Stuttering in relation to anxiety, temperament, and personality: Review and analysis with focus on causality

G Model JFD-5548; No. of Pages 17

ARTICLE IN PRESS Journal of Fluency Disorders xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Journal of Fluency Disorders

Review

Stuttering in relation to anxiety, temperament, and personality: Review and analysis with focus on causality Per A. Alm ∗ Department of Neuroscience, Speech and Language Pathology, Uppsala University, Sweden

a r t i c l e

i n f o

Article history: Received 25 November 2013 Received in revised form 27 January 2014 Accepted 31 January 2014 Available online xxx

Keywords: Stuttering Anxiety Temperament ADHD Social cognition

a b s t r a c t Anxiety and emotional reactions have a central role in many theories of stuttering, for example that persons who stutter would tend to have an emotionally sensitive temperament. The possible relation between stuttering and certain traits of temperament or personality were reviewed and analyzed, with focus on temporal relations (i.e., what comes first). It was consistently found that preschool children who stutter (as a group) do not show any tendencies toward elevated temperamental traits of shyness or social anxiety compared with children who do not stutter. Significant group differences were, however, repeatedly reported for traits associated with inattention and hyperactivity/impulsivity, which is likely to reflect a subgroup of children who stutter. Available data is not consistent with the proposal that the risk for persistent stuttering is increased by an emotionally reactive temperament in children who stutter. Speech-related social anxiety develops in many cases of stuttering, before adulthood. Reduction of social anxiety in adults who stutter does not in itself appear to result in significant improvement of speech fluency. Studies have not revealed any relation between the severity of the motor symptoms of stuttering and temperamental traits. It is proposed that situational variability of stuttering, related to social complexity, is an effect of interference from social cognition and not directly from the emotions of social anxiety. In summary, the studies in this review provide strong evidence that persons who stutter are not characterized by constitutional traits of anxiety or similar constructs. Educational Objectives: This paper provides a review and analysis of studies of anxiety, temperament, and personality, organized with the objective to clarify cause and effect relations. Readers will be able to (a) understand the importance of effect size and distribution of data for interpretation of group differences; (b) understand the role of temporal relations for interpretation of cause and effect; (c) discuss the results of studies of anxiety, temperament and personality in relation to stuttering; and (d) discuss situational variations of stuttering and the possible role of social cognition. © 2014 Elsevier Inc. All rights reserved.

Contents 1. 2.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Concepts and interpretation of data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Interpretation of data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.1. Effect size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

00 00 00 00 00

∗ Correspondence to: BMC, Box 593, SE-751 24 Uppsala, Sweden. Tel.: +46 18 471 47 43; fax: +46 18 51 15 40. E-mail addresses: [email protected], [email protected] 0094-730X/$ – see front matter © 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jfludis.2014.01.004

Please cite this article in press as: Alm, P.A. Stuttering in relation to anxiety, temperament, and personality: Review and analysis with focus on causality. Journal of Fluency Disorders (2014), http://dx.doi.org/10.1016/j.jfludis.2014.01.004

G Model JFD-5548; No. of Pages 17

ARTICLE IN PRESS

2

P.A. Alm / Journal of Fluency Disorders xxx (2014) xxx–xxx

2.2.2. Shape of distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.3. Causal relations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Constitutional temperamental traits in persons who stutter? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Shyness at age two. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Level of cortisol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3. Easy versus difficult temperament . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4. Studies using the Children’s Behavior Questionnaire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5. Studies using the Behavioral Style Questionnaire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6. Experimental studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6.1. Studies of inhibitory control and behavioral inhibition in CWS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6.2. Studies of emotional reactions in CWS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.7. Retrospective assessment of traits of childhood ADHD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.8. Other studies of ADHD in relation to stuttering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.9. Quality of life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.10. Summary, constitutional temperamental traits in persons who stutter? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Will emotional reactivity increase the risk for persistent stuttering? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Development of social anxiety? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. Does anxiety or temperamental traits “drive” the motor symptoms of stuttering? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7. Effects of treatment: will changes in fluency affect anxiety, and vice versa? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8. Situational variations of stuttering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.1. Situational variability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.2. Situational variability in preschool CWS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.3. Variation of the audience size, and talking all alone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.4. Effects of strong emotions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.5. Interpretations of the “talk-alone-effect”? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.5.1. Stuttering as a threshold phenomenon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.5.2. Interference from social cognition? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.5.3. Differential effects of concurrent tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.6. Summary, outline of hypothesis of situational variations of stuttering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.1. Summary of results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

1. Introduction The symptoms of stuttering have led to many different theories about the cause of the speech disruptions. It has often been observed that stuttering changes from time to time, from situation to situation. For example, persons who stutter typically have the experience that it is easier to speak when alone than speaking in front of a group of people. It is conceivable that this type of observation has become extrapolated to also account for the cause of stuttering, leading to the theory that stuttering persons stutter because they tend to be more nervous and anxious than others. In psychology the term “temperament” is often used to signify the part of the personality that is assumed to be constitutional, biologically determined (e.g., Rothbart, Ahadi, & Evans, 2000). Others have used the term temperament differently, for example as “individual differences in emotion-based habit patterns” (Cloninger, 1994, p. 266). Personality and temperament in relation to stuttering have been the focus of a significant amount of research. When Charles Van Riper (1982) reviewed the literature on stuttering he found no evidence for any substantial difference in personality. Still, during the last two decades the assumption that persons who stutter tend to have a more emotionally sensitive/reactive temperament has become widespread. Emotionally reactive temperament means having strong emotional reactions in various situations, for example reactions of frustration, fear, or anger. The supposed sensitivity/reactivity in persons who stutter has been included in a theoretical framework of a multifactorial model, in which emotional sensitivity/reactivity is suggested to interact with linguistic or motor problems. Further, these traits have also been proposed to increase the risk of developing chronic stuttering, because sensitive/reactive persons are assumed to respond stronger to disruptions in speech fluency, for example with more muscular tension (e.g., Guitar, 2005; Walden et al., 2012). Lately a series of articles has claimed that stuttering in adults often is associated with personality disorders (Iverach, O’Brian, et al., 2009), mental health disorders (Iverach, Jones, et al., 2009), and certain traits of personality such as high neuroticism (Iverach et al., 2010). The review below will begin with a discussion of some fundamentals, such as the concepts used in this type of research. Thereafter, each of the main sections will be dedicated to different questions, in the following order: (a) Are there any specific constitutional traits in persons who stutter? (b) Will emotional reactivity increase the risk for persistent stuttering? (c) Is there a development of social anxiety in persons who stutter? (d) Does anxiety or temperamental traits ‘drive’ the motor symptoms of stuttering? (e) Will changes in fluency affect the level of anxiety, and vice versa? The review will end with a discussion of situational variability of stuttering. It is proposed that social cognition is a factor which tends to interfere with speech motor control in persons who stutter, rather than the emotional reaction of social anxiety.

Please cite this article in press as: Alm, P.A. Stuttering in relation to anxiety, temperament, and personality: Review and analysis with focus on causality. Journal of Fluency Disorders (2014), http://dx.doi.org/10.1016/j.jfludis.2014.01.004

G Model JFD-5548; No. of Pages 17

ARTICLE IN PRESS P.A. Alm / Journal of Fluency Disorders xxx (2014) xxx–xxx

3

2. Concepts and interpretation of data 2.1. Concepts The terms “personality” and “temperament” are central to the studies reviewed in this article. Some researchers use strict definitions of personality and temperament. Temperament is typically regarded as the constitutional, biologically determined aspect of personality. However, in practice we seldom have very reliable methods of determining what is really constitutional and what may be an effect of experience or of neurological influences such as birth complications, head injuries, etc. The terms personality and temperament are theoretical constructs which are deduced from observations of behavior and indications of emotions and cognition. In the present review the terms personality and temperament will be used in a relatively unspecific sense. Another terminological issue concerns the range of terms used in studies of aspects of anxiety, for example neuroticism, shyness, emotional reactivity, discomfort, sensitivity, vulnerability, fear, withdrawal, insecurity, and discomfort. When discussing anxiety in the present review also related concepts will be included. 2.2. Interpretation of data 2.2.1. Effect size In research on personality and temperament the common practice is to test for statistically significant group differences. If p is below 0.05 there is at least 95% chance that there is a group difference greater than zero. However, a group difference close to zero does not really provide any information about characteristics of the groups, and may therefore be of marginal interest. More information is needed to be able to interpret to what extent the trait actually characterizes persons who stutter. One such measure is the effect size. When comparing two groups it is often convenient to use the measure of effect size called Cohen’s d (Becker, 2000). Cohen’s d describes the difference between the means of the two groups, divided by the pooled standard deviations of the groups. The result is a measure which is independent of the scale used for measurement, and independent of the group size. If normal distribution of the groups is assumed, the degree of group overlap can be illustrated as in Fig. 1. This graph shows an example with d = 0.63. It can be seen that the two groups are largely overlapping. In other words, knowing which group a person belongs to does not provide much information about the individual persons. 2.2.2. Shape of distribution In Fig. 1, both groups have a normal distribution. Fig. 2 shows another type of distribution, bimodal, which is characterized by two peaks. In statistics the two peaks are described as major mode and minor mode. This distribution indicates the existence of a subgroup with somewhat different characteristics compared with the major group. In most studies of temperament and personality related to stuttering there is no information regarding the shape of the distribution. A real example of bimodal distribution is shown in Fig. 3, with data from Alm and Risberg (2007) regarding traits of childhood attention deficit hyperactivity disorder (ADHD). If the distribution is bimodal it may be the case that the majority have a completely normal result while a subgroup differs.

Fig. 1. Illustration of group overlap with effect size Cohen’s d = 0.63. If normal distribution is assumed this figure illustrates the size of the group difference for “lower perceptual sensitivity”, see Table 2.

Fig. 2. Example of bimodal distribution, with major mode and minor mode, indicating the existence of subgroup.

Please cite this article in press as: Alm, P.A. Stuttering in relation to anxiety, temperament, and personality: Review and analysis with focus on causality. Journal of Fluency Disorders (2014), http://dx.doi.org/10.1016/j.jfludis.2014.01.004

G Model JFD-5548; No. of Pages 17

ARTICLE IN PRESS

4

P.A. Alm / Journal of Fluency Disorders xxx (2014) xxx–xxx

Fig. 3. Histograms of scores on a modified version of the Wender Utah Rating Scale (WURS), indicating childhood traits of ADHD in adults (Alm & Risberg, 2007), for adults who stutter (above) and control subject (below). The scores are presented as mean score per item. The group of adults who stuttered showed a bimodal distribution, with indications of a subgroup with high scores for childhood traits of ADHD. A WURS score of 1.44 corresponds to the cutoff level for identification of likely childhood ADHD according to Ward et al. (1993).

