Description of the condition
Non-communicable diseases, principally cardiovascular diseases, diabetes, certain forms of cancer and chronic respiratory diseases, accounted for an estimated 36 million deaths worldwide in 2008, 63% of all global deaths (World Health Organization 2012). Major risk factors for these conditions are in part determined by patterns of health behaviour that are in principle modiﬁable, including people’s selection and consumption of food, alcohol and tobacco products (United Nations 2011). Identifying interventions that are effective in achieving sustained health behaviour change has therefore become one of the most important public health challenges of the 21
Description of the intervention
It is increasingly recognised that the physical environments that surround us can exert considerable influences on health behaviour and that altering these environments may therefore provide a catalyst for behaviour change (Das 2012). In a recent scoping review, we described a range of interventions for changing health behaviour that involve altering the properties or placement of objects or stimuli within small-scale or micro-environments (such as shops, restaurants, bars or homes) to change the way in which available choices are presented to people. These have been termed ‘choice architecture’ interventions (Hollands 2013a; Hollands 2013b).
The size of a portion or package is a modifiable property of food, alcohol and tobacco products that may influence their selection and consumption. In the case of food and alcohol products, the size of an item of tableware used to consume such products may similarly influence their selection and consumption. Examples include the portion size of alcoholic drinks served in bars or of foods served in restaurants or in the home, such as portions of a pasta dish served to restaurant customers (Diliberti 2004), the size or shape of plates or glasses that these products are served on and in (Shah 2011), and the number of cigarettes in packets sold in shops or the length of each cigarette (Russell 1980). In this context, the intervention involves manipulation of the size or physical dimensions of a food, alcohol or tobacco product, its packaging or the tableware used in its consumption, and comparisons of interest are between products, packages or items of tableware that differ in terms of these properties.
How the intervention might work
There are considerable influences on behaviour that are beyond individuals’ deliberative control. Indeed, it has been suggested that most human behaviour occurs outside of awareness, cued by stimuli in environments and resulting in actions that may be largely unaccompanied by conscious reflection (Marteau 2012; Neal 2006). This proposition has led to increasing policy and research attention being placed on interventions with mechanisms of action that are less dependent on the conscious engagement of the recipients, including interventions that involve altering properties of objects or stimuli within the small-scale environments that surround and cue behaviour, such as the sizes of products that we engage with (Hollands 2013a).
A number of mechanisms of action have been proposed to explain how the size of products may affect their consumption (Steenhuis 2009). It has been suggested that as the amount of a product made available for consumption is increased, individuals will continue to perceive each increasing amount as an appropriate quantity to consume. This phenomenon may be explained by several mediating factors including personal and social norms about what constitutes a suitable amount of a product to consume, for example, larger portions of food have become increasingly prevalent making it increasingly unlikely that smaller portions are viewed as normal or appropriate for a single serving (Young 2002), and a variety of cognitive biases. Such biases include ‘unit bias’, described as the tendency for individuals to engage most comfortably with a product as a single entity independent of its size, meaning that they are predisposed to consume the entirety of a product even as it changes size (Geier 2006). In addition, the presentation of food and alcohol products often entails the use of tableware, such as plates, glasses and cutlery. Not only does the size of tableware have the potential to directly influence the amount of a product available for consumption (Pratt 2012), its physical dimensions can elicit various perceptual biases (Wansink 2005) that may impact on perceptions of quantity and in turn determine levels of consumption. All of these mechanisms may also influence product selection (with or without purchasing), which is recognised as an important intermediate outcome in pathways to consumption.
To our knowledge, extant research has focused on effects of the size of food and drink products to a much greater extent than for tobacco products (Hollands 2013a). Whilst the underlying mechanisms of any observed effects may be assumed to be broadly similar, smokers are known to titrate their received dose of nicotine to regulate the level in the body, which might attenuate any effects of interventions that alter the size of tobacco products (Kozlowski 1986).
Why it is important to do this review
A recent scoping review of evidence for the effects of choice architecture interventions identified a substantial number of randomised controlled trials that have investigated the effects of different portion, package or tableware sizes on selection and consumption behaviours (Hollands 2013a). The majority of these studies focused on food products, but because both tobacco and alcohol use also involve the selection and consumption of products, similar interventions may have the potential to change these behaviours via similar mechanisms. To our knowledge, evidence from these studies has yet to be synthesised using rigorous systematic review methods that include assessment of risk of bias and investigation of potential effect modifiers, nor to encompass effects on important health behaviours other than diet-related behaviours, namely alcohol and tobacco use. As such, we do not yet have reliable estimates of the effects of altering the sizes of portions, packages or tableware on product selection and consumption, nor of the influence of factors that may modify any effects. Both are necessary to inform the selection and design of effective public health interventions.
