articleThis content is more than 5 years old.

Risk factors at age 8

Allan Tyrer | 01.06.2011

Will an eight-year old child still stammer as a teenager? About half do not. New research examines how to predict whether or not a child will recover.

Professor Peter Howell and Dr Steve Davis have published research on predicting at age 8 whether around age 13-15 a particular child will still have a stammer or will have recovered. Approximately 50% of children still stammering at age 8 years will recover by the teenage years. The study used data from 132 children to develop a model of persistence or recovery, with data from a further 74 children being used to validate the model. Recovery or persistence was determined by three separate reports: one from the parents of the child who stammered, a similar report from the child, and a set of ratings from a trained researcher.

The researchers found that severity of stammering predicted recovery or persistence in around four out of five cases.

Of the risk factors assessed at age 8, the only one found to be a significant predictor for recovery or persistence in teenage years was the severity of stammering. Other risk factors found not to be significant in predicting persistence or recovery were: head injury, age at stammering onset, family history of stammering, handedness, whether a second language was spoken in the home, and gender. The researchers found that severity of stammering predicted recovery or persistence in around four out of five cases.

Severity of stammering at age 8 was measured by Stuttering Severity Instrument Version 3 (SSI-3). The paper comments that the sensitive procedures used in the study lead to higher SSI-3 values than other procedures that are used in clinics.

The researchers suggest that risk factors for persistence and recovery near onset may not be the same as those later on. They quote Yairi and Ambrose: "Although very early severity is unrelated to outcome, later severity is a primary indication for concern."

The children participating in the study were from primary health care trusts around London who were referred to specialist clinics and confirmed as stammering. One of the caveats in the paper is that it is possible that the risk factors predict whether a child will respond successfully to treatment or not, rather than predicting persistence versus spontaneous recovery. A 'no-treatment' group was not used because it would be unethical to withhold treatment. Inclusion of a no-treatment group would be important for establishing the effect of receiving/not receiving treatment. The study was not designed to consider the effectiveness of therapy either for those who recovered or those who did not.

The starting point of age 8 was chosen because many children attend clinic at this age. The paper comments that starting the study at age 8 years does not mean that diagnosis and prediction of stammering persistence should be deferred to till then.

This study was supported by the Wellcome Trust, London. The researchers also thank the families and children who participated in the study for their long involvement and commitment to the project.

Howell P, Davis S. Predicting persistence of and recovery from stuttering by the teenage years based on information gathered at age 8 years. J Dev Behav Pediatr. 2011 Apr;32(3):196-205. doi:10.1097/DBP.0b013e31820fd4a9

Update October 2012: UCL has since published research on predicting recovery of children aged 5: Screening test for stuttering 'closer' (link to

Neuroimaging perspective

By Lu Chunming

Howell and Davis (2011) reported a very interesting study that showed that stuttering severity obtained at age 8 years can be used to predict persistence of, and recovery from, stuttering several years later. This study provides several interesting implications for the examination of the neural mechanism of stuttering.

1) The results of this study suggested that risk factors for persistence and recovery near onset are not the same as those during the later course of the disorder. This suggestion corresponds with the neuroimaging findings in stuttering. These findings showed that adults who have a long history of stuttering have some different brain functional and anatomical anomalies compared with children who have a relatively short history of stuttering (Brown et al., 2005; Chang et al., 2008; Watkins et al., 2008). Thus, it is necessary to examine the brains of people with different lengths of history of stuttering and to reveal what is the brain mechanism that underlies the risk factors for persistence and recovery.

2) While this study mentions the importance of the predictive information for clinicians, it is also worth noting the problem of some children being misclassified. Recently, it has been shown that people can be successfully classified as stuttering speakers or non-stuttering speakers based on brain functional activation while performing speaking tasks (Lu et al., 2010). So, it would be worth exploring whether brain functional or anatomical information can significantly increase the rate of correct prediction of persistence and recovery when it is combined with behavioural information. This kind of study will bridge the gap between brain research and clinical treatment.

Lu Chunming, State Key Laboratory of Cognitive Neuroscience and Learning, Beijing Normal University, China

1.Brown S, Ingham RJ, Ingham JC, Laird AR, Fox PT. 2005. Stuttered and fluent speech production: an ALE meta-analysis of functional neuroimaging studies. Hum Brain Mapp. 25: 105-117.
2.Chang S-E, Erickson KI, Ambrose NG, Hasegawa-Johnson MA, Ludlow CL. 2008. Brain anatomy differences in childhood stuttering. NeuroImage. 39: 1333-1344.
3.Watkins KE, Smith SM, Davis S, Howell P. 2008. Structural and functional abnormalities of the motor system in developmental stuttering. Brain. 131: 50-59.
4.Lu C, Chen C, Ning N, Ding G, Guo T, Peng D, Yang Y, Li K, Lin C. (2010). The Neural Substrates for Atypical Planning and Execution of Word Production in Stuttering. Experimental Neurology, 221(1): 146-156.

From the Summer 2011 edition of Speaking Out, page 18