0196-206X/06/2704-0341

Developmental and Behavioral Pediatrics Vol. 27, No. 4, August 2006

Copyright * 2006 by Lippincott Williams & Wilkins, Inc. Printed in U.S.A.

Review Article

Suffering in Silence: Why a Developmental

Psychopathology Perspective on
Selective Mutism Is Needed
SHARON L. COHAN, M.S.
Department of Psychology, San Diego State University, San Diego, CA
Department of Psychiatry, University of California San Diego, La Jolla, CA
JOSEPH M. PRICE, PH.D.
Department of Psychology, San Diego State University, San Diego, CA
MURRAY B. STEIN, M.D., M.P.H.
Department of Psychology, San Diego State University, San Diego, CA
Department of Psychiatry, University of California San Diego, La Jolla, CA
VA San Diego Healthcare System, San Diego, CA

ABSTRACT. A developmental psychopathology perspective is offered in an effort to organize the existing

literature regarding the etiology of selective mutism (SM), a relatively rare disorder in which a child consistently
fails to speak in 1 or more social settings (e.g., school) despite speaking normally in other settings (e.g., home).
Following a brief description of the history, prevalence, and course of the disorder, multiple pathways to the
development of SM are discussed, with a focus on the various genetic, temperamental, psychological, and
social/environmental systems that may be important in conceptualizing this unusual childhood disorder. The
authors propose that SM develops due to a series of complex interactions among the various systems
reviewed (e.g., a strong genetic loading for anxiety interacts with an existing communication disorder, resulting
in heightened sensitivity to verbal interactions and mutism in some settings). Suggestions are provided for
future longitudinal, twin/adoption, molecular genetic, and neuroimaging studies that would be particularly
helpful in testing the pathways perspective on SM. J Dev Behav Pediatr 27:341 Y355, 2006. Index terms:
selective mutism; elective mutism; developmental pathways; developmental psychopathology.

The etiology of selective mutism (SM), a relatively rare or on treatment of children with the disorder.4Y6 Because
disorder in which a child does not speak in 1 or more several competing theories have been put forward to
social settings despite having developed appropriate explain the development of SM (i.e., psychodynamic
language skills, has been debated since the syndrome was conflicts, anxiety/behavioral inhibition, and developmental
first identified in the late 19th century. A number of causes delays), a developmental psychopathology perspective
have been hypothesized, including exposure to childhood may be particularly helpful in organizing what is currently
trauma, oppositional behaviors, developmental disorder, known about the disorder. This perspective allows us to
and extreme shyness. Although SM is considered to be a form a cohesive theory of SM based on the extant lit-
condition first diagnosed in childhood, few researchers or erature and to identify areas in need of further research. By
theorists have taken a developmental perspective on the alerting clinicians to the various risk factors and develop-
disorder. Two recent review articles on the topic did not mental pathways for the disorder, this framework may also
explicitly consider developmental issues,1,2 whereas others aid in the identification and treatment of children at risk
focused on similarities between SM and anxiety disorders3 for SM.

METHOD
Address for reprints: Sharon L. Cohan, M.S., Anxiety and Traumatic
Stress Disorders Research Program, University of California San Diego, The present literature review was undertaken to provide a
8950 Villa La Jolla Drive, Suite C207, La Jolla, CA 92037; e-mail: developmental psychopathology perspective regarding the
scohan@ucsd.edu. etiology of SM. Following the procedure for integrated

341

Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
342 COHAN ET AL JDBP/August, Vol. 27, No. 4

literature reviews,7 we searched PubMed and PsychINFO periods;16 therefore, SM should not be diagnosed within
(1840 to current) databases using the key words ‘‘selective the first month of school. Although some level of shyness
mutism’’ and ‘‘elective mutism.’’ We limited our review and reticence in novel social situations is to be expected,
to articles published in peer-reviewed journals. Clinical the clinical syndrome of SM is not diagnosed unless there
case studies, book chapters, and articles published in is a consistent failure to speak in certain settings that
languages other than English were excluded. References significantly interferes with the child’s psychosocial
of the articles found through PubMed and PsychINFO functioning or social communication.
were also searched to identify additional relevant citations. There have been very few community-based studies of
Seventeen articles were used to provide background SM, so the true prevalence of the disorder remains
information regarding the history, phenomenology, and uncertain. One recent epidemiological study found a
treatment of SM. An additional 22 were descriptive or prevalence rate of 1.9% in a sample of Finnish second
case-control studies of children with SM. Details regarding graders.17 However, this finding was based on DSM-III-R
these studies are presented in Table 1. criteria that did not require evidence of impairment in
psychosocial functioning; lower rates have been found
using the current diagnostic criteria. A more recent study
HISTORY, DIAGNOSIS, AND PREVALENCE OF
conducted in the United States found a prevalence rate of
SELECTIVE MUTISM 0.71% for DSM-IV SM among kindergarten and first and
Accounts of the history of SM note that the syndrome second graders in a large urban school district,18 suggest-
was first identified in 1877 by German physician Adolf ing that the range in prevalence estimates may be due to
Kussmaul, who used the term aphasia voluntaria to the use of different criteria for defining SM in each study.
describe children who voluntarily refused to speak despite In support of this notion, another recent community study
having normal speech and language abilities.1,8 Several using DSM-IV criteria found a nearly identical prevalence
other terms have been used to describe this pattern of rate of 0.76% among Israeli preschoolers.19
behavior, including suppressed speech, voluntary mutism, Prevalence estimates may also vary based on the age
speech avoidance, and speech phobia.9 In 1934, Swiss range included in the sample. For example, a rate of only
psychiatrist Moritz Tramer named the condition elective 0.18% was reported in a study of children between the
mutism, reflecting his belief that children with the ages of 7 and 15 from 2 school districts in Sweden.15
condition were electing not to speak in certain situa- Higher estimates have been found in samples of younger
tions.1,2 This name was retained in the Diagnostic and children, suggesting that SM is more common or more
Statistical Manual of Mental Disorders (DSM-III) and frequently identified in early to middle childhood. This
International Statistical Classification of Diseases (ICD- may be due to entry into the school system, at which time
10).10,11 However, the name was changed to ‘‘selective there is a greater expectation for children to speak and to
mutism’’ in more recent revisions of the DSM (DSM- develop social relationships with their teachers and peers.
IV),12,13 reflecting many contemporary beliefs that the School has been identified as the situation in which
disorder is not primarily due to oppositional behavior. children first show signs of impairment because of SM.2
Current diagnostic criteria in both the DSM-IV-TR13 Selective mutism appears to be slightly more common
and ICD-1011 include (a) consistent failure to speak in among girls, but these findings are still mixed. A study of
specific social situations in which there is an expectation 100 German and Swiss children meeting criteria for ICD-
for speaking (e.g., at school) despite speaking in other 10 elective mutism found a male-female sex ratio of
situations (e.g., at home); (b) the disturbance interferes 1:1.6.20 Similar sex ratios have been reported,21Y23 but one
with educational or occupational achievement or with study found a slight preponderance of boys with the
social communication; (c) the disturbance lasts at least disorder,24 whereas others have reported equal sex
1 month; and (d) the disturbance is not better accounted for ratios.18,19 The elevated rates of boys with SM found in
by a communication disorder or by a lack of knowledge of, clinical samples may reflect a bias in referrals, with
or comfort with, the spoken language required in the social mutism being seen as more problematic in boys relative
situation. According to DSM-IV-TR, associated features to girls. There has also been some suggestion that sex
can include ‘‘excessive shyness, fear of social embarrass- differences become more apparent among older children
ment, social isolation and withdrawal, clinging, compul- with the disorder.10,13
sive traits, negativism, temper tantrums, or controlling or
oppositional behavior, particularly at home.’’13(p126)
COURSE OF SELECTIVE MUTISM
It is important to note that, by approximately 5 years of
age, most children are expected to have developed normal Results from the only recent follow-up study of children
language proficiency14 and to initiate conversations with with SM provide preliminary information regarding the
others. At this stage, a range of behaviors from shyness stability of SM over time. Remschmidt et al25 collected
and reticence in social settings to gregariousness may be follow-up data, an average of 12 years after referral,
observed in normally developing children. For example, regarding a sample of clinically referred patients with
the rate of shyness or reticence was 89 in 10,000 in school- ICD-9 and DSM-III-R elective mutism. On average, age of
based screening of children aged 7 to 15 years, whereas the onset was 3 years, and age at referral was approximately
rate of typical SM was 18 in 10,000.15 Others have noted 8 years. At follow-up, 12% of these patients still met
that transient mutism may be common during transitional diagnostic criteria for the disorder, 20% experienced mild

Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 Table 1. SM Studies Included in the Present Review

Sex of
Age of SM SM Comparison Study design Comorbid diagnoses and
Author (Year) sample No. participants sample Referral source group and method related results
Andersson and Mean, 37 SM (ICD-10) 46% Inpatient and Children with Retrospective Atypical Asperger, obsessive-compulsive
Thomsen 9.43 T 3.8 yr 37 Psychiatric Female outpatient emotional record review disorder, SP, specific developmental
(1998)24 controls psychiatric hospital disorders disorder
Bergman et al Kindergarten to 12 SM (DSM-IV) 50% Nominated by Children without Cross-sectional SM children more symptomatic on measures
(2002)18 second grade 12 Community Female teachers SM teacher report of social anxiety and TRF internalizing,
Controls with 6-mo withdrawn, anxious depressed subscales
follow-up
Black and Uhde Mean, 30 SM 70% Nominated by school None Cross-sectional SP, avoidant disorder, simple phobia,
(1995)21 8.4 T 2.0 yr (DSM-III-R) Female counselors parent and oppositional defiant disorder, separation
teacher report, anxiety, overanxious disorder, enuresis,
structured encopresis, obsessive compulsive disorder,
interview tic disorder
Cunningham Mean, 52 SM (DSM-IV) 60% Assessment center Children without Cross-sectional SM children rated as more anxious and
et al (2004)22 7.16 T 1.88 yr 52 Community Female SM parent report, obsessive and had a greater number of
Controls structured somatic complaints; SM children also
interview showed deficits in social skills relative to
controls
Dummit et al Mean, 50 SM 72% SM advocacy group None Cross-sectional Social phobia, avoidant disorder, separation
(1997)8 8.2 T 2.7 yr (DSM-III-R; Female parent report, anxiety, overanxious disorder, simple
upon review, structured phobia, attention-deficit/hyperactivity
all met criteria interview and disorder, oppositional defiant disorder,
for DSM-IV) clinician rating trichotillomania, enuresis, encopresis,
scales unspecified communication disorders
Elizur and Mean, 4.6 yr 19 SM (DSM-IV) 47% Nominated by Children without Cross-sectional SM children elevated on ratings of social
Perednik 19 Community Female preschool teachers SM parent report anxiety/phobia, behavior problems,
(2003)19 Controls neurodevelopmental delay, and speech
Developmental Psychopathology of SM

problems relative to controls; 32% of SM

children scored in clinically significant
range for social anxiety/phobia
Ford et al Mean, 11.2 yr 153 SM 68% SM advocacy group None Retrospective SM children showed anxiety symptoms on
(1998)40 (DSM-III-R) Female self-report and YSR and CBCL; those who no longer had
parent report SM experienced problems conversing and
felt discomfort in social settings; some
evidence that SM individuals were inhibited
as infants
Kopp and Mean, 10.6 yr 5 SM (DSM-IV) 60% Nominated by None Cross-sectional Comorbid conditions not reported
Gillberg Female teachers teacher report,

Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
(1997)15 neuropsychiatric
examination on
2 children
(Continued on next page)
343
 Table 1. (Continued)
344

Age of SM Sex of SM Comparison Study design Comorbid diagnoses and

Author (Year) sample No. participants sample Referral source group and method related results
Kolvin and Mean, 6.83 yr 24 SM (ICD-10) 54% Pediatric, speech, Children with Record review SM children showed more speech
Fundudis 84 Controls Female child psychiatry, speech delays delays and current disturbances,
(1981)23 with speech and child guidance and normal higher rates of behavior problems,
delays 102 clinics children and elimination disorders; family
Normal history data indicated higher
Controls rates of psychiatric disorders in
SM families
Kristensen Mean, 54 SM 59% School psychology Nonreferred Cross-sectional SM children showed higher rates of
(2000; 2001, 9.0 T 3.4 yr (DSM-IV) 108 Female service and children therapist, communication disorders (expressive
2002);29,51 Community outpatient without SM parent, and language, mixed expressive-receptive
Kristensen controls psychiatry clinics teacher report; language, phonological disorder,
and Torgersen structured stuttering), developmental disorders
(2001, interview; (Asperger, developmental coordination
2002)33,47 children disorder, mild mental retardation),
completed elimination disorders (enuresis,
cognitive/ encopresis), and anxiety disorders
language (separation anxiety, social phobia,
assessments generalized anxiety disorder,
obsessive-compulsive disorder)
Kumpulainen Second grade 39 SM 61% Nominated by None Cross-sectional 34% of SM children had under-
et al (1998)17 (range, (DSM-III-R) Female teachers teacher-report developed language skills; majority
7Y10 yr) of SM children rated by teachers
as shy, withdrawn, serious
COHAN ET AL

Manassis et al Mean, 14 SM (DSM-IV) 64% Outpatient anxiety Children Cross-sectional Rates of comorbid diagnoses did not
(2003)27 10.07 T 2.37 yr 9 Socially Female disorders clinic with SP parent and differ between SM and SP groups;
phobic teacher report, SM children showed greater
controls clinician ratings, language impairments than
structured SP children
interview,
nonverbal
cognitive and
speech/language
assessments
McInnes et al Mean, 7 SM (DSM-IV) 7 Not Outpatient anxiety Children Cross-sectional In narrative language assessment,
(2004);52 9.7 T 1.8 yr Socially reported disorders clinic with SP verbal narrative SM children produced significantly
subset of phobic language shorter narratives than SP children
Manassis et al controls assessments
(2003)27
Remschmidt Mean, 8.7 T 45 SM (ICD-10) 49% Inpatient and Normal Longitudinal Complete remission in 39%, partial

Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
et al (2001)25 3.6 yr at Female outpatient reference (12-yr follow-up) remission in 29%, mild improvement
referral psychiatry clinic, group for self-, physician, in 20%, and no change in 12%
child guidance published and parent report, of SM children; in comparison to
clinic inventories interview, psycho- normative scores, SM group rated
pathological as less independent, less
assessment, and motivated, less self-confident, and
JDBP/August, Vol. 27, No. 4

biographic inventory less mature

 Steinhausen and Mean, 38 SM 68% Inpatient and Children with Family history Significantly more taciturnity and higher
Adamek 8.98 T 2.55 (ICD-10) 31 Female outpatient emotional study, parent rates of psychiatric disorders in SM
(1997) Controls with psychiatry disorder plus interview families; SM found almost exclusively
‘‘behaviorally clinic, self-help articulation/ in relatives of SM children (nonsignificant
similar group expressive result); no difference in rates of speech/
phenotype’’ language language disorders
disorders
Steinhausen and Mean, 100 SM 61% Inpatient and None Cross-sectional Evidence of prenatal and neonatal
Juzi (1996)20 8.73 T 3.76 yr (ICD-10) Female outpatient and retrospec- complications, delayed toileting and
psychiatry tive record motor development, communication
clinics, self- review, clinician disorders in 38% of children
help group ratings of (articulation and expressive language
personal and disorders, receptive language
sociodemo- disorders, stuttering, cluttering),
graphic data, emotional disorders (separation
behavioral anxiety, sleep disorders, eating
abnormalities, disorders, oppositional defiance,
and develop- hyperactivity, elimination disorders,
mental history, and stereotypies), and relationship
direct inter- problems; 85% of sample rated as
view, parent shy, 66% anxious, and 33%
report depressed
Vecchio and Mean, 15 SM (DSM-IV) 53% Public and Children with Cross-sectional, All SM children diagnosed with SP,
Kearney 6.58 T 1.89 yr 15 Anxious Female private anxiety child, parent, 53% also had an additional anxiety
(2005)30 (entire sample) controls 15 (entire schools, child disorders and and teacher disorder (separation anxiety,
Community sample) and family community report, generalized anxiety, specific phobia);
Controls services controls structured SM and anxiety disordered controls
interview showed significantly higher
internalizing symptoms relative to
community controls
Wilkins (1985)16 Mean, 9.8 yr 24 SM (ICD-10) 71% Children’s Children with Retrospective SM children showed higher rates of
24 Controls Female hospital emotional record review delayed speech and articulation
with emotional disorders difficulties; SM children more
Developmental Psychopathology of SM

disorders frequently described as anxious,

depressed, and manipulative
Yeganeh et al Mean, 23 SM (DSM-IV) 57% Anxiety research Children with SP Cross-sectional, Clinicians and observers rated SM
(2003)49 9.5 T 1.72 yr 23 Socially Female centers child and children higher in severity of social
phobic parent report, fears/distress, child self-reports did
controls clinician ratings, not differ between groups; SM
structured children scored higher on CBCL
interview, delinquency scale
behavioral
assessments
SP, social phobia; TRF, Teacher Report Form; YSR, Youth Self Report; CBCL, Child Behavior Checklist. Citation for each is number 50 in reference list.

Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
345
346 COHAN ET AL JDBP/August, Vol. 27, No. 4

improvement, 29% experienced partial remission, and 39% slightly but were still symptomatic and remained impaired
no longer met diagnostic criteria. Patients whose symp- relative to control children.
toms were unchanged at follow-up reported a duration of Results from a longer-term follow-up study also suggest
between 20 and 30 years, whereas those who experienced that significant functional impairment remains over time.
complete or partial remission at follow-up reported an Twenty-five of the patients followed up by Remschmidt
average duration of 9 years (range, 3Y15 years). Nineteen et al25 completed a standardized interview regarding
percent of those patients who were asymptomatic or their subjective experience of SM. The majority (88%) of
significantly improved at follow-up (showed only mild these patients viewed their SM as a serious problem that
psychosocial or speech disturbances that did not interfere resulted in intense suffering. Many of these patients also
with daily functioning) experienced an abrupt amelioration reported that their SM was associated with anxiety (56%),
of mutism, but for the majority (81%), improvement skepticism and mistrust of others (56%), and feelings of
occurred gradually and 19% experienced periods of shame/insufficiency (72%). Patients who were younger
relapse. All of the patients who were only mildly improved than 18 years at follow-up rated themselves significantly
at follow-up (experienced mutism in some settings that lower in terms of independence, self-esteem, maturity,
interfered with daily functioning) continued to experience and achievement motivation in school situations relative
difficulty in certain social settings. These difficulties to the reference group for the Mannheim Biographic
included occasional mutism, fear of unknown situations, Interview. Those older than 18 years at follow-up rated
fear of talking to strangers, and fear of specific speaking themselves significantly lower on scales measuring
situations (e.g., on the telephone or in public settings such extraversion/social open-mindedness and psychological
as offices). robustness/stress tolerance.
Another study reporting follow-up data on 24 treatment- Additional evidence for the functional impact of SM
seeking children with SM found that nearly 13% showed was reported by Dummit et al8 who found evidence of
marked improvement (mutism was no longer evident), significant impairment in a sample of clinically referred
33% were moderately improved (some residual evidence children with SM. Mean CGAS scores (50.4 T 9.4) were
of mutism), and 54% showed little to no improvement similar to those found in another sample of referred chil-
(mutism still evident) in the 5 to 10 years after treatment.23 dren27 and in the community sample of Bergman et al,18
Importantly, all but one of the children who improved did reflecting a ‘‘moderate degree of interference in function-
so by age 10 years, indicating that those who fail to show ing in most social areas or severe impairment of function-
improvements by middle childhood are experiencing a ing in one area.’’26 Scores on the Liebowitz Social
more persistent form of the disorder. These findings Anxiety Scale28 indicated that these children also showed
suggest that although some patients with SM show clinically significant levels of social anxiety and phobic
remarkable improvement over time, many continue to avoidance in speaking and nonspeaking situations. Other
experience discomfort in speaking situations, and a studies have also reported evidence of social skills deficits
considerable number experience a chronic course of in children with SM relative to community controls.22
mutism. These data are from clinical samples and may
therefore reflect the most severe cases of SM; additional
COMORBID DISORDERS
prospective studies using community samples are needed
to better understand the stability and long-term impact of Data from a number of studies suggest that SM is often
the disorder. comorbid with other psychiatric disorders. Several studies
have found high rates of anxiety disorders, most fre-
quently, social phobia.8,21,24,29,30 Communication disorders
FUNCTIONAL IMPACT OF SELECTIVE MUTISM are also believed to be common among SM children,8,20,29
Studies explicitly examining the functional impact of as are elimination disorders.29 Some studies have also
SM are rare, but some recent investigations have attempted found evidence for developmental disorders24,29 and
to address this issue. In their community study of children oppositional defiant disorder.8,20,21 Because SM frequently
with SM, Bergman et al18 found that children with SM occurs in combination with other disorders, significant
were significantly more impaired than healthy comparison levels of functional impairment are likely to be found in
children. Overall functioning as measured by the Chil- most children presenting with the disorder. It is important
dren’s Global Assessment Scale26 was in the moderately to note, however, that comorbid conditions must be ruled
impaired range for children in the SM group (mean CGAS out as the primary cause of a child’s mutism. The extent
score = 47.92), whereas comparison children scored in the to which developmental and psychological factors are
minimally impaired range (mean CGAS = 77.75). Teacher implicated in the etiology of SM will be discussed in fur-
ratings revealed that not speaking often interfered with ther detail.
academic and social functioning in the SM group; there
was no evidence of impairment caused by speaking
THE DEVELOPMENTAL
problems in the control group. Teacher reports also re-
PSYCHOPATHOLOGY PERSPECTIVE
vealed that children with SM were significantly more
likely to have been referred for special services (58%) rela- Since its inception, the field of developmental psycho-
tive to control children (0%). Data collected at 6-month pathology has recognized the need for integrated theories
follow-up indicated that children with SM had improved of both normal and abnormal development that are

Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 Developmental Psychopathology of SM 347

sensitive to the complexities of human experience.31 The these are most important in the development of the
developmental psychopathology perspective does not disorder. Indeed, some authors have suggested that
advocate a particular theoretical orientation or explanation children with SM represent a heterogeneous group33 and
for understanding disorders of childhood such as SM. On that the current diagnostic classification may need to be
the contrary, developmental psychopathology seeks to reconsidered to reflect new research in this area.3 Accord-
combine clinical and developmental perspectives on the ing to a developmental psychopathology perspective,
biological, psychological, and social processes implicated heterogeneity in the manifestation of a given form of
in both adaptive and maladaptive functioning. Because this psychopathology suggests the existence of several possible
approach advocates an integration of previously distinct developmental pathways that could contribute to the mani-
research traditions, it is an ideal perspective for organizing festation of the disorder.
the accumulated knowledge regarding the etiology of SM.
The developmental psychopathology perspective is
particularly relevant to behavioral developmental pedi- DEVELOPMENTAL PATHWAYS TO
atrics, as it provides an integrative theoretical framework SELECTIVE MUTISM
for understanding the complexities of pathological and
Early theories regarding the etiology of SM emphasized
normal development throughout the life span. According
to Sroufe and Rutter,32 development is a series of psychodynamic issues such as family neurosis and unre-
solved intrapsychic conflicts.3 There has also been some
qualitative reorganizations within and among the bio-
suggestion that SM develops as a response to early trauma
logical, social, emotional, cognitive, representational, and
and/or stressors such as divorce, death of a loved one, and
linguistic systems of a given individual. These reorganiza-
frequent moves.5 Oppositional and/or controlling behavior
tions occur throughout the life span and can be seen as
has also been noted in some studies.23,34 More recently,
complex interactions between the changing individual and
theorists in this area have shifted away from dynamic
his or her changing environmental context. Within this
conceptualizations of the disorder, instead emphasizing
framework, psychopathology results from maladaptive
similarities between SM and behaviorally inhibited tem-
attempts to resolve the challenges of a given develop-
perament9 and anxiety disorders such as social phobia.21
mental period, which, in turn, leads to a lack of integration
among the component systems. Failure to meet challenges Still others have noted that children with SM exhibit
higher rates of neurodevelopmental delays2 and that the
of earlier developmental periods can lead to difficulty later
disorder appears to be more common among immi-
in development because poor integration among systems
may not leave the individual adequately equipped to grants.19,35 From a developmental psychopathology per-
spective, each of these factors may be important in
handle these challenges. Therefore, an understanding of
understanding the etiology of SM in a given individual.
multiple risk and protective factors operating throughout
Therefore, a comprehensive review of what is currently
development is important to effectively intervene and
known about the genetic, temperamental, psychological,
increase the likelihood of adaptive outcomes.
developmental, and social/environmental factors associ-
Building on this framework, theorists in developmental
ated with SM is vital to understanding the various
psychopathology have advocated a ‘‘multiple pathways’’
pathways that could lead to the disorder.
approach to understanding child psychopathology that
emphasizes the importance of multiple interacting causal
Genetic Factors
processes in the development of disordered outcomes. In
any individual child, there are likely to be multiple There is some evidence that SM is familial, suggesting a
contributors to disordered outcomes, and for children with genetic component to the disorder. In a pilot study of 30
the same outcome, different contributors may be operating children meeting criteria for DSM-III-R SM, Black and
throughout development. The principle of equifinality Uhde21 found that 15% of parents and 19% of siblings had
states that a variety of pathways may lead to the same a history of SM. Anxiety disorders were also quite
outcome. In contrast, the principle of multifinality suggests common in this study. Forty-four percent of parents had
that the same risk factor may lead to different outcomes, a history of social phobia or avoidant disorder, and 21% of
depending on the organization of the system in which it is siblings had histories of these disorders. This study did not
operating. Therefore, it is expected that there will be a include a comparison group, so the extent to which rates
number of possible pathways by which a given child may of these disorders are elevated in the relatives of se-
come to exhibit pathology (i.e., SM), and there may be lectively mute children as compared with psychiatric or
individual differences in the specific features of the normal controls is still unclear. However, a more recent
disturbance that are exhibited.31 study reported significantly increased rates of shyness/
Within this perspective, risk and vulnerability factors social anxiety in parents of 54 clinically referred children
identified at multiple levels of analysis (e.g., physiological, with SM as compared with parents of 108 nonreferred
genetic, behavioral, psychological, environmental, and matched control children. Kristensen and Torgersen33
sociological) are incorporated into a single model. The found that 38.9% of mothers and 31.4% of fathers of
pathways approach may be especially helpful in under- children with SM endorsed shyness and/or social anxiety
standing the etiology and course of SM because a number as compared with only 3.7% of control mothers and 0.9%
of different risk and vulnerability factors have been of control fathers. History of SM was reported for 9.3%
identified and there is little consensus regarding which of of parents in the SM group. Control parents were not

Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
348 COHAN ET AL JDBP/August, Vol. 27, No. 4

asked regarding history of SM, so statistical comparisons classified as having ‘‘behaviorally inhibited’’ tempera-
could not be made. ments.8,9,40 Behavioral inhibition was first identified by
Family history data also suggest high rates of commu- Kagan et al44 in the Harvard Infant Study. It describes an
nication, depressive, and anxiety disorders in family enduring temperament in which children display fear,
members of children with SM. In a case-control study avoidance, or withdrawal from novel people, situations, or
comparing 24 children with SM to 84 speech-retarded and objects. The construct not only overlaps somewhat with
102 normal controls, Kolvin and Fundudis23 found a high extreme shyness, but also encompasses inhibition in the
rate of depression among fathers (21%) and mothers (17%) presence of novel nonsocial objects (e.g., toys and games).
of the selectively mute children. Similarly, Andersson and It is thought to reflect a lower threshold for activation in
Thomsen24 found a family history of psychopathology limbic and hypothalamic brain structures and the sympa-
(mostly depression) in one third of parents of 37 clinically thetic nervous system in response to unfamiliar, challeng-
referred selectively mute children. They also noted that, in ing, or threatening stimuli. Behavioral inhibition appears
the majority of SM cases, parents (59.0%) and 1 or more to manifest differently at different developmental stages.
siblings (35.1%) exhibited shyness and speech difficulties Toddlers tend to withdraw from unfamiliar people and
in social situations. cling to caregivers; they often cease spontaneous play and
Steinhausen and Adamek36 conducted a case-control vocalization. Preschoolers show hesitancy and restraint
study of 38 children with SM and 31 children diagnosed when interacting with an unfamiliar peer or adult, often in
with what they called a ‘‘behaviorally similar phenotype’’ the form of inhibited speech. In the early elementary
(an emotional disorder and comorbid communication dis- period, behavioral inhibition is expressed as reticence with
order). Maternal reports suggested increased rates of unfamiliar adults and social isolation from peers. This type
taciturnity among first-, second-, and third-degree relatives of inhibited behavior has been documented in both labo-
of the selectively mute children as compared with the ratory and school settings. Interestingly, Kagan et al44
control group. Although disorders of speech and language found that 1 of the most sensitive indices of behavioral
were common in both groups, significantly higher rates of inhibition in 5-year-old children was lack of spontaneous
psychiatric disorders were found in the relatives of SM speech in the presence of an unfamiliar adult.
children. In a more recent study, Kristensen and Torgersen33 There is increasing evidence that children identified as
also examined personality traits and symptoms of social behaviorally inhibited in the first 3 years of life are at
anxiety/shyness in parents of 54 SM children. They used the greater risk for anxiety disorders in later childhood45 and
Millon Clinical Multiaxial Inventory37 to assess personality for social phobia in particular.46 This line of research has
features and clinical syndromes in parents of clinically prompted some researchers to suggest that SM and social
referred children relative to parents of nonreferred matched phobia represent stages in a developmental progression of
control children. Results indicated significantly higher behaviorally inhibited temperament.18 Dummit et al8(p658)
scores on scales measuring features of avoidant and went even further, suggesting that SM could represent
schizotypal personality in mothers of SM children and on ‘‘the extreme end of a continuum of temperament and
scales measuring anxiety in fathers of SM children. social behavior that has a biological basis.’’
High rates of disorders within a family are often taken A number of studies have identified characteristics of
as evidence of a strong genetic contribution; however, selectively mute children that are indicative of behavioral
environmental factors may also account for this pattern. inhibition at various stages of child development. For
Future studies are needed to clarify the role of genetic instance, Ford et al40 recruited 153 children from the
versus environmental factors in the development SM. Due membership rolls of a large SM support and advocacy
in part to the relative rarity of the disorder, there have been group. Most parents in this study reported that, as infants or
no published family-genetic or twin studies of SM to date. toddlers, their children did not respond well to new stimuli,
Anecdotal and case reports of twins suggest a high rate of and they had difficulty handling transition or change. At
concordance for the disorder in twin pairs.38,39 Twins have school age, these children spoke with less frequency,
been overrepresented in several studies of SM.8,40 Unfortu- volume, and spontaneity than is usual, in addition to
nately, the zygosity of twin pairs included in these studies showing mutism in school and community settings. In their
could not be accurately determined, and heritability study of second graders, Kumpulainen et al17 found that
analyses were not conducted.41 Future systematic twin teachers described the majority of SM children as shy
studies are needed to obtain estimates of the heritability of (68%), withdrawn (63%), and serious (58%). This finding is
SM. Molecular genetic studies examining candidate genes similar to results from the study of Steinhausen and Juzi,20
(e.g., such as the serotonin transporter gene promoter in which the most frequent personality characteristic
polymorphism, shown in several studies to be associated reported for SM children was shyness (85% of sample).
with increased risk for shyness)42,43 are also needed to In addition, in a recent case-control study examining
determine the relative influence of genes and environment temperamental characteristics of children with SM using a
in the development of SM. validated temperament inventory, it was found that, in
comparison to children without the disorder, SM children
were rated as shyer (showing more inhibition and
Temperamental Factors
awkwardness in social situations) and less social (showing
Several researchers have noted that selectively mute a preference for being alone rather than with others) by
children share a number of characteristics with children both parents.47 No other studies have directly assessed

Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 Developmental Psychopathology of SM 349

