Before the headache

DISCUSSION

The current study was a hypothesis-generating study designed to investigate whether infant colic may be related to migraine. The data show a significantly higher prevalence of colic in children of mothers with migraine, suggesting that infantile colic may be a childhood periodic syndrome that is a migraine precursor. Infant colic (excessive crying in an otherwise healthy infant) has an incidence of 5%−19% in prospective studies.^13,16 Normal infant crying increases in the first weeks of life, peaks at 6−8 weeks of age, and tapers off by age 3 months.^14,17 Infant colic is a highly intensified version of this same temporal pattern, often with bouts of prolonged, inconsolable crying.^13,17 Colic is classically defined as at least 3 hours of crying, at least 3 days a week, for 3 weeks (Wessel criteria).¹⁸ Practically applied, the more succinct modified Wessel criteria of crying for at least 3 hours a day, at least 3 days a week, for the past week, are often used in practice and in research.^19,20

The underlying pathogenesis of colic is not well understood, despite much research.^9,13 The term “colic” implies an abdominal derivation for the infants' distress, although no clear evidence for that localization exists. Some have implicated excessive intestinal gas²¹ or a problem with the infant's milk. However, a randomized placebo-controlled trial of simethicone, a medication that speeds the transit of intestinal gas, showed no benefit over placebo.²² In addition, type of infant feeding, be it breast milk or formula, does not affect the rate of colic.²³ There is no evidence for lactose intolerance in colicky infants^21,24 and although supplementation with probiotics appeared promising in a small study,²⁵ the effect could not be reproduced in a larger study.²⁶ Cow's milk protein allergy may cause excessive crying in a subset of infants²⁷; however, these infants' symptomatology may distinguish them from those with idiopathic colic.²⁸ Markers of dietary protein hypersensitivity and intestinal damage, such as fecal α₁-antitrypsin concentrations and fecal hemoglobin, are not elevated in infants with colic.²³ Moreover, counseling parents about how best to respond to their infants' colicky crying is more effective in treating colic than eliminating cow or soy protein, and reintroducing cow or soy protein does not worsen crying.²⁹ Whereas colicky infants cry more in the late afternoon and evening hours,^14,17 young infants feed every few hours throughout the day and night, and it is challenging to identify a gastrointestinal etiology that would have this temporal pattern.

Two previous retrospective studies reported a link between childhood migraine and infant colic,^9,10 and young infants with cerebral injuries are prone to excessive crying, suggesting a potential neurologic etiology to colic.³⁰ It is important that we ultimately determine the underlying pathophysiology of colic so that these infants can be treated appropriately. Excessive infant crying is frustrating for caregivers³¹ and is a risk factor for shaken baby syndrome.^14,32,33 The peak in hospitalizations for shaken baby syndrome falls temporally right after the peak in infant crying.¹⁴

In the current study, maternal migraine was associated with a more than 2-fold increase in the prevalence of infant colic in this study. Overall, 14% of the infants in this study had colic, which is in keeping with previously reported colic prevalence estimates of 5%−19%.^13,16 Similarly, the migraine prevalence of 18% in the mothers is highly in keeping with the 18% 1-year migraine prevalence estimate among women in North America.¹¹

The method for determining maternal migraine in this study was either a positive screen on ID Migraine or maternal report of a physician diagnosis of migraine. Physician diagnosis of migraine is almost always correct, with 98% accuracy.³⁴ However, only 48% of adult migraineurs in the United States have ever received a physician diagnosis of migraine,³⁵ so relying on physician diagnosis alone is inadequate, and it is a strength of this study to have also identified maternal migraineurs via a validated survey instrument.