Interventions that involve changes to the physical environment, such as product sizing, as opposed to those that involve providing information to encourage health behaviour change, may also have the potential to reduce health inequalities if they rely less on recipients’ levels of literacy, numeracy and cognitive control, which have been found to be lower in population subgroups experiencing higher levels of social and material deprivation (Kutner 2006; Marteau 2012; Spears 2010; Williams 2003). Despite evidence that behaviours with the potential to undermine health are socially patterned (for example lower socioeconomic groups consume less fruit and vegetables (Giskes 2010)), potential differences in behavioural responses to product sizing interventions between socioeconomic subgroups remain unclear. Also, to our knowledge no studies of the effects of product sizing interventions have been conducted in low or middle income country populations (Hollands 2013a). This review will therefore include a focus on identifying evidence for differential effects of product sizing interventions between socioeconomic subgroups and national settings, highlight current gaps in this aspect of the evidence base, and draw implications for the potential of such interventions to impact on health inequalities.
This systematic review is timely given current interest in the topic within public health policy circles. There is evidence from the USA and Europe that portion sizes have been increasing since the 1970s (Young 2002; Young 2012). There have also been recent attempts to regulate the size of products in order to reduce their consumption and improve public health, such as New York City Mayor Michael Bloomberg’s proposed ban on the sale of sugary drinks larger than 16 oz (473 ml) (Gabbatt 2013). In the UK, there are recent examples of companies reducing the portion sizes of confectionery and sugary drinks as part of their voluntary pledges under the UK Government’s Public Health Responsibility Deal. This systematic review can contribute to a better evidence-based understanding of the likely impact of such policies.
- To assess the effects of interventions that involve manipulation of and comparison of different sizes or sets of physical dimensions of a portion, package, individual unit or item of tableware on unregulated (ad libitum) selection or consumption of food, alcohol or tobacco products in adults and children.
- To assess the extent to which the effects of such interventions may be modified by:
a) study-level characteristics, such as target product type (food, alcohol, tobacco) and whether the target of the manipulation is a portion, package, individual unit or item of tableware;
b) intervention characteristics, such as magnitude of the difference in size; and
c) participant characteristics, such as age, gender and socioeconomic status.
Criteria for considering studies for this review
Types of studies
Randomised controlled trials (experiments) with between-subjects (parallel group) or within-subjects (cross-over) designs, conducted in either laboratory or field settings. Non-randomised studies will be excluded because our recent scoping review indicated that a sufficient number of eligible randomised controlled trials will be available to address our aim to synthesise reliable estimates of intervention effects (Hollands 2013a). A key issue is that, compared with randomised controlled trials, non-randomised studies rely on more stringent often non-verifiable assumptions to be validated to confer confidence that the risk of systematic differences between comparison groups beyond the intervention of interest (that is confounding) is sufficiently low to allow valid inferences to be made about causal effects.
Types of participants
Adults and children who directly engage with the manipulated products. There will be no exclusion criteria relating to socio-demographic or clinical characteristics or prognostic factors. Studies involving non-human participants (that is animal studies) will be excluded.
Types of interventions
Interventions that are eligible to be considered in this review are those that involve the comparison of at least two sizes or sets of visible physical dimensions (that is volume, shape, height, width, depth) of a portion of a food (including non-alcoholic beverages), alcohol or tobacco product, its package or individual unit size, or an item of tableware used to consume it. ‘Portion’ refers to the overall amount (volume or weight, or both) of a product that is presented for selection or consumption (for example 200 g versus 300 g of pasta, 275 ml versus 440 ml of beer, or a half-length cigarette versus a full-length cigarette). ‘Package’ refers to the packaging of the product, including that used for service, consumption or storage (for example boxes, bags, cans and bottles). ‘Individual unit size’ refers to the size or physical dimensions of a unit of a product that is presented within a given portion (for example the size of individual sweets or candies or biscuits or cookies). ‘Tableware’ refers to crockery, cutlery, or glassware used for serving or consuming food or drink (for example plates, bowls, knives, forks, spoons or glasses). Packages and tableware as defined in this way have the capacity to limit or increase the portion or individual unit size of the consumed product and therefore influence the corollary effect of size on selection or consumption behaviours, or both. Eligible comparisons are conditions which provide alternative (that is smaller or larger) package, portion, individual unit or tableware sizes of the same (that is no difference in content of the substance consumed other than quantity) food, alcohol or tobacco product.
