About the Rome IV Criteria for Infant Colic
Three clinical criteria, and none of them requires timing the crying. The infant is under five months of age when symptoms start and stop. There are recurrent and prolonged periods of crying, fussing or irritability, reported by caregivers, that occur without obvious cause and cannot be prevented or resolved by caregivers. And there is no evidence of failure to thrive, fever or illness. The old 'rule of threes' — three hours a day, three days a week, three weeks — survives only in the research definition, which additionally requires three or more hours of crying and fussing on three or more days in seven, confirmed by a prospectively kept 24-hour behaviour diary.
Formula
Infant colic (clinical) = onset and resolution before 5 months of age
AND recurrent prolonged crying/fussing/irritability without obvious cause,
unpreventable and unresolvable by caregivers
AND no failure to thrive, fever or illness
Research definition adds:
≥ 3 hours crying + fussing per day on ≥ 3 days in 7 (screening interview)
AND ≥ 3 hours confirmed on a prospective 24-hour behaviour diary- Fussing
- Rome IV defines it as behaviour that is not quite crying but not awake and content either. Infants fluctuate between the two, which is why the criteria never ask anyone to distinguish them.
- Without obvious cause
- Hunger, a soiled nappy, temperature and pain all have to have been considered and excluded. The criterion assumes an examination has taken place.
- Cannot be prevented or resolved by caregivers
- The soothing-resistant quality is characteristic and is often what distresses families most — the sense that nothing they do works is itself part of the syndrome.
- Rome IV deliberately removed the crying-duration threshold from the clinical criteria; requiring a family to time and document crying before offering help served no clinical purpose.
- The age rule covers both onset and offset, which is unusual among Rome IV criteria.
- Crying in normal infants follows a curve that peaks at around six weeks and declines thereafter — colic follows the same curve, shifted upwards.
- The research definition requires a prospective diary, not a retrospective estimate, because caregiver recall of crying duration is unreliable.
Interpreting the result
Meeting criteria should shift the consultation from the infant to the family. Explain the crying curve explicitly, including the peak and the expected decline, because families who understand that the crying is at its worst now and will improve cope substantially better than those who do not. Ask directly about how the caregivers are managing, screen for maternal depression, and say plainly that it is safe to put the baby down somewhere safe and step away when overwhelmed — this specific advice is a documented protective factor against abusive head trauma and is frequently omitted. On interventions: a probiotic, Lactobacillus reuteri DSM 17938, has meta-analytic evidence of benefit, though the effect was clearest in exclusively breastfed infants and less consistent in formula-fed ones. Cow's milk protein exclusion helps a minority and should be trialled with a planned rechallenge rather than adopted indefinitely. Simethicone, anticholinergics, acid suppression, lactase drops and manipulative therapies do not have evidence supporting routine use. Where criteria are not met, the failing criterion directs the assessment: fever or faltering growth means the baby needs a diagnosis, and crying beyond five months warrants a fresh look rather than an extension of the label.
| Score | Band | What it means | Action |
|---|---|---|---|
| Criteria met | Infant colic | Recurrent unexplained crying in a well infant under 5 months of age | Support the caregivers, explain the crying curve, screen for depression, give explicit safe-coping advice |
| Criteria met and research threshold satisfied | Infant colic — research criteria satisfied | Three or more hours of crying and fussing on three or more days in seven, diary-confirmed | No difference in management; relevant only to trial eligibility |
| Criteria not met — illness or faltering growth | Not colic | Fever, failure to thrive or evidence of illness | This needs a diagnosis — consider infection, cow's milk protein allergy, obstruction and non-accidental injury |
| Criteria not met — outside the age window | Outside the definition | Crying starting or persisting beyond 5 months of age | Reassess rather than extending the label; the normal crying curve has fallen away by this point |
What the Infant Colic needs (4 inputs)
- Infant under 5 months of age when symptoms start and stop
- Rome IV frames the age rule around both onset and resolution, which is unusual. Crying that begins or persists after five months is not infant colic and needs a different explanation.
