supermarket) Tours, 1984–2015 (n ¼ 8)
Reference (ResearchDesign, Class,a
Quality Ratingb)
Participant Sample Sizeand DescriptiveCharacteristics
Intervention Title,Recruiting, TheoreticalBasis, Design, Objective,
Tour Length, BasicComponents
Primary Outcomes:Tool Characteristics
and Timing
Additional Outcomes:Tool Characteristics
and TimingPrimary and Additional
Results
Baic and Thompson(2007)20
Design: Non-controlled trialClass: DNeutral
Sample size: n ¼ 58participants; n ¼ 10follow-up surveyrecipients and 9 returned
Descriptive characteristics:Equal amount of men andwomen
Heart Healthy ToursRecruiting: In-store posters,local print media, healthprofessional referrals
Theoretical basis: NRObjective: Assessment offeasibility andacceptability
Tour periods: n ¼ 1Tour length: NRTour size: 5–15 attendeesBasic components:Focused on heart-healthyfoods; label education;food preparation
Knowledge: Assessedn ¼ 10 clients recruitedfrom health professionalsvia anonymous mailsurvey; 1-mo posttour
Behavior: Measured withthat survey
Interest: 1 question askedto n ¼ 48 posttour
Information novelty: 1question asked to n ¼ 48posttour
Primary ResultsKnowledge: Newinformation was reportedin open-answer format(n ¼ 10)
Behavior: 100% (n ¼ 9)reported increasing dietvariety, 89% (n ¼ 9) felt itwas easier to followhealthy diet, (n¼ 9) madedietary changes
Additional ResultsInterest: 98% (n¼ 48) foundtour interesting
Information novelty: 75%(n¼ 48) reported learningnew information
Carson and Hedl (1998)26
Design: Non-randomizedcontrolled trial
Class: CNeutral
Sample size: n ¼ 315women completedpretest; n ¼ 128 womencompleted pretest andattended tour(s); n ¼ 114women attended tour(s)and completed 2-monthfollow-up
Descriptive characteristics:Average age 28 y, parentof 2 children,unemployed with somehigh school education.Majority Caucasian andWIC recipients
Smart Shoppers TourRecruiting: Peer-leaderssolicited participantsfrom WIC clinics andother assistanceprograms
Theoretical basis: Theory ofPlanned Behavior andSelf-efficacy Theory
Objective: Increase healthyfood purchases usinglabels and unit costs(facilitated by changes inattitudes, self-efficacy,and knowledge)
Tour periods: n ¼ 3Tour length: 1 hTour size: NR
Knowledge: 16-questiontest with word/picturematching; at recruitmentand 2 mo after tour
Behavior: 2 items, agree/disagree to statements: Iuse the nutrition label andI compare unit cost;home food inventory oftour-specific foods atrecruitment and2 months after tour
Attitude: 2 items, agree/disagree to: Healthyfoods do not taste good,and Healthy foods costmore, at recruitment and2 mo after tour
Intention: 5 items, agree/disagree to statementsabout intending to buyhealthy foods, atrecruitment and 2 moafter tour
Self-efficacy: 1 questionasked, whether agree/disagree to: Have skills tobuy healthy foods forfamily, at recruitment and2 mo after tour
Primary Results:Knowledge: Mean quizscores increased 8%from 11.96 (of 16) to12.89, with effect size of0.39 (P < .001)
Behavior: Increasednutrition label use(P ¼ .005); no change inunit cost use
Additional ResultsAttitudes: Increaseddisagreement thathealthy foods cost more(P¼ .007) and no changeregarding attitude thathealthy foods taste good
Intention: Increased
(continued)
JournalofNutritio
nEducatio
nandBehavior�
Volume48,
Number
8,2016
Nikolau
set
al547
Table. Continued
Reference (ResearchDesign, Class,a
Quality Ratingb)
Participant Sample Sizeand DescriptiveCharacteristics
Intervention Title,Recruiting, TheoreticalBasis, Design, Objective,
Tour Length, BasicComponents
Primary Outcomes:Tool Characteristics
and Timing
Additional Outcomes:Tool Characteristics
and TimingPrimary and Additional
Results
Basic components:Focused on grains,produce, low-fat meatand dairy; meal planning;shopping tips; taste tests.Led by trained peers inEnglish and Spanish
Subjective norms: 2questions, askedwhether agree/disagreeto: Their family thinks it isgood to eat a certainamount of fruits andvegetables each day, atrecruitment and 2 moafter tour
intentions to buy wholegrains (P¼ .006) and low-fat milk (P ¼ .002). Nochange regarding fruits,vegetables, or low-fatmeat.
