Emergence of a Developmental Explanation for Prolonged Crying in 1‐ to 4‐Month‐Old Infants

Abstract

Surveys in several countries report that approximately one-fifth of infants have prolonged, unexplained periods of crying in the first 4 months of age (1–6). The prolonged crying usually stops and infants develop normally (7–9). The crying worries many parents, however, leading them to seek professional help and incur substantial health service costs (10,11). The crying also can trigger other adversities, including parental exhaustion and depression (12–15), premature cessation of breast-feeding because parents attribute the crying to hunger (16), excessive infant weight gain because parents overfeed to manage the crying (17,18), and, in a small number of cases, shaken baby syndrome or other child abuse (19–24). A better understanding of infant crying can help parents, improve health service cost-effectiveness, and reduce adverse outcomes. As documented in the introduction to this supplement, these unexplained bouts of crying in young infants traditionally have been attributed to gastrointestinal (GI) disturbance and pain, leading to the clinical designation infant colic. We review the evidence that has challenged this GI explanation of the causes of the crying and examine the emergence of an alternate, developmental explanation of the crying and its implications. CHALLENGES TO THE GI EXPLANATION OF PROLONGED CRYING IN 1- TO 4-MONTH-OLD INFANTS Most Infants With Prolonged Unexplained Crying Lack Organic Disturbances Although organic disturbances can cause infants to cry, such cases are rare. For example, in a study of all of the infants who were presented to a Canadian paediatric hospital for crying or irritability during 1 year, only 12 of 237 (5.1%) were found to have a serious underlying organic aetiology (25). The critical issue, then, is not whether organic disturbances can cause infant crying, but how often they do so in practice and the types of organic factors. Three reviews of the evidence have concluded that no more than 5% to 10% of infants taken to clinicians because of prolonged unexplained crying in early infancy have organic disturbances (3,26,27). Because the crying occurs mostly in the early postnatal weeks, one possibility is that adversities during pregnancy or birth could make some infants irritable. Maternal stress and cigarette smoking during pregnancy in particular have been examined. Some studies have found that these are correlated with infant crying (28–31), but others have not (29,32–35), highlighting the difficulties involved in disentangling causation in such research (36). Whether these or other pre- or perinatal adversities cause crying remains uncertain, but it is noteworthy that most infants with prolonged crying have mothers who do not smoke, suggesting that this is not a common cause (37). Another, potentially more common, organic explanation for infant crying, corneal abrasions caused by infants scratching their eyes, was examined (38). The authors found abrasions in 49% of 1- to 12-week-old infants, but no differences in the amount of crying between those with and without abrasions. The most likely organic cause for prolonged crying is food intolerance, as described in other articles in this supplement and elsewhere (39,40). One implication is that diagnostic protocols are needed to identify this small but important group of infants. For example, Moravej et al found a positive cow's-milk skin test result in only 3 of 114 (2.6%) infants with prolonged crying (41). All 3 responded positively to the elimination of cow's milk from the mother's diet, whereas a randomised controlled trial found that the elimination of cow's milk did not reduce crying in the infants with a negative skin test. It should be emphasised that this study's method is controversial [see (42) and Heine's articles in this supplement]. Their findings need replication in larger samples. However, they are valuable in that they point to the need for studies designed to recognise diverse etiological pathways and infant groups. The second implication of these findings is that 90% to 95% of 1- to 4-month-old infants who cry excessively do not have an organic disturbance, so that protocols need to be developed to target and help this much larger group. Lack of Evidence That the Crying Sounds Abnormal or Signals Pain The GI explanation implies that prolonged crying in young infants is caused by digestive dysfunction and associated abdominal pain (43). To support this implication, an early study found that the cries of 16 infants selected for their prolonged crying had a higher fundamental frequency (audible pitch) and were more dysphonated (harsh sounding) than the cries of control infants (44); however, the group of prolonged criers included prematurely born infants, who tend to have high-pitched cries (45,46). The same research group was unable to replicate these findings in a group of full-term prolonged criers, although in this case their cries’ “mean resonant frequency” was