What helps most when you have diarrhea or a stomach bug?

Diarrhea and vomiting drain water and body salts (electrolytes). WHO describes oral rehydration solution (ORS) as clean water with sugar and salt, absorbed in the small intestine to replace what was lost⁴.

Which fluid? In a Cochrane review that pooled eight trials in children, a reduced-osmolarity ORS (lower in sodium and glucose than the original WHO formula) meant fewer unplanned IV fluid infusions than the older formula (odds ratio 0.59), with no added risk of low blood sodium¹. Use a commercial sachet or ready-made solution exactly as directed; US guidance warns that home-mixed salt-and-sugar solutions can contain serious errors and recommends standard commercial products where they are available². If none can be had, our home oral rehydration recipe shows how to measure a sugar-and-salt mix exactly; for a child it is only a stopgap while you get advice, and a child with signs of dehydration should be seen by a health professional^2,4. Coconut water is not a substitute: its sodium was very low and varied with the age of the coconut⁶.

For mild illness in children who are not dehydrated, preferences matter. In a Toronto trial of 647 children aged 6 to 60 months, half-strength apple juice followed by preferred fluids led to fewer treatment failures than an electrolyte solution (16.7 percent versus 25.0 percent)³. Large amounts of soft drinks and full-strength juice are still discouraged².

Food. Keep breastfeeding, return to the usual diet, and do not withhold food for more than 24 hours². Adults follow the same logic: replace water, electrolytes, and nutrients, preferably by mouth⁵. For ideas, see what to eat and drink when ill.

Probiotics. A Cochrane review of 82 studies (12,127 people) found that, in the highest-quality trials, probiotics made little or no difference to diarrhea lasting 48 hours or longer⁹. Two large North American trials in children found no benefit^10,11. A meta-analysis of Saccharomyces boulardii trials found diarrhea about a day shorter, but the evidence was of low to very low quality¹². See probiotic strains.

Zinc. WHO recommends a 10 to 14 day course of dispersible zinc tablets for children⁴. A Cochrane review of 33 trials (10,841 children) found that zinc shortened diarrhea by about half a day in children over six months, more in malnourished children, but did not help babies under six months or well-nourished children in low-risk settings, and it increased vomiting⁸. See zinc.

Warning signs. WHO advises seeing a health professional for diarrhea lasting 14 days or longer, blood in the stool, or signs of dehydration, such as sunken eyes, restlessness or irritability, drinking eagerly, or skin that stays pinched up⁴. In adults, bloody or mucoid stools, a weakened immune system, or a recent hospital stay also call for evaluation⁵.

Oral rehydration. Hahn and colleagues' Cochrane review compared reduced-osmolarity ORS with the older WHO standard formula in children with acute diarrhea¹. Unscheduled IV fluids, the primary outcome, were reported in 11 trials; a meta-analysis of eight gave an odds ratio of 0.59 (95% CI 0.45 to 0.79) with no heterogeneity. Eleven trials reported stool output, which was lower, and vomiting was less frequent in the six trials that measured it. Low blood sodium occurred in three of six trials that looked, with no obvious difference between arms. The evidence comes mostly from children admitted to hospital and compares two ORS formulas; the benefit of ORS over no treatment rests on decades of practice and guidelines, not on a placebo-controlled trial².

Juice versus electrolyte solution. Freedman's single-center, single-blind trial randomized 647 children aged 6 to 60 months with gastroenteritis and minimal dehydration (68 percent showed no dehydration at all)³. Treatment failure within seven days was 16.7 percent with half-strength apple juice then preferred fluids and 25.0 percent with an electrolyte maintenance solution (difference -8.3 percentage points); IV rehydration was needed in 2.5 versus 9.0 percent. Hospitalization and the frequency of diarrhea and vomiting did not differ. The result applies to mild illness in a high-income setting and should not be stretched to dehydrated children or young infants.

Coconut water. Fagundes Neto and colleagues measured coconut water at 5 to 12 months of maturity: sodium averaged 2.9 mEq/L through month seven and 12.5 mEq/L afterward, glucose and osmolality varied widely, and at no stage did sodium and glucose reach levels of potential value as an ORS⁶.

Probiotics. Collinson's Cochrane review included 82 studies and 12,127 participants (11,526 children and 412 adults)⁹. Risk of bias was high or unclear in many, heterogeneity was marked, and funnel plots showed publication bias. In trials at low risk of bias, diarrhea lasting 48 hours or longer had a risk ratio of 1.00 (0.91 to 1.09; two trials, 1,770 people) and duration was 8.64 hours shorter with a very wide interval (29.4 hours shorter to 12.1 hours longer; six trials, 3,058 people). The Canadian trial randomized 886 children to Lactobacillus rhamnosus R0011 plus L. helveticus R0052 or placebo for five days: moderate-to-severe gastroenteritis occurred in 26.1 versus 24.7 percent (odds ratio 1.06; 0.77 to 1.46)¹⁰. The US trial randomized 971 children to Lactobacillus rhamnosus GG or placebo: 11.8 versus 12.6 percent (relative risk 0.96; 0.68 to 1.35)¹¹. By contrast, Szajewska's S. boulardii analysis of 29 randomized trials found diarrhea 1.06 days shorter (95% CI 0.79 to 1.32; 23 trials, 3,450 children), but only 17 percent of trials concealed allocation adequately, heterogeneity was high (I² 90 percent), and quality was very low¹². Large, high-quality trials of two bacterial products are null, while many small yeast trials are positive; strain, quality, and publication bias all remain open. Safety also differs by organism: in Finnish hospitals, at least 20 of 46 patients with Saccharomyces bloodstream infection had been taking S. boulardii (odds ratio 14 against a bacteremia or candidemia control group)⁷.

Zinc. Lazzerini's review covered 33 trials in 10,841 children aged one month to five years⁸. In acute diarrhea in children over six months, zinc shortened illness by 11.46 hours (95% CI 3.19 to 19.72; low certainty) and lowered the chance of diarrhea still present on day seven (risk ratio 0.73; moderate certainty); in children with signs of malnutrition, the reduction was 26.39 hours (high certainty). In babies under six months there was no effect, and vomiting rose in both age groups (risk ratio 1.57 in older children). The authors support zinc where deficiency or malnutrition is common and not in well-nourished children in low-risk settings.

Adults. Most trials above are in children. For adults, the American Family Physician review notes that most acute diarrhea in places with adequate sanitation is uncomplicated and self-limited, that oral rehydration is preferred, and that bloody or mucoid diarrhea, immune compromise, or recent hospitalization call for further evaluation⁵. We did not find adult trials comparing ORS with other fluids, so adult advice largely extends the pediatric evidence and expert consensus.

Bottom line. Confidence 4 of 5. Replacing fluid and electrolytes and continuing to feed are well established. Juice for mild illness is supported by one good trial, probiotics are contested, and zinc helps in specific settings.