See corresponding editorial on page 1284.
Reducing energy intake and energy density for a sustainable diet: a study based on self-selected diets in French adults1–3 Gabriel Masset, Florent Vieux, Eric Olivier Verger, Louis-Georges Soler, Djilali Touazi, and Nicole Darmon
INTRODUCTION
“Sustainable diets,” as defined by the FAO, need to meet nutritional (health), environmental, affordability, and cultural acceptability criteria (1) based on the sustainable development framework. According to the latest FAO data and the Intergovernmental Panel on Climate Change, agriculture and deforestation accounted for approximately one-third of global greenhouse gas emissions
1460
(GHGE)4 in 2010 (2, 3). Therefore, a reduction in food-related GHGE is needed to achieve national and international GHGE reduction targets (4, 5). The production and consumption of foods of animal origin, particularly ruminant meat, result in increased emissions of greenhouse gases (6, 7) and increased energy (8) and water (9) use relative to plant-based foods. Consequently, experts in environmental sciences argue that diets containing reduced amounts of animal-origin foods would produce less GHGE (10–12). It was estimated that several healthy dietary patterns, such as the Mediterranean (13) or the Nordic diet (14), or diets designed to have a low amount of meat (15, 16) produce less GHGE than do currently observed Western diets. According to studies based on diet modeling, reducing diet-related GHGE while ensuring nutritional adequacy is theoretically feasible (17–20). Two studies based on predictive public health models have strengthened the hypothesis that low-GHGE diets containing a lower content of animal-origin foods and a higher content of plantbased foods could prevent chronic disease and mortality (21, 22). Furthermore, increased consumption of plant-based foods could help reduce the energy density of diets (23) and may therefore help to control total energy intake, which is an important driver of the rising prevalence of obesity (24, 25). 1
From the Unite´ Mixte de Recherche “Nutrition, Obesity and Risk of Thrombosis,” Institut National de la Recherche Agronomique, Institut National de la Sante´ et de la Recherche Me´dicale, Aix- Marseille Universite´, Marseille, France (GM, FV, and ND); AgroParisTech, Nutrition Physiology and Ingestive Behavior, Paris, France and Danone Research, Global Nutrition Department, Palaiseau, France (EOV); and UR “Aliss,” Ivry sur Seine, France (L-GS and DT). 2 Supported by the French National Research Agency under the OCAD project ANR-11-ALID-0002 and the French Environment and Energy Management Agency (ADEME). FV received a PhD grant from the INRA divisions of Nutrition, Chemical Food Safety, and Consumer Behavior and of Social Sciences, Agriculture and Food, Rural Development, and Environment. EOV received a PhD grant from Danone Research and from the Association Nationale de la Recherche et de la Technologie (CIFRE 474/2010). 3 Address correspondence to N Darmon, UMR 1062 INSERM/1260 INRA/ Universite´ d’Aix-Marseille, Nutrition Obe´site´ et Risque Thrombotique, Faculte´ de Me´decine de la Timone, 27 Boulevard Jean Moulin, 13385 Marseille Cedex 05, France. E-mail: [email protected]. 4 Abbreviations used: GHGE, greenhouse gas emissions; INCA2, Second Individual and National Survey on Food Consumption; PANDiet, probability of adequate nutrient intake. Received October 16, 2013. Accepted for publication March 5, 2014. First published online April 2, 2014; doi: 10.3945/ajcn.113.077958.
Am J Clin Nutr 2014;99:1460–9. Printed in USA. Ó 2014 American Society for Nutrition Supplemental Material can be found at: http://ajcn.nutrition.org/content/suppl/2014/05/13/ajcn.113.0 77958.DCSupplemental.html
Downloaded from ajcn.nutrition.org at MCMASTER UNIVERSITY on January 5, 2015
ABSTRACT Background: Studies on theoretical diets are not sufficient to implement sustainable diets in practice because of unknown cultural acceptability. In contrast, self-selected diets can be considered culturally acceptable. Objective: The objective was to identify the most sustainable diets consumed by people in everyday life. Design: The diet-related greenhouse gas emissions (GHGE) for self-selected diets of 1918 adults participating in the cross-sectional French national dietary survey Individual and National Survey on Food Consumption (INCA2) were estimated. “Lower-Carbon,” “Higher-Quality,” and “More Sustainable” diets were defined as having GHGE lower than the overall median value, a probability of adequate nutrition intake (PANDiet) score (a measure of the overall nutritional adequacy of a diet) higher than the overall median value, and a combination of both criteria, respectively. Diet cost, as a proxy for affordability, and energy density were also assessed. Results: More Sustainable diets were consumed by 23% of men and 20% of women, and their GHGE values were 19% and 17% lower than the population average (mean) value, respectively. In comparison with the average value, Lower-Carbon diets achieved a 20% GHGE reduction and lower cost, but they were not sustainable because they had a lower PANDiet score. Higher-Quality diets were not sustainable because of their above-average GHGE and cost. More Sustainable diets had an above-average PANDiet score and a below-average energy density, cost, GHGE, and energy content; the energy share of plant-based products was increased by 20% and 15% compared with the average for men and women, respectively. Conclusions: A strength of this study was that most of the dimensions for “sustainable diets” were considered, ie, not only nutritional quality and GHGE but also affordability and cultural acceptability. A reduction in diet-related GHGE by 20% while maintaining high nutritional quality seems realistic. This goal could be achieved at no extra cost by reducing energy intake and energy density and increasing the share of plant-based products. Am J Clin Nutr 2014;99:1460–9.
IDENTIFYING THE MOST SUSTAINABLE FRENCH DIETS
MATERIALS AND METHODS
Population sample and dietary data The dietary intakes were derived from the 7-d food records of a nationally representative random sample of adults (n = 2624; age .18 y) participating in the INCA2 cross-sectional dietary survey Enqueˆte Individuelle et Nationale sur les Consommations Alimentaires (Individual and National Survey on Food Consumption), conducted in 2006–2007 by ANSES (French Agency for Food, Environmental and Occupational Health Safety) (28). The sampling method involved a 3-stage stratified random sampling. To ensure the representative nature of the sample, a statistical adjustment was performed for the region, town size, age, sex, occupation of the household head, household size, and seasonal variables. After underreporters were excluded by using the Black equations (29), the current analysis was conducted on a final sample of 1918 adults (776 men and 1142 women). The study was approved by the French Data Protection Authority Commission (Commission Nationale Informatique et Liberte´s). All of the foods declared as consumed by the participants during the survey (n = 1314 foods and beverages, including water) were categorized into 8 food groups and 27 food subgroups within a food nutrient composition database associated with the survey. In addition, to assess the share of plant-based foods in the diets, plant-based foods were identified as fruit, vegetables, and nuts; cereals, including breakfast cereals; potatoes; legumes; vegetable oils and margarines; and mixed dishes without meat, fish products, or eggs. The total weight and energy intake (in kcal/d) were calculated for each participant by summing the quantity and energy content of all of the foods and beverages consumed by the individual. The diet weight (in g/d) for the solid foods was calculated by excluding foods typically consumed as beverages, such as milk, juices, and other drinks, as proposed by Ledikwe et al (30). The dietary energy density was then calculated (in kcal/100 g) by dividing the energy intake from solid food by the solid diet weight. Food prices were obtained from the 2006 Kantar-World Panel purchase database (31), which gives the annual expenditure and quantity purchased for each food item available on the market in
a representative sample of 12,000 French households. The mean prices were calculated by dividing the annual expenditures by the quantities purchased. Food prices were expressed in euros per 100 g of edible food (ie, after the changes in weight associated with the trimming or cooking processes were taken into account by using the appropriate conversion factors). The diet costs per day (€/d) were then estimated by multiplying the reported edible weight of each food/beverage by its unit cost and summing over all foods/beverages consumed by that individual. In contrast with energy density, the diet cost was calculated accounting for all of the foods consumed, including beverages. The diet costs per 1000 kcal were also estimated. The PANDiet index for nutritional quality and identification of Higher-Quality diets The adequacy of nutrient intake for each individual was calculated by using the PANDiet index, based solely on nutrients. The PANDiet is composed of an “adequacy” subscore that includes positive nutrients and a “moderation” subscore for the nutrients to be limited, as described in full elsewhere (27). Briefly, the algorithm calculates the probability of adequacy for each nutrient on a scale from 0 to 1, where 1 represents a 100% probability that the usual intake is adequate by satisfying the requirement or is not excessive, according to a reference value. The probability calculation accounts for the number of days of diet recording, average intake, and day-to-day variability in intake, nutrient reference value, and interindividual variability. The reference values were the French nutritional recommendations for adults (32) or the European Union values (33, 34) when specific recommendations were lacking (see Supplementary Table 1 under “Supplemental data” in the online issue). The PANDiet index generates a summary score for all of these probabilities, ranging from 0 to 100; a higher score reflects better nutritional adequacy of the diet (27). Hence, the PANDiet index accounts for day-to-day variability in nutrient intake and does not generate a priori hypotheses for the dietary content of the diet. The PANDiet has been shown to relate to other indicators of nutritional quality among participants in the French Etude Nationale Nutrition Sante´ and US NHANES dietary surveys (27). In the current study, Higher-Quality diets were defined as those with a PANDiet score higher than the sexspecific median score. Estimating diet-related GHGE and identifying Lower-Carbon diets The estimation of diet-related GHGE was based on the selection of 391 foods among those most consumed by INCA2 participants, as previously described by Vieux et al (26). Briefly, within each food subgroup, the food items were ranked in decreasing order based on the percentage of participants who consumed them, and at least one food item was selected from among the most widely consumed items in each subgroup. This process identified 391 representative foods from the 1314 items initially listed in the food database. These 391 foods covered 71% of the total weight intake and 66% of the total energy intake in the INCA2 study population. An environmental consulting firm, Greenext Service (Paris, France), assigned the GHGE values for the 391 foods following
Downloaded from ajcn.nutrition.org at MCMASTER UNIVERSITY on January 5, 2015
However, “cultural acceptability”—a key dimension of the FAO’s definition of “sustainable diets” (1)—was typically not considered in the studies described above. In an analysis of diets actually consumed by French adults participating in the representative Second Individual and National Survey on Food Consumption (INCA2), Vieux et al (26) did not find an association between a higher nutritional quality and lower GHGE, which thus calls into question the relevance of studies based on theoretical and modeling approaches. To help implement the sustainability concept into practice, this study identified existing diets that combined higher nutritional quality and lower GHGE. Among participants in the nationally representative French INCA2 dietary survey, More Sustainable diets were identified based on diet-related GHGEs derived from a life cycle analysis conducted on almost 400 widely consumed foods (26) and on the probability of adequate nutrient intake (PANDiet) index [a measure of diet quality in terms of global nutrient adequacy (27)].
1461
1462
MASSET ET AL
Statistical analyses: identifying More Sustainable diets and the main drivers for reaching sustainability More Sustainable diets were defined as those with a PANDiet score higher than the median value and diet-related GHGE value lower than the median value. The dietary characteristics of the Higher-Quality, Lower-Carbon, and More Sustainable diets were compared with the mean values for the whole population, referred to below as the Average diets, by using an analysis of means that accounted for the nonindependence of the samples (41). Ageadjusted analyses were conducted for both sexes. All of the analyses were sex-specific because differences in GHGE have been previously observed to arise from differences in energy and quantity intakes between men and women (42). The analyses were performed by using specific survey data procedures (surveymeans and surveyfreq) in SAS 9.3 software (SAS Institute Inc).
RESULTS
The characteristics of the Lower-Carbon (daily GHGE ,4511 and 3437 g CO2 equivalents in men and women, respectively), Higher-Quality (PANDiet higher than 62.0 in men and women), and More Sustainable diets compared with the Average diets are shown in Table 1 for men and women separately. Men and women with a Higher-Quality diet and men with a More Sustainable diet were older than average. The diets of 181 men and 230 women did not fulfill the Lower-Carbon or Higher-Quality criteria and therefore were not classified in any of the 3 identified diet categories. The percentage variations of the selected characteristics between the Average diets and the 3 specific diets are shown in Figure 1. The Lower-Carbon diets had daily GHGE reduced by 20% and 21% compared with the Average diets in men and women, respectively. Their GHGE per kcal were reduced by 8% and 7% in men and women, respectively. Compared with the Average diets, the Lower-Carbon diets contained a lower amount of food and provided less energy intake. The lower total intake was associated with lower nutritional adequacy, as measured by the PANDiet, and a lower diet cost, per day and per kcal.