We will exclude the following.
- Interventions in which the size may be altered indirectly as a result of a higher-level intervention but is not directly and systematically altered (e.g. organisational-level interventions to encourage the introduction of small-scale environmental changes to alter product selection or consumption). While changes in portion sizing may be introduced as a result of the higher-level intervention, this is not directly manipulated to safeguard implementation fidelity.
- Interventions in which the behavioural responses of participants (i.e. selection or consumption levels or rates) are regulated by either explicit instructions to participants or by some other action of the researcher (e.g. participants exposed to a product are given instructions on how much they should consume or a target rate of consumption). In such cases, selection or consumption of the manipulated product cannot be considered unregulated (ad libitum).
- Studies of interventions that only involve comparisons of packages, portions, individual units or tableware of different types or functions. For example, we will exclude comparisons between different eating utensils (e.g. straw versus spoon; chopsticks versus fork) whilst comparisons between differently-sized eating utensils will be eligible for inclusion (e.g. small spoon versus large spoon).
- Studies of interventions in which there are concurrent interventions unrelated to sizing that are intrinsically confounded with the comparison(s) of interest. For example, a two arm study in which one group receives a specified portion size and the other group receives a smaller portion plus a concurrent nutritional labelling intervention.
Types of outcome measures
Eligible studies will include a measure of unregulated (ad libitum) selection (with or without purchasing) or consumption of the manipulated food, alcohol or tobacco product. In the specific case of a manipulation of one or more components of a plate of food or meal, studies will include a measure of the selection or consumption of either the plate of food or meal of which the manipulated component is a part, or one or more non-manipulated components of that plate of food or meal. By unregulated, we refer to behaviour of participants that is not regulated by either explicit instructions or by some other action of the researcher.
Our choice of eligible outcome constructs reflects a focus on the assessment of the effects of eligible interventions in terms of the types and amounts of food, alcohol and tobacco people consume, coupled with recognition that selection (with or without purchasing) is an important intermediate outcome in pathways to consumption. We anticipate there will be a range of measures of consumption and selection outcome constructs within the included studies and we present some examples of likely measures below.
1. Consumption (intake) of a product
We will assess:
- quantity of energy, nutrients or substances consumed;
- quantity of product(s) consumed (e.g. count, proportion, weight, volume).
Measurement may involve an objective measure in which the amount of a product consumed is calculated by subtracting the amount remaining after consumption from the amount presented to the participant. Alternatively, it may involve direct observation of the individual by outcome assessors or a participant-reported measure.
2. Selection of a product
a) Without purchase
b) With purchase
We will assess:
- quantity of energy, nutrients or substances selected;
- quantity of product(s) selected (e.g. count, proportion, weight, volume).
Depending on the study setting, a product may be selected with or without this act enjoining a purchase (that is a transfer of money to the vendor of the product). Measurement of selection may involve direct observation of the individual by outcome assessors, a participant-reported measure, or administrative data supplied by a retailer.
Examples of quantity of energy, nutrients or substances consumed or selected
Food: fat, sugars, salt, energy (calories or kilojoules)
Alcohol: units of alcohol
Tobacco: tar or carbon monoxide
Examples of measures of quantity of products consumed or selected
Food: weight or number of sandwiches or biscuits, or volume of soup
Alcohol: volume or bottles of wine or beer
Tobacco: number of cigarettes or weight of remaining cigarette butts
To supplement study eligibility criteria, we have developed a provisional conceptual model (Figure 1). The conceptual model is design-oriented in the sense that it is intended to help direct the review process (Anderson 2011) by providing a simplified visual representation of the causal system of interest, the proposed causal pathway between eligible interventions and their outcomes (behavioural endpoints), and potential moderators of that relationship (effect modifiers) given that differential effects are plausible (Anderson 2013). The initial draft conceptual model will be used to inform the development of search strategies, data extraction forms and a provisional framework for the statistical analysis of the data collected from the eligible studies (see Search methods for identification of studies and Data collection and analysis). We propose to revise the conceptual model iteratively as we encounter evidence from eligible studies during the course of the review process, and will document all revisions including the rationale for each revision and supporting evidence. As such, initial and subsequent iterations of the conceptual model will be used as a reference point for the design (that is protocol) and conduct (that is post-protocol) of all stages of the systematic review up to and including evidence synthesis, and as a conceptual basis for explicit reporting of the methods and assumptions employed within the synthesis (Anderson 2013). We anticipate that, in practice, iterative refinement of the conceptual model will primarily involve incorporating unanticipated potential effect modifiers that we encounter when collecting the data from the eligible studies, with a view to including these in the proposed meta-regression analysis.
