Food environments in Malta: Associations with store size and area-level deprivation

Food Policy 71 (2017) 39–47

Contents lists available at ScienceDirect

Food Policy journal homepage: www.elsevier.com/locate/foodpol

Food environments in Malta: Associations with store size and area-level deprivation Daniel Cauchi ⇑,1, Triantafyllos Pliakas, Cécile Knai London School of Hygiene and Tropical Medicine, 15-17 Tavistock Place, WC1H 9SH London, United Kingdom

a r t i c l e

i n f o

Article history: Received 30 March 2016 Received in revised form 4 June 2017 Accepted 12 July 2017

Keywords: Food environment Grocery store Childhood obesity NEMS-S Malta

a b s t r a c t Food environments are potential targets for interventions to reduce obesity prevalence, particularly in island settings that are typically dependent on food imports. This observational study aimed to characterise the availability, quality and price of foods and beverages in a nationally representative sample of grocery stores in Malta using the Nutrition Environment Measures Survey for Stores (NEMS-S) instrument, and to examine the association between area-level density of different types of food stores and the likelihood of children living in these areas being overweight or obese. Fieldwork was carried out between March and May 2014. There was a strong positive correlation between store size and NEMS-S score (p = <0.001), suggesting that smaller grocery stores generally offered a smaller range of products and fewer healthy food/beverage options than larger supermarkets. Across all stores, median prices of certain ‘healthier’ versions of foods were more expensive than their less healthy alternatives. A significant association between risk of childhood overweight, and density of confectionery stores in children’s locality of residence, was found (OR 1.19; 95% CI: 1.04, 1.37). These baseline findings highlight opportunities to improve the food environment in Malta to support more healthful eating, and may be of particular interest to public health practitioners in island settings. Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction Malta has one of the most obese populations worldwide (Ng et al., 2014). A recent national anthropometric survey which measured the height and weight of all school-aged children (up to 16 years of age) in Malta concluded that approximately 40% are overweight or obese according to WHO criteria (Grech et al., 2016). Maltese adults are also among the most overweight within the WHO European region (Verschuuren et al., 2015), with 2008 self-reported survey data suggesting that 58.5% were overweight (BMI > 25 kg/m2) and 22.3% were obese (BMI > 30 kg/m2) (Health Information and Research Directorate, 2008). Given the potential impact of the food environment – referring to the circumstances surrounding the procurement and consumption of food (Glanz et al., 2005) – on population weight (Chow et al., 2009; Holsten, 2009), measuring the nature of local food environments is important to determine what is available for purchase and develop obesity prevention strategies (Cerin et al., 2011; Cobb et al., 2015; ⇑ Corresponding author at: LG13, LSHTM, 15-17 Tavistock Place, London WC1H 9SH, United Kingdom. E-mail addresses: [email protected], [email protected] (D. Cauchi). 1 Present address: Health Promotion and Disease Prevention Directorate, 5B, The Emporium, C. de Brocktorff Street, Msida, Malta. http://dx.doi.org/10.1016/j.foodpol.2017.07.004 0306-9192/Ó 2017 Elsevier Ltd. All rights reserved.

Kelly et al., 2011). For example, greater availability of grocery stores may improve overall dietary quality and lower obesity prevalence (Morland et al., 2006; Powell et al., 2007). Conversely, a high density of outlets providing less healthy food such as fast food restaurants and convenience stores has been associated with increased obesity (Jeffery et al., 2006). Marketing studies clearly show that the amount of shelf space, together with the location and number of displays allocated to food products within a store, are integral elements of the consumer food environment which have a significant impact on sales and potential consumption (Briesch et al., 2009; Curhan, 1972; Glanz et al., 2012; Inman et al., 2009; Wilkinson et al., 1982). The relationship between food availability, dietary habits, and weight status has not been clearly established, however availability and price of foods are likely to influence dietary behaviour and eventual risk of developing obesity (Cobb et al., 2015; Holsten, 2009; Mackenbach et al., 2014). This is also the case for children and adolescents, with a number of studies showing at least one positive association between food environment exposure and children’s dietary outcomes (Engler-Stringer et al., 2014). Close proximity to convenience stores and a high density of fast-food outlets in adolescents’ home and school environments has been associated with unhealthy dietary behaviour (He et al., 2012).

40

D. Cauchi et al. / Food Policy 71 (2017) 39–47

Local food environments may also be related to neighbourhood socioeconomic level (Larson et al., 2009; Wang et al., 2007), as affluent areas may have better access and greater availability of healthy food compared to deprived neighbourhoods (Moore and Diez Roux, 2006). Objective documentation of potentially obesogenic food environments permits a more comprehensive understanding of potential dietary influences, consistent with an ecological approach (Egger and Swinburn, 1997). In turn, environmental interventions aimed at facilitating healthier food choices through the creation of supportive food environments may achieve more effective results than behavioural interventions that require significant personal motivation (Caspi et al., 2012; Story et al., 2008). This is one of few studies to review an ‘island food environment’. Pacific island research indicates that islands may be disproportionately influenced by the global food system due to their reliance on imported foods (Snowdon and Thow, 2013). Malta is similarly reliant on food imports (Cauchi et al., 2015), hence systematic documentation and assessment of local food environments may offer important insight into their potential effect upon food purchasing and dietary behaviour. Furthermore, research on the food environment in Malta has been specifically recommended as part of a national obesity strategy (Superintendence of Public Health, 2012). The aim of this study was to systematically document the food environment in a representative sample of localities in the Maltese islands. We characterised the environment in grocery stores – the most common outlets for food purchases (National Statistics Office (Malta), 2010) – and tested for differences by area-level deprivation. We also examined associations between area-level density of different types of food stores and the likelihood of overweight or obesity in children living in these areas, hypothesising that the density of stores which typically sell energy-dense food and beverages (e.g. fast food; sweets, soft drinks etc.) in a particular locality would be positively correlated with risk of being overweight or obese for children residing in that locality.

2. Method The ‘minimal’ approach proposed by the International Network for Food and Obesity/non-communicable diseases Research, Monitoring and Action Support (INFORMAS) network was adopted (Lee et al., 2013; Ni Mhurchu et al., 2013). INFORMAS proposes a flexible approach towards the benchmarking of retail food environments that allows participating countries to monitor different indicators depending on local resources and capacity (Swinburn et al., 2013). The minimal approach involves monitoring availability and accessibility of key foods and beverages in one key type of food outlet for which evidence exists of an association with dietary outcomes, in a limited number of locations/areas. The foods to be monitored should link clearly with risk of obesity and NCDs, including healthy items such as fruits and vegetables; and unhealthy items such as sugar-sweetened beverages, energydense nutrient-poor foods (e.g. confectionery), and fast foods. Final selection of specific foods to be monitored should be based on local food consumption data. A contextual analysis (Cauchi et al., 2015) was first carried out to determine the local context, as recommended by INFORMAS. Grocery stores were identified as the food outlet type accounting for >66% of food purchases in Malta (National Statistics Office (Malta), 2010). Consequently, two food environment sub-types, including the community (i.e. number, type, location, and accessibility of different types of food outlets in an area) and consumer (i.e. conditions that individuals encounter within grocery stores, including nutritional quality, price,