2.2.3. Causal relations If a group difference is found the next step in the research process is to interpret the difference, in terms of cause and effect. For instance, if it is proposed that anxiety is a causal factor in the onset of stuttering it has to be shown that the anxiety exists before the onset. In studies of stuttering it is sometimes assumed that stuttering and the investigated variable have a direct cause-and-effect relationship. For example, if a difference in language ability is found it might be assumed that stuttering is a result of language impairment. However, the relation may instead be based on a common underlying factor, such as a difference in certain parts of the brain or in certain neurotransmitters, possibly resulting in several co-occurring symptoms. When interpreting group differences it is essential to not only consider direct causal relations, but also to consider possible underlying factors. 3. Constitutional temperamental traits in persons who stutter? To determine which temperamental traits are constitutional, one might observe children as young as possible. The older the children we study, the more likely it is that traits have been influenced by the experiences of stuttering. Studies of preschool children will be reviewed. 3.1. Shyness at age two The studies presented in Reilly et al. (2009) and in Reilly et al. (2013) are of special interest because the data comes from a prospective study of 1619 children. This study was based on data of children prior to the onset of stuttering. At the age of three a total of 137 children had been diagnosed with stuttering, at least for a period. The study included an estimate of shyness, and it was found that at the age of two there was no significant difference in shyness between children who would start to stutter and those who would not. In fact, the slight non-significant group difference was in the opposite direction, with the group of children who stutter (CWS) being marginally less shy (M = 15.7 for controls and M = 15.5 for the CWS). In summary, this study indicates that shyness at age two is not related to increased risk for later onset of stuttering. 3.2. Level of cortisol In a study of children who do not stutter (CWNS) Kagan, Reznick, and Snidman (1987) reported that shy and behaviorally inhibited children tended to have higher levels of the stress hormone cortisol. van der Merwe, Robb, Lewis, and Ormond (2011) investigated the level of salivary cortisol in seven preschool CWS, with mean age 4.1 years, and seven CWNS matched for age and gender. Baseline samples were collected at two occasions. The mean result from these two occasions was 9.25 nmol/L in the CWS, compared with 10.0 nmol/L for the CWNS. The group of CWS did not have elevated baseline levels of cortisol. Another sample of cortisol was collected after 10 min of conversation. Again the result did not indicate higher levels of stress in the group of CWS – if anything the result indicated the opposite.

Please cite this article in press as: Alm, P.A. Stuttering in relation to anxiety, temperament, and personality: Review and analysis with focus on causality. Journal of Fluency Disorders (2014), http://dx.doi.org/10.1016/j.jfludis.2014.01.004

G Model

ARTICLE IN PRESS

JFD-5548; No. of Pages 17

P.A. Alm / Journal of Fluency Disorders xxx (2014) xxx–xxx

5

3.3. Easy versus difficult temperament In three studies the temperament of preschool CWS have been investigated in terms of easy versus difficult temperament. One study had the result that CWS have a more easy temperament (Lewis & Golberg, 1997), one study indicated that they had a more difficult temperament (Wakaba, 1997), and the third study found no group difference (Williams, 2006). Considering these highly divergent results it appears unlikely that any substantial group difference in terms of easy versus difficult traits exist between preschool CWS and CWNS. 3.4. Studies using the Children’s Behavior Questionnaire A widely used instrument is the Children’s Behavior Questionnaire, CBQ, with questions to parents. In the present review the results from two studies of CWS are merged, resulting in a total of 96 CWS and 96 CWNS (Eggers, De Nil, & Van den Bergh, 2010; Embrechts, Ebben, Franke, & van de Poel, 2000). The mean age of the CWS was 5.07, ranging from 3.0 to 8.1 years. Table 1 shows results for four traits, which may be related to anxiety. These are shyness, sadness, fear, and discomfort. The p-values (from Z-tests) and effect sizes for the merged group were estimated based on the published data. None of the scales showed any significant group differences. Interestingly, the two scales showing the largest non-significant group difference, shyness and sadness, showed a tendency that was opposite to what might have been expected, with marginally lower mean scores for shyness and sadness in the group of CWS. This is in line with the result from Reilly et al. (2009) as discussed above. For fear and discomfort the groups were very similar. The results do not show that CWS actually would tend to be less shy or sad than other children, but it does provide strong evidence that shyness or sadness are not characterizing traits of CWS, at this age. Table 2 shows a summary of traits from Embrechts et al. (2000) and Eggers et al. (2010) which may be related to characteristics of attention deficit disorder, ADHD. According to DSM-V (American Psychiatric Association, 2013) the symptoms of ADHD can be divided into symptoms of inattention and symptoms of hyperactivity and impulsivity. Diagnosis can be differentiated into three types of ADHD: Predominantly inattentive type, predominantly hyperactive-impulsive type, and combined type. Inattention includes symptoms like: Do not seem to listen when spoken to, difficulties finishing tasks which are not intrinsically stimulating, and failing to give attention to details. Symptoms of hyperactivity-impulsivity include: running around, fidgeting, talking excessively, and difficulties with waiting. Five traits presented in Table 2 showed significant group differences, while the remaining four traits were relatively close to significance (p from 0.11 to 0.16). All scales showed higher mean scores for the group of CWS. The largest group difference was lower perceptual sensitivity for the group of CWS, related to items such as “Is [not] quickly aware of some new items in the living room.” This tendency may reflect inattention. Other significant group differences concerned lower inhibitory control, lower ability for attentional shifting, higher activity level, and higher motor activation. All of these aspects are consistent with traits in the ADHD spectrum. This picture is further supported by the close to significant aspects: higher impulsivity, higher anger/frustration, higher approach, and lower attentional focusing. In summary, the merged analysis of the studies of Embrechts et al. (2000) and Eggers et al. (2010), with a total of 96 CWS, did not find any tendency toward higher scores for traits related to anxiety. Rather a non-significant tendency in the opposite direction occurred. Significant group differences were found regarding aspects which are related to ADHD, with Table 1 Results from Embrechts et al. (2000) and Eggers et al. (2010), merged data, in total 96 CWS and 96 matched CWNS. The dimensions presented below are aspects which may be related to anxiety. A negative effect-size means that the scores were lower in the group of CWS. The p-values were estimated from Z-tests. Dimension

Effect size, Cohen’s d

p-Value

Shyness Sadness Fear Discomfort

−0.13 −0.07 0.03 0.02

0.67 0.59 0.91 0.95

Table 2 Results from Embrechts et al. (2000) and Eggers et al. (2010), merged data, in total 96 CWS and 96 matched CWNS. The dimensions presented below are aspects which may be related to ADHD. The p-values were estimated from Z-tests. Tendency of group of CWS

Effect size, Cohen’s d

p-Value

Lower perceptual sensitivity (Sample item: “Is [not] quickly aware of some new items in the living room.”) Lower inhibitory control Poorer attentional shifting (n = 58) Higher activity level Higher motor activation (n = 58) Higher impulsivity Higher anger/frustration Higher approach Lower attentional focusing

0.63 0.35 0.49 0.40 0.39 0.23 0.22 0.22 0.20

0.00002 0.014 0.01 0.006 0.04 0.11 0.12 0.13 0.16

Please cite this article in press as: Alm, P.A. Stuttering in relation to anxiety, temperament, and personality: Review and analysis with focus on causality. Journal of Fluency Disorders (2014), http://dx.doi.org/10.1016/j.jfludis.2014.01.004

G Model JFD-5548; No. of Pages 17

ARTICLE IN PRESS

6

P.A. Alm / Journal of Fluency Disorders xxx (2014) xxx–xxx

the group of CWS showing lower perceptual sensitivity (possible inattention), lower inhibitory control, poorer attentional shifting, and higher activity level. 3.5. Studies using the Behavioral Style Questionnaire The Behavioral Style Questionnaire (BSQ) is similar to the CBQ, in that it consists of questions to parents regarding the behavior of their children, resulting in scores on nine different traits. Anderson, Pellowski, Conture, and Kelly (2003) used the BSQ to investigate the temperament of 31 CWS, with mean age 4.0 years (ranging from 3.0 to 5.3 years). Three of the nine traits showed significant differences between the group of CWS and the CWNS, see Table 3. The largest group difference occurred for “distractability”. An example of low distractability would be when a child does not stop an activity when someone tries to catch his/her attention. The group of CWS achieved lower mean scores for this trait. The effect size calculated as Cohen’s d was 0.69, i.e., the mean of the two groups differed 0.69 standard deviations. This trait may be related to the strongest finding of the CBQ studies discussed in the previous section, where the group of CWS showed lower perceptual sensitivity (e.g., “is [not] quickly aware of some new item in the living room”). Both results may be related to inattention, and possibly to a tendency to become absorbed by something of specific interest, for example a certain computer game. The second largest group difference in Anderson et al. (2003) was for “lower adaptability to novelty”, described as the ease or difficulty with which behaviors can be changed in a desired way, like new routines. Poor adaptability is commonly associated with ADHD (Miodovnik et al., 2012). The third significant difference was “lower rhythmicity of biological patterns”, such as sleep, hunger, etc. Sleep disorders and disturbances of circadian rhythmicity are frequent in ADHD (Snitselaar et al., 2013), and so are eating disorders (Fernandez-Aranda et al., 2013). In summary, all three significant group differences in Anderson et al. (inattention, poor adaptability, and lower rhythmicity of biological patterns) appear to relate to traits which are common in cases of ADHD. Considering indications of elevated traits of ADHD in groups of CWS (see above, Eggers et al., 2010; Embrechts et al., 2000), a plausible interpretation is that also the results from Anderson et al. (2003) reflect traits related to aspects of ADHD. In the article by Anderson et al. the trait “less distractability” was described as “hypervigilance”. However, hypervigilance rather signifies a state involving increased distractibility, hypersensitivity, and anxiety, as described in Pflugshaupt et al. (2005): “. . . hypervigilance, . . . describes the tendency to constantly scan the environment for any signs of threat” (p. 107). The term “hypervigilance” in Anderson et al. may have given some readers an incorrect impression regarding the results of the study. The BSQ was also used by Karrass et al. (2006) in a study of 65 CWS and 56 CWNS (mean age 4.1 years, range 3.0–5.9). For the analysis, questions from the BSQ were grouped to form measures of “emotional reactivity”, “emotional regulation”, and “attention regulation.” These aspects showed significant group differences, see Table 4. The largest group difference was found for attention regulation. This was in agreement with the studies reviewed above, indicating traits of ADHD (Anderson et al., 2003; Eggers et al., 2010; Embrechts et al., 2000). The two other aspects, being less able to regulate emotions and being more emotionally reactive have also been proposed to underlie symptoms of ADHD, as discussed in Martel and Nigg (2006) and Walcott and Landau (2004). 3.6. Experimental studies 3.6.1. Studies of inhibitory control and behavioral inhibition in CWS Three published studies have attempted to measure inhibitory control and similar aspects using experimental setups. Eggers, De Nil, and Van den Bergh (2013) used a Go/NoGo button press task, with 30 CWS with mean age 7.4 years (range 4.8–9.9). The main group difference was found regarding premature responses. 14 out of the 30 CWS had at least one premature response, compared with only 2 out of 30 matched CWNS. This was interpreted as a stronger tendency toward impulsivity, among 47% of the CWS. Table 3 Significant group differences reported in Anderson et al. (2003), Behavioral Style Questionnaire (BSQ). 31 preschool CWS and 31 matched CWNS. Dimension

Effect size, Cohen’s d

p-Value

Lower distractability. (E.g., the child does not stop an activity when someone tries to catch his/her attention.) Lower adaptability Lower rhythmicity of biological patterns (hunger, sleep etc.)

