Nova classification
The NOVA classification system categorizes foods and beverages into four groups based on the extent and purpose of industrial food processing, distinguishing unprocessed or minimally processed foods (such as fruits, vegetables, grains, and freshly caught or butchered meats), processed culinary ingredients (such as oils, butter, sugar, and salt), processed foods (such as canned vegetables with added salt, fruits in syrup, cheeses, and freshly made breads), and ultra-processed products (industrial formulations typically containing five or more ingredients, including substances not used in home cooking like hydrogenated oils, modified starches, protein isolates, and cosmetic additives for flavor, color, or texture).[1][2] Developed in the early 2000s by Brazilian epidemiologist Carlos Augusto Monteiro and colleagues at the University of São Paulo, NOVA shifts focus from nutrients or calories to processing as a determinant of dietary quality and health impacts, positing that ultra-processed foods (group 4) are inherently obesogenic and linked to non-communicable diseases due to their hyper-palatability, poor satiety, and displacement of minimally processed foods in diets.[3][1] Adopted by international bodies including the World Health Organization, Food and Agriculture Organization, and Pan American Health Organization for dietary guidelines, NOVA has informed policies like front-of-pack labeling in Brazil and Chile, with epidemiological studies associating higher ultra-processed food intake—often exceeding 50% of calories in high-income countries—with increased risks of obesity, cardiovascular disease, type 2 diabetes, and all-cause mortality, based on observational data from cohorts tracking self-reported diets.[3][4] However, randomized controlled trials, such as those matching ultra-processed and minimally processed diets for macronutrients and energy density, have shown greater ad libitum consumption and weight gain on ultra-processed arms, supporting causal claims beyond correlation.[5] Despite its influence, NOVA faces criticism for subjectivity in classifying products (e.g., grouping nutrient-dense items like fortified yogurts with sugary snacks in group 4), neglecting nutritional composition and bioavailability in favor of processing alone, and lacking mechanistic precision on how specific formulations drive health outcomes, potentially stigmatizing entire categories without accounting for fortification or reformulation potential.[6][7] Peer-reviewed critiques argue it oversimplifies food science by ignoring evidence-based nutrient profiling systems and may hinder innovation in healthier processed alternatives, though proponents counter that processing effects transcend nutrient matching.[8][9]Origins and Development
Initial Proposal in 2009
The NOVA food classification system originated with a proposal by Brazilian epidemiologist Carlos Augusto Monteiro in a 2009 commentary published in Public Health Nutrition.[10] Monteiro argued that conventional nutritional guidance, which emphasizes isolated nutrients or whole foods without regard to processing, fails to address the primary driver of diet-related diseases such as obesity, namely the industrial processing of foods that alters their nutritional profile, digestibility, and addictive potential.[11] He proposed shifting analytical focus to the extent and purpose of processing, positing that excessive processing—particularly through additives, emulsifiers, and formulations for hyper-palatability—displaces nutrient-dense foods in diets and contributes disproportionately to non-communicable diseases.[11] Monteiro's initial framework divided foods into three groups, distinguishing unprocessed or minimally processed items from those subjected to varying degrees of industrial intervention.[11] Group 1 encompassed unprocessed or minimally processed foods, defined as whole foods subjected only to procedures like removal of inedible parts, freezing, drying, or pasteurization to extend shelf life while preserving inherent nutritional qualities and sensory attributes.[11] Examples included fresh fruits, vegetables, grains, milk, and meat, which Monteiro viewed as the foundation of healthful diets when combined with culinary preparation.[11] Group 2 consisted of processed culinary ingredients extracted from Group 1 foods or nature, such as oils, butter, sugar, and salt, intended for use in small quantities to season or cook Group 1 foods rather than as standalone consumables.[11] These were seen as traditional aids to home cooking that do not fundamentally alter the nutritional matrix of meals. Group 3, termed "ultra-processed products," included formulations of Group 2 ingredients reformulated with Group 1 elements, plus industrial additives like flavors, colors, emulsifiers, and preservatives to create durable, ready-to-eat or heat products optimized for convenience, profitability, and sensory appeal.[11] Examples cited were soft drinks, biscuits, chicken nuggets, and packaged snacks; Monteiro described these as "confections" engineered to be energy-dense yet nutrient-poor, habit-forming, and capable of displacing minimally processed foods in global supplies, as evidenced by FAO data showing their rise correlating with obesity trends.[11] This tripartite classification aimed to guide public health policy by highlighting how ultra-processed products, absent from traditional diets, proliferate through marketing and displace balanced meals, urging a reevaluation of food guidelines to prioritize processing over nutrient profiling.[11] Monteiro's proposal laid the groundwork for subsequent empirical studies linking higher consumption of Group 3 items to adverse health outcomes, though it initially lacked the formal NOVA acronym and the finer distinction later drawn between moderately processed and ultra-processed categories.[11]Subsequent Refinements and Global Adoption