The Higher-Quality diets contained higher amounts of food, provided more energy, and were associated with higher GHGE per day than were the Average diets. In women only, GHGE were also higher when expressed per 1000 kcal. The Higher-Quality diets had a lower energy density and contained a higher share of plant-based foods than did the average diets. In both sexes, the Higher-Quality diets had the highest cost, both per day and per kcal. The More Sustainable diets were consumed by 181 men and 229 women (23% and 20%, respectively). In comparison with the Average diets, their solid weight was not different, but energy intake was reduced by 8% and 10% in men and women, respectively, and the energy density was also reduced. The daily GHGE for the More Sustainable diets were lower than those for the Average diets (219% in men and 217% in women) but were higher than those for the Lower-Carbon diets. When calculated per kcal, the reduction in GHGE for the More Sustainable diets compared with the Average diets was only 11% in men and 7% in women. The More Sustainable diets had the highest content of plant-based foods, both in energy and in weight. The daily cost for the More Sustainable diets was significantly lower than that for the Average diets, although the cost per kcal was not different. The mean intakes by food group and subgroup for the Average, Lower-Carbon, Higher-Quality, and More Sustainable diets are shown in Table 2. Consistent with a lower diet weight and energy intake, the intakes for all food groups and subgroups were equal or lower for the Lower-Carbon diets than for the Average diets (except for sauces and condiments in men). The greatest reductions were observed for meat, fish, and eggs (w20% in both sexes), especially for ruminant and deli meats and alcoholic drinks (w40% reduction). The Higher-Quality diets contained more fruit, vegetables, and nuts; starchy foods; fresh dairy products; oils and margarine; vegetarian mixed dishes; and breakfast cereals than the Average diets. For the Higher-Quality diets, the total content of meat, fish, and eggs was not different for men and was higher for women relative to the Average diets. In both sexes, the intake of butter, cream, and soft drinks was lower for the Higher-Quality diets than for the Average diets. Similar to the Higher-Quality diets, the intake of starchy foods was higher for the More sustainable diets than for the Average diets. The intake of fruit, vegetables, and nuts was also higher for the More Sustainable diets than for the Average diets. For the More Sustainable diets, the intake of meats and eggs was lower than for the Average diets. The intake of fish and fish products was higher in men and not different in women from that of the Average diets. The intake of dairy products was not different between the More Sustainable and Average diets, with a lower content of cheese in the More Sustainable diets. In addition, the More Sustainable diets had the lowest content of butter, cream, alcoholic drinks, salty snacks, desserts, and mixed dishes with animal ingredients. The percentage energy contribution for all meats was reduced in the More Sustainable diets relative to the Average diets (P , 0.001 in both sexes) (Figure 2A). The energy contribution for dairy products was not different between the More Sustainable and Average diets in men and women. In men, the energy contribution for fish products was increased in the More Sustainable diets. In both sexes, the energy contribution for starchy foods, vegetables, and nuts was higher than average for the More Sustainable diets, whereas the energy contribution for mixed
Downloaded from ajcn.nutrition.org at MCMASTER UNIVERSITY on January 5, 2015
the international ISO 1404x life cycle analysis standards (35, 36) (ie, including the whole life cycle of the foods, from farm production to consumption, and waste management for packaging, excluding emissions arising from indirect land use change and consumers’ transport from retail to home). With use of a top-down approach combining French trade and production data (37, 38) and standard life cycle inventory data [eg, Ecoinvent (39)], the GHGE values assigned by the in-house-developed Greenext life cycle analysis tool reflected the average food products as consumed in the French market (40). To estimate the GHGE associated with total dietary intake (ie, not limited to the 391 selected foods), the consumption of foods for which GHGE values were not estimated was reported based on the GHGE values estimated for foods within the same subgroup (26). Lower-Carbon diets were defined as those with a total dietrelated GHGE lower than the sex-specific median value. Dietary GHGE were also assessed per 1000 kcal consumed to better compare diets with varying amounts of energy.
49.0 25.2 2500 2922 1446 162 62.5 4691 1901 8.63 3.48 52.1 41.6
6 6 6 6 6 6 6 6 6 6 6 6 6 0.67 013 23.2 30.1 16.4 1.12 0.31 49.6 12.8 0.10 0.03 0.48 0.44
50.1 24.9 2210 2599 1297 162 61.8 3764 1743 7.29 3.34 53.3 43.3
6 6 6 6 6 6 6 6 6 6 6 6 6 0.97 0.20* 22.9*** 36.7*** 18.6*** 1.62 0.44* 31.0*** 15.3*** 0.09*** 0.04*** 0.67 0.69***
Lower-Carbon (n = 388) 51.4 25.4 2574 3151 1617 149 68.2 4813 1891 9.04 3.56 57.2 47.1
6 6 6 6 6 6 6 6 6 6 6 6 6 1.03** 0.19* 29.4*** 41.1*** 18.9*** 1.20*** 0.30*** 61.0** 17.3 0.12*** 0.04* 0.66*** 0.65***
Higher-Quality (n = 388) 52.2 24.8 2297 2813 1475 149 68.6 3797 1690 7.77 3.45 59.6 50.1
6 6 6 6 6 6 6 6 6 6 6 6 6 1.46* 0.29* 37.0*** 48.1* 25.3 1.99*** 0.41*** 39.3*** 23.0*** 0.12*** 0.07 0.80*** 1.01***
More Sustainable (n = 181) 45.2 24.0 1855 2582 1238 147 62.3 3548 1953 6.69 3.67 53.0 41.2
6 6 6 6 6 6 6 6 6 6 6 6 6 0.58 0.18 15.0 24.7 11.1 1.03 0.27 35.2 15.0 0.06 0.03 0.43 0.42
Average (n = 1142) 44.0 23.5 1605 2251 1054 150 60.2 2803 1813 5.59 3.57 53.3 41.7
6 6 6 6 6 6 6 6 6 6 6 6 6
0.79 0.19 19.4*** 35.3*** 11.5*** 1.66* 0.42*** 22.7*** 17.6*** 0.08*** 0.04** 0.56 0.60
Lower-Carbon (n = 571)
Downloaded from ajcn.nutrition.org at MCMASTER UNIVERSITY on January 5, 2015
47.6 24.3 1901 2829 1399 133 68.4 3753 2011 7.17 3.84 57.1 45.9
6 6 6 6 6 6 6 6 6 6 6 6 6
0.80*** 0.27 21.0** 31.4*** 15.1*** 1.05*** 0.26*** 50.9*** 23.1*** 0.08*** 0.04*** 0.45*** 0.54***
Higher-Quality (n = 569)
Women
47.0 24.0 1665 2570 1219 134 68.1 2947 1818 6.20 3.79 58.2 47.5
6 6 6 6 6 6 6 6 6 6 6 6 6
1.20 0.25 22.9*** 54.8 16.2 1.59*** 0.42*** 28.3*** 23.8*** 0.13*** 0.07 0.65*** 0.87***
More Sustainable (n = 229)
1 All values are means 6 SEs. The Average diet represents mean intakes derived from the sex-specific whole population, the Lower-Carbon diets were those with diet-related GHGE under the overall median (4511 and 3437 g CO2eq in men and women, respectively), the Higher-Quality diets were those with a probability of adequate nutrient intake score above the overall median (62.0 in both men and women), and the More Sustainable diets were those combining the Lower-Carbon and Higher-Quality criteria. A total of 181 men and 230 women were not classified in the Lower-Carbon, Higher-Quality, or More Sustainable diets. Difference between the Average diet and the Lower-Carbon, Higher-Quality, and More Sustainable diets: *P , 0.05, **P , 0.01, ***P , 0.001 (analysis of means). CO2eq, carbon dioxide equivalent; GHGE, greenhouse gas emissions; PANDiet, probability of adequate nutrient intake. 2 Dietary energy density was calculated by dividing the energy intake from solid food by solid diet weight. 3 As of March 12, 2014: US$1 = 0.72€.
Age (y) BMI (kg/m2) Total energy (kcal/d) Total weight (g/d) Solid weight (g/d) Energy density (kcal/100 g)2 PANDiet GHGE (g CO2eq/d) GHGE (g CO2eq/1000 kcal) Diet cost (€/d)3 Energy cost (€/1000 kcal)3 Weight from plant-based foods (%) Energy from plant-based foods (%)
Average (n = 776)
Men
TABLE 1 Characteristics of the Average (whole sample), Lower-Carbon, Higher-Quality, and More Sustainable diets1
IDENTIFYING THE MOST SUSTAINABLE FRENCH DIETS
1463
1464
MASSET ET AL
dishes and drinks was reduced. No difference was found for foods high in fat, salt, or sugar. Meats, including ruminant meat, were the largest contributor (in %) to daily GHGE in the Average and the More Sustainable diets (Figure 2B). Including fish and dairy products, foods of animal origin (excluding mixed dishes) accounted for w45% of daily GHGE and 25% of total energy in the Average diets. Their contribution to daily GHGE was reduced in the More Sustainable diets because of the significant reduction in ruminant-derived GHGE (P , 0.001 in both sexes). Mixed dishes were also important contributors to dietary GHGE (most likely because of the presence of animal products in many prepared meals), but their contribution to daily GHGE was not different between the Average and More Sustainable diets. The increased content of starchy foods, fruit, vegetables, and nuts in the More Sustainable diets compared with the Average diets (Table 2) was associated with an increase in their contribution to daily GHGE (P , 0.001 in both sexes).
DISCUSSION
The originality of the current study lies in its identification of More Sustainable diets based on existing diets in randomly selected individuals in a nationally representative population sample rather than theoretical diets (10, 13, 21) or special diets, such as those adopted by vegetarian or vegan subjects (15). An important result of the current study was that w20% of French adults had diets that could be called More Sustainable because they combined a higher nutritional quality and lower GHGE with no increase in diet cost. The observed shifts between the Average and the More Sustainable diets were in agreement with most food-based dietary guidelines, which indicated that following the existing public health–oriented recommendations could be compatible with reducing dietary GHGE. However, similarly to Vieux et al (26), we showed that higher nutritional quality was not necessarily associated with lower diet GHGE. Higher-Quality diets were characterized by a predominance of
Downloaded from ajcn.nutrition.org at MCMASTER UNIVERSITY on January 5, 2015
FIGURE 1. Daily percentage variation between the sex-specific means of the whole population [0 on the vertical axis; n (men/women) = 776/1142] and means for the Lower-Carbon (n = 388/571), Higher-Quality (n = 388/569), and More Sustainable (n = 181/229) diets for energy intake (A), solid weight (B), greenhouse gas emissions (C), PANDiet score (D), total diet cost (E), and percentage of energy from plant-based foods (F). Analysis of means method: *P , 0.05, **P , 0.01, ***P , 0.001. Lower-Carbon diets were those with diet-related greenhouse gas emissions under the overall median (4511 and 3437 g CO2 equivalents in men and women, respectively); Higher-Quality diets were those with a probability of adequate nutrient intake score above the overall median (62.0 in both men and women); More Sustainable diets were those combining the Lower-Carbon and Higher-Quality criteria. A total of 411 participants were not classified in the Lower-Carbon, Higher-Quality, or More Sustainable diets. PANDiet, probability of adequate nutrient intake.
