Skip to main content

Table 2 Study characteristics and main findings of included studies categorized by TIPPME intervention type

From: The effects of nudges on purchases, food choice, and energy intake or content of purchases in real-life food purchasing environments: a systematic review and evidence synthesis

Author Year Country Nudge description Study design Study size Intervention duration Setting SEP Study outcome(s) Outcome assessment Main finding(s) Quality assessment
Information nudges (symbols)
 Cawley et al. [19] 2015 USA Supermarket items were assigned with stars indicating their relative healthiness Pre-post 168 supermarkets >  1 year Supermarket N/A Purchases healthy items (any stars);
Purchases of unhealthy items (no stars)
Point of sale system 1a. Purchases of healthy items were not affected
1b. Purchases of unhealthy items decreaseda
Moderate
 Dubbert et al. [20] 1984 USA Labels indicating low-calorie choices were placed besides serving location Pre-post 6970 customers >  1 week & ≤ 1 month Cafeteria N/A Purchases of vegetables;
Purchases of salad;
Purchases of entrees;
Caloric content of meal purchased
Point of sale system and observer reported 2a. Increased vegetable purchasesa
2b. Increased salad purchasesa
2c. Entrée purchases not affected
2d. Caloric content of meals purchased not affected
Weak
 Elbel et al. [21] 2013 USA Unhealthy items were assigned a tag stating ‘less healthy’ Pre-post 3680 purchases >  1 week & ≤ 1 month Small food store Store catered to low-income, minority and immigrant population Purchases of healthy items;
Caloric content of purchases
Point of sale system 3a. Probability of purchasing healthy items increaseda
3b. Caloric content of items purchased decreaseda
Strong
 Eldridge et al. [22] 1997 USA Menu boards indicated healthy items with a green check-mark Pre-post 7 cafeterias 6 < months ≤12 Cafeteria N/A Purchases of all targeted items Point of sale system 4. Purchases of targeted items were not affected Moderate
 Freedman et al. [23] 2011 USA Healthy foods were identified with a promotional logo on shelf-tags Pre-post 1 small food store 1 < month(s) ≤ 6 Small food store N/A Purchases of all targeted items Point of sale system 5. Purchases of targeted items were not affected Moderate
 Hobin et al. [24] 2017 Canada Supermarket items were assigned with stars indicating their relative healthiness CT 44 intervention supermarkets; 82 control supermarkets 6 < months ≤12 Supermarket Supermarkets were located in area where 13.1% had no secondary school diploma Purchases of healthy items;
Caloric content of purchases
Point of sale system 6a. Average mean star rating per product purchased increased, so healthy purchases increaseda
6b. Caloric content of purchases was not affected
Strong
 Hoefkens et al. [25] 2011 Belgium Healthy meal suggestions were assigned with stars (0–3 stars) Pre-post 224 customers >  1 week & ≤ 1 month Cafeteria N/A Meal choice (0–3 stars) and energy intake Questionnaire 7a. Meal choice was not affected;
7b. Energy intake was not affected
Weak
 Johnson et al. [26] 1990 USA Labels indicating low-calorie choices were placed besides serving location Pre-post 413 customers >  1 week & ≤ 1 month Cafeteria N/A Caloric content of purchases Observer reported 8. Caloric content of purchases was not affected Weak
 Lassen et al. [27] 2014 Norway Healthy choices were labelled with the Keyhole symbol CT 270 customers 6 < months ≤12 Cafeteria 59% employed as office and administrative personnel or as technical staff Energy density of consumed foods Digital photographic method 9. Energy density decreaseda Strong
 Levin et al. [28] 1996 USA Low-fat entrees were labelled with a heart-shaped symbol CT 2 cafeterias 6 < months ≤12 Cafeteria N/A Purchases of targeted items Point of sale system 10. Purchases of targeted items increaseda Moderate
 Sproul et al. [29] 2003 USA Healthy entrees were labelled with a promotional logo, which additionally provided nutritional information Pre-post 1 cafeteria 1 < month(s) ≤ 6 Cafeteria N/A Purchases of targeted entrees Point of sale system 11. Purchases of targeted entrees were not affected Moderate
 Sutherland et al. [30] 2010 USA Supermarket items were assigned with stars indicating their relative healthiness Pre-post 168 supermarkets >  1 year Supermarket N/A Purchases of star-labelled items Point of sale system 12. Purchases of star-labelled items increaseda Moderate
