Table 3.

Summary of intervention studies comparing smaller multipacks versus larger package(s) containing same total serving size of energy-dense, nutrient-poor snack and drink.

First Author, Year of Publication, Country, Risk of Bias	Study Sample	Setting	Package Size Comparison Groups	Potential Moderators or Mediators	Outcome Measures (Measures Used)	Findings
Argo, 2012 Study 2 [48] Canada Low	207 undergraduate students (123 females)	Laboratory face-to-face	Candy-coated chocolates Two larger packs Eight smaller packs	Package design (transparent/opaque) Appearance self-esteem (ASE) 1 Gender	Consumption (direct weighing)	Participants consumed significantly more from the smaller multipacks than larger packs, which was fully contributed by those with low ASE. No effect was found among those with high ASE. When packaging was transparent (vs. opaque), participants consumed significantly more (42 g/100%) from the smaller multipacks than larger packs. Gender was not a moderator.
Bui, 2017 Study 3 [49] The USA Low	67 undergraduate students (35 females) Mean age 27 years	Laboratory face-to-face	Bite-sized chocolate chip cookies One larger pack (16 pieces per pack) Four smaller packs (4 pieces per pack)	Gender	Consumption (direct weighing)	No significant effect of package size on consumption was found. Gender was not a moderator.
Codling, 2020 [50] The UK Low	166 households Mean age 31 years	Free living	Wine 750 mL bottle 500 mL bottle	The order of receiving each package condition (crossover)	Consumption (recording empty bottles)	Participants (households) consumed significantly less wine in 14 days (173 mL/4%) and had a lower rate (6%) of consumption from the 500 mL bottles than 750 mL bottles. The order of receiving each package size condition was not a moderator.
Do Vale 2008 Study 2 [51] The Netherlands High	140 undergraduate students (59 females)	Laboratory face-to-face	Potato chips Two 200 g packs Nine 45 g packs	Self-regulatory concern 2	Consumption (direct weighing)	No significant effect of package size on consumption was found. The activation of self-regulatory concern led to lower intake from larger packs (but not from smaller multipacks).
Haire, 2014 [52] The USA High	64 university students (30 females) Mean age 23.7 years	Free living	Mini-pretzel Two 283 g packs Twenty-two 26 g packs	Weight (22.2 kg/m2 in normal weight group; 29.8 kg/m2 in overweight group) Dietary restraint status	Consumption (direct weighing)	Overweight or obese participants consumed significantly less (97 g/361 kcal (48%)) from the smaller multipacks than larger pack. No significant effect of package size was found among normal weight participants. Dietary restraint was not a moderator.
Holden, 2015 Study 1 [53] Australia Low	108 university students (58 females)	Laboratory face-to-face	M&M’s One 200 g pack Four 50 g packs	Manipulated diet consciousness 3 Measured diet consciousness 4	Consumption (direct weighing)	Participants consumed significantly more (10 g/67%) from the smaller multipacks than larger pack, which was contributed by those with activated diet consciousness. When diet consciousness concern was activated, participants consumed significantly more (29 g/161%) from the smaller multipacks than larger pack. No effect was found when diet consciousness was not activated. No significant effect of package size was found among those with higher diet consciousness.
Holden, 2015 Study 2 [53] Australia High	114 university students (64 females)	Laboratory face-to-face	M&M’s One 200 g pack Four 50 g packs	Diet consciousness 5 Diet prime (food focus) 6	Consumption (direct weighing)	Diet consciousness was activated in all participants, no significant effect of package size on consumption was found. Food-focused diet prime was a moderator. Participants’ tendency to overconsume from the smaller multipacks disappeared when food-focused diet prime was provided prior to eating.
John, 2017 Study 1 [38] The USA Low	362 drink purchasers (out of 623 participants) Mean age 24 years	Laboratory computer-based	Sugary drinks (iced tea or lemonade One 680 mL cup Two 340 mL cups One 454 mL cup (control)	None	Consumption (direct weighing) Likelihood of purchase (computer task)	No significant effect of package size on consumption was found. Participants in the two 340 mL cups condition had a significant higher likelihood of purchase compared to those in the one 680 mL cup condition.
