Dietary fat is the lipid component of foods and beverages, consisting predominantly of triglycerides, with smaller amounts of phospholipids, sterols, and other lipid-soluble substances. Along with carbohydrate and protein, it is a major energy-yielding component of human diets. Most digestible dietary fat provides approximately 9 kilocalories, or 37 kilojoules, per gram. Its constituent fatty acids differ in chain length, degree of saturation, and molecular configuration—properties that influence both food characteristics and biological effects.
Dietary fat supplies essential fatty acids, supports the absorption of fat-soluble vitamins, and contributes to cell membranes and physiological signaling. After digestion and absorption, its components enter pathways of transport, storage, and metabolism. Its nutritional significance depends not only on the amount consumed but also on fatty-acid composition, food sources, the surrounding dietary pattern, and which nutrients it replaces.
Research on dietary fat connects food chemistry, physiology, and public health, particularly through its relationship with blood lipoproteins and cardiovascular disease. The subject encompasses the composition and processing of edible fats, their biological functions, methods of measuring intake, the development of nutrition research, and the standards used in dietary guidance and food regulation.
Chemical composition and physical properties
Dietary fat is a nutritional category encompassing the lipids supplied by foods, principally triglycerides together with smaller amounts of other lipid compounds. It is not synonymous with body fat: the latter usually means stored triglycerides, whereas adipose tissue is a living tissue containing fat-storing cells, blood vessels, connective tissue, and immune cells. The biochemical category of lipids is broader still, including membrane components and signaling molecules that are not necessarily substantial dietary constituents. Unlike proteins, lipids do not share a single repeating molecular backbone; their classification rests largely on their hydrophobic or amphipathic character.
Triglycerides, also called triacylglycerols, consist of a three-carbon glycerol molecule whose three hydroxyl groups are esterified with fatty acids. Each ester linkage joins a glycerol oxygen to a fatty acid’s carbonyl carbon. The three attached fatty acids may be identical, but naturally occurring food triglycerides commonly contain two or three different fatty acids. Their positions on glycerol, designated -1, -2, and -3, distinguish molecules even when the overall fatty-acid composition is the same. Triglycerides generally account for more than 95% of the mass of refined edible oils. Other food lipids include phospholipids, cholesterol, plant sterols such as beta-sitosterol and campesterol, and small quantities of free fatty acids, monoacylglycerols, and diacylglycerols. Fat-soluble vitamins and pigments, including carotenoids, may also occur in the lipid fraction.
Although triglycerides contain oxygen-bearing ester groups, their long hydrocarbon chains dominate their behavior: they are largely nonpolar and have very low solubility in water. Common glycerophospholipids instead contain two fatty-acid chains and a phosphate-bearing polar head group. This amphipathic structure allows them to associate with both water and oil and to accumulate at their interface. An emulsion disperses droplets of one liquid within another rather than dissolving them molecularly. Milk and mayonnaise are oil-in-water emulsions, whereas butter is predominantly a water-in-oil emulsion. Phospholipids, proteins, and other emulsifiers can stabilize droplets by forming interfacial layers that hinder their coalescence; simply mixing oil and water does not produce a permanently stable emulsion.
Melting behavior depends on both fatty-acid composition and the architecture of the triglycerides containing those fatty acids. Longer, relatively straight chains generally have stronger intermolecular attractions and pack more efficiently than shorter or kinked chains. Cis double bonds introduce bends that commonly reduce packing efficiency and lower melting temperature. As an illustration, the free fatty acid stearic acid melts at about 69–70 °C, whereas oleic acid, with the same number of carbon atoms but one cis double bond, melts at about 13–14 °C. These values do not directly predict a food oil’s melting behavior, because its fatty acids are chiefly esterified in mixtures of triglycerides.
Food fats therefore usually soften and melt across a temperature range rather than at one sharply defined temperature. Their texture depends on the proportion of solid crystals to liquid oil, crystal size, and the network those crystals form. Triglycerides can crystallize in different arrangements, commonly designated alpha, beta-prime, and beta, with different stability and melting properties; fine beta-prime crystals are often desirable in smooth spreads. Cooling rate and storage temperature can change this structure without changing chemical composition. “Fat” and “oil” consequently usually describe predominantly solid and liquid states at a stated temperature, not separate chemical classes: coconut oil can solidify in a cool room and liquefy when warmed.
Classification of dietary fatty acids
Dietary fatty acids are classified along several independent structural dimensions: carbon-chain length, number and position of carbon–carbon double bonds, and configuration around those bonds. Saturated fatty acids contain no carbon–carbon double bonds; palmitic acid has 16 carbon atoms and is written , while stearic acid has 18 and is written . Monounsaturated fatty acids contain one double bond, as in oleic acid, . Polyunsaturated fatty acids contain two or more: linoleic acid is , and alpha-linolenic acid is . In this shorthand, the number before the colon gives the total carbon count, including the carboxyl carbon, and the number after it gives the number of carbon–carbon double bonds. These two numbers alone do not specify a fatty acid completely, because double-bond positions and configurations can differ.
Chain-length categories have boundaries that vary among biochemical, nutritional, and clinical conventions. One convention groups fatty acids with fewer than six carbons as short-chain, those with 6–12 as medium-chain, those with 13–21 as long-chain, and those with 22 or more as very-long-chain. Other schemes include six-carbon acids among short-chain fatty acids or use different boundaries for long-chain and very-long-chain groups. Butyric acid, , occurs in milk fat; caproic, caprylic, and capric acids contain 6, 8, and 10 carbons, respectively. Coconut oil contains medium-chain fatty acids, including substantial lauric acid, , although its composition differs from commercial medium-chain triglyceride preparations composed mainly of eight- and ten-carbon fatty acids. Palmitic, stearic, oleic, and linoleic acids are long-chain examples. Chain-length terminology therefore needs to be read alongside the actual carbon count.