191 84 28 32 46 202 90 71 41 354 69 146 78 58 2.1 301 221 13 67 57 15 23 19 213 131 82 1476 768 85 368 256 129 3.4 122 4.5
6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 2.8 1.9 1.1 1.2 1.3 8.1 7.1 3.2 1.4 9.1 3.7 6.8 2.4 2.2 0.2 4.8 3.7 1.0 2.1 1.5 0.6 0.7 1.1 5.6 4.1 4.6 24.1 22.2 7.0 10.2 10.2 3.1 0.3 2.9 0.7
155 71 27 26 30 175 74 66 35 319 64 128 71 55 2.1 280 207 12 61 57 13 23 21 191 112 79 1303 678 74 379 171 119 2.3 113 4.4
6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 2.7*** 2.1*** 1.4 1.0*** 1.1*** 8.0* 7.0 3.7 1.2*** 10.4*** 6.0 7.2** 3.1** 2.7 0.3 6.4*** 5.6** 1.2 2.5** 2.3 0.7* 1.0 1.9 7.3** 5.2*** 5.9 32.5*** 28.3** 9.1 14.3 14.2*** 3.5** 0.4*** 3.3** 0.9
Lower-Carbon (n = 388) 189 83 34 28 44 222 97 84 41 456 84 205 92 72 2.8 342 255 15 72 59 13 26 20 224 123 101 1535 859 61 381 233 125 3.2 116 6.6
6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 3.6 2.3 1.5*** 1.8** 1.5 8.6* 7.6 5.4*** 2.0 14.0*** 5.6*** 11.4*** 3.3*** 3.7*** 0.3*** 6.9*** 5.5*** 1.3* 3.6* 2.4 0.8*** 1.1*** 1.9 9.2 6.4 8.4*** 35.1* 32.1*** 8.7** 13.8 14.1 4.4 0.3 4.2 1.3**
Higher-Quality (n = 388) 158 72 33 25 28 195 86 74 35 428 84 184 86 72 2.5 329 248 15 66 62 11 27 24 190 93 97 1338 753 46 398 140 113 1.8 105 6.3
6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 4.0*** 3.0*** 2.1* 1.5*** 1.8*** 9.4 9.6 5.0 2.0** 14.0*** 9.7 10.1*** 4.2* 4.5*** 0.3 10.2** 8.7*** 1.9 4.3 4.1 1.0*** 1.5** 3.8 11.6* 8.2*** 10.2 44.2*** 38.4 10.7*** 21.4 14.6*** 5.8** 0.3*** 5.5** 1.6
More Sustainable (n = 181) 138 60 28 18 31 195 82 87 27 358 75 143 81 57 1.6 198 139 8.8 50 57 13 23 20 172 87 85 1344 807 53 421 63 120 2.8 112 5.3
6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 2.0 1.3 0.8 0.6 0.8 4.7 4.0 2.3 0.7 7.6 3.3 4.9 1.9 1.7 0.1 3.2 2.3 0.6 1.6 2.2 0.5 0.6 2.1 4.2 1.9 3.8 21.5 19.0 5.9 16.4 3.5 2.7 0.2 2.6 0.5
Average (n = 1142) 109 51 23 15 21 161 66 74 21 295 67 112 68 47 1.4 178 124 8.5 46 49 11 20 18 150 74 76 1197 721 47 388 41 112 2.0 105 4.7
6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 2.0*** 1.3*** 0.9*** 0.6*** 0.8*** 5.5*** 4.6*** 3.3** 0.8*** 8.0*** 3.7* 4.4*** 2.7*** 2.3*** 0.2 5.0*** 3.0*** 1.0 2.4** 2.6*** 0.6*** 0.8*** 2.3 5.1*** 2.8*** 4.5* 30.3*** 29.0*** 6.6 16.9* 2.8*** 3.3** 0.3** 3.3** 0.6
Lower-Carbon (n = 571)
Downloaded from ajcn.nutrition.org at MCMASTER UNIVERSITY on January 5, 2015
143 62 33 16 31 232 98 108 26 454 92 189 100 71 2.2 216 151 9.1 56 62 11 26 24 177 75 102 1431 892 32 448 59 116 1.9 105 8.4
6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 2.4** 1.6 1.3*** 0.8*** 1.2 6.8*** 6.3*** 3.6*** 1.0 9.9*** 4.9*** 7.7*** 3.1*** 2.4*** 0.3*** 4.6*** 3.5*** 0.8 2.4** 4.1* 0.6*** 0.8*** 3.9 6.5 2.6*** 5.8*** 29.5*** 29.7*** 4.8** 19.0* 4.4 3.7 0.2*** 3.7** 1.0***
Higher-Quality (n = 569)
Women
110 51 28 12 19 198 80 96 22 388 82 157 88 60 2.0 197 139 8.5 49 56 8.4 23 24 160 61 99 1350 851 33 428 38 109 1.6 99 8.6
6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6
2.3 *** 1.7 *** 1.7 0.9 *** 1.2 *** 8.4 8.5 4.9 1.1*** 11.6* 6.1 6.5 5.0 4.3 0.3 6.1 4.2 1.2 4.2 5.5 0.6*** 1.1 5.2 8.3 3.5*** 7.7 52.2 55.4 5.8** 29.3 4.0*** 4.3** 0.2*** 4.5** 1.3**
More Sustainable (n = 229)
1 All values are means 6 SEs. The Average diet represents mean intakes derived from the sex-specific whole population, the Lower-Carbon diets were those with diet-related GHGE under the overall median (4511 and 3437 g CO2eq in men and women, respectively), the Higher-Quality diets were those with a probability of adequate nutrient intake score above the overall median (62.0 in both men and women), and the More Sustainable diets were those combining the Lower-Carbon and Higher-Quality criteria. A total of 181 men and 230 women were not classified in the Lower-Carbon, Higher-Quality, or More Sustainable diets. Difference between the Average diet and the Lower-Carbon, Higher-Quality, and More Sustainable diets: *P , 0.05, **P , 0.01, ***P , 0.001 (analysis of means). CO2eq, carbon dioxide equivalent; GHGE, greenhouse gas emissions.
Meat, fish, and eggs Pork, poultry, eggs Fish and fish products Deli meats Ruminant meat Dairy products Milk Yogurt Cheese Fruit, vegetables, and nuts Processed fruit and juices Fresh fruit Cooked vegetables Raw vegetables Dried fruit and nuts Starchy foods Grains Legumes Poatoes Added fats and condiments Butter, cream Oils and margarine Sauces and condiments Mixed dishes With ingredients of animal origin Vegetarian Drinks Water Soft drinks Hot drinks Alcoholic drinks Foods high in fat, salt, and sugar Salty snacks Desserts Breakfast cereals
Average (n = 776)
Men
TABLE 2 Food group and subgroup content (g) of the Average (whole sample), Lower-Carbon, Higher-Quality, and More Sustainable diets1
IDENTIFYING THE MOST SUSTAINABLE FRENCH DIETS
1465
1466
MASSET ET AL
Downloaded from ajcn.nutrition.org at MCMASTER UNIVERSITY on January 5, 2015 FIGURE 2. Energy contribution (A) and greenhouse gas emission contribution (B) of all main food groups to the Average [n (men/women) = 776/1142] and the More Sustainable (n = 181/229) diets. Analysis of means method: *P , 0.05, **P , 0.01, ***P , 0.001. The Average diet represents mean intakes derived from the whole population; More Sustainable diets were those combining the Lower-Carbon (daily GHGE below the overall median) and Higher-Quality (PANDiet nutritional quality score above the overall median) criteria. GHGE, greenhouse gas emissions; PANDiet, probability of adequate nutrient intake.
low-energy-density foods such as starches, fruit, and vegetables that need to be eaten in large quantities to meet energy requirements (42). This explains why, despite their low GHGE per 100 g (43), Higher-Quality diets had higher GHGE than did the Average diets, both per day (men and women) and per 1000 kcal (women only). Conversely, the lower nutritional quality of the Lower-Carbon diets may have been partly attributable to their
lower energy content. Men and women with More Sustainable diets had 19% and 17% lower daily GHGE compared with the population average, respectively. These figures are close to the 20% GHGE reduction target set in the European Union 2020 roadmap (4), and this reduction was associated with relatively minor diet changes compared with some changes proposed in previous studies (10–13, 15, 17).
IDENTIFYING THE MOST SUSTAINABLE FRENCH DIETS
(47). To include such indicators in the dietary assessment, planning, or modeling analyses, geolocalized data at the food level are needed. A strength of this study was that most dimensions of sustainable diets, as defined by the FAO (1), were considered. Cultural acceptability could not be measured but was ensured through the analysis of self-selected diets. The diet cost was used as a proxy for affordability; the More Sustainable diets presented no extra cost compared with the population average. Nutritional quality was assessed by using the comprehensive PANDiet index. Variations in total intake may result from different energy needs or from diet misreporting. For instance, the men with a More Sustainable diet were older than the overall population, which possibly explained their lower energy intakes. However, age-adjusted analyses did not change the conclusions (data not shown). In addition, people seem to have more difficulty in reducing portion sizes than in modifying their diet structure (48). Hence, the results regarding GHGE reduction in More Sustainable participants should be considered cautiously, and reaching the 20% reduction target may not be feasible with minor dietary changes if the energy intake is kept constant. However, moderation of total food intake may have a beneficial public health effect, considering the rising prevalence of overweight and obesity (49). Reducing the energy intake while keeping the weight intake constant can be achieved by reducing the energy density of the diet, and the World Cancer Research Fund recommends reducing the solid energy density of the diet (50). In epidemiologic studies, a low energy density correlates with a low energy intake (51, 52), which highlights the relation between these 2 factors and the relevance of combining them. To better understand the motivations of the individuals with More Sustainable diets and to help policymakers design more efficient public health strategies, further studies should focus on the determinants and perceived barriers for dietary choices (53). In contrast with our conservative approach, diet scenario analyses allow for assessments with more flexibility. The diet optimization approach used in the United Kingdom, France, Spain, Sweden, and New Zealand (17–19) could represent a valuable tool in this respect, provided that sound acceptability constraints ensure some practicality when designing low-GHGE and nutritionally adequate diets. The current analysis showed that approximately one-fifth of French adults achieved a sustainable diet with a higher nutritional quality and GHGE decreased by almost 20% at no extra cost compared with the population average. These encouraging results confirm that reducing diet-related GHGE and increasing nutritional adequacy is possible through frugality and wiser dietary choices such as a lower intake of meats and alcoholic drinks, higher intake of plant-based foods, and moderate total food intake. However, in these observed diets the compatibility between diet healthiness and low GHGE was not straightforward, which indicated that the development of sustainable dietary guidelines needs to include both nutritional and environmental dimensions to effectively ensure that consumers opt for More Sustainable diets. We thank all of the participants in the INCA2 study and the INCA2 study team for providing the data and necessary support. The authors’ responsibilities were as follows—GM: analyzed the data and drafted the manuscript; GM, ND, L-GS, EOV, and FV: assisted in drafting the manuscript; FV, EOV, and DT: provided the essential materials for the analyses; and ND: had primary responsibility for the final content. All of the
Downloaded from ajcn.nutrition.org at MCMASTER UNIVERSITY on January 5, 2015
Two main factors were identified to result in More Sustainable diets: reduced energy intake and reduced energy density. Indeed, when expressed per kcal, GHGE were reduced by w10% in the More Sustainable diets, which suggests that half of the 20% total GHGE reduction could be attributed to a quantity effect and half to a structural effect. First, the total energy intake was 8% and 10% lower for the More Sustainable diets than for the Average diets in men and women, respectively. Energy intake was thus confirmed as a main driver of dietary GHGE (42). Second, More Sustainable diets had a lower energy density than the Average diets, which accounted for a modified diet structure. The magnitude of the differences was modest, but most differences were statistically significant. The More Sustainable diets contained the highest content of plant-based foods, particularly starchy foods. These results reinforce the rationale for increasing the consumption of plant-based foods and reducing that of animal products to achieve more environment-friendly and healthy diets (10, 13, 15–17). In addition to foods of animal origin, alcoholic beverage consumption was highly associated with dietary GHGE, which strengthened previous findings (14, 44). A better understanding of the environmental pressure associated with alcohol consumption could provide extra support for policymakers in the attempt to limit alcohol consumption in the general population. This study had several limitations. The More Sustainable diets were not sustainable in an absolute sense. They were only defined according to the combination of 2 sustainability criteria (lower GHGE and higher nutritional quality), and those criteria were relative (ie, considered with regard to the respective sex-specific medians) and thus relied on the INCA2 data for comparison. Nutritional quality was assessed by using the PANDiet index, a score based on nutrient recommendations only, without preconceived views on the dietary structure of Higher-Quality diets. The mean PANDiet score for all INCA2 participants was comparable with that in the French Etude Nationale Nutrition Sante´ study and slightly higher than that in US NHANES (27). In contrast with nutrient recommendations, there are no reference values to define low-GHGE diets. We opted for a pragmatic approach that could be applied to any population and sustainability criteria. Diet-related GHGE were assessed by using the internationally standardized life cycle analysis method, and our estimates were comparable with previous estimates based on consumption data (22). However, our estimates did not include emissions arising from indirect land use change and did not account for food losses and waste and may, therefore, have underestimated the actual values (45, 46); such underestimation could explain our lower GHGE estimates when compared with some estimates that included additional factors (12–15, 17). According to the FAO food balance sheet data, the mean food supply for France was 3520 kcal per capita per day in 2007. The INCA2 participants consumed 2162 kcal/d on average, which gave a very gross estimate of 39% for waste at the national level. In the current study, the environmental dimension of sustainability was assessed only through GHGE. A previous study showed that food GHGEs strongly correlate with 2 other indicators of environmental impact (water eutrophication and air acidification) (43), which suggests that the current results would also apply for these indicators. However, food production was shown to account for the vast majority of the global water footprint (9) and is intrinsically linked to ecosystem biodiversity
1467
1468
MASSET ET AL
authors read and approved the final manuscript. No conflicts of interest were reported. 20.