temperamental characteristics, but this preliminary evi- public. Results from another community-based study
dence does suggest an association between behavioral directly comparing children with SM to children with
inhibition and SM. Because the existing studies have been anxiety disorders found evidence of strong similarities
cross-sectional in nature, the extent to which behavioral between the 2 groups.30 All of the SM children in this
inhibition acts as a vulnerability factor for the develop- study met criteria for social phobia. There is also some
ment of SM remains unclear. Future prospective longi- evidence that social anxiety persists even after children have
tudinal studies of youngsters identified as behaviorally recovered from their SM.21,40 Results such as these have
inhibited are needed to determine the extent to which these prompted some researchers to suggest that SM is simply a
children may go on to develop SM and which risk factors more severe variant or a symptom of social phobia.48
are most likely to interact with temperamental variables to In an effort to investigate this hypothesis, Yeganeh et al49
cause the disorder. compared 23 children with comorbid SM and social phobia
to 23 age-matched children with social phobia alone.
Results indicated greater severity of clinician-rated social
Psychological Factors
fears and social distress among the SM group. Children with
The 2 major psychological factors that have been SM were also rated as being more anxious and less skilled
associated with SM are anxiety and oppositionality. on behavioral tasks that involved peer interaction. However,
Whereas associations between SM and anxiety appear to the 2 groups did not differ in terms of self-rated anxiety
be quite consistent in the literature, the findings with regard during the behavioral tasks. Yeganeh et al49 also found that
to oppositional or controlling behavior remain mixed. there were no significant group differences on Child
A number of studies have found evidence for a general Behavior Checklist subscales,50 with the exception of
association between SM and anxiety symptoms and for a delinquency. Children in the SM group scored higher than
specific relationship between SM and social anxiety. those with social phobia alone on this scale, although their
Steinhausen and Juzi20 reported that 66% of their sample scores were still less than the cutoff usually considered
was characterized as anxious in situations that were clinically significant. This finding suggests that low levels
unrelated to social communication. Cunningham et al22 of oppositional behavior may be characteristic of children
also found that SM children were rated by their parents with SM, but not those with social phobia alone.
and teachers as more anxious in comparison to community Similar results were found by Ford et al,40 who reported
controls. Other studies have reported higher scores for that items reflecting oppositional behavior were among the
internalizing problems in children with SM.19,40 Child Behavior Checklist50 items most frequently endorsed
In their pilot study of 30 referred children with SM, by parents of children with SM (e.g., argues a lot; stubborn,
Black and Uhde21 found that 97% met DSM-III-R sullen, or irritable; disobedient in school; whining; temper
diagnostic criteria for social phobia or avoidant disorder, tantrums or hot temper; and does not answer when people
suggesting a very strong link between SM and clinically talk to him/her). Kolvin and Fundudis23 also noted that
significant social anxiety. A third of this sample also met children with SM exhibited significantly more behavior
criteria for another additional anxiety disorder (specific problems relative to control children. A number of other
phobia, overanxious disorder, and separation anxiety studies have reported varying rates of oppositional behavior
disorder). Strikingly similar results were reported by and/or oppositional defiant disorder among children with
Dummit et al,8 who found that all of the 50 clinically SM. In a treatment study of 20 children with the disorder,
referred children with DSM-III-R SM included in their Krohn et al34 found 90% of the sample to be controlling,
study met criteria for either social phobia or avoidant negative, or oppositional. Most other descriptive studies
disorder. An additional 48% of these children also met have found lower rates of oppositionality. For example,
criteria for separation anxiety disorder, overanxious dis- Steinhausen and Juzi20 reported that one fifth of their
order, or specific phobia. A more recent case-control sample could be described as oppositional defiant. Sim-
study29 found higher rates of anxiety disorders among ilarly, Dummit et al8 found that only 3 of the 50 selectively
selectively mute children (74.1%) as compared with mute children in their study met criteria for oppositional
healthy control children (9.3%). The most common defiant disorder. Black and Uhde21 also reported that 3
anxiety disorder reported by SM children was social children in their study were diagnosed with oppositional
phobia (66.7%), followed by separation anxiety disorder defiant disorder, but they stated that these behaviors were
(31.5%). None of the control children met criteria for not a primary concern for any of the parents in their study.
social phobia or had more than 1 anxiety disorder, whereas Recent case-control studies have failed to find signifi-
and 37% of the SM children received multiple anxiety cant differences in teacher-rated oppositional behavior in
disorder diagnoses. selectively mute children relative to age-matched controls,
There has been some suggestion that the high rates of but parents describe SM children as having slightly higher
comorbid anxiety disorder diagnoses found in these studies ratings of behavior problems and being less cooperative
are a result of the clinically referred nature of the samples than control children.18,19,22 Some authors have cautioned
under study. However, in their school-based study of that the avoidance behaviors exhibited by selectively mute
kindergarten and first- and second-grade students, Bergman children may be misinterpreted as being controlling or
et al18 still found significantly higher rates of teacher-rated manipulative, when they are actually an expression of
social phobia symptoms among selectively mute children, shyness or anxiety.49 The extent to which the oppositional
even after excluding items specifically related to speaking in behaviors reported in the literature reflect this type of

Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
350 COHAN ET AL JDBP/August, Vol. 27, No. 4

misunderstanding remains unclear, and further research is Manassis et al27 found that these 2 groups showed similar
needed to clarify this issue. levels of anxiety and academic ability, but children with
SM showed impairments on a speech and language
assessment relative to those with social phobia. Children
Developmental Factors
with SM also showed deficits in expressive narrative
Although organic causes for SM are listed as exclusionary abilities when compared with socially phobic children
criteria in the DSM-IV, an association has consistently been matched on age, nonverbal cognitive skills, and receptive
found between SM and developmental disorder/delay. language ability.52
Kolvin and Fundudis23 reported evidence of developmental Kristensen and Torgersen33,47 have also reported some
immaturity and an excess of speech abnormalities in their evidence that children with SM alone and children who are
sample of selectively mute children. The SM children in diagnosed with both SM and a communication disorder may
their study began speaking significantly later than children represent subgroups with different etiologies. Their research
in the control group, and a greater proportion of SM chil- suggests that those with the comorbid condition may
dren (50%) continued to manifest speech difficulties as actually represent a less distressed group. In contrast to
compared with controls (9%). A similar finding was re- those with SM alone, children with comorbid SM and
ported by Andersson and Thomsen,24 who found speech communication disorders appear to be characterized by
difficulties in nearly 50% of the selectively mute children greater emotional stability and higher sociability. Parents of
they studied. Thirty-eight percent of the children included in children with the comorbid condition did not differ from
the study of Steinhausen and Juzi20 met criteria for control parents on measures of temperament or personality
premorbid speech and language disorders, including artic- disorder, whereas parents of children with SM alone had
ulation disorders, expressive language disorders, receptive significantly greater scores on scales measuring distress,
language disorders, and stuttering. Delayed motor develop- fear, and activity level. Mothers of children with SM alone
ment was also evident in 18% of the sample, and 24% had higher scores on measures of dysthymic, schizoid,
experienced delays in toilet training. In contrast, studies avoidant, and schizotypal personality features. These results
focusing on social anxiety have found lower rates of have led Kristensen and Torgersen to suggest that there are
language disorder/delay (10Y11%).8,21 2 separate developmental pathways that result in SM, one
Kristensen29,51 has published a series of reports examin- related more to developmental disorder/delay and the other
ing developmental disorder/delay in children with SM. In associated more with neuroticism and social withdrawal.47
his initial report, he found that 68% of the children with
SM met criteria for a diagnosis reflecting developmental
Social/Environmental Factors
disorder/delay, in contrast to only 13% of control chil-
dren.29 Higher rates of elimination disorders were also Early psychodynamic views of SM emphasized expo-
found in the selectively mute children (31.5%) as sure to trauma in the etiology of the disorder. This
compared with controls (9.3%). Nearly half of these conceptualization has decreased in influence as cognitive-
children had a comorbid anxiety disorder diagnosis behavioral therapy and pharmacotherapy have become the
(46.3%). More recently, Kristensen51 examined nonspe- treatments of choice for SM. However, results from a
cific markers of neurodevelopmental disorder/delay, small number of studies have indicated an association
including history of prenatal and perinatal difficulties, between traumatic stress and the onset of some cases of
problems with motor coordination, and presence of minor SM. Andersson and Thomsen found that traumatic expe-
physical abnormalities in this same group of children. In riences severe enough to be noted in clinical case records
comparison to healthy controls, selectively mute children were present in more than one third of the selectively mute
exhibited significantly greater delays in gross and/or fine children included in their sample.24 These traumatic
motor function, and their mothers reported more prenatal experiences occurred during the same period as the
and perinatal difficulties. Results from neurodevelop- development of speech, suggesting that traumatic stress
mental assessments also indicated significant delays in could not be ruled out as a potential risk factor for the
motor skills (measured by several age-adjusted motor disorder. Unfortunately, Andersson and Thomsen did not
tests) and greater number of physical abnormalities in the publish rates of traumatic exposure for their psychiatric
SM group. Separate analyses of SM children with and control group, making it difficult to determine if the rate of
without comorbid communication disorders showed no traumatic stress was elevated in the SM group. Descriptive
significant differences on these variables. studies of children with SM have generally reported lower
Because the majority of cases of SM are also charac- rates of traumatic exposure. Steinhausen and Juzi20
terized by anxiety, Kristensen concluded that SM could be reported that a small proportion (8%) of their sample
best understood as a ‘‘symptom of anxiety reflecting experienced a traumatic event that was associated with
different vulnerabilities, rather than a distinct disorder. onset of SM symptoms. Black and Uhde21 also docu-
The common denominator may be that their neurodevel- mented histories of physical abuse, sexual abuse, and
opmental immaturity makes these children more vulner- neglect in their pilot study of 30 children with SM.
able to ‘everyday’ traumas and that they tend to react to Thirteen percent of the sample reported traumatic experi-
novelty with anxiety and withdrawal.’’28(p10) Further ences of this type. The onset of SM preceded the trauma in
evidence for this position comes from a recent study half of these cases and was unclear for the remaining
comparing children with SM to those with social phobia. cases. Based on these findings, the authors concluded that

Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 Developmental Psychopathology of SM 351

there was no evidence of an immediate temporal or causal confident speakers in all social settings. However, children
relationship between traumatic experience and SM. with SM do not show this progression and remain mute in
Other environmental factors, such as family stress and certain (usually public) settings. Toppelberg et al35 noted
instability, have also been noted as potential risk factors that, when evaluating language minority children for SM, it
for SM. Kumpulainen et al17 found evidence for exposure may be important to observe nonspeaking behaviors over
to stressful life experiences (e.g., death of an important a longer duration than the 1 month required for the
person, change of school, and alcoholism in the family) in diagnosis of SM. In these children, SM is likely when
47% of the selectively mute children in their study; in the nonspeaking behavior appears disproportionate to the
16%, the situation occurred just before the onset of SM. child’s knowledge/exposure to the new language, is
Steinhausen and Juzi20 reported that 24% of their sample present in both languages, and is often accompanied by
experienced a stressful life event (e.g., loss of a significant shyness or anxiety in novel situations. A potential pathway
person, acute intrafamilial crisis, and onset of disease in to SM for bilingual children may begin with behaviorally
the child or significant other) before the onset of the inhibited temperament which interacts with immigration
disorder. Kristensen29 also found a greater incidence of and the stress of second-language acquisition to occasion
movement of household or change in school setting among the development of SM. An understanding of the risk
children with SM relative to control children. Black and factors for SM in bilingual children is likely to become
Uhde21 reported that 4 children in their sample developed increasingly important for pediatricians and other treating
SM after life stressors such as moving to a new home, professionals, as these children represent a growing propor-
starting school, or witnessing family conflict. Overt marital tion of the youth population in the United States and abroad.
conflict has also been found to be higher in families of
children with SM relative to control families.19 Results
LIMITATIONS OF THE PRESENT REVIEW
from these studies provide some evidence for the existence
of another pathway to SM that is specifically related to The goal of the present review was to elucidate potential
trauma and/or environmental stress. pathways to the development of SM by organizing the
An additional stressor that may be related to develop- existing SM literature in terms of the various genetic,
ment of SM is immigration status. Descriptive studies temperamental, psychological, and social/environmental
suggested that immigrants may have been overrepresented factors that have been implicated in the etiology of the
among children with SM,8,20 but until recently, there were disorder. A strength of the developmental psychopathology
no controlled studies examining this variable. In 2003, perspective is that it encourages integration of previously
Elizur and Perednik19 conducted the first community- distinct research areas; however, some limitations to the
based study to explicitly examine the role of immigration present literature review should be noted. This review was
as a risk factor for development of SM. They found that limited to articles published in English; international
the prevalence rate of SM among immigrants (2.2%) was attention has been focused on SM since it was first
nearly 4 times that of the prevalence rate among native identified, and a number of recent studies may have been
families (0.47%). Immigrant children with SM appeared to overlooked. Although case studies have provided an
have a less complicated presentation, with higher rates of important source of information regarding phenomenology
social anxiety but few other comorbid symptoms. In and treatment of SM, these were also excluded from our
contrast, native children with the disorder had high rates review. Finally, because treatment was not the primary
of both social anxiety and neurodevelopmental disorder/ focus of this review and this topic has been covered
delay. They appeared to be more impaired socially relative elsewhere,4Y6 we did not include a comprehensive review
to immigrant children with the disorder, prompting the of the treatment literature.
authors to suggest that ‘‘their SM could be triggered by a
variety of stressors that interact with a more generalized
SUMMARY OF PATHWAYS IN THE
maladjustment, while in immigrants, SM could be more
DEVELOPMENT OF SELECTIVE MUTISM
specifically related to language demands.’’19(p1457) This
theory is consistent with a pathways perspective on SM. Our review of the SM literature suggests that the disorder
As previously noted, DSM-IV criteria state that SM can result from several diverse pathways reflecting complex
should not be diagnosed if the child lacks sufficient interactions among multiple genetic, temperamental,
knowledge or comfort with the language spoken in their psychological, developmental, and social/environmental
new host country. Therefore, an understanding of the systems. The presence of risk and vulnerability factors
developmental stages in second-language acquisition is may predispose certain children to develop SM, but these
necessary to correctly diagnose SM in immigrant popula- same factors could also lead to different outcomes. In line
tions. Toppelberg et al35 provided some guidelines for the with this developmental psychopathology perspective,
differential diagnosis of SM in language minority immi- some authors have suggested a diathesis-stress model of
grant children. They described a normal period of non- SM, in which different pathways are involved for the
speaking during the acquisition of a second language that is development of so-called ‘‘pure selective mutism’’ and
common among children aged 3 to 8 years and typically SM occurring in combination with other disorders.19,47
lasts less than 6 months but may be longer in younger This distinction has important implications for under-
children. Children showing normal second-language acqui- standing the functional impact of SM and is also likely
sition pass through this nonverbal period and become to be relevant when developing interventions with SM

Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
352 COHAN ET AL JDBP/August, Vol. 27, No. 4

children. Figure 1 illustrates possible pathways for the information regarding SM. These shortcomings are under-
development of SM in children. standable, as recruitment of children with a relatively
Because anxiety appears to be a characteristic feature of rare psychiatric disorder is a formidable challenge for
most selectively mute children, this may be a particularly researchers. However, the success of recent community-
important vulnerability factor for the development of the based studies is encouraging for those seeking to conduct
disorder. Therefore, a hypothetical pathway to SM may research with this population.
begin with a child who experiences a high level of anxiety Several lines of research would be helpful in testing the
(because of a strong genetic loading for anxiety, behavior- pathways perspective on SM. First, longitudinal prospec-
ally inhibited temperament, unstable home environment, tive studies of children at high risk for the disorder (based
or some combination of these factors). The child may then on behaviorally inhibited temperament and/or family
have a heightened sensitivity to verbal interactions with history of SM, communication disorders, or social phobia)
others (because of developmental immaturity, a commu- are needed to further identify and clarify the temper-
nication disorder, or immigrant status), which is triggered amental, environmental, and psychological variables that
by an environmental stressor (interpersonal trauma, move may interact to occasion the development of SM. Little is
to a new school, or death of a loved one), resulting in a currently known regarding the course of the disorder, and
failure to speak in a given setting (usually school) despite research into factors that distinguish transient from
the ability to do. The failure to engage verbally in social persistent mutism would add considerably to our under-
situations may be interpreted by parents or teachers as standing of the developmental psychopathology of SM.
oppositional behavior. The nonspeaking behavior may also Because there has been some suggestion that most children
reinforce a pattern of withdrawal from peers, resulting in will ‘‘grow out’’ of the disorder, longitudinal investiga-
heightened anxiety regarding social interactions and a tions are necessary to clarify the extent to which SM
potential diagnosis of social phobia. It is important to note spontaneously remits in later childhood or adolescence. In
that the trajectory of SM can be altered at each stage of its addition, this type of research could provide important
development. Thus, some children may find that the information regarding the relationship between SM and
disorder remits fairly quickly, whereas others experience social anxiety because retrospective studies have suggested
SM for many years and continue to report discomfort in that adults who were selectively mute during childhood
speaking situations well into adulthood. experience clinically significant social anxiety later in
life.21 There are few well-controlled studies of children
with the disorder, so the addition of both normal and
psychiatric control groups to this type of investigation
FUTURE RESEARCH DIRECTIONS
would help to tease out the extent to which certain risk and
The SM literature has grown considerably during the vulnerability factors are specific to SM.
past decade, but there are still relatively few studies A number of the studies reviewed here suggest that SM
investigating the etiology and course of this unusual is a familial disorder which may have a genetic compo-
disorder. Among other methodological difficulties, much nent. However, twin and adoption studies are still needed
of the existing research is marked by small sample sizes to assess the role of genetics and shared/nonshared
and lack of adequate comparison groups. The majority of environmental influences in the etiology of the disorder.
these studies are cross-sectional, and several researchers Large-scale investigations of this type have been difficult
have relied on retrospective chart reviews to gather to carry out, given the relative infrequency of SM in the

FIGURE 1. Potential pathways to the development of SM in children.

Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 Developmental Psychopathology of SM 353

population. Because of their large numbers, national has several important implications for the diagnosis and
advocacy groups may be helpful in identifying and treatment of children with SM. It may be particularly
recruiting potential participants for twin and adoption important for pediatricians working with these children to
studies of SM. In addition to twin and adoption studies, be alert for the early risk factors and symptoms of SM (e.g.,
molecular genetic studies would be extremely useful to family history of SM and/or social phobia, and behaviorally
examine genetic variation associated with anxiety disor- inhibited temperament) and to monitor the presence of
ders, communication disorders, and other developmental additional risk factors as these children develop (e.g.,
delays. The use of parent-child triads in family-based neurodevelopmental delays, social anxiety, and immigra-
association analysis designs53 can provide an efficient tion). Because pediatricians are the most likely to see these
strategy for determining the role of molecular genetic children from infancy throughout their development, they
candidates (e.g., serotonin transporter gene promoter are in a good position to identify SM at its early stages and
polymorphism) in susceptibility to SM. to encourage parents to seek treatment for their children.
Although family history is implicated as a risk factor for However, clinicians should be cautioned that transient
SM, there has been little research into the factors that mediate mutism during transitional periods (e.g., school entry and
familial transmission and predict outcome in children with second-language acquisition) may be part of a normal
SM. One promising area of investigation comes from the developmental trajectory for some children and should not
field of affective neuroscience. The last several years have be diagnosed as SM. An understanding of some typical
seen dramatic increases in the use of neuroimaging tech- pathways to SM (e.g., from a preexisting genetic vulner-
niques to investigate brain structure and function in normal ability to behavioral inhibition in infancy and social phobia
and psychiatric populations. Recent investigations indicate later in childhood) may help pediatricians distinguish
reduced amygdala volumes in children with anxiety disorders between children who show signs of reticence that falls
relative to normal controls.54 Functional neuroimaging within the normal range of speaking behaviors from those
techniques have been used to examine brain function in who are likely to be showing early symptoms of SM.
adult patients with anxiety disorders as compared with Awareness of the relationship between early developmental
normal controls, often using paradigms based on fear deviations and later SM could potentially be used to prevent
conditioning or processing of emotional faces. For example, a more persistent and impairing form of the disorder.
recent research suggests that adults with generalized social Many clinicians have found SM to be intractable using
phobia exhibit greater amygdala activation in response to conventional psychosocial treatment approaches,2,4 and appli-
angry and contemptuous facial emotional expressions cation of the pathways perspective to individual SM cases
relative to healthy controls.55 However, the extent to which may also be useful to enhance treatment outcomes for these
children who have social phobia and/or SM exhibit children. New cognitive-behavioral interventions have
similarly abnormal responses has not been systematically recently been developed for SM,57,58 but these were designed
studied. It would be particularly informative to explore based on similarities between SM and social anxiety. The
similarities and differences in brain activation among SM extent to which this sort of intervention is successful across a
children relative to healthy controls and children who have variety of ages and with immigrants or children who have
social phobia and communication disorders. Research into comorbid behavior problems and/or communication disor-
the specific brain circuitry involved in SM and related ders is unclear. In line with the multifactorial etiological
conditions may prove fruitful in understanding the devel- perspective on SM, Mendlowitz58 has suggested adding
opmental psychopathology of these disorders.56 different components to the standard cognitive-behavioral
Several other studies reviewed here also suggest a strong package to tailor treatment to more complicated clinical
role for immigrant status as a risk factor for SM, and this presentations. For example, children who present with
area is of increasing research interest. However, the communication deficits may need interventions that focus
mechanism by which immigrant children are at increased on language skills in addition to the standard anxiety
risk for SM remains unclear. The next step in this line of management approaches to SM treatment. Alternatively,
research should be to investigate mediating variables that children whose SM appears to be related to oppositional
may help explain the relationship between immigration behaviors may require a parent management component.
status and SM (e.g., second-language learning, language There is some evidence that psychopharmacological
fluency, ethnic identification, acculturation, minority status, interventions are useful in the treatment of SM. There have
socioeconomic status, and perceived discrimination). This been a small number of controlled and noncontrolled trials
type of investigation would be best carried out using a of selective serotonin reuptake inhibitors (fluoxetine and
community sample of SM children recruited from schools sertraline)59Y61 and monoamine oxidase inhibitors (phenel-
in a large metropolitan area to maximize recruitment of zine).62 Interestingly, in the open trial of fluoxetine of
immigrant children with the disorder, who may not have Dummit et al,60 treatment response was inversely correlated
access to or be referred for treatment at university clinics. with age. This finding suggests that early intervention may
be key for the success of both psychosocial and psycho-
pharmacological interventions. However, the children in
IMPLICATIONS FOR DIAGNOSIS
each of these trials almost always had an additional anxiety
AND TREATMENT
disorder diagnosis. In some cases, children with SM who
The developmental psychopathology emphasis on mul- did not meet criteria for an anxiety disorder were screened
tiple interacting causal processes in the development of SM out.60 The results from these studies are a useful starting

Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
354 COHAN ET AL JDBP/August, Vol. 27, No. 4

point, but it remains unclear whether the use of psychotropic delays and/or communication disorders may not benefit
medications is indicated in children who show SM in the from this type of pharmacological treatment.
absence of significant social anxiety. Moreover, recent Although there have been reports of some successful
concerns and the US Food and Drug Administration warn- cognitive-behavioral and psychopharmacological treat-
ings about selective serotonin reuptake inhibitor adminis- ments for children with SM, the presence of multiple risk
tration to children63 are likely to limit the use of this class of factors and comorbid conditions has made treating these
drugs, at least for the near future. Therefore, an under- children difficult. Armed with a greater understanding of
standing of individual variation in developmental pathways the various pathways to SM, we may be able to design
and comorbid conditions is needed to guide treatment more effective prevention and intervention programs to
planning. For example, those who have developmental help children who have previously suffered in silence.