|Figure 1. Design-oriented conceptual model|
Within the provisional conceptual model (Figure 1) we distinguish between three sets of potential effect modifiers: study-level characteristics; intervention characteristics; and participant characteristics. Within our proposed analytic framework for quantitative synthesis of data collected from the included studies (see Data collection and analysis), potential effect-modifying impacts of study-level characteristics can only be investigated based on between-study comparisons, whereas potential effect-modifying impacts of intervention characteristics can be investigated based on within-study comparisons between participant groups (for example between different arms of a randomised controlled trial). Potential effect-modifying impacts of participant characteristics may be investigated based on either between-study comparisons or within-study comparisons, depending on the level of reporting of results by participant subgroups within the included studies.
Search methods for identification of studies
We will conduct electronic searches for eligible studies within each of the following databases:
- Cochrane Central Register of Controlled Trials (CENTRAL) in The Cochrane Library (1992 to present);
- Cochrane Public Health Group Specialised Register;
- MEDLINE (OvidSP) (including MEDLINE In-Process) (1946 to present);
- EMBASE (OvidSP) (1980 to present);
- PsycINFO (1806 to present);
- Applied Social Sciences Index and Abstracts (ProQuest) (1987 to present);
- Food Science and Technology Abstracts (Web of Knowledge) (1969 to present);
- Science Citation Index Expanded (Web of Knowledge) (1900 to present);
- Social Sciences Citation Index (Web of Knowledge) (1956 to present);
- Trials Register of Promoting Health Interventions (EPPI Centre) (2004 to present).
We have developed a MEDLINE search strategy by combining sets of controlled vocabulary and free-text search terms based on the eligibility criteria described above (see Criteria for considering studies for this review). This was externally peer-reviewed by an information retrieval specialist and Co-convenor of the Cochrane Information Retrieval Methods Group and revised based on their peer-review comments. We tested the MEDLINE search strategy for its sensitivity to retrieve a reference set of 48 records of reports of potentially eligible studies known to be indexed in MEDLINE that were identified within our preceding, broader scoping review of interventions within physical micro-environments (Hollands 2013a). The final MEDLINE search strategy is presented in Appendix 1.
We will adapt the final MEDLINE search strategy for use to search each of the other databases listed above based on close examination of the database thesauri and scope notes. There will be no restrictions for publication date, publication format or language. No study design filters will be incorporated. Full details of the final search strategies will be provided in an appendix to the published review.
Searching other resources
We will conduct electronic searches of the following grey literature resources using search strategies adapted from the final MEDLINE search strategy as described above:
- Conference Proceedings Citation Index - Science (Web of Knowledge) (1990 to present);
- Conference Proceedings Citation Index - Social Science & Humanities (Web of Knowledge) (1990 to present);
- OpenGrey (1997 to present).
We will search trial registers (ClinicalTrials.gov and the World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP)) to identify registered trials, and the websites of key organisations in the area of health and nutrition including the following:
- UK Department of Health;
- Rudd Centre for Food Policy and Obesity, USA;
- Centres of Disease Control and Prevention, USA;
- International Obesity Task Force;
- EU Platform for Action on Diet, Physical Activity and Health.
In addition, we will search the reference lists of all eligible study reports and undertake forward citation tracking (using Google Scholar and PubMed) to identify further eligible studies or study reports. If non-English language articles are found, we will use Google Translate in the first instance to determine potential eligibility. If an article cannot be excluded on this basis, we will have the article translated by a native language speaker or professional translation service.
Data collection and analysis
Selection of studies
Title and abstract records retrieved by the electronic searches will be imported to EPPI Reviewer 4 (ER4) systematic review software (Thomas 2010). Duplicate records will be identified, reviewed manually and removed using ER4’s automatic de-duplication feature with the similarity threshold set to 0.85. Duplicate screening of title and abstract records retrieved by the electronic searches will be undertaken by two researchers working independently (GJH, IS). Title and abstract records will be coded as ‘provisionally eligible’, ‘excluded’ or ‘duplicate’ by applying the eligibility criteria described above (see Criteria for considering studies for this review). Any disagreements in the coding of title and abstract records will be identified and resolved by discussion to reach consensus between the two researchers (GJH, IS).
