Estimating the effect of moving meat-free products to the meat aisle on sales of meat and meat-free products: A non-randomised controlled intervention study in a large UK supermarket chain

Carmen Piernas; Brian Cook; Richard Stevens; Cristina Stewart; Jennifer Hollowell; Peter Scarborough; Susan A Jebb

doi:10.1371/journal.pmed.1003715

. 2021 Jul 15;18(7):e1003715. doi: 10.1371/journal.pmed.1003715

Estimating the effect of moving meat-free products to the meat aisle on sales of meat and meat-free products: A non-randomised controlled intervention study in a large UK supermarket chain

Carmen Piernas ^1,^*, Brian Cook ¹, Richard Stevens ¹, Cristina Stewart ¹, Jennifer Hollowell ¹, Peter Scarborough ², Susan A Jebb ¹

PMCID: PMC8321099 PMID: 34264943

Abstract

Background

Reducing meat consumption could bring health and environmental benefits, but there is little research to date on effective interventions to achieve this. A non-randomised controlled intervention study was used to evaluate whether prominent positioning of meat-free products in the meat aisle was associated with a change in weekly mean sales of meat and meat-free products.

Methods and findings

Weekly sales data were obtained from 108 stores: 20 intervention stores that moved a selection of 26 meat-free products into a newly created meat-free bay within the meat aisle and 88 matched control stores. The primary outcome analysis used a hierarchical negative binomial model to compare changes in weekly sales (units) of meat products sold in intervention versus control stores during the main intervention period (Phase I: February 2019 to April 2019). Interrupted time series analysis was also used to evaluate the effects of the Phase I intervention. Moreover, 8 of the 20 stores enhanced the intervention from August 2019 onwards (Phase II intervention) by adding a second bay of meat-free products into the meat aisle, which was evaluated following the same analytical methods.

During the Phase I intervention, sales of meat products (units/store/week) decreased in intervention (approximately −6%) and control stores (−5%) without significant differences (incidence rate ratio [IRR] 1.01 [95% CI 0.95–1.07]. Sales of meat-free products increased significantly more in the intervention (+31%) compared to the control stores (+6%; IRR 1.43 [95% CI 1.30–1.57]), mostly due to increased sales of meat-free burgers, mince, and sausages. Consistent results were observed in interrupted time series analyses where the effect of the Phase II intervention was significant in intervention versus control stores.

Conclusions

Prominent positioning of meat-free products into the meat aisle in a supermarket was not effective in reducing sales of meat products, but successfully increased sales of meat-free alternatives in the longer term.

A preregistered protocol (https://osf.io/qmz3a/) was completed and fully available before data analysis.

Carmen Piernas and co-workers study positioning and marketing of meat-free products in supermarkets.

Author summary

Why was this study done?

Reducing meat consumption could bring health and environmental benefits, but there is limited evidence of interventions that may be effective.
Physical environments within which food choices are made can exert a significant influence on food selection. Supermarkets account for a large proportion of all United Kingdom retail grocery sales, and interventions to reduce meat selection store have the potential to reach a large number of consumers. But evidence on the effectiveness of in-store “nudges” to shift meat purchasing behaviours is lacking.
We assessed the immediate and longer-term effects of repositioning meat-free products (plant-based mince, burgers, meatballs, and sausages) into the meat aisle in a major UK supermarket on the purchases of meat and meat-free products.

What did the researchers do and find?

Prominent positioning of meat-free products in the meat aisle did not result in decreased sales of equivalent meat products (the primary outcome measure).
However, the intervention led to a significant increase in sales of meat-free products, which was sustained over time. Intensification of the intervention in the Phase II period further increased sales of meat-free products, with no impact on sales of meat products.
A post hoc analysis by store affluence suggested that the effects on sales of meat-free products were of greater magnitude in stores located in areas of average or below average affluence.

What do these findings mean?

Redesigning microenvironments in grocery stores could help shift habitual purchasing behaviours to reduce the demand for meat.
The limited effectiveness of this positioning intervention in meeting the primary goal, reduced meat consumption, suggests that it may be necessary to make structural changes that explicitly target a reduction in meat purchases to achieve global targets that are compatible with human health and environmental sustainability.

Introduction

Globally, the consumption of meat is rising, driven by population growth and increasing economic development. [1] There is evidence that consuming red and processed meat is associated with adverse health effects, including heart disease and cancer [2,3]. Furthermore, livestock production negatively affects the natural environment through greenhouse gas emissions and other outcomes [4]. Reducing meat consumption could bring health and environmental benefits, but there is little research to date on effective interventions to achieve this [5].

It is now widely recognised that the physical environments within which food choices are made can exert significant influence on food selection and are major determinants of dietary behaviour and obesity [6–8]. With regard to supermarkets, evidence suggests that current practices tend to encourage purchasing of energy-dense processed products [9,10]. Still, supermarkets may be a particularly promising setting for interventions aiming to influence consumer demand for meat because they account for approximately 87% of all UK retail grocery sales and their potential to reach a large number of consumers [11]. However, the evidence on the effectiveness of in-store “nudges” to shift meat purchasing behaviours is lacking. A systematic review of choice architecture interventions to reduce meat consumption found that 4 interventions that repositioned meat products to be less prominent at point of purchase showed promising reductions in the demand for meat. However, these studies were all conducted in buffet/canteen settings with restricted choices and food purchased for immediate consumption, with none in supermarket settings [12].

Systematic reviews that have drawn on broader evidence relating to interventions to shape food purchasing have found moderate- to low-quality evidence that product placement strategies (e.g., availability or prominent positioning (end of aisles, checkouts, or at eye level)) implemented in food retail environments can influence dietary behaviours [13,14]. Also, “cross-category merchandising” (where complementary products are colocated in a store) is an established sales mechanism to increase sales [15,16]. Research has shown that the provision of meat-free alternatives can help people to reduce their consumption of meat [17]; hence, positioning meat-free alternative products in store alongside meat products presents a potential opportunity to encourage switching from meat to meat-free alternatives.

This study aimed to evaluate an in-store positioning intervention in which a major UK supermarket moved 26 meat-free products (plant-based mince, burgers, meatballs, and sausages) into the meat aisle. Through a collaboration agreement with the supermarket, we obtained aggregated weekly store-level sales data, comprising sales of meat and meat-free products for 20 intervention stores and 88 matched control stores. Using this large dataset of supermarket sales, we aimed to assess the immediate and longer-term effects of repositioning meat-free products in the meat aisle on the purchases of meat and meat-free products.

Methods

Data source

Data on store-level weekly sales (units purchased and £) of meat and meat-free products were obtained from August 2017 to December 2019 (123 weeks) for 108 stores located across the UK (20 intervention and 88 matched control stores), resulting in 13,284 total store weeks of data. No stores were excluded from primary or secondary analyses due to the overall completeness of the data. For one store which had missing data for 2 non-consecutive weeks in August and September 2018 (0·01% store weeks), values were imputed using the average of each month. The study used only aggregate sales data to evaluate actions implemented by a supermarket and was exempt from ethical review and approval.

Store selection and matching

The retail partner’s finance and data teams determined the criteria for selecting the intervention and control stores. Intervention sites were all stores where there was unproductive space in the meat aisle. In other words, the meat sales did not justify the volume of inventory or number of product bays. For this study, 20 stores were selected based on the retailer’s estimates of the opportunity cost of removing meat products, as well as operational considerations. The stores were selected to have a broad spread with regard to the following characteristics: affluence (based on top 2 Acorn consumer classification categories within a 10-minute drive) [18]; sales of meat-free products over a 12-week period ending September 25, 2018; and store size (large (>50,000 sq ft), medium (35 to 50,000 sq ft), and small (<35,000 sq ft).

The retail partner used proprietary analytics software to select 88 unique control stores with similar purchasing patterns to the intervention stores based on individual-level purchasing data captured by their loyalty card scheme. For this study, the retailer identified up to 5 matched control stores for each intervention store. This was based on patterns of customer purchasing of meat-free alternatives over the 52 weeks preceding the intervention period as well as data on the store format, visits per customer, spend per visit, and customer affluence. A small number of control stores was initially matched to more than 1 intervention store. In these instances, we manually selected only 1 control per intervention store. Therefore, of the 20 intervention stores, 9 had 5 uniquely matched controls, 10 had 4 uniquely matched controls, and 1 had 3 uniquely matched control stores. The final sample of 108 stores included in this analysis were representative of the retail partner store population, with stores distributed across all UK regions, including small, medium, and large supermarket sites.

Intervention

The intervention was implemented in 20 stores on the week commencing January 27, 2019, for 12 weeks. In the control stores, meat-free products remained where originally located (usually in their own meat-free section outside the meat aisle) without changes to their position. The intervention consisted of moving 26 uniquely barcoded meat-free products within 4 meat-free categories (50% sausages, 30% burgers, 12% meatballs, and 8% mince) from their usual section into a newly created meat-free bay within the meat aisle. This new bay of meat-free products covered the whole bay, including top, middle, and bottom shelves. This meat-free section replaced a bay where either meat was previously sold (16 stores) or where the bay previously contained chilled condiments (3 stores) or was unused (1 store). The meat-free bay was the end bay in the meat aisle in 13 stores and in the middle of the meat aisle in 7 stores.

Additionally, the meat-free bay had a promotional header board above it in most stores (with the words “Meat Alternatives” and a meat-free burger image) and 2 point of sale displays (aisle fins) that included the following text, developed by the retail partner:

Simple swaps—More delicious meat alternatives can be found across the store
Plant power—Easy switches for an alternative source of protein.

