A Hit and a Miss

Ultra-Processed Products

Packaged food often contains ingredients that keep it creamy, sweet, colorful or fresh for longer. Some of these ingredients can also affect the bacteria living in our intestines. The concern is what happens when those bacteria change, or when the gut's protective lining becomes less effective. Different additives work differently, so I looked at them one at a time.

Emulsifiers can affect the space between bacteria and the lining. Emulsifiers stop oil and water from separating in foods such as dressings and ice cream. Inside our intestines, a layer of mucus helps keep bacteria away from the living cells underneath. If bacteria move closer to those cells, the immune system has more reason to react. Two emulsifiers, carboxymethylcellulose and polysorbate-80, caused this kind of change in a mouse experiment, along with low-level inflammation and worse colitis in susceptible mice. Colitis is inflammation of the colon, the large intestine.

A thinner margin of protection could mean more irritation. The mucus layer is useful because we need gut bacteria without having them constantly crowding the intestinal surface. Changes in that community can also change the substances the bacteria make. One emulsifier, carboxymethylcellulose or CMC, has been tested in a small human feeding trial. People eating it had some changes in their gut bacteria and slightly more discomfort after meals; in two people, bacteria also moved closer to the lining. That short trial raised a concern, but it did not establish lasting inflammatory disease.

Sweeteners may change more than a food's sweetness. Some sweeteners reach gut bacteria, and changes in those bacteria may affect how the body handles sugar. That could mean a different blood-sugar response even though the sweetener itself supplies little or no sugar. A human trial found poorer glucose handling with saccharin and sucralose, with responses varying between people. Other sweeteners changed aspects of the bacterial community without the same glucose result. This is a possible effect on blood-sugar control, rather than evidence that every sweetener damages the stomach.

Preservatives can hold back some bacteria more than others. A preservative stops microbes from growing in food. If it affects gut bacteria unevenly too, it could leave some species with less competition and shift the balance. In laboratory tests, Bacteroides coprocola was about 580 times more sensitive to sodium nitrite than Enterococcus faecalis. That was a comparison between two particular bacteria, not a rule that preservatives kill all helpful bacteria. The amount reaching the gut after a meal also matters; these tests did not measure that.

Some colorings can set off immune reactions in experiments. The immune system responds when cells detect something potentially harmful. Tiny particles of titanium dioxide, a white coloring, activated one of these alarm systems in laboratory-grown cells and worsened colitis in mice already given a chemical that injured the colon. Gut bacteria can also break down dyes into other substances. Red 40 and Yellow 6 triggered colitis through this route in mice with an unusually active immune signal. These findings concern those particular conditions, rather than every person eating a colored food.

Dairy Products

As babies, we usually make plenty of lactase, the enzyme that breaks down lactose, the sugar in milk. Many of us make less of it as we grow older, which can make milk harder to digest. Others keep making enough throughout adulthood. Milk also contains proteins, including casein, that can cause problems for different reasons. Understanding which part is involved helps explain why two people can react very differently to the same glass of milk.

Undigested lactose can lead to gas, cramps and diarrhea. Lactase works in the small intestine, breaking lactose into smaller sugars we can absorb. When there isn't enough lactase, more lactose reaches the colon. Bacteria there feed on it and make gas, while the unabsorbed sugar draws extra water into the bowel. That combination can leave you bloated, crampy or needing the toilet. This is lactose intolerance: trouble digesting the sugar, rather than an immune attack on the gut. NIDDK explains this process in the sources below.

Casein may affect how food moves through the gut. Casein is a milk protein that digestion breaks into smaller pieces. One version, called A1 beta-casein, can release a fragment known as BCM-7. This fragment can attach to receptors, the receiving points on cells, involved in the contractions that move food along. Animal research suggests it can slow those contractions; slower movement is one possible explanation for discomfort. A small trial in adults who already reported milk intolerance found more symptoms and slower gut movement with mixed A1/A2 milk than with A2-only milk. The fragment's role in those human symptoms is still being investigated, and A2 milk still contains lactose.