0.69 0.63 0.57

0.008 0.016 0.03

Table 4 Significant group differences reported in Karrass et al. (2006). 65 preschool CWS and 56 CWNS (not matched for gender). Dimension

Effect size, Cohen’s d

p-Value

Poorer attention regulation Less able to regulate emotions More emotionally reactive

0.55 0.44 0.37

<0.01 <0.05 <0.05

Please cite this article in press as: Alm, P.A. Stuttering in relation to anxiety, temperament, and personality: Review and analysis with focus on causality. Journal of Fluency Disorders (2014), http://dx.doi.org/10.1016/j.jfludis.2014.01.004

G Model JFD-5548; No. of Pages 17

ARTICLE IN PRESS P.A. Alm / Journal of Fluency Disorders xxx (2014) xxx–xxx

7

Schwenk, Conture, and Walden (2007) studied 13 CWS age 3–5, and 14 matched CWNS. The experimental setup consisted of a remotely controlled motorized video camera, making a low-intensity noise when moving. The camera was moved several times when the child was talking with the caregiver, and the number of times the child looked at the camera when moving was registered. The results showed that the group of CWS on average looked at the camera a higher number of times (35% of all occasions the camera moved) than the CWNS (13%). The published data indicate that the group difference was based on a subgroup of CWS looking repeatedly at the camera when it moved. This may well be considered a normal behavior, but might also reflect some tendency toward impulsivity and low inhibitory control. A recent study by Choi, Conture, Walden, Lambert, and Tumanova (2013), however, reported the opposite result regarding behavioral inhibition. The study involved 26 CWS and 28 CWNS age 3.0–5.7 years, with mean age 4.0 years. The children had a conversation with an unfamiliar experimenter, and the measured variable was the time latency to the sixth spontaneous comment from the child (with the time for stuttering subtracted). There were no statistically significant difference between the groups in this regard, but the children with very long latency included a higher number of CWS and the group with very short latency included a higher number of CWNS. Further, the CWS with long latency also tended to show more severe stuttering. The authors summarized the conclusions of the study, as follows: “In summary, preschool-age CWS, when compared to their CWNS peers, are more likely to exhibit extremely high behavioral inhibition and less likely to exhibit extremely low behavioral inhibition. Furthermore, CWS’ behavioral inhibition appears to be associated with higher frequency of stuttered disfluencies. Findings of this study provide further support for recent theories of childhood stuttering . . . that propose that temperamental characteristics and/or related emotional processes are associated with childhood/developmental stuttering.” (p. 181) An alternative interpretation is that children with severe stuttering may hesitate more to speaking with unfamiliar persons, simply as a result of their severe speech difficulties. When looking at the data (Fig. 1) it is shown that the correlation between latency to speak and stuttering severity is based almost entirely on 5 outliers of the 21 CWS. The validity of this study as an investigation of behavioral inhibition in CWS can be questioned, because the measure of behavioral inhibition is the latency for speech initiation, and difficulty to initiate speech is a core symptom of stuttering. 3.6.2. Studies of emotional reactions in CWS An experimental study of emotional reactions in CWS was recently published by Ntourou, Conture, and Walden (2013). 18 CWS and 18 CWNS with mean age 4.3 years were studied in two different situations: one baseline neutral task and one task involving some frustration, when being unable to open a transparent box with an attractive toy. The child was alone in these situations and verbal and behavioral responses were registered using video. Emotional reactivity was defined as the duration of signs of negative or positive affect, including facial expression, verbalization, non-verbal sounds, and other behaviors. Attempts of emotion regulation were defined as self-speech and distraction. The CWS showed significantly more signs of negative affect compared with the CWNS, during both conditions. For both groups the signs of negative affect increased during the frustrating condition, more so for the CWS, but the group difference of the response did not reach statistical significance. It is of interest to note that the amount of negative or positive affect shown by the CWS was not related to the amount of stuttering in their speech. Furthermore, the CWS did not exhibit a greater percentage of stuttering during the frustrating condition compared with the control condition. The CWS showed more self-speech than the CWNS during the control condition (mean 3.6 versus 1.2 utterances), but not during the frustrating condition (mean 6.9 versus 5.8). The results that the signs of affect were not related to the symptoms of stuttering, and that the frustrating condition did not increase stuttering, suggest that emotional reactivity is not a main factor behind the breakdown of speech in CWS. Therefore, if some CWS show elevated indications of affect this may be more likely to be a co-occurring trait, for example as a part of (subclinical) traits of ADHD as discussed above. 3.7. Retrospective assessment of traits of childhood ADHD In Alm and Risberg (2007) the tendency toward childhood traits of ADHD was estimated retrospectively in adults who stutter (AWS), using a scale intended for the retrospective diagnosis of childhood ADHD, the Wender Utah Rating Scale, (WURS, Ward, Wender, & Reimherr, 1993). Seven out of 25 original items were excluded, because the questions were related to speech and social situations, all of which may be affected by stuttering. The results showed a strong bimodal tendency (see Fig. 3) with 13 out of 32 persons in the stuttering group showing higher scores than any participant in the control group. A WURS score of 1.44 corresponds to the cut-off level for identification of likely childhood ADHD according to Ward et al. (1993). Only 3 of the AWS had scores above this level, which suggests that most cases with high scores were “subclinical”, meaning that they would not fulfill the clinical criteria for ADHD. The dominating traits in the AWS with high WURS scores were inattention and daydreaming. The result in Alm and Risberg (2007) showed a close to significant tendency toward higher trait anxiety (in adult age) in the group reporting childhood traits of ADHD (p = 0.076). Many of the AWS showing high scores of childhood ADHD had reported some type of neurological incident before the onset of stuttering, such as head injury resulting in hospital care or unconsciousness, birth complications, or premature birth. In the half of the group of AWS with the highest scores for childhood ADHD 13 out of 16 reported this type of preonset incident, compared with only 4 out of 16 in the half of the Please cite this article in press as: Alm, P.A. Stuttering in relation to anxiety, temperament, and personality: Review and analysis with focus on causality. Journal of Fluency Disorders (2014), http://dx.doi.org/10.1016/j.jfludis.2014.01.004

G Model JFD-5548; No. of Pages 17

ARTICLE IN PRESS

8

P.A. Alm / Journal of Fluency Disorders xxx (2014) xxx–xxx

group of AWS with the lowest scores for ADHD (p = 0.004). These results are in line with the study of adolescents who stutter by Blood, Blood, Maloney, Meyer, and Qualls (2007). This study included 36 adolescents who stuttered of which 44% were reported to have a co-occurring disorder, such as attention deficit disorder, language disorder, central auditory processing disorder, behavioral disorder, reading problems, or some neurological disorder. The cases with such co-occurring problems tended to show an elevated level of anxiety (mean score 62.1) compared with fluent controls (mean score 51.7), while the cases without such problems essentially showed normal levels of anxiety (mean score 53.4). The results in Alm and Risberg (2007) seem to indicate that tendencies of ADHD in persons who stutter are linked to preonset neurological incidents, which also may have a causal role in the onset of stuttering. In contrast, Oyler (1994) reported that 40% (10 children) out of 25 CWS age 7–12 had a family history of attention deficits or hyperactivity, compared with 4% of the CWNS. This result suggests the possibility of genetic transmission for traits of ADHD in persons who stutter. A difference between the two studies was the age of the participants, with retrospective data from AWS in Alm and Risberg, and data from school children in the study by Oyler. A possible interpretation is that traits of ADHD in persons who stutter can have a genetic background or be caused by a neurological incident (or a combination of both), but stuttering related to neurological incidents may tend to be more persistent. This would result in a higher percentage of neurological incidents in samples of AWS compared with CWS. It is also possible that stuttering is linked with traits of ADHD through the speech disorder cluttering, which often has been associated with tendencies similar to ADHD (St.Louis, Myers, Bakker, & Raphael, 2007). In turn, cluttering often appears to have a relatively strong genetic component (Drayna, 2011). 3.8. Other studies of ADHD in relation to stuttering The studies reviewed above point toward the existence of traits of ADHD in at least a subgroup of CWS. Studies trying to estimate the prevalence of ADHD among persons who stutter have reached varying results, depending on the age of the group, methodology etc. The results range from 4% (Arndt & Healey, 2001), 5.9% (Blood, Ridenour, Qualls, & Hammer, 2003), 17% (Felsenfeld, van Beijsterveldt, & Boomsma, 2010), 18% (Biederman et al., 1993), 26% (Riley & Riley, 2000), up to 58% (Donaher & Richels, 2012). However, the last figure, 58%, is not for diagnosed ADHD but for the number of children who show symptoms that could warrant a referral to a specialist for possible diagnosis (the age of the children in this study was from 3.9 to 17.2 years, with a mean of 10.4). A tutorial on ADHD in relation to stuttering and to treatment of stuttering was published by Healey and Reid (2003) and is also discussed in Donaher (2011). Felsenfeld et al. (2010) analyzed data on 20,445 twins in the Netherlands, in a cohort study from 1987 onwards. Based on data from the parents at age 5 a total of 4.0% were classified as “probable stuttering” (PS) and 2.7% as “highly nonfluent but not stuttering” (HNF). Parents answered questions related to ADHD in their children at age 5 and 7, and teachers answered questions related to ADHD in these children at age 7. Both the PS and the HNF groups showed higher mean scores related to ADHD compared with typically fluent children. The size of the group differences was comparable to what had been reported above, with Cohen’s d effect size in the range of 0.38–0.58. For the HNF group and for the teachers’ ratings, the differences were more related to inattention than to hyperactivity. Based on similarities between monozygotic twins, dizygotic twins, and unrelated children the pattern of heritability was analyzed. Considering this analysis the authors proposed that stuttering and traits of ADHD in these children may arise from a common set of mechanisms. Felsenfield et al. emphasized that most children in this study with high scores related to ADHD were “subclinical”, i.e., would not get a clinical diagnosis of ADHD, but the tendencies may provide important information regarding the possible pathogenic mechanisms of the development of stuttering. Donaher and Richels (2012) analyzed family history and history of neurological incidents in relation to traits of ADHD. Seven children had a family history of recovering stuttering, and all of them fulfilled the criteria for referral regarding possible ADHD. In contrast, 15 children had a family history of only persistent stuttering, and only 40% of them fulfilled the criteria for possible ADHD. This data points to the possibility that recovering versus persistent stuttering may show differences regarding traits of ADHD. This is in line with the proposal of Ambrose, Cox, and Yairi (1997), that different genes may be related to recovering versus persistent stuttering. One hypothetical mechanism of recovering stuttering is discussed in Alm and Risberg (2007, pp. 353–355). All children tend to have a peak in the number of cerebral dopamine receptors at about the age of two to three. Differences in the balance between different receptor types (D1 and D2) may theoretically result in periods of reduced inhibition of behaviors, and at the same time may increase the risk for motor control instability. The very limited data from humans point toward a possible gender difference in this regard, with a higher risk for this type of hyperactivity in boys. In conclusion, it may be the case that preschool hyperactivity in fact is a positive prognostic factor in childhood stuttering. 3.9. Quality of life Reilly et al. (2013) recently published new data from the cohort discussed in Section 3.1, now from the children at age 4. In relation to temperament and emotional aspects two statistically significant differences were found. However, these differences were in the opposite direction to what might be expected. The quality of life was assessed using the PedsQL, Pediatric Quality of Life Inventory (PedsQL, 2013), based on reports from parents. It was reported that the mean quality of life for social functioning and functioning in preschool was higher for the group of children who stuttered or had stuttered Please cite this article in press as: Alm, P.A. Stuttering in relation to anxiety, temperament, and personality: Review and analysis with focus on causality. Journal of Fluency Disorders (2014), http://dx.doi.org/10.1016/j.jfludis.2014.01.004