The intervention was initially planned by the retail partner for 12 weeks until April 2019 (referred to as intervention Phase I). However, 12 of the 20 intervention stores continued the Phase I intervention until December 2019. For the remaining 8 intervention stores, the retailer intensified the intervention (referred to as intervention Phase II) in the week commencing July 28, 2019 by adding the entire range of meat-free products into a second meat-free bay within the meat aisle. This second bay of meat-free products included additional meat-free products (e.g., falafels, sausage rolls, tofu, and slices), which were stocked alongside the previous 26 products included in the Phase I intervention. These 8 stores were selected by the retailer based on each stores having sufficient meat-free products in stock for the second bay. The Phase II intervention continued in these 8 stores until December 2019.

Outcome measures

The primary outcome measure was average total weekly sales (units) of total meat products (aggregated measure of mince, burgers, meatballs, and sausages). Secondary outcomes included average total weekly sales (units) of total meat-free products (aggregated measure of meat-free mince, burgers, meatballs, and sausages); total weekly sales (£) of total meat products and total meat-free products; total weekly sales (units and £) of individual categories (mince, burgers, meatballs, and sausages) of meat products and meat-free products; total weekly sales (units and £) of other meat-free products not included above (e.g., tofu and falafel); and total weekly sales (units and £) of different aggregated product categories that were chosen as control products, including fish, nondairy milk, vegetables, fruit, and personal care products.

Other measures

The retailer also provided data on demographic characteristics relating to the surrounding population. These data were originally derived using the Acorn categorisation system, which uses postcode data as well as commercial and governmental data to classify geographical areas across the UK [18]. These measures were included as covariates in the analyses: affluence coded as average, >average, <average; age group coded as older versus younger; ethnicity coded as White, Asian, or Other; and area density coded as urban, more urban, or less urban.

Statistical analysis

A preregistered protocol (https://osf.io/qmz3a/) was completed and fully available from April 2019 before obtaining data for analyses (S1 Appendix). A more detailed statistical analysis plan was completed in January 2020 after data cleaning but before data analysis (S2 Appendix). Following the conduct and evaluation of a natural experiment, a power analysis was not conducted, and the final number of trial stores with matched control stores was chosen by the retailer.

For the primary analysis we used 2 prespecified 12-week time periods: (a) a pre-intervention baseline period (week commencing September 2, 2018 to November 18, 2018); and (b) a Phase I intervention period (week commencing February 3, 2019 to April 21, 2019), which excludes the week of January 27, 2019 to allow for the intervention to be fully implemented across all stores. The pre-intervention baseline period and the intervention period were defined a priori with the retail partner to avoid months where meat and meat-free purchases are known to be atypical (e.g., Christmas, January, and summer).

We investigated differences in store demographic characteristics between intervention versus control stores using chi-squared tests. We also investigated differences in aggregated store-level weekly sales (units and £) between intervention versus control stores over the pre-intervention baseline period (September 2, 2018 to November 18, 2018) as well as in January 2018 and January 2019 using Student t tests.

The primary outcome analysis evaluated the effect of the Phase I intervention by comparing changes in items/units sold per week of total meat products between intervention versus control stores during the Phase I intervention period (week commencing February 3, 2019 to April 21, 2019). We used a hierarchical negative binomial model (to account for the non-normal distribution of the outcome data—units sold/week) to obtain incidence rate ratios (IRRs), with fixed effects adjustment for store affluence, store age group, store ethnicity, store area, average units sold per week over the 12-week pre-intervention baseline period (week commencing September 2, 2018 to week commencing November 18, 2018), and a random effects term for matching group. Analyses of all the other secondary outcomes followed the same approach except for using a hierarchical normal model where the outcomes were sales (£) and followed a normal distribution.

In a prespecified secondary analysis, we conducted interrupted time series analyses on all available data from August 1, 2017 to week commencing December 1, 2019 using sales (units, £) of total meat and meat-free products [19]. For evaluation of the Phase I intervention, we plotted (a) the average sales of meat and meat-free products across all 20 intervention stores; and (b) the average sales of meat and meat-free products across all 88 control stores, together with fitted linear trend lines before and after the Phase I intervention began in the week commencing February 3, 2019. To assess whether differences visible in this graph were statistically significant between intervention and control stores, we used a difference-in-difference approach, calculating the mean difference in sales (units or £) at each week between intervention and control group, and testing whether this time series of differences changed after intervention time using a linear regression model. We used a Chow-type test for any difference (in either intercept or slope, or both) after versus before and used Newey–West standard errors with lag 2 to allow for autocorrelation in the time series. In sensitivity analyses, we tested the effect of controlling for seasonality. We used the same approach to evaluate the Phase II intervention, but we only used data from the 20 intervention stores (12 Phase I and 8 Phase II stores) from the of February 3, 2019 to December 1, 2019. In addition to the hierarchical mixed models, we used interrupted time series to plot (a) the average meat and meat-free sales across the 12 Phase I intervention stores; and (b) the average meat and meat-free sales across the 8 Phase II stores, together with fitted linear trend lines before and after the Phase II intervention began in the week commencing July 28, 2019.

In a post hoc sensitivity analysis, we repeated the same hierarchical mixed models for the primary outcome (unit sales) adjusting for an alternative baseline period (week commencing February 4, 2018 to April 22, 2018), which matched the intervention period (February 3, 2019 to April 21, 2019). In another post hoc exploratory analysis, we used interaction terms (store affluence * intervention) in the hierarchical negative binomial model described above to compare the effect of the intervention across store affluence groups. We used R 3.6.0 to produce graphics and Stata version 16 for the remaining statistical tests. All statistical tests were conducted at a 5% significance level.

Fidelity evaluations

We trained and employed community researchers to carry out in-store fidelity evaluations of all Phase I and Phase II intervention stores and a sample of control stores on 2 unannounced occasions throughout the intervention periods (1 week day and 1 weekend day). These visits aimed to assess whether the intervention was implemented as planned and collected information using a survey and photographs of the participating stores.

Results

This analysis included data from 108 stores, the majority of which were located in areas of average (49%) or above average affluence (39%) and below average urbanisation (61%). Approximately half of the stores were located in areas where the main ethnicity was White (51%), followed by Asian (41%). There were no significant differences in any of these characteristics between all intervention stores compared to control (Table 1), neither between intervention stores which implemented the Phase I intervention (n = 12) compared to those that implemented the Phase II intervention (n = 8, Table A in S3 Appendix).

Table 1. Demographic store characteristics relating to the surrounding population.

	Total stores n = 108		Intervention stores n = 20		Control stores n = 88		χ² test
	n	%	n	%	n	%	P value
Store size
Small	49	45	9	45	40	45	0.994
Medium	33	31	6	30	27	31
Large	26	24	5	25	21	24
Age
Older	52	48	11	55	41	47	0.497
Younger	56	52	9	45	47	53
Affluence
Less than average	13	12	2	10	11	13	0.953
Average	53	49	10	50	43	49
More than average	42	39	8	40	34	39
Ethnicity
White	55	51	11	55	44	50	0.327
Asian	44	41	9	45	35	40
Other	9	8	0	0	9	10
Area density
Less urban	66	61	12	60	54	61	0.510
Average urban	40	37	7	35	33	38
More urban	2	2	1	5	1	1

Open in a new tab

A trend analysis over the months preceding the implementation of the intervention (August 2017 to January 2019) showed that sales of meat products and equivalent meat-free products followed similar trends in both the intervention and control stores (Fig A in S3 Appendix). The selected meat products (mince, burgers, meatballs, and sausages) accounted for up to 14 times more sales than their equivalent meat-free products. There were clear seasonal changes in meat and meat-free products in the December to January period, with another peak of sales of meat products around April to May. There were no significant differences in average weekly sales of meat or equivalent meat-free products over the 12 week pre-intervention baseline period (week commencing September 2, 2018 to week commencing November 18, 2018) or the months of January 2018 and January 2019, between intervention and control stores (Table B in S3 Appendix).

After the implementation of the intervention Phase I (February to April 2019), sales of meat products (units per store per week) decreased in both intervention (approximately −6%) and control stores (−5%), but these changes were not significantly different (IRR 1.01 [95% CI 0.95 to 1.07], Fig 1, Table C in S3 Appendix). Changes in sales (units per store per week) of individual categories of mince, burgers, meatballs, or sausages were not significantly different either in the intervention stores compared to the control stores, nor were there differences when the outcomes were expressed as £ sales of meat products (Table C in S3 Appendix).

Fig 1 — *IRRs (95% CI) comparing changes in intervention stores vs. control stores using hierarchical negative binomial models with fixed effect adjustment for store affluence, store age group, store ethnicity, store area and average units sold per week in 12-week pre-intervention period (week commencing September 2, 2018 to week commencing November 18, 2018), and a random effects term for matching group. IRR, incidence rate ratio.

Sales of equivalent meat-free products (units per store per week) increased significantly more in the intervention stores (approximately +31%) compared to the control stores (+6%, IRR 1.43 [95% CI 1.30 to 1.57]. This change was mostly due to increased sales of meat-free burgers, mince, and sausages (Fig 1, Table C in S3 Appendix). Significant changes were also observed when the outcomes were expressed as £ sales, especially for sales of meat-free burgers and sausages (Table C in S3 Appendix). Changes in sales (units per store per week) of other meat-free products or products that were chosen as control products (including fish, nondairy milk, vegetables, fruit, and personal care) did not differ significantly between the intervention and control stores (Fig 1, Table D in S3 Appendix), but there were significant increases in sales (£ per store per week) of other meat-free products (£72, 95% CI 75 to 87) and nondairy milk (£17, 95% CI 4 to 30). A post hoc sensitivity analysis was performed on sales (units per store per week) of meat and meat-free products adjusting for an alternative baseline period, which matched more closely the intervention period, with both analyses showing consistent results (Table E in S3 Appendix).