 Vyth et al. [31] 2011 The Netherlands Healthy sandwiches, soups, and fresh fruit were identified with a promotional logo RCT 13 intervention cafeterias; 12 control cafeterias >  1 week & ≤ 1 month Cafeteria N/A Purchases of healthy sandwiches;
Purchases of healthy soups;
Purchases of fresh fruit
Point of sale system 13a. Purchases of healthy sandwiches were not affected
13b. Purchases of healthy soups were not affected
13c. Fruit purchases increaseda
Moderate
 Mazza et al. [32] 2017 USA Emoticons highlighted healthy items Pre-post 1 cafeteria 1 < month(s) ≤ 6 Cafeteria N/A Purchases of healthy beverages
Purchases of healthy chips
Point-of-sale system 14a. Purchases of healthy chips were not affected
14b. Purchases of healthy beverages were not affected
Moderate
 Steenhuis et al. [33] 2004 The Netherlands In the labelling program, low-fat products were identified with a promotional logo. RCT 17 cafeterias were randomly assigned to either of 4 conditions (including control and labelling program) 1 < month(s) ≤ 6 Worksite cafeteria 2% low educational level Purchases of low-fat items (milk, butter, cheese, meat, desserts). Point of sale system and questionnaire 15a. Purchases of low-fat desserts increaseda
15b. Purchases of milk were not affected
15c. Purchases of butter were not affected
15d. Purchases of cheese were not affected
15e. Purchases of meat were not affected
Weak
Information nudges (nutrition information)
 Cioffi et al. [34] 2015 USA Nutrition labels were added to a selection of pre-packaged meals and snacks Pre-post 20 small food stores 6 < months ≤12 Small food store N/A Purchases of low calorie foods;
Purchases of high calorie foods;
Caloric content of purchases
Point of sale system 1a. Purchases of low calorie foods increaseda
1b. Purchases of high calorie foods decreased
1c. Caloric content of items purchased decreaseda
Moderate
 Hammond et al. [35] 2015 Canada Calorie labels were added to all cafeteria menu boards and food stations Pre-post 159 customers ≤ 1 week University cafeteria N/A Caloric content of purchases;
Calories consumed
Questionnaire 2a. Caloric content of purchases decreaseda
2b. Calorie intake decreaseda
Weak
 Milich et al. [36] 1976 USA Foods were labelled with their caloric value Pre-post 450 customers ≤ 1 week Hospital cafeteria N/A Caloric content of purchases Observer reported 3. Caloric content of purchases decreased; (p = 0.06) Weak
 Vanderlee et al. [37] 2014 Canada Energy, sodium and fat content were displayed on digital menu boards, as well as a health logo for healthier items CT 497 customers at intervention site; 506 customers at control site 1 < month(s) ≤ 6 Hospital cafeteria 14% low educational level (high school or less)
15% low income (<$CAN 40000)
Calorie intake Questionnaire 4. Caloric intake decreaseda Weak
 Aron et al. [38] 1995 UK Foods were provided with nutrition labels CT 65 intervention customers; 35 control customers ≤ 1 week University cafeteria N/A Calorie intake Questionnaire 5. Caloric intake increaseda Weak
 Chu et al. [39] 2009 USA Simplified nutrition labels were posted at the point of selection for entrée dishes Pre-post 1 cafeteria >  1 week & ≤ 1 month University cafeteria N/A Caloric content of purchases Point of sale system 6. Caloric content of purchases was not affected Moderate
 Webb et al. [40] 2011 USA Calorie information was posted on menu boards or was provided only on posters placed away from the point of decision. CT 1 experimental cafeteria; 1 control cafeteria 1 < month(s) ≤ 6 Hospital cafeteria 13% low educational level (< eighth grade, some high school and high school graduate) Purchases of healthy side dishes;
Purchases of healthy snacks;
Purchases of healthy entrees
Point of sale system 7a. Purchases of sides dishes increaseda
7b. Purchases of snacks increaseda
7c. Purchases of entrees were not affected
Moderate
 Chen et al. [41] 2017 Taiwan Entrees and side dishes were labeled with traffic-light labels Pre-post 276 customers for first survey; 205 customers for second survey 6 < months ≤12 Worksite cafeteria N/A Choice for green-labelled food;
Attempt to avoid red-labelled food
Questionnaire 8a. Choice for green entrée increaseda;
8b. Attempt to avoid red-coloured items was not affected.