Omega, also written n-, notation identifies the position of the first double bond counted from the terminal methyl end of the chain, opposite the carboxyl group. In an omega-3 fatty acid, that bond lies between the third and fourth carbons counted from the methyl end; in an omega-6 fatty acid, it lies between the sixth and seventh. The omega-3 family includes alpha-linolenic acid, ; eicosapentaenoic acid (EPA), ; and docosahexaenoic acid (DHA), . The omega-6 family includes linoleic acid, , and arachidonic acid, . Omega notation identifies a family rather than every double-bond position. By contrast, delta notation counts from the carboxyl end: oleic acid has a cis double bond at . Alpha-linolenic acid and EPA therefore belong to the same omega family despite differing in both chain length and double-bond number.
Cis and trans describe the geometry around a carbon–carbon double bond. In the cis configuration, the hydrogen atoms attached to the double-bonded carbons lie on the same side; in the trans configuration, they lie on opposite sides. A cis bond generally introduces a more pronounced bend in the chain, whereas a trans bond permits a straighter configuration. Oleic acid is cis-; its trans counterpart is elaidic acid. Both remain monounsaturated: trans configuration does not make a fatty acid chemically saturated.
Industrial trans fatty acids arise chiefly during partial hydrogenation, which both saturates some double bonds and rearranges some remaining bonds into trans configurations or different positions. Fully hydrogenating an oil instead largely removes its double bonds. Trans fatty acids also form through microbial biohydrogenation in the rumen of cattle, sheep, and goats and consequently occur in their milk and body fat. Vaccenic acid, trans-, is a characteristic ruminant example. Industrial and ruminant fats contain overlapping individual fatty acids but typically different distributions of positional and geometric isomers.
Food sources and patterns of consumption
Dietary fat occurs both as a concentrated ingredient and within the structure of foods. Vegetable oils, butter, ghee, lard, and other rendered animal fats supply predominantly fat, whereas nuts, seeds, avocados, dairy products, meat, eggs, and fish supply fat alongside water, protein, and other constituents. Their fat contents vary substantially: most cooking oils contain approximately 100 g of fat per 100 g, butter about 80 g, and many nuts approximately 45–75 g. Whole milk contains roughly 3–4 g per 100 g, while reduced-fat and skimmed products contain less. Meat fat varies with species, cut, trimming, and marbling; in eggs, nearly all the fat is in the yolk. Manufactured foods—including biscuits, pastries, chocolate, ice cream, sauces, and fried snacks—combine fats with other ingredients, so their fat content depends on formulation and preparation.
The characteristic fatty-acid composition differs among these sources, but no ordinary edible fat consists exclusively of one fatty-acid class. Olive oil commonly contains approximately 70% monounsaturated fatty acids, principally oleic acid, while conventional canola oil contains about 60%. Soybean, corn, and conventional sunflower oils contain substantial proportions of polyunsaturated fatty acids, especially linoleic acid; high-oleic sunflower and safflower varieties are important exceptions. Butterfat contains roughly two-thirds saturated fatty acids, palm oil about one-half, and coconut oil approximately 80–90%, expressed as proportions of total fatty acids rather than of the whole food. Palm oil, extracted from the fruit flesh, differs from palm-kernel oil, whose higher saturated-fat content more closely resembles coconut oil. Cultivar, animal feed, growing conditions, and processing can all affect these profiles.
Sources of omega-3 fatty acids also differ by food type. Flaxseed, chia seeds, walnuts, and their oils provide alpha-linolenic acid (ALA); canola and soybean oils also contribute ALA, although at lower concentrations than flaxseed oil. Oily fish such as salmon, sardines, herring, and mackerel provide the longer-chain fatty acids eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA). Lean white fish generally contain less total fat and, per serving, usually less EPA and DHA. Marine microalgae produce these fatty acids and supply commercial algal oils, including DHA-rich products. Fish composition varies with species, season, wild or farmed origin, and feed. Thus, describing a food as an “omega-3 source” does not specify which fatty acids it supplies or their amounts.
“Added fats” are fats introduced during manufacture, cooking, or serving—for example, oil used to fry vegetables, butter spread on bread, or shortening incorporated into pastry. Intrinsic fats are already present in the food, as in walnuts, milk, egg yolk, or meat. This distinction concerns how fat enters a dish, not a separate chemical category: olive oil added to a salad and fat within an olive share related constituents. It also complicates consumption measurement. Oil purchased for frying is not necessarily all eaten, while fat in restaurant dishes or mixed foods may not be apparent from the dish’s name.
Geographic patterns reflect agricultural production, trade, prices, culinary traditions, and industrial demand. Olive oil is prominent in many Mediterranean cuisines; soybean oil is widely used in the United States and Brazil; palm oil is important in Indonesia and Malaysia; and mustard or rapeseed oils feature in parts of South Asia. These are regional tendencies rather than uniform national diets. Industrial food production can change exposure without changing household cooking practices: reformulating biscuits, spreads, or frying oils alters the fats consumed through manufactured foods. National food-supply statistics therefore describe availability rather than actual ingestion, because they also encompass waste and do not show how intake is distributed among individuals.
Processing, cooking, and food functionality
Oil extraction separates lipids from plant tissues by mechanical pressing, solvent extraction, or a combination of the two; industrial soybean processing commonly uses hexane to recover oil from prepared flakes. Rendering releases animal fat by heating fatty tissues and separating the melted fat from water and solids. Subsequent refining removes unwanted constituents rather than deliberately changing fatty-acid saturation. Degumming removes phospholipids and associated material, while alkali neutralization removes free fatty acids as soaps. Bleaching generally uses adsorbent earths, sometimes with activated carbon, to remove pigments, residual soaps, and trace contaminants; it does not ordinarily mean treatment with household-type bleach. Deodorization uses steam under vacuum, commonly at approximately 180–260 °C, to strip volatile odor compounds and, in physical refining, free fatty acids. These operations can also remove some tocopherols and other minor constituents.
