DEVELOPMENTAL MEDICINE & CHILD NEUROLOGY ORIGINAL ARTICLE

Neurodevelopmental, emotional, and behavioural problems in

Duchenne muscular dystrophy in relation to underlying dystrophin
gene mutations
VALERIA RICOTTI 1 | WILLIAM P L MANDY 2 | MARIACRISTINA SCOTO 1 | MARIKA PANE 3 |
NICOLAS DECONINCK 4,5 | SONIA MESSINA 6 | EUGENIO MERCURI 1,3 | DAVID H SKUSE 2 |
FRANCESCO MUNTONI 1,*

1 Dubowitz Neuromuscular Centre, UCL Institute of Child Health, London, UK. 2 Behavioural and Brain Sciences Unit, UCL Institute of Child Health, London, UK. 3
Department of Paediatric Neurology, Catholic University, Rome, Italy. 4 Department of Paediatric Neurology, NMRC, Universitair Ziekenhuis Gent, Gent; 5 Paediatric
Neurology Department, H^opital Universitaire des Enfants Reine Fabiola, Universite Libre de Bruxelles, Brussels, Belgium. 6 Department of Neurosciences and Nemo
Sud Clinical Centre, University of Messina, Messina, Italy.
Correspondence to Francesco Muntoni at Dubowitz Neuromuscular Centre, UCL Institute of Child Health, 30 Guilford Street, London WC1N 1EH, UK. E-mail: [email protected]

*Additional authors listed at the end of the article.

This article is commented on by Hendriksen and Vles on page 13 of this issue.

PUBLICATION DATA AIM Duchenne muscular dystrophy (DMD) is associated with neuropsychiatric disorders. The
Accepted for publication 3rd August 2015. aim of the study was to characterize the DMD neuropsychiatric profile fully and to explore
Published online 14th September 2015. underlying genotype/phenotype associations.
METHOD One hundred and thirty males with DMD (mean age 9y 10mo, range 5–17y) in four
ABBREVIATIONS European centres were included and completed IQ assessment and a neurodevelopmental-
3Di-sv Developmental, Diagnostic and screening questionnaire. Of these, 87 underwent comprehensive neuropsychiatric assessment
Dimensional Interview – short using structured diagnostic interview and parent-reported questionnaires.
version RESULTS The overall mean score on the neurodevelopmental questionnaire was significantly
ASD Autistic spectrum disorder abnormal compared with the general population of children (p<0.001). On average,
CBCL Child Behavior Checklist intelligence was below the population mean, with intellectual disability observed in 34 males
DMD Duchenne muscular dystrophy (26%). Autistic spectrum disorder was identified in 18 (21%), hyperactivity in 21 (24%), and
SCDC Social Communication Disorder inattention in 38 (44%). Clinical levels of internalizing and externalizing problems were
Checklist observed in 21 (24%) and 13 (15%) respectively. Over a third of males scored more than two
measures of emotional, behavioural, or neurodevelopmental problems. Males with mutations
at the 30 end of the DMD gene affecting all protein isoforms had higher rates of intellectual
disability and clusters of symptoms.
INTERPRETATION Males with DMD are at very high risk of neuropsychiatric disturbance, and
this risk appears to increase with mutations at the 30 end of the gene. Patterns of symptom
clusters suggest a DMD neuropsychiatric syndrome, which may require prompt evaluation
and early intervention.

Duchenne muscular dystrophy (DMD) is an X-linked reces- cardiac muscles, in the neurons in the cortex, and in cerebel-
sive neuromuscular disorder affecting one in 5000 live male lar Purkinje cells respectively. However, mutations progres-
births, which causes progressive muscle weakness leading to sively further along the gene affect increasingly more
loss of ambulation in the mid-adolescent years. Affected isoforms. Mutations between exons 31 and 44, in addition to
males generally present in the first few years of life with disrupting expression of the long isoforms, will also disrupt
motor symptoms and enlarged calves. However, neurodevel- Dp260 (expressed mostly in the retina); mutations between
opmental disorders are increasingly recognized features and exons 45 and 62 will in addition disrupt both Dp140 (ex-
can be the initial presenting symptoms.1,2 DMD occurs as a pressed in the brain and kidneys) and Dp116 (expressed in
result of mutations in the dystrophin gene. The large dys- Schwann cells); while the rare mutations downstream of
trophin gene contains 79 exons plus seven promoters. These exon 63 will affect all dystrophin products including the
tightly regulated internal promoters generate a range of dif- shortest Dp71 (most abundant brain isoform) (Fig. 1).3
ferent protein isoforms, with diverse expression in tissues. The role of dystrophin isoforms in the brain remains lar-
Mutations in the 50 end of the gene (i.e. mutations from gely unclear; nevertheless, it is recognized that the brain is
exons 1–31) only affect the three longest isoforms, Dp427M, affected by the lack of dystrophin and notably that mutations
Dp427C, and Dp427P, which are expressed in skeletal and disrupting the brain isoforms Dp140 and Dp71 are more