promotion, and placement) food environments were selected for monitoring and assessment. 2.1. Sampling strategy The ‘minimal’ INFORMAS approach proposes collection of data from representative locations within a country, however, no specific sampling strategy is recommended. A review of the literature was therefore carried out to identify validated environmental audit instruments that can be applied to Malta. The tool developed by the ‘Consortium for the PREVention of OBesity through effective nutrition and physical activity actions’ (EURO-PREVOB) was deemed to be the most appropriate. This instrument objectively assesses multiple environmental factors, focuses specifically on the obesogenicity of communities, and has been carried out in diverse European countries (Pomerleau et al., 2012). Pilot testing indicated that the instrument was suitable for the Maltese context. Following the validated environmental audit methods outlined by Pomerleau et al., all localities/local councils (n = 68) in Malta were ranked by median socioeconomic (SE) deprivation and stratified into quintiles. A diverse representative sample of ten localities was randomly selected, with two localities in each quintile undergoing a field audit. Quintile one (Q1) represents the most affluent localities, whereas quintile five (Q5) refers to the most deprived localities. Google EarthTM was used to plan walking routes in areas of around 0.5 km2 (range: 0.3–0.7 km2) in each locality, as recommended by EURO-PREVOB (Pomerleau et al., 2012) (Appendix 1). Each area encompassed the densest residential neighbourhoods within each locality, and in most cases contained a primary and/ or secondary school. Interviews with secondary school students living in these areas were carried out as part of a larger PhD project. Only stores in these areas were assessed. 2.1.1. Community food environment The location and frequency of all food stores within each audited area were recorded. Stores were categorized based on adapted, pre-existing food outlet classification criteria (Lake et al., 2010) (Appendix 2). Density (per km2) of different store types in the audited areas was calculated and categorised into low, medium or high density, based on the observed distribution of each store type across all ten audited areas. Additionally, a database of all food outlets licensed up to January 2014 was obtained from the Environmental Health Directorate (Environmental Health Directorate, 2014). The total surface area occupied by each locality was calculated using Earth Point for Google Earth (http://www. earthpoint.us/Shapes.aspx). The Environmental Health Directorate database was then used to derive the overall density of different food store types in the ten localities, including those areas which were not audited. 2.1.2. Consumer food environment Between March and May 2014, one small (1 checkout), one medium (2–4 checkouts) and one large (5 checkouts) grocery store was randomly selected in each locality to undergo consumer food environment assessment (n = 30) as recommended by the EURO-PREVOB consortium (Pomerleau et al., 2012). This ensured that a representative sample of grocery stores was obtained, given the limited resources available to the authors. The number of checkouts was used as a proxy for store size (Glanz et al., 2007; Horowitz et al., 2004; Kerr et al., 2012). For the purposes of this survey, the term ‘grocery stores’ referred to food outlets selling fresh fruit and vegetables (F&V) alone or in addition to other food products (Glanz and Yaroch, 2004), including specialized green grocers, discount stores and supermarkets. Data were collected through in-person visits by DC, assisted by a fieldworker.

D. Cauchi et al. / Food Policy 71 (2017) 39–47

41

2.2. Screening instruments The Nutrition environment Measures Survey for Stores (NEMSS) instrument to measure the consumer environment in grocery stores was adapted for use in Malta (Glanz et al., 2007). NEMS-S is a widely disseminated tool having high reliability and validity (Glanz et al., 2007) that has been adapted for use in several countries (Martins et al., 2013; Vilaro and Barnett, 2013). It allows for comparisons between ‘healthy’ and ‘regular’ (i.e. ‘less healthy’) versions of foods to be made, as recommended by INFORMAS. Definitions of what is ‘healthy’ were based on existing literature, including considerations of fat and sugar levels per 100 g of food (Glanz et al., 2007; Mullis et al., 1990). One of the authors completed an online training course in the use and adaptation of NEMS-S (Honeycutt et al., 2010). Following consultation with nutritionists at the Ministry for Health in Malta and taking into account local literature on affordable food baskets (McKay et al., 2012), elements of the EURO-PREVOB tool (Pomerleau et al., 2012) and INFORMAS food pricing module (Ni Mhurchu et al., 2013) were added to the NEMS-S Malta instrument (Fig. 1). This was also pilot-tested and revised further to more accurately reflect the Maltese context. For example, the original list of NEMS-S items was expanded to include additional categories including canned and frozen F&V, rice and pasta, all of which are very frequently consumed in Malta according to food consumption data (Malta Standards Authority, 2010). Items not typically available in Malta, such as cantaloupes and corn, were substituted by more common items in their respective categories, such as plums and potatoes. Onions and zucchini were added to the ‘fresh vegetables’ category; cheese and yogurt were added to the ‘dairy’ category; representative brand names of packaged items such as juices, breads and crisps were modified to reflect locally available brands, and USC volumes of liquids (e.g. gallons or ounces of milk, juice and sugar sweetened beverages) were changed to reflect typical volumes in the metric system. The NEMS-S scoring system was adapted to reflect these modifications. The final NEMS-S (Malta) instrument contains thirteen food and beverage categories. The most common brands were assessed whenever available. Quality of F&V was assessed based on appearance (i.e. bruising or over-ripeness), and the cheapest price for each category was collected as Euros (€) per kg or Litre. Shelf space allocated to milk was also assessed. NEMS-S composite scores were then calculated for food availability, quality and price. A lower price for a healthier item was scored positively, whereas a higher price subtracts a point from the price dimension. Higher scores denote greater availability of healthier options; better or equal price for healthy choices; and higher F&V quality. 2.3. Cohort BMI data The Malta School Health Service provided measured anthropometric data for three national cohorts of primary schoolchildren with a mean age of 8.3 years (SD ± 1.3), collected in 2008, 2010 and 2013 (Sant’Angelo, V., personal communication). Body mass index (BMI) percentile-for-age equivalents was calculated for 2623 children residing in the ten audited localities using International Obesity Task Force (IOTF) cut-offs (Cole and Lobstein, 2012). BMI > 25 kg/m2 corresponds roughly to the 90th centile, whereas BMI > 30 kg/m2 approximately corresponds to the 99th centile. 2.4. Data analysis Analysis was undertaken using SPSS v.23 and STATATM 14.1. NEMS-S scores, ranges and mean standard deviations were calculated for grocery stores. Non-parametric analyses of data were

Fig. 1. Adaptation of the NEMS-S instrument to the Maltese context.

performed as preliminary analysis indicated that the data were not normally distributed. These data were further stratified by store size and by area-level SE deprivation, and Spearman’s coefficient and Chi-Square independence tests for association were used to explore associations between all variables. Kruskal-Wallis tests were performed in order to test whether there is a socioeconomic gradient to the availability, price or quality of foods across quintiles, with the underlying hypothesis that in more deprived areas ‘healthy’ foods will be more expensive and ‘unhealthy’ items cheaper. In addition, Wilcoxon Rank-Sum tests were conducted so that price comparisons between ‘healthy’ and ‘regular’ (i.e. ‘less healthy’) versions of the same food item could be made. Post hoc pairwise comparisons were performed using Dunn’s procedure with a Bonferroni correction. Adjusted p-values are presented; values of <0.05 are considered statistically significant. Logistic regression was also applied to investigate whether density of different food store types in the localities where children participating in the cohort study lived, was associated with risk of being overweight or obese in children. Since home address data were unavailable, it remains possible that children may have resided at the edge, or outside of, the audited areas within each locality, which has implications for their potential exposure to the community food environment. In order to minimise bias for each locality, the density of all food stores licensed by the Environmental Health Directorate in each locality and therefore sited within in each locality’s total surface area, (including areas that were not audited), was calculated and used in the analysis. Clustering of individuals within localities was taken into account using robust standard errors.

3. Results 3.1. Differences in store density, NEMS-S scores and food prices by SE quintile and store size In all localities, only 1 large grocery store – typically a large supermarket – was identified per locality, often located at the outskirts of the audited area. While few medium sized grocery stores were identified, the absolute number of small grocery stores across all localities varied, however there were no statistically significant differences in the densities of any grocery store size between quintiles. This distribution of grocery stores, with two or three medium to large grocery stores in each locality supplemented by a varying number of small grocery stores, is likely to be typical of localities in Malta. Table 1 shows descriptive statistics for density of food store

42

D. Cauchi et al. / Food Policy 71 (2017) 39–47

Table 1 Food store density (in km2) and NEMS-S scores by area-level SE quintile (M = Mean ± SD; Md = Median; R = Range).