Milk protein can also trigger an immune reaction. In someone with a milk allergy, the immune system reacts to milk proteins, which can include casein. That reaction can cause abdominal pain, vomiting or diarrhea, as well as symptoms elsewhere in the body. Removing lactose does not remove those proteins. There is also research into how diet affects the intestinal barrier: the closely joined cells that control what passes from the gut into the body. A diet using casein already broken into smaller pieces improved this barrier in diabetes-prone rats. Several parts of their diet differed, so that finding doesn't show that ordinary casein opens gaps in a person's gut.

Some families keep the ability to digest milk. Inherited gene variants can keep lactase production going after childhood. This is common in northern Europe and in some African and Middle Eastern populations with a long history of herding and drinking milk. Skin color cannot tell us how much lactase someone makes. One theory is that milk's calcium was especially useful where less sunlight meant less vitamin D. Vitamin D helps the body absorb calcium. It is a possible explanation for an inherited trait becoming more common over generations, rather than a person's body making lactase because it needs vitamin D. Genetic research documents the different populations; a later European study also points to famine and infections as possible forces behind the change.

Wheat & Gluten

The concern with wheat isn't only celiac disease. A component of gluten called gliadin can affect the seals between intestinal cells. Researchers have seen this response in tissue from people with and without celiac disease. That is the connection to zonulin shown in the illustration. Other wheat proteins can also affect immune cells, so it helps to understand the different parts of the grain.

Gliadin can trigger the release of zonulin. Gluten is a group of proteins that gives dough its stretch, and gliadin is one of its components. The cells lining our intestine sit next to each other, joined by seals called tight junctions. Zonulin is a protein that helps regulate those seals. In the pathway studied, gliadin prompts intestinal cells to release zonulin, which signals the junctions to loosen. More material can then pass between the cells; this is what increased intestinal permeability means. The linked cell and tissue study found this response in non-celiac samples too, although it was shorter and smaller than in celiac samples.

Looser seals can expose immune cells to more material. Beneath the intestinal surface are immune cells that help defend us. If more food fragments or bacterial material get through, those cells may react and release inflammatory signals. A barrier change alone doesn't tell us how much inflammation follows. In celiac disease, gluten also triggers an immune reaction that damages the small projections used to absorb nutrients. A later biopsy experiment found increased permeability in all four groups tested, including non-celiac controls. This supports a response beyond celiac disease, but the experiment did not show lasting injury in everyone after eating a meal.

Wheat's protective proteins can activate immune alarms. Plants make proteins called amylase-trypsin inhibitors, or ATIs, that help protect them against pests. Some ATIs can also activate a receiving point on immune cells called TLR4. When that alarm switches on, the cells release messages that call other immune cells into action. Repeated or excessive activation could contribute to inflammation. Researchers observed this response in human cells, intestinal samples and mice. That gives us a possible mechanism, but it does not show that every wheat meal causes this reaction in people.

Another wheat protein can stick to cell surfaces. Wheat germ agglutinin, usually shortened to WGA, belongs to a group of proteins called lectins. Lectins attach to certain sugars on cells. In laboratory experiments, WGA stimulated immune cells and, at higher concentrations, made a layer of intestinal cells less effective at holding substances back. A weaker barrier could allow more material to reach immune cells underneath. These were cells exposed directly to WGA in a dish; the experiment did not show the same damage after people ate ordinary cooked wheat.

Whole Grains & Their Antinutrients

Whole grains keep the outer bran and the germ, the part of the seed that can grow into a new plant. These parts contain fiber and nutrients, but also compounds that can bind minerals or make parts of a meal harder to digest. They are often called antinutrients. The name can sound alarming, so it helps to explain what each one actually does. Reduced nutrient absorption and inflammation are different problems.

Phytate can hold onto minerals in a meal. Phytate is a form in which plants store phosphorus. It can bind minerals such as iron and zinc while food is being digested, leaving less available for us to absorb. This can matter more when someone relies heavily on the same foods and already gets few of those minerals. It doesn't mean a grain scrapes or burns the intestinal lining. The linked review describes mineral binding and how preparation can change the amount of phytate.