G Model JFD-5548; No. of Pages 17

ARTICLE IN PRESS P.A. Alm / Journal of Fluency Disorders xxx (2014) xxx–xxx

9

previously compared with children who had never stuttered. The differences were statistically significant, with effect sizes calculated as Cohen’s d of 0.38 and 0.22, respectively. (For social functioning the mean score was 91.7 for the group of CWS and 87.2 for the group of CWNS.) Interestingly, the standard deviation was somewhat smaller for the group of CWS than for the group of CWNS, indicating that the result was not based on a subgroup of CWS. It is difficult to interpret this result, whether it reflects a real constitutional difference, whether it is a sign of social adaptation of CWS, or whether it is a bias in the reports from the parents. Anyway, some conclusions may be drawn. Firstly, considering the overlap of the two groups it seems clear that most CWS at age 4 do not differ at all from their peers who did not stutter regarding social functioning. Secondly, there is no indication at all that shyness, social phobia, or other psychological problems affecting social relations are characteristic of CWS at age 4. 3.10. Summary, constitutional temperamental traits in persons who stutter? Reports consistent with traits of ADHD in a subgroup of CWS have repeatedly been found in this review. In the DSM-V (American Psychiatric Association, 2013) the symptoms of ADHD is divided into two main categories, related to inattention and to hyperactivity-impulsivity. The following statistically significant group differences in the reviewed studies may be related to inattention: (a) lower perceptual sensitivity (Eggers et al., 2010; Embrechts et al., 2000); (b) inattention/daydreaming (Alm & Risberg, 2007); (c) poorer attentional shifting/regulation (Eggers et al., 2010; Karrass et al., 2006); and (d) lower distractability (Anderson et al., 2003). The following group differences may be related to hyperactivityimpulsivity: (a) higher activity level and motor activation (Eggers et al., 2010; Embrechts et al., 2000); (b) lower inhibitory control (Eggers et al., 2010, 2013; Embrechts et al., 2000; Schwenk et al., 2007); and (c) poorer emotional regulation and higher emotional reactivity (Karrass et al., 2006). Other aspects with relation to ADHD were lower adaptability and lower rhythmicity of biological patterns (Anderson et al., 2003). In Tables 2 and 3, above, effect size measures for these variables are summarized. The effect sizes (Cohen’s d) for the traits which may be related to inattention were in the range of 0.49–0.69, with a mean of 0.59. For traits which may be related to hyperactivity-impulsivity the effect sizes were smaller, ranging from 0.35 to 0.44, mean 0.39. This indicates that the dominating traits of ADHD among CWS are related to inattention, with somewhat lower effect size for traits related to hyperactivity-impulsivity. Information regarding the shape of the distribution is missing in most publications on traits of temperament or personality in persons who stutter. The results from Alm and Risberg (2007) and Eggers et al. (2013) suggests that less than half of CWS may show some tendencies of ADHD. There is a need for information regarding the shape of the distributions in more studies. It is important to emphasize that “traits of ADHD” do not mean that these children fulfills the criteria for clinical diagnosis of ADHD according to DSM-V. As discussed above, most CWS are likely to be no different from other children regarding traits of ADHD, but a subgroup appears to show elevated traits, at a subclinical or clinical level. No study in the review showed significant group difference in traits related to anxiety. Data regarding shyness and sadness instead showed a counterintuitive tendency toward slightly lower shyness and sadness in the group of CWS, though not statistically significant (Reilly et al., 2009; and merged results from Eggers et al., 2010; Embrechts et al., 2000). This slight tendency may well be a real effect, considering the indications of a subgroup of CWS with traits of hyperactivity. In conclusion, this review provides evidence that a subgroup of CWS shows subclinical traits of ADHD, particularly related to inattention but also to hyperactivity-impulsivity. The effect sizes for these traits are around 0.5 (see Fig. 1 for illustrations of effect sizes), and may affect somewhat less than half of the CWS, though there is a need for more information regarding the distribution. Only a small minority of the CWS are likely to show symptoms of ADHD which are sufficiently strong to warrant a clinical diagnosis. No indications of elevated traits of anxiety have been found in the studies of CWS during the preschool age. (However, this does not exclude that some preschool CWS experience social anxiety as a result of their speech problems.) Available data provide consistent indications that anxiety is not a constitutional trait of persons who stutter. Emotional reactivity and emotion regulation may be different in groups of CWS compared with CWNS, probably as a part of ADHD tendencies with hyperactivity-impulsivity in a subgroup of CWS. The possible causal relation between emotional reactivity and development of stuttering will be discussed in next section. 4. Will emotional reactivity increase the risk for persistent stuttering? It has been proposed that certain temperamental traits, especially emotional reactivity, would tend to increase the risk that early childhood stuttering continues and becomes persistent (e.g., Karrass et al., 2006). A central idea seems to be that many children experience disfluencies during childhood, but it is the reaction of the child that determines the course of development. According to this model, sensitive children may react to the disfluencies with stronger emotions and higher levels of muscular tension, which in turn could exacerbate the stuttering and make it persistent. An important factor in relation to this hypothesis is that there is a very high frequency of recovery from stuttering during the preschool years, with higher likelihood of recovery soon after onset. The mean age of onset appears to be between 2.5 and 3 years (Mansson, 2000; Yairi & Ambrose, 1992, 1999). If children with high emotional reactivity would tend to persist to stutter, and children with lower emotional reactivity would tend to recover, the result would be a gradual accumulation of children with high emotional reactivity in the group of CWS. At a group level the difference in emotional reactivity compared with CWNS would therefore increase with age. The study by Karrass et al. (2006) provides this type of data. The results showed that groups of CWS age 3–4 tended to show higher emotional reactivity than groups of CWNS. However, for Please cite this article in press as: Alm, P.A. Stuttering in relation to anxiety, temperament, and personality: Review and analysis with focus on causality. Journal of Fluency Disorders (2014), http://dx.doi.org/10.1016/j.jfludis.2014.01.004

G Model JFD-5548; No. of Pages 17

ARTICLE IN PRESS

10

P.A. Alm / Journal of Fluency Disorders xxx (2014) xxx–xxx

the oldest children, age 5, the CWS actually had somewhat lower scores for emotional reactivity compared with the CWNS. Though this group difference at age 5 was not statistically significant it is in the opposite direction to what was predicted by the hypothesis. In other words, this result, if correct, falsifies the hypothesis that high emotional reactivity increases the risk for persistent stuttering. It would be good to get data from further studies, but the data from Karrass et al. (2006) does not indicate that emotional reactivity is a risk factor for persistent stuttering. In contrast, the data from this study suggests that if emotional reactivity has any prognostic value it would be the opposite, that it would predict a better chance of recovery. Could this be possible? As discussed in Section 3.8, Donaher and Richels (2012) reported that CWS with a family history of recovering stuttering showed higher occurrence of traits related to ADHD than children with a family history of only persistent stuttering. If this can be corroborated it may turn out that hyperactivity and emotional reactivity in CWS are positive prognostic factors. In summary, the hypothesis that a certain trait contributes to persistence of stuttering can be tested by studies of the trait in different age groups during childhood. If the trait contributes to persistence of stuttering an increase of the trait in the group of CWS should be shown with older age. The available data from Karrass et al. (2006) points in the opposite direction, with the lowest emotional reactivity in the oldest age group. Until any study showing the expected pattern has been published it cannot be claimed that certain temperamental traits contributes to persistence of stuttering. 5. Development of social anxiety? In studies of AWS it has repeatedly been reported that many of them experience elevated levels of anxiety (Alm & Risberg, 2007; Blumgart, Tran, & Craig, 2010; Craig, 1990; Craig, Hancock, Tran, & Craig, 2003; Ezrati-Vinacour & Levin, 2004; Messenger, Onslow, Packman, & Menzies, 2004) and other negative mood states (Tran, Blumgart, & Craig, 2011). More detailed analysis indicates that the increased anxiety is limited to social situations involving speech, resulting in fear of negative social evaluation because of the stuttering (Messenger et al., 2004), and that treatment of stuttering focused on speech may reduce “trait anxiety” in AWS to normal levels (Craig, 1990). A recent meta-analysis of group differences between AWS and controls found an effect size of 0.82 for social anxiety and 0.57 for “trait” anxiety, with higher anxiety in the group of AWS (Craig & Tran, in press). If the anxiety is limited to social situations involving stuttering it may be hypothesized that the anxiety has developed as a result of previous negative social experiences related to stuttering. It is therefore of interest to look at studies of anxiety in persons who stutter of different ages. Studies related to social anxiety in CWS during preschool age were reviewed in Section 3. There were no indications at all regarding elevated levels of shyness in young CWS, indicating that social anxiety related to stuttering typically has not developed at this time. In school age CWS the results are somewhat mixed. Craig et al. (1996) did not find elevated state or trait anxiety in a study of CWS age 9–14. Oyler (1994) reported traits of vulnerability and sensitivity in a group of 7–12 year old CWS. In a questionnaire study of CWS and CWNS age 7–11 years De Nil and Brutten (1991) found more negative attitudes toward speech in CWS already at age 7, with increasing group difference with increasing age. In adolescents who stutter elevated social anxiety has been reported, compared with matched controls who did not stutter (Blood, Blood, Tellis, & Gabel, 2001; Mulcahy, Hennessey, Beilby, & Byrnes, 2008). In Blood et al. (2007) 44% of the adolescents who stuttered were reported to have a co-occurring disorder. The cases without co-occurring disorder showed relatively normal anxiety scores (effect size Cohen’s d = 0.38), while the group with co-occurring disorder showed substantial anxiety compared with the group of adolescents who did not stutter (effect size 1.54). This study indicates that stuttering adolescents with some additional disorder are especially at risk for development of social anxiety. According to a study by Erickson and Block (2013) is it common that adolescents who stutter experience heightened communication apprehension and try to keep their stuttering secret. It is of interest to note that this type of development may not be specific for stuttering: Beitchman et al. (2001) reported elevated frequency of social phobia in young adults with language impairment. 6. Does anxiety or temperamental traits “drive” the motor symptoms of stuttering? If the motor symptoms of stuttering are driven by emotional reactions one would expect to find correlations between the severity of the overt motor symptoms and anxiety or certain temperamental traits. However, empirical studies do not appear to have shown such correlation (Eggers et al., 2010; Mulcahy et al., 2008; Oyler, 1994; Tumanova, Zebrowski, Throneburg, & Kulak Kayikci, 2011). This may possibly be an effect of inadequate methods to investigate this question, but so far it seems that no such relation has been established. 7. Effects of treatment: will changes in fluency affect anxiety, and vice versa? In scientific research, an experiment is a method of investigation of cause and effect, with one factor being manipulated to see the effect on other variables. Treatment of stuttering which focuses on one single aspect may be regarded as a type of scientific experiment, in which the effects of changing this aspect is investigated. For example, if treatment exclusively targets speech fluency we may study the effect that fluency has on social anxiety. Conversely, if the treatment only targets anxiety we may study the effect of reduced anxiety on speech fluency. There are studies of fluency shaping for AWS, in which no techniques specifically aimed at anxiety have been included. The results regarding effects on anxiety have been somewhat mixed. In some older papers no reduction of anxiety was reported Please cite this article in press as: Alm, P.A. Stuttering in relation to anxiety, temperament, and personality: Review and analysis with focus on causality. Journal of Fluency Disorders (2014), http://dx.doi.org/10.1016/j.jfludis.2014.01.004