Hydrogenation deliberately changes fatty-acid double bonds by reacting oil with hydrogen, usually using a nickel catalyst. Partial hydrogenation raises the solid-fat content but can also move double bonds and convert some from cis to trans configuration. Full hydrogenation removes nearly all double bonds, producing a hard, highly saturated fat with little residual trans fat; this material can be blended with liquid oils for further processing. Fractionation instead separates higher- and lower-melting triglyceride fractions through controlled crystallization and filtration, without chemically saturating their fatty acids. Palm olein, the more liquid fraction, and palm stearin, the harder fraction, illustrate this process. Their fatty-acid proportions differ because different triglycerides partition into the two fractions.
Interesterification redistributes fatty-acid groups among triglyceride molecules or among positions on their glycerol backbones. Chemical catalysts generally produce a comparatively random redistribution, whereas some enzymes preferentially exchange fatty acids at particular positions. Unless combined with another process, interesterification does not change the overall fatty-acid inventory or itself convert cis bonds to trans bonds. Nevertheless, it changes which triglycerides are present and therefore their crystallization, melting range, and solid-fat content at a given temperature. Manufacturers can use blends of liquid oil and fully hydrogenated fat to obtain a spreadable or bakery fat without partial hydrogenation. Crystal size and crystal form also matter: fine crystals can produce smooth textures, whereas large crystals may produce graininess.
During frying, fat transfers heat to food at temperatures commonly around 160–190 °C, allowing surface dehydration and browning while the food interior remains cooler. In pastry, fat coats flour particles and limits their hydration and the development of a continuous gluten network—the “shortening” effect. Discrete layers of fat also help separate dough sheets in laminated pastry. During cake mixing, suitably plastic fat retains small air bubbles that expand during baking; completely liquid oil does not provide the same crystal-supported aeration. Margarine is usually a water-in-oil emulsion supported by a fat-crystal network, whereas mayonnaise is an oil-in-water emulsion stabilized principally by egg-yolk components. In both, droplet size and interactions among droplets influence viscosity and stability. Fat also dissolves many aroma compounds, affecting their retention and release during chewing and warming.
Hydrolytic rancidity results from hydrolysis of triglyceride ester bonds, releasing free fatty acids. Water, heat, and lipase activity can contribute; released short-chain fatty acids help explain the strong rancid odor of deteriorated butter. Oxidative rancidity instead involves reactions of unsaturated fatty acids with oxygen. Radical-chain oxidation first produces lipid hydroperoxides, which can decompose into aldehydes, ketones, and other compounds responsible for stale or paint-like odors. Light can promote oxidation through photosensitizers, while heat and trace metals accelerate relevant reactions. Antioxidants such as tocopherols can interrupt radical propagation, but their effectiveness depends on concentration, the food matrix, and processing conditions.
Smoke point marks visible smoke formation under specified conditions, not a comprehensive measure of cooking stability. Free fatty acids and impurities can lower it, so refining often raises smoke point even without substantially changing fatty-acid saturation. Oxidation can nevertheless occur below that temperature. During prolonged or repeated frying, oxygen exposure, food moisture, and heating promote hydrolysis, oxidation, and polymerization; degradation products accumulate and the oil may become more viscous or foam. Fatty-acid composition, antioxidant content, heating duration, and previous use therefore describe cooking behavior more fully than smoke point alone.
Digestion, absorption, and transport
Digestion of dietary triglycerides begins before food reaches the small intestine. Lingual and gastric lipases initiate hydrolysis, while muscular mixing in the stomach disperses fat into droplets. Most digestion occurs in the small intestine, where fat entering the duodenum stimulates release of the hormone cholecystokinin, promoting gallbladder contraction and pancreatic enzyme secretion. Secretin stimulates bicarbonate-rich secretions that help neutralize acidic stomach contents. Bile, produced by the liver, supplies bile salts and phospholipids that help disperse fat and stabilize droplets, increasing the surface accessible to digestive enzymes. Bile salts are not enzymes: their principal contribution is to organize otherwise poorly water-soluble lipids within the intestinal fluid.
Pancreatic triglyceride lipase acts at the oil–water interface, preferentially cleaving the ester bonds at positions 1 and 3 of triglycerides. Its main products are therefore two free fatty acids and a 2-monoacylglycerol. Colipase, released by the pancreas as a precursor and activated by trypsin, anchors lipase at the droplet surface in the presence of bile salts. Other pancreatic enzymes digest additional food lipids: phospholipase A2 removes a fatty acid from phospholipids, while pancreatic carboxyl ester lipase hydrolyzes cholesterol esters and other lipid esters. These complementary reactions release products that can be absorbed, rather than requiring intact dietary triglycerides to cross the intestinal lining.
Bile salts combine with fatty acids, monoacylglycerols, cholesterol, lysophospholipids, and fat-soluble vitamins to form mixed micelles. These small aggregates carry poorly water-soluble substances through the unstirred aqueous layer adjacent to the intestinal brush border. Micelles do not enter intestinal absorptive cells, or enterocytes, intact; their lipid components leave the aggregates and enter cells through diffusion and protein-mediated processes. Most bile salts remain in the intestinal lumen until they are recovered in the terminal ileum and returned to the liver through the portal circulation, a recycling process called enterohepatic circulation. Within enterocytes, long-chain fatty acids are activated to fatty acyl-CoA and largely reassembled with monoacylglycerols into triglycerides in the endoplasmic reticulum.