higher (47). A third study (48) also found no difference in fundamental frequency and dysphonation between the cries of prolonged criers and control infants before or after feedings. When the “most vehement” cry utterance (lasting about 1 second) was selected, however, the most vehement utterances of prolonged criers did not decline in fundamental frequency and dysphonation after a feeding, whereas control infants’ cries did decline in these features. A fourth study recorded the cries of prolonged criers continuously throughout 24 hours, comparing them with the cries of controls and carrying out separate analyses of those cries judged by parents to be “colic cries” (49,50). Most of the crying of both groups of infants was intermittent fussing rather than continuous crying. The crying bouts of the prolonged criers were objectively longer and more intense sounding (a higher cry:fuss ratio), but not more sudden in onset, than those of controls. The parent-judged colic crying bouts were not higher in fundamental frequency or dysphonation than the hunger cries of other babies. Taken together, these 4 studies provide little evidence that prolonged criers have cries that are distinct in 2 primary acoustic properties: fundamental frequency or dysphonation. Neither is there a consensus on which properties are distinct. Acoustic studies generate multiple indices, some of which are likely to differ between groups by chance (51). Whether infant cries’ “mean resonant frequency” and “most vehement utterance” are signs of organic impairment and whether they are the source of parents’ anxiety remain unknown. A related question is whether the sound of infants’ cries provides reliable information that they are in pain. Here, too, some early studies appeared to support this claim (52), although others did not (53). Methodologically robust studies, typically involving inoculations, have found that variability within and between infants’ responses to pain is substantial and strongly influenced by infant age and environmental factors, such as whether they are held or have tastes in their mouths (27,54–58). That is not to deny the existence of an archetypal pain cry. Rather, the point is that it is far easier to identify it as a pain cry when the needle is seen going into an infant's foot than if all that is available is the sound of the cry (59). A careful review of this research as a whole concluded that infant cries are “graded signals” that convey an infant's degree of distress but not the precise cause of the cry (60). From an evolutionary standpoint, the cry needs only to provide a “biological siren” that tells a parent whether to come quickly (61); it does not need to identify the precise cause because, once nearby, parents can use contextual information to determine their action (61). More needs to be learned about the way in which somatic, neurological, and vocal system disorders influence infants’ cry sounds, how the sounds map onto infants’ psychological states, and how the communicative functions of cries change with age. Studies have begun to document the complexities involved (62–64). A conservative conclusion for our purposes here is that there is little evidence at present that 1- to 4-month-old infants identified by prolonged crying have distinct or abnormal sounding cries or that they are in pain. Inadequacies of the Rule of Threes and Rome III Criteria for Selecting Clinical Cases As the introduction to this supplement notes, these 2 sets of criteria have been put forward by expert clinical groups to allow infants with “colic crying” to be selected in a uniform way. The 2 approaches are similar in emphasising the amount of crying as a defining feature, with minor differences. The Rule of Threes for a fussy/colicky infant requires infants to have paroxysms of fussing, crying, or irritability lasting a total of >3 hours/day for ≥4 days/week, whereas in extremely fussy/colicky infants the paroxysms continue for >3 weeks or become so severe that the paediatrician determines that medication is indicated (65). The Rome III criteria require crying lasting only 3 hours for 3 days/week (43). The 2 schemes share the requirement that the crying should have a sudden and unexplained onset, termed paroxysmal (65). The Rule of Threes requires evidence that an infant is “otherwise healthy and well fed,” whereas the Rome III criteria stipulate that there is no evidence of failure to thrive. An implication is that infants who fail to gain weight normally because of feeding or other problems should not be diagnosed as having colic. In practice, few studies have examined the onset of infants’ cry periods, although 1 that did so (noted above) found no evidence that infants with prolonged crying or parent-judged colic crying bouts had an abnormally sudden onset to their cries (49,50). Instead, most studies have focused on the amount of crying as the defining feature of colic cases (66). As discussed in detailed reviews (37,67,68), the requirement that an infant fusses/cries for ≥3 hours/day