Self-efficacy: Increase inagreement (P ¼ .002)
Subjective norms: Nochange
Crawford and Kalina(1993)21
Design: Non-controlled trialClass: DNegative
Sample size: n ¼ 48participants
Descriptive characteristics:NR
The Shop Smart Tour
Recruiting: In-storeadvertisements, mailedfliers, word-of-mouth
Theoretical basis: NRObjective: Increase foodchoices that decreaserisk of nutrition-relateddiseases
Tour periods: n ¼ 1Tour length: 1.5 hTour size: 8–10 attendeesBasic components: Basedon Supermarket NutritionEducation Kit; labeleducation; alternativefood products, cost-effectiveness; foodpreparation
Knowledge: NRBehavior: Assessmentwhether dietary changeintention had beenrealized. Purchasingbehavior assessed 3-moposttour
Intention: Assessedintentions regardingspecific behaviors andfood groups posttour
Primary ResultsBehavior: People generallymade intended dietarychanges with exceptionof decreasing dietary fat.Greater number ofproducts were beingpurchased
Additional ResultsIntention: Expressedintentions included:33.3% read food labels,29.2% reduced dietaryfat and 22.9% reducedred meat consumption.41.7% incorporated low-fat cheese/yogurt intodiet, 29.2% incorporatedlegumes, and 22.9%incorporated wholegrains
Lafferty et al (2006)22
Design: Non-controlled trialClass: DNeutral
Sample size: n ¼ 25students (grades 4–5)and n ¼ 27 parents andchildren
No titleRecruiting: NRTheoretical basis: NRObjective: Improve
Knowledge: Asked labelterms and location ofwhole-grain productswithin store pre- and
Intention: Assessedintentions to ask forparents to purchasewhole-grain products
Primary ResultsKnowledge: Parents andchildren both significantlyincreased knowledge,
548Nikolau
set
alJournalofNutritio
nEducatio
nandBehavior�
Volume48,
Number
8,2016
Descriptive characteristics:NR
knowledge of and abilityto identify whole-grainproducts
Tour periods: n ¼ 1Tour length: NRTour size: 25–27 attendeesBasic components: Lecturebefore tour, focused onidentifying whole-grainproducts; Student groupincluded taste-test;Parent–child group hadpackage term activity andrecipes provided
posttourBehavior: NR
(student group) orintentions to purchaseproducts (parent–childgroup); pre- and posttour
Skills: Ability to differentiatewhole-grain from refined-grain products based oningredient lists andnutrition facts panels,pre- and posttour
but in different areas(P < .01 overall)
Additional ResultsIntention: Parents hadgreater intention topurchase children’srequested whole-grainproducts (P NR)
Skills: Parents improvedability to identify whole-grain products (P < .003)
Sadler et al (2003)23
Design: Non-controlled trialClass: DNeutral
Sample size: n ¼ 459participants gaveevaluation; n ¼ 223provided purchasinginformation
Descriptive characteristics:NR
Healthy Heart Store ToursRecruiting: Posters andannouncements instores, advertisements inlocal press and in offices
Theoretical basis: NRObjective: Provide tour forthose interested in hearthealth
Tour periods: n ¼ 1Tour length: 1.5 hTour size: # 10 attendeesBasic components: 15-minlecture before tour;emphasis on specificaisles and products,followed by formalquestion and answerperiod
Knowledge: NRBehavior: Purchasingbehavior evaluated withdata from loyalty card; 7-wk period before and 7-wk period after tour
Motivation for attendance: 1question askedparticipants why theychose to attend, posttour
Primary ResultsBehavior: Increase inhealthier spreadpurchases by 4%(P < .05) and fruit andvegetable purchasesdecreased by 5% (P NR)
Additional ResultsMotivation for attendance:Most frequent reason(42%) was healthproblem present inparticipants or someonethey knew
Silzer et al (1994)27
Design: Non-randomizedcontrolled trial
Class: CPositive