Previous studies on the relationship between migraine and infant colic were retrospective and therefore subject to recall bias.^9,10 This study minimized recall bias by collecting data on colic prevalence during the period of infancy when colic occurs. The magnitude of the association between colic and maternal migraine in this study is consistent with what has been described previously. In a small retrospective study by Jan and Al-Buhairi,⁹ 52% of the children with migraine had a history of colic vs 20% of control children. Because our study uses maternal history of migraine as a means for inferring possible familial migraine tendency in the infants, it is notable that in the study of Jan and Al-Buhairi,⁹ a history of infant colic in a first-degree relative was more likely in the migraineur children, and those children with a history of colic were more likely to have a family history of migraine.⁹ In the second study, a history of colic was also more likely in children with migraine than in headache-free control children (38.4% vs 26.9%) or in children with tension-type headache (38.4% vs 25.2%).¹⁰

One previous study attempted to look at a similar question prospectively.⁸ Infants were evaluated for evidence of hyperreactivity, which included parental report of frequent spontaneous crying; The Wessel criteria for infant colic were not used. Infants' responses to stimuli, as well as their soothability, were also assessed. At follow-up 10.8 years (on average) later, 52.9% of the infants in the hyperactive group had migraine compared with an 11% prevalence of migraine in the control infants (p < 0.01). Childhood periodic syndromes were also more likely in the hyperactive group (64.5% vs 12.5%, p < 0.001).⁸

Why infants with migrainous genetics may be more likely to have colic is unclear, although an increased sensitivity to stimuli is one possible explanation. Increased sensitivity to stimuli is a hallmark of migraine; migraineurs describe heightened sensitivity to sound and light during attacks⁶ and are more likely to be sensitive to motion and cold-induced headache even between attacks.^6,36,37 Transitioning from in utero life to extrauterine life brings a flood of new and intensified stimuli. It is possible that babies with migrainous genetics are the ones who are the most sensitive to this transition and express this sensitivity through excessive crying. In one study, counseling parents to decrease infant stimulation improved colic.³⁸

Although new stimuli are present at birth, infants' ability to perceive these stimuli increases rapidly in the first weeks of life, given their rapid neurodevelopment. For example, between birth and 2 months of life an infant's visual acuity more than doubles, going from an estimated 20/200 to 20/80.³⁹ This marked increase in perceptual ability could explain why colic peaks at 6−8 weeks of life rather than immediately in the neonatal period.

This study has several limitations. For one, it is possible that a phonophobic, migrainous mother may be more likely to overestimate the length of her baby's crying and to assess her infant as colicky. However, the accuracy of migrainous mothers in perceiving whether their infants had colic was no different from that of nonmigrainous mothers. In future studies, prospective crying diaries would be a more reliable means to assess infants' crying duration.

Because this study was preliminary and targeted the parents of very young infants, the survey was kept short. We did not collect data on other epidemiologic variables that could interact with migraine in the relationship with colic, such as history of mood disorders in the parents or stress within the family.^15,20 Our data on family history were limited in that to evaluate truly for a genetic association between migraine and colic a 3-generation family tree would have been optimal, however impractical in a well-baby visit setting. Future studies would also benefit from a means of comparing demographic characteristics of survey responders with those of survey nonresponders.

The main limitation of this study was that the marker of migraine genetics used for the infants was maternal history of migraine rather than development of migraine in the children themselves. Long-term follow-up studies are needed to confirm whether infants with colic are indeed more likely to go on to develop migraine and other childhood periodic syndromes, such as abdominal migraine, as they grow through childhood.

In the meantime, because we do not yet know the reason for distress in colicky babies, perhaps the name of the diagnosis should more closely mirror our limited knowledge of the organ systems involved. “Paroxysmal fussing of early infancy,” a modification of the term originally used in the 1954 article of Wessel et al.¹⁸ may be more accurate than “colic.” Assuming that the infants' distress is abdominal in localization due to the infants' movements is analogous to watching a toddler have a tantrum and interpreting his or her flailing limbs as indicative of diffuse extremity pain. Even if the localization is abdominal, the etiology may still be migrainous and analogous to abdominal migraine. If colic is indeed migrainous rather than gastrointestinal in etiology, one wonders what the number needed to treat with age-appropriate migraine therapy would be to prevent a case of shaken baby syndrome.