تغذیه با شیر مادر ممکن است در برابر لکنت زبان پایدار محافظت کند

Purpose This study investigated the hypothesis that breastfeeding in infancy might protect against persistent stuttering in children. Method We collected new data from the mothers of current and past participants in the Illinois Stuttering Research Program on their children's feeding history during infancy. We obtained 47 usable responses, for 17 children with persistent stuttering and 30 children who recovered naturally after a period of stuttering. Results A chi-squared test for linear trend revealed a significant relationship between breastfeeding duration and the likelihood of natural recovery for the boys in the sample. Mothers of children in the persistent group were no more likely to report early feeding difficulties which might have suggested an underlying oral motor deficit in children predisposed toward persistent stuttering. Conclusions Our results offer preliminary support for the idea that breastfeeding may confer a measure of protection against persistent stuttering. The fatty acid profile of human milk, with its potential to affect both gene expression and the composition of neural tissue, may explain this association. Further research is called for. Learning outcomes: The reader will be able to discuss at least one reason why human milk may make a difference in neurodevelopment generally and with regard to stuttering outcomes specifically. Additionally, the reader will be able to describe the relationship between breastfeeding duration and stuttering recovery observed in this sample.

Approximately 3–5% of preschool-aged children will begin stuttering. In some cases the onset of stuttering is dramatic: a previously fluent child will wake up one morning unable to utter a fluent sentence. An accompanying puzzle is the phenomenon of natural recovery: that same child may stutter severely for weeks or months and then gradually return, without intervention, to normally fluent speech. Preschool-aged children who begin to stutter will recover naturally approximately three-quarters of the time (Yairi & Ambrose, 1999), but predictor variables and causal factors are poorly understood. This study was driven by the hypothesis that breastfeeding could confer a measure of dose-related protection against persistent stuttering. The following sections will describe the rationale for this hypothesis, summarize the existing literature on breastfeeding and speech-language development, and briefly review the etiology of stuttering. 1.1. Human milk and neurodevelopment Fluent speech, with its 140,000 neuromuscular events per second (Darley, Aronson, & Brown, 1975), depends on normal neurodevelopment, and neurodevelopment can be influenced by diet in early infancy. It is widely recognized that the neurological system changes rapidly during the first two years of life; the question less often considered is what, precisely, is required to build a brain. A newborn's brain weighs, on average, 350 g; a year later it will weigh 1100 g (Lawrence & Lawrence, 2005). More than half the solid weight of that newly built tissue will be lipid, and dietary fat intake exerts a significant influence on its composition (Farquharson, Jamieson, Logan, Cockburn, & Ainslie Patrick, 1992). The rationale for this study is that the differing fatty acid profiles of human milk and infant formula have the potential to affect children predisposed to stuttering via two different mechanisms: first, by subtly altering the composition and thus the function of the brain itself, and second, by influencing gene expression (Jump, 2004 and Oddy, 2006). Human milk contains two fatty acids that have been identified as particularly important for early neurodevelopment: the omega-3 (subsequently abbreviated as n-3) fatty acid docosahexaenoic acid (DHA) and the omega-6 (n-6) fatty acid arachidonic acid (AA). DHA is the fatty acid most prevalent in the mammalian brain; DHA levels within the brain are determined by dietary levels (Innis, 2007). DHA and AA are present in both gray and white matter, including myelin (Nettleton, 1995). Across most of the lifespan the body can synthesize DHA and AA, as well as other long-chain fatty acids, from shorter-chain fatty acids. In early infancy, however, this synthesis process is not adequate for optimal brain development; research shows that the rate at which DHA is incorporated into brain tissue outstrips the rate at which it can be synthesized. In infants who lack a dietary source of DHA, other fatty acids are incorporated into the brain to compensate for the dearth of DHA (Farquharson et al., 1992). Thus differences in early diet can have long-lasting effects. In her 2006 study of the long-term effects of breastfeeding among Australian children, Oddy (2006) states, “Dietary alterations of n-3 and n-6 [fatty acids] can trigger dramatic alterations in brain lipid composition associated with changes in physical properties of membranes, alterations in enzyme activities, receptors, carrier mediated transport and cellular interactions” (p. 181). Differences in the fatty acid composition of brain tissue could contribute to differences in cell-to-cell communication, alterations in the function of synaptic membranes, and subtle impairments in nerve conductance and neurotransmission (Lauritzen et al., 2001 and Oddy, 2006). In contrast to human milk, most varieties of infant formula do not contain DHA/AA. Supplementation of infant formula with DHA/AA began in 2001 in the US, but the increased cost of EFA-fortified formula means that many children continue to receive the unsupplemented version. Furthermore, the impact of formula supplementation on neurodevelopment is controversial. Studies of long-term neurodevelopmental effects have reported no clear benefit to supplemented formula, suggesting that modified fatty acid profiles are not the sole determinant of neurodevelopmental outcomes (de Jong et al., 2010 and Smithers et al., 2010; see also a 2003 meta-analysis by Koo). Indeed, it would be reductive to argue that fatty acids are exclusively responsible for the differences observed in populations of breastfed and formula-fed children. Readers are referred to Riordan (2005) for further information on the properties of human milk that may influence neurodevelopment, and are encouraged to keep in mind that these observed differences may well be a synergistic effect of multiple nutrients, or may be related to as-yet-unidentified human milk constituents. Though the mechanism is incompletely understood, research does show that diet affects the composition of neural tissue. Farquharson et al. (1992) obtained necropsy gray matter samples from infants who died of SIDS, comparing exclusively breastfed babies with exclusively formula-fed babies. All the formula-fed babies had significantly less DHA in their cerebrocortical tissue than their breastfed counterparts; those fed formula were observed to have a corresponding increase in n-6 series fatty acids. Because of the ethical concerns raised by manipulation of the nutrient content of an infant's diet, animal studies can provide some additional information on this topic. Rodent studies are particularly useful for this purpose because of similarities between rodent pups’ prenatal/postnatal brain growth patterns and those of human infants. Lim, Hoshiba, and Salem (2005) compared adult rats given varying types of milk during infancy. In findings that paralleled those of Farquharson and colleagues in their study of human infants (1992), they reported that early diets high in n-6 fatty acids were clearly associated with brain tissue high in n-6 fatty acids. In the rat pups, this difference in neural tissue composition was associated with significant performance deficits, a finding that raises questions about the potential long-term ramifications of early diets high in n-6 fatty acids in human populations. Neurotransmission can be affected by fatty acid profiles; in addition, the fatty acid composition of a cell membrane can have significant effects on the expression of genes within that cell. Oddy reports: “n-6 and n-3 fatty acids directly govern the transcription rate of specific genes…. This means that the n-6 and n-3 in our cell membranes exert a significant influence on the way a given genetic profile is expressed” (p. 180). For researchers studying genetically mediated phenomena, including certain forms of speech-language impairment, this is an observation of critical import. Research into the impact of human milk on speech-language development has examined children from infancy onward. A 1999 study by Vestergaard et al. points to a connection between breastfeeding and early pre-speech development. In this prospective cohort study, babies who were exclusively breastfed for a longer period produced variegated babbling at earlier ages. The effect persisted after control for multiple potential confounding variables, including social class, maternal education, prenatal smoking, birthweight, gestational age, and number of prior illnesses. Taylor and Wadsworth (1984) studied the long-term effects of breastfeeding on a variety of domains. One of the questions in their survey asked parents whether their 5-year-olds had a “stammer, stutter, or other [speech] problem.” Based on this single question, they found no correlation between infant feeding choices and stuttering. Research into children's language skills has shown a consistent association between breastfeeding and improved language development (Dee et al., 2007, Gibson-Davis and Brooks-Gunn, 2006 and Oddy, 2006). Tomblin, Smith, and Zhang (1997) found that children breastfed for <9 months faced a significant increase in their risk of specific language impairment (SLI). All these authors reported significant effects after control for potential confounding variables. Covariate control is a persistent problem with breastfeeding research in cultures where middle-class educated women are most likely to breastfeed their children: it is difficult to determine whether outcomes of interest are related to breastfeeding, to maternal education, to socioeconomic advantage, or to synergistic effects of multiple factors. This difficulty was circumvented in a series of studies by Kramer and colleagues. They selected more than 17,000 pregnant women who planned to breastfeed after their children were born, and randomized them to two groups of healthcare providers (HCPs): one with extensive training in supporting the breastfeeding dyad, one with only the usual training. The children under the care of knowledgeable HCPs breastfed significantly longer than the control children, and the investigators have followed these children longitudinally to assess group differences across a variety of domains. In a 2008 study, they reported a statistically significant verbal IQ difference of 0.5 SD between the two groups of children at age 6.5. Some researchers have also investigated a link between diet in early infancy and subsequent diagnosis with an autistic spectrum disorder. Tanoue and Oda (1989) compared age at weaning for children with autism and typically developing controls, and concluded that breastfeeding duration was significantly longer in the control group. In 2006 Schultz and colleagues corroborated these findings, reporting an odds ratio of 4.41 for children without a dietary source of long-chain fatty acids. Taken together, these studies indicate that it is reasonable to assess for an association between breastfeeding in infancy and later communication skills. Research into the long-term effects of breastfeeding in other domains lends further support to this claim, since a number of genetically linked conditions occur less frequently in breastfed individuals (Léon-Cava, Lutter, Ross, & Martin, 2002). 1.2. Stuttering etiology To elucidate further the rationale underlying our hypothesis, we offer a brief review of stuttering etiology. Stuttering is known to have multiple etiologic strands: a genetic component underlies subtle neurological differences in people who stutter. Observers have long remarked that stuttering appears to run in families, and researchers have begun to elucidate the nature of stuttering's genetic component. In very early stuttering, a male-to-female ratio of approximately 2:1 has been reported; for persistent stuttering the ratio shifts dramatically, to the range of 4:1–6:1 (Ambrose, Cox, & Yairi, 1997). Kidd, 1980 and Kidd, 1984 hypothesized that this gender difference might point to a lower threshold for males, indicating that either more genetic loading or more environmental pressures or both were necessary for persistent stuttering to be expressed in females. This idea suggests that environmental variables, such as breastfeeding, might have different effects in boys than in girls. Howie (1981) reported a 63% concordance rate for stuttering in monozygotic twins, and a 19% concordance rate in dizygotic twins. This relatively high discordance rate demonstrates the importance of the interaction between genetic and environmental variables in determining stuttering outcomes, since more than a third of the twins with identical genetic material had divergent stuttering outcomes. More recently, Dworzynski, Remington, Rijsdijk, Howell, and Plomin (2007) reported stuttering concordance rates for children in the Twins Early Development Project and estimated heritability in 4-year-olds at .65, with most of the variability in stuttering outcomes arising from nonshared environmental variables. The environmental factors that contribute to stuttering outcomes, however, are poorly understood. To the best of our knowledge, diet has never been considered as a potentially relevant environmental influence. Most prior research on environmental variables has emphasized factors related to communication or emotional regulation (see, as examples, Kloth, Janssen, Kraaimaat, & Brutten, 1998, and Johnson, Walden, Conture, & Karrass, 2010; see also Cox, Seider, and Kidd, whose 1984 study found no prenatal or medical factors that predisposed children to stutter). Our focus in the present study is an environmental variable with potential neurobiological effects; it thus addresses a gap in the existing literature. A genetic predisposition toward stuttering is an important element of developmental stuttering, but genotype alone offers no certainty that an individual will begin to stutter. For the phenotype to appear, there must be an underlying neurological difference, the specifics of which are still speculative. Alm, for instance, suggested that in fluent speech the basal ganglia provide internal timing cues, while stuttering is a product of inadequate cues (2004). Other research has suggested that developmental stuttering may be associated with atypical myelination during the first year of life (Cykowski, Fox, Ingham, Ingham, & Robin, 2010), reduced white matter integrity (Chang et al., 2008 and Watkins et al., 2008), or other factors contributing to anomalies in speech timing (Sommer, Koch, Paulus, Weiller, & Büchel, 2002). Other researchers have proposed theories less tethered to specific neural structures. Bosshardt (2006) summarized two potential models. In one group of theories, it is hypothesized that an underlying instability exists in the speech motor control system. Smith and Kleinow (2000), for instance, found subtle differences in the kinematic parameters of the fluent speech of stuttering adults; Peters, Hulstijn, and van Lieshout (2000) described people who stutter as falling on “the weak end of the speech motor skill continuum” (p. 113). A second group of theories posits a difference that is more cognitive or linguistic than motoric. First articulated by Perkins, Kent, and Curlee in 1991, these theories propose that the linguistic and paralinguistic components of speech are handled by different neural systems, and that stuttering results when they are not synchronized. Bosshardt suggested that adults who stutter are slower to code phonological and semantic data, and concluded that speech planning and production are less modularized in individuals who stutter than they are in normally fluent individuals. The purpose of this paper is not to argue for or against any of these proposed etiologies. Rather, we propose that they could all be associated with the subtle alterations in the structure and function of neural tissue observed in children without a dietary source of long-chain n-3 fatty acids like DHA in early infancy. This study was designed to test the hypothesis that breastfeeding would provide dose-related protection against persistent stuttering, i.e., that shorter breastfeeding duration would be associated with higher rates of persistent stuttering. We hypothesized (1) that longer duration of breastfeeding would be associated with higher rates of recovery, and (2) that the observed effect would be more pronounced among the male participants, based on the findings of Broad, 1972 and Broad, 1975 and Broad and Duganzich (1983) and boys’ increased vulnerability to persistent stuttering.