Tannins can make protein and starch harder to break down. Tannins are plant compounds that can attach to proteins and interfere with digestive enzymes, the substances that break food into smaller pieces. If less food is broken down, less may be absorbed at that stage of digestion. But that does not automatically mean more gas or inflamed tissue. In a simulated digestion experiment, sorghum tannins reduced protein and starch digestion; the later fermentation stage also produced less gas and fewer fermentation products. It was a laboratory model, rather than a meal tested in people.

Saponins can interact with the covering around cells. Saponins are plant compounds that can behave a little like soap, which is why some make foam in water. Certain saponins can interact with cell membranes and make them easier for substances to cross. Oat saponins increased passage through isolated rat intestinal tissue in one experiment. But when the researchers tested live rats, they did not find the same increase in intestinal permeability. That difference matters: the tissue result alone cannot tell us that eating oats damages the gut.

Preparation changes what reaches the gut. Soaking, sprouting and fermentation can help break down phytate, while cooking can change other compounds and make starch easier to digest. The result depends on the grain and the method used. A fermented dough and an unprepared grain are different foods by the time we eat them. Research on these methods supports paying attention to preparation, but there isn't one reduction percentage that applies to all antinutrients or a guaranteed link between eating whole grains and gut injury.

Dense Acellular Carbohydrates

Flour and many foods made from it contain a lot of carbohydrate packed into a small amount of food. One concern is that processing changes how easily gut bacteria can get at that carbohydrate. If it changes which bacteria thrive, it might also affect digestion and inflammation. This is the idea Ian Spreadbury explored in his paper. It remains a hypothesis about a chain of events, rather than a proven explanation for everyone's symptoms.

Processing removes some of the food's original structure. In a plant, much of the carbohydrate sits inside cells with walls around them. Milling grain into flour breaks many of those walls and releases their contents. Spreadbury called these carbohydrates acellular, meaning they are no longer enclosed in intact plant cells. This can leave more carbohydrate readily available to digestive enzymes and microbes. Cooking changes food structure too, so the difference isn't simply raw versus cooked, and there isn't a single carbohydrate percentage that makes a food safe or harmful.

More available fuel could favor different bacteria. Gut bacteria use the food available to them to grow. Spreadbury proposed that frequent meals of concentrated, exposed carbohydrate could favor a different bacterial community, including in the upper digestive tract. Some microbes might multiply more easily while others lose ground. As they break down carbohydrate, they produce gas and other substances. More fermentation in the wrong place could mean bloating, pressure and discomfort. This is the proposed bacterial step in his theory; it does not establish that eating flour causes small intestinal bacterial overgrowth.

Overgrowth can interfere with digestion as well as cause gas. When too many bacteria are present in the small intestine and cause symptoms, the condition is called SIBO. Bacterial activity there can contribute to bloating, abdominal pain and diarrhea, and in severe cases interfere with nutrient absorption. That can lead to weight loss or deficiencies, including vitamin B12 deficiency. These are recognized consequences of diagnosed SIBO, described in the clinical guideline below. Gas alone cannot identify it, and the guideline does not establish acellular carbohydrates as its cause.

Bacterial material could keep setting off immune alarms. Some bacteria carry a substance called LPS on their outer surface. The intestinal lining normally helps keep it on the gut side. If more gets across that barrier, immune cells can detect it through a receiving point called TLR4. They then send out signals that bring other immune cells into action. That is useful when the body needs to fight infection, but repeated stimulation can keep inflammation going. Spreadbury proposed that a changed bacterial community could increase this exposure. The complete path from acellular food to this inflammation in people is still unproven.

The proposed effects go beyond an uncomfortable belly. The gut sends messages to the brain through nerves, including the vagus nerve. Fat tissue also releases leptin, a hormone that tells the brain about stored energy and helps regulate appetite. Spreadbury suggested that ongoing bacterial inflammatory signals could make parts of this communication less responsive to leptin. If the messages become less effective, a person might eat more and gain weight. He also proposed links to insulin resistance, when cells respond less well to insulin and blood sugar becomes harder to control. These are possible wider consequences in his hypothesis, rather than established results of eating these foods.