© 2015 Mac Keith Press DOI: 10.1111/dmcn.12922 77

frequently associated with lower IQ scores.4–6 Furthermore, What this paper adds
central nervous system (CNS) involvement in DMD goes • DMD males with mutations progressively downstream have an overall more
beyond disability arising from impaired intellectual develop- severe neurodevelopmental profile.
ment. Several studies have shown that males with DMD pre-
sent with symptoms of neurodevelopmental disorders, with METHODS
higher prevalence than in the general paediatric population, Patient populations
including autistic spectrum disorder (ASD) reported to Families attending the neuromuscular departments in three
occur in up to 19% of males with DMD,7 attention-deficit– counties (UK, Italy, and Belgium) and four centres were
hyperactivity disorder (ADHD) in up to 31%,8 and obses- recruited over 12 months. Inclusion criteria were as fol-
sive–compulsive disorder in up to 5%.9 lows: (1) males with DMD with proven mutation in the
In our multicentre international study, we aimed to eluci- dystrophin gene; mutations were classified according to the
date the neuropsychiatric phenotype of DMD by using a Leiden Muscular Dystrophy database (http://www.dmd.nl);
comprehensive battery of standardized, well-validated neu- (2) aged between 5 years and 16 years; and (3) willing to
rodevelopment assessments. In so doing we aimed to offer take part into the study and with good understanding of
the most comprehensive picture so far of the neuropsychi- the language spoken in the country of the assessment.
atric elements of DMD, by capturing with parent reporting a Ethics Committee approval was obtained by all participat-
wide array of CNS characteristics including (1) intellectual ing centres.
difficulties, (2) neurodevelopmental disorders (ASD,
ADHD), and (3) emotional behavioural problems such as Measures
internalizing and externalizing behaviours. We aimed to Social Communication Disorder Checklist
ascertain how these different characteristics cluster in the Families attending the neuromuscular outpatient depart-
same individual, exploring the possibility of a ‘DMD ment were asked to complete a validated, 12-item measure,
neuropsychiatric syndrome’. Finally, we aimed to learn about the Social Communication Disorder Checklist (SCDC),
the association between the neuropsychiatric profile and the which is highly sensitive to neurodevelopmental disorders
genotype of DMD, in particular referring to those regions of especially in the social communication domain.10,11 The
interest disrupting the different dystrophin isoforms. SCDC is a three-point Likert scale, with each item scoring

Chromosome X

Xp21

5’
3’
Dp427M
Dp427B Dp260 Dp140 Dp116 Dp71
Dp427P

Exon 1 Exon 31 Exon 63

Dp427 B, M, P : cortical neurons (B), Dp260 : brain, retina layer Dp71 : brain, liver,
skeletal and cardiac muscles (M), and heart heart and retina
Purkinje cells (P)

Dp140 : brain, retina and

kidneys

Dp116 : Schwann cells

Figure 1: Genomic organization of the dystrophin gene.