Following the initial proposal, the NOVA classification underwent refinements through subsequent publications that clarified definitions, expanded examples, and addressed boundary issues. In 2016, Monteiro and colleagues provided an updated framework with more precise criteria for categorizing foods, particularly distinguishing industrial techniques in ultra-processed formulations and incorporating additional case studies to resolve ambiguities in group assignments.[12] These updates emphasized the purpose of processing—such as extending shelf life or enhancing palatability via additives—while maintaining the four-group structure without altering core principles.[13] The system's global adoption accelerated after its integration into Brazil's Dietary Guidelines for the Brazilian Population in 2014, marking the first national policy to explicitly base recommendations on NOVA groups: prioritizing unprocessed or minimally processed foods (group 1) as the dietary foundation, using processed culinary ingredients (group 2) in moderation, limiting processed foods (group 3), and avoiding ultra-processed products (group 4).[14] This approach influenced subsequent guidelines in other countries, including Uruguay's 2016 edition, which similarly advised minimizing ultra-processed items. International organizations further propelled adoption; the Pan American Health Organization (PAHO) incorporated NOVA into its 2014-2019 Plan of Action for Childhood Obesity Prevention and subsequent reports on ultra-processed foods in Latin America, using it to quantify regional consumption patterns and advocate for reformulation policies.[15] The Food and Agriculture Organization (FAO) endorsed the system in a 2019 technical report, applying it to assess diet quality and health outcomes across diverse populations, which facilitated its use in over 1,000 epidemiological studies worldwide by 2025.[1] By the mid-2020s, NOVA informed public health strategies in Europe and North America, including Canada's 2022-2023 food labeling discussions and the UK's 2023 nutrition profiling efforts, though implementation varied due to debates over nutritional nuance versus processing focus.[16]Classification Categories
Group 1: Unprocessed or Minimally Processed Foods
Group 1 of the NOVA classification includes unprocessed foods, defined as edible parts of plants (such as fruits, seeds, leaves, stems, roots, tubers), animals (such as muscle, offal, bone, blood, eggs, milk), fungi, algae, or water, obtained directly from nature without any alteration beyond harvesting or basic collection.[3] Minimally processed foods within this group result from physical processes applied to unprocessed foods, including removal of inedible or unwanted parts, crushing, grinding, milling, drying, roasting, pasteurization, freezing, vacuum-packing, or non-alcoholic fermentation, as long as no added substances like salt, sugar, oils, or emulsifiers are introduced.[3] These processes aim to increase durability, facilitate preparation, or ensure safety while preserving the food's intrinsic nutritional properties and avoiding industrial formulations.[17] Examples of unprocessed foods encompass fresh apples, raw carrots, whole grains like rice or wheat kernels, uncooked beef cuts, fresh eggs, and plain milk.[18] Minimally processed variants include frozen vegetables without additives, ground nuts, pasteurized milk, dried beans, or milled flour from whole grains, provided no extraneous ingredients are incorporated during processing.[17] [18] Such foods form the foundation of diets emphasizing whole, recognizable ingredients, distinguishing them from higher NOVA groups that involve added culinary or industrial components.[19] Boundary cases arise when minimal processing verges on adding non-food elements; for instance, roasted coffee beans qualify as minimally processed due to heat application alone, but roasted nuts with added salt shift to Group 2 or higher.[17] The classification, originally proposed by researchers at the University of São Paulo in 2009 and refined in subsequent publications, relies on ingredient lists and processing descriptions rather than nutritional content, allowing consistent categorization across global food systems despite variations in industrial practices.[3] This approach has been adopted by organizations like the Food and Agriculture Organization for dietary guideline development as of 2019.[1]Group 2: Processed Culinary Ingredients
Processed culinary ingredients in the NOVA classification, designated as Group 2, consist of substances derived from unprocessed or minimally processed foods (Group 1) or directly from nature through industrial processes including pressing, refining, grinding, milling, or drying.[1] These ingredients are intended for use in the preparation, cooking, or seasoning of dishes primarily composed of Group 1 foods, rather than for standalone consumption.[19] Unlike Group 1 items, Group 2 products undergo extraction or purification to isolate components such as fats, sugars, or salts, but without addition of other Group 2 substances or complex formulations characteristic of higher groups.[3] Examples of Group 2 ingredients include vegetable oils extracted from seeds, nuts, or fruits like olives; animal fats such as butter and lard; sweeteners like sugar, honey, and syrups; salt; and other extracts like vinegar and starch.