Interrupted time series analyses of the main intervention Phase I (February to April 2019) showed consistent results (Fig 2, Figs B and C in S3 Appendix), with a decreasing trend in baseline sales (units and £ per store per week) of meat products before and after the intervention and no significant difference in the trend in sales of meat products after implementation (P = 0.7 for differences between intervention versus control stores). For meat-free products, there was a modest declining trend in sales before intervention implementation, but a significant increase in sales of meat-free products after the intervention (P < 0·001 for differences between intervention versus control stores). This was followed by an increasing secular trend in meat-free sales in both intervention and control stores. These results were robust in sensitivity analyses that adjusted for seasonal changes, e.g., those observed over the months of December and January (Fig D in S3 Appendix).

Hierarchical mixed methods (Tables F and G in S3 Appendix) as well as interrupted time series analyses (Fig 3) were also used to evaluate the Phase II intervention, where 8 of the 20 original intervention stores moved the entire meat-free bay into the meat aisle, while the remaining 12 stores continued as in the Phase I intervention. In interrupted time series analyses, there was a declining trend in sales (units per store per week) of meat products before the intervention in both Phase I and Phase II stores and a small significant level change in meat products after implementation (P < 0·001 for differences between Phase I versus Phase II stores), followed by a slightly increasing trend in meat products after the intervention in both Phase I and Phase II stores. For meat-free products, there was an increasing trend in sales (units per store per week) in the Phase II stores and a significant change in meat-free products after implementation of Phase II (P < 0·001 for differences between Phase I versus Phase II stores) followed by an increasing trend in both Phase I and Phase II stores.

Post hoc observational analysis of the Phase I intervention by store affluence showed that changes in sales of meat products (units per store per week) were not significantly different in the intervention stores compared to the control stores regardless of the store affluence (P_interaction = 0·738, Fig 4). However, among stores of average or below average affluence, sales of equivalent meat-free products increased markedly in the intervention stores (approximately +46%) compared to the control stores (+5%, IRR 1.61 [95% CI 1.45 to 1.78]), while this increase was less pronounced among stores of above average affluence (intervention stores (+15%) and control stores (+7% IRR 1.29 [95% CI 1.08 to 1.56]); P_interaction = 0·021, Fig 4).

Fig 4 — *Changes in intervention vs. control stores were compared using hierarchical negative binomial models with fixed effect adjustment for store affluence, store age group, store ethnicity, store area and average units sold per week in 12-week pre-intervention period (week commencing September 2, 2018 to week commencing November 18, 2018) and a random effects term for matching group. To compare the effect of the intervention across affluence groups, interaction terms (affluence groups * intervention) were included in the model: P_interaction for meat products = 0·738; P_interaction for meat-free products = 0.021.

Fidelity evaluations

The intervention was implemented in all 20 Phase I stores as planned, with 62% of them having at least 80% of items in the meat-free planogram in stock. Of the 20 stores, 15 stores had at least 1 meat-free product missing (no product shelf label), with 1 store missing 10 meat-free products. Point of sale displays were present in 13 stores. One Phase II intervention store did not add a second meat-free bay, and all other stores had considerable variation in the number and type, of meat-free products stocked within the additional bay over the Phase II intervention period. See Table H in S3 Appendix for a detailed breakdown of the fidelity evaluation results.

Discussion

This non-randomised controlled intervention study in a large UK supermarket chain provided evidence that prominent positioning of meat-free products in the meat aisle did not result in decreased sales of equivalent meat products (the primary outcome measure). However, the intervention led to a significant increase in sales of meat-free products, which was sustained over time. Intensification of the intervention in the Phase II period further increased sales of meat-free products, with no impact on sales of meat products. Our post hoc analysis by store affluence suggested that the effects on sales of meat-free products were of greater magnitude in stores located in areas of average or below average affluence.

There is limited and mixed evidence on the effectiveness of interventions to reduce the demand for meat where meat-free items are made more prominent [20,21], but no previous studies have tested this in real supermarket environments. Our analysis did not detect a significant effect on sales of meat products, despite the fact that the meat-free section replaced a bay where meat was previously sold in most of the intervention stores. However, sales of meat were clearly in decline over the preceding 18 months in both intervention and control stores, which may have limited the potential for the intervention to further reduce meat sales. This is consistent with declining trends in consumption of red and processed meat observed in the UK population [22,23]. Additionally, the subset of meat products studied here included a combination of fresh processed and unprocessed products (i.e., mince, burgers, meatballs, and sausages), and we cannot rule out the possibility that other sources of meat, such as carcass cuts, would have changed as a result of the intervention, and we assessed sales only of refrigerated and not frozen meat and meat-free products.

Prominent positioning of meat-free products had a significant long-lasting effect on sales of these products. Recent systematic reviews support the value of positioning strategies in retail environments to improve the nutritional quality of food purchases [13,14]. Evidence suggests that there is a greater likelihood of change when more options are available [24], and repeated exposure can also elicit increased acceptability or establish a new social norm about the presence of new products [25]. However, this study reinforces previous evidence that these nudges to encourage sales of specific products may not be sufficient to displace choices where the goal is to decrease purchases [16,26]. In practice, the range of meat-free products that were repositioned in the meat aisle had similar culinary properties and use to their equivalent meat products, and we hypothesised that they would act as direct substitutes for meat. However, the perceived palatability, price, level of processing, or acceptability of the meat-free alternatives by the purchaser or other household members may be a barrier to this substitution [27,28]. Our results provide preliminary evidence that the intervention has potential to be particularly effective in lower and middle affluence populations, which has been observed with similar policies to improve nutrition [29], although evidence supporting the potential of this kind of intervention to reduce inequalities is insufficient [14].

A strength of this study is the use of a real-world setting and objective sales data to evaluate the intervention in a natural experiment, where consumers are not aware of the intervention being tested, and, therefore, better reflects actual rather than intended behaviour. Although stores were not randomised to intervention and control conditions, we showed that they were well matched. The choice of multiple control stores with matching store sales and area characteristics (plus adjustment for confounding) may result in less biased estimates of effect. We analysed data over a long time period before and after the intervention was implemented, which allowed us to control for previous trends and seasonal changes and to explore sustained effects after the intervention. Our retail partner is a national supermarket chain with one of the largest shares of the grocery sector, and its offerings are targeted to a general and broad customer base, which increases the generalisability of the findings to the broader UK population. This study also demonstrates how academic collaborations with the food industry can yield important information to inform public health interventions.

Some methodological limitations are inherent to natural field experiments. In-store environments are inundated with multiple cues that influence food choices and are especially challenging to control for in commercial environments. For example, stores could have implemented price promotions on meat products while positioning the meat-free alternatives, although due to national pricing policies, we would expect these to affect both intervention and control stores to a similar extent. We were also unable to look fully at substitution effects because we had data for only a limited range of products, which were part of the intervention or hypothesised as substitutes or as control products. For example, participants may have changed purchases of prime meat cuts in the same aisle or of frozen meat or meat alternatives located elsewhere in the store. We were not able to access any information regarding availability of products to understand if there were changes in the proportion of meat-free and meat products available on display, which could also explain the results. We also relied on the store implementation plans, and our fidelity evaluations revealed some deviations from the original plan in regards to signposting and positioning within the bays. However, the effect of these deviations was to reduce the likely effectiveness of the intervention.

Dietary change at the population level focusing on reductions in saturated fat, free sugars, or salt or to increase fruit and vegetable intake to reduce the risk of noncommunicable diseases is proving to be slow and especially challenging when using standard policy strategies aiming at targeting conscious decision-making processes of human behaviour [30]. There is less evidence pertaining to specific interventions to reduce meat or meet sustainability goals, although it has been suggested that physical microenvironments or social norms may override conscious intentions to follow better food practices [31,32] and that redesigning microenvironments could help shift habitual behaviours to reduce the demand for meat [8,33,34]. However, our research suggests that prominent positioning of meat-free products next to equivalent meat-products is not enough to reduce meat purchases. Positioning interventions aim at guiding food choices rather than reducing personal autonomy through structural changes. These kinds of interventions have high acceptability to governments, since they have no cost to the state, to retailers since they focus on enhancing rather than limiting sales, and the public who favour nudging over price interventions [35]. However, their limited effectiveness in meeting the primary goal, reduced meat consumption, suggests that it may be necessary to make structural changes that explicitly target meat purchases, for example, to reduce the space dedicated to meat or remove incentives to purchase meat to achieve global targets that are compatible with human health and environmental sustainability [36].

In conclusion, an intervention to make meat-free products more prominent by locating them in the meat aisle next to equivalent meat products was not effective in reducing sales of meat products, but successfully increased sales of meat-free alternatives in the long term. This research contributes to the under-explored field of effective in-store interventions to change food purchasing patterns whether to meet health or sustainability goals and supports prior evidence that positioning interventions can increase sales of prominently placed products within supermarket contexts. However, further research is needed into effective, scalable, and sustainable approaches that can reduce the demand for meat and encourage more sustainable patterns of consumption.

Disclaimers: The views expressed in this publication are those of the authors and not necessarily those of the National Institute for Health Research or the Department of Health and Social Care.

Supporting information

S1 Appendix. Study protocol.

(PDF)

Click here for additional data file.^{(1MB, pdf)}

S2 Appendix. Study statistical analysis plan.

(PDF)

Click here for additional data file.^{(902.5KB, pdf)}

S3 Appendix. Supporting information tables and figures.