78 Developmental Medicine & Child Neurology 2016, 58: 77–84

0, 1, or 2; higher scores indicate greater symptom severity. the repetitive and stereotyped behaviour scale and/or the
A total score of at least eight is indicative of likely neu- communication impairment scale are categorized as meet-
rodevelopmental disturbance, whereas scores greater than ing criteria for ASD.
15 are strongly suggestive of ASD. In addition, parents completed the Conners’ Parent Rat-
ing Scale. The short version of the Conners’ Parent Rating
Cognitive assessment Scale is a 45-item questionnaire addressing issues such as
Intelligence was measured, where possible (n=92, 71%), aggression, learning problems, inattention, hyperactivity/
with the Wechsler Intelligence Scales for Children – impulsivity, executive functioning, and peer relations. Chil-
Fourth Edition (WISC-IV).12 A truncated version of the dren who had t-scores in the abnormal range (t<65) on the
WISC-IV was administered (Vocabulary, Similarities, hyperactivity and/or inattention indexes were categorized
Matrix Reasoning, Block Design, Digit Span, and Letter– as meeting criteria for ADHD by parent report.19
Number Sequencing) to reduce participant burden. The
Digit Span and Letter–Number Sequencing subtests were Emotional behavioural problems
used to produce a Working Memory Index score; Similari- Parents completed the Child Behavior Checklist (CBCL).
ties and Vocabulary subtests produced a Verbal Compre- The CBCL is a parent-report measure, with scales measur-
hension Index score; and the Matrix Reasoning and Block ing internalizing symptoms (anxiety and depression) and
Design subtests generated a Perceptual Reasoning Index externalizing symptoms (aggression and hyperactivity). It is
score. General intelligence was measured using the WISC- normed for use with young people between the ages of 4
IV General Ability Index score. This is a measure of over- years and 18 years. In line with the CBCL manual, inter-
all intellectual ability that is particularly suitable for males nalizing, externalizing, and total t-scores greater than or
with DMD as, compared with the Full-scale IQ score of equal to 63 were deemed to be in the clinical range.20
the WISC-IV, it is less influenced by processing speed,
working memory deficits, and motor problems, which are Statistical methods
common in this population.13 In addition, three males To compare the SCDC mean score of the current sample
(2%) who were aged five when they joined the study, and with published population norms, a one-sample t-test was
so were too young for the WISC-IV, were assessed using used. One-way ANOVA was used to compare mean scores
the Wechsler Preschool and Primary Scale of Intelligence on continuous measures in the three genotype-defined sub-
– Third Edition.14 groups of the sample. Post hoc analyses were conducted
In males reluctant to complete a Wechsler intelligence using Gabriel’s procedure as this is well designed to
test or those who were non-verbal (n=20, 15%), general accommodate unequal group sizes. Fisher’s exact tests were
intelligence was assessed with the Raven’s Coloured Pro- used to make group comparisons for categorical variables.
gressive Matrices.15 This is a widely used, well-normed
cognitive assessment for children aged 5 to 11. In the cur- RESULTS
rent study, raw scores were converted to IQ scores with a We recruited 130 males with DMD in four centres: UK
mean of 100 (SD 15).16 Males with a General Ability Index (n=73); Italy (Rome, n=29; Messina, n=16); Belgium
or a Raven’s Coloured Progressive Matrices IQ score equal (n=12). Mean age was 9 years 10 months (range 5–16y)
to or less than 70 were categorized as being in the intellec- including both ambulant and non-ambulant participants.
tual difficulty range.17 An additional 15 males (12% of All males were treated according to the standards of care
overall sample) were not able to complete a WISC-IV, a for DMD.1 The overall grouping for genotype was 41
Wechsler Preschool and Primary Scale of Intelligence, or patients with mutations upstream of exon 30, 75 males
Raven’s Coloured Progressive Matrices, because of being with mutations in between exons 31 and 62, and 14 males
non-verbal and having developmental delay. After inspec- with mutations downstream of exon 63. There was no dif-
tion of their clinical notes, they were all estimated to have ference in mean age between the genotype subgroups.
an intellectual difficulty, on the basis of reports of their
language, functioning, and presentation in clinic. SCDC
The mean total SCDC score for 130 males with DMD
Neurodevelopmental assessments was 7.8 (95% CI 6.8–8.9), which was significantly higher
Autistic spectrum disorder was assessed with the Develop- (p<0.001) than the mean score of healthy comparison indi-
mental, Diagnostic and Dimensional Interview – short viduals (2.9, 95% CI 2.2–3.6) previously reported,21 indi-
version (3Di-sv).18 The 3Di-sv is a semi-structured, cating greater levels of impairment in this sample with
parent-report, diagnostic interview comprising 53 items. It DMD. The males with mutations predicted to disrupt all
has high interrater reliability and test–retest reliability short/brain dystrophin isoforms, including Dp71 (i.e. with
coefficients.18 The clinical cut-off scores on the 3Di-sv mutations downstream exon 63), had a total mean score of
indexes are reciprocal social interaction impairment, 11.5; 11.7 (95% CI 8.5–15.4); males with mutations disrupting
repetitive and stereotyped behaviour, 5; and communica- Dp260, Dp140, Dp116 (i.e. between exons 31 and 62) had
tion impairment, 8. Children who score above the cut-off a mean score of 7.3 (95% CI 5.9–8.6); and males with
point on the reciprocal social interaction scale and either Dp427 mutations (i.e. upstream of exon 30) had a mean

DMD Neuropsychiatric Profile and Genotype/Phenotype Associations Valeria Ricotti et al. 79

Table I: Raw scores for the Developmental, Diagnostic and Dimensional Interview-short version (3Di-sv), Conners Parent Rating Scale (Conners 3),
Child Behavior Checklist (CBCL), and general intelligence (Wechsler Intelligence Scales for Children – Fourth Edition/Ravens Coloured Progressive
Matrices) in the total population and according to genotype subgroups

Mean score (95% CI)

Upstream of Downstream of
Assessment exon 30 Exons 31–62 exon 63 Total p Post hoc
a
Cognitive function
General Ability Index 96.0 (87.9–104.1) 91.7 (86.7–96.7) 74.8 (62.7–86.9) 91.8 (87.7–95.8) 0.042 31>63
Working Memory Index 86.6 (78.2–94.9) 81.7 (75.8–87.5) 57.5 (24.4–90.7) 81.3 (76.2–86.3) 0.011 31, 31–62>63
3Di-sv
Social 7.6 (6.1–9.2) 8.8 (7.3–10.3) 10.5 (4.9–16.0) 8.6 (7.5–9.7) 0.396
Communication 7.8 (5.9–9.7) 7.76 (6.2–9.2) 10.20 (4.35–16.1) 6.6 (6.9–9.7) 0.516
Repetitive stereotypical behaviour 4.1 (2.8–5.7) 3.2 (2.4–4.0) 4.4 (1.5–7.6) 3.6 (2.8–4.2) 0.387
Conners 3 (parents)
Inattentive 62.1 (57.3–66.9) 59.5 (57.2–67) 63.6 (47.5–79.6) 60.6 (57.1–64.1) 0.706
Hyperactivity 53.4 (50.1–56.9) 53.4 (49.5–58.3) 63.1 (52.6–73.7) 54.5 (51.5–57.5) 0.240
CBCL
Internalizing Problems 9.4 (6.3–12.6) 8 (7.3–11.7) 6.3 (3.5–11.1) 9.5 (7.8–11.3) 0.998
Externalizing Problems 8.31 (5.2–11.3) 9.98 (7.4–12.5) 8.7 (1.5–15.9) 8.6 (7.5–11.2) 0.708
a
Upstream of exon 30, n=25; mutations between exons 31 and 62, n=52; Downstream of exon 63, n=6. Non-verbal patients were not
included.