Reassembled triglycerides are packaged with cholesterol, phospholipids, and apolipoprotein B-48 into chylomicrons, with microsomal triglyceride transfer protein supporting particle assembly. Chylomicrons leave enterocytes and enter intestinal lymphatic capillaries called lacteals. They pass through the lymphatic system and thoracic duct into venous blood, initially bypassing the liver’s portal circulation. In blood, they acquire exchangeable apolipoproteins, including apolipoprotein C-II, which activates lipoprotein lipase. This enzyme acts on the luminal surface of capillaries, particularly in adipose tissue, skeletal muscle, and the heart, hydrolyzing chylomicron triglycerides. Released fatty acids can enter nearby tissues for storage or energy metabolism. The resulting triglyceride-depleted remnants retain apolipoprotein B-48 and are taken up by the liver through pathways involving apolipoprotein E.
Transport varies with fatty-acid chain length. Short-chain fatty acids and much of the absorbed medium-chain fraction pass more directly into portal blood, with fatty acids circulating largely bound to albumin, rather than requiring chylomicron packaging. This distinction is not absolute: chain length, meal composition, and intestinal metabolism influence the route, and the 12-carbon fatty acid laurate participates substantially in chylomicron transport. Impaired bile delivery, as in biliary obstruction, reduces micellar solubilization; pancreatic exocrine insufficiency reduces enzymatic hydrolysis; and intestinal mucosal disease or extensive resection reduces absorption. These disturbances can produce steatorrhea—excess fat in stools—and impair absorption of vitamins A, D, E, and K. Ileal disease or resection can additionally disrupt bile-salt recycling, reducing the bile-salt pool available for subsequent meals.
Energy metabolism and storage
Dietary fat is conventionally assigned a metabolizable energy value of approximately 9 kilocalories per gram, equivalent to about 37 kilojoules per gram. This general Atwater factor estimates energy available to the body after accounting for digestive and excretory losses; it is not the same as the gross heat released when fat is completely burned in a bomb calorimeter, typically around 9.4 kilocalories per gram. Actual metabolizable energy varies with fatty-acid composition, digestibility, and food structure. Fat therefore supplies more than twice the conventional energy per gram of carbohydrate or protein, each assigned approximately 4 kilocalories per gram. These values describe nutritional energy availability, not the amount conserved directly as adenosine triphosphate (ATP).
Before oxidation, a fatty acid is activated to fatty acyl–coenzyme A by an acyl-CoA synthetase. The reaction converts ATP to AMP and pyrophosphate, costing the equivalent of two high-energy phosphate bonds. Long-chain acyl groups enter the mitochondrial matrix through the carnitine shuttle: carnitine palmitoyltransferase I transfers the acyl group to carnitine, a translocase carries acylcarnitine across the inner membrane, and carnitine palmitoyltransferase II regenerates acyl-CoA inside. Repeated cycles of beta-oxidation then shorten the chain by two carbon atoms, producing acetyl-CoA and reduced electron carriers. Acetyl-CoA enters the citric acid cycle, while electrons from NADH and FAD-linked reactions support oxidative phosphorylation. Under standard biochemical accounting, complete oxidation of the 16-carbon saturated fatty acid palmitate yields approximately 106 ATP net; this estimate depends on assumed coupling efficiencies rather than representing a fixed yield in every cell.
The principal long-term energy reserve is triglyceride stored in lipid droplets within adipose tissue. After meals, fatty acids released from circulating triglyceride-rich lipoproteins can enter adipocytes and be esterified with glycerol-3-phosphate. Insulin promotes storage by increasing adipocyte glucose uptake, supporting triglyceride synthesis, and suppressing intracellular fat breakdown. During fasting or increased energy demand, adipose triglyceride lipase, hormone-sensitive lipase, and monoacylglycerol lipase sequentially hydrolyze stored triglycerides. Catecholamine signaling through beta-adrenergic receptors can stimulate this lipolysis through cyclic AMP and protein kinase A. Released nonesterified fatty acids circulate largely bound to albumin and supply tissues such as skeletal muscle and liver; glycerol is used chiefly by the liver and kidneys. Some released fatty acids are re-esterified rather than immediately oxidized.
The liver directs incoming fatty acids toward oxidation, triglyceride storage, or export in very-low-density lipoproteins. During prolonged fasting or marked carbohydrate scarcity, low insulin and increased fatty-acid delivery favor hepatic ketogenesis. As oxaloacetate is increasingly directed toward gluconeogenesis, acetyl-CoA from fatty-acid oxidation is diverted into ketone bodies, principally acetoacetate and beta-hydroxybutyrate. These circulate as fuels for extrahepatic tissues, including the brain during prolonged fasting. The liver produces ketone bodies but cannot use them for energy because it lacks succinyl-CoA:acetoacetate CoA transferase, the enzyme required for their utilization.
Fatty acids also arise through endogenous synthesis, particularly in the liver. Cytosolic acetyl-CoA is converted to malonyl-CoA by acetyl-CoA carboxylase; fatty acid synthase then uses malonyl-CoA and NADPH to produce predominantly palmitate. Malonyl-CoA inhibits carnitine palmitoyltransferase I, helping coordinate synthesis with mitochondrial oxidation. Dietary fat intake therefore does not translate directly into an equal increase in body-fat stores: net storage reflects the balance among dietary delivery, endogenous synthesis, oxidation, and mobilization over time. Stored triglycerides can contain fatty acids derived either from food or from synthesis using other energy substrates.