Store type Small grocery stores Medium grocery stores Large grocery stores All food stores Supermarkets Takeaways Bars Restaurants or cafeterias Mobile vendors NEMS-S scores Availability

Price

Quality

Total NEMS score

Q1

Q2

Q3

Q4

Q5

Difference (p value)

M: 3.0 ± 0.85 R: 2.4–3.6 M: 1.2 ± 0.8 R: 0.6–1.8 M: 0.6 ± 0.0 R: 0.6–0.6 M: 27.50 ± 27.1 R: 8.3–46.7 M: 18.33 ± 11.8 R: 10–26.7 M: 4.17 ± 3.5 R: 1.7–6.7 M: 4.17 ± 5.5 R: 1.7–6.7 M: 49.17 ± 53.0 R: 11.7–86.7 M: 2.50 ± 1.2 R: 1.7–3.3

M: 3.3 ± 0.42 R: 3.0–3.6 M: 0.6 ± 0.0 R: 0.6–0.6 M: 0.6 ± 0.0 R: 0.6–0.6 M: 21.33 ± 7.5 R: 16.0–26.7 M: 16.33 ± 0.5 R: 16.0–16.7 M: 15.33 ± 1.9 R: 14.0–16.7 M: 16.00 ± 5.7 R: 12.0 – 20.0 M: 40.67 ± 24.5 R: 23.3–58.0 M: 9.67 ± 5.2 R: 6.0–13.3

M: 7.2 ± 0.0 R: 7.2 – 7.2 M: 0.6 ± 0.0 R: 0.6–0.6 M: 0.6 ± 0.0 R: 0.6–0.6 M: 44.57 ± 10.5 R: 37.1–52.0 M: 18.42 ± 7.9 R: 12.9–24.0 M: 12.85 ± 10.1 R: 5.7–20.0 M: 14.42 ± 2.2 R: 12.9–16.0 M: 1.43 ± 2.0 R: 0.0–2.9 M: 7.71 ± 8.9 R: 1.4–14.0

M: 4.2 ± 3.40 R: 1.8–6.6 M: 0.6 ± 0.0 R: 0.6–0.6 M: 0.6 ± 0.0 R: 0.6–0.6 M: 35.33 ± 16.0 R: 24.0–46.7 M: 17.00 ± 9. 9 R: 10.0–24.0 M: 12.33 ± 8.0 R: 6.7–18.0 M: 6.33 ± 0.57 R: 6.0–6.7 M: 3.00 ± 4.2 R:0.0–6.0 M: 9.67 ± 8.9 R: 3.3–16.0

M: 5.4 ± 0.0 R: 5.4–5.4 M: 0.6 ± 0.0 R: 0.6–0.6 M: 0.6 ± 0.0 R: 0.6–0.6 M: 45.67 ± 24.9 R: 28.0–63.3 M: 22.33 ± 6.1 R: 18.0–26.7 M: 12.33 ± 6.1 R: 8.0–16.7 M: 35.66 ± 3.3 R: 33.3–38.0 M: 29.66 ± 19.3 R: 16.0–43.3 M: 5.33 ± 19 R: 4.0–6.7

0.214

M: 15.0 ± 3.2 Md: 15.5 R: 10–19 M: 1.8 ± 3.1 Md: 3.5 R: 3 to 4 M:6.0 ± 0 Md: 6.0 R: 6–6 M: 31.7 ± 5.0 Md: 29.5 R: 27–39

M: 14.3 ± 4.6 Md: 14.5 R: 8–19 M: 1.0 ± 1.3 Md: 0.5 R: 0–3 M: 5.3 ± 0.8 Md: 5.5 R: 4–6 M: 27.7 ± 7.7 Md: 28.5 R: 17–36

M: 12.5 ± 4.9 Md: 11.0 R: 7–19 M: 1.7 ± 1.9 Md: 2.0 R: 1 to 4 M: 5.7 ± 0.5 Md: 6.0 R: 5–6 M: 26.7 ± 9.9 Md: 25.0 R: 13–39

M: 16.2 ± 4.9 Md: 15.0 R: 10–23 M: 0.5 ± 2.6 Md: 1.5 R: 3 to 4 M: 5.5 ± 0.8 Md: 6.0 R: 4–6 M: 30.5 ± 8.1 Md: 28.0 R: 22–41

M: 13.3 ± 5.4 Md: 14.0 R: 6–19 M: 0 ± 2.1 Md: 0.5 R: 3 to 2 M: 5.8 ± 0.4 Md: 6.0 R: 5–6 M: 26.7 ± 8.3 Md: 30.5 R: 13–33

types and grocery store NEMS-S scores by SE quintile. There were no statistically significant differences between means for density of any food store type, and no significant differences in median NEMS scores, food prices, food quality, shelf space and variety of options available for specific food items, between SE quintiles, suggesting an absence of a socioeconomic gradient with regards to density of food stores or the healthfulness of grocery stores. However, there was a strong positive correlation between store size and NEMS-S total (rs(28) = 0.778; p = <0.001) and availability scores (rs(28) = 0.787; p = <0.001), indicating more healthful consumer food environments in larger stores. Differences in mean NEMS-S scores by grocery store size emerged (Table 2), including statistically significant differences in mean NEMS-S availability scores (v2(2) = 17.542, p = <0.001) and mean NEMS-S total score (v2(2) = 18.202, p = <0.001). Post-hoc analysis revealed statistically significant differences in mean NEMS-S total scores between the small and medium (p = 0.032), and small and large grocery stores (p = <0.001), but not between the medium and large stores. Additionally, there was a statistically significant difference in mean NEMS-S availability score between the small and large stores (p = <0.001), but not between any other group combination. These findings suggest that in Malta, area-level socioeconomic deprivation has minimal impact on food availability. Instead, it is the size of grocery stores that influences the ‘healthfulness’ of the community food environment. As expected, medium and large grocery stores differ substantially from smaller grocery stores in terms of availability and variety of food on offer.

3.2. Differences in availability, number of options, and food quality by store size and SE quintile No significant associations between the availability of any food item and SE quintile were observed. Larger store size was

0.406 1.00 0.480 0.812 0.561 0.094 0.151 0.626

0.730

0.340

0.210

0.810

moderately to strongly associated with greater choice (i.e. more options available for a single product) of several foods including fresh vegetables (p = 0.047); wholegrain rice (p = 0.003); ‘healthy’ cereals (p = <0.001); canned vegetables (p = 0.004); chocolate bars (p = <0.001) and packaged extruded snacks (p = 0.004). Furthermore, there were moderate to strong associations between larger store size and availability of baked crisps (p = 0.036); wholemeal pasta (p = 0.010); brown rice (p = 0.014); wholemeal bread (p = 0.018); diet Cola (p = 0.036); ’healthy’ cereal (p = 0.002); lean beef mince (p = <0.001); low fat baked goods (p = 0.007); and low fat cheddar cheese (p = 0.001). Baked crisps were consistently absent from small grocery stores. Larger stores offered significantly more healthy options than small grocery stores. Fresh F&V quality was high overall, hence this variable was excluded from further analysis. While median shelf space allocated to whole or skimmed milk did not differ significantly across SE quintiles, larger stores offered significantly more shelf space to whole milk than smaller stores (v2(2) = 13.363; p = 0.001). Differences in median shelf space allocated to skimmed milk (v2(2) = 8.937; p = 0.011), were significant only between the small and large (p = 0.010) stores. Additional food variable data are presented in Appendix 3.

3.3. Relationships between food prices, SE quintiles and store size Relatively few associations were identified between SE quintile or store size and food prices. A moderately positive Spearman‘s correlation between SE quintile and the median unit price of grapes (p = 0.007) and plums (p = 0.018) was found. Increasing deprivation was also moderately associated with a reduction in the median price of cabbage (p = 0.007); 1.5 L diet (p = 0.020) and regular (p = 0.031) Cola drink bottles; and 100% orange juice (p = 0.050).