Table II: Prevalence of cognitive function, neurodevelopmental disorders, and emotional behavioural problems for all mutations and according to
genotype

Genotype subgroups (mutations) n (%; 95% CI)

Upstream of exon Exon 31–62 (Dp260, Downstream of

30 (Dp427) Dp140, Dp116) exon 63 (Dp71) Total, n (%; 95% CI) p

Cognitive function (n=130)

Intellectual disability 6/41 (15; 7–29) 19/75 (25; 17–37) 9/14 (64; 36–85) 34/130 (26; 19–34) 0.002
Neurodevelopmental disorders (n=87)
Scores above threshold for autistic 4/26 (15; 6–35) 11/54 (20; 11–34) 3/7 (43; 13–79) 18/87 (21; 13–31) 0.264
spectrum disorder
Scores above threshold for inattention 13/26 (50; 31–69) 22/54 (41; 28–55) 3/7 (43; 13–79) 38/87 (44; 33–54) 0.753
Scores above threshold for hyperactivity 3/26 (11; 4–31) 14/54 (26; 16–39) 4/7 (57; 21–87) 21/87 (24; 16–34) 0.046
Emotional behavioural problems (n=87)
Scores above threshold for 6/26 (23; 10–44) 14/54 (26; 16–39) 1/7 (14; 2–63) 21/87 (24; 16–34) 0.929
internalizing problems
Scores over threshold for 2/26 (8; 2–27) 10/54 (19; 10–31) 1/7 (14; 2–63) 13/87 (15; 9–24) 0.404
externalizing problems

score of 7.6 (95% CI 5.8–9.3). ANOVA confirmed a sig- A group effect was seen (p=0.011), with a greater change of
nificant group effect (p=0.045). Post hoc testing using Gab- intellectual disability associated with mutations further
riel’s procedure showed that this reflected higher scores of along the gene (Table I). Of the 14 individuals with muta-
males with mutations downstream of exon 63, compared tions affecting Dp71 (i.e. downstream of exon 63), nine
with those with mutations upstream of exon 30 (p=0.024). had intellectual disability (64%) compared with 25% (19/
Also, there was a trend (p=0.071) towards the group with 75) of males with mutations disrupting Dp260, Dp140,
mutations downstream of exon 63 having higher SCDC Dp116, and 15% (6/41) of males with mutation disrupting
scores than males with mutations between exons 31 and only the long dystrophin isoforms Dp427 (Table II).
62.
Neurodevelopmental disorders
Cognitive function Of the 130 participants, 87 completed the full assessment
We observed that 34 out of 130 males with DMD (26%) protocol (UK, n=62; Italy, n=13; Brussels, n=12). There
were estimated to have an intellectual disability. For 19 of was no bias based on clinical presentation in selecting these
these, intellectual disability was assigned on the basis of an 87 males: the primary reason why the remaining males did
IQ score below 70, and for 15 it was estimated from clini- not undergo the full assessment was time constraint during
cal notes as they presented with too severe a language and the routine clinical appointment. The 87 males were repre-
developmental delay to complete an IQ test. Working sentative of the whole cohort: they did not differ signifi-
Memory Index was, on average, strongly affected (Table I). cantly from the remaining individuals in relation to age,

