Selection of Studies
A total of 16 studies (see Table 1) was identified for inclusion in the current review that reported on a total of 620 participants with diagnoses of MPS III. One study (Paper 15) involved two control groups, MPS I-Hurler Syndrome with eight participants and MPS IIIA with nine participants. Of the 620 participants with MPS III (see Table 2), 286 had confirmed sub-type A (MPS IIIA), 124 sub-type B (MPS IIIB), 56 had sub-type C (MPS IIIC), nine had sub-type D (MPS IIID) and 145 did not specify which sub-type. Four of the studies originated from the USA (Papers 2, 9, 11 and 15). Case report studies (Papers 3, 7, 9, 12, 13, 14 and 16) originated from Saudi Arabia, the Netherlands, Poland, Israel, India and Turkey, whereas cross-sectional studies (Papers 1, 2, 4, 5, 6, 8 and 10) originated from Denmark, the UK, the Netherlands, Sweden, France and Spain.
The studies varied in their participant descriptions (see Table 2). All studies reported gender and age; however, only nine studies provided information surrounding ethnicity or nationality (Papers 3, 5, 6, 7, 8, 9, 12, 13 and 16). Half of the studies (Papers 1, 5, 7, 8, 9, 12, 13 and 16) provided additional data regarding the relationship between participants and their families. Paper 1 noted that 41 of 73 participants were from sibling relationships and Paper 8 reported on 70 of 111 participants who were from multiplexFootnote 1 families. Consanguinity was another factor to which some studies referred to. Paper 5 highlighted that six of their 20 participants were from consanguineous families and Paper 13 described the case of a child from a consanguineous family. Whilst these two studies noted consanguinity, other studies (Papers 7, 9, 12 and 16) explicitly stated that participants were from non-consanguineous families.
Of the 16 studies, seven studies (Papers 1, 2, 4, 5, 6, 8 and 10) described and analysed cognitive, behavioural and motor difficulties within MPS III (Table 2). Seven of the studies were case reports (Papers 3, 7, 9. 12, 13, 14 and 16) which described presentations of MPS III and issues pertaining to its diagnosis and misdiagnosis. Paper 15 compared participants with MPS IIIB to participants with MPS IIIA and MPS IH (Hurler syndrome) and two studies (Papers 11 and 15) formally assessed symptoms of ASD in participants with MPS IIIA and MPS IIIB.
Scores on the QATSDD ranged between 22 and 74% with a mean score of 49% (see Table 3).
As previously mentioned, almost half of the studies (n = 7) reviewed were single case reports (Papers 3, 7, 9. 12, 13, 14 and 16); consequently, these were mostly rated as being of moderate methodological quality. In contrast, other studies were group-level studies (Papers 5, 11 and 15), with the three highest methodological quality studies coming from the USA (Papers 2, 11 and 15). Paper 16 obtained the lowest rating owing to a lack of information surrounding the choice of data collection, brief reference to the theoretical framework, justification for method and analysis and a lack of reference to the study’s strengths and limitations. In all 16 studies there was a lack of discussion surrounding the involvement of service users in their study designs.
Across the studies, seven obtained a moderate score for quality (Papers 1, 3, 5, 6, 12, 13 and 14) and eight were deemed to be of good quality (Papers 2, 4, 7, 8, 9, 10, 11 and 15). There was no observed relationship between quality and year of publication. Furthermore, as can be expected cross-sectional designs scored consistently better than the other designs reviewed. As this is the first review of the literature surrounding the symptoms of ASD in MPS III, all studies were retained to present a comprehensive picture of the available research.
Symptoms of ASD
All of the 16 studies referred to behaviours characteristic of ASD, with ten studies (Papers 3, 5, 6, 7, 8, 11, 12, 13, 14 and 15) making a specific reference to ASD (Table 1). However, whilst these studies made specific reference to ASD, some case reports described observed behaviours in detail (Papers 9 and 14). In addition, some studies used the term ‘autistic features’ or ‘autistic like’ and failed to provide detailed descriptions (Papers 3,7 and 13). The lack of detailed descriptions was also noted in some of the cross-sectional studies (Papers 2, 4, 6, 8 and 10). For example, a relative strength of Paper 8 was its larger sample size when compared to the other studies, yet it did not describe symptoms of ASD in great detail.