43

D. Cauchi et al. / Food Policy 71 (2017) 39–47 Table 2 NEMS-S scores by grocery store size (M = Mean ± SD; R = Range). Store size

Availability Price Quality Total score

Small

Medium

Large

Diff. (p value)

M: 10.0 ± 3.3 R: 6–17 M: 0.9 ± 2.3 R: 3 to 4 M: 5.7 ± 0.7 R: 4–6 M: 21 ± 5.5 R: 13–28

M: 14.3 ± 2.6 R: 10–19 M: 1.0 ± 2.3 R: 3 to 4 M: 5.6 ± 0.7 R: 4–6 M: 29.7 ± 4.2 R: 24–37

M: 18.5 ± 2.9 R: 12–23 M: 0.5 ± 2.6 R: 3 to 4 M: 5.7 ± 0.5 R: 5–6 M: 35.2 ± 5.4 R: 23–41

<0.001

There were moderate to strong positive associations between store size and the price of watermelon (p = 0.007); plums (p = 0.035); peaches (p = 0.043); apples (p = 0.046) and broccoli (p = 0.019); and negative correlations between store size and the unit prices of regular beef mince (p = 0.001); regular cheddar cheese (p = 0.001) and 1.5 L bottles of regular Cola drink (p = 0.012).

3.4. Differences in food prices by store size, and between ‘healthy’ and ‘less healthy’ food pairs Table 3 shows food items which exhibited significant differences in unit price by store size, whereas Table 4 illustrates differences in median unit price between pairs of ‘healthy’ food items and their regular or ‘less healthy’ alternatives across all stores. These pairs of items were recommended by INFORMAS and/or EURO-PREVOB as indicators that typically highlight the existence of a price premium in order to purchase the ‘healthy’ version of the food or beverage item.

3.5. Area-level food store density and risk of overweight or obesity BMI percentile-for-age data for 2623 children indicated that 34% (n = 888) were overweight or obese according to IOTF criteria. Risk of overweight or obesity was modelled against area-level density of different food store types in their locality of origin after adjusting for SE quintile and clustering. Table 5 shows results for selected store types having the highest density across audited localities. Only confectionery store density was significantly associated with a higher risk of being overweight or obese (OR 1.19; 95% CI: = 1.04, 1.37; p = 0.013).

Table 3 Chi-Square independence tests for food items exhibiting a significant difference in unit price by store size. Food item (unit)

Chi-square test

Pairwise comparisons (median price in €)

Cheddar cheese, regular (kg) Cola drink, regular (1.5 L bottle) Vegetable mix, frozen (kg) Beef mince, regular (kg)

v2(2) = 9.320,

Strawberries (kg)

v2(2) = 6.795,

Watermelon (kg)

v2(2) = 6.396,

Difference between small and large stores only (p = 0.009) Difference between small and large stores only (p = 0.033) Difference between small and large stores only (p = 0.018) Difference between the small and large (p = 0.034); and small and medium (p = 0.009) stores No significant pairwise comparisons observed Difference between small and large stores only (p = 0.035)

p = 0.009 v2(2) = 6.722, p = 0.035 v2(2) = 7.666, p = 0.022 v2(2) = 10.96, p = 0.004

p = 0.033 p = 0.041

0.860 0.870 <0.001

4. Discussion This study aimed to explore intersections between the community and consumer food environments in Malta. Our findings support research suggesting that residents of deprived areas may not necessarily pay more for healthy food compared to residents of more affluent areas (Cummins and Macintyre, 2002; Winkler et al., 2006), although this should not negate the disproportionate impact of food expenditure on household budgets of less affluent consumers seeking diets of high nutritional quality (Darmon and Drewnowski, 2015). Our results also support studies indicating that the size of stores where consumers shop has a greater impact on food prices and availability than area-level deprivation (Cummins et al., 2010; Laska et al., 2010). Larger grocery stores provided a substantially more ‘healthful’ overall consumer environment than smaller stores, having greater availability of healthier items at often cheaper prices. With the exception of some seasonal fruit, the price of most F&V did not differ significantly by store size or deprivation. Quality of fresh produce is an additional structural factor that might influence purchasing behaviour (Cummins et al., 2009), however the quality of fresh F&V in this study was high overall. Substituting unhealthy food with healthier alternatives (e.g. replacing sugar-sweetened drinks with low-calorie alternatives) is frequently proposed as a way of improving dietary behaviour and reducing population weight levels, and people are more likely to buy healthy foods when they cost less (Horgen and Brownell, 2002). However, the evidence suggests that low energy-dense diets of high nutritional quality tend to cost more than less nutritious, energy-dense diets (Drewnowski and Specter, 2004); that healthy foods are perceived to be expensive (Kearney and McElhone, 1999); and that price is more important than nutritional quality when making food choices (Glanz et al., 1998). Although no systematic research in this regard has yet been carried out in Malta, it is likely that the same principles apply to the Maltese consumer. Our study partially supports the belief that choosing the healthier option for certain frequently-consumed foods such as pasta, bread, beef and cheese (Malta Standards Authority, 2010) might prove to be a financial burden for consumers. Median prices of healthier versions of these products (i.e. wholegrain pasta, wholemeal bread, lean beef mince and low-fat cheddar cheese) were significantly higher than the prices of their regular counterparts (i.e. white pasta, white bread, regular beef mince, regular cheddar cheese). This represents a potential disincentive to substitute regular versions of such foods for healthier alternatives. On the other hand, regular breakfast cereal and whole milk – popular breakfast options among the Maltese (Malta Standards Authority, 2010) – were significantly more expensive than low-sugar, high-fibre cereal and skimmed milk. Thus, substitution in the case of cereal-based breakfasts could make financial sense. In addition, the median paired prices of plain yoghurt, juice products, rice and sausages did not differ significantly, suggesting that simple

44

D. Cauchi et al. / Food Policy 71 (2017) 39–47

Table 4 Differences in median unit price of ‘healthy’ and ‘less healthy’ versions of select paired food products. Item name (unit)

Healthy (€/unit)

Regular (€/unit)

Difference (€/unit)

P value

Cereal: unsweetened vs sweetened (kg) Pasta: wholemeal vs white (kg) Rice: brown vs white (kg) Beef mince: lean vs regular (kg) Cheese, cheddar: low-fat vs regular (kg) Bread: brown vs white (400 g loaf) Sausages: low fat vs regular (kg) Yoghurt, plain: fat-free vs regular (kg) Coca-Cola: diet vs regular (1.5 L bottle) Juice: 100% juice vs juice drink (1 L carton) Milk: skimmed vs whole (1 L carton)

7.74 2.37 3.85 7.95 10.9 0.95 3.8 1.73 1.5 1.29 0.81

8.85 1.67 3.64 7.1 7.23 0.85 4.81 2.2 1.5 1.3 0.86

0.42 0.71 – 1 4.1 0.15 – – – 0.01 0.05

0.001 <0.001 0.146 0.028 0.004 <0.001 0.581 0.257 0.655 0.016 <0.001

Table 5 Unadjusted and adjusted (for area-level deprivation) odds ratios for being overweight or obese by area-level food store density. Area density

Unadjusted (95% C.I.) p-value

Adjusted (95% C.I.) p-value

All stores selling fast fooda Low (reference) 1 Medium 0.85 (0.70, 1.04); 0.110 High 0.94 (0.77, 1.14); 0.530

1 0.89 (0.68, 1.16); 0.372 0.97 (0.73, 1.31); 0.861

All stores selling fruit and vegetablesb Low (reference) 1 Medium 1.03 (0.86, 1.24); 0.734 High 1.08 (0.81, 1.44); 0.614

1 1.04 (0.85, 1.29); 0.689 0.94 (0.82, 1.09); 0.441

Small supermarket Low (reference) 1 Medium 0.99 (0.81, 1.22); 0.942 High 1.19 (0.96, 1.46); 0.108