80 Developmental Medicine & Child Neurology 2016, 58: 77–84

IQ, SCDC score, and genotype (all p>0.09). The genotype males with mutations predicted to affect the Dp427 alone,
grouping was as follows: 26 participants with mutations so they inversely correlated with intellectual disability
upstream of exon 30; 54 with mutations between exons 31 (Table II).
and 62; and seven with mutations downstream of exon 63.
In this entire cohort, we observed a high prevalence of Clustering of neurodevelopmental, behavioural, and
individuals who were above the threshold scores on the emotional symptoms
standardized parent-report instruments used. A fifth (21%, In light of the multitude of neurodevelopmental, emo-
n=18) of the males scored in the ASD range, 21 (24%) tional, and behavioural problems observed in our popula-
showed clinically significant hyperactivity, and 38 (44%) tion, we looked at how many of these cluster in the same
had severe difficulties with inattention (Table II). The participant. We found a high prevalence of clustering of
mean raw scores are also reported (Table I). With the psychiatric symptoms: 37% of males (n=32/87) scored in
exception of hyperactivity, where a group difference was the clinical range on more than one measure of emotional,
observed, with mutations at the 30 end of the gene being behavioural, or neurodevelopmental problems (Table III).
more severely affected (p=0.039), we did not find any sta- Twelve males (14%) scored up on four or more of these
tistical difference between groups. However, of clinical measures. Compared with males with mutations affecting
interest, a higher proportion of males with mutations only the long Dp427 isoforms (n=3, 12%), those with a
affecting the shorter isoforms met the threshold scores for mutation affecting the short isoforms (n=17, 28%) were
neurodevelopmental disorders as opposed to males with more than twice as likely to have clinical problems on
mutations only affecting Dp427 (Table II). three or more measures, although this difference was not
significant (p=0.162). In Table SI (online supporting infor-
Emotional behavioural problems mation), we report the correlation within psychiatric
As shown in Table II, emotional and behavioural problems comorbidities. Despite the high prevalence of neuropsychi-
were highly prevalent in our cohort of 87 males with atric disturbances, it is also noteworthy that 32 (37%)
DMD. Around a quarter (n=21) had scores in the clinical males with DMD had no symptoms detected.
range for internalizing problems (i.e. anxious, depressive,
and over-controlled); and approximately one in six (n=13) DISCUSSION
scored in the clinical range for externalizing behavioural We report a large, multicentre, and international cohort
problems (i.e. aggressive, hyperactive, non-compliant, and study using well-validated parent-report measures of neu-
under-controlled.). On the CBCL total problems scale, 15 rodevelopmental, behavioural, and emotional difficulties to
males (17%, 95% CI 10–26) scored above the conventional screen males with DMD with a wide spectrum of genetic
clinical threshold (t≥63), and 13 (15%, 95% CI 8–23) mutations.
scored above a more stringent threshold (t≥67) used in pre- We found a significantly elevated mean total SCDC
vious research with the CBCL in a group of males with score, suggestive of high rates of neurodevelopmental dis-
DMD.22 Unlike intellectual disability and neurodevelop- turbance in this population.10 In relation to general intelli-
mental disorders, internalizing and externalizing problems gence, we confirm previous findings reported in the
were not related to genotype: they were common in the literature: cognitive function was lower than in the general
males with mutations at the 30 end of the gene, and among population, with Working Memory Index most strongly
affected.23,24 Including 15 males who were non-verbal/pro-
foundly delayed, we estimated that 26% (n=34) had intel-
lectual disability. On the basis of detailed and standardized
Table III: Clustering of neurodevelopmental, emotional, and behavioural
parent-reported measures, we observed a high prevalence
problems for all mutations and according to genotype
of neurodevelopmental disorders: 21% of males scored
Mutations Mutations above the threshold for ASD, 24% for hyperactivity, and
between between
exons 1–30, exons 31–79, All mutations,
44% for inattention. These figures are much higher than
Symptoms n (%; 95% CI) n (%; 95% CI) n (%; 95% CI) those in the general paediatric population.25–28 Our study
fits with previous observations on the high prevalence of
0 8 (31; 16–51) 24 (39; 28–52) 32 (37; 27–48)
1 9 (35; 19–55) 14 (23; 14–35) 23 (26; 18–37) ADHD, reported as 32% in a cohort of Italian males with
2 6 (23; 11–44) 6 (10; 4–21) 12 (14; 8–23) DMD assessed according to DSM-IV criteria.8 In relation
3 2 (8; 2–27) 6 (10; 4–21) 8 (9; 5–18) to ASD, previous studies supported an association with
4+ 1 (4; 1–24) 11 (18–10–30) 12 (14; 8–23)
Total n 26 61 87 dystrophinopathies: Hinton et al.7 reported that 19% of
Duchenne/Becker muscular dystrophies met the criteria for
Problems included intellectual disability, autistic spectrum disorder,
attention-deficit–hyperactivity disorder, and internalizing and exter-
ASD in the Autism Diagnostic Interview – Revised. Our
nalizing problems. A total of 32 (37%) males with Duchenne muscu- findings are based on a diagnostic interview with parents;
lar dystrophy across all mutations had a clustering of two or more although the 3Di is a reliable, highly sensitive, and specific
symptoms. Males with mutations affecting the short brain isoforms
had a higher proportion of clustering of three or more symptom
tool for ASD, it may overestimate the diagnosis. Therefore
(28%, n=17) than males with mutations at the 50 end of the gene these findings need further exploration with in-depth
(12%, n=3); however, this was not statistically significant. objective assessments that include direct observation with

DMD Neuropsychiatric Profile and Genotype/Phenotype Associations Valeria Ricotti et al. 81

