Abstract
Background. Infants' healthy growth and development are predicated, in part, on regular functioning of the gastrointestinal (GI) tract. In the first 6 months of life, infants typically double their birth weights. During this period of intense growth, the GI tract needs to be highly active and to function optimally. Identifying modifiable causes of GI tract dysregulation is important for understanding the pathophysiologic processes of such dysregulation, for identifying effective and efficient interventions, and for developing early prevention and health promotion strategies. One such modifiable cause seems to be maternal smoking, both during and after pregnancy.
Purpose. This article brings together information that strongly suggests that infants' exposure to tobacco smoke is linked to elevated blood motilin levels, which in turn are linked to an increased risk of GI dysregulation, including colic and acid reflux. We base this hypothesis on evidence supporting a link between maternal smoking and infantile colic (IC) and on additional evidence proposing increased motilin release, attributable to exposure to tobacco smoke and its metabolites, as a physiologic mechanism linking maternal smoking with infantile GI dysregulation.
Methods. We critically review and synthesize epidemiologic, physiologic, and biological evidence pertaining to smoking and colic, smoking and motilin levels, and motilin and colic.
Results. Six studies have investigated the link between maternal smoking and IC, but IC was defined according to Wessel's rule of threes (crying for ≥3 hours per day, ≥3 days per week, for ≥3 weeks) in only 1 of these studies. The remaining studies used definitions that ranged from less-stringent variations of Wessel's criteria to definitions that would suggest excessive crying but not necessarily colic. Results from 5 of these studies suggest that there is an independent association between maternal smoking and excessive crying, as well as IC. Recent studies of the GI system provide strong, but indirect, corroborating evidence suggesting physiologic pathways through which maternal smoking can be linked to IC. This physiologic evidence can be outlined as follows: (1) smoking is linked to increased plasma and intestinal motilin levels and (2) higher-than-average levels of motilin are linked to elevated risks of IC. Although these findings from disparate fields suggest a physiologic mechanism linking maternal smoking with IC, the entire chain of events has not been examined in a single cohort. A prospective study, begun in pregnancy and continuing through the first 4 months of life, could provide definitive evidence linking these disparate lines of research. Key points for such a study are considered.
Conclusions. New epidemiologic evidence suggests that exposure to cigarette smoke and its metabolites may be linked to IC. Moreover, studies of the GI system provide corroborating evidence that suggests that (1) smoking is linked to increased plasma and intestinal motilin levels and (2) higher-than-average intestinal motilin levels are linked to elevated risks of IC. In the United States, nearly one-half of all women smokers continue to smoke during their pregnancies. This amounts to ∼12% of all women who give birth. Moreover, it is estimated that 50% to 80% of employed adults have regular exposure to environmental tobacco smoke and that >30% of nonsmokers live with smokers. If, as we suspect, exposure to cigarette smoke increases the risk of colic, then this would provide additional incentives to parents to abstain from smoking. Decreased exposure to tobacco smoke can be expected to provide widespread and long-term health benefits to maternal and child populations.
Infants' healthy growth and development are predicated, in part, on regular functioning of the gastrointestinal (GI) tract. In the first 6 months of life, infants typically double their birth weights.1 During this period of intense growth, the GI tract needs to be highly active and to function optimally. Identifying modifiable precursors of GI tract dysregulation is important for understanding the pathophysiologic processes of such dysregulation, for identifying effective and efficient interventions, and, most importantly, for developing early prevention and health promotion strategies.2 One such modifiable precursor seems to be maternal smoking, both during and after pregnancy.
Our aims for this article are twofold. First, we present evidence supporting a link between maternal smoking and infantile GI tract dysregulation. Second, we present evidence to propose motilin release as a physiologic mechanism linking maternal smoking and infantile GI dysregulation (Fig 1). There is evidence linking exposure to tobacco smoke with a number of GI dysfunctions, including gastroesophageal reflux, esophagitis, and infantile colic (IC).3–8 This evidence is most well developed for IC; therefore, we focus on the link between maternal smoking and IC and on the evidence suggesting physiologic plausibility of the proposed link between maternal smoking and IC.