1 1.01 (0.8, 1.26); 0.948 1.10 (0.95, 1.27); 0.190

Confectionery store Low (reference) 1 Medium 1.12 (0.89, 1.41); 0.328 High 1.15 (0.97, 1.36); 0.112

1 1.05 (0.96, 1.14); 0.275 1.19 (1.04, 1.37); 0.013

Convenience store Low (reference) 1 Medium 1.15 (0.91, 1.45); 0.230 High 1.06 (0.84, 1.35); 0.601

1 1.17 (0.92, 1.48); 0.205 1.09 (0.90, 1.31); 0.372

a Includes all stores offering takeaway including: takeaway outlets; bars where snack food is served (i.e. snack bars); fast food restaurants; mobile vendors selling fast food; and restaurants with takeaway option. b Includes green grocers; mobile vendors selling F&V; small discount stores and large supermarkets.

substitution of regular with healthier versions of these items would not pose a disproportionate financial burden upon consumers. The median price of frozen F&V did not vary by store size or deprivation quintile, hence substituting fresh F&V with their generally cheaper frozen and packaged alternatives may be one way of minimizing price premiums without comprising the nutritional profile of one’s diet. Such small changes could accumulate to have a potentially significant overall impact on diet, provided that substantial investment in health literacy and nutrition knowledge is made to empower consumers to make healthier choices (Superintendence of Public Health, 2012). The breakfast cereal, pasta and cheese brands assessed in this study are imported, whereas most of the bread, yogurt, milk and beef mince brands are locally manufactured. Coca-Cola beverages and the fruit juice brand assessed in this study are also locally manufactured. This means that an opportunity exists to revise the prices of certain locally manufactured, commonly consumed foods and beverage items in such a way as to dis-incentivise consumption of their regular, ‘less healthy’ versions and promote consumption of low-fat or low-sugar alternatives. With regards to soft drinks, diet and sugarsweetened Coca-Cola were set at the same price point in each store, obviating any financial incentive to purchase the low calorie

version. Taste preference and nutritional knowledge thus remain the primary reasons for selecting one option over the other. Our results also indicate that 1.5 L bottles of diet and sugarsweetened Cola drinks are cheaper in areas of greater deprivation and in larger stores: this finding warrants further investigation and might indicate that these items are positioned as ‘loss leaders’, being purposely under-priced in order to attract customers into stores (Nestle, 2015). A higher density of confectionery stores – which mainly sell sweets, pastries and chocolate – is associated with increased odds of being overweight or obese for children living in these localities. Interestingly, the density of other stores popularly thought to have protective (e.g. F&V stores) or harmful effects (e.g. fast food outlets or convenience stores), particularly within a local context that often blames fast food consumption for Malta’s high obesity prevalence (Health Promotion and Disease Prevention Directorate, 2015), did not show any significant association. The literature in this regard is inconsistent, as the relationship between food availability, dietary habits, and weight status has not been clearly established (Cobb et al., 2015; Holsten, 2009; Mackenbach et al., 2014). However, sales of unhealthy foods in supermarkets have been associated with greater overweight and obesity prevalence outside of Malta, particularly in areas of greater SE deprivation (Howard Wilsher et al., 2016). 4.1. Policy implications Increasing awareness that interventions targeting individual activity and dietary behaviours have limited effectiveness has led to the recognition that broader environmental changes may be driving the obesity epidemic, and to calls for a multi-level environmental approach to obesity interventions (Swinburn et al., 2011). There is a nascent debate around the feasibility of introducing taxation of unhealthy foods and beverages in Malta (Superintendence of Public Health, 2012) – fiscal policies have been shown to be effective elsewhere (Powell et al., 2013), and are further supported by encouraging preliminary reports of reductions in the sale of taxed beverages following the imposition of an excise tax on sugar sweetened beverages in Mexico (Colchero et al., 2016). Currently, a price premium for healthy versions of certain food categories exists in Malta, with only limited improvements in food choices possible without increasing overall food cost for the consumer. However, it is encouraging to note that some of these price premiums apply to popular, locally manufactured food items. Policies designed to encourage local production and sale of healthy foods at low prices may help to counteract the disadvantage of relying on imported products whose price is difficult to control. Such an approach is likely to require the design and implementation of complementary informational policies that make it easier for people to distinguish between healthier and less healthy foods within given categories (Barlagne et al., 2015).

D. Cauchi et al. / Food Policy 71 (2017) 39–47

Strategies to encourage owners of smaller grocery stores to stock more healthful food options (e.g. assistance with purchasing freezers; improving direct links with farmers etc.) have also been shown to be successful (Cavanaugh et al., 2014; Gittelsohn et al., 2012), and should be actively pursued locally. Furthermore, research suggests that increased frequency of shopping for groceries may negatively influence the healthfulness of food purchases by increasing consumer exposure to highly processed, energy-dense foods (Rudi and Çakır, 2017), hence policies that limit the accessibility and visibility of unhealthy foods in grocery stores (e.g. restricting availability of sweets and confectionery items at supermarket checkouts) may be effective in improving consumer purchasing behaviour and overall diet quality. Lastly, the diversity of food stores may be a more important predictor of weight outcomes than store density, so that policies designed to increase the relative proportion of outlets serving healthy versus unhealthy foods should be considered (Centre for Disease Control and Prevention, 2011; Polsky et al., 2015). Recent efforts to establish farmers’ markets across Malta to increase food retail diversity and access to fresh F&V (Health Promotion and Disease Prevention Directorate, 2015) may help to improve this diversity. 4.2. Limitations This study has several limitations. Although nationally representative, the relatively small sample size may have limited ability to detect significant associations. The foods assessed were not randomly selected and do not reflect the entire range of foods available to consumers. Moreover, the NEMS-S protocol requires use of branded products, possibly underestimating the true price differential between healthy and regular options since consumers on a limited budget might preferentially purchase cheaper, nonbranded products that lack healthier alternatives. Only grocery stores were evaluated, hence results do not consider the full consumer food environment. This may be of importance: butcher stores in Malta, for example, typically offer packaged savoury snacks in addition to meat products. Seasonal changes in price, quality and availability of foods could not be determined due to the cross-sectional nature of the study. There is also uncertainty as to which is the most appropriate unit of analysis to characterise food store diversity (Black and Macinko, 2008), as neighbourhoods of differing affluence may be found within the same locality (NSO, 2012). Furthermore, Malta’s small size and the population’s relative reliance on private transport (Cauchi et al., 2015) renders identification of the consumer food environment where food is actually purchased problematic. Although studies have hinted that socioeconomically disadvantaged individuals may shop frequently and in small amounts at ‘traditional’ food outlets close to their residences (Webber et al., 2010), selective purchasing behaviour from different locations and store types may be taking place (e.g. fresh F&V may be judiciously bought from vendors in the neighbourhood, whereas packaged foods are sourced from large supermarket conveniently located on the commute to the workplace) (Webber et al., 2010). In addition, there is a dearth of quantitative research on individual food purchasing and food consumption behaviour of the Maltese (Cauchi et al., 2015), hence a definite relationship between the local food environments, individual dietary behaviour, and weight outcomes cannot be established. Lastly, our results regarding area-level food store density and risk of overweight or obesity should be interpreted with caution. It is possible that some stores listed in the database of licensed food outlets were inadvertently misclassified, and the lack of data on other relevant covariates such as actual location of children’s residence; the distance between homes and food stores in their neighbourhood; parental weight status and SES prevented adjusting for these variables in the analysis.

