Speech and Anxiety Management With Persistent Stuttering: Current Status and Essential Research

Abstract

Purpose

The purpose of this review article is to provide an overview of the current evidence base for the behavioral management of stuttering and associated social anxiety.

Method

We overview recent research about stuttering and social anxiety in the context of contemporary cognitive models of social anxiety disorder. That emerging evidence for self-focused attention and safety behavior use with those who stutter is considered in relation to current treatment approaches for stuttering: speech restructuring and social anxiety management.

Results

The emerging information about social anxiety and stuttering suggests a conflict between the two clinical approaches. For those clients who wish to control their stuttering and where speech restructuring is deemed the most suitable approach, it is possible that speech restructuring may (a) induce or increase self-focused attention, (b) promote the use of safety behaviors, and (c) become a safety behavior itself. This conflict needs to be explored further within clinical and research contexts.

Conclusions

The issues raised in this review article are complex. It appears that evidence-based speech treatment procedures are in conflict with current best-practice treatment procedures that deal with social anxiety. In this review article, we propose directions for future research to inform the development of improved treatments for those who stutter and recommendations for interim clinical management of stuttering.

Stuttering, being a disorder of speech production, can have a profound effect on a person's ability to communicate. For many who stutter, verbal output and language complexity are restricted. For instance, anticipation and fear of stuttering can prompt those affected to avoid using particular words and the intricacies of language associated with engagement (Lee et al., 2015; Spencer et al., 2005, 2009). Overall, quality of life can be impacted by stuttering (Craig, 2010; Craig et al., 2009; Cummins, 2010; Koedoot et al., 2011; Yaruss, 2010). For many, educational and occupational potential is restricted (Gerlach et al., 2018; McAllister et al., 2012; O'Brian et al., 2011). Many people who stutter will also experience various mental health conditions (Craig et al., 2015; Derogatis, 1994; Iverach, Jones, et al., 2009a, 2009b; Iverach et al., 2010; Iverach, O'Brian, et al., 2009; Manning & Beck, 2013), of which social anxiety disorder is the most prevalent (Blumgart et al., 2010; Iverach, O'Brian, et al., 2009).

Treatment for persistent stuttering typically involves teaching a client to change the way they talk so that they can communicate with greater ease, dealing with the psychological effects of the disorder or a combination of both. Charles Van Riper was one of the pioneers for stuttering treatment and described procedures that targeted both issues—fluency and anxiety (Van Riper, 1973).

To date, for those clients seeking help to minimize stuttering, the treatment approach with the strongest evidence base is speech restructuring (Onslow & Menzies, 2010). Speech restructuring involves teaching a novel speech technique to control stuttering during everyday speech. The resulting speech is intended to sound as natural as possible and acceptable enough for the speaker to use when needed.

Subsequent to early documented treatment procedures that aimed to target speech and anxiety (Van Riper, 1973), cognitive models of social anxiety disorder have informed the development of current gold standard evidence-based cognitive behavior therapy (CBT) procedures for social anxiety. CBT involves identifying and challenging unhelpful cognitions and behaviors and replacing these with coping strategies (Clark, 2001; Clark et al., 2003; Clark & Wells, 1995).

The thesis of this review article is that the emerging information about social anxiety and stuttering suggests a potential conflict between the two clinical approaches. Speech restructuring procedures for the control of stuttering appear to be in conflict with psychological treatment approaches for anxiety associated with stuttering. For instance, some speech treatment procedures may induce or maintain anxiety for some people who stutter who are prone to experience anxiety. Accordingly, we provide an overview of the current evidence-based speech and anxiety treatments for persistent stuttering as well as current research about stuttering and anxiety that signals a potential conflict between the two clinical approaches. We draw attention to essential research that is needed to allow the field to develop an optimal understanding about the nature of stuttering and to provide treatment for it. Finally, we propose interim clinical applications for the management of stuttering and anxiety.

Treatment for Persistent Stuttering

Treatment for persistent stuttering has a long history. Several prominent researchers and clinicians in the field foreshadowed current evidence-based procedures. One such pioneer in the field was Van Riper (1973), who described a treatment based on learning theory, psychotherapy, and stuttering modification procedures that involved dealing with fears and avoidance, in addition to modifying stuttering behaviors. However, there is limited clinical trial evidence available for those early treatment procedures (Blomgren et al., 2005; Georgieva, 2014). Therefore, it is not possible to evaluate the efficacy of those studies according to overall standards for clinical trials (World Health Organization, 2018) or according to generally accepted standards for trials of stuttering treatment (Onslow et al., 2008). Another treatment program for persistent stuttering that incorporates procedures to target speech behaviors and anxiety is the Comprehensive Stuttering Program (Boberg & Kully, 1994; Langevin & Boberg, 1993; Langevin et al., 2006). However, rather than employing stuttering modification procedures, the Comprehensive Stuttering Program uses speech restructuring procedures alongside components to manage anxiety.

The current empirical evidence for treating the speech motor component of stuttering is heavily weighted to speech restructuring. This treatment approach has many variants, described by terms including fluency shaping, prolonged speech, and smooth speech. Speech restructuring has a long history, dating back centuries (Packman et al., 2000). Its first formal description in modern clinical terms seems to have occurred during the 18th century (Bormann, 1969). From the 1960s, the discovery of the effects of delayed auditory feedback led to the development and clinical trials of modern variants of this technique (R. J. Ingham, 1984). At the end of the 1970s, a meta-analysis established the merits of this technique over other approaches (Andrews et al., 1980). Decades later, the situation has not changed. It appears that speech restructuring produces changes to the acoustic features of speech, which override the unstable speech motor system of those who stutter, thereby enabling fluent speech (Mallard & Westbrook, 1985; Onslow et al., 1994; Packman et al., 2007; Robb et al., 1985; Shenker & Finn, 1985). Many clinical trials of speech restructuring procedures have been reported during past decades, and preliminary trials have been reported for Internet-based speech restructuring treatments that do not require a clinician (for an overview, see Onslow, 2020).