Proposed mechanisms for effect of maternal smoking on infant risk for colic.
MATERNAL SMOKING AND IC
IC
IC is a syndrome characterized by paroxysms of irritability, inconsolable crying, and screaming accompanied by clenched fists, drawn-up legs, and a red face.9 The most commonly accepted, but not necessarily most commonly used, diagnostic criteria are based on the “rule of threes” proposed by Wessel,10 which requires the syndrome to persist for (1) ≥3 hours per day, (2) ≥3 days per week, and (3) ≥3 weeks. IC presents a few weeks after birth and peaks at 5 to 8 weeks of age; it usually resolves spontaneously by 4 months of age. Among newborns in Western countries, the prevalence of colic is estimated to be between 5% and 28%.11
The available evidence suggests that IC has multiple independent causes. IC has been attributed variously to infants' difficulty self-regulating, type of feeding, exposure to cow's milk proteins (especially α-lactalbumin) in formula or breast milk, exposure to tobacco smoke, and maternal depression or anxiety.9 In cases of maternal depression or anxiety,12 any association may be attributable to the effect of the colicky infant on the mother, rather than the converse. In his classic study, Illingworth13 found the incidence of colic to be independent of maternal factors such as age, parity, and pregnancy history, as well as infant factors such as gender, weight, feeding habits, allergies, and weight gain. In self-reported studies of colic or colicky behavior, the odds of reported IC were slightly higher for infants with older, better educated, and more affluent parents.14 Reporting bias cannot be ruled out in those studies, because less affluent women with limited education may be reluctant to report problems for fear of appearing incompetent. The type of feeding (ie, breastfeeding versus bottle-feeding) has been reported to be correlated with colic, but the evidence regarding which type of feeding is positively associated with IC is contradictory.13–15 Despite at least 3 reports to the contrary,16–18 it is reasonable to suggest that cow's milk allergies are linked with colic for some infants. The positive evidence was collected from a series of studies that used the double-blind, crossover design to assess the association between exposure to cow's milk and IC.19–21 In those studies, symptoms of IC were reduced for both formula-fed infants and breastfed infants when cow's milk was eliminated from the infants' or mothers' diets, and symptoms reappeared when infants were orally challenged with bovine whey protein. In all studies in that series, IC was defined on the basis of Wessel's rule of threes. In contrast, in 2 of the studies with null findings, the diagnosis of colic was based on a history of persistent crying for no apparent reason17 or IC was defined as unexplained crying and restlessness (in a population-based survey).16 Because the diagnosis of IC was not based on Wessel's rule of threes, excessive crying might have been misclassified as IC. In a third study with null findings, a double-blind crossover design was used and the diagnosis of IC corresponded to Wessel's rule of threes. However, intolerance to cow's milk protein was evaluated with serum immunoglobulin E measurements and radioallergosorbent tests.18 It is difficult to detect specific immunoglobulin E among infants <1 year of age, and the secondary development of tolerance is usually not established among young infants.22 Therefore, the presence of cow's milk allergies cannot be ruled out completely in that study.
Evidence of a Link Between Maternal Smoking and IC
We searched Medline for original research articles, review articles, commentaries, and letters on maternal smoking and colic published between January 1, 1966, and March 10, 2004. This search was conducted in 3 steps. First, using the keywords “colic” and “infantile colic,” we produced a list of ∼4900 references. Next, using the keywords “maternal smoking,” “cigarette smoking,” and “tobacco,” we produced a list of ∼116000 references. Finally, we identified 25 articles that appeared in both lists. We reviewed those articles to identify those that specifically addressed maternal smoking and colic and to identify additional studies that we might have missed. This review led to a list of 7 references for 6 studies, including non-English reports that were not cited in Medline (Table 1). Two of the references were for the English abstract23 and the Danish report24 of the same study; our review of this study was based on the English report and our e-mail communication with the first author (March 2004). Another study25 had only an abstract available in English; our review of this study was based on e-mail communication with the study's first author (March 2004).