45

4.3. Conclusion This is the first study to characterise the community and consumer food environment in Malta, providing baseline information for monitoring of food availability and prices. Findings may guide national policy efforts to improve the food environment in Malta, and may be of particular interest to public health professionals in other islands. The data presented here suggest that availability and price of most food items do not vary by area-level deprivation. However, small grocery stores often have lower availability of certain healthy food options and are more expensive compared to medium and large stores. Despite these findings, no firm conclusions can be made regarding the potential impact of price or structural modifications to the food environment upon dietary behaviour in Malta. Experimental or longitudinal research that accounts for neighbourhood selfselection and evaluates environmental influences on individual purchasing patterns is required for causal inference to be determined (Gordon-Larsen, 2014; Kelly et al., 2011). Work to determine the most appropriate measures for socioeconomic status and units of analyses when characterising food environments is needed. Food environment studies set within a wider range of settings including schools and workplaces, and studies that assess individual travel patterns to determine food availability throughout the day rather than relying on residential address, are crucial (Vandevijvere et al., 2013). Acknowledgements Financial support for this study was provided by a grant from the Malta Government Scholarship Scheme. The funding agreement ensured the authors’ independence in designing the study, interpreting the data, writing, and publishing the report. The authors would also like to thank Dr. Victoria Farrugia Sant’Angelo (School Health Services, Primary Health Care Department), who provided children’s BMI data used in this study; the Environmental Health Directorate of Malta, which provided store licensing data; and fieldworkers John Cauchi, Karl Spiteri, and Bianca Van Der Craats. The constructive suggestions provided by two anonymous referees and the responsible editor are gratefully acknowledged. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.foodpol.2017.07. 004. References Barlagne, C., Bazoche, P., Thomas, A., Ozier-Lafontaine, H., Causeret, F., Blazy, J.-M., 2015. Promoting local foods in small island states: the role of information policies. Food Policy 57, 62–72. http://dx.doi.org/10.1016/j.foodpol. 2015.09.003. Black, J.L., Macinko, J., 2008. Neighborhoods and obesity. Nutr. Rev. 66, 2–20. http:// dx.doi.org/10.1111/j.1753-4887.2007.00001.x. Briesch, R.A., Chintagunta, P.K., Fox, E.J., 2009. How does assortment affect grocery store choice? J. Mark. Res. 46, 176–189. http://dx.doi.org/10.1509/ jmkr.46.2.176. Caspi, C.E., Sorensen, G., Subramanian, S.V.V., Kawachi, I., 2012. The local food environment and diet: a systematic review. Health Place 18, 1172–1187. http:// dx.doi.org/10.1016/j.healthplace.2012.05.006.The. Cauchi, D., Rutter, H., Knai, C., 2015. An obesogenic island in the Mediterranean: mapping potential drivers of obesity in Malta. Public Health Nutr. 18, 3211– 3223. http://dx.doi.org/10.1017/S1368980015000476. Cavanaugh, E., Green, S., Mallya, G., Tierney, A., Brensinger, C., Glanz, K., 2014. Changes in food and beverage environments after an urban corner store intervention. Prev. Med. (Baltim) 65, 7–12. http://dx.doi.org/10.1016/j. ypmed.2014.04.009. Centre for Disease Control and Prevention, 2011. Children’s Food Environment State Indicator Report.

46

D. Cauchi et al. / Food Policy 71 (2017) 39–47

Cerin, E., Frank, L.D., Sallis, J.F., Saelens, B.E., Conway, T.L., Chapman, J.E., Glanz, K., 2011. From neighborhood design and food options to residents’ weight status. Appetite 56, 693–703. http://dx.doi.org/10.1016/j.appet.2011.02.006. Chow, C.K., Lock, K., Teo, K., Subramanian, S.V., McKee, M., Yusuf, S., 2009. Environmental and societal influences acting on cardiovascular risk factors and disease at a population level: a review. Int. J. Epidemiol. 38, 1580–1594. http:// dx.doi.org/10.1093/ije/dyn258. Cobb, L.K., Appel, L.J., Franco, M., Jones-Smith, J.C., Nur, A., Anderson, C.A.M., 2015. The relationship of the local food environment with obesity: a systematic review of methods, study quality, and results. Obesity (Silver Spring) 23, 1331– 1344. http://dx.doi.org/10.1002/oby.21118. Colchero, M.A., Popkin, B.M., Rivera, J.A., Ng, S.W., 2016. Beverage purchases from stores in Mexico under the excise tax on sugar sweetened beverages: observational study. BMJ 352, h6704. http://dx.doi.org/10.1136/bmj.h6704. Cole, T.J., Lobstein, T., 2012. Extended international (IOTF) body mass index cut-offs for thinness, overweight and obesity. Pediatr. Obes. 7, 284–294. http://dx.doi. org/10.1111/j.2047-6310.2012.00064.x. Cummins, S., Macintyre, S., 2002. A systematic study of an Urban foodscape: the price and availability of food in greater Glasgow. Br. Food J. 10, 545–553. http:// dx.doi.org/10.1080/004209802200001139. Cummins, S., Smith, D.M., Aitken, Z., Dawson, J., Marshall, D., Sparks, L., Anderson, a. S., 2010. Neighbourhood deprivation and the price and availability of fruit and vegetables in Scotland. J. Hum. Nutr. Diet. 23, 494–501. http://dx.doi.org/ 10.1111/j.1365-277X.2010.01071.x. Cummins, S., Smith, D.M., Taylor, M., Dawson, J., Marshall, D., Sparks, L., Anderson, A.S., 2009. Variations in fresh fruit and vegetable quality by store type, urbanrural setting and neighbourhood deprivation in Scotland. Public Health Nutr. 12, 2044–2050. http://dx.doi.org/10.1017/S1368980009004984. Curhan, R., 1972. The relationship between shelf space and unit sales in supermarkets. J. Mark. Res. Darmon, N., Drewnowski, A., 2015. Contribution of food prices and diet cost to socioeconomic disparities in diet quality and health: a systematic review and analysis. Nutr. Rev. 73, 643–660. http://dx.doi.org/10.1093/nutrit/nuv027. Drewnowski, A., Specter, S.E., 2004. Poverty and obesity: the role of energy density and energy costs. Am. J. Clin. Nutr. 79, 6–16. Egger, G., Swinburn, B., 1997. An ‘‘ecological” approach to the obesity pandemic. BMJ 315, 477–480. Engler-Stringer, R., Le, H., Gerrard, A., Muhajarine, N., 2014. The community and consumer food environment and children’s diet: a systematic review. BMC Public Health 14, 522. http://dx.doi.org/10.1186/1471-2458-14-522. Environmental Health Directorate, 2014. Applications for food trading licenses (2003–2013). Gittelsohn, J., Rowan, M., Gadhoke, P., 2012. Interventions in small food stores to change the food environment, improve diet, and reduce risk of chronic disease. Prev. Chronic Dis. 9, E59. Glanz, K., Bader, M.D.M., Iyer, S., 2012. Retail grocery store marketing strategies and obesity: an integrative review. Am. J. Prev. Med. 42, 503–512. http://dx.doi.org/ 10.1016/j.amepre.2012.01.013. Glanz, K., Basil, M., Maibach, E., Goldberg, J., Snyder, D., 1998. Why Americans eat what they do: taste, nutrition, cost, convenience, and weight control concerns as influences on food consumption. J. Am. Diet. Assoc. 98, 1118–1126. http://dx. doi.org/10.1016/S0002-8223(98)00260-0. Glanz, K., Sallis, J.F., Saelens, B.E., Frank, L.D., 2007. Nutrition Environment Measures Survey in stores (NEMS-S): development and evaluation. Am. J. Prev. Med. 32, 282–289. http://dx.doi.org/10.1016/j.amepre.2006.12.019. Glanz, K., Sallis, J.F., Saelens, B.E., Frank, L.D., 2005. Healthy nutrition environments: concepts and measures. Am. J. Heal. Promot. 19, 330–333. http://dx.doi.org/ 10.4278/0890-1171-19.5.330. Glanz, K., Yaroch, A.L., 2004. Strategies for increasing fruit and vegetable intake in grocery stores and communities: policy, pricing, and environmental change. Prev. Med. (Baltim) 39, 75–80. http://dx.doi.org/10.1016/j.ypmed.2004.01.004. Gordon-Larsen, P., 2014. Food availability/convenience and obesity. Adv. Nutr. 809– 817. http://dx.doi.org/10.3945/an.114.007070.(7). Grech, V., Aquilina, S., Camilleri, E., Camilleri, K., Busuttil, M.-L., Sant’Angelo, V.F., Calleja, N., 2016. The Malta childhood national body mass index study – a population study. J. Pediatr. Gastroenterol. Nutr. 1. http://dx.doi.org/10.1097/ MPG.0000000000001430. He, M., Tucker, P., Gilliland, J., Irwin, J.D., Larsen, K., Hess, P., 2012. The influence of local food environments on adolescents’ food purchasing behaviors. Int. J. Environ. Res. Public Health 9, 1458–1471. http://dx.doi.org/10.3390/ ijerph9041458. Health Information and Research Directorate, 2008. European Health Interview Survey 2008 – Summary Statistics. Directorate of Health Information and Research, Malta, Msida. Health Promotion and Disease Prevention Directorate, 2015. Food and Nutrition Policy and Action Plan for Malta 2015-2020. Holsten, J.E., 2009. Obesity and the community food environment: a systematic review. Public Health Nutr. 12, 397–405. http://dx.doi.org/10.1017/ S1368980008002267. Honeycutt, S., Davis, E., Clawson, M., Glanz, K., 2010. Training for and dissemination of the Nutrition Environment Measures Surveys (NEMS). Prev. Chronic Dis. 7, A126. Horgen, K.B., Brownell, K.D., 2002. Comparison of price change and health message interventions in promoting healthy food choices. Health Psychol. 21, 505–512. Horowitz, C.R., Colson, K.A., Hebert, P.L., Lancaster, K., 2004. Barriers to buying healthy foods for people with diabetes: evidence of environmental disparities. Am. J. Public Health 94, 1549–1554.