tools such as the Autism Diagnostic Observation Schedule. In relation to the detailed neurodevelopmental assessment
However, it is undisputable that ADHD and ASD have of a subsample (n=87), our analyses were somewhat limited
traditionally been an underdiagnosed comorbidity of by the small numbers of males with mutations affecting the
DMD, which is a concern given the high levels of func- 50 end of the gene, which meant we only had the statistical
tional impairment associated with both disorders. power to detect very large effects. Nevertheless, in com-
Emotional behavioural problems, such as anxiety, affec- mon with previous findings for ADHD, we observed
tive disorder, and oppositional/aggressive behaviour, have higher levels of hyperactivity among males with the 50
also been recognized in DMD.22,29 One study reported mutation. We observed a non-significant trend for these
that on a screening behavioural questionnaire, 32% of same males to have a higher incidence of ASD, which will
DMD families responded with scores above the normal be worth investigating in larger samples.
cut-off, compared with 0% to 2% of other disabling neu- In contrast to neurodevelopmental problems, emotional
romuscular disorders, including spinal muscular atrophy.30 behavioural problems were evenly distributed across the
Only a small proportion of these patients with DMD had genotype subgroups. Our findings support the notion that
been investigated for neurodevelopmental disorders. In our mutations towards the 30 end of the dystrophin gene,
large cohort with DMD, we observed that families which disrupt not only the long products, but also in turn
reported also a high prevalence of internalizing and exter- the short brain-expressed isoforms Dp140 and Dp71, have
nalizing problems, 24% and 15% respectively. These find- a more devastating effect on the neurocognitive phenotype
ings reflect the high level of anxiety and mood disorders, (Table II).4,5 Furthermore, the long dystrophin products
alone or in combination, observed in clinical practice. Dp427 must also play a significant role in the association
Although not uncommon for neurodevelopmental disor- of CNS comorbidities. Proximal mutations are sufficient to
ders to cluster together,25 in our cohort with DMD we cause cognitive and neurobehavioural problems, and are
observed a high prevalence of combination of neuropsychi- therefore important for the neurodevelopment of the brain.
atric symptoms: over a third of males presented with at This should not be surprising considering that Dp472 is
least two comorbidities, and seven males with DMD with largely expressed in the cortex, hippocampus, and cerebel-
intellectual disability (8%) scored in the range of all four lum, localizing to neuronal GABAergic synapses. In the
disorders assessed. Such striking patterns of clustering of mdx mouse, lack of Dp427 is associated with reduced
symptoms suggest the existence of a ‘DMD neuropsychi- receptor clustering, impairing specific amygdala GABAer-
atric syndrome’, which in clinical practice has a significant gic transmission and enhancing defensive behaviour in
impact on the condition and its management. In future, it response to danger.32 In contrast, Dp140 localizes in glial
will be important to evaluate DMD further with objective cells and Dp71 is expressed in the perivascular astrocytes,
neuropsychiatric inventories, and to use non-DMD neu- which is required for anchoring aquaporin-4. Loss of
rodevelopmental comparison groups to see if there is any Dp71 is associated with reduced levels of aquaporin-4, thus
distinctive pattern of clustering observed in DMD. altering transmembrane water permeability.33,34
Mean life expectancy for DMD is now in the late twen- We could speculate that, when in addition to Dp472 the
ties, with many individuals surviving into the third and short dystrophin products are also disrupted, a ‘compen-
even the fourth decade of life; it is anticipated this will fur- satory/protective’ mechanism produced by these abun-
ther shift in future generations. Therefore, improving neu- dantly CNS-expressed isoforms may be lost, increasing the
ropsychological well-being is of paramount importance for risk of associated neuropsychiatric disorders and account-
participation and quality of life. Intervening from early ing for the more devastating profile observed in males with
stages, during the crucial initial years of schooling, target- Dp71 mutations.
ing specific intellectual disability and/or behavioural issues, Yet much remains unclear about the effects of dystrophin
first requires prompt and appropriate recognition and disruption in the CNS, and how different isoforms con-
understanding of the disorder. tribute to pathology. Structural abnormalities in the brain
In our study, we explored further a relationship between have been recently described: changes in the cerebral grey
genotype (i.e. mutation location along the gene) and neu- matter volume and white matter microstructure in males
ropsychiatric profile. On the 12-item SCDC questionnaire with DMD have been described using quantitative magnetic
completed by the families, we found that males with Dp71 resonance imaging.35 Previous studies using magnetic reso-
mutations (i.e. downstream of exon 63) had overall the nance spectroscopy have indicated altered phosphorus
highest SCDC mean total score (11.7), with more than half metabolite ratios in the cerebral cortex of 19 males with
of the males scoring in the abnormal range of at least 8. In DMD.36 In an animal model, disrupted cerebral diffusivity
relation to intelligence, confirming previous reports,5,31 was demonstrated to be associated with lack of dystrophin.37
males with DMD with mutations affecting an increasing Finally, novel pharmaco-gene therapies aiming at restor-
number of shorter isoforms had a higher frequency of ing dystrophin expression and delaying the course of the
intellectual disability (p=0.002), with 63% of Dp71 muta- disease progression are currently in different phases of
tions being most severely affected. Working Memory Index experimentation. Some of these compounds, such as the
also showed a significant group effect (p=0.011), with the tricyclo-DNA antisense molecules, are also capable of
group missing the short isoform being the worst affected. crossing the blood–brain barrier and restoring dystrophin