Moderate
 Sonnenberg et al. [42] 2013 USA Food and beverages were labelled red, yellow, or green on either the menu board, shelf, or directly on the packaging. Pre-post 389 customers 1 < month(s) ≤ 6 Hospital cafeteria N/A Purchases of green items
Purchases of red items
Point of sale system 9a. Healthy (green) item purchases were not affected
9b. Unhealthy (red) item purchases were not affected
Strong
 Whitt et al. [43] 2017 USA Items were labelled green (healthy), yellow (neutral) or red (unhealthy). Pre-post 1 small food store 1 < month(s) ≤ 6 Small food store N/A Purchases of green items
Purchases of red items
Point of sale system 10a. Purchases of healthy (green) items increaseda
10b. Purchases of unhealthy (red) items decreaseda
Moderate
Information nudges (signage)
 Allan et al. [44] 2015 UK Signs visually arranged snacks and drinks from least caloric to most caloric, with arrows indicating their location in store RCT > 20,000 purchases 1 < month(s) ≤ 6 Small food store N/A Purchases of high-calorie snacks
Purchases high-calorie drinks
Point of sale system 1a. Purchases of high calorie snacks decreaseda;
1b. Purchases of high calorie drinks were not affected.
Moderate
 Buscher et al. [45], study 1 2001 Canada Signs with promotional prompts were located at the cafeteria entrance and in front of the targeted foods Pre-post 2280 students potentially exposed to the intervention >  1 week & ≤ 1 month University cafeteria N/A Purchases of vegetable basket
Purchases of pretzels
Purchases of yoghurt
Purchases of fruit basket
Point of sale system and hand-counts 2a. Vegetable basket purchases were not affected
2b. Pretzel purchases increaseda
2c. Yoghurt purchases increaseda
2d. Fruit basket purchases were not affected
Moderate
 Buscher et al. [45], study 2 2001 Canada Signs with promotional prompts were located at the cafeteria entrance and in front of the targeted yoghurt Pre-post 2280 students potentially exposed to the intervention >  1 week & ≤ 1 month University cafeteria N/A Purchases of yoghurt Point of sale system 3. Yoghurt purchases increaseda Moderate
 Montuclard et al. [46] 2017 USA A water sign was taped to the cafeterias soda dispensers and coffee dispensers Pre-post 357 students pre-intervention survey; 301 students post-intervention survey 1 < month(s) ≤ 6 University cafeteria N/A Choice for water Questionnaire 4. Choice for water increaseda Moderate
 Ogawa et al. [47] 2011 Japan Health and nutrition information related to consumption of fruits and vegetables was displayed on posters near fruit/vegetable display and/or checkout counter CT 1 intervention supermarket; 1 control supermarket 1 < month(s) ≤ 6 Supermarket N/A Purchases of fruits
Purchases of vegetables
Point of sale system 5a. Vegetable purchases increaseda
5b. Fruit purchases were not affected.
Moderate
 Policastro et al. [48] 2017 USA Water consumption was promoted through signage promoting swapping soda for water Pre-post 2393 students covering 6730 transactions ≤ 1 week University cafeteria N/A Choice for water Point of sale system 6. Water purchases increaseda Moderate
 Scourboutakos et al. [49] 2017 Canada Posters promoted water and fruit and vegetable consumption Pre-post 368 to 510 students per data collection day 1 < month(s) ≤ 6 University cafeteria N/A Purchases of water
Purchases of fruits
Purchases of vegetables
Observer reported 7a. Purchases of water increaseda
7b. Purchases of fruit increaseda
7c. Purchases of vegetables increaseda
Weak
 Mazza et al. [32] 2018 USA A health message stating the % of daily calories contained in beverages, and required exercise to burn calories of chips Pre-post 1 cafeteria >  1 week & ≤ 1 month Hospital cafeteria N/A Purchases of healthy chips
Purchases of healthy beverages
Point of sale system 8a. Purchases of healthy chips were not affected
8b. Purchases of healthy beverages were not affected
Moderate
 Mazza et al. [32] 2018 USA A health message stating the % of daily calories contained in chips, and required exercise to burn calories of beverages Pre-post 1 cafeteria >  1 week & ≤ 1 month Hospital cafeteria N/A Purchases of healthy chips
Purchases of healthy beverages
Point of sale system 9a. Purchases of healthy chips increaseda
9b. Purchases of healthy beverages were not affected
Moderate
 Payne et al. [50], study 1 2015 USA Messages on grocery carts stated the number of fruits and vegetable items customers of that particular store normally purchased CT 396,017 individual person transactions >  1 week & ≤ 1 month Supermarket Supermarkets were located in area with 7% unemployment and
24% only high school education
Purchases of fruits and vegetables Point-of-sale system 10. Fruit and vegetable purchases increaseda Strong