Essential fatty acids and physiological functions
Humans require dietary linoleic acid (LA; 18:2 n-6) and alpha-linolenic acid (ALA; 18:3 n-3) because human desaturase enzymes cannot introduce the double bonds needed to synthesize these fatty acids from oleic acid. In particular, humans lack the Δ12 and Δ15 desaturase activities found in plants. LA and ALA therefore supply the parent structures of the omega-6 and omega-3 families, respectively. “Essential” denotes this biochemical dependence on an external supply, not rarity in food or a requirement for large quantities. LA is abundant in many vegetable oils, while ALA occurs in foods such as flaxseed, walnuts, and rapeseed oil; abundance and physiological essentiality are separate properties.
Within each family, elongases add carbon atoms and desaturases introduce additional double bonds. LA can be converted through gamma-linolenic and dihomo-gamma-linolenic acids to arachidonic acid (20:4 n-6). ALA can yield eicosapentaenoic acid (EPA; 20:5 n-3), docosapentaenoic acid, and docosahexaenoic acid (DHA; 22:6 n-3). DHA synthesis involves further elongation and desaturation followed by chain shortening in peroxisomes. These pathways preserve the omega-family identity: omega-6 fatty acids do not become omega-3 fatty acids. Conversion of ALA to EPA is limited, and conversion to DHA is generally still smaller. Human tracer studies show substantial variation with sex, physiological state, background diet, and study method; no single conversion percentage applies universally. Preformed EPA and DHA bypass these conversion steps.
Fatty acids incorporated into membrane phospholipids influence cell-membrane organization, protein activity, and signaling. Phospholipase enzymes can release arachidonic acid and other polyunsaturated fatty acids from these membranes. Cyclooxygenase and lipoxygenase pathways then produce eicosanoids, including prostaglandins, thromboxanes, and leukotrienes, which participate in vascular tone, platelet activity, immune responses, and other local processes. EPA supplies alternative eicosanoid substrates, while EPA and DHA are precursors of specialized pro-resolving mediators, including resolvins; DHA also gives rise to protectins and maresins. Their effects depend on the particular molecule, receptor, tissue, and stage of a response. Describing all omega-6 fatty acids as inflammatory and all omega-3 fatty acids as anti-inflammatory obscures this diversity: arachidonic acid also supplies mediators involved in restraining or resolving inflammation.
Dietary fat also supports absorption of the fat-soluble vitamins A, D, E, and K. Products of fat digestion, together with bile salts, form mixed micelles that carry these vitamins through the watery intestinal environment toward absorptive cells; subsequent transport commonly involves chylomicrons. This function is not exclusive to essential fatty acids. DHA has more specific structural roles: it is highly enriched in retinal photoreceptor membranes and is an important component of neuronal membranes in the brain. Its membrane properties support visual and neural function. Arachidonic acid likewise contributes to developing tissues and cellular signaling.
Recognized essential-fatty-acid deficiency can produce scaly dermatitis, hair loss, impaired growth in infants and children, and poor wound healing. Biochemically, deficiency promotes synthesis of Mead acid (20:3 n-9) from oleic acid. An elevated Mead-acid-to-arachidonic-acid ratio, often called the triene-to-tetraene ratio, is an established biomarker, interpreted alongside clinical findings and the broader fatty-acid profile. Deficiency has been documented during prolonged fat-free intravenous feeding and can occur with severe fat malabsorption or markedly restricted lipid provision.
Infancy, pregnancy, and lactation involve lipid requirements for growth, tissue development, and transfer between mother and child. Fetal DHA accumulation is especially rapid late in gestation, and human milk supplies LA, ALA, arachidonic acid, and DHA, with composition varying in part with maternal diet and stores. Premature birth interrupts placental fatty-acid transfer before term. In long-term artificial nutrition, lipid provision must therefore account for essential-fatty-acid supply as well as energy; the fatty-acid composition of enteral formulas and intravenous lipid emulsions determines which essential fatty acids and longer-chain derivatives they deliver.
Blood lipoproteins and cardiovascular health
Dietary cholesterol is cholesterol consumed in food; circulating cholesterol is transported in blood within lipoprotein particles. LDL cholesterol (LDL-C) and HDL cholesterol (HDL-C) describe the amounts of cholesterol carried by low-density lipoprotein and high-density lipoprotein, respectively—not different cholesterol molecules or direct measurements of particle number. Dietary cholesterol can affect circulating concentrations, but intestinal absorption, hepatic synthesis, and clearance also regulate them, so intake does not translate directly into an equivalent rise in blood cholesterol. LDL and other apolipoprotein B–containing particles can enter and become retained in arterial walls, contributing to atherosclerosis. HDL participates in cholesterol transport, but a higher HDL-C concentration alone does not establish protection: medicines that raise HDL-C have not consistently reduced cardiovascular events.
Controlled feeding experiments show that replacing saturated fatty acids with cis-polyunsaturated or cis-monounsaturated fatty acids generally lowers LDL-C. One mechanism involves increased hepatic LDL-receptor–mediated clearance relative to diets rich in LDL-raising saturated fatty acids. Individual saturated fatty acids nevertheless differ. Lauric, myristic, and palmitic acids raise LDL-C relative to cis-unsaturated fats, whereas stearic acid has a substantially smaller effect and is often approximately neutral in these comparisons. Lauric acid also raises HDL-C, illustrating why changes in total cholesterol alone can obscure different lipoprotein responses. Foods contain mixtures of fatty acids, and their effects can also depend on food structure and other constituents; a food’s saturated-fat content is therefore informative but not a complete description of its cardiovascular effects.
Industrial trans fatty acids, historically supplied chiefly by partially hydrogenated oils, have an adverse lipoprotein profile. Compared with cis-unsaturated fats, they raise LDL-C and lower HDL-C; controlled feeding studies also document adverse changes in the ratio of total to HDL cholesterol. Ronald Mensink and Martijn Katan’s 1990 feeding trial helped establish that trans fats were not metabolically equivalent to cis-unsaturated fats. Prospective cohort studies subsequently associated higher trans-fat intake with greater coronary heart disease risk. The evidence connecting industrial trans fats with coronary disease combines these clinical associations with controlled evidence on lipoproteins rather than relying on long-term trials deliberately assigning high trans-fat consumption.