Full-text copies of corresponding study reports will be obtained for all records coded as ‘provisionally eligible’ at the title and abstract screening stage. Duplicate screening of full-text study reports will be undertaken by two researchers working independently (GJH, IS). Full-text study reports will be coded as ‘eligible’ or ‘excluded’ by applying the eligibility criteria described above (see Criteria for considering studies for this review), with the reasons for exclusion recorded. Any disagreements in the coding of full-text study reports or reasons for exclusion will be identified and resolved by discussion to reach consensus between the two researchers (GJH, IS). A third researcher (DO) will act as arbiter in the event that any coding disagreements cannot be resolved between the two researchers (GJH, IS). Bibliographic details of study reports excluded at the full-text screening stage will be provided, along with the primary reason for exclusion, in a ’Characteristics of excluded studies’ table within the published review. Multiple full-text reports of the same study will be identified, linked and treated as a single study. Full-text reports comprising multiple eligible studies will be identified and each study will be treated separately. We will document the flow of records and studies through the systematic review process and report this using a PRISMA flow diagram (Moher 2009).
Data extraction and management
An electronic data extraction form will be developed based on the Cochrane Public Health template and modified to allow extraction of all data required for this review. An initial draft of this form will be piloted using a selection of included studies, to ensure that it enables reliable and accurate extraction of appropriate data, and amended in consultation with the review team. Data pertaining to the characteristics of included studies will be extracted by one researcher. Outcome data will be extracted in duplicate by two researchers working independently (GJH, IS). If a study with more than two intervention arms is included, only outcome data pertaining to the intervention and comparison groups that meet the eligibility criteria described above will be included in the review, but the table 'Characteristics of included studies’ will include details of all intervention and comparison groups present in the study. Any discrepancies in extracted outcome data will be identified and resolved by checking against the study report, discussion and consensus between two researchers (GJH, IS), with a third researcher (DO) acting as arbiter in case of any unresolved discrepancies. Key unpublished data that are missing from reports of included studies will be sought by contacting the study authors.
At the outset we intend to collect the data summarised below. This represents a maximum core dataset that we can reasonably anticipate will be required based on our study eligibility criteria and the design-oriented conceptual model (Figure 1). This dataset will inevitably evolve as the review develops. For example, the process of extracting data from the included studies may identify unanticipated potential effect modifiers (moderators or mediators) that prompt revisions to our design-oriented conceptual model, as described above.
Study-level characteristics (between-study comparisons)
- Study design: between-subjects design, within-subjects design
- Study (intervention) setting: laboratory, field; for at home consumption versus away from home consumption
- Product type: food (including non-alcoholic beverages), alcohol, tobacco
- Product healthiness: Food Standards Agency (FSA) score (Rayner 2005) at level of specific product or, if not possible, at level of product category
- Target of manipulation: portion, package, individual unit, tableware
- Type of manipulation: size (including volume) or shape
- Manipulation from a standard size: no or yes
- If applicable, direction of the change relative to standard size: smaller or larger
- If applicable, selection with purchasing or selection without purchasing
- Concurrent intervention components (e.g. nutritional labelling)
- Socioeconomic status context (low, high)
- Information required to inform risk of bias assessments
- Information on funding source and potential conflicts of interest from funding
Intervention characteristics (within-study comparisons)
- Magnitude of the absolute difference in size (e.g. difference in quantity): smaller size always coded as Intervention 1 and larger size as Intervention 2
- Magnitude of the relative difference in size (e.g. percentage difference in quantity): smaller size always coded as Intervention 1 and larger size as Intervention 2
- Gender: male, female
- Body mass index (BMI), body weight or body weight status
- Behavioural characteristics (e.g. dietary restraint)
- Biological state (e.g. hunger, satiety)
- Other clinical characteristics (morbidities such as cardiovascular diseases, diabetes, psychiatric disorders)
- Occupational status (e.g. employed, unemployed)
- Other proxy measures of socioeconomic status (e.g. home ownership, geographic location)
These participant characteristics cover several categories of social differentiation relevant to health equity, namely: age, ethnicity, gender, occupation, education, income and other proxy measures of socioeconomic status. The incorporation of study-level data on these participant characteristics into our proposed meta-regression analysis (see ‘Data synthesis’) is in part intended to enable us to draw inferences concerning any differential effects of the intervention on health equity (Welch 2012). For example, within our proposed meta-regression (see ‘Data synthesis’) we intend to enter proxy measures of socioeconomic status as participant characteristics that may moderate the observed effects of the intervention on product selection and consumption. In addition, to complement investigations based on participant characteristics, we will use the most commonly available measure of socioeconomic status to construct a binary study-level covariate of ‘socioeconomic status context’ (see ‘Study-level characteristics’, above) that will serve as a proxy for the overall study context in terms of baseline levels of social and material deprivation amongst study participants. Analysis of this study-level covariate as a potential effect modifier will allow us to investigate specifically whether eligible interventions are more or less effective in a study context characterised by high versus low levels of social and material deprivation.