(DOCX)

Click here for additional data file.^{(876.9KB, docx)}

Data Availability

This research was conducted according to a framework collaboration agreement between the University of Oxford and the food retailer. Access to the study dataset by external researchers is not permitted without the expressed written consent of the retailer as this is defined as confidential information in the agreement.

Funding Statement

This analysis was funded by the Wellcome Trust, Our Planet Our Health (Livestock, Environment and People – LEAP), award number 205212/Z/16/Z. SAJ is a National Institute of Health Research (NIHR) senior investigator and is funded by NIHR Oxford Biomedical Research Centre. SAJ and CP are funded by the Oxford and Thames Valley NIHR Applied Research Centre. The retail partner funded the cost of the intervention itself and extracted the data for analysis. They selected the intervention and control stores but had no further part in developing the protocol, conducting the analysis, or drafting the paper. The principal funder of the study and the participating retailer had no role in study design, data collection, data analysis, data interpretation, preparation of the manuscript or decision to publish.

References

1.York R, Gossard MH. Cross-national meat and fish consumption: exploring the effects of modernization and ecological context. Ecol Econ. 2004;48(3):293–302. doi: 10.1016/j.ecolecon.2003.10.009 [DOI] [Google Scholar]
2.Afshin A, Sur PJ, Fay KA, Cornaby L, Ferrara G, Salama JS, et al. Health effects of dietary risks in 195 countries, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet . doi: 10.1016/S0140-6736(19)30041-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Key TJ, Appleby PN, Bradbury KE, Sweeting M, Wood A, Johansson I, et al. Consumption of Meat, Fish, Dairy Products, and Eggs and Risk of Ischemic Heart Disease: A Prospective Study of 7198 Incident Cases Among 409 885 Participants in the Pan-European EPIC Cohort. Circulation. 2019;139(25):2835–45. doi: 10.1161/CIRCULATIONAHA.118.038813 [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Steinfeld H, Gerber P, Wassenaar T, Castel V, Rosales M, Rosales M, et al. Livestock’s long shadow: environmental issues and options. Food & Agriculture Org.; 2006. [Google Scholar]
5.Bianchi F, Dorsel C, Garnett E, Aveyard P, Jebb SA. Interventions targeting conscious determinants of human behaviour to reduce the demand for meat: a systematic review with qualitative comparative analysis. Int J Behav Nutr Phys Act. 2018;15(1):102. doi: 10.1186/s12966-018-0729-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Caspi CE, Sorensen G, Subramanian SV, Kawachi I. The local food environment and diet: a systematic review. Health Place. 2012;18(5):1172–87. Epub 2012/06/22. doi: 10.1016/j.healthplace.2012.05.006 ; PubMed Central PMCID: PMC3684395. [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Holsten JE. Obesity and the community food environment: a systematic review. Public Health Nutr. 2009;12(3):397–405. Epub 2008/05/15. doi: 10.1017/S1368980008002267 . [DOI] [PubMed] [Google Scholar]
8.Hollands GJ, Shemilt I, Marteau TM, Jebb SA, Kelly MP, Nakamura R, et al. Altering micro-environments to change population health behaviour: towards an evidence base for choice architecture interventions. BMC Public Health. 2013;13(1):1218. doi: 10.1186/1471-2458-13-1218 [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Bond MECB, Loff B. It’s not about choice: the supermarket and obesity. Med J Aust. 2012;197(7):371. doi: 10.5694/mja12.10923 [DOI] [PubMed] [Google Scholar]
10.The Obesity Health Alliance. Out of place: The extent of unhealthy foods in prime locations in supermarkets. 2018. [Google Scholar]
11.Department for Environment Food and Rural Affairs. Food Statistics Pocketbook 2014. London: 2015. 2015. Report No. [Google Scholar]
12.Bianchi F, Garnett E, Dorsel C, Aveyard P, Jebb SA. Restructuring physical micro-environments to reduce the demand for meat: a systematic review and qualitative comparative analysis. Lancet Planet Health. 2018;2(9):e384–e97. doi: 10.1016/S2542-5196(18)30188-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Hollands GJ, Carter P, Anwer S, King SE, Jebb SA, Ogilvie D, et al. Altering the availability or proximity of food, alcohol, and tobacco products to change their selection and consumption. Cochrane Database Syst Rev. 2019;9. [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Shaw SC, Ntani G, Baird J, Vogel CA. A systematic review of the influences of food store product placement on dietary-related outcomes. Nutrition Reviews. 2020. doi: 10.1093/nutrit/nuaa024 [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Cameron AJ, Charlton E, Ngan WW, Sacks G. A Systematic Review of the Effectiveness of Supermarket-Based Interventions Involving Product, Promotion, or Place on the Healthiness of Consumer Purchases. Curr Nutr Rep. 2016;5(3):129–38. doi: 10.1007/s13668-016-0172-8 [DOI] [Google Scholar]
16.Foster GD, Karpyn A, Wojtanowski AC, Davis E, Weiss S, Brensinger C, et al. Placement and promotion strategies to increase sales of healthier products in supermarkets in low-income, ethnically diverse neighborhoods: a randomized controlled trial. Am J Clin Nutr. 2014;99(6):1359–68. doi: 10.3945/ajcn.113.075572 [DOI] [PubMed] [Google Scholar]
17.Holloway T, Salter AM, McCullough FS. Dietary intervention to reduce meat intake by 50% in University students–a pilot study. Proceedings of the Nutrition Society. 2012;71(OCE2):E164. Epub 2012/10/19. doi: 10.1017/S0029665112002212 [DOI] [Google Scholar]
18.Acorn. The Acorn user guide: Consumer classification 2020. Available from: https://www.caci.co.uk/sites/default/files/resources/Acorn%20User%20Guide%202020.pdf.
19.Bernal JL, Cummins S, Gasparrini A. Interrupted time series regression for the evaluation of public health interventions: a tutorial. Int J Epidemiol. 2016;46(1):348–55. doi: 10.1093/ije/dyw098 [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Campbell-Arvai V, Arvai J, Kalof L. Motivating sustainable food choices: The role of nudges, value orientation, and information provision. Environ Behav. 2014;46(4):453–75. [Google Scholar]
21.Kongsbak I, Skov LR, Nielsen BK, Ahlmann FK, Schaldemose H, Atkinson L, et al. Increasing fruit and vegetable intake among male university students in an ad libitum buffet setting: A choice architectural nudge intervention. Food Qual Prefer. 2016;49:183–8. [Google Scholar]
22.Bates B, Colllins D, Cox L,Nicholosn S, Page P, Roberts C, Steer T, & Swan G. National Diet and Nutrition Survey Years 1 to 9 of the Rolling Programme (2008/2009-2016/2017): Time trend analysis and income analyses. London: Public Health England. 2019. [Google Scholar]
23.Godfray HCJ, Aveyard P, Garnett T, Hall JW, Key TJ, Lorimer J, et al. Meat consumption, health, and the environment. Science. 2018;361(6399). Epub 2018/07/22. doi: 10.1126/science.aam5324 . [DOI] [PubMed] [Google Scholar]
24.Chernev A. Product assortment and consumer choice: An interdisciplinary review. Foundations and Trends in Marketing. 2011;6(1):1–61. [Google Scholar]
25.Dalenberg JR, Nanetti L, Renken RJ, de Wijk RA, ter Horst GJ. Dealing with consumer differences in liking during repeated exposure to food; typical dynamics in rating behavior. PLoS ONE. 2014;9(3):e93350. doi: 10.1371/journal.pone.0093350 [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Adam A, Jensen JD, Sommer I, Hansen GL. Does shelf space management intervention have an effect on calorie turnover at supermarkets? J Retail Consum Serv. 2017;34:311–8. [Google Scholar]
27.Hoek AC, Luning PA, Weijzen P, Engels W, Kok FJ, de Graaf C. Replacement of meat by meat substitutes. A survey on person- and product-related factors in consumer acceptance. Appetite. 2011;56(3):662–73. doi: 10.1016/j.appet.2011.02.001 [DOI] [PubMed] [Google Scholar]
28.Apostolidis C, McLeay F. It’s not vegetarian, it’s meat-free! Meat eaters, meat reducers and vegetarians and the case of Quorn in the UK. Social Business. 2016;6(3):267–90. [Google Scholar]
29.Ejlerskov K, Sharp SJ, Stead M, Adamson AJ, White M, Adams J. Socio-economic and age variations in response to supermarket-led checkout food policies: a repeated measures analysis. Int J Behav Nutr Phys Act. 2018;15(1):125. doi: 10.1186/s12966-018-0755-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Marteau TM, Hollands GJ. Fletcher PC. Changing human behavior to prevent disease: the importance of targeting automatic processes. Science. 2012;337(6101):1492–5. doi: 10.1126/science.1226918 [DOI] [PubMed] [Google Scholar]
31.Marteau TM. Towards environmentally sustainable human behaviour: targeting non-conscious and conscious processes for effective and acceptable policies. Philos Trans A Math Phys Eng Sci. 2017;375(2095):20160371. doi: 10.1098/rsta.2016.0371 [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Stoll-Kleemann S, Schmidt UJ. Reducing meat consumption in developed and transition countries to counter climate change and biodiversity loss: a review of influence factors. Reg Environ Chang. 2017;17(5):1261–77. [Google Scholar]
33.Hollands GJ, Bignardi G, Johnston M, Kelly MP, Ogilvie D, Petticrew M, et al. The TIPPME intervention typology for changing environments to change behaviour. Nat Hum Behav. 2017;1(8):0140. [Google Scholar]
34.Sunstein CR, Reisch LA. Behaviorally green: Why, which and when defaults can help. New Perspectives for Environmental Policies Through Behavioral Economics: Springer; 2016. p. 161–94. [Google Scholar]
35.Reynolds J, Archer S, Pilling M, Kenny M, Hollands GJ, Marteau T. Public acceptability of nudging and taxing to reduce consumption of alcohol, tobacco, and food: a population-based survey experiment. Soc Sci Med. 2019;236:112395. doi: 10.1016/j.socscimed.2019.112395 [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Willett W, Rockström J, Loken B, Springmann M, Lang T, Vermeulen S, et al. Food in the Anthropocene: the EAT–Lancet Commission on healthy diets from sustainable food systems. Lancet. 2019;393(10170):447–92. doi: 10.1016/S0140-6736(18)31788-4 [DOI] [PubMed] [Google Scholar]

PLoS Med. doi: 10.1371/journal.pmed.1003715.r001

Decision Letter 0

Raffaella Bosurgi

1 Feb 2021

Dear Dr Piernas,

Thank you for submitting your manuscript entitled "Estimating the effect of moving meat-free products to the meat aisle on sales of meat and meat-free products: a non-randomised controlled intervention study in a large UK supermarket chain" for consideration by PLOS Medicine.

