Corrigendum: Modeled sustainability impacts of increasing pork consumption among adults in the United States

1IntroductionCurrent food systems present challenges for all domains of sustainability, which include nutrition/health, environment, economy, and society. In the United States (US), less than 5% of the population meets dietary recommendations,(1) which accounts for nearly half of all deaths from cardiometabolic diseases(2) and is the leading modifiable risk factor for overall mortality.(3) Suboptimal diet quality also accounts for 11% of disability-adjusted life years(3) and nearly 20% of total direct medical costs.(4) Healthier diets are often more expensive than less healthy ones(5) and are unaffordable for lower income minority groups,(6) which has contributed to worsening disparities in diet quality(1) and healthcare costs.(4) At the same time, the US food system places a heavy burden on the environment. It is responsible for nearly 20% of total national greenhouse gas emissions (GHGE)(7) and over 25% of land use and freshwater withdrawals.(7) On a per capita basis, GHGE are over 70% higher in the US than the global average.(8)There is growing urgency among consumers,(9) clinicians,(10) policymakers,(11) and other stakeholders(12) to identify sustainable food choices to stay within planetary boundaries. There is a heightened focus on dietary shifts within the protein foods group,(12) given the wide range of sustainability impacts across different foods within this category.(13) Prior modeling research has focused on replacing beef with poultry or plant proteins, or replacing animal proteins with plant proteins, which have resulted in higher diet quality and reductions in GHGE and diet cost.(14-16) However, less is known about the role of pork in sustainable diet patterns, given that it is often aggregated with other animal proteins or not evaluated at all.(17) This oversight prevents a comprehensive understanding of how dietary shifts within the protein food category can impact sustainability outcomes, given that pork has a prominent place in US diets. Pork accounts for nearly 25% of daily intake of meat and poultry by weight (1.45 oz), and consumption has increased by 14% since 2014.(18) Pork also has a central place in consumer diets throughout the world, which is supported by US production and exports. Pork is the most consumed meat in the world, exceeding 115 million tons annually,(19) and consumption is expected to grow by 8% by 2033.(20) The US is the third largest producer of pork, accounting for 11% of global production,(21) and the second largest exporter of pork, accounting for 31% of exports.(19) Prior research has shown that pork consumption is associated with greater likelihood of meeting many micronutrient requirements,(22) and has a higher digestible indispensable amino acid score (DIAAS) than most other protein foods.(23) Pork also outperforms some other protein foods in other sustainability metrics.(13, 17) For example, GHGE from pork (per kg of food) are 6-fold lower than for beef,(13) and pork is the lowest cost protein food (per 50 g of protein) behind eggs and beans.(17) Thus, further research is needed to evaluate whether replacing different protein foods with pork can improve sustainability outcomes.To address this gap, the present study models the sustainability impacts of replacing beef, poultry, seafood, eggs, and legumes with pork in a nationally representative sample of US adults. Sustainability is comprehensively evaluated using multiple impacts to understand trade-offs, which include GHGE, cumulative energy demand, water scarcity footprint, agricultural land, fertilizer nutrients, pesticides, diet cost, and diet quality. This research can help inform important clinical and policy discussions related to diet sustainability in the US.(11, 24)2Methods2.1Dietary dataData on individual-level intake of foods and nutrients, as well as sociodemographic characteristics, were acquired from the National Health and Nutrition Examination Survey (NHANES), 2011-2018.(25) Data are collected continuously from approximately 5,000 non-institutionalized participants each year using a clustered, stratified, multi-stage sampling design, and data are released in two-year cycles.(25) Some demographic groups are oversampled to increase reliability and precision for subgroup analysis.(26) Dietary data collection is administered by a trained interviewer that uses the Automated Multiple Pass Method, a computer-assisted 24-hour recall, to minimize participant burden and increase reliability and validity of the data.(27, 28) Data from one day of dietary recall was appropriately used to estimate per capita intake.(29) 2.2Food categories and serving sizesAs part of the automated dietary recall procedure, each food (including mixed dishes) reported consumed by NHANES participants is automatically assigned an 8-digit numerical identifier (known as a food code) based on the predominant ingredient in that food, using the US Department of Agriculture (USDA) Food and Nutrient Database for Dietary Studies (FNDDS).(30) In the present study, the FNDDS food codes from 2011-2018 were used to group foods into the following protein dish categories: beef, pork, poultry, seafood, eggs, and legumes (Supplemental Table 1). Foods that are typically consumed in small amounts were excluded from these categories because their serving sizes are not comparable to other foods within these categories, such as bacon (beef, pork, and Canadian), dried beef, spareribs, cracklings, pork skin, and miscellaneous parts. Foods that contain multiple types of meat were also not assigned to a food category. However, this study included the sustainability impacts of all foods when estimating total daily per capita impacts. For each protein dish category, the median gram weight across all eating occasions for all foods was used to represent the median serving size (Supplemental Table 2). 