Sample size: n ¼ 61females in treatmentgroup, 12 lost to follow-up (n ¼ 49 final follow-uptreatment group); n ¼ 63females in control group
Descriptive characteristics:Majority aged 30–39 ywith at least high schooleducation
Supermarket SafariNutrition Education Tours
Recruiting: NR; participantscalled health department
Theoretical basis: NRObjective: Identify changesin food selection andpreparation after tour,compared with control
Tour periods: n ¼ 1
Knowledge: NRBehavior: 16-question skilland behavior checklist;frequency and yes/no/donot know format askedtypes of food purchased,information read onlabels, and foodpreparation practices;
None Primary ResultsBehaviors: Gain in behaviorsurvey score wassignificantly higher intreatment groupcompared with controlgroup (P < .001)
Additional ResultsNone
(continued)
JournalofNutritio
nEducatio
nandBehavior�
Volume48,
Number
8,2016
Nikolau
set
al549
Table. Continued
Reference (ResearchDesign, Class,a
Quality Ratingb)
Participant Sample Sizeand DescriptiveCharacteristics
Intervention Title,Recruiting, TheoreticalBasis, Design, Objective,
Tour Length, BasicComponents
Primary Outcomes:Tool Characteristics
and Timing
Additional Outcomes:Tool Characteristics
and TimingPrimary and Additional
Results
Tour length: 2 hTour size: 10–12 attendeesBasic components: Led bydietitian; focused on low-fat, low-sodium foodsand high-fiber foods;label education; foodpreparation
pretour and 1-moposttour via mail
Smith and Kalina (2004)24
Design: Non-controlled trialClass: DNeutral
Sample size: n ¼ 947participants; n ¼ 496returned pre and postattitude surveys; n ¼ 45students (in grade 3classes) returnedknowledge quizzes
Descriptive characteristics:68% did not reportethnicity and 11%reported special dietaryneeds
Kids Shop Smart ToursRecruiting: NRTheoretical basis: NRObjective: Impact attitudestoward trying new varietyof foods and developability to recognize 4 foodgroups of Canada’s FoodGuide to Healthy Eating
Tour periods: NRTour length: NRTour size: 20 attendeesBasic components: NR,described in separatearticle not accessible attime of review
Knowledge: Quiz withquestions about 4 foodgroups and examplesthat would be in each,asked to treatment andcontrol grade 3 classes1 week after tour
Behavior: NR
Attitude: Questions aboutattitude regarding tryingnew foods and eating avariety of foods with aLikert-type scale asked tocaregivers, 1 wk pre- andposttour
Primary ResultsKnowledge: Averagescores (out of 20) were10.24 for treatment groupand 11.4 for controlgroup (difference notsignificant)
Additional ResultsAttitude: Regarding tryingnew foods, pretour meanof 4.6 (out of 7)decreased to 4.59posttour, (change notsignificant). Pretour meanregarding eating variety offoods was 4.54 andincreased to 4.61posttour (change notsignificant)
van Assema et al (1998)25
Design: Non-controlled trialClass: DNeutral
Sample size: n ¼ 419 givenposttour assessment;n ¼ 145 given 5- to 7-wkfollow-up behaviorassessment; n ¼ 310given pretour knowledgeassessment; n ¼ 305given posttourknowledge assessment;
No title, part of Beware ofFat campaign
Recruiting: Pressconference, pressreleases, localnewspaperadvertisements, postersand pamphlets; localgroups were contacted
Knowledge: 9-questionquiz, assessed at pre-and posttour and 5- to 7-week follow-up phonecall
Behavior: Questions on fatintake reduction; 5- to 7-week follow-up phonecall
Intention: Questions onintentions to reduce fatintake, posttour
Awareness: Questions onawareness of fat intake,posttour
Primary ResultsKnowledge: Student groupincreased from 5.3 to 6.4posttour (P< .01) and 6.2at follow-up (P NR). Adultscores increased from5.8 to 6.8 posttour(P < .01) and 6.6 atfollow-up (P NR).
