Identification of Stuttering in Preschool Children: A Multifactorial Longitudinal Study in Development

Identification Of Stuttering In Preschool Children: a multifactorial longitudinal study in development

by Hans M´┐Żnsson
from Bornholm, Denmark


In recent years researchers all over the world have examined small children who stutter. Characteristic of this research is the fact that the section and number of participating children is dependant on parents and /or professionals dealing with small children perhaps incidentally, contact a research centre, typically a university clinic - or the other way round: that the research centre contacts children's institutions and parents in the specific region.

Whether the collected data forming the basis of this research may be considered representative for the population as a whole, may therefore be justly questioned.

The most serious attempt to reveal the prevalence, incidence and recovery of stuttering of a representative section of a population is the study of Andrews and Harris, published in the book ’The Syndrome of stuttering’ in 1964. This study took place in Newcastle, England and the data was collected over a period of 16 years - from 1946 to1962 - including approx. 1000 children.

The Andrews and Harris study is often referred to, when prevalence, incidence and remission is discussed.

The study certainly has its strong point regarding its very comprehensive time span and contents, but a number of aspects of the study have to be characterized as insufficient -such as:

  • it may be questioned to which extent the collection of data is based on qualified professional knowledge of stuttering,
  • there are no audio- or video recordings of stuttering, and
  • no definition of the terms stuttering, recovery and persistency is found.
  • Finally it may be stated that the study is highly based on a very traditional scientific view on the stuttering phenomenon.

Therefore there is a basic need to carry out present research on the emergence and development of stuttering, encircling the prevalence, incidence, remission and a number of other factors in a representative section of a population.

The Bornholm Programme attempts to contribute to the achievement of this goal.


Among other things the Danish island Bornholm is characterized by a geographical demarcation and relatively high population stability - although the last decade has shown some decrease of population, so that the present number of inhabitants is approx. 45.000.

Another characteristic of Bornholm is that since 1985 an extensive logopedic initiative has been carried out with reference to strengthening an early and preventive effort. In essence, prevention has a high priority on Bornholm, a fact that is exemplified by the island's attempt to screen all of its 3-year-old children.

This preventative procedure involves the offering, to all parents of children that have just turned 3 years of age, a visit to their home by a local speech-language pathologist (SLP). During this visit, the SLP evaluates the child's level of development in various areas, with particular attention paid to the child’s linguistic and communicative competence. In connection with this home visit the SLP has a thorough interview with the parents about the child and other aspects that the parents might find meaningful to discuss with a SLP. A very large majority of parents receive this screening offer.

In 1995 and the following years a pilot project was conducted, focusing on the topic of stuttering among small children on Bornholm. Project results and implications have been published in Journal of Fluency Disorders 2000/1 (p. 47-57).



In the years 1998 and 1999 where approx. 1000 children were screened, the ordinary 3-year screening model has been expanded so that the SLP has paid special attention to and asked about three risk indicators relating to stuttering:

  1. Does your child stutter at present?
  2. Has your child stuttered previously? Even short periods of stuttering are of importance.
  3. Are there present or previous examples of stuttering in the family?

If one or more of these questions have been answered affirmatively the child is registered as a ’child at risk’ and an extensive interview of the parents was carried out in immediate connection with the 3-year screening.

At the same time the parents were asked to let the child go through a more comprehensive test in the near future - including a sound recording of the child’s speech - and participate in a supplementary parental interview, conducted by the leader of the project.

The stuttering development of the children at risk and additional aspects is monitored for a minimum of 4 years.

In the test - and interview material the following areas are included, which the project attempts to throw further light upon a large number of factors.

The implemented intervention towards the children at risk is subject to an ongoing registration so that the study also throws light upon the effect of the logopedic effort which has occurred. (Månsson, 1998).

Theoretical anchorage:

As a theoretical frame of reference particularly Anne Smith and Ellen Kelly’s model of stuttering has been of inspiration (Curles & Siegel, p.204-217)

Smith and Kelly emphasize that stuttering is a multifactorial, non-linear, and dynamic phenomenon.

Their basic viewpoint is that stuttering may contain several - perhaps many different elements - in very different connections, importance and displacement of single factors. By focusing on one or more elements and excluding others and by regarding development of stuttering as a process characterized by great predictability and linearity often occurring in traditional models of stuttering, important underlying dynamic processes may be concealed, which may explain a development and remission of stuttering.

If for instance focus is on the linguistic capacity of the child, the dynamic point of view may be perceived by the following conceptual framework, which I have developed, illustrating that the different aspects of the language is at an interacting and changing relation to each

other (fig 1):


The conceptual framework used by the writer to encapsulate and organize the manifold variables and their interactions conjectured to be involved with the onset, development and maintenance of stuttering in children. While at present it may not be possible to descriptively and/or experimentally study the relevance or influence of all such variables, past and present research and speculation warrant their inclusion in the writer's conceptual framework for studying childhood stuttering.

But the linguistic development of the child exists in a mutual state of dependence on numerous other factors.

The conceptual framework stresses that stuttering may be under the influence of internal as well as external impacts. Therefore it is not sufficient to consider stuttering merely as a result of unbalance between the various components of the language.