REFERENCES 21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
Downloaded from ajcn.nutrition.org at MCMASTER UNIVERSITY on January 5, 2015
1. Food and Agriculture Organization. Annex I: International scientific symposium biodiversity and sustainable diets—final document. In: Burlingame B, Dernini S, eds. Sustainable diets and biodiversity— directions and solutions for policy, research and action. Rome, Italy: FAO, 2012:294. 2. Tubiello FN, Salvatore M, Rossi S, Ferrara A, Fitton N, Smith P. The FAOSTAT database of greenhouse gas emissions from agriculture. Environ Res Lett 2013;8:015009. 3. Smith P, Martino D, Cai Z, Gwary D, Janzen H, Kumar P, McCarl B, Ogle S, O’Mara F, Rice C, et al. Agriculture. In: Metz B, Davidson O, Bosch P, Dave R, Meyer L, eds. Climate change 2007: mitigation. Contribution of Working Group III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. New York, NY: Cambridge University Press, 2007. 4. European Parliament, Council of the European Union. Decision no 406/2009/Ec of the European Parliament and of the Council of 23 April 2009 on the effort of Member States to reduce their greenhouse gas emissions to meet the Community’s greenhouse gas emission reduction commitments up to 2020. Brussels, Belgium: European Parliament, Council of the European Union, 2012:136–48. 5. Legislation UK. Climate Change Act 2008, code 2008 c. 27. London, United Kingdom: HMSO, 2008. 6. Steinfeld H, Gerber P, Wassenaar T, Castel V, Rosales M, de Haan C. Livestock’s long shadow—environmental issues and options. Rome, Italy: FAO, 2006. 7. Carlsson-Kanyama A, Gonza´lez AD. Potential contributions of food consumption patterns to climate change. Am J Clin Nutr 2009;89 (suppl):1704S–9S. 8. Walle´n A, Brandt N, Wennersten R. Does the Swedish consumer’s choice of food influence greenhouse gas emissions? Environ Sci Policy 2004;7:525–35. 9. Vanham D, Bidoglio G. A review on the indicator water footprint for the EU28. Ecol Indic 2013;26:61–75. 10. Stehfest E, Bouwman L, van Vuuren DP, den Elzen MGJ, Eickhout B, Kabat P, Vuuren DP, Elzen MGJ. Climate benefits of changing diet. Clim Change 2009;95:83–102. 11. Committee on Climate Change. Reducing emissions from agriculture and land use, land-use change and forestry. In: The fourth carbon budget—reducing emissions through the 2020s. London, United Kingdom: Committee on Climate Change, 2010:295–329. 12. Audsley E, Brander M, Chatterton J, Murphy-Bokern D, Webster C, Williams A. How low can we go? An assessment of greenhouse gas emissions from the UK food system and the scope for reduction by 2050. Godalming, United Kingdom: FCRN-WWF-UK, 2010. 13. Tukker A, Goldbohm RA, de Koning A, Verheijden M, Kleijn R, Wolf O, Pe´rez-Domı´nguez I, Rueda-Cantuche JM, Perez-Dominguez I. Environmental impacts of changes to healthier diets in Europe. Ecol Econ 2011;70:1776–88. 14. Saxe H, Larsen TM, Mogensen L. The global warming potential of two healthy Nordic diets compared with the average Danish diet. Clim Change 2013;116:249–62. 15. Berners-Lee M, Hoolohan C, Cammack H, Hewitt CNN. The relative greenhouse gas impacts of realistic dietary choices. Energy Policy 2012;43:184–90. 16. Friel S, Dangour AD, Garnett T, Lock K, Chalabi Z, Roberts I, Butler A, Butler CD, Waage J, McMichael AJ, et al. Public health benefits of strategies to reduce greenhouse-gas emissions: food and agriculture. Lancet 2009;374:2016–25. 17. Macdiarmid JI, Kyle J, Horgan GW, Loe J, Fyfe C, Johnstone A, McNeill G. Sustainable diets for the future: can we contribute to reducing greenhouse gas emissions by eating a healthy diet? Am J Clin Nutr 2012;96:632–9. 18. Wilson N, Nghiem N, Ni Mhurchu C, Eyles H, Baker MG, Blakely T. Foods and dietary patterns that are healthy, low-cost, and environmentally sustainable: a case study of optimization modeling for New Zealand. PLoS ONE 2013;8:e59648. 19. Thompson S, Gower R, Darmon N, Vieux F, Murphy-Bokern D, Maillot M. A balance of healthy and sustainable food choices for
France, Spain, and Sweden. Godalming, United Kingdom: WWF-UK, 2013. Briggs ADM, Kehlbacher A, Tiffin R, Garnett T, Rayner M, Scarborough P. Assessing the impact on chronic disease of incorporating the societal cost of greenhouse gases into the price of food: an econometric and comparative risk assessment modelling study. BMJ Open 2013;3:e003543. Scarborough P, Allender S, Clarke D, Wickramasinghe K, Rayner M. Modelling the health impact of environmentally sustainable dietary scenarios in the UK. Eur J Clin Nutr 2012;66:710–5. Aston LM, Smith JN, Powles JW. Impact of a reduced red and processed meat dietary pattern on disease risks and greenhouse gas emissions in the UK: a modelling study. BMJ Open 2012;2:e001072 . Rolls BJ, Drewnowski A, Ledikwe JH. Changing the energy density of the diet as a strategy for weight management. J Am Diet Assoc 2005; 105(suppl):S98–103. Bes-Rastrollo M, van Dam RM, Martinez-Gonzalez MA, Li TY, Sampson LL, Hu FB. Prospective study of dietary energy density and weight gain in women. Am J Clin Nutr 2008;88:769–77. Vergnaud A-C, Estaquio C, Czernichow S, Pe´neau S, Hercberg S, Galan P, Bertrais S. Energy density and 6-year anthropometric changes in a middle-aged adult cohort. Br J Nutr 2009;102:302–9. Vieux F, Soler L-G, Touazi D, Darmon N. High nutritional quality is not associated with low greenhouse gas emissions in self-selected diets of French adults. Am J Clin Nutr 2013;97:569–83. Verger EO, Mariotti F, Holmes BA, Paineau D, Huneau J-F. Evaluation of a diet quality index based on the probability of adequate nutrient intake (PANDiet) using National French and US Dietary Surveys. PLoS ONE 2012;7:e42155. Dubuisson C, Lioret S, Touvier M, Dufour A, Calamassi-Tran G, Volatier J-L, Lafay L. Trends in food and nutritional intakes of French adults from 1999 to 2007: results from the INCA surveys. Br J Nutr 2010;103:1035–48. Black AE. Critical evaluation of energy intake using the Goldberg cutoff for energy intake:basal metabolic rate. A practical guide to its calculation, use and limitations. Int J Obes Relat Metab Disord 2000; 24:1119–30. Ledikwe JH, Blanck HM, Khan LK, Serdula MK, Seymour JD, Tohill BC, Rolls BJ. Dietary energy density determined by eight calculation methods in a nationally representative United States population. J Nutr 2005;135: 273–8. Kantar Worldpanel. French household consumer panel—Kantar worldpanel. Available from: http://www.kantarworldpanel.com/global/Sectors (cited 5 May 2013). Martin A, ed. Apports nutritionnels conseille´s pour la population franc¸aise. [Recommended nutritional intakes for the French population.] 3rd ed. Paris, France: Lavoisier, 2001. (in French). EC Scientific Committee for Food. Nutrient and energy intakes for the European Community (opinion expressed on 11 December 1992). Luxembourg, Luxembourg: EC Scientific Committee for Food, 1993. EFSA Panel on Dietetic Products, Nutrition, and Allergies. Scientific Opinion on Dietary Reference Values for carbohydrates and dietary fibre. EFSA J 2010;8(3). International Standard. ISO 14040:2006 environmental management— life cycle assessment—principles and framework. Geneva, Switzerland: ISO, 2006. International Standard. ISO 14044:2006 Environmental management— life cycle assessment—requirements and guidelines. Geneva, Switzerland: ISO, 2006. National Institute of Statistics and Economic Studies (Insee). Definitions and methods—statistical operation: survey on industrial energy consumption (EACEI). Available from: http://www.insee.fr/en/methodes/ default.asp?page=sources/ope-enq-conso-energie-industrie-eacei.htm (cited 10 May 2013). French Department of Ecology Sustainable Development and Energy. SitraM database. Available from: http://www.statistiques.developpementdurable.gouv.fr/donnees-ligne/r/flux-marchandises-sitram-i.html?tx_ttnews [tt_news]=20519&cHash=a891e4085d89a9486f97d0282957ec1a (cited 10 May 2013). Althaus H, Doka G, Dones R, Heck T, Hellweg S, Hischier R, Nemecek T, Rebitzer G, Spielmann M, Wernet G. In: Frischknecht R, Jungbluth N, eds. Overview and Methodology - Data v2.0 - ecoinvent report No. 1. Du¨bendorf, CH: Ecoinvent, 2007. Greenext. Engagements & me´thodologie. (Commitments & methods.) Available from: http://www.greencode-info.fr/ (cited 12 November 2012).
IDENTIFYING THE MOST SUSTAINABLE FRENCH DIETS 41. Nelson PR. Additional uses for the analysis of means and extended tables of critical values. Technometrics 1993;35:61–71. 42. Vieux F, Darmon N, Touazi D, Soler LG. Greenhouse gas emissions of self-selected individual diets in France: changing the diet structure or consuming less? Ecol Econ 2012;75:91–101. 43. Masset G, Soler L-G, Vieux F, Darmon N. Identifying sustainable foods: the relationship between environmental impact, nutritional quality, and prices of foods representative of the French diet. J Acad Nutr Diet 2014 (in press). 44. Meier T, Christen O, Semler E, Jahreis G, Voget-Kleschin L, Schrode A, Artmann M. Balancing virtual land imports by a shift in the diet. Using a land balance approach to assess the sustainability of food consumption. Germany as an example. Appetite 2014;74:20–34. 45. Food and Agriculture Organization. Food wastage footprint—impacts on natural resources. Rome, Italy: FAO, 2013. 46. Schmidinger K, Stehfest E. Including CO2 implications of land occupation in LCAs—method and example for livestock products. Int J Life Cycle Assess 2012;17:962–72.
1469
47. FAO. Building on gender, agrobiodiversity and local knowledge— a training manual. Rome, Italy: FAO, 2005. 48. Macdiarmid JI, Loe J, Kyle J, McNeill G. “It was an education in portion size.” Experience of eating a healthy diet and barriers to long term dietary change. Appetite 2013;71:411–9. 49. Hill JO, Peters JC, Wyatt HR. Using the energy gap to address obesity: a commentary. J Am Diet Assoc 2009;109:1848–53. 50. World Cancer Research Fund and American Institute for Cancer Research. Food, nutrition, physical activity and the prevention of cancer: a global perspective. Washington, DC: World Cancer Research Fund and American Institute for Cancer Research, 2007. 51. Westerterp-Plantenga MS. Modulatory factors in the effect of energy density on energy intake. Br J Nutr 2004;92(suppl):S35–9. 52. Stubbs J, Ferres S, Horgan G. Energy density of foods: effects on energy intake. Crit Rev Food Sci Nutr 2000;40:481–515. 53. Ma¨kiniemi J-P, Vainio A. Barriers to climate-friendly food choices among young adults in Finland. Appetite 2014;74:12–9.