- Recurrent and prolonged crying, fussing or irritability without obvious cause, that caregivers cannot prevent or resolve
- 'Fussing' is defined in Rome IV as distressed vocalisation that is neither crying nor contented settling. The two blur into one another in practice, which is why the criteria treat them together rather than trying to separate them.
- No evidence of failure to thrive, fever or illness
- A thriving, afebrile, well baby. This is the criterion that turns a label into a safe one, and it requires an examination rather than a history alone.
- Research only — 3 or more hours of crying/fussing on 3 or more days in 7, confirmed by a 24-hour diary
- The old rule of threes, retained by Rome IV only for research. It requires both a screening interview and a prospectively kept behaviour diary, and it is not needed for the clinical diagnosis.
What it returns
- Criteria met or not met
- All three clinical criteria are required.
- Whether the research threshold is also satisfied
- Reported separately. It has no bearing on clinical management and exists so that trial populations are comparable.
How it is calculated
Normal infant crying is not flat. It rises from birth to a peak at around five to six weeks and then declines steadily through the following months, and this curve is remarkably consistent across cultures and caregiving styles. Colic sits on that curve rather than beside it: the same shape, at a greater amplitude. That single observation explains a great deal — why colic resolves without treatment, why it resolves at roughly the same age regardless of what was tried, and why almost any intervention started at the peak appears to work. It also explains the age criterion. Rome IV bounds the diagnosis at five months because that is where the normal crying curve has fallen away, and crying persisting past it is no longer explicable as an exaggerated version of a normal developmental pattern. Attempts to identify a gastrointestinal mechanism — gas, immaturity of the gut microbiome, cow's milk protein intolerance, altered motility — have produced suggestive findings but nothing that accounts for the majority of cases, which is why Rome IV describes the behaviour and declines to explain it.
Facts & figures
| Intervention | Evidence | Practical position |
|---|---|---|
| Caregiver support and explanation of the crying curve | Consistently valuable; addresses the actual risk | The core of management, and the part most often rushed |
| Lactobacillus reuteri DSM 17938 | Meta-analysis shows benefit, clearest in exclusively breastfed infants | Reasonable to offer, particularly in breastfed infants |
| Cow's milk protein exclusion | Helps a minority | Time-limited trial with a planned rechallenge, never open-ended |
| Simethicone | No consistent benefit over placebo | Harmless but not useful; do not present it as treatment |
| Acid suppression | No benefit for crying; infection risk | Avoid — crying is not evidence of reflux disease |
| Anticholinergics | Effect reported but serious adverse events | Not used in infants |
| Manipulative therapies | No reliable evidence | Not recommended |
Most interventions started at the crying peak appear to work, because the crying was about to decline anyway. That is why placebo-controlled data matter more here than clinical impression.
Evidence
Derivation — Rome Foundation, neonate/toddler committee
2016Consensus criteria from the Rome IV committee on childhood functional gastrointestinal disorders in the neonate and toddler, published in Gastroenterology in 2016.
Consensus-derived. The substantive change from earlier definitions was to move the Wessel 'rule of threes' out of the clinical criteria into a separate research definition, on the grounds that requiring a crying diary before offering help was clinically pointless.
Meta-analysis — Lactobacillus reuteri
2018Individual participant data meta-analysis of randomised trials of Lactobacillus reuteri DSM 17938 in infants with colic, reported by Sung and colleagues in 2018.
Treated infants were more likely to respond, with the effect clearest in exclusively breastfed infants and less consistent in those receiving formula.
How it compares
Infant Colic vs Infant regurgitation
Two benign conditions of the same age group that frequently coexist — and attributing the crying to the regurgitation is how infants end up on acid suppression.
Both peak in the first months, both present with a well baby and a distressed family, and many infants have both. The temptation is to treat one as the cause of the other, which produces a prescription for acid suppression that randomised trials show does not reduce crying. An infant who regurgitates and cries usually has two common benign conditions at once. Treating them as one condition creates an intervention where none is indicated.
Infant Colic vs Infant dyschezia
Crying with a clear pattern — ten minutes of straining before a soft stool — is dyschezia, not colic, and the distinction changes what you tell the family.