550Nikolau
set
alJournalofNutritio
nEducatio
nandBehavior�
Volume48,
Number
8,2016
n¼
86given5-to
7-w
kfollow-upkn
owledge
assessment
Descrip
tivecharacteristics:
Majorityfemale,prim
ary
foodsh
opper,rangeof
ages(includingstudents
age<
16,andwomen
aged>
40y),prim
arily
low
educatio
nlevel,in
householdswith
$2
family
members
Theoreticalb
asis:
NR
Objective:Increase
knowledgeand
awareness
offatintake
/content;create
intentio
nto
reducefatintake
Tourperio
ds:
n¼
1Tourlength:1–2
hToursize:8attendees
Basiccomponents:Brie
flecture
before
tour;ledby
dietitian;focusedon
differentfoodgroups;
usedposters
and
products
andsm
alltasks
toke
epparticipants
engaged;tastetest
of
cheeseswith
differentfat
content;materia
lsgiven
toparticipants
totake
home
Behavior:63%
ofstudents
and75%
ofadults
reportedreducingfat
intake
AdditionalResults
Intentio
n:45%
ofstudents
and70%
ofadults
reportedintentio
nto
increase
purchasesof
low-fatitems
Awareness:Most
were
more
aware
oftheirown
fatintake
NRindicatesnotreported;WIC,SpecialSupplementalNutritionProgram
forWomen,Infants,andChild
ren.
aStudieswere
classifiedasA,B,C,orDaccordingto
thestrength
oftheresearchdesign,basedontheAcademyofNutritionandDietetics’EvidenceAnalysisManual’s1
9
hierarchyandclassificationsystem;bAqualitycriteriachecklistfrom
theAcademyofNutritionandDietetics’EvidenceAnalysisManual19wasusedto
rate
eachstudyas
negative,neutral,orpositive.Ratingswere
basedonanswers
to4relevancequestionsand10valid
ityquestionsaccordingto
inform
ationreportedwithineacharticle.Ifall
relevanceandmostvalid
ityquestionswere
checkedasyes,anarticlewasratedpositive.In
contrast,negativeratingswere
assignedwhenaminority
ofvalid
ityquestions
wasansweredpositively,andneutralratingswere
assignedwhenthere
wasamixture
ofpositiveandnegativeanswers.
Journal of Nutrition Education and Behavior � Volume 48, Number 8, 2016 Nikolaus et al 551
incomecompleted 31-hour tours.26 In-crease in knowledge was reported byparents and children after attending atour focused on whole grains.22 A car-diac dietary pattern was emphasizedin 1 1.5-hour tour that led to greaterpurchasing of heart-healthy foodsamong participants (the characteristicsof which were not reported).23 Adultwomen reported positive behaviorchange after attending 1 2-hour touron general healthful dietary behav-iors.27 Children and adults reported astatistically significant increase inknowledge after completing a 1 1.5-hour grocery store tour focused onlow-fat foods.25 Additional details ofeach study are found in the Table.
Theoretical Basis
Only 1 article mentioned a theoreticalbasis for the grocery store tour.26 TheSmart Shoppers Tour was based on theTheory of Planned Behavior and Self-efficacy Theory with the goal ‘‘to in-crease the purchase of healthy foodsby mothers on limited incomes (p.323).’’26 This study foundpositive changesin attitudes and intentions related totour goals.26 Although they did notreport the theoretical basis, 3 addi-tional studies measured intentionsand reported positive change in partic-ipants after attending tours.21,22,25
Participant attitudes were measuredin 1 additional study, but no change wasfound after participants attended atour.24 Of the 4 studies that foundpos-itive changes in these theoretical me-diators of dietary behavior change, allreported additional positive changeswhen they measured participantknowledge or behavior.21,22,25,26
DISCUSSION
The objective of this review was toidentify attributes of grocery store toursthat promote and sustain improve-ments in knowledge and behaviorchange. Low-quality studies preventedthe ability to answer the primaryresearch questions. Limited evidencesuggests that grocery store tours havethe potential to increase knowledgeand improve behaviors, but suchstudies are unique regarding multipledesign parameters, and commonalitiesacross studies do not exist to guidepractices adequately at present. When
Identification of Need◊ Conduct a community nutrition needs assessment to identify whether a grocery store tour is
necessary (and what type of nutrition education intervention is desired) for the target audienceTour Components
◊ When available and appropriate for the target audience, use a previously existing curriculum to facilitate comparability of results across previous studies and program reports of grocery store tours
◊ If an appropriate curriculum is not available: » Ground a new tour in a theoretical framework of behavioral change» Develop the tour based on appropriate dietary guidance for the target audience » Establish content validity with expert review
Research Design◊ Apply a strong design, such as randomized controlled trial or clustered randomized trial, to
optimize strength of outcomes and conclusions from the grocery store tour◊ Conduct a power analysis to identify the sample size needed to detect the tour’s effect◊ Obtain approval for research with human participants before initiating the tour◊ Conduct statistical analyses, beyond simply reporting of participant characteristics
Assessment◊ Tailor the evaluation to the dose and content of the tour◊ Measure outcomes using tools with established validity and reliability◊ If valid and reliable measurement tools or techniques are not available:
» Develop and pilot test new assessment materials to demonstrate validity and reliability of these instruments
◊ Follow-up with participants at least three months after the tour to evaluate retention of change on outcome measures
Evaluation and Dissemination◊ Statistical significance and tests should be reported with enough detail to be replicable◊ Share the findings of the intervention in appropriate peer-reviewed avenues
Additional Resources◊ Cates S, Blitstein J, Hersey J, Kosa K, Flicker L, Morgan K, Bell L. Addressing the Challenges of
Conducting Effective Supplemental Nutrition Assistance Program Education (SNAP-Ed) Evaluations: A Step-by-Step Guide. Prepared by Altarum Institute and RTI International for the U.S. Department of Agriculture, Food and Nutrition Service. 2014.