Downloaded from ajcn.nutrition.org at MCMASTER UNIVERSITY on January 5, 2015
Reducing energy intake and energy density for a sustainable diet: a study based on self-selected diets in French adults1–3 Gabriel Masset, Florent Vieux, Eric Olivier Verger, Louis-Georges Soler, Djilali Touazi, and Nicole Darmon
INTRODUCTION
“Sustainable diets,” as defined by the FAO, need to meet nutritional (health), environmental, affordability, and cultural acceptability criteria (1) based on the sustainable development framework. According to the latest FAO data and the Intergovernmental Panel on Climate Change, agriculture and deforestation accounted for approximately one-third of global greenhouse gas emissions
1460
(GHGE)4 in 2010 (2, 3). Therefore, a reduction in food-related GHGE is needed to achieve national and international GHGE reduction targets (4, 5). The production and consumption of foods of animal origin, particularly ruminant meat, result in increased emissions of greenhouse gases (6, 7) and increased energy (8) and water (9) use relative to plant-based foods. Consequently, experts in environmental sciences argue that diets containing reduced amounts of animal-origin foods would produce less GHGE (10–12). It was estimated that several healthy dietary patterns, such as the Mediterranean (13) or the Nordic diet (14), or diets designed to have a low amount of meat (15, 16) produce less GHGE than do currently observed Western diets. According to studies based on diet modeling, reducing diet-related GHGE while ensuring nutritional adequacy is theoretically feasible (17–20). Two studies based on predictive public health models have strengthened the hypothesis that low-GHGE diets containing a lower content of animal-origin foods and a higher content of plantbased foods could prevent chronic disease and mortality (21, 22). Furthermore, increased consumption of plant-based foods could help reduce the energy density of diets (23) and may therefore help to control total energy intake, which is an important driver of the rising prevalence of obesity (24, 25). 1
From the Unite´ Mixte de Recherche “Nutrition, Obesity and Risk of Thrombosis,” Institut National de la Recherche Agronomique, Institut National de la Sante´ et de la Recherche Me´dicale, Aix- Marseille Universite´, Marseille, France (GM, FV, and ND); AgroParisTech, Nutrition Physiology and Ingestive Behavior, Paris, France and Danone Research, Global Nutrition Department, Palaiseau, France (EOV); and UR “Aliss,” Ivry sur Seine, France (L-GS and DT). 2 Supported by the French National Research Agency under the OCAD project ANR-11-ALID-0002 and the French Environment and Energy Management Agency (ADEME). FV received a PhD grant from the INRA divisions of Nutrition, Chemical Food Safety, and Consumer Behavior and of Social Sciences, Agriculture and Food, Rural Development, and Environment. EOV received a PhD grant from Danone Research and from the Association Nationale de la Recherche et de la Technologie (CIFRE 474/2010). 3 Address correspondence to N Darmon, UMR 1062 INSERM/1260 INRA/ Universite´ d’Aix-Marseille, Nutrition Obe´site´ et Risque Thrombotique, Faculte´ de Me´decine de la Timone, 27 Boulevard Jean Moulin, 13385 Marseille Cedex 05, France. E-mail: [email protected]. 4 Abbreviations used: GHGE, greenhouse gas emissions; INCA2, Second Individual and National Survey on Food Consumption; PANDiet, probability of adequate nutrient intake. Received October 16, 2013. Accepted for publication March 5, 2014. First published online April 2, 2014; doi: 10.3945/ajcn.113.077958.
Am J Clin Nutr 2014;99:1460–9. Printed in USA. Ó 2014 American Society for Nutrition Supplemental Material can be found at: http://ajcn.nutrition.org/content/suppl/2014/05/13/ajcn.113.0 77958.DCSupplemental.html
Downloaded from ajcn.nutrition.org at MCMASTER UNIVERSITY on January 5, 2015
ABSTRACT Background: Studies on theoretical diets are not sufficient to implement sustainable diets in practice because of unknown cultural acceptability. In contrast, self-selected diets can be considered culturally acceptable. Objective: The objective was to identify the most sustainable diets consumed by people in everyday life. Design: The diet-related greenhouse gas emissions (GHGE) for self-selected diets of 1918 adults participating in the cross-sectional French national dietary survey Individual and National Survey on Food Consumption (INCA2) were estimated. “Lower-Carbon,” “Higher-Quality,” and “More Sustainable” diets were defined as having GHGE lower than the overall median value, a probability of adequate nutrition intake (PANDiet) score (a measure of the overall nutritional adequacy of a diet) higher than the overall median value, and a combination of both criteria, respectively. Diet cost, as a proxy for affordability, and energy density were also assessed. Results: More Sustainable diets were consumed by 23% of men and 20% of women, and their GHGE values were 19% and 17% lower than the population average (mean) value, respectively. In comparison with the average value, Lower-Carbon diets achieved a 20% GHGE reduction and lower cost, but they were not sustainable because they had a lower PANDiet score. Higher-Quality diets were not sustainable because of their above-average GHGE and cost. More Sustainable diets had an above-average PANDiet score and a below-average energy density, cost, GHGE, and energy content; the energy share of plant-based products was increased by 20% and 15% compared with the average for men and women, respectively. Conclusions: A strength of this study was that most of the dimensions for “sustainable diets” were considered, ie, not only nutritional quality and GHGE but also affordability and cultural acceptability. A reduction in diet-related GHGE by 20% while maintaining high nutritional quality seems realistic. This goal could be achieved at no extra cost by reducing energy intake and energy density and increasing the share of plant-based products. Am J Clin Nutr 2014;99:1460–9.
IDENTIFYING THE MOST SUSTAINABLE FRENCH DIETS
MATERIALS AND METHODS
Population sample and dietary data The dietary intakes were derived from the 7-d food records of a nationally representative random sample of adults (n = 2624; age .18 y) participating in the INCA2 cross-sectional dietary survey Enqueˆte Individuelle et Nationale sur les Consommations Alimentaires (Individual and National Survey on Food Consumption), conducted in 2006–2007 by ANSES (French Agency for Food, Environmental and Occupational Health Safety) (28). The sampling method involved a 3-stage stratified random sampling. To ensure the representative nature of the sample, a statistical adjustment was performed for the region, town size, age, sex, occupation of the household head, household size, and seasonal variables. After underreporters were excluded by using the Black equations (29), the current analysis was conducted on a final sample of 1918 adults (776 men and 1142 women). The study was approved by the French Data Protection Authority Commission (Commission Nationale Informatique et Liberte´s). All of the foods declared as consumed by the participants during the survey (n = 1314 foods and beverages, including water) were categorized into 8 food groups and 27 food subgroups within a food nutrient composition database associated with the survey. In addition, to assess the share of plant-based foods in the diets, plant-based foods were identified as fruit, vegetables, and nuts; cereals, including breakfast cereals; potatoes; legumes; vegetable oils and margarines; and mixed dishes without meat, fish products, or eggs. The total weight and energy intake (in kcal/d) were calculated for each participant by summing the quantity and energy content of all of the foods and beverages consumed by the individual. The diet weight (in g/d) for the solid foods was calculated by excluding foods typically consumed as beverages, such as milk, juices, and other drinks, as proposed by Ledikwe et al (30). The dietary energy density was then calculated (in kcal/100 g) by dividing the energy intake from solid food by the solid diet weight. Food prices were obtained from the 2006 Kantar-World Panel purchase database (31), which gives the annual expenditure and quantity purchased for each food item available on the market in
a representative sample of 12,000 French households. The mean prices were calculated by dividing the annual expenditures by the quantities purchased. Food prices were expressed in euros per 100 g of edible food (ie, after the changes in weight associated with the trimming or cooking processes were taken into account by using the appropriate conversion factors). The diet costs per day (€/d) were then estimated by multiplying the reported edible weight of each food/beverage by its unit cost and summing over all foods/beverages consumed by that individual. In contrast with energy density, the diet cost was calculated accounting for all of the foods consumed, including beverages. The diet costs per 1000 kcal were also estimated. The PANDiet index for nutritional quality and identification of Higher-Quality diets The adequacy of nutrient intake for each individual was calculated by using the PANDiet index, based solely on nutrients. The PANDiet is composed of an “adequacy” subscore that includes positive nutrients and a “moderation” subscore for the nutrients to be limited, as described in full elsewhere (27). Briefly, the algorithm calculates the probability of adequacy for each nutrient on a scale from 0 to 1, where 1 represents a 100% probability that the usual intake is adequate by satisfying the requirement or is not excessive, according to a reference value. The probability calculation accounts for the number of days of diet recording, average intake, and day-to-day variability in intake, nutrient reference value, and interindividual variability. The reference values were the French nutritional recommendations for adults (32) or the European Union values (33, 34) when specific recommendations were lacking (see Supplementary Table 1 under “Supplemental data” in the online issue). The PANDiet index generates a summary score for all of these probabilities, ranging from 0 to 100; a higher score reflects better nutritional adequacy of the diet (27). Hence, the PANDiet index accounts for day-to-day variability in nutrient intake and does not generate a priori hypotheses for the dietary content of the diet. The PANDiet has been shown to relate to other indicators of nutritional quality among participants in the French Etude Nationale Nutrition Sante´ and US NHANES dietary surveys (27). In the current study, Higher-Quality diets were defined as those with a PANDiet score higher than the sexspecific median score. Estimating diet-related GHGE and identifying Lower-Carbon diets The estimation of diet-related GHGE was based on the selection of 391 foods among those most consumed by INCA2 participants, as previously described by Vieux et al (26). Briefly, within each food subgroup, the food items were ranked in decreasing order based on the percentage of participants who consumed them, and at least one food item was selected from among the most widely consumed items in each subgroup. This process identified 391 representative foods from the 1314 items initially listed in the food database. These 391 foods covered 71% of the total weight intake and 66% of the total energy intake in the INCA2 study population. An environmental consulting firm, Greenext Service (Paris, France), assigned the GHGE values for the 391 foods following
Downloaded from ajcn.nutrition.org at MCMASTER UNIVERSITY on January 5, 2015
However, “cultural acceptability”—a key dimension of the FAO’s definition of “sustainable diets” (1)—was typically not considered in the studies described above. In an analysis of diets actually consumed by French adults participating in the representative Second Individual and National Survey on Food Consumption (INCA2), Vieux et al (26) did not find an association between a higher nutritional quality and lower GHGE, which thus calls into question the relevance of studies based on theoretical and modeling approaches. To help implement the sustainability concept into practice, this study identified existing diets that combined higher nutritional quality and lower GHGE. Among participants in the nationally representative French INCA2 dietary survey, More Sustainable diets were identified based on diet-related GHGEs derived from a life cycle analysis conducted on almost 400 widely consumed foods (26) and on the probability of adequate nutrient intake (PANDiet) index [a measure of diet quality in terms of global nutrient adequacy (27)].
1461
1462
MASSET ET AL
Statistical analyses: identifying More Sustainable diets and the main drivers for reaching sustainability More Sustainable diets were defined as those with a PANDiet score higher than the median value and diet-related GHGE value lower than the median value. The dietary characteristics of the Higher-Quality, Lower-Carbon, and More Sustainable diets were compared with the mean values for the whole population, referred to below as the Average diets, by using an analysis of means that accounted for the nonindependence of the samples (41). Ageadjusted analyses were conducted for both sexes. All of the analyses were sex-specific because differences in GHGE have been previously observed to arise from differences in energy and quantity intakes between men and women (42). The analyses were performed by using specific survey data procedures (surveymeans and surveyfreq) in SAS 9.3 software (SAS Institute Inc).
RESULTS
The characteristics of the Lower-Carbon (daily GHGE ,4511 and 3437 g CO2 equivalents in men and women, respectively), Higher-Quality (PANDiet higher than 62.0 in men and women), and More Sustainable diets compared with the Average diets are shown in Table 1 for men and women separately. Men and women with a Higher-Quality diet and men with a More Sustainable diet were older than average. The diets of 181 men and 230 women did not fulfill the Lower-Carbon or Higher-Quality criteria and therefore were not classified in any of the 3 identified diet categories. The percentage variations of the selected characteristics between the Average diets and the 3 specific diets are shown in Figure 1. The Lower-Carbon diets had daily GHGE reduced by 20% and 21% compared with the Average diets in men and women, respectively. Their GHGE per kcal were reduced by 8% and 7% in men and women, respectively. Compared with the Average diets, the Lower-Carbon diets contained a lower amount of food and provided less energy intake. The lower total intake was associated with lower nutritional adequacy, as measured by the PANDiet, and a lower diet cost, per day and per kcal.