Howard Wilsher, S., Harrison, F., Yamoah, F., Fearne, A., Jones, A., 2016. The relationship between unhealthy food sales, socio-economic deprivation and childhood weight status: results of a cross-sectional study in England. Int. J. Behav. Nutr. Phys. Act. 13, 21. http://dx.doi.org/10.1186/s12966-016-0345-2. Inman, J., Winer, R., Ferraro, R., 2009. The interplay among category characteristics, customer characteristics, and customer activities on in-store decision making. J. Mark. 73, 19–29. Jeffery, R.W., Baxter, J., McGuire, M., Linde, J., 2006. Are fast food restaurants an environmental risk factor for obesity? Int. J. Behav. Nutr. Phys. Act. 3, 2. http:// dx.doi.org/10.1186/1479-5868-3-2. Kearney, J.M., McElhone, S., 1999. Perceived barriers in trying to eat healthier– results of a pan-EU consumer attitudinal survey. Br. J. Nutr. 81 (Suppl. 2). S1337. Kelly, B., Flood, V.M., Yeatman, H., 2011. Measuring local food environments: an overview of available methods and measures. Heal. Place 17, 1284–1293. http:// dx.doi.org/10.1016/j.healthplace.2011.08.014. Kerr, J., Sallis, J.F., Bromby, E., Glanz, K., 2012. Assessing reliability and validity of the GroPromo audit tool for evaluation of grocery store marketing and promotional environments. J. Nutr. Educ. Behav. 44, 597–603. http://dx.doi.org/10.1016/j. jneb.2012.04.017. Lake, A.a., Burgoine, T., Greenhalgh, F., Stamp, E., Tyrrell, R., 2010. The foodscape: classification and field validation of secondary data sources. Health Place 16, 666–673. http://dx.doi.org/10.1016/j.healthplace.2010.02.004. Larson, N.I., Story, M.T., Nelson, M.C., 2009. Neighborhood environments. disparities in access to Healthy foods in the U.S. Am. J. Prev. Med. 36 (74–81), e10. http:// dx.doi.org/10.1016/j.amepre.2008.09.025. Laska, M.N., Hearst, M.O., Forsyth, A., Pasch, K.E., Lytle, L., 2010. Neighbourhood food environments: are they associated with adolescent dietary intake, food purchases and weight status? Public Health Nutr. 13, 1757–1763. http://dx. doi.org/10.1017/S1368980010001564. Lee, A., Mhurchu, C.N., Sacks, G., Swinburn, B., Snowdon, W., Vandevijvere, S., Hawkes, C., L’Abbé, M., Rayner, M., Sanders, D., Barquera, S., Friel, S., Kelly, B., Kumanyika, S., Lobstein, T., Ma, J., Macmullan, J., Mohan, S., Monteiro, C., Neal, B., Walker, C., Informas, 2013. Monitoring the price and affordability of foods and diets globally. Obes. Rev. 14, 82–95. http://dx.doi.org/10.1111/ obr.12078. Mackenbach, J.D., Rutter, H., Compernolle, S., Glonti, K., Oppert, J.-M., Charreire, H., De Bourdeaudhuij, I., Brug, J., Nijpels, G., Lakerveld, J., 2014. Obesogenic environments: a systematic review of the association between the physical environment and adult weight status, the SPOTLIGHT project. BMC Public Health 14, 233. http://dx.doi.org/10.1186/1471-2458-14-233. Malta Standards Authority, 2010. National Food Consumption Survey, Malta. Martins, P.a., Cremm, E.C., Leite, F.H.M., Maron, L.R., Scagliusi, F.B., Oliveira, M.a., 2013. Validation of an adapted version of the nutrition environment measurement tool for stores (NEMS-S) in an Urban Area of Brazil. J. Nutr. Educ. Behav. 45, 785–792. http://dx.doi.org/10.1016/j.jneb.2013.02.010. McKay, L., Sammut, J., Farrugia, K., Piscopo, S., 2012. A Minimum Budget for a Decent Living. Moore, L.V., Diez Roux, A.V., 2006. Associations of neighborhood characteristics with the location and type of food stores. Am. J. Public Health 96, 325–331. http://dx.doi.org/10.2105/AJPH.2004.058040. Morland, K., Diez Roux, A.V., Wing, S., 2006. Supermarkets, other food stores, and obesity: the atherosclerosis risk in communities study. Am. J. Prev. Med. 30, 333–339. http://dx.doi.org/10.1016/j.amepre.2005.11.003. Mullis, R.M., Snyder, M.P., Hunt, M.K., 1990. Developing nutrient criteria for foodspecific dietary guidelines for the general public. J. Am. Diet. Assoc. 90, 847– 851. National Statistics Office (Malta), 2010. Household Budgetary Survey 2008. Valletta. Nestle, M., 2015. Soda Politics: Taking on Big Soda (and Winning). Oxford University Press, New York. Ng, M., Fleming, T., Robinson, M., Thomson, B., Graetz, N., Margono, C., Mullany, E.C., Biryukov, S., Abbafati, C., Abera, S.F., Abraham, J.P., Abu-Rmeileh, N.M.E., Achoki, T., AlBuhairan, F.S., Alemu, Z.A., Alfonso, R., Ali, M.K., Ali, R., Guzman, N.A., Ammar, W., Anwari, P., Banerjee, A., Barquera, S., Basu, S., Bennett, D.A., Bhutta, Z., Blore, J., Cabral, N., Nonato, I.C., Chang, J.-C., Chowdhury, R., Courville, K.J., Criqui, M.H., Cundiff, D.K., Dabhadkar, K.C., Dandona, L., Davis, A., Dayama, A., Dharmaratne, S.D., Ding, E.L., Durrani, A.M., Esteghamati, A., Farzadfar, F., Fay, D. F.J., Feigin, V.L., Flaxman, A., Forouzanfar, M.H., Goto, A., Green, M.A., Gupta, R., Hafezi-Nejad, N., Hankey, G.J., Harewood, H.C., Havmoeller, R., Hay, S., Hernandez, L., Husseini, A., Idrisov, B.T., Ikeda, N., Islami, F., Jahangir, E., Jassal, S.K., Jee, S.H., Jeffreys, M., Jonas, J.B., Kabagambe, E.K., Khalifa, S.E.A.H., Kengne, A.P., Khader, Y.S., Khang, Y.-H., Kim, D., Kimokoti, R.W., Kinge, J.M., Kokubo, Y., Kosen, S., Kwan, G., Lai, T., Leinsalu, M., Li, Y., Liang, X., Liu, S., Logroscino, G., Lotufo, P.A., Lu, Y., Ma, J., Mainoo, N.K., Mensah, G.A., Merriman, T.R., Mokdad, A. H., Moschandreas, J., Naghavi, M., Naheed, A., Nand, D., Narayan, K.M.V., Nelson, E.L., Neuhouser, M.L., Nisar, M.I., Ohkubo, T., Oti, S.O., Pedroza, A., Prabhakaran, D., Roy, N., Sampson, U., Seo, H., Sepanlou, S.G., Shibuya, K., Shiri, R., Shiue, I., Singh, G.M., Singh, J.A., Skirbekk, V., Stapelberg, N.J.C., Sturua, L., Sykes, B.L., Tobias, M., Tran, B.X., Trasande, L., Toyoshima, H., van de Vijver, S., Vasankari, T. J., Veerman, J.L., Velasquez-Melendez, G., Vlassov, V.V., Vollset, S.E., Vos, T., Wang, C., Wang, S.X., Weiderpass, E., Werdecker, A., Wright, J.L., Yang, Y.C., Yatsuya, H., Yoon, J., Yoon, S.-J., Zhao, Y., Zhou, M., Zhu, S., Lopez, A.D., Murray, C. J.L., Gakidou, E., 2014. Global, regional, and national prevalence of overweight and obesity in children and adults during 1980–2013: a systematic analysis for the Global Burden of Disease Study 2013. Lancet 6736, 1–16. http://dx.doi.org/ 10.1016/S0140-6736(14)60460-8.