2.3Diet quality measurementThe Healthy Eating Index-2020 (HEI-2020) was used to measure diet quality because it measures adherence to the Dietary Guidelines for Americans.(31) The HEI-2020 includes nine components to encourage (total fruit, whole fruit, total vegetables, greens and beans, whole grains, dairy, total protein foods, seafood and plant proteins, and the ratio of unsaturated to saturated fats) and four components to limit (refined grains, sodium, added sugars, and saturated fats). The intake of most components is energy-adjusted to 1,000 kcal, and is scored against predefined minimum and maximum values, with intermediate intakes scored proportionally (Supplemental Table 3). Each component is scored from 0-5 or 0-10, and higher scores represent more favorable intakes. For each participant, scores for all components are summed to generate a total score out of 100.(31) 2.4Greenhouse gas emissions, cumulative energy demand, and water scarcity footprintThe database of Food Impacts on the Environment for Linking to Diets (dataFIELD) provided information on GHGE, cumulative energy demand (CED), and water scarcity footprint (WSF) for each food in NHANES.(13, 32) These data were compiled from 321 food environmental life cycle assessments (LCA) published from 2005-2016 using a systematic review, which resulted in 1,645 combinations of food types and production scenarios.(33) Data represent most regions of the world, with the majority from Europe(33). Nearly all studies accounted for agricultural production, 51% accounted for post-farmgate processing, 19% accounted for distribution and retail, and 6% accounted for the consumer-level impacts. For each food, environmental impacts from multiple studies were averaged and matched to commodities in the US Environmental Protection Agency (EPA) Food Commodity Intake Database (FCID), which provides information on the amount of approximately 500 commodity ingredients in each food in NHANES.(34) The present study used an updated version of FCID that aligns with NHANES 2011-2018.(35) 2.5Agricultural land, fertilizer nutrients, and pesticidesFoodprint 2.0 was used to estimate the agricultural resource requirements associated with individual-level diet patterns in NHANES.(36) These agricultural resources include land (including all types of cropland and pasture land), fertilizer nutrients (sum of nitrogen, phosphorus-P2O5, potash-K2O, and sulfur), and pesticides (sum of herbicides, insecticides, and fungicides). Foodprint 2.0 is a biophysical simulation model that represents the US food system as a series of integrated processes. Embedded data and calculations are used to transform NHANES foods in their as-consumed forms into agricultural commodities and the agricultural resources needed to produce these commodities (Supplemental Document 1). Foodprint 2.0 accounts for population size, international food trade, loss and waste, food composition, food processing conversions, livestock feed requirements, crop and livestock yields, availability of agricultural land, suitability of agricultural land for food production, multi-use crops (i.e., crops that are used to produce multiple products from equivalent mass), multi-use cropland (i.e., cropland used to produce multiple crops during different parts of the year), and application rates for fertilizer nutrients and pesticides. All parameters represent US national averages.2.6Diet costThe USDA Economic Research Service (ERS) Purchase-to-Plate Price Tool (PPPT) provided information on prices for each NHANES food.(37) These data were collected from retail checkout scanners and represent nearly 50% of all retail food sales in the US.(38) USDA ERS staff matched these scanner data to NHANES foods using machine learning, and removed the cost associated with loss and waste so the final data reflect the cost associated with the consumed portion only.(39) Food prices from PPPT only represent food-at-home (FAH) prices and there are no publicly available data on national average food-away-from-home (FAFH) prices for each NHANES food. Therefore, PPPT assigns FAH prices for all NHANES foods regardless of whether participants reported consuming that food at home or away from home. This will severely underestimate total diet cost because consumers typically face higher prices for FAFH than FAH, and other data show that FAFH accounts for approximately 50% of consumer food expenditures.(40) Therefore, the present study derived FAFH prices using a methodology previously demonstrated(5, 41) and described below. Data on FAH and FAFH spending were acquired from the National Household Food Acquisition and Purchase Survey (FoodAPS),(42) and were used to derive a coefficient that converted FAH prices (from PPPT) to FAFH prices for each of the FAFH reported consumed by NHANES participants. From April 2012 through January 2013, FoodAPS used a multi-stage survey design to collect information from US households on the price of all purchased foods from receipts and scanned barcodes.(42) Survey-weighted mean FAH and FAFH prices were estimated for each major food group (meat, poultry, seafood, eggs, dairy, fats and oils, fruits and vegetables, sweets, grains, non-alcoholic beverages, and other foods), and these were used to derive a coefficient that represents the ratio of FAFH-to-FAH prices for each food group. These coefficients were multiplied by the price of each FAFH in PPPT to estimate its FAFH price. For example, if the price of a given beef product was $3.50 (from PPPT), and if the mean price of FAFH beef was 3.17 times greater than the mean price of FAH beef (from FoodAPS), the adjusted price of that given beef product would be estimated as $11.11 ($3.50 × 3.17). 