◊ Glasgow RE, Vogt TM, Boles SM. Evaluating the public health impact of health promotioninterventions: The RE-AIM framework. Am J Public Health. 1999;89(9):1322-1327.
◊ Monsen ER. Research: Successful Approaches. Chicago, IL: American Dietetic Association. 1992. ◊ United States Department of Agriculture. Nutrition Education: Principles of Sound Impact Evaluation.
Washington, DC: Food and Nutrition Service. September 2005.
Figure 2. Methodological considerations and recommendations for designing futuregrocery store (or supermarket) tour interventions.
552 Nikolaus et al Journal of Nutrition Education and Behavior � Volume 48, Number 8, 2016
measured, theoretical mediators ofbehavior change were positively changed,which suggests that foundations inbehavior change theory would be rec-ommended when creating futuresupermarket tours. To develop high-quality evidence of effective practices,valid and reliable methodologicaltools and high-quality study designsshould be used in the future. Super-market tours are a unique mode ofeducation that requires collaborationwith grocers as well as the transporta-tion of educators and participants toan atypical site, and are often con-ducted with small-size groups. Theirdistinct requirements and popularuse warrant the demand to produceevidence of their effectiveness.
The most salient observation fromthis review and synthesis of studiesis the lack of reported qualitiescommonly anticipated in scientific arti-cles. Two investigations did not reportstatistical analyses, andthus their resultswere only descriptive in nature withlimited ability to make inferences.Noneof thepublications included state-ments acknowledging institutional re-view for research involving humans.One research team that partnered withan elementary school mentionedapproval from an education director.24
This suggests that tours were designedfor programmatic and practical pur-poses, with lesser emphasis on effectivesystematic assessment. None of the arti-cles addressed the reliability andvalidityof instruments, which cannot beassumed.No studies reported power an-alyses for sample size needs. The studiesperformed with small samples wereassumed to be underpowered, and ef-fects would be more difficult to detect.Finally, the lack of control groups wasan inherent limitation of the non-controlled trial design used by the ma-jority of studies included in this review.No study used a study designwith anA-or B-quality classification.
Increased knowledgewas reported inseveral studies.20,22,25,26 However, eachassessment was custom-created for thetours' specific focus, andthuswas incon-sistent across interventions. Therefore,synthesis of results should be madewith caution. The short intervals be-tween tour and assessment should alsobe considered, because a lack of longer-term follow-up prevented the ability toaddress the impact of knowledge reten-tion beyond 3 months after the tour.
Retention of participant knowledgegain beyond the short term cannot beassumed without longer-term testing.
In all studies that assessed behavior,participants had at least 1 positivebehavior change after attending a su-permarket tour. However, concernsregarding social-desirability bias werepresent with many of the studies usingself-report measures. The corrobo-rating positive results from the studythat used objective purchasing datafrom receipts and loyalty card informa-tion lent further evidence to suggestthat positive behavior change mayhave resulted from attending grocerystore tours. However, the longest post-tour follow-up of 3 months was in astudy that did not report statistical an-alyses,21 so sustained change has yet tobe determined indisputably.
