CCBYNC Open access
Analysis

Downsizing: policy options to reduce portion sizes to help tackle obesity

BMJ 2015; 351 doi: https://doi.org/10.1136/bmj.h5863 (Published 02 December 2015) Cite this as: BMJ 2015;351:h5863

Portion size matters
Click here to see an infographic, showing the impact of portion size, and how policies could be developed in response.

  1. Theresa M Marteau, professor of behaviour and health1,
  2. Gareth J Hollands, senior research associate 1,
  3. Ian Shemilt, senior research associate1,
  4. Susan A Jebb, professor of diet and population health2
  1. 1Behaviour and Health Research Unit, University of Cambridge, Cambridge, UK
  2. 2Nuffield Department of Primary Care Health Sciences, University of Oxford, UK
  1. Correspondence to: T M Marteau tm388{at}cam.ac.uk
  • Accepted 21 October 2015

Larger portions of food increase consumption. Theresa Marteau and colleagues suggest ways to reduce their size, availability, and appeal

The worldwide prevalence of obesity and overweight has risen substantially over the past three decades with no country yet achieving a reduction.1 International and national ambitions to “end childhood obesity”2 and “reduce non-communicable diseases by 25% by 2025”3 are unmatched by policies that could realise them. The causes of obesity are complex but overconsumption of food and sugary drinks is a critical proximal determinant, driven in part by large portion sizes. The importance of developing interventions and policies to reduce the size, availability, and appeal of large portions is underscored by the compelling evidence that people eat and drink more from larger portions.4

The problem

The size of portions, packages, and tableware has increased over the past 50 years (fig 1). Our recent Cochrane review shows that people consistently consume more food or non-alcoholic drinks when offered larger sized portions or packages, or when using larger items of tableware.4 The size of this effect suggests that eliminating larger portions from the diet could reduce average daily energy consumed by 12-16% among UK adults and by 22-29% among US adults. Our estimates are in line with those generated in another review on portion size using different methods.5

Figure1

Fig 1 US and UK posters illustrating changes in portion, package, and tableware sizes since the 1950s. Reproduced with permission from Centers for Disease Prevention and Control and World Cancer Research Fund

Crucially, portion size is a modifiable determinant of dietary energy intake. Although clearer guidance on healthy portion sizes for a range of foods and drinks is awaited,6 most national and international policies to prevent obesity highlight a need to reduce portion sizes.7 8 Indeed, a recent economic analysis ranked reduced portion size as having the highest potential to reduce the population health burden of obesity.9

The mechanisms underlying the “portion size effect” are not fully understood.10 11 However, it seems that the social and personal norms for what constitutes a suitable amount to consume are shaped by food portions we routinely encounter in supermarkets, restaurants, or the home, including images used in marketing. As exposure to larger portions has become more common, these sizes have come to be viewed as appropriate, with consumption correspondingly increasing. This suggests that reductions in portion size might, over time, recalibrate consumption norms, even if there were some initial resistance from consumers and industry. The effect may often operate without awareness,10 12 making default portion, package, and tableware size an effective psychophysical barrier to our ability to regulate the amounts of food and energy we consume.4 This explains why providing information about the effect13 or guidance, such as labelling, to indicate an appropriate portion size seems to have only a small effect.14 15

Policy options

Given that larger portions are now an established part of the highly competitive food market, change will require active intervention. The abundance of large portion sizes reflects a synergy of public demand with commercial interests; buyers want filling portions at competitive prices—particularly of highly palatable, usually energy dense foods—and industry benefits from cost savings when supplying and packaging larger portions combined with promotional strategies to increase producers’ market share. There has been little research on interventions to reduce portion size and existing studies have shown mixed effects.11 It seems likely, however, that effective interventions will reduce demand and supply, setting up a virtuous circle, to recalibrate portion sizes. Achieving such change will require support from the public, industry, and politicians. We consider what interventions might work, where and how they can be made, and who needs to act (fig 2).

Figure2

Fig 2 Environments in which changes are needed to reduce exposure to larger portion sizes, the means of achieving these changes, and the change enablers

What interventions might work?

The box shows interventions with the potential to reduce portion sizes. Interventions in physical environments include reducing the availability of larger sizes for energy dense foods and drinks and the tableware used for their consumption, designing packaging to demarcate single portion sizes, and limiting portion sizes used in advertisements.