2.7Impacts per gram and per servingThe price, GHGE, CED, WSF, land, fertilizer nutrients, and pesticides per gram of each NHANES food were calculated by dividing the total impacts associated with each food by its gram weight. The amount of each HEI-2020 component per gram of each NHANES food was also calculated in the same way. After accounting for food loss and waste (see below), these were multiplied by the median serving size (in gram weight) of each protein dish category to estimate the impacts per serving of each category (Supplemental Tables 4 and 5). 2.8Food loss and wasteData on GHGE, CED, and WSF (from dataFIELD); and food prices (from PPPT); do not include the environmental impacts and cost associated with food that is lost and wasted at the retail and consumer levels. Not accounting for food loss and waste can underestimate consumer food demand,(43) environmental impacts,(44) and diet cost(41) by up to 37-42%. To fill these gaps, data on data on commodity-level loss and waste were acquired from the USDA Loss-adjusted Food Availability data system (LAFA),(45) and matched to commodities in the FCID, which links to NHANES. This procedure has been demonstrated previously(5, 41, 43, 44, 46, 47) and additional details on sources of uncertainty and embedded assumptions are described elsewhere.(5, 47) 2.9Diet modelingA food substitution model was developed to estimate the effects of iteratively replacing 1-3 servings of each protein dish (beef, poultry, seafood, eggs, and legumes) with 1-3 servings of a pork dish on daily per capita sustainability impacts, which include GHGE, CED, WSF, land, fertilizer nutrients, pesticides, diet cost, and diet quality. Substitutions were performed separately for each protein dish. The serving size of each protein dish varied from 92 g (eggs) to 117 g (beef), so the serving size used for pork varied accordingly (Supplemental Table 2). For example, each serving of beef (117 g) was replaced by 117 g of pork, whereas each serving of poultry (110 g) was replaced by 110 g of pork. These substitutions were made based on observed serving sizes (i.e., from NHANES 24-h recalls) to reflect the observed quantity of dishes consumed during eating occasions in which Substitutions were only made for whole For example, participants consumed servings of beef at a substitution for pork, and participants consumed servings of beef a substitution for pork. This modeling for intake of foods, which be more in than their evaluate the of the modeling procedure, were used to understand whether servings sizes than observed serving different Each dish consumed by NHANES participants was assigned a serving size based on the by the US Food and Table daily per capita GHGE, CED, WSF, land, fertilizer nutrients, pesticides, diet cost, and diet quality were estimated at and were estimated using models adjusted for energy intake and NHANES survey cycle in mean impacts and substitutions were using at of same substitution and substitution NHANES design and survey were used to for the sampling design and to produce nationally representative was used for data and total of NHANES participants provided dietary data from were excluded from the if were not dietary energy or sustainability impact CED, WSF, land, fertilizer nutrients, pesticides, or diet that was from the mean The final sample included participants. The mean of participants was and more than half were The majority an ratio and were greenhouse gas emissions, cumulative energy demand, and water scarcity footprintThe mean daily intake (in of beef, pork, poultry, seafood, and dishes at and substitutions are in Table GHGE when replacing up to servings of beef and seafood with pork and Table for all up to servings of poultry, seafood, eggs, and legumes with pork to an increase in GHGE of up to with the increase observed for poultry substitution for all when replacing up to servings of beef and seafood and increased by up to when replacing poultry, eggs, and legumes for all WSF by up to 6% when replacing beef and increased by up to when replacing each of the other protein foods for all resource land, fertilizer nutrients, and up to servings of beef with pork to a in land fertilizer nutrients, and pesticides by up to and and Table for all poultry, seafood, eggs, or legumes was associated with a in land fertilizer nutrients, and pesticides for all cost and diet cost by when replacing beef or seafood with pork, and increased by up to 5% when replacing poultry, eggs, or legumes and Table HEI-2020 scores by up to replacing poultry with pork, and by replacing beef, seafood, eggs, or legumes Each protein dish category represents a of all foods within that category, foods with multiple Thus, protein dish substitutions can to in intake of Table and show how these foods diet quality by the impact of protein dish substitutions on HEI-2020 component 1-3 servings of each protein dish for pork was associated with lower intake of to higher HEI-2020 component and greater intake of total protein to higher For most protein dish intake increased to higher saturated intake increased to higher lower scores for beef and unsaturated intake to lower higher score for and consumption of seafood and plant proteins to 6% lower in intake were observed for total and to lower score for replacing each protein dish with pork. serving sizes were estimated using the published by the US which represent serving sizes than observed serving In serving sizes in to the (Supplemental Table For beef GHGE, land, and pesticides were up to 8% lower than the For poultry, seafood, and land, fertilizer nutrients, and pesticides were up to higher than the For all other the was to the this nationally representative study of over US substitution of protein foods (beef, poultry, seafood, eggs, and legumes) with pork to in sustainability impacts. The were observed when beef