While capable of controlling stuttering in the short term, many who have received speech restructuring treatment do not maintain treatment benefits in the long term (Andrews & Craig, 1988; Block et al., 2006; Craig, 1998; Craig & Calver, 1991; Craig & Hancock, 1995; Martin, 1981). Relapse rates range from 20% to 70% (Boberg & Kully, 1994; Craig & Hancock, 1995; Martin, 1981). Several variables have been associated with relapse: pretreatment stuttering severity (Andrews & Craig, 1988; Block et al., 2006; Craig, 1998; Huinck et al., 2006), locus of control (Bloodstein & Bernstein Ratner, 2008; Craig & Andrews, 1985), and attitude to communication (Guitar, 1976; Guitar & Bass, 1978). These reports highlight the so-called “revolving door” effect, where clients will re-present to speech clinics several times during their lives.

Although many researchers have attempted to establish predictors for the high rates of relapse following speech restructuring treatment, the most compelling evidence that has emerged is the presence of anxiety-related disorders (Craig & Hancock, 1995; Iverach, Jones, et al., 2009a). Subsequent to indications implicating self-reported relapse with anxiety, Craig and Hancock (1995) and Iverach, Jones, et al. (2009a) provided further evidence in support of this finding. In the Iverach, Jones, et al. (2009a) report, 64 adults received speech restructuring treatment. Up to two thirds of the participants were diagnosed with one or more mental health disorders, the majority being anxiety disorders. The critical finding of this study was that the presence of any mental health condition was associated with an increase in stuttering severity and avoidance of speaking situations 6 months posttreatment. This finding occurred in the absence of any differences in stuttering severity between those with and those without mental health disorders prior to treatment. This was the first empirical study to connect having a mental health disorder with long-term treatment outcomes.

Stuttering and Social Anxiety Disorder

More recent research indicates that adults who stutter are at high risk of developing social anxiety disorder. Up to 60% of those who seek treatment for stuttering warrant a diagnosis of social anxiety disorder (Blumgart et al., 2010; Iverach, O'Brian, et al., 2009; Menzies et al., 2008; Stein et al., 1996). Iverach, O'Brian, et al. (2009) found that a large cohort of adults seeking treatment for stuttering had 16- to 34-fold increased odds of developing a diagnosis of social anxiety disorder compared with community controls. Blumgart et al. (2010) reported that up to 46% of a combined sample of community and treatment-seeking participants who stuttered warranted a diagnosis of the condition. Moreover, using diagnostic interviews, up to 44% of adults seeking treatment for stuttering (Stein et al., 1996) and 60% of participants taking part in a clinical trial of CBT for stuttering (Menzies et al., 2008) met the criteria for a social anxiety disorder diagnosis. These high rates of social anxiety disorder are not constrained to adults who stutter. A report by Iverach, Jones, et al. (2016) found that 24% of children aged 7–12 years had the disorder. This high prevalence of social anxiety disorder is understandable, given that stuttering interferes significantly with verbal communication.

Social Anxiety Disorder

Social anxiety disorder is a common but serious psychological condition characterized by an extreme and debilitating fear of negative evaluation from others during feared social situations (American Psychiatric Association, 2013). Those with the disorder fear being humiliated in front of others, leading to an overestimated fear of negative consequences (American Psychiatric Association, 2013). The debilitating nature of the condition can impact significantly on daily functioning and subsequent quality of life (Sareen et al., 2006; Schneier et al., 1994; Slade et al., 2009; Stein & Kean, 2000; Stein et al., 2005). Social anxiety disorder can affect the capacity to develop and sustain relationships and fulfill educational and occupational potential. Furthermore, it is associated with a number of comorbid conditions, including depression and substance abuse (Kessler, 2003; Lampe et al., 2003; Marks, 1969; Turner et al., 1986; Wittchen et al., 1999). Social anxiety disorder can be a chronic condition (Liebowitz et al., 1985; Turner et al., 1986; Wittchen et al., 1999), with lifetime prevalence estimates ranging from 5% to 13% (Grant et al., 2005; Kessler et al., 2005; Kirmizioglu et al., 2009; Ruscio et al., 2008; Stein & Kean, 2000).

Treatment for Social Anxiety Associated With Stuttering

CBT is the most empirically evaluated and efficacious treatment approach for anxiety disorders, including social anxiety disorder (Clark et al., 2003; Mattick et al., 1989; Tolin, 2010; for reviews, see Acarturk et al., 2009; Rodebaugh, Holaway, & Heimberg, 2004). CBT involves identifying and challenging negative and unhelpful thoughts and beliefs. Procedures then target those thoughts through psychoeducation, cognitive restructuring, exposure therapy, and behavioral experiments.

Based on gold standard CBT procedures, a program has been developed specifically to deal with social anxiety experienced by adults who stutter (Menzies et al., 2008). This treatment targets specific negative thoughts and concerns associated with stuttering (St Clare et al., 2009) and is based on contemporary theories of social anxiety disorder (Clark & Wells, 1995). The procedures in this CBT program are broadly consistent with procedures developed by Van Riper (1973). Several studies have evaluated the merits of using the Menzies et al. (2008) CBT program with adults who stutter. In a randomized controlled trial, Menzies et al. (2008) administered CBT before speech treatment to determine whether it had any effect on stuttering severity. The participants who received CBT before speech treatment showed improved scores for Global Assessment of Functioning (American Psychiatric Association, 1994) immediately posttreatment, and the trend for improvement continued at 12 months posttreatment. In addition, the CBT program was associated with significantly less avoidance of speaking situations than for the control group, and this trend was retained at 12 months posttreatment. For those participants who received CBT before speech treatment, the CBT did not appear to affect speech outcomes for percent syllables stuttered, as measured by blinded observers. However, as would be expected, both groups demonstrated lower scores for percent syllables stuttered following speech treatment.