Existing Studies of Maternal Smoking and IC
Of the 6 studies, only 14 used Wessel's criteria to define IC (Table 1). The remaining studies used definitions that ranged from less-stringent variations of Wessel's criteria26 to definitions that would suggest excessive crying and not necessarily colic.27, 28 These distinctions are noteworthy, because relatively small differences in definitions can result in considerable variability in the estimated prevalence of IC and can also result in the identification of different subsets of infants and corresponding risk groups.29, 30
Studies of Excessive Crying
In a cross-sectional study of 253 infants examined at routine health surveillance visits at 3 months of age, Said et al27 investigated the associations between smoking of ≥1 cigarette per day by either parent and regular crying with apparent pain after feeding and crying for ≥2 hours in the early evening. Mothers were interviewed, and information on socioeconomic status, parity, birth weight, delivery conditions, feeding method, education, and parents' smoking habits was collected. In bivariate analyses, smoking of ≥1 cigarette per day by either parent, socioeconomic status, and feeding method were all associated with apparently painful crying after feeding (clinical indices were not reported) (P < .001, χ2 test). The association between parental smoking and crying after feeding remained significant after controlling for socioeconomic status and feeding method; however, the analytic strategy in the study was not well described. There was no association between parental smoking and crying in the early evening. This study also found that, among infants of mothers who did not smoke, there was a positive association between the extent of paternal smoking and the likelihood of colic. Although these findings are suggestive, this study over-relied on exploratory statistical methods, to the exclusion of more-appropriate multivariate methods. Moreover, the lack of relative measures of risk makes it difficult to determine the public health significance of the reported association between parental smoking and regular crying in the early evening.
In a second cross-sectional study, the link between colic-like behaviors and parental smoking was examined among a convenience sample of 80 infants.28 Visiting nurses in Stockport, England, recruited parents of 2 infants from their caseload, 1 who exhibited “nonspecific colic-like” behavior and 1 who did not. These nurses judged infants to be colicky or to have been colicky during the preceding few months on the basis of a list of colic symptoms provided by the investigators, which apparently included crying despite all basic needs being met, continuous crying for extended periods, and not being particularly pacified by handling and cuddling (the symptoms included in the questionnaire were not specified in the article; however, the introduction to the study suggested that these were the queried symptoms). The parents of the infants reported the number of cigarettes smoked in the home. In bivariate analyses, no association was found between the number of cigarettes smoked at home and colic-like behavior; multivariate analyses were not conducted. However, the small sample size and poor assessment of symptoms make these findings suspect.
In a third cross-sectional study, a random sample of 16% of women who gave birth in Norway in 1985 reported at 4 months after the birth, via mailed questionnaire, their demographic characteristics, postpartum smoking habits, breastfeeding, and infant disorders.23 Among mothers who smoked ≥5 cigarettes per day, 40% of breastfed infants exhibited excessive crying for more than 2 to 3 hours per day at least 4 days per week, compared with 26% of infants' of nonsmokers (P < .05). This association was independent of maternal education and socioeconomic status. This association was not observed among bottle-fed infants. In this study, the combination of smoking and breastfeeding was associated with a 50% increase in the incidence of excessive crying, whereas no association was found with the 2 factors separately. This was a well-designed, population-based study; however, the lack of relative measures of risk makes it difficult to draw inferences regarding the public health significance of the findings.
The fourth study was a retrospective investigation of a randomly selected sample of Danish children born in the autumn of 1995.25 Personal interviews were conducted with mothers at their homes when the infants were either 8 or 36 weeks of age. Eighty-seven percent of the mothers in the sample answered an extensive battery of questions, including questions on pregnancy complications, type of feeding, emotional health, and current number of cigarettes smoked per day. Mothers were also asked if their infants had had an “attack” of colic. Mothers who smoked ≥15 cigarettes per day were more likely than others to report that the infant had had an attack of colic (risk ratio [RR]: 1.30; 95% confidence interval [CI]: 1.05-1.61). This was a well-designed, population-based study. Although the study was retrospective in design, recall bias was an unlikely source of bias, because questions regarding smoking were included among many other health-related questions.