The Higher-Quality diets contained higher amounts of food, provided more energy, and were associated with higher GHGE per day than were the Average diets. In women only, GHGE were also higher when expressed per 1000 kcal. The Higher-Quality diets had a lower energy density and contained a higher share of plant-based foods than did the average diets. In both sexes, the Higher-Quality diets had the highest cost, both per day and per kcal. The More Sustainable diets were consumed by 181 men and 229 women (23% and 20%, respectively). In comparison with the Average diets, their solid weight was not different, but energy intake was reduced by 8% and 10% in men and women, respectively, and the energy density was also reduced. The daily GHGE for the More Sustainable diets were lower than those for the Average diets (219% in men and 217% in women) but were higher than those for the Lower-Carbon diets. When calculated per kcal, the reduction in GHGE for the More Sustainable diets compared with the Average diets was only 11% in men and 7% in women. The More Sustainable diets had the highest content of plant-based foods, both in energy and in weight. The daily cost for the More Sustainable diets was significantly lower than that for the Average diets, although the cost per kcal was not different. The mean intakes by food group and subgroup for the Average, Lower-Carbon, Higher-Quality, and More Sustainable diets are shown in Table 2. Consistent with a lower diet weight and energy intake, the intakes for all food groups and subgroups were equal or lower for the Lower-Carbon diets than for the Average diets (except for sauces and condiments in men). The greatest reductions were observed for meat, fish, and eggs (w20% in both sexes), especially for ruminant and deli meats and alcoholic drinks (w40% reduction). The Higher-Quality diets contained more fruit, vegetables, and nuts; starchy foods; fresh dairy products; oils and margarine; vegetarian mixed dishes; and breakfast cereals than the Average diets. For the Higher-Quality diets, the total content of meat, fish, and eggs was not different for men and was higher for women relative to the Average diets. In both sexes, the intake of butter, cream, and soft drinks was lower for the Higher-Quality diets than for the Average diets. Similar to the Higher-Quality diets, the intake of starchy foods was higher for the More sustainable diets than for the Average diets. The intake of fruit, vegetables, and nuts was also higher for the More Sustainable diets than for the Average diets. For the More Sustainable diets, the intake of meats and eggs was lower than for the Average diets. The intake of fish and fish products was higher in men and not different in women from that of the Average diets. The intake of dairy products was not different between the More Sustainable and Average diets, with a lower content of cheese in the More Sustainable diets. In addition, the More Sustainable diets had the lowest content of butter, cream, alcoholic drinks, salty snacks, desserts, and mixed dishes with animal ingredients. The percentage energy contribution for all meats was reduced in the More Sustainable diets relative to the Average diets (P , 0.001 in both sexes) (Figure 2A). The energy contribution for dairy products was not different between the More Sustainable and Average diets in men and women. In men, the energy contribution for fish products was increased in the More Sustainable diets. In both sexes, the energy contribution for starchy foods, vegetables, and nuts was higher than average for the More Sustainable diets, whereas the energy contribution for mixed
Downloaded from ajcn.nutrition.org at MCMASTER UNIVERSITY on January 5, 2015
the international ISO 1404x life cycle analysis standards (35, 36) (ie, including the whole life cycle of the foods, from farm production to consumption, and waste management for packaging, excluding emissions arising from indirect land use change and consumers’ transport from retail to home). With use of a top-down approach combining French trade and production data (37, 38) and standard life cycle inventory data [eg, Ecoinvent (39)], the GHGE values assigned by the in-house-developed Greenext life cycle analysis tool reflected the average food products as consumed in the French market (40). To estimate the GHGE associated with total dietary intake (ie, not limited to the 391 selected foods), the consumption of foods for which GHGE values were not estimated was reported based on the GHGE values estimated for foods within the same subgroup (26). Lower-Carbon diets were defined as those with a total dietrelated GHGE lower than the sex-specific median value. Dietary GHGE were also assessed per 1000 kcal consumed to better compare diets with varying amounts of energy.
49.0 25.2 2500 2922 1446 162 62.5 4691 1901 8.63 3.48 52.1 41.6
6 6 6 6 6 6 6 6 6 6 6 6 6 0.67 013 23.2 30.1 16.4 1.12 0.31 49.6 12.8 0.10 0.03 0.48 0.44
50.1 24.9 2210 2599 1297 162 61.8 3764 1743 7.29 3.34 53.3 43.3
6 6 6 6 6 6 6 6 6 6 6 6 6 0.97 0.20* 22.9*** 36.7*** 18.6*** 1.62 0.44* 31.0*** 15.3*** 0.09*** 0.04*** 0.67 0.69***
Lower-Carbon (n = 388) 51.4 25.4 2574 3151 1617 149 68.2 4813 1891 9.04 3.56 57.2 47.1
6 6 6 6 6 6 6 6 6 6 6 6 6 1.03** 0.19* 29.4*** 41.1*** 18.9*** 1.20*** 0.30*** 61.0** 17.3 0.12*** 0.04* 0.66*** 0.65***
Higher-Quality (n = 388) 52.2 24.8 2297 2813 1475 149 68.6 3797 1690 7.77 3.45 59.6 50.1
6 6 6 6 6 6 6 6 6 6 6 6 6 1.46* 0.29* 37.0*** 48.1* 25.3 1.99*** 0.41*** 39.3*** 23.0*** 0.12*** 0.07 0.80*** 1.01***
More Sustainable (n = 181) 45.2 24.0 1855 2582 1238 147 62.3 3548 1953 6.69 3.67 53.0 41.2
6 6 6 6 6 6 6 6 6 6 6 6 6 0.58 0.18 15.0 24.7 11.1 1.03 0.27 35.2 15.0 0.06 0.03 0.43 0.42
Average (n = 1142) 44.0 23.5 1605 2251 1054 150 60.2 2803 1813 5.59 3.57 53.3 41.7
6 6 6 6 6 6 6 6 6 6 6 6 6
0.79 0.19 19.4*** 35.3*** 11.5*** 1.66* 0.42*** 22.7*** 17.6*** 0.08*** 0.04** 0.56 0.60
Lower-Carbon (n = 571)
Downloaded from ajcn.nutrition.org at MCMASTER UNIVERSITY on January 5, 2015
47.6 24.3 1901 2829 1399 133 68.4 3753 2011 7.17 3.84 57.1 45.9
6 6 6 6 6 6 6 6 6 6 6 6 6
0.80*** 0.27 21.0** 31.4*** 15.1*** 1.05*** 0.26*** 50.9*** 23.1*** 0.08*** 0.04*** 0.45*** 0.54***
Higher-Quality (n = 569)
Women
47.0 24.0 1665 2570 1219 134 68.1 2947 1818 6.20 3.79 58.2 47.5
6 6 6 6 6 6 6 6 6 6 6 6 6
1.20 0.25 22.9*** 54.8 16.2 1.59*** 0.42*** 28.3*** 23.8*** 0.13*** 0.07 0.65*** 0.87***
More Sustainable (n = 229)
1 All values are means 6 SEs. The Average diet represents mean intakes derived from the sex-specific whole population, the Lower-Carbon diets were those with diet-related GHGE under the overall median (4511 and 3437 g CO2eq in men and women, respectively), the Higher-Quality diets were those with a probability of adequate nutrient intake score above the overall median (62.0 in both men and women), and the More Sustainable diets were those combining the Lower-Carbon and Higher-Quality criteria. A total of 181 men and 230 women were not classified in the Lower-Carbon, Higher-Quality, or More Sustainable diets. Difference between the Average diet and the Lower-Carbon, Higher-Quality, and More Sustainable diets: *P , 0.05, **P , 0.01, ***P , 0.001 (analysis of means). CO2eq, carbon dioxide equivalent; GHGE, greenhouse gas emissions; PANDiet, probability of adequate nutrient intake. 2 Dietary energy density was calculated by dividing the energy intake from solid food by solid diet weight. 3 As of March 12, 2014: US$1 = 0.72€.
Age (y) BMI (kg/m2) Total energy (kcal/d) Total weight (g/d) Solid weight (g/d) Energy density (kcal/100 g)2 PANDiet GHGE (g CO2eq/d) GHGE (g CO2eq/1000 kcal) Diet cost (€/d)3 Energy cost (€/1000 kcal)3 Weight from plant-based foods (%) Energy from plant-based foods (%)
Average (n = 776)
Men
TABLE 1 Characteristics of the Average (whole sample), Lower-Carbon, Higher-Quality, and More Sustainable diets1
IDENTIFYING THE MOST SUSTAINABLE FRENCH DIETS
1463
1464
MASSET ET AL
dishes and drinks was reduced. No difference was found for foods high in fat, salt, or sugar. Meats, including ruminant meat, were the largest contributor (in %) to daily GHGE in the Average and the More Sustainable diets (Figure 2B). Including fish and dairy products, foods of animal origin (excluding mixed dishes) accounted for w45% of daily GHGE and 25% of total energy in the Average diets. Their contribution to daily GHGE was reduced in the More Sustainable diets because of the significant reduction in ruminant-derived GHGE (P , 0.001 in both sexes). Mixed dishes were also important contributors to dietary GHGE (most likely because of the presence of animal products in many prepared meals), but their contribution to daily GHGE was not different between the Average and More Sustainable diets. The increased content of starchy foods, fruit, vegetables, and nuts in the More Sustainable diets compared with the Average diets (Table 2) was associated with an increase in their contribution to daily GHGE (P , 0.001 in both sexes).
DISCUSSION
The originality of the current study lies in its identification of More Sustainable diets based on existing diets in randomly selected individuals in a nationally representative population sample rather than theoretical diets (10, 13, 21) or special diets, such as those adopted by vegetarian or vegan subjects (15). An important result of the current study was that w20% of French adults had diets that could be called More Sustainable because they combined a higher nutritional quality and lower GHGE with no increase in diet cost. The observed shifts between the Average and the More Sustainable diets were in agreement with most food-based dietary guidelines, which indicated that following the existing public health–oriented recommendations could be compatible with reducing dietary GHGE. However, similarly to Vieux et al (26), we showed that higher nutritional quality was not necessarily associated with lower diet GHGE. Higher-Quality diets were characterized by a predominance of
Downloaded from ajcn.nutrition.org at MCMASTER UNIVERSITY on January 5, 2015
FIGURE 1. Daily percentage variation between the sex-specific means of the whole population [0 on the vertical axis; n (men/women) = 776/1142] and means for the Lower-Carbon (n = 388/571), Higher-Quality (n = 388/569), and More Sustainable (n = 181/229) diets for energy intake (A), solid weight (B), greenhouse gas emissions (C), PANDiet score (D), total diet cost (E), and percentage of energy from plant-based foods (F). Analysis of means method: *P , 0.05, **P , 0.01, ***P , 0.001. Lower-Carbon diets were those with diet-related greenhouse gas emissions under the overall median (4511 and 3437 g CO2 equivalents in men and women, respectively); Higher-Quality diets were those with a probability of adequate nutrient intake score above the overall median (62.0 in both men and women); More Sustainable diets were those combining the Lower-Carbon and Higher-Quality criteria. A total of 411 participants were not classified in the Lower-Carbon, Higher-Quality, or More Sustainable diets. PANDiet, probability of adequate nutrient intake.