Your manuscript has now been evaluated by the PLOS Medicine editorial staff with relevant expertise and I am writing to let you know that we would like to send your submission out for external peer review.

However, before we can send your manuscript to reviewers, we need you to complete your submission by providing the metadata that is required for full assessment. To this end, please login to Editorial Manager where you will find the paper in the 'Submissions Needing Revisions' folder on your homepage. Please click 'Revise Submission' from the Action Links and complete all additional questions in the submission questionnaire.

Please re-submit your manuscript within two working days, i.e. by .

Once your full submission is complete, your paper will undergo a series of checks in preparation for peer review. Once your manuscript has passed all checks it will be sent out for review.

Feel free to email us at plosmedicine@plos.org if you have any queries relating to your submission.

Kind regards,

Raffaella

Dr Raffaella Bosurgi MSc, PhD

Executive Editor, PLOS Medicine

rbosurgi@plos.org

https://twitter.com/raffi74

Remote based in London, UK

PLOS

Empowering researchers to transform science

PLoS Med. doi: 10.1371/journal.pmed.1003715.r002

Decision Letter 1

Raffaella Bosurgi

22 Mar 2021

Dear Dr. Piernas,

Thank you very much for submitting your manuscript "Estimating the effect of moving meat-free products to the meat aisle on sales of meat and meat-free products: a non-randomised controlled intervention study in a large UK supermarket chain" (PMEDICINE-D-21-00493R1) for consideration at PLOS Medicine.

Your paper was evaluated by the PLOS MED editorial team editors present (Raffaella Bosurgi, Richard Turner, Caitlin Moyer, Beryne Odeny). It was also sent to independent reviewers, including a statistical reviewer. The reviews are appended at the bottom of this email and any accompanying reviewer attachments can be seen via the link below:

[LINK]

In light of these reviews, I am afraid that we will not be able to accept the manuscript for publication in the journal in its current form, but we would like to consider a revised version that addresses the reviewers' and editors' comments. Obviously we cannot make any decision about publication until we have seen the revised manuscript and your response, and we plan to seek re-review by one or more of the reviewers.

In revising the manuscript for further consideration, your revisions should address the specific points made by each reviewer and the editors. Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments, the changes you have made in the manuscript, and include either an excerpt of the revised text or the location (eg: page and line number) where each change can be found. Please submit a clean version of the paper as the main article file; a version with changes marked should be uploaded as a marked up manuscript.

In addition, we request that you upload any figures associated with your paper as individual TIF or EPS files with 300dpi resolution at resubmission; please read our figure guidelines for more information on our requirements: http://journals.plos.org/plosmedicine/s/figures. While revising your submission, please upload your figure files to the PACE digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at PLOSMedicine@plos.org.

We expect to receive your revised manuscript by Apr 12 2021 11:59PM. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

***Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.***

We ask every co-author listed on the manuscript to fill in a contributing author statement, making sure to declare all competing interests. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. If new competing interests are declared later in the revision process, this may also hold up the submission. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT. You can see our competing interests policy here: http://journals.plos.org/plosmedicine/s/competing-interests.

Please use the following link to submit the revised manuscript:

https://www.editorialmanager.com/pmedicine/

Your article can be found in the "Submissions Needing Revision" folder.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information for obtaining the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

We look forward to receiving your revised manuscript.

Sincerely,

Dr Raffaella Bosurgi,

Executive Editor

PLOS Medicine

plosmedicine.org

-----------------------------------------------------------

Requests from the editors:

Comments from the reviewers:

Reviewer #1: This non-randomised controlled intervention study aims to evaluate whether prominent positioning of meat-free products in the meat aisle was associated with a change in weekly mean sales of meat and meat-free products.

Comments:

Did the authors undertake any sample size calculations (a priori) to provide an estimate of statistical power for the study of 20 stores with matched controls?

A randomised control design would have been more rigorous and robust, with greater strength of conclusions in terms of inferring causality and minimising the risk of confounding.

However, it is good that a matched-control design has been followed, which goes some way towards understanding what might have happened in the scenario of no intervention.

Do the 108 stores included represent the wider population of stores for this retail partner?

"Intervention sites were all stores where there was unproductive space in the meat aisle. In other words, the meat sales did not justify the volume of inventory or number of product bays. For this study, 20 stores were selected based on the retailer's estimates of the opportunity cost of removing meat products, as well as operational considerations.

The stores were selected to have a broad spread with regard to the following characteristics: 1) affluence (based on top 2 Acorn consumer classification categories[17] within a 10 minute drive, in combination with meat-free sales performance (assessed in the 12 week period ending 25 September 2018)); and 2) store size (large, medium, small)."

Whilst Table 1 shows little difference between the characteristics of the 20 intervention stores and their matched controls, Figure 2 (left) suggests that intervention and control groups had different purchasing behaviours pre-intervention?

Figure 2 also shows that control stores saw a slope change, despite no intervention - what might be the reasons for this?

"The primary outcome analysis used a hierarchical negative binomial model to compare changes in weekly sales (units) of meat products sold in intervention vs control stores during the main intervention period (Phase I-February 2019-April 2019). Interrupted time series analysis was also used to evaluate the effects of the phase I intervention."

The modelling technique applied is appropriate for the data and research question.

Can the authors please clarify whether they compared weekly sales, or changes in weekly sales?

Seasonality is almost certainly at play in this data, as acknowledged by the authors.

It is noted that: "The preintervention baseline period and the intervention period were defined a priori with the retail partner to avoid months where meat and meat-free purchases are known to be atypical (e.g. Christmas, January, summer)." This is good practice, but does not explore whether baseline months might be different to Phase I and Phase II intervention months, due to reasons other than different products in the meat aisle (such as seasonality).

The authors go on to say that: "In sensitivity analyses we tested the effect of controlling for seasonality and changes in purchases over January." Can the authors please present these sensitivity analyses in the supplementary material?

Given the longitudinal data available prior to intervention, this seasonality effect across a complete year could be explored? ("Data on store-level weekly sales (units purchased and £) of meat and meat-free products were obtained for August 2017 to December 2019 (123 weeks) for 108 stores (20 intervention and 88 matched control stores), resulting in 13,284 total store-weeks of data").

An alternative approach might be to compare the same months as a sensitivity analysis? i.e. Feb-Apr 2018 compared to Feb-Apr 2019?

The authors have suitably included covariates in their analysis in an attempt to account for various potential sources of confounding that may be at play:

"These measures were included as covariates in the analyses: affluence coded as average, >average, <average; age group coded as older vs younger; ethnicity coded as White, Asian, Other; area density coded as urban, more urban, less urban. "

"To assess whether differences visible in this graph were statistically significant between intervention and control stores, we used a difference-in-difference approach, calculating the mean difference in sales (£) at each week between intervention and control group, and testing whether this time series of differences changed after intervention time using a linear regression model. We used a Chow-type test for any difference (in either intercept or slope, or both) after vs. before, and used Newey-West standard errors with lag to allow for autocorrelation in the time series"

The authors have calculated aggregate differences between control and intervention groups, and compared the linear regressions before vs. after intervention.

This is a novel approach which attempts to draw the most from the data at hand. However, it is not necessarily the most intuitive methodology, and may be difficult to interpret given the other possible differences between control and intervention groups.

Did the authors also consider testing for differences in linear regressions for intervention groups and control groups, before vs. after (i.e. comparing intervention to intervention, or control to control, before and after intervention)? This additional analysis may help to provide some depth and improved understanding as to what is changing over time, and how.

"Eight of the 20 stores enhanced the intervention from August 2019 onwards (Phase II intervention) by moving the entire meat-free section into the meat aisle, which was evaluated following the same analytical methods."

Thanks to Phase II of the study, the authors now have access to a cohort of stores with a) no intervention, b) some intervention and c) full intervention. Did they consider running analyses across all three of these groups in an attempt to understand if a dose-response relationship could be measured?

Similar to the non-randomisation for the original 20 stores which may cause a bias in the data, there may be some issue regarding the selection of these 8 stores: "These eight stores were selected by the retailer based on sufficient meat-free products in stock." What might this imply about the sample of 8 stores with enhanced intervention? Is there a bias here, that might suggest the enhanced intervention was more likely to be effective?

Figure 3 shows differences between the Phase I and enhanced intervention group before Phase II started. The Phase I stores saw a step and slope change, despite no additional intervention. Why might this be?

"A small number of control stores was initially matched to more than one intervention store. In these instances, we manually selected only one control per intervention store. Therefore, of the 20 intervention stores, nine had five matched controls, ten had four matched controls, and one had three matched control stores. "

Can the authors please clarify - each control store has been uniquely matched to one intervention store? (i.e. the 88 controls refer to 88 distinct stores)?