Mediators (mechanisms of action)
We will extract text data relating to proposed or observed mediators of effects (mechanisms of action), such as:
- personal norm or perception that a certain portion size is appropriate;
- perception of value for money.
It is anticipated that some eligible primary studies will include more than one eligible measure of each target outcome construct, specifically: (a) more than one eligible measure of selection for a given eligible comparison; or (b) more than one eligible measure of consumption for a given eligible comparison; or both. For each included study in which (a) or (b) applies, we will extract outcome data for the (a) primary selection or (b) primary consumption outcome(s) as (pre)-specified by the study authors. If the study authors did not (pre)-specify a single (primary) (a) selection or (b) consumption outcome, we will apply the following criteria to select the (a) selection or (b) consumption measure for which the data will be extracted from a list of all available measures. We will select the (a) selection or (b) consumption outcome that represents the largest-scale, most proximal measure of the target outcome construct in the context of the specific intervention at hand. For example, as regards the largest-scale measure, if a study manipulated the size of a portion of vegetable served as one component of a plated entrée, and measured the effects of a large versus a small vegetable portion size in terms of: (i) the amount of that vegetable consumed from the plated entrée, and (ii) the total amount of food consumed from the plated entrée, then we would select (ii) as the consumption outcome measure for which the data will be extracted. As regards the most proximal measure, if, for example, a study reported measures of both energy intake and the amount of food eaten (for example in grams) we would select energy intake as the measure most proximal to the target construct of consumption. We will make each selection in advance of data extraction and thus blinded to the outcome data. We will record details of all selection and consumption outcomes measures in each included study and document them in the 'Characteristics of included studies' table.
For all outcome data we will collect information on: outcome variable type (dichotomous, continuous); outcome variable definition; unit of measurement (if relevant); timing of measurement (immediate (that is ≤ 1 day) or longer term (> 1 day)); and type of measure (objective, self-report). For dichotomous outcomes we will extract event rates in each comparison group using 2 x 2 tables. For continuous outcomes we will extract mean differences, or mean changes in final measurements from baseline measurements, for each comparison group with their standard deviations (or, if standard deviations are missing, standard errors, 95% confidence intervals or the relevant information on P values or t-values); we will also indicate whether a high or low value is favourable. For included studies with factorial designs, in which the independent and combined effects of other manipulations are studied alongside a size manipulation, we will combine comparison groups so that any independent or combined effect of the co-occurring manipulation is averaged across the comparison of interest to allow investigation of the independent effects of the size manipulation.
Assessment of risk of bias in included studies
Risk of bias in the included studies will be assessed using the Cochrane risk of bias tool (Higgins 2011) addressing seven specific domains, namely: sequence generation and allocation concealment (selection bias); blinding of participants and providers (performance bias); blinding of outcome assessors (detection bias); incomplete outcome data (attrition bias); selective outcome reporting (reporting bias); and other sources of bias. Other sources of bias that will be assessed in this review will include baseline comparability of participant characteristics between groups and consistency in intervention delivery (that is whether information and specific instructions provided to participants were standardised between conditions and whether participant (non-)compliance with protocol was appropriately managed).
The Cochrane risk of bias tool will be applied to each included study by two researchers working independently (GJH, IS) and justifications for judgments of risk of bias (high, low or unclear) will be recorded, where possible, in the form of verbatim text extracted from study reports. Key information that is needed to inform risk of bias assessments but is missing from the reports of included studies will be sought by contacting the study authors. Any discrepancies in judgements of risk of bias or justifications for judgements will be identified and resolved by discussion to reach consensus between two researchers (GJH, IS), with a third researcher (DO) acting as arbiter in the case of any unresolved discrepancies.
Completed risk of bias tables will be presented in the published systematic review, including justificatory information for each judgement (as ‘Support for judgement’). Where judgements are based on either assumptions or information provided outside publicly available documents (for example supplementary information provided by study authors) this will be explicitly stated. We will include a summary assessment of risk of bias for each specific outcome included in our statistical analysis as a covariate in the final stage of the meta-regression analysis (see Data synthesis). We will also consider the summary risk of bias in determining the strength of inferences drawn from the results of the data synthesis and in developing conclusions and any recommendations concerning the design and conduct of future research.