The studies reviewed thus far relied on the parents' perceptions of the severity of their infants' crying. Parents' perceptions of their children's crying has important practical implications, because such perceptions, independent of a medical diagnosis, often motivate concerned calls to the pediatrician and even visits to the emergency department. However, to establish links between potential antecedents and an outcome such as IC, it is important to have uniform definitions among studies. This can be achieved by using a widely accepted definition, such as Wessel's. To date, only 2 studies have used this definition, both of which were well-designed, population-based studies.
Studies of Colic
In a population-based, retrospective study conducted in the Netherlands between 1997 and 1998,26 physicians and nurses interviewed mothers of a nationally selected sample of infants, 1 to 6 months of age, regarding smoking, colic, and a host of related variables. IC was defined, on the basis of a modification of Wessel's criteria, as crying for ≥3 hours per day on >3 days in the week preceding the interview. Among completely formula-fed infants, maternal smoking of ≥15 cigarettes was associated with colic (RR: 1.8; 95% CI: 1.2–2.7), independent of socioeconomic status, obstetric history, and age of the infant.
Finally, in a population-based prospective study,4 mothers who delivered a singleton infant between May 1991 and February 1992 at a regional Danish hospital and who still resided in the region 8 months after the birth answered either self-administered or health visitor-administered questionnaires on sociodemographic factors, their pregnancy, and postpartum health-related behaviors, including smoking. In the self-administered questionnaires, colic was defined as several hours of crying per day for several days with legs drawn up toward the abdomen, distended abdomen, and excessive flatus. For health visitors, colic was defined as at least 3 hours of crying per day for at least 3 days per week for >3 weeks or >1.5 hours of crying per day in 6 out of 7 days and included the characteristic behaviors of drawn-up legs, abdominal distention, and excessive flatus. Compared with infants of nonsmokers, infants of mothers who smoked at least 15 cigarettes during pregnancy had a twofold increase in their risk of IC (RR: 2.1; 95% CI: 1.4–3.2). Infants of mothers who smoked after the birth but not during pregnancy were also twice as likely to have colic (RR: 2.0; 95% CI: 1.3–3.0). Surprisingly, infants of mothers who smoked both during pregnancy and after the birth were only 50% more likely to have colic, compared with infants of nonsmokers (RR: 1.5; 95% CI: 1.1–2.0).
In sum, there seems to be a link between maternal smoking and excessive crying as well as IC. With 1 exception, these studies suggest this link to be independent of feeding type. To date, only 1 study has considered both prenatal and postnatal exposure to tobacco smoke. There are currently no studies in which exposure to environmental tobacco smoke (ETS) has been directly measured and quantified.
PHYSIOLOGIC LINK BETWEEN MATERNAL SMOKING AND IC
Recent studies of the GI system provide strong but indirect corroborating evidence suggesting physiologic pathways through which maternal smoking can be linked to IC. This physiologic evidence can be outlined as follows: (1) smoking is linked to increased plasma and intestinal motilin levels and (2) higher-than-average levels of motilin are linked to elevated risks of IC.
GI Activity Among Humans and Motilin
GI activity among humans consists of 2 major contractile states, ie, the digestive and interdigestive states.31, 32 In the digestive state, repetitive contractions occurring simultaneously at all levels of the intestine mix and churn the nutrients with secretions and repeatedly present nutrients to the GI mucosa. The interdigestive or fasting state is characterized by the migrating motor complex (MMC), which consists of 3 phases. Phase I is a 15- to 60-minute period with little or no contractile activity, which is followed by a 15- to 60-minute period of intermittent and irregular contractions (phase II). Phase II is followed by a series of intense phasic contractions lasting ∼5 minutes (phase III), which completes the MMC cycle.