"A preregistered protocol (https://osf.io/qmz3a/) was completed and fully available from April 2019 before obtaining data for analyses. A more detailed statistical analysis plan was completed in January 2020 after data cleaning but before data analysis. "

Can the authors please provide these within the supplementary material?

The text of the manuscript suggests that one particular retail partner or supermarket chain is included in the analysis. It is appreciated that possibly the supermarket chain cannot be named, however, can the authors please discuss how the results seem here may or may not be applicable to other chains? (different market and clientele, pricing structures etc. may affect generalisation of the study findings?).

Of note, the title of the Supplementary Material document names the retail partner.

Given the authors have (presumably deliberately) not specified this within the manuscript, should this title be changed for commercial confidentiality reasons?

Supplementary Figure 1 shows identical weekly sales (where the lines completely overlap)- is this correct?

Reviewer #2: I congratulate the team for the design, delivery and analysis of this impressive experiment - this is exactly the sort of information that is needed to help understand the potential impact of food industry-led interventions to change dietary practices in the UK. This has clearly been a hugely difficult study to conduct given what seems to have been quite careful "management" of the study by your industry partner and a change to the protocol mid-way through the intervention. My comments below relate largely to the accurate and appropriate reporting of your study.

1. The very helpful online protocol calls this a "pilot non-randomised before and after study". Indeed, the protocol states that one of the secondary aims of your "pilot" study is to "to obtain data needed to design and conduct a larger, more definitive in-store intervention study". Is this the case, in which should this study be better re-titled to acknowledge its pilot nature? It's noteworthy for example that there is no sample size calculation in the manuscript (you justify this in the protocol by stating this is a pilot study).

2. The online protocol relates to the study as originally planned - and this does not entirely reflect what actually happened (i.e. Phase 2). If this was an RCT then this would of course have been a major protocol change and the protocol would have need to be update. I realise that this was not an RCT but I wonder if it would have been appropriate to update your protocol to reflect what actually happened.

3. The description of the intervention(s) could be improved to help the reader. There is currently insufficient distinction between the description of Phase 1 "to their own meat-free bay in the meat aisle (l.122-123) and Phase 2 "moving the entire bay of meat-free products in the mat aisle" (l.136-137). It's not clear what "the entire bay" consists of and how that is different from "their own meat-free bay". "The meat-free bay was the last bay in the meat aisle…" surely depends on the direction in which you access the aisle - maybe "end bay" would be an alternative? It's not clear what happened in the stores between April 2019 and July 2019 (i.e. l.132-139 needs to be re-written).

4. The reader would, I'm sure, welcome a bit more description of "Acorn categories" (l.159) to help understand more clearly the metrics used to define store characteristics.

5. This reader (at least) would like to see the statistical analysis plan. It is highlighted in the text (l.164-165) but does not seem to have been made available. This is particularly important as the analysis you present in the manuscript deviates in several areas from that in your pre-specified protocol.

6. The fidelity evaluations are clearly important not only to help interpret the study but also to aid wider implementation (i.e. lessons learnt). The protocol suggests that you planned to collect a lot of detailed information about stocking and positioning of meat-free products (including photographs) and the manuscript summarises this as "to assess whether the intervention was implemented as planned". This is critical information that allows the reader to understand more about the implementation of the intervention. The 5 lines on fidelity in the result section provide insufficient and non-systematic detail on this important issue. It's not clear how the reader is supposed to interpret a finding such as "with one store missing 10 meat-free products". Was any of the fidelity information used to inform additional sensitivity analysis?

7. Table 1 is hard to interpret as the categories have not been clearly defined.

8. The August 2017-January 2019 trend analysis (supplementary Figure 1) is interesting. I am surprised that while you have noted in the text that the intervention and control stores "followed similar trends" you did not report that there is a clear and consistent difference between the groups over the entire 17 month period i.e. the control stores sold ~1-200 more units of meat than the intervention stores. To me this suggests that the matching didn't work perfectly.

9. Some of the changes identified in meat-free product sales are tiny - for example in Phase 1, meat-free mince changed from 24 units at baseline to 27 units (while the control dropped from 25 to 24). This shows up as a big effect in Figure 1 but I would suggest you consider more carefully how this is reported. Clearly the size of the change for meat-free burgers and sausages is more substantial.

10. You need to be more consistent (or at least systematic) in the way you report the primary results that I assume come from the pre-specified Phase 1 of the study. For example, why in l.242-244 to you switch from your pre-specified primary outcome "change in weekly sales (units)" to "sales (£ per store per week)"? It is not appropriate to switch outcomes in this manner.

11. The primary outcome of the study is clearly pre-specified in the protocol and repeated in the manuscript as "change in weekly sales (units)". In your manuscript you report your primary outcome (units sold) for most of Phase 1 but then use a different outcome "£ per store per week" for Phase 1 time series analyses and all Phase 2 analysis (including the very eye-catching Figures 2 and 3). Is there a reason that you have switched away from your pre-specified primary outcome for these analyses - I could not find a justification in the manuscript and again I really do not think that it is appropriate to switch outcomes in this manner without a very solid justification. It seems to me that you should reproduce Figures 2 and 3 (and the associated text) using your pre-specified primary outcome (units) and these should be the primary results that you report in this paper.

12. The opening paragraph of your discussion that reports your primary findings is currently hard to interpret without access to analysis reporting your pre-specified primary outcome (units).

Reviewer #3: This study aimed to evaluate the effect of positioning meat-free substitutes in the same aisle as the meat products. This study makes an important contribution to the literature. There are aspects of the manuscript that require further clarification and additional details. I have also outlined some methodological/presentation suggestions for the authors to consider.

Key points:

1. The intervention could be described in greater detail in the abstract, introduction and methods sections in terms of its relevance to existing literature. It is described as increasing the prominence of the positioning of meat-free products but it is unclear how this relates to existing literature which largely describes in-store prominence as being front of store, end-of-aisle and checkouts. The novelty of this study is the revised shelf positioning of the products but this point doesn't appear to be adequately explained in the manuscript, particularly the positioning of the meat-free products along the aisle in relation to the meat products (front, mid, end of aisle and if each meat-free alternative was co-located with the meat versions or whether the meat-free section was altogether). Also whether the positioning of the meat-free products was at eye-level, or bottom shelf or above eye-level. Detailing these positioning factors of the intervention in greater detail could help improve the interpretation of findings (i.e. why some meat-free products sold more than others and why meat product sales did not decline) and aid future intervention design.

2. This point links to that above in terms of description of numbers of products available. Greater availability/variety of products has been shown to increase sales of these products in previous literature. It would therefore be helpful for the reader to understand whether there was an increase in the availability/variety of meat-free products as part of the intervention or whether the same number of products were available at baseline, Phase I and Phase II, and how these numbers compared to the control stores. Additionally, what were the numbers/proportion of products available (both meat-free and meat) in the intervention position (both Phases) and in the pre-intervention and the control conditions; were these the same across all intervention stores? (Lines 121-126)

3. There is no description of the control condition in the methods section. Providing these details will enable the reader to better understand the intervention design, particularly the in-store and shelf positioning, and number, of meat-free and meat products available for both intervention conditions and the control condition. A full description of the control condition (and any variation across the 88 stores) should be added to the methods section. Also where were intervention and control stores located?

4. There are no sample size calculations provided for the experiment. The existing body of literature suffers from a lack of description of power calculations that account for clustering at the store level and this is an essential addition for new studies in this field, particularly for sizeable studies such as this one. It would also be helpful if the authors expanded on the reason for varying numbers of control stores; this is alluded to as a strength of the study in the discussion but no reference or clear rationale is provided.

Abstract/Introduction:

5. Line 74-78: These sentences could be more specific in terms of my key points 1 and 2 above and how this study relates to existing literature on store/shelf prominence and the marketing practices used by supermarkets. The reference to co-location of complementary products seems fundamental to the intervention design and it would be helpful if this was more clearly explained and referenced by similar studies (e.g. Foster 2014 AJCN and de Wijk 2016 Plos One - probably others from marketing literature).

Methods:

6. Line 94 - it would be helpful to additional details about the store with two weeks of missing data. Was it intervention/control store? When did the two weeks fall in terms of seasons and the intervention implementation? There is great variability in retail sales weeks so taking the week before and after may not always be the best approach. If it was a control store - would have imputing from an average of the other control stores been an alternative approach?

7. Lines 104-108; lines 158-161 - The authors conclusions about potential impact on inequalities is based on a measure that of socioeconomic status (SES) that may be unfamiliar to journal readers. Additional details about the affluence measure and how it relates to more typical measures of SES used by researchers would improve understanding and relevance of the associated findings. Additionally, the descriptions of the other variables to determining store matching and confounding variables could be improved to enable clear interpretation of these variables. (i.e. was is meant by meat-free sales performance - total, % sales, alternative? Can examples of store size be provided? What is average affluence? What is the cut-off for young and old? What is meant by average urban, more or less urban?)

8. Lines 109-117 - Use of a computerised matching system provides an objective methods for matching which is beneficial - though reliant on the supermarket rather than the research team. It is difficult to understand how the research matching linked to this matching through the supermarket store details system. Could this description be revisited and a rationale provided.

9. Lines 140-146 - would be better placed in another sub-section as they don't directly relate to the intervention description.

10. Related to key-points 1 and 2, it would be helpful if sensitivity analyses were conducted to assess differences in changes to availability of meat-free and meat products across intervention/control stores. (i.e. lines 123-125 state that meat-free replaced meat products suggesting potential decrease in meat product availability and if proportions available altered across stores this additional information about differential intervention effects would make a useful contribution to existing literature).