Interventions to target portion size

Physical environment*
Food and drink
  • Sizing—Make default serving sizes smaller for energy dense foods and drinks—eg, reduce size of single serve confectionery and serving size of chips and cakes in canteens

  • Availability—Reduce availability of larger portion and package sizes†—eg, remove largest serving size of drinks; increase availability of smaller portion and package sizes—eg, offer option of smaller portions to diners in restaurants

  • Placement—Place larger portion sizes in stores and cafes less accessibly†—eg, portion size limits at checkouts, aisle ends, and special displays

  • Design—Demarcate single portion sizes in packaging through wrapping or visual cues†—eg, individual wrapping of biscuits

  • Marketing—Restrict portion and package sizes used in advertisements and other marketing

Tableware (plates, cups, glasses, and cutlery)
  • Sizing—Make smaller tableware the default for self service and served foods and drinks†

  • Availability—Increase availability of smaller tableware and reduce availability of larger tableware for home use

  • Design—Develop tableware that maximises the mechanisms underlying the portion size effect—eg, shallow plates, straight sided glasses, cutlery that holds smaller mouthfuls

Economic environment
  • Restrict pricing practices whereby larger portion and package sizes cost less in relative (and sometimes absolute) monetary terms than smaller sizes†

  • Restrict price promotions on larger portion and package sizes†

  • Price tableware in relation to size

  • *Subheadings taken from a typology for interventions in physical microenvironments16

  • †Actions most consistent with evidence from our systematic review4

Price is cited by shoppers as a major influence on product choice.17 This makes foods and drinks in larger portions and packages more appealing because they often cost less in relative (and sometimes absolute) terms. However, restrictions on such pricing practices may be limited by differences in the true cost of production and Competition Commission rules.

While the evidence for the effectiveness of changes in physical and economic environments in relation to portion size is limited, there is strong evidence that interventions that reduce availability and increase price reduce tobacco and alcohol consumption.18 19

Reducing portion sizes across the whole diet to realise large reductions in consumption may mean reverting to sizes of portions and tableware similar to those in the 1950s (fig 1). This would involve reductions of over 50% for some energy dense products, far greater than the typical 5% reductions currently offered and negotiated with the food industry.20

Key uncertainties

Most of the existing evidence on the effects of reducing portion size comes from studies of very large portions. We therefore cannot be certain that reducing smaller portions would be as effective in reducing food consumption.

In addition, we do not know by how much large portion sizes can be reduced before becoming unacceptably small. While the portion size effect seems to operate without awareness,10 12 overt actions to reduce portion sizes, particularly when prices do not also decrease, may prompt consumer resistance.21 This may not recede until social norms are recalibrated and pricing is adjusted to reflect a smaller size.22

Reducing exposure to larger portion sizes could also have unintended compensatory effects, encouraging consumption of multiple smaller portions or additional foods. At present there is no strong evidence for this. The few primary studies that have assessed this directly found no such effects—for example, consumption of a reduced size of breakfast did not affect the amount of food consumed over the rest of the day.23

Where can these interventions be made?

Public sector environments

Implementation of portion size interventions will be easier in public sector organisations, such as schools, hospitals, military bases, and prisons, than in commercial environments. Intervening in health related environments could have a particular potency, removing the “health halo” that comes from providing less healthy foods, including larger portions, in these environments.24 The public sector in England spends £1.2bn a year on food and drink and the food procurement plan for public bodies in England, which makes the provision of healthy, tasty, and sustainable food priorities alongside value for money, could include default procurement of smaller portion sizes.25

Commercial sector environments

Interventions in commercial environments pose major challenges. For example, the attempt to introduce a 16 ounce (454 mL) limit on the size of sugar sweetened beverages sold in food outlets in New York City met with much resistance and has been unsuccessful.26 There has been an effort to reduce portion size in England through voluntary agreements as part of the Public Health Responsibility Deal. For example, Mars, Nestlé, and Mondelez, the three largest chocolate manufacturers have committed to limiting the energy content of single serve confectionery to 250 kcal. However, broader change across companies and products is fragmented, and there has been no evaluation of the net reduction in energy intake as a consequence of portion size reductions. Some cinema chains in England have voluntarily removed their largest cup size of soft drinks so that the maximum is now 32 ounces, but this is still a large amount and the fact that not all cinemas have signed up illustrates the limits of a voluntary approach.

