What to Eat When You Have a Stomach Bug

Ask anyone what to eat when you can't keep anything down and you get four letters back. BRAT: bananas, rice, applesauce, toast. It's the top answer in the big threads on the question, and it's probably what somebody handed you when you were eight.

‍

The American Academy of Pediatrics retired that advice in 2021, and what replaced it is less of a rule. Eat whatever you normally eat, in smaller amounts, whenever you feel like eating. Drink more than you think you need, in small sips. Dairy is fine unless it bothers you. See a doctor for bloody diarrhea, a fever over 102°F, diarrhea past three days, or if you can't keep liquids down at all.

‍

The rest of this is why four letters survived for decades without anyone testing them.

‍

The evidence behind the four letters

‍

In 2021 the AAP wrote that the diet "is no longer considered useful." In 2024 Cleveland Clinic ran an article recommending it by name. Both of those are positions held by serious institutions, and they sit side by side because nobody has ever run the experiment.

‍

The closest thing to a serious published review of the BRAT diet comes from Debora Duro and Christopher Duggan, both at Children's Hospital Boston and Harvard Medical School. Their line on the evidence is one sentence: "To date, no clinical trials have been conducted to assess its effectiveness, although some data exist to evaluate the role of bananas and rice in treating diarrhea."

‍

The European pediatric gastroenterology society (ESPGHAN) turns that into a formal recommendation: "The bread, rice, apple, toast (BRAT) diet has not been studied and is not recommended." The Infectious Diseases Society of America, whose 2017 guideline is the one that explicitly covers adolescents rather than toddlers, says it this way: "Although the BRAT (bananas, rice, applesauce, and toast) diet and the avoidance of dairy are commonly recommended, supporting data for those interventions are limited. Instructing patients to refrain from eating solid food for 24 hours also does not appear to be useful."

‍

So BRAT was never studied, which is a different claim from BRAT having been studied and failed. What got it dropped is nutrition. Duro and Duggan modelled the diet against what a healthy toddler eats and found it "provides approximately 300 calories per day less," with very little fat, fiber, or protein. Those numbers are modelled for a hypothetical two-year-old rather than measured in anybody, and the two published cases of actual harm were both small children who stayed on the diet for weeks. A 16-year-old eating toast for a day and a half isn't going to develop a deficiency.

‍

Eating toast because it's what you can face is fine. Trouble starts if you treat those four foods as the treatment and stay on them after you could be eating properly.

‍

What replaced it

‍

IDSA's recommendation runs the other direction: "Resumption of an age-appropriate usual diet is recommended during or immediately after the rehydration process is completed." The World Gastroenterology Organisation puts a clock on it, saying that "the practice of withholding food for > 4 hours is inappropriate." The NIH's digestive-disease institute is the bluntest of them: "Research shows that following a restricted diet does not help treat viral gastroenteritis. Most experts do not recommend fasting or following a restricted diet when you have viral gastroenteritis." The AAP page that retired BRAT says it in five words: "Fasting is not a treatment for diarrhea."

‍

In practice that means smaller portions of whatever you normally eat, whenever you feel like eating, instead of a list of approved foods you work through in order. The AAP describes the target as "a mix of fruits, vegetables, meat, yogurt, and complex carbohydrates" within about a day of getting sick.

‍

The evidence under that advice is thinner than the confidence of the wording suggests. The Cochrane review behind it found an absence of harm rather than a benefit: across 12 trials, eating early made no difference to unscheduled IV fluids, vomiting, or persistent diarrhea, and it couldn't reach a conclusion on whether eating early shortens anything. Almost every trial in this entire field was run in children under five, most of them in hospital, and a search for diet or refeeding trials in adolescents with acute gastroenteritis comes back empty. That's still what the professional bodies say, and it was worked out on toddlers.

‍

Drinking is the harder half

‍

What puts people in an emergency room with a stomach bug is fluid, not food. The CDC's norovirus page lists decreased urination, dry mouth and throat, and feeling dizzy when you stand up, and notes that most people recover in one to three days.

‍

Oral rehydration solution (a drink mixed to a specific concentration of salt and sugar) is what every guideline names first. It's mixed that precisely because glucose and sodium get absorbed together in the small intestine, so the sugar is there to pull the salt and the water across rather than to make it taste good. WHO's formula is 75 mmol/L of sodium, 75 mmol/L of glucose, and a total osmolarity of 245 mOsm/L. Pedialyte, the version sold in American drugstores, is 250. The CDC's own table lists apple juice at 730 and Coca-Cola Classic at 650, both under a heading reading "Commonly used beverages (not appropriate for diarrhea treatment)."

‍

Then the advice splits, and it splits between two current guidelines rather than between experts and myths. IDSA's table footnote says that "popular beverages that should not be used for rehydration include apple juice, Gatorade, and commercial soft drinks." The American College of Gastroenterology's adult guideline says nearly the opposite, in a strong recommendation: "Most individuals with acute diarrhea or gastroenteritis can keep up with fluids and salt by consumption of water, juices, sports drinks, soups, and saltine crackers."