Kerameas, 2015 Study 1 [54] Australia Low	87 female undergraduate students Mean age 20 years	Laboratory face-to-face	Cookies 30 g or 90 g total serving size: One 30 g/90 g cookie in one larger bag Three 10 g/30 g cookies in three smaller bags	Perceived norm of appropriate intake 7	Consumption (direct weighing)	Participants consumed significantly less from the multiple smaller packages (17 g/24%) than a larger pack. Participants in the 30 g total serving size conditions (additional cookies were available) consumed significantly less compared to those in the 90 g total serving size conditions. The perceived norm of appropriate intake was a mediator. Participants reported a lower perceived norm of appropriate intake when served the multiple smaller packages than a larger package.
Mantzari, 2017 [32] The UK Low	16 household representatives (12 females) Mean age 33 ± 6.6 years	Free living	Cola 1500 mL, 1000 mL, 500 mL, 250 mL bottles	None	Consumption (recording empty bottles)	No powered significance testing was undertaken as it was a feasibility study. The average weekly household consumption when provided with 250 mL, 500 mL, 1000 mL, and 1500 mL bottle size was 7878 ± 3861 mL, 8595 ± 3559 mL, 8331 ± 3963 mL, 8010 ± 3977 mL, respectively.
Mantzari, 2020 [55] The UK Low	16 households Mean age 40 ± 2.7 years	Free living	Wine 750 mL bottles 375 mL bottles	The order of receiving each package condition (crossover)	Consumption (recording empty bottles)	No powered significance testing was undertaken as it was a feasibility study. Household consumption in 2 weeks was 8.4 mL lower when receiving smaller bottles than when receiving larger bottles. The order of receiving each package condition could be a possible moderator. In four weeks, households receiving smaller bottles first overall consumed 1020 mL less wine than those receiving the larger bottles first.
Raynor, 2007 [56] The USA High	24 adults (12 female) Mean age 20 ± 1.6 years	Free living	A snack box with potato chips, crackers, mini cookies, M&M’s 142–227 g packs with smaller/larger total serving size 28–48 g packs with smaller/larger total serving size	Gender Weight	Consumption (recording empty packages)	The total serving size had a significant effect on consumption, regardless of package size. No significant effect of package size on consumption was found. Gender and weight were not moderators.
Roose, 2017 Study 2 [57] Belgium High	188 university students (88 females) Mean age 22 years	Laboratory face-to-face	Brownies One larger bag of 6 brownies Three smaller bags (2 brownies per bag)	Self-control conflict 8 Dietary restraint status 9	Consumption (direct weighing)	Participants consumed significantly more (13 g/30%) from smaller multipacks than from a larger pack, which was fully contributed by restrained eaters. The self-control conflict was a mediator. Participants experienced less self-control conflict when consuming from the smaller multipacks than a larger pack.
Scott, 2008 Study 2 [18] The USA High	343 university students	Laboratory face-to-face	M&M’s One 200 kcal pack regular-sized M&M’s Four 50 kcal pack of mini M&M’s	Dietary restraint status	Consumption (direct weighing) Perception of energy content (questionnaire)	Participants consumed significantly less from smaller multipacks with mini M&M’s than a larger pack with regular-sized M&M’s. Unrestrained eaters consumed significantly less (48 kcal/38%) from smaller multipacks with mini M&M’s than a larger pack with regular-sized M&M’s. Restrained eaters tended to consume more (12 kcal/12%) from smaller multipacks than a larger pack (not statistically significant). Participants perceived the energy content of smaller multipacks to be significantly greater than that of a larger pack; they also perceived mini M&M’s in smaller multipacks to be more similar to diet foods than regular-sized M&M’s in a larger pack.
Scott, 2008 Study 3 [18] The USA High	96 undergraduate students	Laboratory face-to-face	Cookies One 240 kcal pack regular-sized cookies (4 pieces per pack) Four 60 kcal pack mini cookies (2 pieces per pack)	Dietary restraint status	Consumption (direct weighing) Perception of predicted consumption (questionnaire)	No significant package size effect (mini cookies in smaller multipacks vs. regular-sized cookies in larger pack) was found. Participants predicted that they would consume less from smaller multipacks than a larger pack. Dietary restraint was not a moderator.