11. Lines 205-208 - could further details about the measures used to assess fidelity be outlined in this section and a table of the assessments be provided in the results section/supplementary materials? These additional details again provide a helpful addition to previous work.

Results:

12. Table 1 - could size of store be added here too? It would be interesting to view the differences in store size across the study stores.

13. Suggest including Supplementary Table 3 in the main body of the results because this contains important detailed information.

14. Figures 2 and 3 seem to be missing a key - difficult to interpret without this when in black and white print. Also, it would be helpful to see the longer-term effect of the intervention in the interrupted time series analyses. Could you test the effect again at time point Nov 2019 not just at the time of implementation?

Discussion:

15. Related to key-points 1 and 2, it would be helpful if the discussion reflected the potential differential effects of positioning and availability and how this relates to existing literature.

16. Lines 39-42 (page 20) these analyses are important to have conducted but the interpretation here seems a little inflated because low affluence stores were combined with mid affluence and these were taken at the store rather than household/individual level. I would recommend revising this sentence in light of these factors.

References:

17. References 17 and 18 appear to be the same reference. It would be beneficial for the reader if additional access details were added (i.e. weblink and date accessed if online or full details of author and publisher of only available in print).

Any attachments provided with reviews can be seen via the following link:

[LINK]

PLoS Med. 2021 Jul 15;18(7):e1003715. doi: 10.1371/journal.pmed.1003715.r003

Author response to Decision Letter 1

5 May 2021

Attachment

Submitted filename: Positioning Plos Med 1st revision.docx

Click here for additional data file.^{(42.6KB, docx)}

PLoS Med. doi: 10.1371/journal.pmed.1003715.r004

Decision Letter 2

Raffaella Bosurgi

24 Jun 2021

Dear Dr. Piernas,

Thank you very much for re-submitting your manuscript "Estimating the effect of moving meat-free products to the meat aisle on sales of meat and meat-free products: a non-randomised controlled intervention study in a large UK supermarket chain" (PMEDICINE-D-21-00493R2) for review by PLOS Medicine.

I have discussed the paper with my colleagues and the academic editor-Alan Dangour I am pleased to say that provided the remaining point raised by Alan Dangour and editorial/production issues are dealt with we are planning to accept the paper for publication in the journal.

Aland Dangour: For total clarity and scientific transparency I would like to make one further request. They have (as is good practice) noted with dates the amendments/updates they have made to their protocol (following our earlier review comments). However they have not done this for their statistical analysis plan (SAP). Given the importance of the SAP I think it would be appropriate for them to note (with dates) the changes they have made to this document during the latter stages of the study.

The remaining issues that need to be addressed are listed at the end of this email. Any accompanying reviewer attachments can be seen via the link below. Please take these into account before resubmitting your manuscript:

[LINK]

In revising the manuscript for further consideration here, please ensure you address the specific points made by each reviewer and the editors. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments and the changes you have made in the manuscript. Please submit a clean version of the paper as the main article file. A version with changes marked must also be uploaded as a marked up manuscript file.

Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. If you haven't already, we ask that you provide a short, non-technical Author Summary of your research to make findings accessible to a wide audience that includes both scientists and non-scientists. The Author Summary should immediately follow the Abstract in your revised manuscript. This text is subject to editorial change and should be distinct from the scientific abstract.

We expect to receive your revised manuscript within 1 week. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

We ask every co-author listed on the manuscript to fill in a contributing author statement. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT.

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript.

Please note, when your manuscript is accepted, an uncorrected proof of your manuscript will be published online ahead of the final version, unless you've already opted out via the online submission form. If, for any reason, you do not want an earlier version of your manuscript published online or are unsure if you have already indicated as such, please let the journal staff know immediately at plosmedicine@plos.org.

If you have any questions in the meantime, please contact me or the journal staff on plosmedicine@plos.org.

We look forward to receiving the revised manuscript by Jul 01 2021 11:59PM.

Sincerely,

Dr Raffaella Bosurgi,

Executive Editor

PLOS Medicine

plosmedicine.org

------------------------------------------------------------

Requests from Editors:

Comments from Reviewers:

Any attachments provided with reviews can be seen via the following link:

[LINK]

PLoS Med. 2021 Jul 15;18(7):e1003715. doi: 10.1371/journal.pmed.1003715.r005

Author response to Decision Letter 2

27 Jun 2021

Attachment

Submitted filename: Reviewer comments.docx

Click here for additional data file.^{(12KB, docx)}

PLoS Med. doi: 10.1371/journal.pmed.1003715.r006

Decision Letter 3

Raffaella Bosurgi

29 Jun 2021

Dear Dr Piernas,

On behalf of my colleagues and the Academic Editor, Alan Dangour , I am pleased to inform you that we have agreed to publish your manuscript "Estimating the effect of moving meat-free products to the meat aisle on sales of meat and meat-free products: a non-randomised controlled intervention study in a large UK supermarket chain" (PMEDICINE-D-21-00493R3) in PLOS Medicine.

Before your manuscript can be formally accepted you will need to complete some formatting changes, which you will receive in a follow up email. Please be aware that it may take several days for you to receive this email; during this time no action is required by you. Once you have received these formatting requests, please note that your manuscript will not be scheduled for publication until you have made the required changes.

In the meantime, please log into Editorial Manager at http://www.editorialmanager.com/pmedicine/, click the "Update My Information" link at the top of the page, and update your user information to ensure an efficient production process.

PRESS

We frequently collaborate with press offices. If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximise its impact. If the press office is planning to promote your findings, we would be grateful if they could coordinate with medicinepress@plos.org. If you have not yet opted out of the early version process, we ask that you notify us immediately of any press plans so that we may do so on your behalf.

We also ask that you take this opportunity to read our Embargo Policy regarding the discussion, promotion and media coverage of work that is yet to be published by PLOS. As your manuscript is not yet published, it is bound by the conditions of our Embargo Policy. Please be aware that this policy is in place both to ensure that any press coverage of your article is fully substantiated and to provide a direct link between such coverage and the published work. For full details of our Embargo Policy, please visit http://www.plos.org/about/media-inquiries/embargo-policy/.

Thank you again for submitting to PLOS Medicine. We look forward to publishing your paper.

Sincerely,

Dr Raffaella Bosurgi

Executive Editor

PLOS Medicine

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 Appendix. Study protocol.

(PDF)

Click here for additional data file.^{(1MB, pdf)}

S2 Appendix. Study statistical analysis plan.

(PDF)

Click here for additional data file.^{(902.5KB, pdf)}

S3 Appendix. Supporting information tables and figures.

(DOCX)

Click here for additional data file.^{(876.9KB, docx)}

Attachment

Submitted filename: Positioning Plos Med 1st revision.docx

Click here for additional data file.^{(42.6KB, docx)}

Attachment

Submitted filename: Reviewer comments.docx

Click here for additional data file.^{(12KB, docx)}

Data Availability Statement

[pmed.1003715.ref001] 1.York R, Gossard MH. Cross-national meat and fish consumption: exploring the effects of modernization and ecological context. Ecol Econ. 2004;48(3):293–302. doi: 10.1016/j.ecolecon.2003.10.009 [DOI] [Google Scholar]

[pmed.1003715.ref002] 2.Afshin A, Sur PJ, Fay KA, Cornaby L, Ferrara G, Salama JS, et al. Health effects of dietary risks in 195 countries, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet . doi: 10.1016/S0140-6736(19)30041-8 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1003715.ref003] 3.Key TJ, Appleby PN, Bradbury KE, Sweeting M, Wood A, Johansson I, et al. Consumption of Meat, Fish, Dairy Products, and Eggs and Risk of Ischemic Heart Disease: A Prospective Study of 7198 Incident Cases Among 409 885 Participants in the Pan-European EPIC Cohort. Circulation. 2019;139(25):2835–45. doi: 10.1161/CIRCULATIONAHA.118.038813 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1003715.ref004] 4.Steinfeld H, Gerber P, Wassenaar T, Castel V, Rosales M, Rosales M, et al. Livestock’s long shadow: environmental issues and options. Food & Agriculture Org.; 2006. [Google Scholar]

[pmed.1003715.ref005] 5.Bianchi F, Dorsel C, Garnett E, Aveyard P, Jebb SA. Interventions targeting conscious determinants of human behaviour to reduce the demand for meat: a systematic review with qualitative comparative analysis. Int J Behav Nutr Phys Act. 2018;15(1):102. doi: 10.1186/s12966-018-0729-6 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1003715.ref006] 6.Caspi CE, Sorensen G, Subramanian SV, Kawachi I. The local food environment and diet: a systematic review. Health Place. 2012;18(5):1172–87. Epub 2012/06/22. doi: 10.1016/j.healthplace.2012.05.006 ; PubMed Central PMCID: PMC3684395. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1003715.ref007] 7.Holsten JE. Obesity and the community food environment: a systematic review. Public Health Nutr. 2009;12(3):397–405. Epub 2008/05/15. doi: 10.1017/S1368980008002267 . [DOI] [PubMed] [Google Scholar]

[pmed.1003715.ref008] 8.Hollands GJ, Shemilt I, Marteau TM, Jebb SA, Kelly MP, Nakamura R, et al. Altering micro-environments to change population health behaviour: towards an evidence base for choice architecture interventions. BMC Public Health. 2013;13(1):1218. doi: 10.1186/1471-2458-13-1218 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1003715.ref009] 9.Bond MECB, Loff B. It’s not about choice: the supermarket and obesity. Med J Aust. 2012;197(7):371. doi: 10.5694/mja12.10923 [DOI] [PubMed] [Google Scholar]