[1] These are obtained via straightforward physical or chemical processes that concentrate naturally occurring elements without introducing additives beyond the base material.[19] For instance, olive oil results from pressing olives, while table sugar arises from refining sugarcane or beets.[3] In the NOVA framework, Group 2 ingredients serve a supportive role in home cooking to enhance flavor, texture, or preservation of Group 1 foods, aligning with traditional culinary practices.[1] The classification posits that their moderate use does not inherently promote unhealthy diets, provided they complement rather than displace minimally processed foods.[19] Boundary distinctions arise with products like honey, which may border Group 1 if unprocessed, but qualify as Group 2 when industrially refined or pasteurized.[3] This group excludes items with added preservatives or those used primarily in industrial food manufacturing, reserving such for Groups 3 and 4.[1]Group 3: Processed Foods
In the NOVA classification system, Group 3 comprises processed foods manufactured through techniques such as canning, bottling, smoking, fermenting, or baking, primarily using ingredients from Group 1 (unprocessed or minimally processed foods) combined with additions of Group 2 substances like salt, sugar, oil, or vinegar to enhance preservation or sensory qualities such as flavor and texture.[5][20] These processes aim to extend shelf life or make foods more palatable while preserving the recognizable form of the original ingredients, without the extensive use of industrial additives, emulsifiers, or artificial flavors characteristic of Group 4.[21][22] Unlike minimally processed foods in Group 1, which undergo only basic alterations like cleaning or freezing, Group 3 products involve deliberate additions and methods that alter composition but maintain a simple ingredient profile, typically 2-3 recognizable components.[5] This category excludes formulations designed for hyper-palatability or convenience through multiple industrial techniques.[23] Common examples include:- Canned or jarred vegetables, fruits, legumes, or fish in brine, water, or their own juices.[5][24]
- Salted, smoked, or cured meats and fish, such as bacon or jerky without additional preservatives.[25]
- Cheeses produced by fermenting milk with salt or rennet.[5]
- Freshly made breads from flour, water, yeast, and salt; or fruits preserved in syrup.[5][25]
- Beer or wine from fermented grains or fruits with minimal additives.[25]
Group 4: Ultra-Processed Foods
Group 4 in the NOVA classification encompasses ultra-processed foods, defined as industrial formulations primarily composed of substances of exclusive industrial use, resulting from multiple sequential processes such as extrusion, molding, and pre-frying.[23] These products typically contain five or more ingredients, including those not commonly found in home cooking, and are designed to be convenient, hyper-palatable, and shelf-stable, often displacing minimally processed foods in diets.[19] The classification emphasizes the extent and purpose of processing, distinguishing ultra-processed items by their reliance on additives like emulsifiers, artificial flavors, colors, and stabilizers to enhance sensory appeal and extend durability.[28] Key characteristics include the use of ingredients such as hydrogenated fats, modified starches, hydrolyzed proteins, high-fructose corn syrup, and soy protein isolates, which are derived from Group 1 or 2 commodities but transformed through industrial means.[29] Unlike processed foods in Group 3, which preserve recognizable forms of whole foods (e.g., canned vegetables or cheeses), ultra-processed products undergo extensive fractionation, reconfiguration, and combination, often rendering original food structures unrecognizable.[23] This level of processing aims to create ready-to-consume or heat items that require no further preparation beyond minimal heating or consumption, prioritizing profitability through low cost and high volume production.[30] Examples of ultra-processed foods include:- Carbonated soft drinks and energy drinks sweetened with added sugars or artificial sweeteners.[24]
- Packaged snacks such as chips, cookies, and extruded cereals with added flavors and preservatives.[19]
- Instant noodles, frozen ready meals, and mass-produced breads containing dough conditioners and emulsifiers.[31]
- Sweetened yogurts, candies, and chocolate bars formulated with multiple stabilizers and flavor enhancers.[24]
Methodology and Defining Criteria
Principles of Food Processing Assessment
The NOVA classification system evaluates food processing through three interrelated principles: the nature of the processes applied (encompassing physical methods like milling or freezing, biological techniques such as fermentation, and chemical alterations including hydrogenation or the addition of synthetic substances); the extent of processing (ranging from minimal interventions to preserve edibility and safety to extensive industrial formulations involving multiple sequential techniques like extrusion, moulding, or pre-frying); and the purpose of processing (distinguishing between extensions of traditional home cooking for preservation or palatability, and industrial designs aimed at enhancing profitability, convenience, and hyper-palatability through displacement of minimally processed foods).[3][32] These principles prioritize the overall transformation of the food item over isolated nutrient profiles, positing that industrial processing inherently alters nutritional integrity, digestibility, and metabolic impacts.