Constraints of included studiesmade it difficult to develop strongconclusions regarding characteristicsthat define an effective grocery storetour. Identifying tour qualities relatedto any of the additional outcomeswas not feasible with the currentresearch and the limited number ofhigh-quality studies available. Studiesthat found significant positive health-related outcomes did not differ appre-ciably from studies that did not reportthese outcomes. This indicates thatqualities related to successful changein positive health-related factors maybe found in unreported or unmea-sured facets of such interventions.
This review of the literature hadrecognized limitations. It is possiblethat inclusion of abstracts and unpub-lished work would have better informed
Journal of Nutrition Education and Behavior � Volume 48, Number 8, 2016 Nikolaus et al 553
the research questions. Only citationswritten in English were included inthe review; this was considered appro-priate because grocery store tours mayhave a different cultural role whenperformed in unique global cultures.Positive aspects of the review, suchas inclusion of all study designs andthe range of publication dates, pro-vided a comprehensive synthesis ofpublished work on supermarket tours.
Basedon the limited evidence, at thistime it would be unwise to develop atheory of how or whether supermarkettours are effective in promoting sus-tained positive health-related change.However, results suggesting positivechange in attitudes and intentionsalong with knowledge and behaviorchanges suggested that the Theory ofPlanned Behavior may be a desirablemodel that should be assessed withstronger experimental designs.Althoughnot reported explicitly by name in thecurrent literature, Social CognitiveTheory is another model that couldmediate the impact of supermarkettours. Tours may influence an individ-ual's self-efficacy related to groceryshopping behaviors, because observa-tional learning takes place in thecontextual grocery store environment.
IMPLICATIONS FORRESEARCH ANDPRACTICE
Because the current literature regardinggrocery store tours is primarily composedof lower-quality study designs, prom-isingknowledgeandbehavioroutcomeresults are not robust. Recommenda-tions for future studies in this area arepresented in Figure 2. Future researchshould use stronger study designs,including randomized controlled tri-als, to show a clear relationship be-tween intervention and any changein participants. Instruments to mea-sure health-related knowledge andbehavior change should be valid andreliable. Knowledge and behaviorshould be measured at least 3 months,if not 6 months, after the tour to eval-uate whether changes are maintained.Encouraging results from studies usingbehavior change theory, support theuse of theories when developing andconducting future research on super-market tours.
ACKNOWLEDGMENTS
This material is based on work that issupported by the National Instituteof Food and Agriculture, US Depart-ment of Agriculture, under AwardNo. 2012-68001-22032. The authorsthank the many program developersand research teams for pioneeringwork in this field. Their contributionsin the use of supermarkets as a novelteaching environment have provideda foundation for additional avenuesfor nutrition education of consumers.
REFERENCES
1. Dietary Guidelines Advisory Committee.Scientific report of the 2015 DietaryGuidelines Advisory Committee. http://health.gov/dietaryguidelines/2015-scientific-report/. Accessed July 2, 2015.
2. Food Marketing Institute. US groceryshopper trends 2015: executive summary.http://www.fmi.org/research-resources/u-s-grocery-shopper-trends-2015. Ac-cessed October 28, 2015.
3. Briefel RR, Johnson CL. Secular trendsin dietary intake in the United States.Annu Rev Nutr. 2004;24:401-431.
4. Poti JM, Popkin BM. Trends in energyintake among US children by eatinglocation and food source, 1977-2006. JAm Diet Assoc. 2011;111:1156-1164.
5. Nielsen SJ, Siega-Riz AM, Popkin BM.Trends in food locations and sourcesamong adolescents and young adults.Prev Med. 2002;35:107-113.
6. Kant AK, Graubard BI. Eating out inAmerica, 1987–2000: trends and nutri-tional correlates. Prev Med. 2004;38:243-249.
7. McGee BB, Johnson GS, Yadrick MK,et al. Food shopping perceptions, be-haviors, and ability to purchase health-ful food items in the lower Mississippidelta. J Nutr Educ Behav. 2011;43:339-348.
8. Blitstein JL, EvansWD.Use of nutritionfacts panels among adults who makehousehold food purchasing decisions. JNutr Educ Behav. 2006;38:360-364.