82 Developmental Medicine & Child Neurology 2016, 58: 77–84

expression through exon skipping in the CNS. This mit, and Santhera Pharmaceutical, receives research support from
approach demonstrates clinical benefit and behavioural Trophos and GlaxoSmithKline, and has received funding for trials
amelioration in the mdx mouse model, indicating that at from AVI, and PTC. DHS is a stockholder in IxDx Ltd., which
least some aspects of the dystrophin-associated neurodevel- owns exclusive rights to 3Di software and to the dissemination of
opmental syndrome might be reversible upon restoration 3Di technology and intellectual property. EM is site principal
of brain dystrophin.38 investigator for the PTC extension study of Ataluren in DMD, for
In conclusion, our findings fuel the growing interest in the TROPHOS clinical trial on SMA, and for the GSK study on
unravelling the as yet unclear role that the disruption of exon skipping. He also receives funds from the Italian Telethon and
different dystrophin isoforms may play in brain develop- SMA Europe. He has acted on the advisory board for Shire, PTC
ment and function. They provide a baseline to evaluate if Therapeutics, and Prosensa. The other authors have stated that
these events can be reversed and to what degree, with they had no interests that might be perceived as posing a conflict or
emerging dystrophin-restoring therapies. bias. VR, WPLM, MS, EM, DHS, and FM oversaw the design of
the study. VR, MS, KE, SM, ND, MP, SLF, EL, SB, and GV con-
A CK N O W L E D G E M E N T S tributed to data collection. VR and WPLM analysed the data. VR
The following authors also contributed to the writing of this paper: wrote the first draft of the manuscript; WPLM contributed to the
Kirsty Entwistle, Stefania La Foresta, Lien Everaert, Simon Bajot, writing of the manuscript. VR, WPLM, MS, KE, SM, ND, MP,
and Gianluca Vita. FM is supported by the Great Ormond Street SLF, EL, SB, SAR, EM, DHS, and FM contributed to the revision
Hospital Children’s Charity and GOSH Biomedical Research cen- of the manuscript.
tre. The support of the Muscular Dystrophy Campaign Centre
Grant to the Dubowitz Neuromuscular Centre is acknowledged. SUPPORTING INFORMATION
The financial support of L’Association Francßaise contre les Myopa- The following additional material may be found online:
thies is also acknowledged (to VR). FM has served on scientific Table SI: Pearson Correlation between neurodevelopmental
advisory boards for AcceleronPharma, Genzyme, AVI BioPharma, disorders, emotional behavioural problems, and intellectual dis-
Debiopharma Group, GlaxoSmithKline, Prosensa, Servier, Sum- ability.

REFERENCES
1. Bushby K, Finkel R, Birnkrant DJ, et al. Diagnosis and of attention-deficit hyperactivity disorder (ADHD), aut- 18. Santosh PJ, Mandy WP, Puura K, Kaartinen M, War-
management of Duchenne muscular dystrophy, part 1: ism spectrum disorder, and obsessive–compulsive disor- rington R, Skuse DH. The construction and validation
diagnosis, and pharmacological and psychosocial man- der. J Child Neurol 2008; 23: 477–81. of a short form of the developmental, diagnostic and
agement. Lancet Neurol 2010; 9: 77–93. 10. Skuse DH, Mandy WP, Scourfield J. Measuring autistic dimensional interview. Eur Child Adolesc Psychiatry 2009;
2. Pane M, Scalise R, Berardinelli A, et al. Early neurode- traits: heritability, reliability and validity of the Social 18: 521–4.
velopmental assessment in Duchenne muscular dystro- and Communication Disorders Checklist. Br J Psychiatry 19. Conners CK, Sitarenios G, Parker JD, Epstein JN. The
phy. Neuromuscul Disord 2013; 23: 451–5. 2005; 187: 568–72. revised Conners’ Parent Rating Scale (CPRS-R): factor
3. Muntoni F, Torelli S, Ferlini A. Dystrophin and muta- 11. Skuse DH, Mandy W, Steer C, et al. Social communica- structure, reliability, and criterion validity. J Abnorm
tions: one gene, several proteins, multiple phenotypes. tion competence and functional adaptation in a general Child Psychol 1998; 26: 257–68.
Lancet Neurol 2003; 2: 731–40. population of children: preliminary evidence for sex-by- 20. Achenbach T. Manual for the Child Behavior Checklist/
4. Felisari G, Martinelli Boneschi F, Bardoni A, et al. Loss verbal IQ differential risk. J Am Acad Child Adolesc Psy- 4 – 18 and 1991 Profile. Burlington, VT: University of
of Dp140 dystrophin isoform and intellectual impairment chiatry 2009; 48: 128–37. Vermont Department of Psychiatry, 1991.
in Duchenne dystrophy. Neurology 2000; 55: 559–64. 12. Wechsler D. Weschler Intelligence Scale for Children – 21. Skuse D, Lawrence K, Tang J. Measuring social-cogni-
5. Daoud F, Angeard N, Demerre B, et al. Analysis of Fourth Edition (WISC-IV). London: Pearson Assess- tive functions in children with somatotropic axis dys-
Dp71 contribution in the severity of mental retardation ment, 2004. function. Horm Res 2005; 64(Suppl. 3): 73–82.
through comparison of Duchenne and Becker patients 13. Raiford SE, Weiss LG, Rolfhus E, Coalson D. WISC- 22. Fee RJ, Hinton VJ. Resilience in children diagnosed
differing by mutation consequences on Dp71 expression. IV General Ability Index (WISC-IV Technical Report with a chronic neuromuscular disorder. J Dev Behav
Hum Mol Genet 2009; 18: 3779–94. No. 4). 2005. Available from: http://images.pearsonclini- Pediatr 2011; 32: 644–50.
6. Lorusso ML, Civati F, Molteni M, Turconi AC, Breso- cal.com/images/assets/WISC-IV/80720_WISCIV_Hr_ 23. Cotton S, Voudouris NJ, Greenwood KM. Intelligence
lin N, D’Angelo MG. Specific profiles of neurocognitive r4.pdf (accessed 1 October 2011), and Duchenne muscular dystrophy: full-scale, verbal,
and reading functions in a sample of 42 Italian boys 14. Wechsler D. Wechsler Preschool and Primary Scale of and performance intelligence quotients. Dev Med Child
with Duchenne muscular dystrophy. Child Neuropsychol IntelligenceTM. London, UK: Pearson, 2004. Neurol 2001; 43: 497–501.
2013; 19: 350–69. 15. Raven J. Raven’s Coloured Progressive Matrices. Lon- 24. Hinton VJ, De Vivo DC, Nereo NE, Goldstein E, Stern
7. Hinton VJ, Cyrulnik SE, Fee RJ, et al. Association of don, UK: Pearson, 2008. Y. Selective deficits in verbal working memory associ-
autistic spectrum disorders with dystrophinopathies. Pe- 16. Strauss E, Sherman EMS, Spreen O. A Compendium of ated with a known genetic etiology: the neuropsycholog-
diatr Neurol 2009; 41: 339–46. Neuropsychological Tests: Administration, Norms, and ical profile of duchenne muscular dystrophy. J Int
8. Pane M, Lombardo ME, Alfieri P, et al. Attention defi- Commentary. New York, NY: Oxford University Press, Neuropsychol Soc 2001; 7: 45–54.
cit hyperactivity disorder and cognitive function in 2006. 25. Simonoff E, Pickles A, Charman T, Chandler S, Loucas
Duchenne muscular dystrophy: phenotype-genotype cor- 17. American Psychiatric Association. Diagnostic and Sta- T, Baird G. Psychiatric disorders in children with autism
relation. J Pediatr 2012; 161: 705–9. tistical Manual of Mental Disorders: DSM-IV-TR. spectrum disorders: prevalence, comorbidity, and associ-
9. Hendriksen JG, Vles JS. Neuropsychiatric disorders in Washington, DC: American Psychiatric Association, ated factors in a population-derived sample. J Am Acad
males with duchenne muscular dystrophy: frequency rate 2000. Child Adolesc Psychiatry 2008; 47: 921–9.