raises considerable practical difficulties. The first is the need to measure crying for at least 1 week, so that the 3- or 4-day requirement can be met. Because objective methods are costly, almost all studies have used parental measures, and because few parents will keep 24-hour cry diaries for 7 consecutive days, have selected infants who cried >3 hours on average in the days measured or who parents recall to have cried 3 hours on most days in the previous week. Recall methods, in particular, inflate the amounts of crying reported so that approximately 50% of infants claimed to cry ≥3 hours/24 hours have been found not to do so when measured directly (2,69–73). It seems likely that parental vulnerabilities influence their subjective reports of how much infants cry (73). We do not know how many infants would meet the Rome or Wessel criteria if accurately assessed, but the result in practice is that wide variability exists within and between studies in how much the participating infants cry. As Barr pointed out, a second obstacle to the use of the Wessel and Rome III criteria is that the rules are arbitrary (67). There is no evidence that an infant who cries for 3 hours/day has a more serious problem or differs in any important way from one who cries for 2 hours 50 minutes/day. The third obstacle is that amounts of infant crying vary substantially between countries. It follows that a single rule based on amount of crying will automatically identify more clinically disturbed infants in some countries than in others. Instead, criteria for selecting atypical cases need to be norm referenced within each country so that infants who, for instance, cry 2 standard deviations more than their peers are selected for clinical attention (68). This diagnostic requirement is not specific to crying—for instance, it applies to measures of overweight—but it is not met by the Rome III or Wessel criteria. The last obstacle to applying these rules is that they assume that crying amount is the chief defining feature of clinical cases. It is doubtful whether an amount of crying, by itself, would worry parents if they are able to stop it. Instead, studies have found that what distresses parents is that crying in 1- to 4-month-old infants is sometimes “unsoothable” and, therefore, uncontrollable by parents (66,74,75). Fujiwara et al (76) found in both Canadian and US studies that “unsoothable crying,” and particularly the maximum bout length of unsoothable crying, was much more strongly associated with caregiver frustration than the amount of daily crying. To be effective, diagnostic schemes need to be supported by evidence of external validity and be reliable and practical to the clinicians who routinely use them (77). The Rule of Threes and Rome III criteria do not meet these requirements. EMERGENCE OF A DEVELOPMENTAL EXPLANATION FOR PROLONGED CRYING IN 1- TO 4-MONTH-OLD INFANTS Both the Rule of Threes and Rome III criteria assume the existence of a distinct group of infants with a clinical condition. The developmental viewpoint is that most infants with prolonged crying are healthy and the crying is a result of normal growth and developmental processes. It follows that most cases presented to clinicians will share the features of normal infants and differ from them by degree, so that normal variations in development or temperament account for differences in crying. This conceptual approach has distinct implications for clinical practice, which are considered after reviewing the following supporting evidence. Developmental and Time-of-Day Peaks in 1- to 4-Month-Old Infants’ Crying Studies that have progressed beyond clinically referred groups to assess community samples have found that babies in general reach a peak in their crying, together with a pronounced evening clustering, at approximately 2 months old (78–81). The peaks also have been found in non-Western societies (82) and in healthy prematurely born infants at a corrected age of approximately 6 weeks (83), prompting the proposal that this crying pattern is a “behavioural universal of infancy” (79). It is noteworthy that some studies have found a crying plateau in the early weeks rather than a peak, but even in these cases, infants’ amount of crying declined substantially, by approximately 40%, by 3 months of age (78,81,84). Unsoothable Bouts of Crying Along with crying peaks, studies have found that bouts of inconsolable or unsoothable crying are a feature of early infancy, virtually disappearing by 5 months of age (50,66,69,74,75). The hard-to-soothe nature of these bouts is an objective infant characteristic. Trained researchers using standard consoling procedures, as well as parents, have been unable to stop the crying (75). Like periods of unexplained crying, the unsoothable bouts are found in infants generally at this age, but they are more common and last longer in infants with prolonged crying (69,74). As noted above, these unsoothable crying bouts seem to be the main trigger for parents’ distress because they feel