Scott, 2008 Study 4 [18] The USA High	393 undergraduate students	Laboratory face-to-face	M&M’s One 200 kcal pack regular-sized M&M’s Four 50 kcal packs mini M&M’s	Dietary restraint status Diet prime (food-focus) 10	Consumption (direct weighing) Perception (perceived caloric content) (questionnaire)	Participants consumed significantly less from smaller multipacks with mini M&M’s than a larger pack with regular-sized M&M’s. Participants perceived smaller multipacks with mini M&M’s to be significantly more similar to diet food and had higher energy content than a larger pack with regular M&M’s. Food focus was a moderator for restrained eaters but not for unrestrained eaters. Restrained eaters consumed less from the smaller multipacks than larger pack when regarding the provided snacks as ‘non-food objects’, whereas they consumed more from the smaller multipacks than larger pack when there was no food focus (control).
Stroebele, 2009 [58] The USA High	59 (41 females) Mean age 37.3 ± 12.0 years	Free living	Crackers, chips, biscuits, cookies Four packs (187–360 g per pack) Accordingly, number of smaller packs (19–26 g per pack) to keep the total serving size consistent	The order of receiving each package size condition (crossover)	Consumption (self-recorded snack diary)	On a weekly basis, participants consumed significantly less (187 g/32%) from smaller multipacks than larger packs. Participants who received smaller multipacks first consumed significantly less snacks (28%) from larger packs later, compared to those who received larger packs first and smaller multipacks later.
Van Kleef, 2014 Study 3 [59] The Netherlands High	165 university students (104 females) Mean age 21 ± 2.4 years	Laboratory face-to-face	Mars chocolate bars, package present or absent Three 51 g bars Fifteen 10 g bars	Perception of impulsiveness 11 Weight	Consumption (direct weighing) Perception (satiety) (questionnaire)	Participants consumed significantly less (51 kcal/23%) from smaller multipacks than larger packs. The perception of impulsiveness was a mediator. Participants counteracted the feelings of impulsiveness by eating less from smaller multipacks. Larger packs were perceived to be more satiating than smaller multipacks. Weight was not a moderator.
Wansink, 2011 [60] The USA High	37 university students (15 females) Mean age 20.3 ± 1.1 years	Laboratory face-to-face	Crackers One 400 kcal pack Four 100 kcal packs	Weight (mean 23.8 ± 3.9 kg/m2)	Consumption (direct weighing)	Participants consumed significantly less (75 kcal/25%) from smaller multipacks than a larger pack, which was fully contributed by overweight participants. No effect was found in normal weight participants. No significant effect of package size on feeling of fullness between package size conditions was found after consumption.
Bui, 2017 Study 1 [49] The USA Low	77 postgraduate students (44 females) Mean age 31 years	Laboratory computer-based	Bite-sized chocolate chip cookies One larger pack (16 pieces per bag) Four smaller packs (4 pieces per bag)	None	Intention to consume (computer task)	No significant effect of package size on intended consumption was found for cookies (which were perceived as an ‘unhealthy food’).
Bui, 2017 Study 2 [49] The USA High	171 (103 females) Mean age 38 years	Laboratory computer-based	Bite-sized chocolate chip cookies Two larger packs (8 pieces per bag) Four smaller packs (4 pieces per bag)	None	Intention to consume (computer task)	No significant effect of package size on intended consumption was found for cookies (which were perceived as an ‘unhealthy food’).
Scott, 2008 Study 3 follow-up [18] The USA High	201 undergraduate students	Laboratory face-to-face	M&M’s One 200 kcal pack regular-sized M&M’s Four 50 kcal packs mini M&M’s	Dietary restraint status	Intention to consume (questionnaire)	Participants intended to eat significantly less (23%) from the smaller multipacks with mini M&M’s than from larger pack with regular-sized M&M’s. Dietary restraint status was not a moderator. However, restrained eaters perceived that considering the consumption of mini M&M’s from smaller multipacks to be significantly more stressful than eating regular-sized M&M’s from a larger pack. This effect was not observed in unrestrained eaters.