[pmed.1003715.ref010] 10.The Obesity Health Alliance. Out of place: The extent of unhealthy foods in prime locations in supermarkets. 2018. [Google Scholar]

[pmed.1003715.ref011] 11.Department for Environment Food and Rural Affairs. Food Statistics Pocketbook 2014. London: 2015. 2015. Report No. [Google Scholar]

[pmed.1003715.ref012] 12.Bianchi F, Garnett E, Dorsel C, Aveyard P, Jebb SA. Restructuring physical micro-environments to reduce the demand for meat: a systematic review and qualitative comparative analysis. Lancet Planet Health. 2018;2(9):e384–e97. doi: 10.1016/S2542-5196(18)30188-8 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1003715.ref013] 13.Hollands GJ, Carter P, Anwer S, King SE, Jebb SA, Ogilvie D, et al. Altering the availability or proximity of food, alcohol, and tobacco products to change their selection and consumption. Cochrane Database Syst Rev. 2019;9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1003715.ref014] 14.Shaw SC, Ntani G, Baird J, Vogel CA. A systematic review of the influences of food store product placement on dietary-related outcomes. Nutrition Reviews. 2020. doi: 10.1093/nutrit/nuaa024 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1003715.ref015] 15.Cameron AJ, Charlton E, Ngan WW, Sacks G. A Systematic Review of the Effectiveness of Supermarket-Based Interventions Involving Product, Promotion, or Place on the Healthiness of Consumer Purchases. Curr Nutr Rep. 2016;5(3):129–38. doi: 10.1007/s13668-016-0172-8 [DOI] [Google Scholar]

[pmed.1003715.ref016] 16.Foster GD, Karpyn A, Wojtanowski AC, Davis E, Weiss S, Brensinger C, et al. Placement and promotion strategies to increase sales of healthier products in supermarkets in low-income, ethnically diverse neighborhoods: a randomized controlled trial. Am J Clin Nutr. 2014;99(6):1359–68. doi: 10.3945/ajcn.113.075572 [DOI] [PubMed] [Google Scholar]

[pmed.1003715.ref017] 17.Holloway T, Salter AM, McCullough FS. Dietary intervention to reduce meat intake by 50% in University students–a pilot study. Proceedings of the Nutrition Society. 2012;71(OCE2):E164. Epub 2012/10/19. doi: 10.1017/S0029665112002212 [DOI] [Google Scholar]

[pmed.1003715.ref018] 18.Acorn. The Acorn user guide: Consumer classification 2020. Available from: https://www.caci.co.uk/sites/default/files/resources/Acorn%20User%20Guide%202020.pdf.

[pmed.1003715.ref019] 19.Bernal JL, Cummins S, Gasparrini A. Interrupted time series regression for the evaluation of public health interventions: a tutorial. Int J Epidemiol. 2016;46(1):348–55. doi: 10.1093/ije/dyw098 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1003715.ref020] 20.Campbell-Arvai V, Arvai J, Kalof L. Motivating sustainable food choices: The role of nudges, value orientation, and information provision. Environ Behav. 2014;46(4):453–75. [Google Scholar]

[pmed.1003715.ref021] 21.Kongsbak I, Skov LR, Nielsen BK, Ahlmann FK, Schaldemose H, Atkinson L, et al. Increasing fruit and vegetable intake among male university students in an ad libitum buffet setting: A choice architectural nudge intervention. Food Qual Prefer. 2016;49:183–8. [Google Scholar]

[pmed.1003715.ref022] 22.Bates B, Colllins D, Cox L,Nicholosn S, Page P, Roberts C, Steer T, & Swan G. National Diet and Nutrition Survey Years 1 to 9 of the Rolling Programme (2008/2009-2016/2017): Time trend analysis and income analyses. London: Public Health England. 2019. [Google Scholar]

[pmed.1003715.ref023] 23.Godfray HCJ, Aveyard P, Garnett T, Hall JW, Key TJ, Lorimer J, et al. Meat consumption, health, and the environment. Science. 2018;361(6399). Epub 2018/07/22. doi: 10.1126/science.aam5324 . [DOI] [PubMed] [Google Scholar]

[pmed.1003715.ref024] 24.Chernev A. Product assortment and consumer choice: An interdisciplinary review. Foundations and Trends in Marketing. 2011;6(1):1–61. [Google Scholar]

[pmed.1003715.ref025] 25.Dalenberg JR, Nanetti L, Renken RJ, de Wijk RA, ter Horst GJ. Dealing with consumer differences in liking during repeated exposure to food; typical dynamics in rating behavior. PLoS ONE. 2014;9(3):e93350. doi: 10.1371/journal.pone.0093350 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1003715.ref026] 26.Adam A, Jensen JD, Sommer I, Hansen GL. Does shelf space management intervention have an effect on calorie turnover at supermarkets? J Retail Consum Serv. 2017;34:311–8. [Google Scholar]

[pmed.1003715.ref027] 27.Hoek AC, Luning PA, Weijzen P, Engels W, Kok FJ, de Graaf C. Replacement of meat by meat substitutes. A survey on person- and product-related factors in consumer acceptance. Appetite. 2011;56(3):662–73. doi: 10.1016/j.appet.2011.02.001 [DOI] [PubMed] [Google Scholar]

[pmed.1003715.ref028] 28.Apostolidis C, McLeay F. It’s not vegetarian, it’s meat-free! Meat eaters, meat reducers and vegetarians and the case of Quorn in the UK. Social Business. 2016;6(3):267–90. [Google Scholar]

[pmed.1003715.ref029] 29.Ejlerskov K, Sharp SJ, Stead M, Adamson AJ, White M, Adams J. Socio-economic and age variations in response to supermarket-led checkout food policies: a repeated measures analysis. Int J Behav Nutr Phys Act. 2018;15(1):125. doi: 10.1186/s12966-018-0755-4 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1003715.ref030] 30.Marteau TM, Hollands GJ. Fletcher PC. Changing human behavior to prevent disease: the importance of targeting automatic processes. Science. 2012;337(6101):1492–5. doi: 10.1126/science.1226918 [DOI] [PubMed] [Google Scholar]

[pmed.1003715.ref031] 31.Marteau TM. Towards environmentally sustainable human behaviour: targeting non-conscious and conscious processes for effective and acceptable policies. Philos Trans A Math Phys Eng Sci. 2017;375(2095):20160371. doi: 10.1098/rsta.2016.0371 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1003715.ref032] 32.Stoll-Kleemann S, Schmidt UJ. Reducing meat consumption in developed and transition countries to counter climate change and biodiversity loss: a review of influence factors. Reg Environ Chang. 2017;17(5):1261–77. [Google Scholar]

[pmed.1003715.ref033] 33.Hollands GJ, Bignardi G, Johnston M, Kelly MP, Ogilvie D, Petticrew M, et al. The TIPPME intervention typology for changing environments to change behaviour. Nat Hum Behav. 2017;1(8):0140. [Google Scholar]

[pmed.1003715.ref034] 34.Sunstein CR, Reisch LA. Behaviorally green: Why, which and when defaults can help. New Perspectives for Environmental Policies Through Behavioral Economics: Springer; 2016. p. 161–94. [Google Scholar]

[pmed.1003715.ref035] 35.Reynolds J, Archer S, Pilling M, Kenny M, Hollands GJ, Marteau T. Public acceptability of nudging and taxing to reduce consumption of alcohol, tobacco, and food: a population-based survey experiment. Soc Sci Med. 2019;236:112395. doi: 10.1016/j.socscimed.2019.112395 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1003715.ref036] 36.Willett W, Rockström J, Loken B, Springmann M, Lang T, Vermeulen S, et al. Food in the Anthropocene: the EAT–Lancet Commission on healthy diets from sustainable food systems. Lancet. 2019;393(10170):447–92. doi: 10.1016/S0140-6736(18)31788-4 [DOI] [PubMed] [Google Scholar]

PERMALINK

Estimating the effect of moving meat-free products to the meat aisle on sales of meat and meat-free products: A non-randomised controlled intervention study in a large UK supermarket chain

Carmen Piernas

Brian Cook

Richard Stevens

Cristina Stewart

Jennifer Hollowell

Peter Scarborough

Susan A Jebb

Roles

Abstract

Background

Methods and findings

Conclusions

Author summary

Why was this study done?

What did the researchers do and find?

What do these findings mean?

Introduction

Methods

Data source

Store selection and matching

Intervention

Outcome measures

Other measures

Statistical analysis

Fidelity evaluations

Results

Table 1. Demographic store characteristics relating to the surrounding population.

Fig 1. Average baseline sales (units/items sold) per store per week and comparison of changes between intervention and control stores over the Phase I intervention period.

Fig 2. Interrupted time series analysis showing level and trend changes in sales (units) of meat and meat-free products before and after Phase I intervention (week commencing February 3, 2019) in intervention stores (n = 20) and control stores (n = 88).

Fig 3. Interrupted time series analysis showing level and trend changes in sales (units) of meat and meat-free products before and after Phase II intervention period (week commencing July 28, 2019) in Phase I stores (n = 12) and Phase II stores (n = 8).

Fig 4. Average baseline sales (units/ items sold) per store per week and comparison of changes between intervention and control stores over the Phase I intervention period by store affluence.

Fidelity evaluations

Discussion

Supporting information

Data Availability

Funding Statement

References

Decision Letter 0

Raffaella Bosurgi

Roles

Decision Letter 1

Raffaella Bosurgi

Roles

Author response to Decision Letter 1

Decision Letter 2

Raffaella Bosurgi

Roles

Author response to Decision Letter 2

Decision Letter 3

Raffaella Bosurgi

Roles

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases