Research in Early Childhood Stuttering

Ehud Yairi
University of Illinois at Urbana Champaign, Winter 1999-2000

To conduct research and develop effective clinical management strategies for any disorder it is important to have reliable data regarding its general incidence and prevalence as well as the incidence in different ethnic or racial sub-populations. The natural development and changes in symptomatology during the disorder's course, and the existence of any subtypes must be documented. This information guides differential diagnoses, prevention programs, selection of treatments for each stage of the disorder, timing of intervention, and the evaluation of treatment efficacy. Such information on early childhood stuttering has been scarce.

For many years, clinicians working with young children who stutter noticed that many preschool children outgrow stuttering without formal clinical intervention while others develop a chronic disorder. As early as 1938, Dr. Bryngelson, a well known pioneer in speech pathology, wrote that a substantial number of children who stutter would not need the help of a speech clinician because the stuttering would disappear of its own accord. Spontaneous (unaided) recovery in stuttering has been the focus of scientific attention and has stirred considerable controversy among scientists during the past several decades, reflecting its critical theoretical and clinical implications. In particular, the questions of how many children recover, how many become chronic, and whether all children who begin stuttering should receive immediate intervention, have been hotly debated.

Considering the intense interest, the scarcity of accurate data for the chronic and recovered groups becomes obvious. Past studies attempting to report such data were limited in scope and relied too heavily on second-hand, unreliable data, such as parents' reports. The Stuttering Research Project at the University of Illinois has sought to answer these questions with reliable data by conducting a longitudinal investigation that includes more than 150 preschool-age children who stutter as well as 60 normally fluent children. The Project, under the direction of Dr. Ehud Yairi with colleagues Drs. Nicoline Ambrose, Elaine Paden, and Ruth Watkins (all, University of Illinois), Nancy Cox and Edwin Cook (both, Chicago School of Medicine), Kelly Hall (Northern Illinois University); Rebecca Throneburg (Eastern Illinois University); and Ofer Amir (Tel Aviv University), and generously supported by the NIDCD.

The project is unique for two reasons. First, there has been a sustained effort to identify children close to the onset of stuttering. Second, through periodic follow-up observations, recordings of speech, and multiple assessments of other factors, the course of stuttering has been successfully documented across several years. With first-hand data from such a large pool of children, it has been possible to delineate comprehensive information on how stuttering begins and how it progresses. The major findings, as have been reported over the last several years in the Journal of Speech, Language, and Hearing Research (JSLHR), question longstanding concepts about the onset and developmental trends of early childhood stuttering.

For example, it has been believed that stuttering onset was always gradual and occurred under uneventful circumstances, that early symptoms included only easy repetition of syllables and words, and that parents helped create the problem by reacting negatively to normal disfluencies. The findings of the Illinois project, however, present a very different picture. Stuttering onset was sudden in at least one-third of the children; was severe in nature, including tense blocks, sound prolongations, and secondary physical characteristics for a good number of children; and, occurred close to a physically or emotionally stressful event for nearly 50% of all cases. The early belief that parents overreact to a child's normal disfluencies is unsupported by Illinois Project data as well. Disfluent speech regarded by parents "stuttering" is actually qualitatively and quantitatively different from normal disfluency, even in the earliest stages of the disorder. It appears that parents' concern is justified when they suspect that their child has begun stuttering.

In addition to data about onset, the investigators have reported other aspects of early stuttering using longitudinal research methods. Factors of interest included percentages of children who spontaneously recover as opposed to developing chronic stuttering, as well as precipitating factors for either of these two subtypes of the disorder. With regards to the number of children who spontaneously recover, a core study tracked children for four years following onset to determine distribution of these children as spontaneously recovered or exhibiting chronic stuttering at the end of the four-year period. The data indicate continuous reductions in frequency and severity of stuttering over time as many children progressed toward recovery. They show that whereas 26% (all of whom received some form of treatment) continued to exhibit chronic stuttering, a large majority (76%) recovered completely without treatment. The tendency to become persistent was greater among boys, and girls tended to recover at earlier ages. This led to the conclusion that for most children seen close to onset, stuttering is often (though not always) a short-lived disorder that disappears apparently on its own, without formal intervention. The investigators maintain that while all children who begin stuttering need to be closely monitored, those likely to persist should be identified early and have priority in receiving available clinical services. Further, the findings make clear that any claim of successful therapy must recognize the strong element of unaided, spontaneous recovery. Future treatment studies should include both adequate control groups and unbiased subject samples.

Isolating factors predictive of risk for chronic stuttering and those of eventual recovery has been another important line of investigation. The scientists associated with the Illinois Project reported several studies of these factors in the 1999 volume of JSLHR. One study examined persistency of stuttering and phonological abilities. Phonology (speech sound system) was assessed when the children were seen at the first visit, close to stuttering onset and before it was possible to determine who would develop chronic stuttering or recover. Phonology was also assessed during the periodic follow-up visits. Findings indicate that the group of children who persisted in stuttering performed less well on all phonological measures and phonological development progressed more slowly than did the group of children who recovered from stuttering. These findings suggest that presence of phonological difficulties in young children in the very early stages of stuttering may be a sign of chronic stuttering risk, but phonological skills alone may not be a sufficient predictor.

In a similar vein, the children's expressive language abilities were examined using several language measures. Results show that all children performed at or above developmental expectations for language, regardless of whether they later became chronic or recovered. These findings suggest that precocious language development may be another risk factor for stuttering, but not one that as yet differentiates persistent from transient stuttering. There seems to be a possibility of varied associations between language proficiency and stuttering development over time. These results yield important implications for future research of the relationship between early language abilities and early stuttering.

In a series of other studies designed to identify factors that may contribute to early identification of chronic and recovered stuttering, the disfluency patterns of the children have been analyzed over time. Findings in this respect are encouraging. It appears that within a year after stuttering onset, the trend of the number of certain disfluency types provides reasonable clues concerning the future course of the child's stuttering. Of particular interest is the finding that the initial severity of stuttering is not a good predictor of its eventual development. Two additional studies have focused on the speaking rate of the children and on specific acoustic features in their fluent speech. Although some positive findings have emerged, considerably more research is needed.

Finally, one of the most important studies of the project concerns the possible genetic bases of stuttering. Through detailed analyses of the incidence of the disorder within the families of the participating children provided strong evidence not only that stuttering, in general, has strong genetic components, but that the two subsets of children who stutter, those who persist and those who recover, have different genetic liabilities for stuttering. In other words, the tendency to persist or recover also tends to be heritable. This, as well as the other findings mentioned above, should help us formulate additional research to obtain clinically reliable diagnostic and prognostic procedures for children with high risk for chronic stuttering. Presently, the University of Illinois Stuttering Research Project with the cooperation of our colleagues from University of Chicago School of Medicine has placed emphasis on conducting a linkage analysis study designed to identify the gene, or genes, that are responsible for stuttering. For more information, please contact Dr. Ehud Yairi at (217) 244-2547 (, or Dr. Nicoline Ambrose at (217) 244-2559 (