Dyschezia describes an infant who screams and strains for at least ten minutes before passing a soft stool, because the coordination between pushing and pelvic floor relaxation has not yet developed. The crying is patterned and predictable, which colic is not. The distinction matters because dyschezia has a specific piece of advice attached — avoid rectal stimulation, which teaches the infant to wait for an external trigger — that does not apply to colic at all.
Infant Colic vs Cow's milk protein allergy
The commonest organic explanation to consider, and the one most likely to be diagnosed without ever being confirmed.
Cow's milk protein allergy can present with crying, and exclusion helps a minority of colicky infants. The problem is procedural rather than diagnostic: exclusion trials are started widely and rechallenges are done rarely, so infants and breastfeeding mothers end up on restricted diets indefinitely on the strength of a trial that nobody closed. If a trial is worth starting, the rechallenge date is worth writing down at the same time. Supporting features — eczema, blood or mucus in the stool, faltering growth — make the diagnosis more likely and the trial more worthwhile.
Pearls & pitfalls
- The consultation is about the family, not the abdomen. Unexplained infant crying is the most consistently identified trigger for abusive head trauma.
- Give explicit safe-coping advice — that it is safe to put the baby down somewhere safe and step away. This is protective and is routinely omitted.
- Explain the crying curve: it peaks around six weeks and declines. Families who know this cope far better.
- Screen the mother for depression. Colic is a recognised risk factor and nobody asks.
- Rome IV does not require timing the crying — the rule of threes is now research-only.
- Fever or faltering growth means this is not colic. Examine the baby properly before applying the label.
- Crying starting or persisting beyond five months falls outside the criteria and needs reassessment.
- Almost any intervention started at the crying peak appears to work, because the crying was about to fall anyway.
- Lactobacillus reuteri DSM 17938 has meta-analytic support, clearest in exclusively breastfed infants.
- Do not prescribe acid suppression for crying. Crying is not evidence of reflux disease and the trials are negative.
Critical actions
- Examine the infant fully, including temperature, growth measurements and a check for occult injury, torsion and hair tourniquets.
- Ask how the caregivers are coping, and screen for maternal depression.
- State explicitly that putting the baby down safely and stepping away is an acceptable response to being overwhelmed.
- Explain the normal crying curve and the expected timeline of resolution.
- Review feeding — volume, technique and, where relevant, maternal diet.
- Consider Lactobacillus reuteri DSM 17938, particularly in exclusively breastfed infants.
- If trialling cow's milk protein exclusion, set the duration and the rechallenge date at the start.
- Avoid acid suppression, anticholinergics and manipulative therapies.
- Arrange follow-up rather than a one-off reassurance, and use it to recheck growth and caregiver wellbeing.
Why this score exists
Retiring the rule of threes was a small change with a revealing rationale. Wessel's criteria had defined colic for half a century, and their arithmetic — three hours, three days, three weeks — was easy to remember and easy to teach. The problem was what applying them required: a family already at the end of their resources being asked to keep a crying diary before anyone would take them seriously. Rome IV split the definition instead, keeping the arithmetic where it is genuinely needed, in trials that must recruit comparable populations, and removing it from the clinic, where it functioned only as a barrier. That is the same design decision the committee made for adult faecal incontinence, and for the same reason: a threshold that improves research comparability can simultaneously make clinical care worse, and the honest solution is two definitions rather than a compromise between them.
About the creator
First author, Rome IV neonate/toddler functional gastrointestinal disorders committee
Chaired the committee that produced the Rome IV criteria for this age band.
Co-author; infant crying research
His work on the normal infant crying curve underpins the age boundary in these criteria.
Co-author; paediatric pain
Contributed to the neonate and toddler chapter of Rome IV.
Limitations
- Consensus criteria with no external validation, and the five-month boundary is a convention drawn from the normal crying curve rather than a measured cut-off.
- Entirely dependent on caregiver report, and caregiver estimates of crying duration are known to be unreliable.