An Internet version of the treatment—iGlebe—has been developed. Several reports have provided evidence for its capacity to eliminate social anxiety disorder diagnoses for adults who stutter (Helgadóttir et al., 2009, 2014a) and to improve responses on measures that assess unhelpful thoughts, avoidance behaviors, fear of negative evaluation, depression, anxiety, stress, and overall quality of life (Menzies et al., 2016; Menzies, O'Brian, et al., 2019; Menzies, Packman, et al., 2019). There is also emerging evidence that the iGlebe program is associated with improved self-reported stuttering severity (Menzies, O'Brian, et al., 2019; Menzies, Packman, et al., 2019).

iBroadway, a modified version of the iGlebe treatment program for adolescents, has also provided encouraging results (Gunn et al., 2019). This report associated the iBroadway program with reduced mental health diagnoses, including social anxiety disorder, and improvement on scores of psychological measures of anxiety, for adolescents who stutter.

Cognitive Models of Social Anxiety Disorder

Two decades after Van Riper (1973) detailed his treatment procedures, cognitive models of social anxiety disorder were developed in an attempt to explain how anxiety is maintained with the condition (Clark & Wells, 1995; Hirsch & Clark, 2004; Rapee & Heimberg, 1997). These models are driven by the need to explain why those with social anxiety disorder fail to unlearn fear, regardless of the many social encounters that do not result in negative consequences. One such prominent model (Clark & Wells, 1995) includes a central component of “processing of the self as a social object” (Clark, 2001, p. 72). When a person enters a feared social situation, anxiety symptoms, such as increased heart and respiration rate, confirm preexisting negative expectations that the situation is dangerous and to be feared. This exaggerated interpretation of danger within a social situation triggers a focus on the physiological arousal associated with anxiety symptoms, along with the negative thoughts and dysfunctional beliefs connected with the social situation. The model proposed by Clark and Wells (1995) informed the development of much of the existing gold standard CBT procedures as described above. In what follows, we explore the components of the Clark and Wells model of social anxiety disorder as it pertains to stuttering.

Self-Focused Attention, Social Anxiety, and Stuttering

According to the Clark and Wells (1995) cognitive model of social anxiety disorder, entering a feared situation triggers a shift toward what is referred to as self-focused attention. This self-focused attention involves an excessive and maladaptive preoccupation toward negative thoughts and physiological symptoms associated with anxiety (Clark & Wells, 1995; Ingram, 1990). Cognitive theories have been supported by evidence clearly demonstrating that self-focused attention plays a role in the maintenance of anxiety with social anxiety disorder (Bögels & Lamers, 2002; Hoffmann, 2000; McManus et al., 2009; Wells & Papageorgiou, 1998; Woody et al., 1997). There are several mechanisms by which self-focused attention can maintain anxiety. An overview of these mechanisms is given below, with reference to studies that have investigated self-focused attention with social anxiety disorder and stuttering.

Self-Focused Attention and Access to Feedback From Social Situations

One prediction of the Clark and Wells (1995) model of self-focused attention that is fundamental to social anxiety disorder is “reduced processing of external social cues when anxious” (Clark, 2001, p. 413). Such social information could include positive feedback that has the potential to disconfirm fears of poor performance and fear of negative social evaluation (Clark, 1999; Clark & Wells, 1995; Ingram, 1990). Maladaptive self-focused attention with social anxiety has been explored extensively, and research has confirmed that self-focused attention is associated with poorer recall of information from the social environment and heightened recall of self-referential information (Daly et al., 1989; D'Argembeau et al., 2006; Hope et al., 1990; Mansell et al., 2003; Mellings & Alden, 2000).

Direct evidence has shown that adults who stutter similarly appear to avoid positive social information during a speaking task (Lowe et al., 2012). In that study, adults who stutter and fluent controls gave a speech to a non–live audience; the audience was prerecorded, and audience members displayed a range of expressions, including showing signs of being interested or uninterested or having neutral expressions. Participants were led to believe that the audience was real and was sitting in another room during the speaking task. Participants who stuttered, compared with fluent controls, looked less at the audience during the speech. Furthermore, a tendency by the participants who stuttered to avoid looking at the interested audience members was associated with increased self-reported anxiety. The avoidance of positive information within social settings for adults who stutter (Lowe et al., 2012) is consistent with findings for social anxiety disorder (Chen et al., 2002; Pishyar et al., 2004, 2008).

Self-Focused Attention and Attentional Biases

The most significant consequence of self-focused attention is that it can result in failure to receive information and feedback from others during social situations. This has been confirmed in laboratory studies with social anxiety disorder using various research methods. One such method to evaluate attention is the probe detection task (MacLeod et al., 1986). For this task, emotionally valenced stimuli are presented on a computer screen. Examples of such stimuli are negative, positive, or neutral words or pictures of faces displaying negative, positive, or neutral expressions. Participants press a button on a device or keyboard following a cue. Participant reaction times are recorded, and latencies are assumed to reflect biases in attentional processing.

The probe detection task has been used extensively in research with social anxiety disorder, demonstrating attentional biases to avoid threatening stimuli (Chen et al., 2002; Mansell et al., 2003, 1999; Pineles & Mineka, 2005). However, Lowe et al. (2016) were unable to replicate the effect using the same design and stimuli with participants who stuttered and were not socially anxious. In contrast, higher trait anxiety was associated with a tendency to look toward threatening stimuli, which is consistent with studies that have investigated attentional biases with general social anxiety (Bradley et al., 1999; Macleod et al., 1986). Interestingly, a similar effect to that reported by Lowe et al. (2016) has recently been found with adolescents who were not socially anxious. Rodgers et al. (2020), using a modified probe detection task, reported that the participants looked toward faces with threatening expressions faster than toward those with neutral expressions and that they looked away from faces with threatening expressions faster than from those with negative expressions. However, McAllister et al. (2015) reported that young people who stutter with higher scores on a social phobia subscale directed attention toward faces with sad expressions. The disparity of results, using tasks that aim to assess attentional biases, highlights nuances with the task design and participant characteristics that have also been demonstrated in the literature about attentional biases and social anxiety.