The replacement nutrient is essential to interpreting saturated-fat research. In an energy-matched comparison, reducing saturated fat necessarily increases another energy source. Replacement with polyunsaturated fat lowers LDL-C and has evidence supporting fewer cardiovascular events; replacement with refined starches or sugars does not provide equivalent evidence of benefit. Relative to unsaturated fat, carbohydrate-rich substitutions can increase triglycerides and lower HDL-C, although carbohydrate quality matters. Whole grains and fiber-rich foods are not interchangeable with refined carbohydrate. Analyses reporting an association with saturated fat without specifying what replaces it can consequently obscure materially different dietary comparisons.
Blood-lipid changes and cardiovascular events are distinct outcomes. Feeding studies can establish changes in LDL-C within weeks, whereas trials measuring myocardial infarction, stroke, or cardiovascular death generally require years and many more participants. A 2020 Cochrane review of 15 randomized trials involving 56,675 participants found that reducing saturated fat for at least two years reduced combined cardiovascular events by approximately 17%, but showed no clear reduction in overall or cardiovascular mortality. Such findings must be interpreted alongside differences in replacement foods, adherence, and concurrent treatment; a favorable lipid change does not establish an identical effect on every clinical endpoint.
Trials of concentrated omega-3 products likewise cannot be treated as direct tests of ordinary fish consumption. In REDUCE-IT, published in 2019, 8,179 statin-treated patients with elevated triglycerides and high cardiovascular risk received prescription icosapent ethyl, a purified EPA product, at 4 grams daily or a mineral-oil comparator; the primary cardiovascular endpoint was reduced by 25% in relative terms. In STRENGTH, published in 2020, 13,078 high-risk patients received a different 4-gram EPA–DHA formulation or corn oil, without a significant cardiovascular benefit. Differences in formulation, comparators, and patient populations limit direct comparison, and neither result can be transferred unchanged to dietary fish or typical over-the-counter supplements.
Body weight and other metabolic outcomes
Dietary fat supplies approximately 9 kilocalories per gram, compared with about 4 kilocalories per gram for carbohydrate or protein. Its high energy density allows relatively small portions to contain substantial energy, especially in foods with little water or fiber. Adding fat without reducing other energy sources can therefore increase total energy intake. However, the proportion of dietary energy supplied by fat does not, by itself, determine whether body weight rises. Sustained weight gain reflects energy intake exceeding expenditure; changes in appetite, physical activity, and energy expenditure affect that balance over time. Dietary fat can be stored efficiently, but neither its storage nor the hormonal response to a meal establishes that fat, or any other single macronutrient, inevitably causes obesity.
Randomized dietary-pattern trials demonstrate the distinction between macronutrient targets and achieved weight loss. In the 2009 POUNDS LOST trial, Frank Sacks and colleagues assigned 811 adults with overweight or obesity to four energy-restricted diets targeting either 20% or 40% of energy from fat and either 15% or 25% from protein. Participants lost approximately 6 kg at six months, followed by partial regain; average losses at two years were similar across the assigned diets. Attendance at counseling sessions was associated with greater weight loss, and achieved macronutrient differences were smaller than prescribed differences. In the 2018 DIETFITS trial, Christopher Gardner and colleagues randomized 609 adults to healthy low-fat or healthy low-carbohydrate diets without a prescribed calorie target. Mean 12-month weight losses were 5.3 and 6.0 kg, respectively, a difference that was not statistically significant. Both groups reported lower energy intake. These results do not establish that all diets are equivalent: they show that fat percentage alone cannot explain outcomes independently of food selection, total intake, and adherence.
Persistent excess energy intake can enlarge adipose stores and, in susceptible individuals, impair their capacity to accommodate additional lipid. Enlarged fat cells, altered immune-cell activity, and changes in adipose signaling can accompany increased release of fatty acids and reduced suppression of fat breakdown by insulin. Lipid accumulation in liver and skeletal muscle can then contribute to insulin resistance, including through lipid intermediates that interfere with insulin signaling. These processes are influenced by genetic susceptibility, fat distribution, activity, and the duration and degree of energy excess; they are not consequences of dietary fat intake alone.
Fatty liver disease likewise reflects several interacting pathways. Liver triglycerides derive from circulating fatty acids, dietary lipid delivered after meals, and fatty acids synthesized within the liver. Excess carbohydrate as well as excess fat can contribute to hepatic lipid accumulation under conditions of energy surplus. Insulin resistance can increase fatty-acid delivery from adipose tissue while hepatic lipid synthesis remains active. Consequently, total dietary fat is not a direct measure of liver-fat accumulation, and changes in liver fat need not parallel changes in total body weight exactly.
The fat content of a food also does not describe its complete nutritional characteristics. Nuts contain substantial fat together with protein, fiber, and cellular structures that can limit lipid release during digestion; oils provide fat without that intact food matrix. A lower-fat food may contain intact whole grains and fiber, or it may contain refined starch and added sugars. Protein content, water content, processing, texture, and eating rate can influence satiation and subsequent intake. Comparisons between dietary patterns therefore require attention to what replaces fat and to the foods supplying each macronutrient, rather than treating all low-fat or high-fat diets as nutritionally interchangeable.