9. Ollberding NJ, Wolf RL, Contento I.Food label use and its relation to dietaryintake among US adults. J Am Diet As-soc. 2011;111(5 suppl):S47-S51.
10. Campos S, Doxey J, Hammond D.Nutrition labels on pre-packaged foods:a systematic review. Public Health Nutr.2011;14:1496-1506.
11. GrahamDJ, Laska MN. Nutrition labeluse partially mediates the relationshipbetween attitude toward healthy eatingand overall dietary quality among col-lege students. J Acad Nutr Diet. 2012;112:414-418.
12. Barnes M. Solving the problem of child-hood obesity within a generation: WhiteHouse Task Force onChildhoodObesityreport to the President. http://www.letsmove.gov/white-house-task-force-childhood-obesity-report-president. Ac-cessed October 28, 2015.
13. Sugarman C. Shopping tours practicemaking food decisions. The WashingtonPost. http://www.washingtonpost.com/archive/lifestyle/wellness/1990/06/12/shopping-tours-practice-making-food-decisions/bc38cf8a-9f19-44ae-94e0-168c025a3006/. Accessed August 15, 2015.
14. Burros M. Eating well. Living Desk.http://www.nytimes.com/1990/08/01/garden/eating-well.html. Accessed August15, 2015.
15. Escaron AL, Meinen AM, Nitzke SA,Martinez-Donate AP. Supermarketand grocery store-based interventionsto promote healthful food choices andeating practices: a systematic review.Prev Chronic Dis. 2013;10:E50.
16. Glanz K, Rimer BK, Viswanath K.Health Behavior and Health Education:Theory, Research, and Practice. 4thed. San Francisco, CA: Jossey-Bass;2008.
17. Centre for Reviews and Dissemination.Systematic Reviews: CRD’s Guidance forUndertaking Reviews in Health Care. 3rded. Layerthorpe, York, UK: Centre forReviews and Dissemination, Univer-sity of York; 2009.
18. Stroup DF, Berlin JA, Morton SC,et al. Meta-analysis of observationalstudies in epidemiology: a proposalfor reporting. J Am Med Assoc. 2000;283:2008-2012.
19. Academy of Nutrition and Dietetics.Evidence Analysis Manual: Steps in theAcademy Evidence Analysis Process. Chi-cago, IL: Academy of Nutrition andDietetics; 2012.
20. Baic S, Thompson JL. Prevent it: Usinggrocery store tours as an educationaltool to promote heart health. ACSMsHealth Fit J. 2007;11:15-20.
21. Crawford SM, Kalina L. The shopsmart tour. J Nutr Educ. 1993;25:100B.
22. Lafferty A, Marquart L, Reicks M.Hunting for whole grains: a super-market tour. J Nutr Educ Behav. 2006;38:197-198.
554 Nikolaus et al Journal of Nutrition Education and Behavior � Volume 48, Number 8, 2016
23. Sadler MJ, Fine G, Richards S, Read L.Healthy heart store tours—a usefulcommunication tool? Nutr Bull. 2003;28:179-186.
24. Smith SC, Kalina L. Evaluation of theKids’ Shop Smart� Tour. Can J DietPract Res. 2004;65:10-14.
25. van Assema P, Cremers S, van Dis I.Nutrition education tours in the super-market: the results of a pilot project inthe Netherlands. In: Worsley A, ed.Multidisciplinary Approaches to FoodChoice. Adelaide: Food Choice Confer-ence; 1996:174-178.
26. Carson JAS, Hedl JJ. Smart shopperstours: outcome evaluation. J NutrEduc. 1998;30:323-331.
27. Silzer JS, Sheeshka J, Tomasik HH,Woolcott DM.An evaluation of Super-market Safari nutrition education tours.J Can Diet Assoc. 1994;55:179-183.
Journal of Nutrition Education and Behavior � Volume 48, Number 8, 2016 Nikolaus et al 554.e1
CONFLICT OF INTEREST
The first author was supported by aKraft Human Nutrition Fellowshipthrough the Division of NutritionalSciences at the University of Illinois,
at the time of this review. The seniorauthor has received research fundingfrom the US Department of Agricul-ture, National Institute of Food andAgriculture; the Bell Institute of Nutri-
tion and Health, General Mills, Inc;Hershey Foods; and the Dairy Man-agement Institute within the past3 years. The other author has notstated any conflicts of interest.