Self-Focused Attention and the Maintenance of Fears and Negative Beliefs

According to the Clark and Wells (1995) model, when a person with social anxiety enters a feared social situation, anxiety symptoms, such as the heart beating fast, feeling hot and flushed, and increased breath rate, confirm negative thoughts that the situation is dangerous and to be feared. This exaggerated interpretation of danger within a social situation triggers a tendency to focus intently toward the physiological arousal associated with anxiety, increasing the salience of anxiety symptoms. Consequently, the information obtained from that inward focus of attention is recalled during postevent evaluations, confirming negative beliefs and expectations of poor performance (Clark & Wells, 1995). The biased recollection of anxiety symptoms in the absence of accurate and potentially positive information confirms fears and dysfunctional beliefs about danger and risk of negative consequences associated with social situations (Clark & Wells, 1995).

There is evidence to suggest that adults who stutter may focus their attention toward negative thoughts about stuttering and the negative consequences of stuttering (Boyle, 2013b; Boyle & Fearon, 2018; Cream et al., 2003; Jackson et al., 2015; Lowe et al., 2015; St Clare et al., 2009; Tudor et al., 2013) and, more specifically, to the physiological arousal associated with moments of stuttering (Lowe et al., 2015). This is not surprising because negative emotions associated with stuttering have been well documented and include feelings of guilt, shame, embarrassment, fear, helplessness, isolation, and frustration (Bloodstein & Bernstein Ratner, 2008; Corcoran & Stewart, 1998; Hugh-Jones & Smith, 1999; Sheehan et al., 1962; Yaruss & Reardon, 2002). Negative thoughts about stuttering have been comprehensively documented in the context of developing the Unhelpful Thoughts and Beliefs About Stuttering scale (Iverach et al., 2011; St Clare et al., 2009). Examples of negative thoughts include “It's impossible to be really successful in life if you stutter,” “I won't be able to keep a job if I stutter,” “No one will like me if I stutter,” and “People will think I'm boring because I have nothing to say.” Furthermore, Boyle (2013a) developed the Self-Stigma of Stuttering Scale. This scale is designed to assess the extent to which a person agrees with and applies stereotypical attitudes toward themselves. The scale includes items that portray negative thoughts about stuttering: “Because I stutter, I feel less capable than people who don't stutter,” “Most people in the public believe that people who stutter are insecure,” and “I believe that most people who stutter are nervous.”

The potential for self-focused attention to be associated with well-documented negative consequences is important to consider for those who stutter. As with social anxiety disorder, the focus of attention toward negative thoughts and emotions by those who stutter is likely to maintain fears of negative consequences of stuttering and, overall, maintain anxiety.

Self-Focused Attention and Spontaneous Imagery

Clark (2001) also states that a prediction of the Clark and Wells (1995) model is that social anxiety disorder causes those affected to experience “distorted observer-perspective images of how they think they appear to others when in feared social situations” (p. 414). For example, these images depict scenes of the self from the perspective of another. Such scenes convey incorrect, exaggerated content, such as seeing the self with a twisted face, with distorted features, or seeming much smaller than others in the image. This has been demonstrated for those with social anxiety disorder (Conway et al., 2004; Hackmann et al., 1998; Kenny & Bryant, 2007; Wells & Papageorgiou, 1999). The recall of scenes from the perspective of the observer is thought to maintain negative fears and beliefs, as it is this information that is recalled during post-event evaluations rather than accurate feedback from the social interaction (Clark & Wells, 1995). Furthermore, such scenes feature prominently during anticipatory processing of social encounters (Clark & Wells, 1995). Combined with faulty predictions of poor performance during social situations, faulty negatively valenced imagery can lead to negative consequences and maladaptive behaviors. For instance, negative anticipatory processing of impending social events may lead to avoidance of the situation. However, if a situation cannot be avoided, anticipatory processing may induce self-focused attention during the event, thereby perpetuating inaccurate processing of the encounter (Clark & Wells, 1995). The cycle of anxiety maintenance is therefore reinforced.

In the case of stuttering, two studies have investigated the recall of memories with adults. Tudor et al. (2013) reported more recall by stuttering participants of recurrent, intrusive, and negative imagery during speaking situations than by controls. Lowe et al. (2015) reported that, when comparing participants who stutter with a control group, the stuttering group was more likely to recall an observer perspective when recalling social events. Furthermore, the participants who stutter recalled images of how they thought they appeared to others rather than memories of images of what, in reality, was within their visual fields during the interaction. For those who stutter, recalling negatively valenced observer-perspective images is likely to maintain fears of poor performance during social interactions and maintain fear of negative evaluation from others. In the same way that negative observer-perspective recall of memories with social anxiety can maintain anxiety, such biased recall of memories is likely to maintain anxiety with those who stutter. However, this clearly requires further investigation.

Safety Behaviors, Social Anxiety, and Stuttering

Cognitive models of social anxiety disorder incorporate the use of safety behaviors and suggest they maintain anxiety (Clark & Wells, 1995). For example, perceived social danger prompts those affected by social anxiety to engage in behaviors or rituals to minimize the occurrence of feared outcomes (Salkovskis, 1991). With social anxiety disorder, a common safety behavior is avoidance, for example, avoiding eye contact with others, avoiding situations, or talking to specific people (Clark & Wells, 1995; Marks, 1969; Ohman, 1986; Wells & Clark, 1997; Wells et al., 1995). As with self-focused attention, using safety behaviors has been demonstrated to contribute to the maintenance of anxiety with social anxiety disorder (Salkovskis, 1991; Wells & Clark, 1997; Wells et al., 1995).