D. Cauchi et al. / Food Policy 71 (2017) 39–47 Ni Mhurchu, C., Vandevijvere, S., Waterlander, W., Thornton, L.E., Kelly, B., Cameron, a.J., Snowdon, W., Swinburn, B., Informas, 2013. Monitoring the availability of healthy and unhealthy foods and non-alcoholic beverages in community and consumer retail food environments globally. Obes. Rev. 14, 108–119. http://dx. doi.org/10.1111/obr.12080. NSO, 2012. Statistics on Income and Living Conditions 2010. National Statistics Office, Malta. Polsky, J.Y., Moineddin, R., Dunn, J.R., Glazier, R.H., Booth, G.L., 2015. Absolute and relative densities of fast-food versus other restaurants in relation to weight status: does restaurant mix matter? Prev. Med. (Baltim) 82, 28–34. http://dx. doi.org/10.1016/j.ypmed.2015.11.008. Pomerleau, J., Knai, C., Foster, C., Rutter, H., Darmon, N., Derflerova Brazdova, Z., Hadziomeragic, A.F., Pekcan, G., Pudule, I., Robertson, A., Brunner, E., Suhrcke, M., Gabrijelcic Blenkus, M., Lhotska, L., Maiani, G., Mistura, L., Lobstein, T., Martin, B.W., Elinder, L.S., Logstrup, S., Racioppi, F., McKee, M., 2012. Measuring the food and built environments in urban centres: reliability and validity of the EURO-PREVOB Community Questionnaire. Public Health 127, 259–267. http:// dx.doi.org/10.1016/j.puhe.2012.12.025. Powell, L.M., Auld, M.C., Chaloupka, F.J., O’Malley, P.M., Johnston, L.D., 2007. Associations between access to food stores and adolescent body mass index. Am. J. Prev. Med. 33, 301–307. http://dx.doi.org/10.1016/j.amepre.2007.07.007. Powell, L.M., Chriqui, J.F., Khan, T., Wada, R., Chaloupka, F.J., 2013. Assessing the potential effectiveness of food and beverage taxes and subsidies for improving public health: a systematic review of prices, demand and body weight outcomes. Obes. Rev. 14, 110–128. http://dx.doi.org/10.1111/obr.12002. Rudi, J., Çakır, M., 2017. Vice or virtue: how shopping frequency affects healthfulness of food choices. Food Policy 69, 207–217. Snowdon, W., Thow, A.M., 2013. Trade policy and obesity prevention: challenges and innovation in the Pacific Islands. Obes. Rev. 14 (Suppl. 2), 150–158. http:// dx.doi.org/10.1111/obr.12090. Story, M., Kaphingst, K.M., Robinson-O’Brien, R., Glanz, K., 2008. Creating healthy food and eating environments: policy and environmental approaches. Annu. Rev. Public Health 29, 253–272. http://dx.doi.org/10.1146/annurev. publhealth.29.020907.090926.

47

Superintendence of Public Health, 2012. A Healthy Weight for Life: A National Strategy for Malta (2012–2020). Swinburn, B.A., Sacks, G., Hall, K.D., McPherson, K., Finegood, D.T., Moodie, M.L., Gortmaker, S.L., 2011. The global obesity pandemic: shaped by global drivers and local environments. Lancet 378, 804–814. http://dx.doi.org/10.1016/S01406736(11)60813-1. Swinburn, B., Sacks, G., Vandevijvere, S., Kumanyika, S., Lobstein, T., Neal, B., Barquera, S., Friel, S., Hawkes, C., Kelly, B., L’Abbé, M., Lee, A., Ma, J., Macmullan, J., Mohan, S., Monteiro, C., Rayner, M., Sanders, D., Snowdon, W., Walker, C., Informas, 2013. INFORMAS (International Network for Food and Obesity/noncommunicable diseases Research, Monitoring and Action Support): overview and key principles. Obes. Rev. 14, 1–12. http://dx.doi.org/10.1111/obr.12087. Vandevijvere, S., Monteiro, C., Krebs-Smith, S.M., Lee, A., Swinburn, B., Kelly, B., Neal, B., Snowdon, W., Sacks, G., Informas, 2013. Monitoring and benchmarking population diet quality globally: a step-wise approach. Obes. Rev. 14, 135–149. http://dx.doi.org/10.1111/obr.12082. Verschuuren, M., Fietje, N., Greenwell, F., Raj, T., Stein, C., 2015. The European Health Report 2015: Targets and Beyond – Reaching New Frontiers in Evidence. Copenhagen: UN City. Vilaro, M.J., Barnett, T.E., 2013. The rural food environment: a survey of food price, availability, and quality in a rural florida community. Food Public Health. Wang, M.C., Kim, S., Gonzalez, A.A., MacLeod, K.E., Winkleby, M.A., 2007. Socioeconomic and food-related physical characteristics of the neighbourhood environment are associated with body mass index. J. Epidemiol. Community Health 61, 491–498. http://dx.doi.org/10.1136/jech.2006.051680. Webber, C.B., Sobal, J., Dollahite, J.S., 2010. Shopping for fruits and vegetables. Food and retail qualities of importance to low-income households at the grocery store. Appetite 54, 297–303. http://dx.doi.org/10.1016/j.appet.2009.11.015. Wilkinson, J., Mason, J., Paksoy, C., 1982. Assessing the impact of short-term supermarket strategy variables. J. Mark. Res. Winkler, E., Turrell, G., Patterson, C., 2006. Does living in a disadvantaged area entail limited opportunities to purchase fresh fruit and vegetables in terms of price, availability, and variety? findings from the Brisbane Food Study. Health Place 12, 741–748. http://dx.doi.org/10.1016/j.healthplace.2005.09.006.