G Model JFD-5548; No. of Pages 17

ARTICLE IN PRESS P.A. Alm / Journal of Fluency Disorders xxx (2014) xxx–xxx

11

from improved fluency (Gregory, 1972; Ingham & Andrews, 1971). But in a more recent study Craig (1990) reported a strong positive effect. Interestingly, in this study the mean score for trait anxiety of the group of AWS was reduced to a somewhat lower level than the level for the adults who did not stutter (AWNS). This observation provides further support for the assumption that anxiety is not a core feature of persons who stutter, but rather a dynamic effect of the experienced and expected problems of speech. Conversely, will reduction of anxiety in itself reduce the motor severity of stuttering? Oliver Bloodstein (1967) discussed this question and reported that anxiety may be eliminated completely or almost completely, but frequently there is a considerable amount of stuttering left. A more recent report comes from Blomgren, Roy, Callister, and Merrill (2005). This was an investigation of treatment based on principles from Sheehan and Van Riper. Six months after the treatment there was still a substantial reduction of anxiety and avoidance, but the overt symptoms of stuttering were about the same as before the therapy. The severity of stuttering was determined both by the Stuttering Severity Instrument, third edition (SSI-3, Riley, 1972) and by self-report. SSI-3 reflects the merged scores for stuttering frequency, duration of stuttering moments, and associated movements and struggle. Any type of motor improvement would result in reduced scores of SSI-3. The result that anxiety was reduced but the SSI-3 scores were almost unaffected therefore speaks against the hypothesis that anxiety is the major drive behind the motor symptoms of stuttering. Menzies et al. (2008) reported a study of cognitive-behavior therapy (CBT) for AWS. The treatment showed a clear effect on social anxiety and psychological well-being, but did not show an effect in terms of improved fluency. Another type of treatment which is aimed specifically at anxiety is medication with benzodiazepines. In a review of Brady (1991) it was concluded that controlled studies of benzodiazepines for stuttering did not show an effect on the motor symptoms of speech. To summarize these studies there are indications that improved fluency (for persons succeeding with fluency shaping) may result in substantial reduction of anxiety. Studies of treatment focusing on anxiety have had important effects of improved psychological well-being, but generally have shown little or no effect on the motor symptoms of speech. If these treatment studies are viewed as scientific experiments it may be concluded that no major causal relation from anxiety to the severity of stuttering has been revealed. Reversed, there are indications of a causal relation from the degree of fluency to anxiety, e.g., Craig (1990). 8. Situational variations of stuttering 8.1. Situational variability The review above has focused on the relation between stuttering and psychological traits (or long-term states). Another question concerns the possible situational variation of stuttering, from moment to moment. For example, persons who stutter often report that it is much more difficult to read aloud in class compared with reading all alone or with only a pet as a listener (Rasskazov & Rasskazov, 2007). It is outside the scope of this paper to provide a complete review of these issues, but in order to give some additional perspective on the main topic the question of situational variability will here be briefly discussed and a theoretical interpretation will be proposed. 8.2. Situational variability in preschool CWS Meyers (1986) assessed the amount of stuttering in 12 preschool CWS, during three different conditions: Talking with (a) their own mother, (b) the mother of another CWS, or (c) the mother of a CWNS. The result was surprising, with no significant difference in the number of stuttered words between the three conditions. In a later study Yaruss (1997) studied stuttering variability in 45 preschool children. They were assessed during five different conditions, including “play with pressure”, which meant that a clinician asked frequent questions, interrupted, and increased time pressure. The result showed statistically significant differences between situations, but the differences were relatively small in absolute measures. The highest frequency of stuttering occurred during “play with pressure”, with a mean of 7.4% “less typical disfluencies”, which can be compared with 6.7% for “play”, and 5.4% for the condition with the least amount of stuttering (picture description). In other words, adding pressure to the play increased the amount of stuttering with about 11%. 8.3. Variation of the audience size, and talking all alone One way to investigate situational influence is to vary the number of listeners during a task, i.e., the size of the audience. This type of study with AWS has yielded mixed results, from no significant effect (Armson, Foote, Witt, Kalinowski, & Stuart, 1997; Young, 1965) to some increase of stuttering with increased audience (Hahn, 1940; Mullen, 1986; Siegel & Haugen, 1964). As mentioned above, persons who stutter often report that they may speak fluently or stutter much less when talking alone (when no one can hear), or when talking to a pet or an infant. This effect was studied experimentally by Svab, Gross, and Langova (1972), with 24 AWS. The participants were recorded in a soundproof chamber after they had been told that they should read a text all alone just to become acquainted with the chamber, and then push a button when they were ready. In this condition the number of stuttered words was reduced with about 60% compared with reading with the experimenter being present. Please cite this article in press as: Alm, P.A. Stuttering in relation to anxiety, temperament, and personality: Review and analysis with focus on causality. Journal of Fluency Disorders (2014), http://dx.doi.org/10.1016/j.jfludis.2014.01.004

G Model JFD-5548; No. of Pages 17

ARTICLE IN PRESS

12

P.A. Alm / Journal of Fluency Disorders xxx (2014) xxx–xxx

In a web-based survey of 1649 persons (Rasskazov & Rasskazov, 2007) about 66% claimed that they speak fluently when all alone. Further, 27% answered that they stutter significantly less when all alone, while 2.6% answered that there is no difference. The reliability of this result is difficult to estimate, for example because some instances of stuttering may be unnoticed, but it seems clear that such an effect exists at least to some degree. 8.4. Effects of strong emotions In some circumstances emotions and cognitions may become dissociated, for example in a situation with strong emotions and no cognitive ambivalence. How does this type of situation affect stuttering? Bloodstein and Bernstein-Ratner (2008, p. 270) meant that an interesting observation is that persons who stutter often tend to speak well when experiencing strong excitement, such as anger, enthusiasm, or fear. To illustrate this phenomenon a case from World War II was presented. Two persons who stuttered claimed that they were only able to communicate fluently when the situation was very dangerous. It seems likely that this type of situation is associated with extreme focus of attention on the task at hand, and that selfreflection and social considerations related to the stuttering were suppressed. One conclusion from this type of example may be that strong emotions, like fear, are not intrinsically negative for speech fluency. Instead it appears likely that strong focused emotions contribute to goal-directed activity and inhibition of competing activity. 8.5. Interpretations of the “talk-alone-effect”? When interviewing persons who stutter about their experiences of reading or talking when no one can hear, some say that it may be sufficient that they begin to think that someone may be listening for the speech to become disrupted. Assuming this is a real effect, how could this type of report be interpreted? This question will be discussed below. 8.5.1. Stuttering as a threshold phenomenon Stuttering appears to be a threshold phenomenon, meaning that fluent speech may be close to the neurophysiological threshold for disruption, but as long as the threshold is not passed no obvious symptoms are shown. This type of non-linearity means that also small perturbations of the speech motor system can have large effects. (In brief, non-linearity means that changes of the output of a system are not proportional to changes of the input. Threshold-type non-linearity means that the response of the system changes drastically at the point of the threshold.) 8.5.2. Interference from social cognition? The review in the present paper provides consistent indications that children who begin to stutter are not characterized by elevated anxiety or social inhibition. This suggests that the fundamental characteristics of CWS are localized to the speech production system and not to the emotional system. Whatever the exact nature of this difference, it appears that it is sensitive to other ongoing activity within the brain. It has often been assumed that it is emotional activity that interferes with the speech process, and thereby modulating the severity of stuttering in different situations. Another possibility is that it is cognitive activity that interferes with speech, especially social cognition. Social cognition involves thoughts about what one thinks of oneself, and what others may think or expect, regarding how one should behave. For persons who are concerned about stuttering it is likely that social situations often involve thoughts about possible scenarios, including what other may think if they stutter and alternative plans how to act. The amount of social cognition about stuttering in a certain situation may be hypothesized to be related to three main factors: (a) the importance and the possible consequences of the situation, (b) the risk for stuttering, and (c) the uncertainty about the best way to act. When talking alone with a pet, the possible consequences of the social behavior are negligible. When talking to the CEO of the company for which one would like to work the potential consequences are much larger. The risk of stuttering may be higher for example in a situation requiring certain words to be uttered, such as when reading aloud. The uncertainty about how to act may be very high when one has no preconceived strategy. In therapy persons who stutter may have learnt strategies such as using soft onsets, or continuing to talk regardless of the stuttering. What goes on in the brain when thinking about stuttering in a social context? Neuroscience research has shown that social cognition and self-reflection is especially related to processing in the medial prefrontal cortex (mPFC), which means the cortex region in the medial wall hidden between the two cerebral hemispheres in the most anterior part of the brain, approximately Brodmann area 10, see Fig. 4 (Amodio & Frith, 2006; Frith & Frith, 2006; Moran, Kelley, & Heatherton, 2013; Moran, Macrae, Heatherton, Wyland, & Kelley, 2006). For example, the region respond when a person is asked to estimate to what extent words such as honest, talented, lazy, or angry describe themselves. The mPFC is part of a network of brain regions which tend to be active during rest and when the brain is not focusing on goal-directed behavior, “the default mode network” (Fox et al., 2005; Moran et al., 2013; Spreng, 2012). This network exists in parallel with a goal-directed network, consisting of a varying set of regions depending on the nature of the task (Bush, Luu, & Posner, 2000; Fox et al., 2005; Spreng, 2012). In many situations these two networks show opposing responses, so that increase of activity in one network is followed by decrease in the other. This contrast tends to be more pronounced when more attention is given to a goal-directed task (Fox et al., 2005), which may reflect competition between processing of sensory information versus processing of internal sources of information, such as emotions and memories (Spreng, 2012). Please cite this article in press as: Alm, P.A. Stuttering in relation to anxiety, temperament, and personality: Review and analysis with focus on causality. Journal of Fluency Disorders (2014), http://dx.doi.org/10.1016/j.jfludis.2014.01.004