It has been suggested that safety behaviors maintain anxiety through several mechanisms. For instance, when safety behaviors are used to reduce anxiety, if the feared outcome does not occur, the person attributes this to the use of the safety behavior. The safety behavior is then repeated in future situations to reduce anxiety, and the cycle of anxiety maintenance continues (Salkovskis, 1991; Wells & Clark, 1997; Wells et al., 1995). For example, a person may avoid talking during meetings at work due to fear of negative evaluation from others. This avoidance reduces anxiety in the immediate situation; however, the same fears continue in future situations, and the person therefore continues to use avoidance strategies to manage the anxiety.

Using safety behaviors can tax attentional resources, which can, in turn, affect performance during social encounters, such as engaging with others. Perceptions of poor performance confirm fears and beliefs, which, in turn, maintain anxiety (Clark & Wells, 1995).

The use of safety behaviors can increase the likelihood of the feared outcome occurring. For example, avoiding eye contact can decrease opportunities to engage in the social encounter, increasing the risk of negative self-perceptions of performance. Additionally, avoiding eye contact can lead to the perception by others of reticence to engage in the social encounter, leading to negative evaluation.

However, the most significant problem with using safety behaviors is that fears are not challenged and, therefore, persist (Lovibond et al., 2009; Salkovskis, 1991; Wells & Clark, 1997). In other words, safety behaviors prevent new learning that fears are irrational and beliefs are inaccurate.

Considering the range of coping behaviors associated with stuttering, many could be considered safety behaviors. For example, there is ample documentation for those who stutter to avoid topics, specific words, and eye contact (Corcoran & Stewart, 1998; Cream et al., 2003; Crichton-Smith, 2002; Hayhow et al., 2002; Jackson et al., 2015; Kraaimaat et al., 2002; Lowe et al., 2012; Mahr & Torosian, 1999; Martens & Engel, 1986; Plexico et al., 2005; Vanryckeghem et al., 2004; Yaruss & Reardon, 2002). Situation avoidance is common among those who stutter. Many who stutter recall that, as children, they avoided situations such as participating in class discussions, answering questions, reading aloud, and seeking help from teachers (Hugh-Jones & Smith, 1999; Klompas & Ross, 2004; Plexico et al., 2005; Silverman & Zimmer, 1982). Furthermore, extreme types of avoidance include truanting, and withdrawing from school and not undertaking further education in order to avoid talking are commonly reported (Crichton-Smith, 2002; Hugh-Jones & Smith, 1999; O'Brian et al., 2011).

There is emerging evidence that, indeed, those who stutter engage in safety-seeking behaviors to avoid feared outcomes. In a recent survey, adults who stutter who enrolled in the iGlebe program (see above), for treatment of social anxiety, reported engaging in behaviors that are considered to be safety behaviors (Lowe et al., 2017). Such behaviors included to “mentally rehearse sentences,” “keep answers short,” and “avoid difficult words.” Cream et al. (2003) reported that adults who stutter rehearse sentences and responses and that they manipulate situations to avoid the unpredictability of being asked questions. Systemic functional linguistic analyses have shown that adults who stutter alter their use of language to limit conversational engagement with others (Lee et al., 2015; Spencer et al., 2005, 2009). Lee et al. (2015) suggest this may be a coping strategy used to avoid potential negative events associated with interacting with others. Further evidence of the use of safety behaviors by adults who stutter is demonstrated by the anticipation of stuttering. The anticipation of stuttering has been shown to result in the use of strategies such as avoidance or changing the content of speech and language, such as substituting words anticipated to be stuttered (Jackson et al., 2015; Vanryckeghem et al., 2004).

An Emerging Clinical and Theoretical Conflict

The field is well under way with treatments for managing persistent stuttering and social anxiety associated with stuttering, and there are encouraging signs that such developments may be translatable and scalable with Internet-based versions (Erickson et al., 2016; Menzies et al., 2016; Van Eerdenbrugh et al., 2018). Concurrent with the more recent development of formalized CBT procedures for social anxiety for those who stutter, research has provided some insights about the underpinning processes that may sustain social anxiety with those who stutter. These processes are self-focused attention and safety behaviors.

In the following sections, we develop the idea that these recent advances raise clinical and theoretical questions about current clinical management of stuttering. We propose that speech restructuring and psychological approaches for stuttering may, in fact, be in direct conflict with one another. This conflict may provide an explanation for the failure of many adults who stutter to maintain their speech treatment benefits. Accordingly, we will conclude with recommendations for interim clinical management of stuttering and research that may contribute to the resolution of this conflict.

Self-Focused Attention and Speech Restructuring Treatment

We speculate that self-focused attention, as discussed earlier, could affect speech performance, even in the absence of techniques to manage stuttering. Adults who stutter, having dealt with stuttering and its consequences throughout life, may have a stronger tendency for self-focused attention than those with social anxiety disorder in general. Anticipating stuttering, monitoring speech, and modifying spoken language in an attempt to avoid stuttering—as discussed earlier—are examples where attention is focused inwardly at the expense of attending to and participating in interactions with others.

Predictably though, self-focused attention is likely to have detrimental effects on the ability to implement speech restructuring techniques. Maladaptive attentional focus toward unhelpful thoughts and beliefs, or other salient symptoms of anxiety, can tax attentional resources available to implement speech restructuring techniques. Therefore, excessive and maladaptive self-focused attention may be one explanation for the Iverach, Jones, et al. (2009a) findings that those participants with mental health and anxiety-related disorders failed to maintain their speech treatment outcomes.

Conversely, another possible explanation for posttreatment relapse rates for particularly anxious participants is that, although speech restructuring can provide control of stuttered speech, it has the potential of also enhancing self-focused attention. Clients who have received speech restructuring treatment need to focus inwardly, in situations where stuttering control is desired, with constant attention to their use of the novel speech pattern (Block et al., 2005; Curlee, 1993; Howie et al., 1981; Langevin et al., 2010). Additionally, during many treatment approaches, clients are taught to monitor and evaluate their stuttering severity to assess change over time during treatment and to monitor speech over time. This self-monitoring may occur at the expense of attending to the social interaction and receiving accurate feedback from others. In the presence of anxiety and associated negative thoughts and beliefs, lack of corrective feedback from attending to the interaction can lead to the maintenance of maladaptive thoughts and beliefs and overall persistence of anxiety.