was replaced with pork, which was associated with reductions in GHGE, land, pesticides, and fertilizer nutrients by seafood with pork to reductions in by 6% and diet cost by All other substitutions to an increase in sustainability impacts of up to These can help inform and research discussions about sustainable diet patterns, which have previously not focused on the role of pork as a protein food. the of pork in consumer it is often not from beef in research and policy This has to in about the role of pork in sustainable diet patterns, which policy some prominent food do pork and do Pork and beef are into the categories meat and meat in the Food Database which is by to the dietary intake of the US The are given that these are in the Dietary Guidelines for which all and For example, pork and beef are aggregated into a category in the USDA Food which is used to for over million that in the Food The US can an important role in data collection that pork and other The National on and Health the National published by the in provides the for this The National for a to improve the of the US food system on all domains of It is into one of which is Nutrition and Food which to data data and research that can be used to inform and food For example, it for a of the National Household Food Acquisition and Purchase Survey updated in which provides information on food and food FoodAPS has been a data for estimating diet for the present The National can also the for so that it pork from other and can be used to a of the Food to Database and the Food Commodity Intake Database The updated in and FCID updated in are commodity ingredient that are used to the ingredients in each NHANES mixed dish so can be with sustainability impacts, and of these pork from other to these are needed to fill about sustainable food in the protein foods present study also has for clinical that pork for other protein foods was associated with lower intake of and higher intake of increased intake of These are with research that that pork consumers and were more to the daily intake for These that pork is a food that for diet and the of consumption of mixed dishes than foods, which eating can an important role in their to their pork dishes to improve overall diet which focus on sodium, or by replacing pork dishes with protein Pork is a ingredient in such as which typically contain amounts of sodium, and are associated with risk of cardiometabolic and some types of to pork, or to protein an to intake and improve overall diet quality. At the same time, to pork are more to daily micronutrient for and most to most other protein foods, pork is associated with greater digestible indispensable amino acid scores and greater of amino in and pork consumption by one per day was associated with up to 8% lower likelihood of among Pork is also the most of protein when as per gram of behind only and it an of quality studies have evaluated the and impacts of protein food these not evaluate pork as a food The present study this by that different protein foods with pork can to in sustainability impacts. in sustainability impacts were only observed when beef was replaced by pork lower GHGE, land, pesticides, and fertilizer and when seafood was replaced by pork lower and lower diet All other substitutions were associated with greater sustainability impacts of up to all associated with poultry, eggs, and legumes with pork. Some of these from modeling research that that diet can be meeting energy and requirements if consumers the lowest price pork than This from the present study, which evaluated the price of pork products purchased by than the lowest cost diet sustainability are observed in the research For example, a that a more diets lower the intake of some The which was to represent healthy and sustainable diet has been associated with higher land use and lower intake of some and is unaffordable for in lower income regions of the In the other research that higher diet quality was associated with higher amounts of food food and some types of agricultural and diets with lower GHGE have been associated with lower intake of some These important for and other about how to this information to the and how consumers can use this information to sustainable food choices in different This study has The dietary data were nationally these to the US Food substitutions were made on the of observed serving than serving which the in which food were used to the of this by using serving sizes by the US which in the most of these were in this study evaluated sustainability impacts that represent of the four domains of sustainability environment, and which to sustainability study also has NHANES participants typically consuming mixed dishes that contain multiple so each dish was to its food and the sustainability impacts were averaged across all dishes within each category. Therefore, it is that the were by ingredients in each dish (i.e., that a of each dish on a weight However, this many in which consumers one of dish with of than For example, it is that most consumers would use pork than eggs in an use pork than There is also wide in sustainability impacts by of and other that not be be an important for to of data some sustainability impacts not be some environmental of and substitution of different protein foods with pork to in sustainability impacts. in sustainability impacts were observed when pork replaced beef GHGE, land, pesticides, and fertilizer and seafood and diet All sustainability impacts increased when pork replaced poultry, eggs, and This an important research because studies have typically not evaluated pork as a protein food, and some major food do not pork from other This to sustainable food The US can an important role in data collection that pork from other

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.