191 84 28 32 46 202 90 71 41 354 69 146 78 58 2.1 301 221 13 67 57 15 23 19 213 131 82 1476 768 85 368 256 129 3.4 122 4.5
6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 2.8 1.9 1.1 1.2 1.3 8.1 7.1 3.2 1.4 9.1 3.7 6.8 2.4 2.2 0.2 4.8 3.7 1.0 2.1 1.5 0.6 0.7 1.1 5.6 4.1 4.6 24.1 22.2 7.0 10.2 10.2 3.1 0.3 2.9 0.7
155 71 27 26 30 175 74 66 35 319 64 128 71 55 2.1 280 207 12 61 57 13 23 21 191 112 79 1303 678 74 379 171 119 2.3 113 4.4
6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 2.7*** 2.1*** 1.4 1.0*** 1.1*** 8.0* 7.0 3.7 1.2*** 10.4*** 6.0 7.2** 3.1** 2.7 0.3 6.4*** 5.6** 1.2 2.5** 2.3 0.7* 1.0 1.9 7.3** 5.2*** 5.9 32.5*** 28.3** 9.1 14.3 14.2*** 3.5** 0.4*** 3.3** 0.9
Lower-Carbon (n = 388) 189 83 34 28 44 222 97 84 41 456 84 205 92 72 2.8 342 255 15 72 59 13 26 20 224 123 101 1535 859 61 381 233 125 3.2 116 6.6
6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 3.6 2.3 1.5*** 1.8** 1.5 8.6* 7.6 5.4*** 2.0 14.0*** 5.6*** 11.4*** 3.3*** 3.7*** 0.3*** 6.9*** 5.5*** 1.3* 3.6* 2.4 0.8*** 1.1*** 1.9 9.2 6.4 8.4*** 35.1* 32.1*** 8.7** 13.8 14.1 4.4 0.3 4.2 1.3**
Higher-Quality (n = 388) 158 72 33 25 28 195 86 74 35 428 84 184 86 72 2.5 329 248 15 66 62 11 27 24 190 93 97 1338 753 46 398 140 113 1.8 105 6.3
6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 4.0*** 3.0*** 2.1* 1.5*** 1.8*** 9.4 9.6 5.0 2.0** 14.0*** 9.7 10.1*** 4.2* 4.5*** 0.3 10.2** 8.7*** 1.9 4.3 4.1 1.0*** 1.5** 3.8 11.6* 8.2*** 10.2 44.2*** 38.4 10.7*** 21.4 14.6*** 5.8** 0.3*** 5.5** 1.6
More Sustainable (n = 181) 138 60 28 18 31 195 82 87 27 358 75 143 81 57 1.6 198 139 8.8 50 57 13 23 20 172 87 85 1344 807 53 421 63 120 2.8 112 5.3
6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 2.0 1.3 0.8 0.6 0.8 4.7 4.0 2.3 0.7 7.6 3.3 4.9 1.9 1.7 0.1 3.2 2.3 0.6 1.6 2.2 0.5 0.6 2.1 4.2 1.9 3.8 21.5 19.0 5.9 16.4 3.5 2.7 0.2 2.6 0.5
Average (n = 1142) 109 51 23 15 21 161 66 74 21 295 67 112 68 47 1.4 178 124 8.5 46 49 11 20 18 150 74 76 1197 721 47 388 41 112 2.0 105 4.7
6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 2.0*** 1.3*** 0.9*** 0.6*** 0.8*** 5.5*** 4.6*** 3.3** 0.8*** 8.0*** 3.7* 4.4*** 2.7*** 2.3*** 0.2 5.0*** 3.0*** 1.0 2.4** 2.6*** 0.6*** 0.8*** 2.3 5.1*** 2.8*** 4.5* 30.3*** 29.0*** 6.6 16.9* 2.8*** 3.3** 0.3** 3.3** 0.6
Lower-Carbon (n = 571)
Downloaded from ajcn.nutrition.org at MCMASTER UNIVERSITY on January 5, 2015
143 62 33 16 31 232 98 108 26 454 92 189 100 71 2.2 216 151 9.1 56 62 11 26 24 177 75 102 1431 892 32 448 59 116 1.9 105 8.4
6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 2.4** 1.6 1.3*** 0.8*** 1.2 6.8*** 6.3*** 3.6*** 1.0 9.9*** 4.9*** 7.7*** 3.1*** 2.4*** 0.3*** 4.6*** 3.5*** 0.8 2.4** 4.1* 0.6*** 0.8*** 3.9 6.5 2.6*** 5.8*** 29.5*** 29.7*** 4.8** 19.0* 4.4 3.7 0.2*** 3.7** 1.0***
Higher-Quality (n = 569)
Women
110 51 28 12 19 198 80 96 22 388 82 157 88 60 2.0 197 139 8.5 49 56 8.4 23 24 160 61 99 1350 851 33 428 38 109 1.6 99 8.6
6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6
2.3 *** 1.7 *** 1.7 0.9 *** 1.2 *** 8.4 8.5 4.9 1.1*** 11.6* 6.1 6.5 5.0 4.3 0.3 6.1 4.2 1.2 4.2 5.5 0.6*** 1.1 5.2 8.3 3.5*** 7.7 52.2 55.4 5.8** 29.3 4.0*** 4.3** 0.2*** 4.5** 1.3**
More Sustainable (n = 229)
1 All values are means 6 SEs. The Average diet represents mean intakes derived from the sex-specific whole population, the Lower-Carbon diets were those with diet-related GHGE under the overall median (4511 and 3437 g CO2eq in men and women, respectively), the Higher-Quality diets were those with a probability of adequate nutrient intake score above the overall median (62.0 in both men and women), and the More Sustainable diets were those combining the Lower-Carbon and Higher-Quality criteria. A total of 181 men and 230 women were not classified in the Lower-Carbon, Higher-Quality, or More Sustainable diets. Difference between the Average diet and the Lower-Carbon, Higher-Quality, and More Sustainable diets: *P , 0.05, **P , 0.01, ***P , 0.001 (analysis of means). CO2eq, carbon dioxide equivalent; GHGE, greenhouse gas emissions.
Meat, fish, and eggs Pork, poultry, eggs Fish and fish products Deli meats Ruminant meat Dairy products Milk Yogurt Cheese Fruit, vegetables, and nuts Processed fruit and juices Fresh fruit Cooked vegetables Raw vegetables Dried fruit and nuts Starchy foods Grains Legumes Poatoes Added fats and condiments Butter, cream Oils and margarine Sauces and condiments Mixed dishes With ingredients of animal origin Vegetarian Drinks Water Soft drinks Hot drinks Alcoholic drinks Foods high in fat, salt, and sugar Salty snacks Desserts Breakfast cereals
Average (n = 776)
Men
TABLE 2 Food group and subgroup content (g) of the Average (whole sample), Lower-Carbon, Higher-Quality, and More Sustainable diets1
IDENTIFYING THE MOST SUSTAINABLE FRENCH DIETS
1465
1466
MASSET ET AL
Downloaded from ajcn.nutrition.org at MCMASTER UNIVERSITY on January 5, 2015 FIGURE 2. Energy contribution (A) and greenhouse gas emission contribution (B) of all main food groups to the Average [n (men/women) = 776/1142] and the More Sustainable (n = 181/229) diets. Analysis of means method: *P , 0.05, **P , 0.01, ***P , 0.001. The Average diet represents mean intakes derived from the whole population; More Sustainable diets were those combining the Lower-Carbon (daily GHGE below the overall median) and Higher-Quality (PANDiet nutritional quality score above the overall median) criteria. GHGE, greenhouse gas emissions; PANDiet, probability of adequate nutrient intake.
low-energy-density foods such as starches, fruit, and vegetables that need to be eaten in large quantities to meet energy requirements (42). This explains why, despite their low GHGE per 100 g (43), Higher-Quality diets had higher GHGE than did the Average diets, both per day (men and women) and per 1000 kcal (women only). Conversely, the lower nutritional quality of the Lower-Carbon diets may have been partly attributable to their
lower energy content. Men and women with More Sustainable diets had 19% and 17% lower daily GHGE compared with the population average, respectively. These figures are close to the 20% GHGE reduction target set in the European Union 2020 roadmap (4), and this reduction was associated with relatively minor diet changes compared with some changes proposed in previous studies (10–13, 15, 17).
IDENTIFYING THE MOST SUSTAINABLE FRENCH DIETS
(47). To include such indicators in the dietary assessment, planning, or modeling analyses, geolocalized data at the food level are needed. A strength of this study was that most dimensions of sustainable diets, as defined by the FAO (1), were considered. Cultural acceptability could not be measured but was ensured through the analysis of self-selected diets. The diet cost was used as a proxy for affordability; the More Sustainable diets presented no extra cost compared with the population average. Nutritional quality was assessed by using the comprehensive PANDiet index. Variations in total intake may result from different energy needs or from diet misreporting. For instance, the men with a More Sustainable diet were older than the overall population, which possibly explained their lower energy intakes. However, age-adjusted analyses did not change the conclusions (data not shown). In addition, people seem to have more difficulty in reducing portion sizes than in modifying their diet structure (48). Hence, the results regarding GHGE reduction in More Sustainable participants should be considered cautiously, and reaching the 20% reduction target may not be feasible with minor dietary changes if the energy intake is kept constant. However, moderation of total food intake may have a beneficial public health effect, considering the rising prevalence of overweight and obesity (49). Reducing the energy intake while keeping the weight intake constant can be achieved by reducing the energy density of the diet, and the World Cancer Research Fund recommends reducing the solid energy density of the diet (50). In epidemiologic studies, a low energy density correlates with a low energy intake (51, 52), which highlights the relation between these 2 factors and the relevance of combining them. To better understand the motivations of the individuals with More Sustainable diets and to help policymakers design more efficient public health strategies, further studies should focus on the determinants and perceived barriers for dietary choices (53). In contrast with our conservative approach, diet scenario analyses allow for assessments with more flexibility. The diet optimization approach used in the United Kingdom, France, Spain, Sweden, and New Zealand (17–19) could represent a valuable tool in this respect, provided that sound acceptability constraints ensure some practicality when designing low-GHGE and nutritionally adequate diets. The current analysis showed that approximately one-fifth of French adults achieved a sustainable diet with a higher nutritional quality and GHGE decreased by almost 20% at no extra cost compared with the population average. These encouraging results confirm that reducing diet-related GHGE and increasing nutritional adequacy is possible through frugality and wiser dietary choices such as a lower intake of meats and alcoholic drinks, higher intake of plant-based foods, and moderate total food intake. However, in these observed diets the compatibility between diet healthiness and low GHGE was not straightforward, which indicated that the development of sustainable dietary guidelines needs to include both nutritional and environmental dimensions to effectively ensure that consumers opt for More Sustainable diets. We thank all of the participants in the INCA2 study and the INCA2 study team for providing the data and necessary support. The authors’ responsibilities were as follows—GM: analyzed the data and drafted the manuscript; GM, ND, L-GS, EOV, and FV: assisted in drafting the manuscript; FV, EOV, and DT: provided the essential materials for the analyses; and ND: had primary responsibility for the final content. All of the
Downloaded from ajcn.nutrition.org at MCMASTER UNIVERSITY on January 5, 2015
Two main factors were identified to result in More Sustainable diets: reduced energy intake and reduced energy density. Indeed, when expressed per kcal, GHGE were reduced by w10% in the More Sustainable diets, which suggests that half of the 20% total GHGE reduction could be attributed to a quantity effect and half to a structural effect. First, the total energy intake was 8% and 10% lower for the More Sustainable diets than for the Average diets in men and women, respectively. Energy intake was thus confirmed as a main driver of dietary GHGE (42). Second, More Sustainable diets had a lower energy density than the Average diets, which accounted for a modified diet structure. The magnitude of the differences was modest, but most differences were statistically significant. The More Sustainable diets contained the highest content of plant-based foods, particularly starchy foods. These results reinforce the rationale for increasing the consumption of plant-based foods and reducing that of animal products to achieve more environment-friendly and healthy diets (10, 13, 15–17). In addition to foods of animal origin, alcoholic beverage consumption was highly associated with dietary GHGE, which strengthened previous findings (14, 44). A better understanding of the environmental pressure associated with alcohol consumption could provide extra support for policymakers in the attempt to limit alcohol consumption in the general population. This study had several limitations. The More Sustainable diets were not sustainable in an absolute sense. They were only defined according to the combination of 2 sustainability criteria (lower GHGE and higher nutritional quality), and those criteria were relative (ie, considered with regard to the respective sex-specific medians) and thus relied on the INCA2 data for comparison. Nutritional quality was assessed by using the PANDiet index, a score based on nutrient recommendations only, without preconceived views on the dietary structure of Higher-Quality diets. The mean PANDiet score for all INCA2 participants was comparable with that in the French Etude Nationale Nutrition Sante´ study and slightly higher than that in US NHANES (27). In contrast with nutrient recommendations, there are no reference values to define low-GHGE diets. We opted for a pragmatic approach that could be applied to any population and sustainability criteria. Diet-related GHGE were assessed by using the internationally standardized life cycle analysis method, and our estimates were comparable with previous estimates based on consumption data (22). However, our estimates did not include emissions arising from indirect land use change and did not account for food losses and waste and may, therefore, have underestimated the actual values (45, 46); such underestimation could explain our lower GHGE estimates when compared with some estimates that included additional factors (12–15, 17). According to the FAO food balance sheet data, the mean food supply for France was 3520 kcal per capita per day in 2007. The INCA2 participants consumed 2162 kcal/d on average, which gave a very gross estimate of 39% for waste at the national level. In the current study, the environmental dimension of sustainability was assessed only through GHGE. A previous study showed that food GHGEs strongly correlate with 2 other indicators of environmental impact (water eutrophication and air acidification) (43), which suggests that the current results would also apply for these indicators. However, food production was shown to account for the vast majority of the global water footprint (9) and is intrinsically linked to ecosystem biodiversity
1467
1468
MASSET ET AL
authors read and approved the final manuscript. No conflicts of interest were reported. 20.