 Payne et al. [50], study 2 2015 USA Messages on grocery carts stated the number of F&V items customers of that particular store normally purchased Pre-post 575,689 individual person
transactions
>  1 week & ≤ 1 month Supermarket Supermarkets were located in area with 7% unemployment and
24% only high school education
Purchases of fruits and vegetables Point-of-sale system 11. Fruit and vegetable purchases increaseda Strong
 Salmon et al. [51] 2015 The Netherlands The presence of a banner with was manipulated (absent/present, which stated that a particular low-fat cheese was the most sold brand of cheese in the supermarket. CT 127 customers N/A Supermarket 10% low educational level (primary school or lower levels of high school) Purchases of low-fat cheese Collection of receipts 12. Purchases of low-fat cheese were not affected Strong
Position nudges
 Kroese et al. [52] 2016 The Netherlands Unhealthy snacks at the check-out counter were replaced by healthy snacks CT 2 intervention stores; 1 control store ≤ 1 week Small food store N/A Purchases of healthy snacks;
Purchases of unhealthy snacks
Point-of-sale system 1a. Healthy snack purchases increaseda
1b. Unhealthy snack purchases were not affected
Moderate
 Meiselman et al. [53], study 1 1994 UK Candy was repositioned from four cash-points to one distant cash-point Pre-post 43 students ≤ 1 week University cafeteria N/A Choice for candy
Energy intake
Questionnaire and food weighing 2a. Candy selection decreaseda
2b. Energy intake was not affected
Weak
 Meiselman et al. [53], study 2 1994 UK Potato chips were repositioned from meal line to distant snack bar Pre-post 60 students >  1 week & ≤ 1 month University cafeteria N/A Choice for potato chips Questionnaire 3. Potato chips selection decreaseda Weak
 Meyers et al. [54] 1980 USA High calorie desserts were placed in the rear position on buffet line Pre-post 4412 food choices were observed ≤ 1 week Hospital cafeteria N/A Choice for high calorie desserts Observer-reported 4. High calorie dessert choice was not affected Moderate
 Rozin et al. [55], study 3 2011 USA Salad bar ingredients were placed on edge position of salad bar vs. middle position Pre-post 1 cafeteria 1 < month(s) ≤ 6 Hospital cafeteria N/A Purchases of salad bar ingredients Food weighing 5. Salad bar purchases increaseda Moderate
 Van Gestel et al. [56] 2017 The Netherlands Unhealthy snacks at the check-out counter were replaced by healthy snacks Pre-post 1 small food store >  1 week & ≤ 1 month Small food store N/A Purchases of healthy snacks Point of sale system 6. Healthy snack purchases increaseda Moderate
 Chapman et al. [57], study 1 2012 UK Confectionery was removed from check-out counters and replaced by fruit Pre-post 1 cafeteria ≤ 1 week University cafeteria N/A Purchases of fruit;
Purchases of confectionery
Point of sale system 7a. Healthy fruit purchases decreaseda
7b. Unhealthy confectionary purchases increaseda
Moderate
 De Wijk et al. [58] 2016 The Netherlands Wholegrain bread was placed near entrance vs. away from entrance CT 2 supermarkets 1 < month(s) ≤ 6 Supermarket N/A Purchases of wholegrain bread Point of sale system 8. Healthy bread purchases were not affected Moderate
 Thorndike et al. [59] 2017 USA Stores improved visibility of fruits and vegetables through new supplies (e.g., baskets, shelving) RCT 3 intervention stores; 3 control stores 1 < month(s) ≤ 6 Small food store Store was located in low-income urban community Purchases of fruits and vegetables Point of sale system (WIC voucher redemption) 9. Fruit and vegetable purchases increaseda Moderate
 Winkler et al. [60] 2016 Denmark Sugar confectionery at one checkout counter was replaced by healthy snacks CT 4 intervention stores; 2 control stores >  1 week & ≤ 1 month Supermarket N/A Purchases of healthy snacks;
Purchases of sugar confectionary
Point of sale system 10a. Healthy snack purchases were generally not affected
10b. Unhealthy purchases were not affected
Moderate
Mixed nudges
 Gittelsohn et al. [61] 2013 USA Environmental changes included demonstrations of healthier cooking methods, taste-tests, and display of point-of-purchase materials (e.g., posters and shelf labels) RCT 98 participants from intervention condition; 47 participants from control condition >  1 year Supermarket Years of schooling; 10.9y (intervention particpants) and 9.3y (control participants) Healthy food purchasing score
Unhealthy food purchasing score
Questionnaire 1a. Healthy food purchasing score was not affected
1b. Unhealthy food purchasing score was not affected
Weak
 Dorresteijn et al. [62] 2013 The Netherlands Environmental changes included signage promoting low-sodium soup and low-fat croissants. Also, margarine was made less accessible whereas butter was made more accessible. Pre-post 1 cafeteria >  1 week & ≤ 1 month Hospital cafeteria N/A Purchases of normal soup