Quickly digested carbohydrates may also bring hunger back sooner. This is a separate issue from the bacterial theory. Digestion turns starch into smaller sugars, and glucose enters the blood. Insulin helps cells use or store it. A quickly digested meal can cause a sharp rise, followed in some people by a larger dip. That later dip may leave them hungry sooner and reaching for more food. Wyatt's large human study linked appetite and earlier eating more closely to the dip two to three hours after a meal than to the initial spike. Lennerz's smaller meal trial also found more later hunger after a meal that raised blood sugar quickly. The response varies with the food, portion and person.

Food structure is only part of the picture. How much we eat and the rest of the diet also affect weight and digestion. Bacteria make substances as they ferment food, including butyrate, so changing the foods they receive can change those products without necessarily causing disease. In the human CARBFUNC trial, diets emphasizing cellular or acellular carbohydrates produced similar reductions in fat around the abdominal organs. A later report found some differences in stool butyrate and modest reflux improvement with the lower-carbohydrate diet, but no clear overall difference in bowel-symptom scores. Neither report demonstrated that acellular carbohydrates caused SIBO or inflammation.

Legumes & Their Antinutrients

Beans, lentils and soy can cause digestive trouble for different reasons. Some carbohydrates make gas when bacteria break them down. Other compounds can interfere with digestion, especially in raw or poorly prepared beans. Raw kidney-bean poisoning is a clear example, but it should not be confused with the bloating someone might feel after a serving of cooked beans.

Raw kidney beans contain a lectin that can cause poisoning. Lectins are proteins that attach to particular sugars on cells. Kidney beans contain one called phytohemagglutinin, or PHA. If enough active PHA is eaten, it can cause nausea, vomiting, abdominal pain and diarrhea. Raw-bean experiments in rats also found poorer nutrient absorption and bacterial overgrowth in the small intestine. A UK report recorded 50 suspected poisoning incidents, nine confirmed by testing. These concerns are about raw or inadequately cooked kidney beans, rather than the same exposure from properly cooked beans.

Some soybean compounds get in the way of protein digestion. The pancreas supplies an enzyme called trypsin that helps cut protein into smaller pieces. Soybeans contain proteins called trypsin inhibitors that block some of that activity. If the inhibitor stays active, protein digestion becomes less efficient and the pancreas may respond by making more digestive enzymes. A rat feeding experiment found a larger pancreas and increased enzyme-production measurements with more active soybean inhibitor. It doesn't establish that a normal serving of cooked soy causes pancreatic damage in people.

Saponins can affect the surface around cells. These plant compounds can interact with cell membranes, the thin coverings that help control what goes into and out of a cell. Certain saponins can make substances pass through intestinal tissue more easily, raising a question about barrier protection. An isolated-intestine experiment found that some did this much more than others; soybean saponins had much less effect. Direct contact with tissue in a laboratory is different from digesting cooked beans, and this experiment did not show permanent holes in people's intestines.

Some bean carbohydrates become food for bacteria. We don't fully digest certain small chains of sugars in legumes. They pass into the colon, where bacteria break them down and release gas. That can leave you bloated or uncomfortable, especially after more than you usually tolerate. This process is fermentation, a normal part of what gut bacteria do. It can cause unpleasant symptoms without showing that the lining is inflamed. NIDDK explains this connection between undigested carbohydrates, bacteria and gas.

Proper preparation removes the raw-kidney-bean risk. Heat destroys the dangerous lectin when the beans are prepared correctly. For raw kidney beans, the FDA advises soaking them for at least five hours, throwing away that water, then boiling them in fresh water for at least 30 minutes. A slow cooker should not replace that boiling step because it may not get hot enough. Canned kidney beans have already been processed. Soaking and cooking can also change other compounds in legumes, although the amount reduced depends on the food and method.

Seed Oils & Inflammatory Fats

The concerns about seed oils involve what happens to the oil during processing, repeated frying and digestion. Some breakdown products can put strain on cells or stimulate immune activity. The research reaches beyond the gut to blood vessels, fat tissue and the liver, although much of it comes from animal experiments. The oil used and the way it was treated matter to the explanation.