G Model JFD-5548; No. of Pages 17

ARTICLE IN PRESS P.A. Alm / Journal of Fluency Disorders xxx (2014) xxx–xxx

13

Fig. 4. Outline of brain regions in the medial frontal cortex with possible relevance for stuttering. SMA, supplementary motor area. CC, cingulate cortex. mPFC, medial prefrontal cortex. BA 10, Brodmann area 10.

Propositional speech is a goal-directed task, requiring processing of the cognitive content of the utterance and of the auditory and the somatosensory feedback. Based on the model outlined above, self-reflection and social cognition is likely to compete with the goal-directed process of speech, both with the content and the speech motor process. It may be the case that the speech network of persons who stutter tend to be unusually vulnerable to this type of interference. For example, much of the current research on stuttering and the brain are in line with the proposal that stuttering tends to be related to unstable bilateral control of speech movements (e.g., Alm, Karlsson, Sundberg, & Axelson, 2013; Chang, Erickson, Ambrose, Hasegawa-Johnson, & Ludlow, 2008; Helm-Estabrooks, Yeo, Geschwind, & Freedman, 1986; Kell et al., 2009). Bilateral speech motor control would mean that left and right speech motor regions in the lateral parts of the frontal lobe would need fast communication through the corpus callosum. This also includes regions in the medial frontal lobe involved in initiation and sequencing of speech movements, such as the anterior cingulate cortex and the supplementary motor area (SMA) (see review in Alm, 2011). The mPFC involved in social cognition includes part of the anterior cingulate cortex and is located close to the SMA. This type of bilateral organization of speech motor control may well be operating close to the threshold for breakdown, so that interference or competition also from normal social cognition may have a disruptive effect. When talking all alone the social evaluation system does not need to be active, but just the thought that someone may be listening might be sufficient to interfere with an unstable speech motor system. In Sections 6 and 7, the findings indicated that reduction of speech-related anxiety in AWS does not seem to have a clear effect on speech fluency. This result may appear to be in conflict with the discussion regarding interference from social cognition. However, alternative explanations may be conceived. One explanation could be that the influence of social cognition on stuttering is non-linear, so that the main disrupting effect occurs already for normal levels of social cognition. Another possibility is that reduced social anxiety after treatment is linked to reduced motivation to invest attention in control of speech. In this way there may be a conflict of goals in treatment of stuttering, so that successful reduction of concerns about stuttering would not necessarily imply reduction of the stuttering (which may not or may be a problem, depending on the goals of the individual). 8.5.3. Differential effects of concurrent tasks Tasks performed simultaneously with speech can either improve stuttering or make it worse. Bosshardt (2006) reported that concurrent tasks demanding phonological encoding and attention increased the amount of stuttering. On the contrary, Arends, Povel, and Kolk (1988) found that a simple hand motor task, to follow a moving dot on a computer screen using a joystick, reduced stuttering. This latter report is in line with observations that almost all types of volitional movement concurrent with speech appear to reduce stuttering (Bloodstein & Bernstein-Ratner, 2008). These observations may also fit with the model outlined in the previous section. Almost any volitional movement or focus of attention on sensory input can be expected to activate the goal-directed network and to inhibit the competing network involving social cognition. However, a concurrent task involving phonological encoding is likely to compete with speech for limited processing capacity (Bosshardt, 2006). 8.6. Summary, outline of hypothesis of situational variations of stuttering To summarize, stuttering is proposed to be a threshold phenomenon, meaning that fluent speech may be close to the neurophysiological threshold for disruption, but no obvious symptom is shown until the threshold is passed. In general, persons who stutter are not characterized by any constitutional trait of shyness or anxiety, but their neuromotor system for speech tends to be unusually sensitive even for normal processes of social cognition. It is suggested that anxiety in itself (the emotion) Please cite this article in press as: Alm, P.A. Stuttering in relation to anxiety, temperament, and personality: Review and analysis with focus on causality. Journal of Fluency Disorders (2014), http://dx.doi.org/10.1016/j.jfludis.2014.01.004

G Model JFD-5548; No. of Pages 17

ARTICLE IN PRESS

14

P.A. Alm / Journal of Fluency Disorders xxx (2014) xxx–xxx

is not a major interfering factor, but social anxiety is often linked to interfering social cognition. Social cognition is especially associated with activity in the medial frontal cortex, i.e., in the wall between the cerebral hemispheres. It is proposed that bilateral speech motor control will tend to be highly sensitive to interference from social cognition. The interfering effect of social cognition is suggested to be non-linear, so that the main interfering effect occurs also for normal social cognition. When talking all alone or with a pet, or during strong emotions, the social cognition can be assumed to be substantially reduced, resulting in reduced interference and improved fluency. Stuttering may be worsened by concurrent tasks which interfere with speech-related processing, such as phonological coding. Fluency can also be improved by increased activation of the goal-directed speech motor system. This activation may be the result of (a) concurrent volitional motor activity, for example hand movements; (b) increased focus of attention on sensory input (e.g., “proprioceptive monitoring”, Van Riper, 1973), (c) sensory stimulation (e.g., chorus reading, metronome, auditory white noise), or (d) increased drive (strong focused emotions). 9. Discussion 9.1. Summary of results The results from the review of traits in persons who stutter provide a quite consistent picture. No studies of preschool CWS in this review reported that shyness, social anxiety, or similar traits are more frequent in the group of preschool CWS. On the contrary, both the merged results from Embrechts et al. (2000) and Eggers et al. (2010) and the results from Reilly et al. (2009) found tendencies toward marginally lower shyness among CWS, though not statistically significant. It may be possible that this is a real but weak effect, reflecting a subgroup of CWS with traits of hyperactivity and impulsivity. It is easy to conceive that hyperactivity and impulsivity can result in low scores for shyness. Data from several studies provide consistent indications that the main temperamental difference between CWS (as a group) and CWNS is an increased tendency toward inattention and hyperactivity-impulsivity in the group of CWS. In most cases this tendency is not sufficient to result in a diagnosis of ADHD. Sparse information regarding the distribution suggests that the traits of ADHD reflect a subgroup of CWS, with the majority of CWS showing no difference from other children in this regard. It is therefore important not to make unwarranted generalizations, for example by claiming that traits of ADHD are a characteristic of CWS – because in most cases this is not correct. Instead it can be said that the data indicates that some CWS show elevated traits of inattention and/or hyperactivity-impulsivity, though in most cases too mild to warrant a diagnosis of ADHD. The largest difference is for inattention, with an effect size of about d = 0.6, and somewhat less for hyperactivityimpulsivity, about d = 0.4. From different studies there are indications that these traits of ADHD may be genetic and/or the result of neurological incidents. As discussed in sections 3.8 and 4 there are some empirical and theoretical indications that traits of hyperactivity-impulsivity may be a positive prognostic factor in childhood stuttering, but there is a need for more specific studies. There are no empirical indications that an emotionally reactive temperament is a risk factor for persistent stuttering, rather the contrary, as discussed in Section 4. It is clear that many persons with persistent stuttering develop social anxiety as a result of their speech problem. There is limited information regarding the typical age for this process. Available data points to the school age and the teenage years as important periods. Stuttering school-children and adolescents with additional disorders appear to have an elevated risk for the development of social anxiety. There does not seem to be a relationship between the severity of the motor symptoms of stuttering and temperament, as discussed in Section 6. Analysis of effects of treatment provides indications that improvement of fluency may result in reduced anxiety, though this has not been found in all studies. Regarding the reverse relation, the overall conclusion appear to be that reduced anxiety in AWS does not in itself lead to significant improvement of fluency. Situational variability of stuttering was discussed. In studies of preschool children the frequency of stuttering was relatively stable over different conditions, though with significantly more stuttering during occasions of added “pressure”. Studies of AWS have yielded mixed results, with a tendency toward more stuttering with larger audience. Data from covert recordings and self-reports indicate that stuttering in the majority of cases is substantially reduced when talking alone with no listener. It is proposed that the direct factor behind this effect is reduction of interfering social cognition. It is further proposed that the effect of social cognition on stuttering is non-linear, so that the main interfering effect occurs also for normal social cognition. 10. Conclusions The available evidence appears sufficiently strong to allow some conclusions: (a) Children who develop stuttering (as a group) are not characterized by temperamental traits such as shyness, social anxiety, or general anxiety. (b) A subgroup of CWS tends to show somewhat elevated traits of inattention and hyperactivity-impulsivity. (c) Many persons with persistent stuttering develop social anxiety as a result of their speech problem. (d) Reduction of anxiety in AWS does not generally in itself lead to significantly improved fluency. Acknowledgement The preparation of this article has been supported by grants from The Dominic Barker Trust, Ipswich, UK. Please cite this article in press as: Alm, P.A. Stuttering in relation to anxiety, temperament, and personality: Review and analysis with focus on causality. Journal of Fluency Disorders (2014), http://dx.doi.org/10.1016/j.jfludis.2014.01.004

G Model JFD-5548; No. of Pages 17

ARTICLE IN PRESS P.A. Alm / Journal of Fluency Disorders xxx (2014) xxx–xxx

15

CONTINUING EDUCATION Stuttering in relation to anxiety, temperament, and personality: Review and analysis with focus on causality QUESTIONS

1. Explain the meaning of an effect size of Cohen’s d = 0.5. (a) It is a measure of statistical significance; d = 0.5 means that it is not significant. (b) It is a measure of treatment effect; d = 0.5 means 50% reduction of symptoms. (c) It is a measure of group overlap; d = 0.5 implies that the means of two group differ with 0.5 (pooled) standard deviation. (d) It is a measure of correlation; d = 0.5 means a moderate correlation. 2. Which of the following is correct? Studies of preschool CWS indicate that (a) Anxiety is a constitutional trait of persons who stutter. (b) CWS show elevated shyness already before the onset of stuttering. (c) Inattention is the trait with the largest group difference between CWS and CWNS. (d) Preschool CWS show high levels of the stress hormone cortisol. 3. The hypothesis that a certain trait increases the risk for childhood stuttering to become persistent can be tested in the following way: (a) By comparing CWS with CWNS regarding this trait. (b) If the trait increases the risk for persistence, the percentage of CWS with this trait will increase with increasing age (because children without the trait have higher likelihood to recover). (c) By investigation of this trait in AWS. (d) By investigation of this trait before the onset of stuttering. 4. Which of the following is correct regarding anxiety in AWS? (a) AWS tend to have developed generalized anxiety, including anxiety for elevators etc. (b) AWS (as a group) have not been shown to differ from controls in measures of anxiety. (c) Stuttering in adults tends to vanish if speech-related anxiety can be overcome. (d) Stuttering may lead to development of speech-related social anxiety in adults. 5. In the review of situational variability (Section 8) it is proposed that (a) It is a myth that stuttering tend to decrease when talking all alone. (b) Normal social cognition may interfere with speech in persons who stutter. (c) Strong emotions are always detrimental for the speech of persons who stutter. (d) A concurrent hand motor task when talking tends to increase stuttering. References Alm, P. A. (2011). Cluttering: A neurological perspective. In D. Ward, & K. Scaler-Scott (Eds.), Cluttering: A handbook of research, intervention, and education (pp. 3–28). London: Psychology Press. Alm, P. A., Karlsson, R., Sundberg, M., & Axelson, H. W. (2013). Hemispheric lateralization of motor thresholds in relation to stuttering. PLOS ONE, 8, e76824. Alm, P. A., & Risberg, J. (2007). Stuttering in adults: The acoustic startle response, temperamental traits, and biological factors. Journal of Communication Disorders, 40, 1–41. Ambrose, N. G., Cox, N. J., & Yairi, E. (1997). The genetic basis of persistence and recovery in stuttering. Journal of Speech, Language, and Hearing Research, 40, 567–580. American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Washington, DC: American Psychiatric Association. Amodio, D. M., & Frith, C. D. (2006). Meeting of minds: The medial frontal cortex and social cognition. Nature Reviews Neuroscience, 7, 268–277. Anderson, J. D., Pellowski, M. W., Conture, E. G., & Kelly, E. M. (2003). Temperamental characteristics of young children who stutter. Journal of Speech, Language, and Hearing Research, 46, 1221–1233. Arends, N., Povel, D. J., & Kolk, H. (1988). Stuttering as an attentional phenomenon. Journal of Fluency Disorders, 13, 141–151. Armson, J., Foote, S., Witt, C., Kalinowski, J. S., & Stuart, A. (1997). Effect of frequency altered feedback and audience size on stuttering. European Journal of Disorders of Communication, 32, 359–366. Arndt, J., & Healey, E. C. (2001). Concomitant disorders in school-age children who stutter. Language, Speech, and Hearing Services in Schools, 32, 68–78. Becker, L. (2000). Effect size (ES). Retrieved from: http://web.uccs.edu/lbecker/Psy590/es.htm Beitchman, J. H., Wilson, B., Johnson, C. J., Atkinson, L., Young, A., Adlaf, E., et al. (2001). Fourteen-year follow-up of speech/language-impaired and control children: Psychiatric outcome. Journal of the American Academy of Child and Adolescent Psychiatry, 40, 75–82. Biederman, J., Faraone, S. V., Spencer, T., Wilens, T., Norman, D., Lapey, K. A., et al. (1993). Patterns of psychiatric comorbidity, cognition, and psychosocial functioning in adults with attention deficit hyperactivity disorder. American Journal of Psychiatry, 150, 1792–1798. Blomgren, M., Roy, N., Callister, T., & Merrill, R. M. (2005). Intensive stuttering modification therapy: A multidimensional assessment of treatment outcomes. Journal of Speech, Language, and Hearing Research, 48, 509–523. Blood, G. W., Blood, I. M., Maloney, K., Meyer, C., & Qualls, C. D. (2007). Anxiety levels in adolescents who stutter. Journal of Communication Disorders, 40, 452–469. Blood, G. W., Blood, I. M., Tellis, G., & Gabel, R. M. (2001). Communication apprehension and self-perceived communication competence in adolescents who stutter. Journal of Fluency Disorders, 26, 161–178. Blood, G. W., Ridenour, V. J., Qualls, C. D., & Hammer, C. S. (2003). Co-occurring disorders in children who stutter. Journal of Communication Disorders, 36, 427–448. Bloodstein, O. (1967). Interpersonal dynamics and the treatment of the stutterer. Journal of Communication Disorders, 1, 58–65. Bloodstein, O., & Bernstein-Ratner, N. (2008). A handbook on stuttering (6th ed.). Clifton Park, NY: Delmar Learning. Blumgart, E., Tran, Y., & Craig, A. (2010). Social anxiety disorder in adults who stutter. Depression and Anxiety, 27, 687–692. Bosshardt, H. G. (2006). Cognitive processing load as a determinant of stuttering: Summary of a research programme. Clinical Linguistics and Phonetics, 20, 371–385.

Please cite this article in press as: Alm, P.A. Stuttering in relation to anxiety, temperament, and personality: Review and analysis with focus on causality. Journal of Fluency Disorders (2014), http://dx.doi.org/10.1016/j.jfludis.2014.01.004

G Model JFD-5548; No. of Pages 17

ARTICLE IN PRESS

16

P.A. Alm / Journal of Fluency Disorders xxx (2014) xxx–xxx

Brady, J. P. (1991). The pharmacology of stuttering: A critical review. American Journal of Psychiatry, 148, 1309–1316. Bush, G., Luu, P., & Posner, M. I. (2000). Cognitive and emotional influences in anterior cingulate cortex. Trends in Cognitive Sciences, 4, 215–222. Chang, S. E., Erickson, K. I., Ambrose, N. G., Hasegawa-Johnson, M. A., & Ludlow, C. L. (2008). Brain anatomy differences in childhood stuttering. Neuroimage, 39, 1333–1344. Choi, D., Conture, E. G., Walden, T. A., Lambert, W. E., & Tumanova, V. (2013). Behavioral inhibition and childhood stuttering. Journal of Fluency Disorders, 38, 171–183. Cloninger, C. R. (1994). Temperament and personality. Current Opinion in Neurobiology, 4, 266–273. Craig, A. (1990). An investigation into the relationship between anxiety and stuttering. Journal of Speech and Hearing Disorders, 55, 290–294. Craig, A., Hancock, K., Chang, E., McCready, C., Shepley, A., McCaul, A., et al. (1996). A controlled clinical trial for stuttering in persons aged 9 to 14 years. Journal of Speech and Hearing Research, 39, 808–826. Craig, A., Hancock, K., Tran, Y., & Craig, M. (2003). Anxiety levels in people who stutter: A randomized population study. Journal of Speech, Language and Hearing Research, 46, 1197–1206. Craig, A., & Tran, Y. (in press). Trait and social anxiety in adults with chronic stuttering: Conclusions following meta-analysis. Journal of Fluency Disorders. De Nil, L. F., & Brutten, G. J. (1991). Speech-associated attitudes of stuttering and nonstuttering children. Journal of Speech and Hearing Research, 34, 60–66. Donaher, J. (2011). ADHD & children who stutter. Accounting for the impact of co-existing ADHD on children who stutter [video]. Memphis, TN: Stuttering Foundation of America. Donaher, J., & Richels, C. (2012). Traits of attention deficit/hyperactivity disorder in school-age children who stutter. Journal of Fluency Disorders, 37, 242–252. Drayna, D. (2011). Possible genetic factors in cluttering. In D. Ward, & K. Scaler-Scott (Eds.), Cluttering: A handbook of research, intervention and education (pp. 29–33). London: Psychology Press. Eggers, K., De Nil, L. F., & Van den Bergh, B. R. (2010). Temperament dimensions in stuttering and typically developing children. Journal of Fluency Disorders, 35, 355–372. Eggers, K., De Nil, L. F., & Van den Bergh, B. R. (2013). Inhibitory control in childhood stuttering. Journal of Fluency Disorders, 38, 1–13. Embrechts, M., Ebben, H., Franke, P., & van de Poel, C. (2000). Temperament: A comparison between children who stutter and children who do not stutter. In H. G. Bosshardt, J. S. Yaruss, & H. F. M. Peters (Eds.), Proceedings of the 3rd World congress on fluency disorders, in Nyborg, Denmark (pp. 557–562). Nijmegen, the Netherlands: Nijmegen University Press. Erickson, S., & Block, S. (2013). The social and communication impact of stuttering on adolescents and their families. Journal of Fluency Disorders, 38, 311–324. Ezrati-Vinacour, R., & Levin, I. (2004). The relationship between anxiety and stuttering: A multidimensional approach. Journal of Fluency Disorders, 29, 135–148. Felsenfeld, S., van Beijsterveldt, C. E., & Boomsma, D. I. (2010). Attentional regulation in young twins with probable stuttering, high nonfluency, and typical fluency. Journal of Speech, Language, and Hearing Research, 53, 1147–1166. Fernandez-Aranda, F., Aguera, Z., Castro, R., Jimenez-Murcia, S., Ramos-Quiroga, J. A., Bosch, R., et al. (2013). ADHD symptomatology in eating disorders: A secondary psychopathological measure of severity? BMC Psychiatry, 13, 166. Fox, M. D., Snyder, A. Z., Vincent, J. L., Corbetta, M., Van Essen, D. C., & Raichle, M. E. (2005). The human brain is intrinsically organized into dynamic, anticorrelated functional networks. Proceedings of the National Academy of Sciences, 102, 9673–9678. Frith, C. D., & Frith, U. (2006). The neural basis of mentalizing. Neuron, 50, 531–534. Gregory, H. H. (1972). An assessment of the results of stuttering therapy. Journal of Communication Disorders, 320–334. Guitar, B. (2005). Stuttering: An Integrated Approach to Its Nature and Treatment. Philadelphia, PA: Lippincott Williams & Wilkins. Hahn, E. F. (1940). A study of the relationship between the social complexity of the oral reading situation and the severity of stuttering. Journal of Speech and Hearing Disorders, 5, 5–14. Healey, E. C., & Reid, R. (2003). ADHD and stuttering: A tutorial. Journal of Fluency Disorders, 28, 79–93. Helm-Estabrooks, N., Yeo, R., Geschwind, N., & Freedman, M. (1986). Stuttering: Disappearance and reappearance with acquired brain lesions. Neurology, 36, 1109–1112. Ingham, R. J., & Andrews, G. (1971). The relation between anxiety reduction and treatment. Journal of Communication Disorders, 289–301. Iverach, L., Jones, M., O’Brian, S., Block, S., Lincoln, M., Harrison, E., et al. (2009). The relationship between mental health disorders and treatment outcomes among adults who stutter. Journal of Fluency Disorders, 34, 29–43. Iverach, L., O’Brian, S., Jones, M., Block, S., Lincoln, M., Harrison, E., et al. (2009). Prevalence of anxiety disorders among adults seeking speech therapy for stuttering. Journal of Anxiety Disorders, 23, 928–934. Iverach, L., O’Brian, S., Jones, M., Block, S., Lincoln, M., Harrison, E., et al. (2010). The five factor model of personality applied to adults who stutter. Journal of Communication Disorders, 43, 120–132. Kagan, J., Reznick, J. S., & Snidman, N. (1987). The physiology and psychology of behavioral inhibition in children. Child Development, 58, 1459–1473. Karrass, J., Walden, T. A., Conture, E. G., Graham, C. G., Arnold, H. S., Hartfield, K. N., et al. (2006). Relation of emotional reactivity and regulation to childhood stuttering. Journal of Communication Disorders, 39, 402–423. Kell, C. A., Neumann, K., von Kriegstein, K., Posenenske, C., von Gudenberg, A. W., Euler, H., et al. (2009). How the brain repairs stuttering. Brain, 132, 2747–2760. Lewis, K. E., & Golberg, L. L. (1997). Measurements of temperament in the identification of children who stutter. European Journal of Disorders of Communication, 32, 441–448. Mansson, H. (2000). Childhood suttering: Incidence and development. Journal of Fluency Disorders, 25, 47–57. Martel, M. M., & Nigg, J. T. (2006). Child ADHD and personality/temperament traits of reactive and effortful control, resiliency, and emotionality. Journal of Child Psychology and Psychiatry and Allied Disciplines, 47, 1175–1183. Menzies, R. G., O’Brian, S., Onslow, M., Packman, A., St Clare, T., & Block, S. (2008). An experimental clinical trial of a cognitive-behavior therapy package for chronic stuttering. Journal of Speech, Language, and Hearing Research, 51, 1451–1464. Messenger, M., Onslow, M., Packman, A., & Menzies, R. (2004). Social anxiety in stuttering: Measuring negative social expectancies. Journal of Fluency Disorders, 29, 201–212. Meyers, S. C. (1986). Qualitative and quantitative differences and patterns of variability in disfluencies emitted by preschool stutterers and nonstutterers during dyadic conversations. Journal of Fluency Disorders, 11, 293–306. Miodovnik, A., Diplas, A. I., Chen, J., Zhu, C., Engel, S. M., & Wolff, M. S. (2012). Polymorphisms in the maternal sex steroid pathway are associated with behavior problems in male offspring. Psychiatric Genetics, 22, 115–122. Moran, J. M., Kelley, W. M., & Heatherton, T. F. (2013). What can the organization of the brain’s default mode network tell us about self-knowledge? Frontiers in Human Neuroscience, 7, 391. Moran, J. M., Macrae, C. N., Heatherton, T. F., Wyland, C. L., & Kelley, W. M. (2006). Neuroanatomical evidence for distinct cognitive and affective components of self. Journal of Cognitive Neuroscience, 18, 1586–1594. Mulcahy, K., Hennessey, N., Beilby, J., & Byrnes, M. (2008). Social anxiety and the severity and typography of stuttering in adolescents. Journal of Fluency Disorders, 33, 306–319. Mullen, B. (1986). Stuttering, audience size, and the other-total ratio: A self-attention perspective. Journal of Applied Social Psychology, 16, 139–149. Ntourou, K., Conture, E. C. G., & Walden, T. A. (2013). Emotional reactivity and regulation in preschool-age children who stutter. Journal of Fluency Disorders, 38, 260–274. Oyler, M. E. (1994). Vulnerability in stuttering children (Doctoral dissertation). UMI Microform, No. AAT9602431. PedsQL. (2013). The PedsQL, Measurement model for the pediatric quality of life inventory. Retrieved from: http://www.pedsql.org/about pedsql.html