Mantzari, 2018 [31] The UK Low	16 household representatives (12 females) Mean age 33 ± 6.6 years	Free living	Cola 1500 mL, 1000 mL, 500 mL, 250 mL bottles	None	Perception of previous consumption (rate and amount) (interview)	Participants believed that their consumption rate and amount was higher with the smallest (250 mL) bottle size due to the perception of more convenient, reduced awareness of the amount consumed, harder for consumption monitoring, and insufficient quantity in each bottle.
Scott, 2008 Study 1 [18] The USA High	385 undergraduate students	Laboratory face-to-face	M&M’s One 200 kcal pack regular-sized M&M’s One 200 kcal pack mini M&M’s Four 50 kcal packs regular-sized M&M’s Four 50 kcal packs mini M&M’s	None	Perception (diet food characteristics and energy content) 12 (questionnaire)	Participants perceived that mini M&M’s in smaller multipacks contain significantly more energy (144 kcal/75%) than regular-sized M&M’s in a larger pack. Participants perceived that mini M&M’s in smaller multipacks to be significantly more similar to ‘diet food’ than regular-sized M&M’s in larger packs.
Van Kleef, 2014 Study 2 [59] The Netherlands High	124 university students (75 female)	Laboratory face-to-face	Mars chocolate bars One 51 g pack Five 10 g packs	None	Perception (perceived energy intake) (questionnaire)	Participants overestimated their energy intake more significantly when eating from smaller multipacks (43% more) than a larger pack (4% more). Participants perceived that finishing the provided chocolates in smaller multipacks as significantly less appropriate, more excessive and more impulsive, and resulted in significantly lower expected satiation and satiety than finishing those provided in a larger pack.

¹ Appearance self-esteem (ASE): the self-worth a person derives from his or her body-image and weight. ² Self-regulatory concern activated group: participants were instructed to complete a body image satisfaction scale and dieting scale and report their weight before the study; self-regulatory concern-inactivated group (control group): participants participated in an unrelated study before the study. ³ Diet consciousness-activated group: participants were instructed to complete a body image questionnaire, self-reported height, and weight before the study to manipulate diet consciousness; Diet consciousness-inactivated group (control group): the same questionnaire was given to participants but after food exposure. ⁴ Measured diet-consciousness: participants were categorised into high and low diet consciousness groups by a dietary restraint scale questionnaire. ⁵ Diet consciousness was activated in all participants; participants were instructed to complete a body image questionnaire and report height and weight before the study. ⁶ Food focus: participants were instructed to evaluate the M&Ms while eating. ⁷ Perceived norm of appropriate intake: the perception of appropriate serving size (the appropriate amount of food to consume per eating occasion). ⁸ Self-control conflict: the offer of tempting food to a consumer who is occupied with restraining food intake (i.e., commitment to a health goal) sparks a self-control threat that evokes feelings of conflict. This conflict experience operates as an alarm that signals the need to restrain food intake. Failing to evoke this conflict leads to a failure to exert self-control, which then contributes to overconsumption. ⁹ Dietary restraint status: linked with individual’s perceived ability to estimate energy in this study. Restrained eaters perceived that they have strong ability to determine energy estimation. Unrestrained eaters perceived that they lack ability to determine energy estimation. ¹⁰ Diet prime conditions: (1) food-focus: participants were instructed to ‘think about the sensory experience of enjoying M&Ms’ such as the texture and taste; (2) non-food focus: participants were instructed to ‘think about the M&Ms as ‘non-food objects’; (3) control condition (no food focus): participants were instructed to ‘think about anything you would like to think’. ¹¹ Perception of impulsiveness: participants were instructed to consider the amount of chocolate they consumed and answer five questions on self-perceived impulsiveness (for example, ‘… am self-indulgent’, ‘… cannot resist the temptation of chocolate’). ¹² Diet food characteristics: measured by 7-point scale on the extent to which they disagree/agree with the statement ‘Overall, the M&Ms in their packages seemed similar to diet foods’.