Measures of treatment effect
For continuous outcomes, we will calculate the standardised mean difference (SMD) with 95% confidence intervals to express the size of the intervention effect in each study relative to the variability observed in that study. For dichotomous outcomes, we will calculate the odds ratio (OR) for each included study to express the size of the relative intervention effect between comparison groups, with the uncertainty in each result being expressed by the confidence interval. We will then re-express the odds ratio as a standardised mean difference by applying the formula described in Section 9.4.6 of the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2011). We will classify included study results according to two categories of timing of outcome measurement: immediate outcomes (that is ≤ 1 day) versus longer-term outcomes (that is > 1 day).
Unit of analysis issues
In the case of cluster-randomised trials, if an analysis is reported that accounts for the clustered study design, the effect will be estimated on this basis. If this is not possible and the information is not available from the authors, then an ’approximately correct’ analysis will be carried out according to current guidelines (Higgins 2011). Estimates of the intra-cluster correlation (ICC) will be imputed using estimates derived from similar studies, if possible other studies included in the review, or by using general recommendations from empirical research. We will report details of any such procedures. If it is not possible to implement these procedures, we will give the effect estimate as presented but report the unit of analysis error. For included studies with a within-subjects design, we will calculate the standardised mean difference for continuous outcomes using the methods described in Section 16.4 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). Final outcome values will serve as the primary unit of analysis. For studies assessing changes from baseline as a result of an experimental manipulation, we will aim to calculate final values based on reported data supplemented by additional data through contact with the authors, if needed.
Dealing with missing data
Data that are missing from reports of included studies will be sought by contacting the study authors. Where data are missing due to participant dropout we will conduct available case analyses and record any issues of missing data within the assessments conducted using the Cochrane risk of bias tool. However, we do not anticipate significant missing data due to participant dropout as we expect the large majority of studies to include only immediate or short-term outcomes.
Assessment of heterogeneity
Statistical heterogeneity in results will be assessed by inspection of a graphical display of the estimated treatment effects from included studies along with their 95% confidence intervals, and by formal statistical tests of homogeneity (χ
Assessment of reporting biases
We will draw funnel plots (plots of effect estimates versus the inverse of their standard errors) to inform assessment of reporting biases. We will conduct statistical tests to formally investigate the degree of asymmetry using the method proposed by Egger et al (Egger 1997). Results of statistical tests will be interpreted based on visual inspection of the funnel plots. Asymmetry of the funnel plot may indicate publication bias or other biases related to sample size, though it may also represent a true relationship between trial size and effect size.
We will describe and summarise the findings of included studies to address the two stated objectives of the review. We will provide a narrative synthesis describing the interventions, participants, study characteristics and effects of eligible interventions upon pre-specified outcomes (see Criteria for considering studies for this review). We will consider presenting the narrative syntheses in disaggregated form by type of product: food, alcohol and tobacco.
Our statistical analysis of the results of included studies will use a series of random-effects models to estimate pooled effect sizes with 95% confidence intervals in terms of each specified outcome. The precise configuration of our proposed statistical analysis will be determined based on the final iteration of our design-oriented conceptual model. Provisionally, our statistical analysis will comprise the following stages.
1. Conduct a standard meta-analysis to estimate pooled effect sizes for all eligible interventions versus all comparators.
2. Perform a meta-regression analysis with type of product (food, alcohol, tobacco) as a covariate.
3. Perform a meta-regression analysis with study-level characteristics as additional covariates.
4. Perform a meta-regression analysis with intervention characteristics as covariates. In this stage we may decide to incorporate multivariate analysis to deal with studies with multiple treatment arms in order for direct comparisons between each treatment arm and a control condition to be modelled, using mvmeta (White 2011). The latter decision will be contingent on the extent to which outcome data are derived from studies with multiple treatment arms (that is if there are few or no included multi-arm studies, multivariate analysis may not be appropriate).
5. Perform a meta-regression analysis with participant characteristics and risk of bias assessment as covariates. Product type, study-level characteristics and intervention characteristics will be re-entered into the model if they are found to be important predictors of effects at stages 2, 3 and 4 respectively.