Additionally, adults who stutter feel different from fluent speakers, and, even after behavioral speech treatment, that perception remains. Before treatment, stuttering is the point of difference; however, following treatment with speech restructuring, it is the fluency technique that discerns them from fluent speakers (Cream et al., 2003; Finn & Ingham, 1994; Metz et al., 1990). This is an important issue in the context of fear of negative evaluation from others that adults who stutter experience (Cream et al., 2003). Furthermore, fear of negative evaluation about how others will perceive the sound of a novel fluency technique is likely to tax attentional resources needed to use the technique. Consequently, this fear of negative evaluation may provide an explanation for the high failure rates to maintain speech treatment outcomes. Those who stutter may be reticent to use a technique to minimize their stuttering due to fear of negative evaluation about how the technique sounds.

Safety Behaviors and Speech Restructuring Treatment

It is well established that safety behaviors can maintain anxiety for those with anxiety disorders (Helbig-Lang & Petermann, 2010). For those who stutter, the use of safety behaviors is also likely to maintain anxiety. In turn, persistent anxiety may lead to a reticence to use speech restructuring techniques. Safety behaviors can impact on the capacity to be present and engaged in social encounters and can increase the risk of negative self-perception of social performance (Clark & Wells, 1995). As proposed by Clark and Wells (1995), biased perceptions of performance can confirm fears and maintain anxiety. Furthermore, using safety behaviors can impact on social performance, thereby increasing the risk of negative evaluation from others, which could confirm fears of poor performance, hence maintaining anxiety (Clark & Wells, 1995). As with anxiety disorders, the use of safety behaviors by those who stutter, in an attempt to avoid negative consequences, could prevent fears being challenged and, hence, could maintain anxiety. Additionally, safety behavior use by those who stutter could have the consequence of affecting speech performance and, more specifically, maintain or exacerbate stuttering.

As has been established as a theme for this review article, speech treatment procedures may be in conflict with psychology theory about safety behavior use. Inadvertently, in the context of speech treatment, speech-language pathologists (SLPs) may recommend that their clients use safety behaviors. This was highlighted in a study by Helgadóttir et al. (2014b), in which 169 Australian SLPs were surveyed. The SLPs were asked to comment on how often they recommended that their clients use certain behaviors. Those behaviors had previously been categorized by clinical psychologists as potential safety-seeking behaviors. The most frequently recommended items on the list included “rehearse opening line of telephone call before making call” and “choose safe or easy people to talk to in socially threatening situations.” Additionally, around half of the respondents reported to recommend suggesting to their clients the following: “If you are feeling anxious, try to avoid difficult syllables,” “If you are feeling anxious, try to avoid difficult words,” and “If you feel like you are having a bad day, skip unnecessary talking.”

However, critical to the thesis of this review article, if speech structuring techniques are sought and used to avoid feared consequences of stuttering, in itself, the treatment could be considered a safety behavior. Indeed, there is evidence that speech restructuring is used as a means of protection from harm (Cream et al., 2003). One possible explanation for the failure of participants to maintain their speech treatment benefits in the Iverach, Jones, et al. (2009a) study is that if speech restructuring is used as a safety behavior, anxiety is maintained, self-focused attention is increased, cognitive resources are reduced, and speech restructuring techniques are unable to be used effectively to control stuttering.

Treatment for stuttering using speech restructuring techniques requires ongoing practice to maintain the skill in order to continue to control stuttering during everyday speaking situations. Many of the safety behaviors from the Helgadóttir et al. (2014b) list were specifically about the use of speech restructuring techniques, such as “practicing speech technique immediately before an important speaking situation.” However, in the context of safety behaviors, if such behaviors are used to moderate anxiety and not merely used to practice the speech motor technique, there is ample reason to foreshadow that such behaviors have the potential to sustain anxiety (Lovibond et al., 2009). Clearly, this matter warrants urgent investigation.

Clinical Applications

Overall, this emerging conflict between speech restructuring procedures and social anxiety needs to be considered in the context of clinical practice. The main issues that have been discussed are that speech restructuring (a) may induce or increase self-focused attention, (b) may promote the use of safety behaviors, and (c) may become a safety behavior itself. Therefore, it appears there is a conflict between speech restructuring treatment and CBT for social anxiety that needs to be dealt with by clinicians during client management. Craig and Hancock (1995) reported that perceived relapse by adults who stutter was associated with high trait anxiety levels, and Iverach, Jones, et al. (2009a) reported that high rates of anxiety and mental health disorders were associated with a failure to maintain treatment benefits. According to these findings, these so-called relapse rates could apply to as many of two thirds of adults who present to clinics for assistance with their stuttering.

For clients with clinically significant social anxiety, speech restructuring used to control stuttering may come at a cost. Speech restructuring techniques could trigger or worsen social anxiety by promoting self-focused attention. In general, focusing on implementing speech techniques is likely to impact on attentional resources available to engage fully in the interaction. Furthermore, postevent processing of the situation is therefore likely to draw on self-perceptions of performance due to lack of accurate information from the situation, as has been proposed with social anxiety disorder (Clark & Wells, 1995). In addition, when speech restructuring techniques are used as a safety behavior, they may prevent the extinction of fear and anxiety for those who stutter. That cost may be increased if the clinical process causes both effects: triggering or worsening social anxiety while concurrently preventing its extinction. Therefore, based on the Iverach, Jones, et al. (2009a) report, speech restructuring may be contraindicated in the presence of social anxiety and may perpetuate the “revolving door” effect with client health care for stuttering. If a client's perceived need to control stuttering is driven by social anxiety, the use of speech restructuring to attain that control may, in effect, maintain their anxiety and, in turn, prevent effective stuttering control. In short, speech restructuring for such clients could potentially be more clinically harmful than helpful. As such, in the event that a client is experiencing social anxiety and could concurrently benefit from control of stuttering, the SLP has a challenging situation to deal with. At present, there is little empirical guidance about this matter, and, as discussed below, more is needed urgently.