How can these interventions be achieved?

A combination of regulatory and non-regulatory measures is likely to be needed.27 Indeed, the food industry may find it difficult to act without regulation given “first mover disadvantage.”28 Including disincentives or sanctions for non-participation in voluntary agreements may also help.29

Effective interventions will also need to take into account industry innovations that may circumvent the intended effects of policy approaches. For example, the agreement of confectionery manufacturers to phase out king size chocolate bars in 2005 led to the introduction of bars containing multiple portions, ostensibly for sharing or consuming at different times.

A 2015 poll reported that 40% of those aged under 25 regularly eat a whole 150 g “sharing bag” of crisps, which can contain up to a third of an adult’s recommended daily intake of calories and salt and almost a quarter of the recommended intake of fat.30 Conversely, very small portions may also increase consumption if they encourage non-consumers to try the product or if small individual units are offered in bulk quantities that may increase frequency of consumption.

Who needs to act?

Although policy makers and the food industry have primary responsibility for action, public acceptability is likely to be an important facilitator. Public acceptance of government intervention to prevent obesity is mixed but stronger when it is focused on children.31 32 More specifically, little is known about the acceptability of reducing portion or package sizes. A newspaper survey of New York residents in 2012 reported 60% opposed the proposed 16 ounce cap on sugary drinks.33 This was during a media campaign, funded by soda manufacturers, highlighting the rights of citizens to purchase soda in sizes “without interference from bureaucrats.” By contrast, a more recent survey of UK and US participants found greater support for this intervention, with 59.9% and 53.5%, respectively, finding it acceptable (unpublished data). Some studies have suggested the acceptability of smaller portion sizes. For example, a study in a US university campus restaurant found that 14-33% of those invited to halve the size of a starchy side dish accepted regardless of whether they were offered a discount to do so.34

While public acceptance seems necessary for governments and the private sector to act, real progress may require more coordinated public demand.35 The introduction of many tobacco control measures reflects the mobilisation of public support, not yet evident for obesity control.36 Strategies that inform and enable communities could increase their demand for change to tackle obesity. Providing information about the effectiveness of interventions increases support for them.37 38 Communities can also be enabled to act through stronger non-governmental organisations (NGOs) as seen in Mexico where, through Bloomberg Philanthropies, NGOs purchased prominent advertising space to effectively counter industry opposition to soda taxes.

Conclusion

The compelling evidence that larger portion sizes of food and non-alcoholic drinks increase consumption is currently unmatched by similarly strong evidence on how to reduce this effect. This requires independent and rigorous evaluation of interventions that aim to reduce the size, availability, and appeal of larger portions. Successful interventions, if implemented at sufficient scale, have the potential to help prevent obesity as part of a wider obesity strategy.27 Aligning the will of the public, private industry, and political leadership is key to progress.

Key messages

  • People consistently consume more food or non-alcoholic drinks when offered larger sized portions or when they use larger tableware

  • Actions in public and private sectors to reduce the size, availability, and appeal of larger portion sizes might help prevent obesity

  • Some interventions will probably require regulation and legislation, facilitated by public demand for change

  • Independent and rigorous evaluation is essential to ensure actions are effective

Notes

Cite this as: BMJ 2015;351:h5863

Footnotes

  • Contributors and sources: This article was prompted by the completion of our Cochrane review4 and the desire to draw out the policy implications of this and other cognate studies. TMM and GJH have expertise in behaviour and behaviour change, GJH and IDS have expertise in evidence synthesis, SAJ has expertise in nutrition and public policy to change diet. TMM is the guarantor of the article.

  • Competing interests: We have read and understood BMJ policy on declaration of interests and declare SAJ is the independent chair of the Public Health Responsibility Deal Food Network in England. The Behaviour and Health Research Unit is part funded by the UK Department of Health policy research programme (PR-UN-0409-10109). The Department of Health had no role in the preparation of this manuscript. The views expressed are the authors and not necessarily those of the funders.

  • Provenance and peer review: Not commissioned; externally peer reviewed.

This is an Open Access article distributed in accordance with the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: http://creativecommons.org/licenses/by/4.0/ .

References

View Abstract