Measuring intake and interpreting research
Dietary-fat intake is usually estimated by combining reported food consumption with food-composition databases, which assign nutrient values to foods, recipes, and portion sizes. Weighed or estimated dietary records document foods as they are consumed, but recording can alter eating behavior. A 24-hour recall reconstructs the preceding day through a structured interview; repeated recalls help characterize day-to-day variation. Food-frequency questionnaires ask about usual consumption over longer periods, often the previous year, but depend on memory and a predefined food list. Database estimates introduce additional uncertainty when recipes, brands, or cooking practices differ from those represented. Grams per day measure an absolute amount, whereas percentage of total energy expresses dietary composition. Using the conventional factor of 9 kilocalories per gram, 70 g of fat in a 2,000-kilocalorie diet supplies approximately of energy.
Laboratory measurements depend on both the food matrix and the analytical definition of fat. Solvent-extraction methods recover lipid fractions that can be weighed after solvent removal; acid hydrolysis before extraction helps release fat bound within some processed foods. The Folch method, published in 1957, and the Bligh–Dyer method, published in 1959, use chloroform–methanol mixtures, although their solvent proportions and extraction procedures differ. For fatty-acid profiling, extracted lipids are commonly converted to fatty-acid methyl esters and separated by gas chromatography, often with flame-ionization detection. Appropriate columns and reference standards are needed to distinguish chain lengths, degrees of unsaturation, and cis–trans isomers. Results reported as percentages of total identified fatty acids are not equivalent to grams per 100 g of food. Solvent choice, incomplete extraction, oxidation, and whether “total fat” is determined gravimetrically or calculated from measured fatty acids can limit comparisons.
Blood and tissue fatty acids provide biomarkers that complement, rather than replace, dietary reports. Plasma triglycerides are relatively responsive to recent meals, whereas erythrocyte fatty-acid composition generally reflects exposure over weeks to months; adipose-tissue composition integrates a longer period, often months to years. These windows are approximate because lipid fractions and individual fatty acids turn over at different rates. Biomarkers are especially informative for fatty acids that humans cannot synthesize, such as linoleic acid, but their concentrations still depend on absorption, oxidation, tissue exchange, and lipid metabolism. Palmitic and oleic acids are less specific indicators of dietary intake because endogenous synthesis contributes substantially. Conversion of alpha-linolenic acid to longer-chain omega-3 fatty acids further complicates interpretation. A relative percentage can also fall because another fatty acid rises, without any decline in its absolute concentration.
Controlled feeding studies specify the foods supplied and can isolate the effects of changing fat composition, particularly on blood lipoproteins. Their duration and artificial setting, however, often limit conclusions about long-term disease. Prospective cohort studies can follow large populations for years, but dietary measurement error and confounding by smoking, physical activity, socioeconomic circumstances, and other foods complicate causal interpretation. Repeated dietary assessments can capture changes and reduce some misclassification. A randomized dietary trial reduces baseline confounding through random allocation, but incomplete adherence, participant dropout, and convergence between comparison diets can weaken the achieved contrast. Provision of foods and biomarker measurements help document that contrast. Changes in LDL cholesterol or other intermediate markers are not themselves measurements of heart attacks, strokes, or mortality; clinical outcomes require sufficiently large and long studies.
Energy adjustment makes the comparison underlying an association explicit. Investigators may express fat as a percentage of energy or use residual methods to distinguish intake from overall food consumption. In an isocaloric substitution model, increasing one energy-providing nutrient necessarily decreases another. For example, replacing 5% of energy from carbohydrate with fat in a 2,000-kilocalorie diet corresponds to approximately 11 g more fat and 25 g less carbohydrate daily. A model holding total energy and protein constant therefore answers a different question from one in which fat replaces protein or adds calories. Interpretation also depends on whether the displaced carbohydrate comes from whole grains or refined starches and sugars, and whether the added fat is predominantly saturated or unsaturated.
Development of scientific understanding
Nineteenth-century chemistry helped establish that food fats were chemically distinct substances rather than interchangeable sources of nourishment. Michel-Eugène Chevreul’s investigations, consolidated in his 1823 treatise on animal fats, distinguished several fatty acids and clarified their relationship to glycerol. In the mid-nineteenth century, Claude Bernard’s experiments established the importance of pancreatic secretions in fat digestion. Later, Wilbur Olin Atwater combined food-composition analysis, measurements of excreted material, and human calorimetry to quantify the energy available from foods. In the 1890s, Atwater and physicist Edward B. Rosa developed a human respiration calorimeter that compared heat production with respiratory exchange and material balances. The resulting Atwater energy factors distinguished metabolizable energy from the heat released by complete combustion: the general factors eventually became approximately 9 kilocalories per gram for fat and 4 each for protein and carbohydrate.
The recognition that particular fats were indispensable came from experiments by George O. Burr and Mildred M. Burr. Their 1929 paper described rats receiving a highly purified, fat-free diet that developed poor growth and scaly skin lesions. Adding certain fats prevented or reversed the disorder, whereas supplying energy and known vitamins did not adequately explain the effect. Their subsequent work, published in 1930, identified linoleic acid as an effective dietary component, establishing the experimental basis for essential fatty acids. These were initially animal findings, not direct demonstrations of human requirements. Later clinical observations, including studies of infants fed diets extremely low in linoleic acid and cases involving prolonged fat-free intravenous nutrition, documented human deficiency. The distinction between fats as fuel and fatty acids as required nutrients therefore emerged through several decades of experimental and clinical work.
During the mid-twentieth century, researchers increasingly separated total fat intake from fatty-acid composition. Controlled feeding experiments by groups including Ancel Keys, Francisco Grande, Joseph Anderson, and D. Mark Hegsted showed that changing dietary fatty acids could alter serum cholesterol even when energy intake was controlled. Equations published by Keys and colleagues and by Hegsted and colleagues in 1965 quantified relationships between dietary composition and serum cholesterol. These studies supported the distinction between saturated and polyunsaturated fats rather than treating all fats as metabolically equivalent. Their principal outcomes were blood measurements, however, not direct demonstrations of changes in heart-attack incidence.