High-temperature refining can create unwanted compounds. Oil is extracted from seeds, then cleaned and often heated to remove odors. Some methods use a solvent that is removed afterward; its use alone doesn't tell us dangerous residues remain. A separate concern is the formation of glycidyl esters and 3-MCPD esters during high-temperature processing. Digestion can release glycidol from the first group. Glycidol can damage genetic material and caused cancer in animal tests; 3-MCPD caused kidney effects in animals. EFSA identified these as processing contaminants to control, with the highest levels in palm oils and fats, rather than exclusively in seed oils.

Some fats can break down into compounds that stress cells. Linoleic acid is an omega-6 fat found in many seed oils. It can react with oxygen and change, a process called oxidation. Some of the resulting substances can react with proteins or alter the messages cells send. One is called 4-HNE; others are grouped under the name oxidized linoleic acid metabolites, or OXLAMs. These names appear in the studies below because researchers have examined how these products affect cells and immunity. Their formation depends on the conditions, rather than every use of linoleic acid producing the same effect.

Repeated frying gives oil more opportunity to break down. Heat and oxygen change oil over time. Reusing it for several rounds of frying can allow reactive breakdown products to build up, including compounds called aldehydes that can interact with cell proteins. This is one reason repeatedly heated oil raises concerns about cell stress and inflammation. Several of the adverse findings below involve deliberately reused oil. They describe a different exposure from opening a fresh bottle or cooking with it once.

Changes in gut bacteria could weaken intestinal protection. Some bacteria can use linoleic acid as fuel. If an oil-rich diet favors bacteria that get close to the intestinal lining, the barrier may become less effective and inflammation easier to trigger. This was a concern in mice fed a high-fat diet rich in soybean oil. Their gut community changed, including expansion of a strain of E. coli associated with inflammatory bowel disease, and their lining allowed more material through. They were also more susceptible to colitis, inflammation of the colon. This was an intestinal finding in mice, rather than a demonstration of human stomach inflammation.

Reused oil may affect the walls of blood vessels. When a vessel wall becomes more inviting to immune cells, those cells can attach and contribute to inflammation there. Reactive oil breakdown products are one possible contributor to this process. Rats fed soybean oil heated five or ten times for six months developed higher blood pressure and more proteins that help immune cells attach to the main artery. The artery also showed tissue changes. The adverse pattern concerned repeatedly heated oil in that experiment; fresh oil did not produce the same pattern.

Fat tissue can become a site of immune activity. Body fat contains immune cells as well as cells that store energy. A fat breakdown product such as 4-HNE can change signals on a cell's surface, potentially drawing an immune response. In a 2024 experiment, a linoleic-acid-rich high-fat diet increased activity of a particular group of immune cells in mouse fat tissue compared with a diet rich in stearic acid, a saturated fat. Related cell experiments linked 4-HNE to a protein appearing on the cell surface that helped stimulate those immune cells. This supports a possible mechanism under those conditions, rather than proving that an ordinary meal causes inflamed fat tissue in people.

Alcohol can make the liver more vulnerable. The liver handles alcohol, and this can create reactive substances that strain its cells. If fat breakdown adds to that strain, the liver's defenses may struggle to keep up, allowing injury and inflammation. In one mouse experiment, a corn-oil-rich diet alongside alcohol exposure produced more liver injury and inflammation than a saturated-fat diet alongside alcohol. The animals also received an alcohol binge. The finding suggests that the dietary fat influenced alcohol-related injury; it does not show that corn oil caused the same liver damage on its own.

Diet can change the breakdown products measured in people. Researchers can look for these products in blood or urine to see how much fat oxidation is taking place. They are clues about a chemical process, rather than a diagnosis of inflammation. A controlled-diet study found an increased urine oxidation marker after a linoleic-acid-rich diet compared with the earlier saturated-fat diet. It did not find a significant difference from a diet rich in oleic acid, the main fat in olive oil, or a change in the measured vessel-inflammation marker. A separate intervention in people with chronic headaches found fewer blood OXLAMs after lowering dietary linoleic acid. These measurements alone don't establish clinical disease from seed oils.