Dealing With This Clinical Conflict

The first stage of dealing with this clinical issue involves assessment. The assessment process clearly needs to consider both speech and psychological domains. Drawing on Baer's (1988, 1990) accepted statement, central to the assessment and treatment process are the complaints of those seeking behavioral treatments. Therefore, if a client is seeking clinical help to minimize their stuttering, to some extent, this should be considered during the clinical interview with the knowledge that there is a sound clinical trial evidence base in support of clinical efforts to achieve such a goal.

However, screening for anxiety and, in particular, social anxiety now seems essential for all cases of persistent stuttering, based on recent evidence demonstrating the high rates present within clinical caseloads (Iverach, Jones, et al., 2009a; Menzies et al., 2008; Stein et al., 1996). For adults, existing anxiety screening resources can be recommended. To assess the expectation of negative evaluation from others, the eight- or 10-item Brief Fear of Negative Evaluation scale can be used (Carleton et al., 2007, 2011, 2006; Collins et al., 2005; Duke et al., 2006; Leary, 1983; Rodebaugh, Woods, et al., 2004; Weeks et al., 2005). Clinically, this measure can be used to assess fear of negative evaluation and potential anxiety in order to determine the need for referral to a psychologist. To screen for disorder-specific potential markers of anxiety, the six-item screening version of the Unhelpful Thoughts and Beliefs About Stuttering scale (Iverach, Heard, et al., 2016) can be used. If needed, the full version of the scale (Iverach et al., 2011; St Clare et al., 2009) is useful as a guide for a clinical interview about the topic or as an indicator of the need for a clinical psychology assessment. This process of assessment may be ongoing during treatment as potential indicators of anxiety or failure to make progress with speech treatment may appear.

If, based on screening and clinical interview, the clinician forms a view that clinically significant anxiety is not an issue with an adult seeking to minimize their stuttering, then, potentially, there would appear to be no problem proceeding with current evidence-based treatments aimed at reducing stuttering. For clients with subclinical levels of social anxiety, it is unknown at present if speech restructuring will induce self-focused attention, safety-seeking responses, and associated effects. In the interim, it is advised to proceed with caution during speech restructuring treatment. It may well be that for some clients, for whom stuttering management is important, a trade-off needs to be established between that need and the potential to develop social anxiety. However, if a client presents with a need for stuttering and social anxiety management, what is yet to be established empirically is how best to proceed, for instance, whether speech restructuring or CBT should be introduced first, whether and how they can be combined, or how treatment procedures may be modified to take account of the potential conflicts described above.

As discussed earlier, Van Riper (1973) and several later studies have reported the use of CBT procedures to deal with anxiety for adults and adolescents who stutter (Andrews & Craig, 1982; Andrews & Feyer, 1985; Blood, 1995; Boberg & Kully, 1994; Craig & Andrews, 1985; Howie et al., 1981; Maxwell, 1982). Typically, during the speech treatment, those reports combined various CBT procedures in an attempt to remove fears and reduce avoidance behaviors. Those procedures included desensitization, such as openly stuttering while attempting to reduce avoidance behaviors, exposure tasks, thought stopping, locus of control, and relaxation procedures. As noted by Craig et al. (1987), the CBT procedures used in some of those early studies were a “developing and experimental aspect of the course, and as a result is not as structured as the behavioral change aspects of the programme” (p. 59).

Van Riper's (1973) text provides a detailed account of anxiolytic procedures included within a broader treatment approach. However, as noted in the introduction, therehave been no clinical trials of those treatment approaches. In subsequent studies (Andrews & Craig, 1982; Andrews & Feyer, 1985; Blood, 1995; Boberg & Kully, 1994; Craig & Andrews, 1985; Howie et al., 1981; Maxwell, 1982), the CBT components that featured with speech restructuring procedures were not consistently documented, nor was there documentation of psychological responses to those CBT procedures. Regardless, it is possible that the CBT procedures used may have been incompatible with the speech treatment procedures, explaining the high relapse rate (Craig et al., 1987). However, that is conjecture only, because those early studies did not report standard participant diagnoses of anxiety disorders. Since those studies, it has been established that a diagnosable social anxiety disorder is the predominant anxiety condition associated with stuttering.

At present, the evidence available for the CBT program for social anxiety with stuttering involved administration either before or after speech restructuring treatment (Menzies et al., 2016, 2008). Interestingly, while the CBT program is capable of removing social anxiety disorder diagnoses, there is no evidence for this treatment reducing stuttering severity when measured by blinded observers. However, preliminary evidence is emerging that, following the CBT program, self-reported stuttering severity decreased in the absence of speech treatment (Menzies, Packman, et al., 2019). Furthermore, in another study by the same research group, participants who received the CBT program as a supplement to speech restructuring treatment reported improved self-ratings of stuttering severity in the absence of such an effect for observer ratings of percent syllables stuttered (Menzies, O'Brian, et al., 2019). Notably, the speech restructuring treatment in the Menzies et al. (2008) and Menzies, O'Brian, et al. (2019) studies was not complete, as all fluency technique practice regimes, hierarchical transfer tasks, and anxiolytic treatment components had been removed. Consequently, as previously noted by Menzies et al. (2008), the results of such studies cannot be compared to speech outcomes from reports of complete speech restructuring treatments.

These issues require further exploration in future studies, but one interpretation so far is that CBT may improve perceptions of performance and, specifically, perceptions of the cost of stuttering, irrespective of observer evaluation. While this can be viewed positively in light of CBT outcomes, the discussion earlier about the potential conflict between speech restructuring and CBT approaches requires consideration. It may be that speech restructuring procedures could be dampening potential CBT effects beyond that reported in preliminary studies, and, potentially, CBT procedures may be clouding speech restructuring outcomes.