The model emphasizes four fields of acknowledgement — biology, neurology, psychology and sociology, which separately and jointly may contribute to a wider frame of understanding of the emergence, development and possible remission of stuttering. These fields of knowledge indicate consequently that the etiology of stuttering may be contained within a continuum where the genetic code is one pole while the cultural and social context make up the other. At the same time the model indicates, that these areas form a dynamic, interacting sound-board for linguistic development.


Present space does not permit discussion of all relevant variables described in the mentioned conceptual framework (fig. 1). Thus, we have chosen to focus on, in an initial attempt to describe this data set, the prevalence, severity and other essential characteristics of stuttering of boys versus girls as viewed from the perspective of their parents.

The selected population consists of 1011 children (fig. 2). Of these 928 participated in the study, because 83 parents/-couples have refused the offer of a 3 year screening of their child. Thus the percentage of participation is 91.8.


At the 3 year screening the local SLP found a number of children where the parents answered in the affirmative to at least one of the three previously asked questions. Data has been collected from the screening and the parental interview regarding all these children who have been referred to me in continuation of the screening - with no exception.

Subsequently, I talked with each child´s parents in an attempt to assess the need or relevance for further, detailed testing of the child as well as inteviewing of the parents. As a result of these discussions with each child's parent(s), the writer excluded all cases where parents had answered in the affirmative to only the question of family disposition (i.e., children whose parents reported a positive family history for stuttering, but who exhibited no apparent stuttering in their speech, past and/or present). In addition some further cases where left out because the SLP and parental description of the non-fluent speech were apparently within normal limits for the child's age. This procedure eliminated a number of children from the total of 928 participating children, resulting in 174 children remaining, or 18.75% of the total participating children.

Thus, what remained in the 174 cases, were children whose speech contained, to a lesser or greater degree, stuttering-like disfluencies. Interestingly, the number of children exhibiting a noticeable amount of stuttering-like disfluency is unusually high compared to previous findings. That observation aside, during 1998 and 1999, the previously mentioned testing and interviewing conducted in the home of these 174 children and their parents, lead to the following observations (selected aspects of which will be presented below):

The boy-to-girl ratio is 98 to 76 - yielding 56.3:43.7 percent.

At the thorough interview the parents were for instance questioned about their view on their child's non-fluent speech at its peak.

The parental assessment of the degree of deviation of non-fluent speech - classified into six levels - is illustrated by figure 3:


Before the parents reflected on this question, they where instructed, that 0 and 1 on the scale is normal speech for the age. 0 refers to speech without error, which is almost hypothetical, whereas 1 represents normal speech with all the small errors characterizing ordinary speech. The remaining 6 categories are more or less deviating from normal speech for the age. 2 and 3 deviate mildly, 4 and 5 deviate moderately, and finally 6 and 7 deviate severely. This means, that the parents did not reflect on the stuttering diagnosis, but whether and to which extent the non-fluent speech deviates from normal speech for the age.

This may be compared with the parent's spontaneous answer to the question, whether the child's non-fluent speech is stuttering or not - asked in connection with the 3 year screening. This is reflected in figure 4:


The parental evaluation of the frequency of the non-fluent speech is contained in fig. 5.


This parameter didn't show significant gender differences.

The parental evaluation of the child's consciousness about and anxiety for the non-fluent speech appears in figure 6.


As mentioned a number of components of the stuttering phenomenon are included in the investigation. Consequently I have chosen the parental judgement of the number of repetitions and the length of the repetition sequence.

The parental evaluation of the number of repetition, when the non-fluent speech was at its peak, is shown in successive figure (7):


The parental evaluation of the length of the repetition sequence, is pictured in the final figure (8).



In the years to come and until 2004 further data concerning the children born in 1995 and 1996 will be gathered. This additional data will throw supplementary light upon the development of stuttering and especially the individual variations in progress and possible recovery of stuttering.

Thus a longitudinal dimension is applied to the cross-sectional study of stuttering in early childhood, which has already been implemented.

In this way the study may contribute to an earlier and more efficient effort towards stuttering through a more broadly facetted comprehension of the stuttering phenomenon and more differentiated prevention - and intervention programme.

The Bornholm Stuttering Programme is supported by grants from:

Egmont Fonden, Statens Humanistiske Forskningsråd, Undervisningsministeriet, Sygekassernes Helsefond, Gangstedfonden, Den Danske Banks Fond, Rosalie Petersens Fond, Socialministeriet, Foreningen af Stammere i Danmark, Lingvistisk-Logopædisk Studielegat, Center for Småbørnsforskning, DPU.


Andrews, G. & Harris, M. (1964). The syndrome of stuttering. London: Heineman.

Månsson, H. (1998): Stammeprojektet på Bornholm 1998-2004. Dansk Audiologopædi vol. 34/3, p. 103-107.

Månsson, H. (2000): Childhood Stuttering: Incidence and development. Journal of Fluency Disorders, vol. 25/1, p. 47-57.

Smith, A. & E. Kelly (1997): Stuttering: A dynamic, multifactorial model. In: Curlee, R. & G.M. Siegel (Ed): Nature and Treatment of Stuttering - New directions. Second Edition. USA, p. 204-217.

August 23, 2000