Summary of findings table
We will use the standard GRADE system to rate the quality of the respective bodies of evidence for (1) selection (with or without purchasing) and (2) consumption outcomes in terms of the extent to which we can be confident that the synthesised estimates of effects are correct. GRADE criteria for assessing quality of evidence encompass study limitations, inconsistency, imprecision, indirectness and publication bias. Justifications underpinning all assessments will be documented and published in the review. We will present this information in a standard summary of findings table developed using GRADEprofiler (Brozek 2008), alongside the number of participants and studies for each outcome and a summary of the intervention effect.
Sensitivity analyses will be conducted to explore the impact of any outcome or moderator data that are imputed due to missing data.
We would like to acknowledge the work of Julie Glanville (University of York) who reviewed our MEDLINE search strategy and Claire Stansfield (EPPI-Centre, University of London) who helped to develop and run our TRoPHI search.
Appendix 1. MEDLINE search strategy
OVID SP MEDLINE 1946 -
1 exp Beverages/
2 exp Drinking Behavior/
3 exp Alcohol Drinking/
4 exp Food Industry/
5 exp Alcohol-Related Disorders/
6 (drink$ or drunk$ or alcohol$ or beverage$1 or beer$1 or lager$1 or wine$1 or cider$1 or alcopop$1 or alco-pop$1 or spirit or spirits or liquor$1 or liquer$1 or liqueur$1 or whisky or whiskey or whiskies or whiskeys or schnapps or brandy or brandies or gin or gins or rum or rums or tequila$1 or vodka$1 or cocktail$1).ti,ab.
7 exp Tobacco/
8 exp Smoking/
9 exp "Tobacco Use Disorder"/
10 (cigar$ or smoke or smokes or smoking or smoker or smokers or smoked or tobacco$).ti,ab.
11 exp Diet/
12 exp Food Industry/
13 exp Food/
14 exp Food Habits/
15 exp Food Preferences/
16 exp Eating/
17 exp Feeding Behavior/
18 exp Eating Disorders/
19 (nutri$ or calori$ or food$ or eat or eats or eaten or eating or ate or meal$ or snack$ or drink$ or drunk$ or beverage$1).ti,ab.
20 exp Food Packaging/
21 exp Food Storage/
22 exp Cooking/ and Eating Utensils/
23 exp Product Packaging/
24 ((siz$ or dimension$ or capacit$ or volume$ or shap$ or height$ or width$ or length$ or depth$ or divide$) adj4 (portion$ or serving$ or product$ or packag$ or packet$ or unit$ or cigar$ or food$ or drink$ or alcohol$ or tableware or drinkware or dinnerware or crockery or plate$1 or platter$1 or tureen$1 or tajine$1 or tagine$1 or bowl$1 or charger$1 or cup$1 or saucer$1 or glass or glasses or mug or mugs or beaker$1 or pitcher$1 or jug$1 or decanter$1 or receptacle$1 or container$1 or dish$ or pot or pots or cutlery or flatware or utensil$1 or knife or $knife or knives or fork$1 or spoon$ or $spoon or tongs or ladle$1 or chopstick$1 or box$ or bag$ or can$ or carton$1 or bottle$ or straw$1)).ti,ab.
29 25 and 28
30 26 and 28
31 27 and 28
34 (rat or rats or mouse or mice or murine or rodent or rodents or hamster or hamsters or pig or pigs or porcine or rabbit or rabbits or animal or animals or dog or dogs or cat or cats or cow or cows or bovine or sheep or ovine or monkey or monkeys).ti,ab.
36 humans/ and animals/
37 35 not 36
38 32 not 37
39 (editorial or case reports or in vitro).pt.
40 38 not 39
Contributions of authors
Draft the protocol - all authors
Develop a search strategy - GJH, IS
Search for trials - GJH, IS
Obtain copies of trials - GJH, IS
Select which studies to include - GJH, IS, DO
Extract data from studies - GJH, IS, HBL
Enter data into RevMan - GJH, IS
Carry out the analysis - YW, JPH
Interpret the analysis - all authors
Draft the final review - all authors
Update the review - all authors
Declarations of interest
Susan Jebb is Chair of the Public Health Responsibility Deal Food Network. All other authors declare no confounding interests.
Sources of support
- Kings College London, UK.Database access
- University of Cambridge, UK.Computer provision, database access
- University of East Anglia, UK.Database access
- University of Bristol, UK.Computer provision
- Plymouth University, UK.Computer provision
- Funded by UK Department of Health Policy Research Programme (107/0001- Policy Research Unit in Behaviour and Health), UK.
- YW was supported by the UK Medical Research Council (MRC) grant to the MRC Clinical Trials Unit Hub for Trials Methodology Research [Grant number MSA7355QP21], UK.