REFERENCES
1. Krysanski VL. A brief review of selective mutism literature. 19. Elizur Y, Perednik R. Prevalence and description of selective
J Psychol. 2003;137:29Y40. mutism in immigrant and native families: a controlled study. J Am
2. Standart S, Le Couteur A. The quiet child: a literature review of Acad Child Adolesc Psychiatry. 2003;42:1451Y1459.
selective mutism. Child Adolesc Ment Health. 2003;8:154Y160. 20. Steinhausen H, Juzi C. Elective mutism: an analysis of 100 cases.
3. Anstendig KD. Is selective mutism an anxiety disorder? Rethinking J Am Acad Child Adolesc Psychiatry. 1996;35:606Y614.
its DSM-IV classification. J Anxiety Disord. 1999;13:417Y434. 21. Black B, Uhde TW. Psychiatric characteristics of children with
4. Anstendig K. Selective mutism: a review of the treatment literature selective mutism: a pilot study. J Am Acad Child Adolesc Psychiatry.
by modality from 1980Y1996. Psychotherapy. 1998;35:381Y391. 1995;34:847Y856.
5. Dow SP, Sonies BC, Scheib D, et al. Practical guidelines for the 22. Cunningham CE, McHolm A, Boyle MH, et al. Behavioral and
assessment and treatment of selective mutism. J Am Acad Child emotional adjustment, family functioning, academic performance,
Adolesc Psychiatry. 1995;34:836Y846. and social relationships in children with selective mutism. J Child
6. Kumpulainen K. Phenomenology and treatment of selective Psychol Psychiatry. 2004;45:1363Y1372.
mutism. CNS Drugs. 2002;3:175Y180. 23. Kolvin I, Fundudis T. Elective mute children: psychological
7. Torraco RJ. Writing integrative literature reviews: guidelines and development and background factors. J Child Psychol Psychiatry.
examples. Hum Resour Dev Rev. 2005;4:356Y367. 1981;22:219Y232.
8. Dummit ES, Klein RG, Tancer NK, et al. Systematic assessment of 24. Andersson CB, Thomsen PH. Electively mute children: an analysis
50 children with selective mutism. J Am Acad Child Adolesc of 37 Danish Cases. Nord J Psychiatry. 1998;52:231Y238.
Psychiatry. 1997;36:653Y660. 25. Remschmidt H, Poller M, Herpertz-Dahlmann B, et al. A follow-up
9. Bergman RL, Piacentini J. Selective mutism. In: Sadock B, Sadock V, study of 45 patients with elective mutism. Eur Arch Psychiatry Clin
eds. Comprehensive Textbook of Psychiatry VIII, vol 2. Baltimore, Neurosci. 2001;251:284Y296.
MD: Williams & Wilkins Press, 2005:3302Y3306. 26. Shaffer D, Gould MS, Brasic J, et al. A Children’s Global Assessment
10. American Psychiatric Association. Diagnostic and Statistical Manual Scale (CGAS). Arch Gen Psychiatry. 1983;40:1228Y1231.
of Mental Disorders, 3rd ed. Washington, DC: American Psychiatric 27. Manassis K, Fung D, Tannock R, et al. Characterizing selective
Association, 1980. mutism: is it more than social anxiety? Depress Anxiety. 2003;
11. World Health Organization. ICD-10: International Statistical 18:153Y161.
Classification of Diseases and Related Health Problems. Geneva: 28. Liebowitz MR. Social phobia. Mod Probl Pharmacopsychiatry.
World Health Organization, 1992. 1987;22:141Y173.
12. American Psychiatric Association. Diagnostic and Statistical Manual 29. Kristensen H. Selective mutism and comorbidity with developmental
of Mental Disorders, 4th ed. Washington, DC: American Psychiatric disorder/delay, anxiety disorder, and elimination disorder. J Am Acad
Association, 1994. Child Adolesc Psychiatry. 2000;39:249Y256.
13. American Psychiatric Association. Diagnostic and Statistical Manual 30. Vecchio JL, Kearney CA. Selective mutism in children: comparison
of Mental Disorders, 4th ed. Washington, DC: American Psychiatric to youths with and without anxiety disorders. J Psychopathol Behav.
Association, 2000. text revision. 2005;27:31Y37.
14. National Institute on Deafness and Other Communication Disorders. 31. Cicchetti D, Cohen DJ. Perspectives on developmental psychopa-
Speech and language: Developmental milestones. 2000 (NIH pub- thology. In: Cicchetti D, Cohen DJ, eds. Developmental Psychopa-
lication no. 00-4781). Available at: http://www.nidcd.nih.gov/health/ thology, vol 1. New York: Wiley & Sons, 1995:13Y20.
voice/speechandlanguage.asp#mychild. Accessed March 6, 2005. 32. Sroufe LA, Rutter M. The domain of developmental psychopa-
15. Kopp S, Gillberg C. Selective mutism: a population-based study: a thology. Child Dev. 1984;55:17Y29.
research note. J Child Psychol Psychiatry. 1997;38:257Y262. 33. Kristensen H, Torgersen S. MCMI-II personality traits and symptom
16. Wilkins R. A comparison of elective mutism and emotional traits in parents of children with selective mutism: a case-control
disorders in children. Br J Psychiatry. 1985;146:198Y203. study. J Abnorm Psychol. 2001;110:648Y652.
17. Kumpulainen K, R.s.nen E, Raaska H, et al. Selective mutism 34. Krohn DD, Weckstein SM, Wright HL. A study of the effectiveness
among second-graders in elementary school. Eur Child Adolesc of a specific treatment for elective mutism. J Am Acad Child
Psychiatry. 1998;7:24Y29. Adolesc Psychiatry. 1992;31:711Y718.
18. Bergman RL, Piacentini J, McCracken JT. Prevalence and descrip- 35. Toppelberg CO, Tabors P, Coggins A, et al. Differential diagnosis
tion of selective mutism in a school-based sample. J Am Acad Child of selective mutism in bilingual children. J Am Acad Child Adolesc
Adolesc Psychiatry. 2002;41:938Y946. Psychiatry. 2005;44:592Y595.

Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 Developmental Psychopathology of SM 355
36. Steinhausen HC, Adamek R. The family history of children with and TRF Profiles. Burlington, VT: University of Vermont Depart-
elective mutism: a research report. Eur Child Adolesc Psychiatry. ment of Psychiatry, 1991.
1997;6:107Y111. 51. Kristensen H. Non-specific markers of neurodevelopmental disorder/
37. Millon T. Manual for the MCMI-II. Minneapolis, MN: National delay in selective mutism: a case-control study. Eur Child Adolesc
Computer Systems, 1987. Psychiatry. 2002;11:71Y78.
38. Gray RM, Jordan CM, Ziegler RS, et al. Two sets of twins with 52. McInnes A, Fung D, Manassis K, et al. Narrative skills in children
selective mutism: neuropsychological findings. Neuropsychol Dev with selective mutism: an exploratory study. Am J Speech Lang
Cogn C Child Neuropsychol. 2002;8:41Y51. Pathol. 2004;13:304Y315.
39. Segal NL. ‘Two’ quiet: monozygotic female twins with selective 53. Purcell S, Sham P, Daly MJ. Parental phenotypes in family-based
mutism. Clin Child Psychol Psychiatry. 2003;8:473Y488. association analysis. Am J Hum Genet. 2005;76:249Y259.
40. Ford MA, Sladeczek IE, Carlson J, et al. Selective mutism: 54. Milham M, Nugent A, Drevets W, et al. Selective reduction in
phenomenological characteristics. Sch Psychol Q. 1998;13:192Y227. amygdala volume in pediatric anxiety disorders: a voxel-based
41. Segal N. Letter to the editor. Neuropsychol Dev Cogn C Child morphometry investigation. Biol Psychiatry. 2005;57:961Y966.
Neuropsychol. 2003;9:157Y158. 55. Stein MB, Goldin PR, Sareen J, et al. Increased amygdala activation
42. Arbelle S, Benjamin J, Golin M, et al. Relation of shyness in grade to angry and contemptuous faces in generalized social phobia. Arch
school children to the genotype for the long form of the serotonin Gen Psychiatry. 2002;59:1027Y1034.
transporter promoter region polymorphism. Am J Psychiatry. 56. Pine DS. Affective neuroscience and the development of social
2003;160:671Y676. anxiety disorder. Psychiatr Clin North Am. 2001;24:689Y705.
43. Battaglia M, Ogliari A, Zanoni A, et al. Influence of the 57. Bergman RL. Diagnosis and assessment of selective mutism. Paper
serotonin transporter promoter gene and shyness on children’s presented at: the 25th Annual Conference of the Anxiety Disorders
cerebral responses to facial expressions. Arch Gen Psychiatry. 2005; Association of America; 2005; Seattle, WA.
62:85Y94. 58. Mendlowitz S. When cognitive behavioral therapy isn’t enough.
44. Kagan J, Reznick JS, Snidman N. The physiology and psychology of Paper presented at: the 25th Annual Conference of the Anxiety
behavioral inhibition in children. Child Dev. 1987;58:1459Y1473. Disorders Association of America; 2005; Seattle, WA.
45. Rosenbaum JF, Biederman J, Bolduc-Murphy EA, et al. Behavioral 59. Black B, Uhde TW. Treatment of elective mutism with fluoxetine: a
inhibition in childhood: a risk factor for anxiety disorders. Harv Rev double-blind, placebo-controlled study. J Am Acad Child Adolesc
Psychiatry. 1993;1:2Y16. Psychiatry. 1994;33:1000Y1006.
46. Biederman J, Hirshfeld-Becker DR, Rosenbaum JF, et al. Further 60. Dummit ES, Klein RG, Tancer NK, et al. Fluoxetine treatment of
evidence of association between behavioral inhibition and social children with selective mutism: an open trial. J Am Acad Child
anxiety in children. Am J Psychiatry. 2001;158:1673Y1679. Adolesc Psychiatry. 1996;35:615Y621.
47. Kristensen H, Torgersen S. A case-control study of EAS child and 61. Carlson JS, Kratochwill TR, Johnson HF. Sertraline treatment of 5
parental temperaments in selectively mute children with and without children diagnosed with selective mutism: a single-case research
a co-morbid communication disorder. Nord J Psychiatry. 2002;56: trial. J Child Adolesc Psychopharmacol. 1999;9:293Y306.
347Y353. 62. Golwyn DH, Sevlie CP. Phenelzine treatment of selective mutism in
48. Black B, Uhde TW. Elective mutism as a variant of social phobia. four prepubertal children. J Child Adolesc Psychopharmacol.
J Am Acad Child Adolesc Psychiatry. 1992;31:1090Y1094. 1999;9:109Y113.
49. Yeganeh R, Beidel DC, Turner SM, et al. Clinical distinctions 63. US Food and Drug Administration. Suicidality in children and
between selective mutism and social phobia: an investigation of adolescents being treated with antidepressant medications. FDA
childhood psychopathology. J Am Acad Child Adolesc Psychiatry. Public Health Advisory October 15, 2004. Available at: http://
2003;42:1069Y1075. www.fda.gov/cder/drug/antidepressants/SSRIPHA200410.htm.
50. Achenbach TM. Integrative Guide for the 1991 CBCL/4-18, YSR, Accessed March 6, 2005.

Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.