While the studies investigating CBT for social anxiety with those who stutter are preliminary, several considerations are necessary for future research, including sample size and the presence of psychological diagnoses, which may impact treatment outcomes. For instance, there has been an emergence of research reporting on the presence of mental health conditions associated with adult stuttering, including anxiety and mood disorders (Iverach, Jones, et al., 2009b; Iverach et al., 2010; Tran et al., 2011). Clearly, further research is urgently needed.

An Essential Research Plan

It is essential to empirically clarify issues raised in this review article about the potential conflict between speech restructuring treatment approaches for stuttering and social anxiety associated with stuttering. Clinically, it will be essential to determine if speech restructuring procedures induce potential maladaptive information processing biases such as self-focused attention and attentional biases, which can maintain anxiety. In the event that this is determined, speech restructuring treatment approaches would need to be modified. One procedure may be to ensure that those who are treated with speech restructuring direct their attention more adaptively toward cues and stimuli in the social environment that can provide accurate and potentially disconfirming feedback about fears. It could be justifiably argued that those who stutter do experience negative consequences as a result of stuttering. However, it is common for those who are socially anxious to overestimate the probability and cost of negative outcomes (Brundage et al., 2017). Typically, strategies taught during CBT involve teaching clients to identify maladaptive cost and probability estimates and modifying these appraisals to more realistic and likely levels. This approach is detailed in Menzies et al. (2009).

It is now reasonable to consider which behaviors those with stuttering might engage with that have the potential to be safety behaviors (Helgadóttir et al., 2014b; Lowe et al., 2017). However, in order to be confirmed as safety behaviors, they require empirical confirmation of their capacity to prevent fear extinction and maintain anxiety. At the outset, such a research plan is complicated by the possibility that not all safety behaviors will be clinically unhelpful to the treatment process (Rachman et al., 2008; Thwaites & Freeston, 2005). The initial and judicious use of safety behaviors may, in fact, assist to engage with treatment and complete necessary treatment tasks, as has been shown with anxiety and phobia disorders (Hood et al., 2010; Milosevic & Radomsky, 2008, 2013). For instance, in the short term, safety behaviors may assist clients to complete practice and exposure tasks, and that may prove to be clinically beneficial in the long term. For example, adults who stutter may, during the early stages of treatment when learning to use the speech restructuring technique, initially choose safe or easy people to talk with in socially threatening situations. However, as proficiency using the fluency technique improves, this strategy of avoiding talking with people who are not considered safe or easy to talk to could become a safety-seeking behavior. For instance, if one always avoids talking to people not considered safe, to avoid negative outcomes, there will be no opportunity to challenge fears and, importantly, it is likely to maintain anxiety. A consequence of only speaking to safe and easy people would profoundly limit the potential to meet new people that could lead to lasting friendships, fulfilling relationships, and life experiences. However, in the first instance, it will be important to explore whether such behaviors maintain anxiety. An experimental laboratory protocol could determine whether potential safety behaviors used by those who stutter do prevent fear extinction.

In any event, it is unknown whether those with persistent stuttering routinely use safety behaviors. This issue could be explored initially with a laboratory protocol involving a speech to strangers, which is a situation often feared by those who stutter (Trotter & Bergmann, 1957; Vanryckeghem et al., 2017). Participants could identify particular behaviors or rituals they use in such a scenario and then be instructed to use or not use the behavior. Furthermore, the use of safety behaviors during a speech could be regressed using measures that would include those that are translatable to clinical practice in speech-language pathology, such as stuttering severity, the Brief Fear of Negative Evaluation scale, the Unhelpful Thoughts and Beliefs About Stuttering scale, and the Self-Stigma of Stuttering Scale. This would enable the exploration of constructs known to be related to safety behaviors, such as anxiety, unhelpful thoughts, and disproportionate fear.

To overcome the potential issue of speech restructuring being a safety behavior, laboratory studies could draw on psychological theory and systematically include exposure tasks while participants learn speech restructuring. An example of such a study could include exposure while learning speech restructuring, alternating between using and not using the technique during speaking tasks. They would then evaluate the outcomes of using and not using the speech restructuring technique in terms of probability and cost of negative evaluation from others about stuttering and use of the speech restructuring technique. Furthermore, such exposure and cognitive restructuring exercises have the potential to enable clients to evaluate the cost and consequences of stuttering more realistically when their ability to use speech restructuring is not as effective, such as when fatigued or unwell.

Such a program of research, as outlined above, should include participants from as young as the early adolescent years. As with every clinical advance with this disorder, the earlier that treatment can be administered during the course of the disorder, the better.

Conclusions

The issues raised in this review article are complex. Many adults who stutter will present to speech clinics and experience social anxiety at levels that warrant a diagnosis of social anxiety disorder. Cognitive theories have confirmed, by empirical research, that several processes are involved in the maintenance of anxiety for social anxiety disorder. These processes form the basis of well-established CBT procedures for anxiety disorders. However, there is emerging evidence that the same processes that are associated with anxiety maintenance for social anxiety disorder appear to also be used by those who stutter. It appears that evidence-based speech treatment procedures that target control of stuttering are in conflict with treatment procedures that deal with social anxiety and can potentially induce or exacerbate those processes involved in anxiety maintenance.

This review article proposes suggestions for future research to investigate this potential conflict. Such research should explore the potential for speech restructuring treatment to increase self-focused attention or to become a safety behavior. The results may inform the direction of research efforts toward modifying speech restructuring treatment approaches or to psychological treatment approaches that may ameliorate these effects. Such research efforts may lead to the development of optimal speech and psychological treatments for those who stutter that have long-term treatment and economic benefits. In the interim, it is recommended that clinicians who treat persistent stuttering continually monitor their clients for anxiety and consider, in consultation with their clients, the most personally presenting concern: managing speech behaviors or anxiety management.

Funding Statement

This research was supported by National Health and Medical Research Council Program Grant 1132370.