Please cite this article in press as: Alm, P.A. Stuttering in relation to anxiety, temperament, and personality: Review and analysis with focus on causality. Journal of Fluency Disorders (2014), http://dx.doi.org/10.1016/j.jfludis.2014.01.004

G Model JFD-5548; No. of Pages 17

ARTICLE IN PRESS P.A. Alm / Journal of Fluency Disorders xxx (2014) xxx–xxx

17

Pflugshaupt, T., Mosimann, U. P., von Wartburg, R., Schmitt, W., Nyffeler, T., & Muri, R. M. (2005). Hypervigilance-avoidance pattern in spider phobia. Journal of Anxiety Disorders, 19, 105–116. Rasskazov, I. U., & Rasskazov, N. M. (2007). Why do so many stutterers fail to stutter when alone and how can this phenomenon be used in treatment? Retrieved from: http://www.mnsu.edu/comdis/isad10/papers/rasskazov10.html Riley, G. D. (1972). A stuttering severity instrument for children and adults. Journal of Speech and Hearing Disorders, 37, 314–322. Reilly, S., Onslow, M., Packman, A., Cini, E., Conway, L., Ukoumunne, O. C., et al. (2013). Natural history of stuttering to 4 years of age: A prospective community-based study. Pediatrics, 132, 460–467. Reilly, S., Onslow, M., Packman, A., Wake, M., Bavin, E. L., Prior, M., et al. (2009). Predicting stuttering onset by the age of 3 years: A prospective, community cohort study. Pediatrics, 123, 270–277. Riley, G., & Riley, J. (2000). A revised component model for diagnosing and treating children who stutter. Contemporary Issues in Communication Science and Disorders, 27, 188–199. Rothbart, M. K., Ahadi, S. A., & Evans, D. E. (2000). Temperament and personality: Origins and outcomes. Journal of Personality and Social Psychology, 78, 122–135. Schwenk, K. A., Conture, E. G., & Walden, T. A. (2007). Reaction to background stimulation of preschool children who do and do not stutter. Journal of Communication Disorders, 40, 129–141. Siegel, G. M., & Haugen, D. (1964). Audience size and variations in stuttering behavior. Journal of Speech and Hearing Research, 7, 381–388. Snitselaar, M. A., Smits, M. G., van der Heijden, K. B., & Spijker, J. (2013). Sleep and circadian rhythmicity in adult ADHD and the effect of stimulants: A review of the current literature. Journal of Attention Disorders (Epub ahead of print). Spreng, R. N. (2012). The fallacy of a “task-negative” network. Frontiers in Psychology, 3, 145. St. Louis, K. O., Myers, F. L., Bakker, K., & Raphael, L. J. (2007). Understanding and treating cluttering. In E. Conture, & R. Curlee (Eds.), Stuttering and related disorders of fluency (pp. 297–325). Thieme. Svab, L., Gross, J., & Langova, J. (1972). Stuttering and social isolation. Effect of social isolation with different levels of monitoring on stuttering frequency. (A pilot study). Journal of Nervous and Mental Disease, 155, 1–5. Tran, Y., Blumgart, E., & Craig, A. (2011). Subjective distress associated with chronic stuttering. Journal of Fluency Disorders, 36, 17–26. Tumanova, V., Zebrowski, P. M., Throneburg, R. N., & Kulak Kayikci, M. E. (2011). Articulation rate and its relationship to disfluency type, duration, and temperament in preschool children who stutter. Journal of Communication Disorders, 44, 116–129. van der Merwe, B., Robb, M. P., Lewis, J. G., & Ormond, T. (2011). Anxiety measures and salivary cortisol responses in preschool children who stutter. Contemporary Issues in Communication Science and Disorders, 38, 1–10. Van Riper, C. (1973). The treatment of stuttering. Englewood Cliffs, NJ: Prentice-Hall. Van Riper, C. (1982). The nature of stuttering (2nd ed.). Englewood Cliffs, NJ: Prentice Hall. Wakaba, Y. Y. (1997). Research on temperament of stuttering children with early onset. In E. C. Healey, & H. F. M. Peters (Eds.), 2nd world congress on fluency disorders August 18–22, 1997, San Francisco, USA, (pp. 84–87). Nijmegen, The Netherlands: Nijmegen University Press. Walcott, C. M., & Landau, S. (2004). The relation between disinhibition and emotion regulation in boys with attention deficit hyperactivity disorder. Journal of Clinical Child & Adolescent Psychology, 33, 772–782. Walden, T. A., Frankel, C. B., Buhr, A. P., Johnson, K. N., Conture, E. G., & Karrass, J. M. (2012). Dual diathesis-stressor model of emotional and linguistic contributions to developmental stuttering. Journal of Abnormal Child Psychology, 40, 633–644. Ward, M. F., Wender, P. H., & Reimherr, F. W. (1993). The Wender Utah Rating Scale: An aid in the retrospective diagnosis of childhood attention deficit hyperactivity disorder. American Journal of Psychiatry, 150, 885–890. Williams, M. J. (2006). Children who stutter: Easy, difficult, or slow to warm up? In Paper presented at the ASHA convention 2006. Yairi, E., & Ambrose, N. (1992). A longitudinal study of stuttering in children: a preliminary report. Journal of Speech and Hearing Research, 35, 755–760. Yairi, E., & Ambrose, N. G. (1999). Early childhood stuttering I: Persistency and recovery rates. Journal of Speech, Language, and Hearing Research, 42, 1097–1112. Yaruss, J. S. (1997). Clinical implications of situational variability in preschool children who stutter. Journal of Fluency Disorders, 22, 187–203. Young, M. A. (1965). Audience size, perceived situational difficulty and stuttering frequency. Journal of Speech and Hearing Research, 8, 401–407.

Per A. Alm, Ph.D., is a researcher and teacher at the Department of Neuroscience, Speech-Language Pathology, Uppsala University, Sweden. The research focuses understanding of the mechanisms of stuttering and cluttering, with an integrative approach on brain functions and their relation to symptoms, speech motor control, psychological interactions, and background factors.

Please cite this article in press as: Alm, P.A. Stuttering in relation to anxiety, temperament, and personality: Review and analysis with focus on causality. Journal of Fluency Disorders (2014), http://dx.doi.org/10.1016/j.jfludis.2014.01.004