What helps a colicky baby: probiotics, fennel, or just time?

What colic is. Colic is full-force crying for at least three hours a day, on at least three days a week, for at least three weeks, in a healthy baby⁹. Nobody knows the cause. Fussing and crying averaged 117 to 133 minutes a day in the first six weeks and fell to about 68 minutes by 10 to 12 weeks, when colic is rare (0.6%)¹. Doctors diagnose it from the history and an examination once alarm signs are absent. Parent education and reassurance are central, and no therapy is strongly recommended².

Soothing. Holding, walking, gentle motion and white noise are common comfort measures, though the evidence is thin. In one trial, carrying colicky babies 2.2 hours more a day did not reduce their crying¹⁰.

Probiotics. L. reuteri DSM 17938 is the best studied. Pooled data from four trials (345 infants) showed less crying, with a number needed to treat of about 2.6 in breastfed babies and no clear effect in formula-fed ones⁴. A large Australian trial of 167 breastfed and formula-fed babies found no benefit, and the probiotic group cried or fussed 49 minutes more⁵. Trial doses were about 100 million colony-forming units a day⁵. See Lactobacillus reuteri.

Fennel and herbal teas. A 2003 trial found a fennel seed oil emulsion ended colic in 65% of babies versus 24% on placebo¹¹, but it has not been replicated. European regulators do not recommend fennel fruit for children under 4 years because safety data are lacking, and estragole in fennel is a genotoxic carcinogen in rodents⁶. Star anise tea caused neurologic reactions in seven infants⁷.

Simethicone and other drugs. Simethicone was no better than placebo in 83 infants⁸. A review of reviews found no benefit or possible harm from simethicone, and no support for acid-blocking drugs¹².

Diet changes. Trials of hydrolyzed formula or low-allergen maternal diets were small and low quality, so none can be recommended; ask your pediatrician before switching formula or restricting your own diet⁹.

Coping. Infant crying is an important trigger for shaken-baby injuries. Education that normalizes crying and encourages parents to walk away when frustrated improved their knowledge³.

Time: the best-documented "treatment"

Wolke's meta-analysis of 28 diary studies (8,690 infants) found no universal crying peak at 6 weeks. Mean fussing and crying was stable at 117 to 133 minutes a day through week 6, then dropped to 68 minutes by weeks 10 to 12. Colic by modified Wessel criteria affected 17% to 25% of babies in the first six weeks, 11% at 8 to 9 weeks and 0.6% at 10 to 12 weeks, and was least common in Denmark and Japan¹. Waiting is hard, but it is also the most reliable course.

Probiotics: two conflicting stories

Sung's 2014 BMJ trial randomized 167 babies under 3 months, breastfed or formula-fed, to L. reuteri DSM 17938 (100 million CFU daily) or placebo for a month. At one month the probiotic group cried or fussed 49 minutes more (95% CI 8 to 90), mainly more fussing in formula-fed babies; secondary outcomes were similar and no adverse events were related to the study⁵. The authors' 2018 individual-participant meta-analysis of four double-blind trials (174 probiotic and 171 placebo infants) reached a different view: crying fell by 25.4 minutes more than with placebo at day 21 (95% CI -47.3 to -3.5), and treatment success was 1.7 times as likely (95% CI 1.4 to 2.2). In breastfed babies the number needed to treat was 2.6 (95% CI 2.0 to 3.6), while the effect in formula-fed babies was not significant. Data on formula-fed infants were insufficient for firm conclusions⁴. Taken together, L. reuteri looks plausible for breastfed babies, unproven for formula-fed ones, and not a general solution.

Fennel and herbal remedies: encouraging, old and risky

Alexandrovich randomized 125 babies aged 2 to 12 weeks at two clinics to fennel seed oil emulsion or placebo. Colic resolved in 65% (40 of 62) versus 23.7% (14 of 59), an absolute risk reduction of 41% (95% CI 25 to 57) and a number needed to treat of 2, and no side effects were reported¹¹. It was small, over twenty years old, and tested an oil emulsion rather than tea. The Cochrane review of dietary modifications found one small study (93 infants) of a fennel, chamomile and lemon balm extract with lower crying (76.9 versus 169.9 minutes a day), but rated all evidence very low quality and could not recommend any intervention⁹. The EMA's 2024 monograph says fennel fruit is not recommended under age 4, notes estragole is a genotoxic carcinogen in rodents, and says estragole intake under age 12 must stay below 1.0 µg/kg a day⁶. Ize-Ludlow described seven infants with adverse neurologic reactions after home-made star anise tea, and found Chinese star anise adulterated with the toxic Japanese species⁷.

Simethicone, other drugs and manual therapy

Metcalf's crossover trial gave 83 infants aged 2 to 8 weeks simethicone and placebo. Symptoms improved in 54% of treatment periods, worsened in 22% and were unchanged in 24%, whatever the treatment; 28% responded only to simethicone and 37% only to placebo⁸. Ellwood's review of 32 reviews and guidelines found probiotics the strongest evidence for breastfed infants (25 to 65 fewer crying minutes a day), weaker low-quality evidence for manual therapies (33 to 76 minutes), no benefit or harm from simethicone and no support for proton pump inhibitors. Three national guidelines unanimously advised education, reassurance and clinical evaluation¹².

Diet and carrying

The Cochrane review pooled 15 small, high-risk-of-bias trials (1,121 infants). One study (90 infants) found 74% responded to a low-allergen maternal diet versus 37%, while hydrolyzed-formula results were inconsistent⁹. Barr's trial of 66 mothers of colicky babies found 56% more carrying changed nothing: infants cried only 3 minutes less a day (95% CI 37 fewer to 32 more)¹⁰, even though extra carrying reduces ordinary crying in healthy infants.

Safety and coping

In a randomized trial of 2,738 mothers, the PURPLE crying materials raised knowledge about crying, slightly raised knowledge about the dangers of shaking, and increased sharing of the advice to walk away if frustrated³.

Bottom line (confidence 3 of 5)

Colic reliably eases with time, and the only intervention with some pooled support is L. reuteri for breastfed babies, with conflicting trials. Fennel and herbal teas lack safety data, and simethicone does not work. Check for alarm signs and keep caregivers supported.