REFERENCES 21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
Downloaded from ajcn.nutrition.org at MCMASTER UNIVERSITY on January 5, 2015
1. Food and Agriculture Organization. Annex I: International scientific symposium biodiversity and sustainable diets—final document. In: Burlingame B, Dernini S, eds. Sustainable diets and biodiversity— directions and solutions for policy, research and action. Rome, Italy: FAO, 2012:294. 2. Tubiello FN, Salvatore M, Rossi S, Ferrara A, Fitton N, Smith P. The FAOSTAT database of greenhouse gas emissions from agriculture. Environ Res Lett 2013;8:015009. 3. Smith P, Martino D, Cai Z, Gwary D, Janzen H, Kumar P, McCarl B, Ogle S, O’Mara F, Rice C, et al. Agriculture. In: Metz B, Davidson O, Bosch P, Dave R, Meyer L, eds. Climate change 2007: mitigation. Contribution of Working Group III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. New York, NY: Cambridge University Press, 2007. 4. European Parliament, Council of the European Union. Decision no 406/2009/Ec of the European Parliament and of the Council of 23 April 2009 on the effort of Member States to reduce their greenhouse gas emissions to meet the Community’s greenhouse gas emission reduction commitments up to 2020. Brussels, Belgium: European Parliament, Council of the European Union, 2012:136–48. 5. Legislation UK. Climate Change Act 2008, code 2008 c. 27. London, United Kingdom: HMSO, 2008. 6. Steinfeld H, Gerber P, Wassenaar T, Castel V, Rosales M, de Haan C. Livestock’s long shadow—environmental issues and options. Rome, Italy: FAO, 2006. 7. Carlsson-Kanyama A, Gonza´lez AD. Potential contributions of food consumption patterns to climate change. Am J Clin Nutr 2009;89 (suppl):1704S–9S. 8. Walle´n A, Brandt N, Wennersten R. Does the Swedish consumer’s choice of food influence greenhouse gas emissions? Environ Sci Policy 2004;7:525–35. 9. Vanham D, Bidoglio G. A review on the indicator water footprint for the EU28. Ecol Indic 2013;26:61–75. 10. Stehfest E, Bouwman L, van Vuuren DP, den Elzen MGJ, Eickhout B, Kabat P, Vuuren DP, Elzen MGJ. Climate benefits of changing diet. Clim Change 2009;95:83–102. 11. Committee on Climate Change. Reducing emissions from agriculture and land use, land-use change and forestry. In: The fourth carbon budget—reducing emissions through the 2020s. London, United Kingdom: Committee on Climate Change, 2010:295–329. 12. Audsley E, Brander M, Chatterton J, Murphy-Bokern D, Webster C, Williams A. How low can we go? An assessment of greenhouse gas emissions from the UK food system and the scope for reduction by 2050. Godalming, United Kingdom: FCRN-WWF-UK, 2010. 13. Tukker A, Goldbohm RA, de Koning A, Verheijden M, Kleijn R, Wolf O, Pe´rez-Domı´nguez I, Rueda-Cantuche JM, Perez-Dominguez I. Environmental impacts of changes to healthier diets in Europe. Ecol Econ 2011;70:1776–88. 14. Saxe H, Larsen TM, Mogensen L. The global warming potential of two healthy Nordic diets compared with the average Danish diet. Clim Change 2013;116:249–62. 15. Berners-Lee M, Hoolohan C, Cammack H, Hewitt CNN. The relative greenhouse gas impacts of realistic dietary choices. Energy Policy 2012;43:184–90. 16. Friel S, Dangour AD, Garnett T, Lock K, Chalabi Z, Roberts I, Butler A, Butler CD, Waage J, McMichael AJ, et al. Public health benefits of strategies to reduce greenhouse-gas emissions: food and agriculture. Lancet 2009;374:2016–25. 17. Macdiarmid JI, Kyle J, Horgan GW, Loe J, Fyfe C, Johnstone A, McNeill G. Sustainable diets for the future: can we contribute to reducing greenhouse gas emissions by eating a healthy diet? Am J Clin Nutr 2012;96:632–9. 18. Wilson N, Nghiem N, Ni Mhurchu C, Eyles H, Baker MG, Blakely T. Foods and dietary patterns that are healthy, low-cost, and environmentally sustainable: a case study of optimization modeling for New Zealand. PLoS ONE 2013;8:e59648. 19. Thompson S, Gower R, Darmon N, Vieux F, Murphy-Bokern D, Maillot M. A balance of healthy and sustainable food choices for
France, Spain, and Sweden. Godalming, United Kingdom: WWF-UK, 2013. Briggs ADM, Kehlbacher A, Tiffin R, Garnett T, Rayner M, Scarborough P. Assessing the impact on chronic disease of incorporating the societal cost of greenhouse gases into the price of food: an econometric and comparative risk assessment modelling study. BMJ Open 2013;3:e003543. Scarborough P, Allender S, Clarke D, Wickramasinghe K, Rayner M. Modelling the health impact of environmentally sustainable dietary scenarios in the UK. Eur J Clin Nutr 2012;66:710–5. Aston LM, Smith JN, Powles JW. Impact of a reduced red and processed meat dietary pattern on disease risks and greenhouse gas emissions in the UK: a modelling study. BMJ Open 2012;2:e001072 . Rolls BJ, Drewnowski A, Ledikwe JH. Changing the energy density of the diet as a strategy for weight management. J Am Diet Assoc 2005; 105(suppl):S98–103. Bes-Rastrollo M, van Dam RM, Martinez-Gonzalez MA, Li TY, Sampson LL, Hu FB. Prospective study of dietary energy density and weight gain in women. Am J Clin Nutr 2008;88:769–77. Vergnaud A-C, Estaquio C, Czernichow S, Pe´neau S, Hercberg S, Galan P, Bertrais S. Energy density and 6-year anthropometric changes in a middle-aged adult cohort. Br J Nutr 2009;102:302–9. Vieux F, Soler L-G, Touazi D, Darmon N. High nutritional quality is not associated with low greenhouse gas emissions in self-selected diets of French adults. Am J Clin Nutr 2013;97:569–83. Verger EO, Mariotti F, Holmes BA, Paineau D, Huneau J-F. Evaluation of a diet quality index based on the probability of adequate nutrient intake (PANDiet) using National French and US Dietary Surveys. PLoS ONE 2012;7:e42155. Dubuisson C, Lioret S, Touvier M, Dufour A, Calamassi-Tran G, Volatier J-L, Lafay L. Trends in food and nutritional intakes of French adults from 1999 to 2007: results from the INCA surveys. Br J Nutr 2010;103:1035–48. Black AE. Critical evaluation of energy intake using the Goldberg cutoff for energy intake:basal metabolic rate. A practical guide to its calculation, use and limitations. Int J Obes Relat Metab Disord 2000; 24:1119–30. Ledikwe JH, Blanck HM, Khan LK, Serdula MK, Seymour JD, Tohill BC, Rolls BJ. Dietary energy density determined by eight calculation methods in a nationally representative United States population. J Nutr 2005;135: 273–8. Kantar Worldpanel. French household consumer panel—Kantar worldpanel. Available from: http://www.kantarworldpanel.com/global/Sectors (cited 5 May 2013). Martin A, ed. Apports nutritionnels conseille´s pour la population franc¸aise. [Recommended nutritional intakes for the French population.] 3rd ed. Paris, France: Lavoisier, 2001. (in French). EC Scientific Committee for Food. Nutrient and energy intakes for the European Community (opinion expressed on 11 December 1992). Luxembourg, Luxembourg: EC Scientific Committee for Food, 1993. EFSA Panel on Dietetic Products, Nutrition, and Allergies. Scientific Opinion on Dietary Reference Values for carbohydrates and dietary fibre. EFSA J 2010;8(3). International Standard. ISO 14040:2006 environmental management— life cycle assessment—principles and framework. Geneva, Switzerland: ISO, 2006. International Standard. ISO 14044:2006 Environmental management— life cycle assessment—requirements and guidelines. Geneva, Switzerland: ISO, 2006. National Institute of Statistics and Economic Studies (Insee). Definitions and methods—statistical operation: survey on industrial energy consumption (EACEI). Available from: http://www.insee.fr/en/methodes/ default.asp?page=sources/ope-enq-conso-energie-industrie-eacei.htm (cited 10 May 2013). French Department of Ecology Sustainable Development and Energy. SitraM database. Available from: http://www.statistiques.developpementdurable.gouv.fr/donnees-ligne/r/flux-marchandises-sitram-i.html?tx_ttnews [tt_news]=20519&cHash=a891e4085d89a9486f97d0282957ec1a (cited 10 May 2013). Althaus H, Doka G, Dones R, Heck T, Hellweg S, Hischier R, Nemecek T, Rebitzer G, Spielmann M, Wernet G. In: Frischknecht R, Jungbluth N, eds. Overview and Methodology - Data v2.0 - ecoinvent report No. 1. Du¨bendorf, CH: Ecoinvent, 2007. Greenext. Engagements & me´thodologie. (Commitments & methods.) Available from: http://www.greencode-info.fr/ (cited 12 November 2012).
IDENTIFYING THE MOST SUSTAINABLE FRENCH DIETS 41. Nelson PR. Additional uses for the analysis of means and extended tables of critical values. Technometrics 1993;35:61–71. 42. Vieux F, Darmon N, Touazi D, Soler LG. Greenhouse gas emissions of self-selected individual diets in France: changing the diet structure or consuming less? Ecol Econ 2012;75:91–101. 43. Masset G, Soler L-G, Vieux F, Darmon N. Identifying sustainable foods: the relationship between environmental impact, nutritional quality, and prices of foods representative of the French diet. J Acad Nutr Diet 2014 (in press). 44. Meier T, Christen O, Semler E, Jahreis G, Voget-Kleschin L, Schrode A, Artmann M. Balancing virtual land imports by a shift in the diet. Using a land balance approach to assess the sustainability of food consumption. Germany as an example. Appetite 2014;74:20–34. 45. Food and Agriculture Organization. Food wastage footprint—impacts on natural resources. Rome, Italy: FAO, 2013. 46. Schmidinger K, Stehfest E. Including CO2 implications of land occupation in LCAs—method and example for livestock products. Int J Life Cycle Assess 2012;17:962–72.
1469
47. FAO. Building on gender, agrobiodiversity and local knowledge— a training manual. Rome, Italy: FAO, 2005. 48. Macdiarmid JI, Loe J, Kyle J, McNeill G. “It was an education in portion size.” Experience of eating a healthy diet and barriers to long term dietary change. Appetite 2013;71:411–9. 49. Hill JO, Peters JC, Wyatt HR. Using the energy gap to address obesity: a commentary. J Am Diet Assoc 2009;109:1848–53. 50. World Cancer Research Fund and American Institute for Cancer Research. Food, nutrition, physical activity and the prevention of cancer: a global perspective. Washington, DC: World Cancer Research Fund and American Institute for Cancer Research, 2007. 51. Westerterp-Plantenga MS. Modulatory factors in the effect of energy density on energy intake. Br J Nutr 2004;92(suppl):S35–9. 52. Stubbs J, Ferres S, Horgan G. Energy density of foods: effects on energy intake. Crit Rev Food Sci Nutr 2000;40:481–515. 53. Ma¨kiniemi J-P, Vainio A. Barriers to climate-friendly food choices among young adults in Finland. Appetite 2014;74:12–9.
Downloaded from ajcn.nutrition.org at MCMASTER UNIVERSITY on January 5, 2015