Motilin, a 22-amino acid hormone, is produced in the duodenal and jejunal mucosa and is present throughout the GI tract of humans and other species.33 Experimental evidence shows that intravenous infusion of the GI hormone motilin induces phase III activity.31 In particular, motilin stimulates gastric and intestinal motility by inducing interdigestive phase III antral and duodenal MMC.34 The only known function of motilin is initiation of the MMC.35
Motilin is released cyclically every 90 minutes in the fasting state. The highest plasma motilin levels are seen during fasting, when the average adult level is 72 ± 6 pmol/L, although individual levels vary over a wide range.36 Gut hormones are present early in fetal development, and adult patterns of motilin distribution in the GI tract are evident by 20 weeks of gestation. Fetal plasma motilin levels are ∼60% of maternal levels by 18 to 22 weeks of gestation. However, MMC is seldom seen before 32 weeks of gestation, suggesting that motilin receptor cells do not mature until late in the third trimester. Motilin levels in infancy are much higher than adult levels, with average preprandial levels of 172 ± 33 pmol/L and an average postprandial trough of 68 ± 12 pmol/L by day 13, and decrease with age.37 Serum levels also differ according to feeding type. Average basal serum motilin levels are higher for formula-fed infants than for breastfed infants (65 ± 44 pmol/L vs 32 ± 8 pmol/L).15
Recently, motilin receptor cells were isolated from the human stomach, duodenum, and colon. This suggests that motilin acts directly on smooth muscle cells. In addition to motilin's direct effects on smooth muscle, changes in motilin levels are partly under the control of the vagus nerve.38, 39 Motilin receptors have been found in the postrema area and on the vagus nerve. Animal studies suggest both local and vagally mediated control of motilin concentrations.40 It is also known that vagus nerve stimulation increases both the number and force of MMC contractions. Therefore, vagal stimulation of motility may be mediated by motilin. This suggests that the difficulty of comforting colicky infants may be partly the result of a self-perpetuating cycle of increased GI motility, continued crying, and, as a result of vagal stimulation, even higher plasma motilin levels.
It is also noteworthy that input from the vagal complex (cranial nerves IX–XII) and the phrenic and thoracic nerves to the intrinsic muscles of the larynx determine acoustic characteristics of the cry.41 Several studies have shown alterations in acoustic cry characteristics among infants with colic.42, 43 It has been suggested that the vagal complex may provide the neurophysiologic link between spasms in the GI tract and changes in acoustic cry characteristics associated with colic.44
Overall, the role of motilin in the central nervous system remains speculative, and there are conflicting reports of vagal stimulation both inhibiting and stimulating the release of motilin.45, 46 It is quite possible that several mechanisms are responsible and that motilin release is dependent on the equilibrium between stimulating and inhibiting factors.47
Linkage of Exposure to Tobacco Smoke With Elevated Plasma Motilin Levels
Two studies have examined the effects of smoking on serum motilin levels; both found a positive association between smoking and plasma motilin levels. In a study by Bell et al,48 heavy smokers, during the course of 1 hour, smoked either 4 high-nicotine cigarettes or 4 low-nicotine cigarettes. Smoking high-nicotine cigarettes, compared with low-nicotine cigarettes, resulted in significantly higher motilin levels and also in shortening of phase I and II of MMC. Beyond its acute effects, smoking seems to have chronic effects on motilin levels. In a second study designed to determine whether high motilin levels are related to smoking habits or only to acute nicotine exposure, serum motilin concentrations of fasting smokers who abstained from smoking for at least 10 hours were compared with serum motilin levels of fasting nonsmokers. Smokers had significantly higher motilin levels than did nonsmokers.49 These studies provide strong but preliminary evidence that smoking is linked to increased plasma motilin levels.
Linkage of High Intestinal Motilin Levels With Higher Risk of Colic
Motilin is known to reduce the small-intestinal transit time in humans and animals. Motilin is actively involved in the change in motility in rabbits with induced colitis.39 Plasma motilin levels are also significantly elevated among human patients with inflammatory bowel disease or Crohn's disease.31, 50 These findings are consistent with evidence that implicates motilin in diarrheal diseases.51, 52
Among humans, there is evidence that colicky infants have higher serum motilin levels than do infants without colic. In a hospital-based, case-control study, Lothe et al51 examined serum levels of 3 gut hormones, ie, motilin, gastrin, and vasoactive intestinal peptide. Case subjects were 40 infants with colic (age: 2–22 weeks), and control subjects were 42 healthy, age-matched infants (age: 3–36 weeks). The following important findings emerged from this study: (1) formula-fed infants, irrespective of colic status, had higher serum motilin levels than did breastfed infants; (2) infants with colic had significantly higher serum motilin levels than did similarly fed control subjects; and (3) vasoactive intestinal peptide and gastrin levels were not elevated, suggesting some specificity with respect to the role of motilin.