‍

Johns Hopkins threads it in a way that makes both make sense: "Keep in mind that sports drinks are high in sugar. They are not appropriate if you are extremely dehydrated. In this case, you will need an oral rehydration solution." A sports drink is a drink. It isn't a rehydration solution, and the distinction stops mattering when you're mildly ill and starts mattering when you're genuinely dehydrated.

‍

The one trial that compared them head to head in adults randomized 75 hospital patients to Gatorade, Pedialyte, or a rehydration solution for 48 hours. Stool frequency, consistency, and body weight improved in all three groups with no difference between them, and the authors concluded that "Gatorade and N-ORS seem to be as effective as Pedialyte in correcting dehydration and in improving bowel symptoms." One thing did separate: low potassium kept turning up in the Gatorade group across both days and mostly cleared in the other two. Sixty of the 75 finished, which the authors flag as a limitation.

‍

Ginger ale is the drink most people reach for, and it's mostly sugar, 33 grams in a 12 oz can. Whether the ginger in it does anything has its own article: Does Ginger Ale Help an Upset Stomach?.

‍

Where the sugar argument gets shaky

‍

The osmolarity story (sugary drink, too concentrated, pulls water into your gut, makes diarrhea worse) is the explanation everybody gives, and the human evidence for it is narrower than the confidence around it. The trials that support it compared a 245 mOsm/L solution against a 311 mOsm/L one, and found the lower one caused fewer unscheduled IV infusions. Nobody has run a trial of cola against rehydration solution in a person with a stomach bug. Going from "245 beats 311" to "730 will make you worse" is a reasonable guess rather than a finding.

‍

The study that did test a high-sugar drink went the other way. In 2016, researchers in Toronto randomized 647 children with mild gastroenteritis to half-strength apple juice followed by whatever they wanted to drink, or to a flavored electrolyte solution. The apple juice group failed treatment less often, 16.7% against 25.0%, and fewer of them needed IV fluids, 2.5% against 9.0%. Apple juice, cut with water, beat the medical drink, while sitting on CDC's own list of beverages that aren't appropriate for treating diarrhea.

‍

The children were 6 to 60 months old, two thirds had no evidence of dehydration to begin with, it ran at a single Toronto hospital, and a 2020 review found it had never been replicated, with five of the six parts of its composite outcome showing no significant difference. The authors' own conclusion is careful: dilute apple juice "may be an appropriate alternative to electrolyte maintenance fluids in children with mild gastroenteritis and minimal dehydration."

‍

If you're mildly sick and drinking normally, what you drink matters less than that you keep drinking it. If you're actually dehydrated, the mixed-to-spec drink is the one with evidence behind it.

‍

Dairy, and whether you need to skip it

‍

Skipping milk is one of the oldest pieces of advice here, and IDSA groups it with BRAT as something "commonly recommended" with "limited" supporting data. The number underneath it comes from a Cochrane review where lactose-free products shortened diarrhea by about 18 hours. Cochrane rated that low-quality evidence itself, and every participant was under five and mostly formula-fed, so what got tested was swapping an infant's entire diet rather than skipping cheese.

‍

There's a version of the dairy warning with better support, and it applies after you recover rather than during. NIDDK notes that "some people recovering from viral gastroenteritis have problems digesting lactose for up to a month or more afterwards." Nemours puts the during-illness rule plainly: "There's no need to avoid dairy unless it makes the vomiting or diarrhea worse."

‍

Yogurt and probiotics

‍

Two large trials published in 2018, covering 943 children in ten American emergency departments and 827 in six Canadian ones, found no benefit at all on any measure. Cochrane's 2020 review went further and explained why the earlier evidence had looked good: "in this topic there is publication bias, with small studies demonstrating a positive effect more likely to be published, which skews the results." KYG's fuller take on what probiotics do and don't do is in Do Probiotics Actually Work?.

‍

When to stop waiting it out

‍

The four thresholds up top are CDC's: bloody diarrhea, diarrhea past three days, a fever over 102°F, and vomiting so often you can't keep liquids down. MedlinePlus adds one that's easier to check on yourself, which is eight hours or more without urinating.

‍

Reaching for Imodium so you can sit through a test is the common move, and IDSA recommends against it for anyone under 18. The deaths in its supporting data all occurred in children under three, so that rule is more cautious than the evidence behind it. The warning that applies at any age is NIDDK's: "if you have bloody diarrhea or fever, signs of infections with bacteria or parasites, don't use over-the-counter medicines to treat diarrhea." Slowing your gut down while something bacterial is sitting in it is the scenario those warnings exist for.

‍

Anything that keeps coming back, or drags on for weeks instead of days, belongs in What's Wrong With My Gut? territory.

‍

Eating again

‍

Come back to normal food when you want it, in smaller amounts than usual, and expect dairy to be the thing most likely to disagree with you for a few weeks.

‍

None of the guidance will hand you a list of approved foods, because there isn't one. The closest thing to a rule is the one the NHS gives for going back to school: two days without vomiting or diarrhea.

‍