DMD Neuropsychiatric Profile and Genotype/Phenotype Associations Valeria Ricotti et al. 83

26. Taylor B, Jick H, Maclaughlin D. Prevalence and inci- 30. Darke J, Bushby K, Le Couteur A, McConachie H. Sur- choroid plexus and microvasculature of the brain. Neuro-
dence rates of autism in the UK: time trend from 2004– vey of behaviour problems in children with neuromuscu- science 2004; 129: 403–13.
2010 in children aged 8 years. BMJ Open 2013; 3: lar diseases. Eur J Paediatr Neurol 2006; 10: 129–34. 35. Doorenweerd N, Straathof CS, Dumas EM, et al. Re-
e003219. 31. Taylor PJ, Betts GA, Maroulis S, et al. Dystrophin gene duced cerebral gray matter and altered white matter in
27. Polanczyk G, de Lima MS, Horta BL, Biederman J, Ro- mutation location and the risk of cognitive impairment boys with Duchenne muscular dystrophy. Ann Neurol
hde LA. The worldwide prevalence of ADHD: a system- in Duchenne muscular dystrophy. PLoS ONE 2010; 5: 2014; 76: 403–11.
atic review and metaregression analysis. Am J Psychiatry e8803. 36. Tracey I, Scott RB, Thompson CH, et al. Brain abnor-
2007; 164: 942–8. 32. Sekiguchi M, Zushida K, Yoshida M, et al. A deficit of malities in Duchenne muscular dystrophy: phosphorus-
28. Hiscock H, Bayer JK, Lycett K, et al. Preventing mental brain dystrophin impairs specific amygdala GABAergic 31 magnetic resonance spectroscopy and neuropsycho-
health problems in children: the Families in Mind popu- transmission and enhances defensive behaviour in mice. logical study. Lancet 1995; 345: 1260–4.
lation-based cluster randomised controlled trial. BMC Brain 2009; 132: 124–35. 37. Goodnough CL, Gao Y, Li X, et al. Lack of dystrophin
Public Health 2012; 12: 420. 33. Nicchia GP, Rossi A, Nudel U, Svelto M, Frigeri A. results in abnormal cerebral diffusion and perfusion
29. Hendriksen JG, Poysky JT, Schrans DG, Schouten EG, Dystrophin-dependent and -independent AQP4 pools in vivo. NeuroImage 2014; 102P2: 809–16.
Aldenkamp AP, Vles JS. Psychosocial adjustment in are expressed in the mouse brain. Glia 2008; 56: 869–76. 38. Goyenvalle A, Griffith G, Babbs A, et al. Functional
males with Duchenne muscular dystrophy: psychometric 34. Haenggi T, Soontornmalai A, Schaub MC, Fritschy JM. correction in mouse models of muscular dystrophy using
properties and clinical utility of a parent-report ques- The role of utrophin and Dp71 for assembly of different exon-skipping tricyclo-DNA oligomers. Nat Med 2015;
tionnaire. J Pediatr Psychol 2009; 34: 69–78. dystrophin-associated protein complexes (DPCs) in the 21: 270–5.

2016 Annual Meeting

European Society for the
Study of Tourette Syndrome
June 8-11, 2006 / Warsaw, Poland
9th European Conference on Tourette Syndrome and
5th International Meeting of Tourette Syndrome Support
and Advocacy Groups

http://tourette-eu.org

84 Developmental Medicine & Child Neurology 2016, 58: 77–84