In a prospective study of 78 infants born during 1 month in 1986 at a Swedish hospital,15 motilin levels for 1-day-old neonates who later developed colic were higher than those for infants who did not develop colic. This is particularly noteworthy because elevated motilin levels occurred before the onset of any colic symptoms. In this study, formula-fed infants had higher basal motilin levels than did breastfed infants, irrespective of colic status, and there were no differences between motilin levels of mothers of colicky infants and those of mothers of infants without colic. The authors speculated that the increased motilin levels were “caused by an affectation of the gut present already at birth.”15(p414)
In sum, evidence suggests that infants with colic have higher motilin levels independent of feeding type, an association that is consistent with evidence that motilin levels are increased in disorders associated with GI dysregulation. Although these findings from disparate fields provide a cohesive model (Fig 1) for the physiologic pathway linking maternal smoking with IC, the entire chain of events has not yet been examined among a single cohort, nor has the link between maternal smoking and IC been replicated in a study that simultaneously considers all sources of prenatal and perinatal exposure to tobacco smoke.
CONSIDERATIONS FOR FUTURE STUDIES
Ideal Study Design
A prospective study begun in pregnancy and continuing through the first 4 months of life could provide definitive evidence linking these disparate lines of research. Assessment of all sources of tobacco exposure would require 2 interviews. Ideally, the first interview to assess maternal smoking during pregnancy would occur at the hospital (before the onset of colic), where it would also be possible to collect cord blood for the measurement of motilin levels. A more in-depth interview conducted 4 to 8 weeks later at home would allow diagnosis of colic and collection of information on maternal smoking, feeding patterns, and other confounding variables.
Comprehensive Assessment of Tobacco Exposure
Infants can be exposed to tobacco smoke and its metabolites while in utero, through breastfeeding, and through exposure to ETS after birth. Accurate estimation of infants' exposure to cigarette smoke and its metabolites is contingent on consideration of all 3 sources of exposure. The first source is in utero exposure. Throughout pregnancy, nicotine is readily transferred across the placenta; consequently, fetuses of mothers who smoke are exposed to relatively higher nicotine concentrations than are their mothers.53–55 Exposure can be assessed with standard questionnaires on smoking, and results can be validated with measurements of serum cotinine levels for a randomly selected group of participants.56 The second source is breastfeeding. Because of its lipophilic nature, nicotine is readily found in human milk, at concentrations that can be 3 times as high as those in the blood; because nicotine is absorbed orally, the infant of a smoking mother is exposed to significant amounts of nicotine if breastfed.57–60 When the infant is formula fed, it is important to note the formula type. The third source is ETS exposure after birth. There is a significant positive association between passive smoking during pregnancy and neonates' nicotine and cotinine levels, which can be as high as those found among infants of mothers who are light smokers.61–64 ETS exposure can be assessed with standard questionnaires.65 Ideally, a randomly selected subsample of participants could receive an additional visit ∼1 week before the home visit, for the purpose of placing an ETS monitor next to the newborn's crib. Data collected from the air monitors could be used to validate data from the questionnaires.
Measurement of Motilin
Serum motilin levels vary according to the volume, type, and acid and fat contents of oral intake and the time of measurement in relation to the last meal. Therefore, it would be ideal to extract motilin from cord blood, which is readily available from umbilical cord stumps routinely collected in the delivery suite. In addition, some women who refrain from smoking during pregnancy may begin to smoke after the birth. For this reason, and because of the possibility of ETS exposure, it would be necessary also to measure infants' motilin levels after birth. It would be preferable to assess infants' motilin levels at 4 to 8 weeks of age, to coincide with the time of peak incidence of colic. For this measurement, whole blood could be collected via heelstick. To account for the covariation between amount and time of feeding and serum motilin levels, the time and amount of each infant's last feeding must also be ascertained.