- 'Recurrent and prolonged' is unquantified in the clinical criteria, which is deliberate but leaves the threshold to individual judgement.
- No severity grading, so an infant crying for one hour and one crying for six receive the same label.
- The criteria describe the behaviour without offering any mechanism, which limits what they can say about treatment.
- Cow's milk protein allergy is not named despite being the commonest organic alternative.
- Nothing in the criteria addresses caregiver distress, which is the main clinical risk and the reason for most consultations.
- The separate research definition means published trial populations are not the same as the clinic population.
If you are the patient
Colic means a baby who cries a great deal, for long stretches, without any obvious reason, and who cannot be settled however hard you try. It is common, it is not caused by anything you have done, and it is not a sign that your baby is unwell — your doctor will check for that. Something worth knowing is that all babies cry more in the early weeks than later. Crying rises from birth, peaks at around six weeks, and then falls away steadily. Babies with colic follow exactly the same pattern, just louder and longer. That is why colic settles on its own, usually well before five months, whatever anyone does. The most important thing is not a treatment — it is that you are supported. Relentless crying is exhausting and demoralising, and it is completely normal to feel overwhelmed by it. If you reach that point, it is safe and sensible to put your baby down somewhere safe, walk into another room, and take a few minutes. That is not neglect; it is the right thing to do, and every doctor will tell you the same. If you are struggling more than that, please say so — support is available and low mood after a birth is common and treatable. As for treatments, a specific probiotic called Lactobacillus reuteri has reasonable evidence, particularly for breastfed babies, and is worth asking about. Some babies improve if cow's milk is removed from the diet, but that should be a proper trial with a plan to reintroduce it, not a permanent change. Reflux medicines have been tested and do not help with crying, so they are not recommended.
Frequently asked questions
What are the Rome IV criteria for infant colic?#
An infant under five months of age when symptoms start and stop; recurrent and prolonged crying, fussing or irritability without obvious cause that caregivers cannot prevent or resolve; and no evidence of failure to thrive, fever or illness. All three are required, and none of them involves timing the crying.
What happened to the rule of threes?#
Rome IV moved it out of the clinical criteria and into a separate research definition. The research version requires three or more hours of crying and fussing on three or more days in seven, confirmed by a prospectively kept 24-hour diary. The committee's view was that asking an already exhausted family to document crying before anyone would help them served no clinical purpose.
When does infant colic resolve?#
Almost always by five months, and usually earlier. Normal infant crying peaks at around six weeks and declines steadily thereafter; colic follows the same curve at a higher amplitude. That is also why interventions started at the peak so often appear to work.
Does any treatment actually work?#
Lactobacillus reuteri DSM 17938 has meta-analytic evidence of benefit, clearest in exclusively breastfed infants. Cow's milk protein exclusion helps a minority and should be a time-limited trial with a planned rechallenge. Simethicone, acid suppression, anticholinergics and manipulative therapies do not have evidence supporting routine use.
Why does the guidance emphasise caregiver support so heavily?#
Because that is where the modifiable risk is. Prolonged unexplained infant crying is the most consistently identified trigger for abusive head trauma, and it is a documented risk factor for maternal depression and early cessation of breastfeeding. Explicitly telling caregivers that it is safe to put the baby down and step away is a protective intervention, and it is routinely left unsaid.
Should a crying baby be given reflux medication?#
No. Crying is not evidence of reflux disease, and randomised trials of acid suppression in crying infants have not shown benefit. Acid suppression in infancy is associated with higher rates of gastrointestinal and lower respiratory tract infection, so there is cost without benefit.
What if the crying starts after five months?#
Then the criteria are not met, and the label should not be stretched. By that age the normal crying curve has fallen away, so persistent or new crying is no longer explicable as an exaggerated developmental pattern and deserves a fresh assessment.
What should be excluded before diagnosing colic?#
Rome IV requires the absence of failure to thrive, fever and illness, which means an examination rather than a history alone. Practically, that includes checking growth, temperature, and looking for occult causes of pain — corneal abrasion, hair tourniquets, testicular torsion, occult fracture and non-accidental injury among them.