Purchases of healthier soup
Purchases of normal croissants
Purchases of healthier croissants
Purchases of butter
Purchases of margarine
Point of sale system and hand-counts 2a. Normal soup purchases were not affected
2b. Healthier soup purchases were not affected
2c. Normal croissant purchases were not affected
2d. Healthier croissant purchases were not affected
2e. Butter purchases increaseda
2 f. Margarine purchases decreaseda
Weak
 Gamburzew et al. [63] 2016 France Environmental changes included shelf labels indicating healthy foods, signage explaining the labelling system, placement strategies, and a taste-testing booth. CT 6625 customers 1 < month(s) ≤ 6 Supermarket N/A Purchases of targeted foods (fruits and vegetables; starches; meat/fish/eggs; mixed dishes and sandwiches; dairy products). Point of sale system 3a. Purchases of fruits and vegetables increaseda
3b. Purchases of starches increaseda
3c. Purchases of meat/fish/eggs were not affected
3d. Purchases of mixed dishes/sandwiches were not affected
3e. Purchases of dairy were not affected
Weak
 Gittelsohn et al. [64] 2010 USA Environmental changes included posters, shelf labels, cooking demonstrations and taste tests. CT 64 intervention participants; 53 control participants 6 < months ≤12 Supermarket For intervention and comparison group, respectively:
Years of schooling, 12.5y and 12.4y;
Percentage unemployed, 35.9 and 18.8%.
Healthy food purchasing score
Unhealthy food purchasing score
Calorie intake
Questionnaires and dietary recall 4a. Healthy food purchasing score was not affected
4b. Unhealthy food purchasing score was not affected
4c. Calorie intake was not affected
Strong
 Foster et al. [65] 2014 USA Environmental changes included placement strategies (multiple facings, prime placement), signage, shelf-tags, cross-promotion of healthy foods, and taste testing. RCT 4 intervention supermarkets; 4 control supermarkets 6 < months ≤12 Supermarket Supermarkets located in low-income, high-minority neighborhoods Purchases of targeted foods (milk, cereal, frozen meals, in aisle-beverages, checkout cooler beverages, water) Point of sale system 5a. Purchases of some targeted products within the milk category increaseda
5b. Purchases of cereals were not affected
5c. Purchases of some targeted products within the frozen meal category increaseda
5d. Purchases of some targeted products within the in-aisle beverages category decreased to a lesser extent in the intervention stores as compared to controla
5e. Purchases of check-out cooler beverages were not affected
5 f. Purchases of water increaseda
Moderate
 Lawman et al. [66] 2015 USA Availability of healthy foods was increased and promoted through banners, shelf labels, and recipes. A subset of stores was provided additional business trainings and mini-grants for storing their inventory of healthy foods (high-intensity intervention). Pre-post 8671 customers at baseline; 5949 customers at follow-up 6 < months ≤12 Small food store N/A Mean energy purchased Observer reported 6. The intervention did not affect mean energy purchased. Moderate
 Levy et al. [67] 2012 USA Foods were labelled green, orange or red. Additionally, a choice architecture intervention was added which increased visibility and accessibility of green-labelled foods and beverages while decreasing the same for certain red-labelled items. Pre-post 4642 customers 1 < month(s) ≤ 6 Hospital cafeteria 28% employed as service workers, administrative support, technicians Purchases of green items
Purchases of red items
Point of sale system 7a. Green item purchases increased
7b. Red item purchases decreased
Strong
 Lowe et al. [68] 2010 USA Environmental changes included increased availability of foods lower in energy density. Additionally, a labelling system was introduced which color-coded food items. Pre-post 49 customers 1 < month(s) ≤ 6 Hospital cafeteria N/A Caloric content of purchases Point of sale system 8. Caloric content of purchases decreaseda Moderate
 Cardenas et al. [69] 2015 Peru Fruit was repositioned from a distant position to a more accessible location near the point of purchase. Additionally, signage highlighted health benefits of fruit consumption. Pre-post 150 customers >  1 week & ≤ 1 month University cafeteria N/A Fruit purchases Hand-counts 9. Fruit purchases were not affected Moderate
 Thorndike et al. [70] 2012 USA Foods were labelled red, yellow and green. Additionally, healthy foods were located to convenient positions and unhealthy options were made less convenient. Pre-post 1 cafeteria 1 < month(s) ≤ 6 Hospital cafeteria N/A Purchases of green items
Purchases of red items
Point of sale system 10a. Purchases of green items increased.