Diagnosis of Colic
It is important to diagnose colic on the basis of objective standards and continuous units of measurement. By asking for objective information, such as the number of hours the child cries on any given day, without reference to colic, researchers can minimize problems pertaining to the social desirability of the mother's responses. The Colic Symptom Checklist, a useful instrument that has been used in related work,42–44 allows quantitative measurement of the severity of colic symptoms and provides an overall score for the number of colic symptoms. The accuracy of a diagnosis of colic would be enhanced with information on competing diagnoses, the most prevalent of which is probably reflux; this is both because some reflux symptoms overlap with those of colic and because the incidence of reflux is linked to exposure to tobacco smoke.4–6 Reflux can be diagnosed with a readily available instrument.66
Confounding Variables
Because the hypothesis is that colic is at least partly a physiologic disorder, it would be useful to rule out related problems (in particular, a maternal history of GI problems). The mother's and infant's use of erythromycin, a motilin agonist that can influence motilin levels, must also be assessed. Maternal stress and mood would be useful to assess, preferably with structured instruments.67, 68 Sociodemographic variables must also be assessed. Key among these variables are those related to parenting skills and resources, such as parity and mother's age.
CONCLUSIONS
In the United States, nearly one-half of all women smokers continue to smoke during their pregnancies.69 This amounts to ∼12% of all women who give birth.70 More than 500000 infants each year are exposed to cigarette smoke in utero. Moreover, it is estimated that 50% to 80% of employed adults have regular exposure to ETS and that >30% of nonsmokers live with smokers.71 This is an important source of exposure, because breast milk of nonsmoking women can contain nicotine if the women are exposed to ETS.60 Although cotinine levels are lower among infants nursed by passive smokers, compared with active smokers, neither nicotine nor cotinine is found in the plasma of infants nursed by nonsmokers without ETS exposure, indicating that ETS exposure alone results in exposure of infants to nicotine through breast milk.72 The home is an important source of ETS. It is conservatively estimated that 40% of all US children <5 years of age are regularly exposed to tobacco smoke after birth, through someone in the household.73
Tobacco exposure, and its physiologic consequences outlined above, could have important effects on child growth and development. Colic in infancy has potentially long-lasting effects on both the mother-child dyad and the entire family. For example, in at least 1 study, children identified as having crying and feeding problems during infancy were more likely to be perceived as vulnerable and having behavior problems at 3 1/2 years of age. In this study, there was some indication that those who were more severely affected were also more likely to experience later adverse effects.74 Infants with colic may also have more sleep problems, both during the colic period and after the colic has resolved,75 as well as more feeding difficulties.9 A longitudinal follow-up study showed that infants who had colic at 3 months displayed more sleep difficulties and temper tantrums at 3 years of age, compared with children who did not have colic.76, 77
As distressing as IC may be for the infant, colic is also quite distressing for the caretaker. Caretakers of infants with colic have been reported as exhausted,78 depressed, and angry.79 These feelings can be expected to influence the strength of the mother-child bond during a critical period of development. To the extent that mothers of colicky infants provide fewer positive responses to their infants, relative to the mothers of noncolicky infants,80–82 colic may hinder mother-child attachment, with serious long-term consequences for the mother-infant dyad44 and the entire family unit.83 Preventing colic can improve caregiver's feelings of self-assurance. If, as we suspect, exposure to cigarette smoke increases the risk of colic, then this would provide additional incentives to parents to abstain from smoking. Decreased exposure to tobacco smoke can be expected to provide widespread, long-term, health benefits to maternal and child populations.84, 85
Acknowledgments
Support was provided to E.D.S. by Flight Attendants Medical Research Institute, by Transdisciplinary Tobacco Use Research Centers, funded by the National Cancer Institute, National Institute for Drug Abuse, and the Robert Wood Johnson Foundation and by Cancer Control and Prevention grant NIH R25T awarded to the Centers for Behavioral and Preventive Medicine, Brown Medical School.
REFERENCES
- Copyright © 2004 by the American Academy of Pediatrics
In this issue