frustrated, helpless, or guilty about their inability to stop the crying (35,66). Evidence of Neurodevelopmental Changes in Infants’ Ability to Regulate Their Crying Unlike GI explanations, in which the focus is what causes crying to start, the focus for developmental explanations is on how infants regulate (control) crying and on changes in regulatory abilities with age. Early studies using a rubber-band-snap apparatus to deliver a standard painful stimulus to the sole of infants’ right feet provided direct evidence for this developmental viewpoint (85). The crying response was brief in the first 2 days after birth, increased and remained prolonged from 2 days to 12 weeks of age, then showed a gradual decrease in the likelihood and intensity of response up to 1 year. Sweet tastes, which are highly effective in preventing crying in response to inoculation pain in newborns, become less successful as infants age (86–89). Many studies have documented the growing control over crying behaviour that develops, so that infants grimace or struggle instead of crying, or stop crying just before a caregiver intervenes, with increasing age (90–93). Specifically, the developmental explanation for the peak in the amount of early crying and unsoothable cry bouts attributes these to the maturational reorganisation of brain systems that normally occurs at approximately 2 months of age, as reflex mechanisms are replaced by systems involving control of behaviour by the cerebral cortex (79,94,95). In particular, the long and unsoothable nature of the bouts has been attributed to a temporary deficit in neurological control of behaviour during this transition, so that infants respond quickly and strongly, or cannot stop crying once crying has started (96,97). Limitations of the Developmental Explanation The developmental viewpoint provides a plausible explanation for the long and unsoothable nature of crying bouts in early infancy, the puzzle of why so many apparently healthy babies should cry without a reason, and the spontaneous resolution of this type of crying with age; however, it has several shortcomings. The first shortcoming is the dearth of direct empirical support. If it is correct, infants of this age should have longer, rather than more frequent, crying bouts. This is what Barr et al found, but they and others have found prolonged criers to have more frequent, as well as, longer crying bouts (50,74). Accurate separation of crying bouts may require more precise measurements than those used thus far, but this remains unproven. The causal relation between neurological changes and regulation of crying (ie, the mechanism involved) has not been delineated. A further question is how this explanation accounts for one of the most striking features of the crying of 1- to 4-month-old infants: its evening clustering. It is also possible that neurodevelopmental processes account for the nature and age of prolonged crying in babies, but other factors in the infant or environment provide the explanation for the onset of crying periods. IMPLICATIONS FOR RESEARCH AND CLINICAL PRACTICE Basic research is needed to test the predictions derived from the developmental explanation, document the mechanisms involved, and explain why outcomes are poor in some cases. There is growing evidence that prolonged crying after 4 months of age, although rare, has much poorer outcomes and the reasons for this need to be established (98–102). Although the details of the developmental explanation need further testing, the key evidence underlying it, documented in 3 separate reviews, is that most infants with prolonged crying are in good health and gain weight and develop normally (3,8,27). Only 5% to 10% of prolonged criers in this age group, or approximately 1% of infants overall, have an organic disturbance. It follows that health services for prolonged infant crying need to adopt a differentiated 3-level approach, as follows: Level 1: Providing parents in general with information and guidance that help them to anticipate, manage, and contain the crying and prevent adverse longer-term outcomes. Level 2: More detailed assessment and targeting of support for cases involving parental vulnerabilities such as depression, social isolation, and lack of support. Level 3: Screening and diagnostic assessments to identify the small number of cases with organic disturbances, together with specialist referral procedures to treat the organic problems. for this including for organic cases have been put forward research is needed to assess whether health services based on these protocols are The may not recall that the for prolonged colic crying in young infants the mother's or infant's or the infant “colic We also need to that other common are to the infant or which are and may be or to or which lack evidence of that prolonged infant crying parents and that parental vulnerabilities are involved in adverse this the problems as social and rather than an infant condition. To be effective, health services for infant crying need to for and social