10b. Purchases of red items decreased
Moderate
 Steenhuis et al. [33] 2004 The Netherlands In the food supply program, the availability of low-fat items increased and was made salient with signage. RCT 17 cafeterias were randomly assigned to either of 4 conditions (including control and food supply program) 1 < month(s) ≤ 6 Worksite cafeteria 2% low educational level Purchases of low-fat items Point of sale system and questionnaire 11. Purchases of low-fat items were not affected Weak
 Thorndike et al. [71] 2014 USA Items were labelled green, yellow or red. Additionally, items were rearranged to make some of the green items more visible and some red items less visible. Pre-post 1 cafeteria >  1 year Hospital cafeteria 29% low educated jobs (service workers, administrative support, technicians) Purchases of green items
Purchases of red items
Point of sale system 12a. Purchases of green-labelled items increaseda
12b. Purchases of red-labelled items decreaseda
Weak
 Seward et al. [72] 2016 USA The full intervention included traffic-light labels, accessibility changes and tray stickers visualizing recommended portions of food types. The minimal intervention only included accessibility changes. CT 4 experimental cafeterias; 2 control cafeterias 1 < month(s) ≤ 6 University cafeteria N/A Purchases of green items
Purchases of red items
Observer reported 13a. Purchases of green items were not affected in neither full or minimal intervention cafeterias.
13b. Purchases of red items were not affected in neither full or minimal intervention cafeterias.
Moderate
 Lee-Kwan et al. [73] 2015 USA During phase 1, menus were revised to emphasize healthy foods with labels. Consecutively, during phase 2, (additional) healthy sides and beverages were introduced and promoted. CT 3 intervention stores; 4 control stores 1 < month(s) ≤ 6 Small food store Carry-outs were based in low-income neighbourhoods Purchases of healthy items Point of sale system 14. Purchases of healthy items were not affected. Moderate
Availability, size, functionality, and presentation nudges
 Diliberti et al. [74] 2004 USA During baseline conditions, the portion size of the entree was the standard 100% portion; in the experimental condition the size was increased to 150%. CT 180 customers >  1 week & ≤ 1 month Cafeteria N/A Energy intake Food weighing Energy intake from the pasta entree, accompaniments, and entire meal increaseda Moderate
 Vandenbroele et al. [75] 2018 Belgium Different sizes of sausages were available: a 150 g portion (default); a 125 g in-between portion; or a small, 100 g portion. CT 161 customers who bought targeted product >  1 week & ≤ 1 month Supermarket N/A Meat purchases Point-of-sale system The introduction of smaller portion size alternatives was associated with less meat being purchaseda Moderate
 Payne et al., study 1 [76] 2016 USA Large green arrows were placed on the floor directing attention to the store’s produce section. CT 1 intervention store; 1 control store >  1 week & ≤ 1 month Supermarket N/A Purchases of fruit & vegetables Point-of-sale system Green arrows on floors were associated with increased fruit and vegetable purchasesa Strong
 Payne et al., study 2 [76] 2016 USA Large green arrows were placed on the floor directing attention to the store’s produce section. CT 1 intervention store; 1 control store >  1 week & ≤ 1 month Supermarket N/A Purchases of fruit & vegetables Point-of-sale system Green arrows on floors were associated with increased fruit and vegetable purchasesa Strong
 Tal et al. [77] 2015 USA Samples (no, apple, or cookie) were offered to participants at the entrance of the store CT 120 customers N/A Supermarket N/A Purchases fruit & vegetables Observer-reported Receiving an apple sample was associated with increased subsequent purchases of fruits and vegetables vs. cookie or no sample (p = 0.06). Weak
  1. aAsterix indicates statistical significance (p < 0.05); Main findings are aggregated across two consecutive intervention phases for which data on statistical significance was not available