The Seven Countries Study, launched in 1958 under Keys’s leadership, extended this research into population comparisons. It enrolled 12,763 men aged 40–59 in 16 cohorts across the United States, Finland, the Netherlands, Italy, Yugoslavia, Greece, and Japan. Standardized examinations and follow-up related dietary patterns, serum cholesterol, and other characteristics to coronary heart disease. Differences in saturated-fat consumption and average cholesterol contributed to the developing diet–heart hypothesis. Because participants and diets were not randomly assigned, the study could not isolate dietary fat from all accompanying social, behavioral, and food-pattern differences. Its findings were interpreted alongside feeding experiments and other epidemiological evidence.
Long-term dietary intervention studies subsequently tested whether changing diets affected clinical disease. Trials differed in whether they reduced total fat, replaced saturated fat with vegetable oils, or changed several foods simultaneously. The Women’s Health Initiative Dietary Modification Trial, whose principal cardiovascular results appeared in 2006, randomized 48,835 postmenopausal women to a low-fat dietary intervention or a comparison group. Over an average 8.1 years, it found no statistically significant reduction in coronary heart disease or stroke. Because the intervention emphasized reducing total fat rather than a specific saturated-to-unsaturated-fat substitution, it addressed a different question from many earlier replacement experiments.
Later research also incorporated fatty-acid measurements in blood fractions and adipose tissue as biomarkers, supplementing dietary questionnaires. Prospective studies increasingly examined foods and dietary patterns, distinguishing, for example, nuts, fish, dairy products, and processed meats rather than grouping them solely by fat content. Nutrient-substitution analyses made the comparison explicit: replacing saturated fat with polyunsaturated fat was not the same exposure as replacing it with refined carbohydrate. These developments shifted research toward the type of fat, its food source, and the nutrient or food displaced.
Dietary reference standards, labeling, and regulation
Dietary reference values support assessment and planning of nutrient intakes in healthy populations; population guidelines translate evidence into objectives for disease prevention. Neither constitutes a personalized medical prescription. An acceptable macronutrient distribution range (AMDR) expresses intake as a percentage of total dietary energy, balancing nutrient adequacy with chronic-disease considerations. The Institute of Medicine’s United States–Canada Dietary Reference Intakes, released in 2002 and published in consolidated form in 2005, established an adult fat AMDR of 20–35% of energy. Different ranges apply to children: 30–40% at ages 1–3 and 25–35% at ages 4–18. These percentages describe energy contributions, not percentages of food weight.
The World Health Organization (WHO) issued a separate guideline in 2023 conditionally recommending that adults limit total fat to 30% of total energy or less to help prevent unhealthy weight gain. Its purpose and evidence framework differ from those of the AMDR; the figures are therefore not interchangeable definitions of nutritional adequacy. At an illustrative intake of 2,000 kilocalories daily, 30% of energy corresponds to approximately 67 grams of fat, using the conventional conversion of 9 kilocalories per gram. This calculation illustrates the standard rather than establishing an individual energy or fat requirement.
WHO’s companion 2023 guideline strongly recommends limiting saturated fatty acids to 10% or less and trans fatty acids to 1% or less of total energy in adults and children aged two years and older. The trans-fat limit includes both industrially produced and ruminant-derived forms. Replacement nutrients are explicit: WHO recommends replacing saturated fat with polyunsaturated fat, and conditionally recommends replacement with monounsaturated fat from plant sources or carbohydrates from foods containing naturally occurring dietary fiber, including whole grains, vegetables, fruits, and pulses. It recommends replacing trans fat with polyunsaturated fat or monounsaturated fat primarily from plant sources. These are substitution recommendations, not simply instructions to remove fat without specifying the resulting diet.
Nutrition-label requirements differ across jurisdictions. In the United States, the Nutrition Facts panel generally declares total fat, saturated fat, and trans fat in grams per serving, and cholesterol in milligrams. Daily Values for adults and children aged four and older are 78 grams for total fat, 20 grams for saturated fat, and 300 milligrams for cholesterol; no Daily Value is established for trans fat. Under rounding rules, less than 0.5 gram of trans fat per serving can be declared as 0 grams, so that declaration does not necessarily mean complete absence. In the European Union, mandatory nutrition declarations ordinarily state fat and saturates per 100 grams or 100 millilitres; per-portion information may supplement them. Trans fat and cholesterol are not part of the mandatory declaration. Regulated claims also differ: the EU permits “low fat” for solid foods containing no more than 3 grams per 100 grams, whereas the US criterion for many foods uses 3 grams or less per reference amount customarily consumed, with additional rules for certain categories.
WHO launched REPLACE in 2018 as a framework for eliminating industrially produced trans fat from the food supply through policy review, healthier replacement fats, legislation, monitoring, awareness, and enforcement. Its best-practice regulatory approaches are a mandatory national limit of 2 grams of industrially produced trans fat per 100 grams of total fat in all foods, or a national prohibition on producing or using partially hydrogenated oils as food ingredients. The product-composition limit is distinct from the dietary limit of 1% of energy.
Implementation has followed different legal routes. Denmark introduced a 2-gram-per-100-grams-of-fat limit in 2003, with implementation beginning in 2004. EU Regulation 2019/649 established the same numerical ceiling for trans fat other than that naturally occurring in animal fat, applying from April 2021. The US Food and Drug Administration determined in 2015 that partially hydrogenated oils were no longer generally recognized as safe, with the principal manufacturing compliance date in June 2018. Enactment, compliance, measured food composition, population intake, and subsequent cardiovascular disease outcomes are separate indicators; adoption of a regulation alone does not establish the magnitude of its health effects.