[33] Assessment begins by inspecting the product's ingredient list, packaging claims, and manufacturing descriptors to determine if it consists primarily of intact or simply modified natural foods (assigning to Group 1), extracted substances for culinary use (Group 2), or recognizable modifications via basic additions like salt or sugar (Group 3).[32] Products featuring five or more ingredients, especially cosmetic additives (e.g., emulsifiers, artificial flavors, colors, or stabilizers not used in home cooking), fractionated food-derived elements (e.g., hydrogenated oils, modified starches, or hydrolyzed proteins), and techniques yielding unrecognizable forms (e.g., reconstituted meat or molded snacks) are classified as ultra-processed (Group 4).[3] This hierarchical evaluation ensures that the dominant processing level dictates the group, with industrial intent—evidenced by formulations engineered for extended shelf life, sensory optimization, and minimal preparation—elevating classification regardless of occasional inclusion of Group 1 elements. Practical application involves cross-referencing against validated examples from NOVA guidelines, such as classifying canned fish in oil as processed (Group 3) due to salting and oil addition for preservation, versus packaged fish sticks as ultra-processed owing to battering, frying, and additives for texture and appeal.[32] While these criteria emphasize empirical observation of processing markers over proprietary formulations (often undisclosed), reproducibility can vary without algorithmic tools, as demonstrated by inter-evaluator agreement rates below 70% for ambiguous items like flavored yogurts.[34] Nonetheless, the system's focus on verifiable industrial signatures supports its use in epidemiological coding of food databases, where processing extent correlates with additive counts and formulation complexity exceeding traditional recipes.[33]Examples and Boundary Challenges
Examples of Group 1 foods include fresh fruits and vegetables, grains such as rice or wheat, fresh or frozen meat, poultry, fish, milk, and eggs, with minimal alterations like pasteurization, freezing, drying, or grinding to preserve them without adding substances.[3] Group 2 encompasses ingredients extracted from Group 1 foods or nature, such as vegetable oils, butter, sugar, honey, salt, and vinegar, used in home cooking to season or prepare dishes.[20] Group 3 consists of foods made by combining Group 1 and 2 items through processes like canning, fermenting, or baking, yielding products such as canned vegetables with salt, cheese, fruits in syrup, salted or sugared nuts, freshly made bread, and beer.[3] Group 4 features formulations of ingredients, often including high-fructose corn syrup, hydrogenated oils, modified starches, and additives like emulsifiers, flavors, and colors, resulting in items like soft drinks, packaged snacks, instant noodles, ready-to-heat meals, and mass-produced breads with preservatives.[23] Boundary challenges arise primarily between Groups 3 and 4, where distinctions hinge on the purpose and extent of processing—preservation and flavor enhancement for Group 3 versus industrial formulations for palatability and convenience in Group 4—but definitions remain ambiguous, leading to inconsistent classifications.[35] For instance, artisanal bread made with flour, water, salt, and yeast qualifies as Group 3, while industrially produced bread with emulsifiers and dough conditioners falls into Group 4, yet small-scale producers using similar additives blur this line, complicating application without detailed ingredient scrutiny.[16] Yogurt poses another ambiguity: plain versions from milk fermentation are Group 3, but those with added sugars, flavors, or stabilizers for shelf life shift to Group 4, with criteria like the presence of non-culinary additives not always clearly delineating industrial intent.[34] Further difficulties occur in mixed dishes or reformulations, where incorporating even minor Group 4 ingredients—such as a flavor enhancer in a traditionally processed sauce—may elevate the entire product, but NOVA lacks explicit rules for ingredient integration or proportional thresholds, fostering subjective interpretations.[35] Professionals report confusion over items like plant-based milks or fortified cereals, where mechanical extraction might suggest Group 2, but added stabilizers indicate Group 4, underscoring the need for iterative, multi-step assessment protocols to enhance transparency and reduce errors.[36] These ambiguities stem from NOVA's reliance on descriptive rather than quantitative criteria, as outlined by Monteiro et al., potentially overlooking variations in production scale or regional practices.[23]Empirical Evidence on Health Associations
Studies Linking Consumption to Outcomes
Numerous prospective cohort studies and meta-analyses have identified associations between higher consumption of ultra-processed foods (NOVA Group 4) and increased risks of adverse health outcomes, including cardiometabolic diseases, obesity, and mortality, after adjustment for potential confounders such as age, sex, smoking, physical activity, and overall energy intake.[37] [38] These associations often exhibit dose-response patterns, with risk escalating as the proportion of ultra-processed foods in the diet rises.[37] A 2023 systematic review and meta-analysis of 28 prospective cohort studies reported the following pooled relative risks (RR) for highest versus lowest ultra-processed food intake categories:| Outcome | Pooled RR (95% CI) | Number of Studies |
|---|---|---|
| Obesity | 1.32 (1.20–1.45) | 13 |
| Type 2 diabetes | 1.37 (1.20–1.56) | 7 |
| Hypertension | 1.32 (1.19–1.45) | 5 |
| Hypertriglyceridemia | 1.47 (1.12–1.93) | 3 |
| Low HDL cholesterol | 1.43 (1.05–1.95) | 3 |