When ASD-related behaviours and symptoms were specified and described in studies, these were categorised as:
Speech, language and communication difficulties Despite differing sample sizes, study designs and quality ratings, all 16 studies consistently reported difficulties with speech, language and communication (Table 2) with nine studies making explicit reference to this as a feature of ASD (Papers 3, 5, 7, 8, 11, 12, 13, 14 and 15). Difficulties in this domain included delayed language and speech development, limited vocabulary, no speech, echolalia, variable or no eye contact and impaired communication skills. The studies differed in reporting of age of onset of these behaviours, which were primarily reported as emerging after 18 months. When participants were assessed with a ‘gold-standard’ ASD assessment tool, the Autism Diagnostic Observation Schedule (ADOS) (Lord et al. 1999), they consistently met ADOS diagnostic criteria on communication domains (Papers 11 and 15).
Repetitive and restricted behaviour Two case reports reported repetitive and restricted behaviours. Paper 7 described a female participant (aged 57 years) with MPS IIIB who exhibited repetitive behaviour during adulthood, whereas Paper 14 described a 7-year-old girl (sub-type of MPS III not determined), presenting with repetitive hand clapping and head banging from the age of 2 years. However, neither study used standardised assessment tools. Two larger studies with sample sizes of 21 (Paper 11) and ten (Paper 15) utilised the ADOS, but reported little repetitive or restricted behaviour.
Social difficulties Ten studies observed social difficulties typical of an ASD presentation and these included aggression in social situations, peer difficulties and difficulties making personal contacts, social immaturity and impaired social interaction (Papers 1, 2, 3, 4, 5, 7, 8, 9, 11 and 14). As previously noted, not all studies adequately identified the age of the emergence of these behaviours, but those that did highlighted that these behaviours were evident from the age of 3 years. Four of the ten studies (Papers 2, 4, 11 and 15) used standardised assessment tools, such as the ADOS and Rutter’s parent checklist (Rutter et al. 1970), to formally assess for such behaviours, but the remaining six studies (a) failed to identify how behaviours were assessed and (b) failed to identify the assessment tool or (c) relied on professional or parent impression.
Diagnosis of ASD
As previously mentioned, two studies (Papers 11 and 15) utilised the ADOS to assess behaviours of children with confirmed diagnoses of MPS IIIA and MPS IIIB; both studies were rated as having good methodological quality. Paper 11 concluded that 13 of the 21 children aged between 1.8 and 8.8 years with MPS IIIA met diagnostic criteria for ASD (as measured by module 1 of the ADOS) and that this was strongly associated with age; 11 children aged over 3.8 years met ADOS diagnostic criteria, compared with only two of ten children aged less than 3.8 years meeting ADOS diagnostic criteria. Paper 15 examined ASD in children with MPS IIIB and concluded that nine of their ten participants met ADOS criteria for ASD between the ages of 6 and 24 years. Both studies reported increased incidences of social/affective behaviour than restricted or repetitive behaviour. As part of their case reports, Paper 9 and Paper 15 described participants who presented with hyperactivity, poor social interactions and repetitive behaviour who consequently received diagnoses of pervasive developmental disorder prior to receiving a diagnosis of MPS III. In a large study of MPS III participants, Paper 8 noted “autism related symptoms” at time of diagnosis of MPS III in 66% of the UK MPS III population (n = 126) and in 76% of the French MPS III population (n = 128), but did not state whether any formal diagnoses of ASD had been made.
Method of Behavioural Assessment
Studies that obtained higher scores on the QATSDD were noted to utilise formal questionnaires, standardised assessment tools and reported reliability and validity. Only two of the studies (Papers 11 and 15) assessed symptoms of ASD formally using the Autism Diagnostic Observation Schedule (ADOS) which was a strength of these studies. These studies more reliably and validly identified symptoms of ASD in individuals with MPS III. Paper 2 reported on the use of several validated and reliable assessment tools, including the Bayley Scale of Infant Development (Bayley 1969), the Stanford-Binet Intelligence Scale (Terman and Merrill 1973) and the Vineland Social Maturity Scale (Doll 1965). The findings of this paper suggested that children with MPS III begin to deteriorate cognitively between the ages of 3.5 and 6.5 years, lose language by age 8 years and demonstrate self-stimulatory behaviour and experience peer difficulties (Nidiffer and Kelly 1983). Whilst these tools do not specifically assess for ASD or repetitive behaviour and restricted interests, they do examine behaviours, such as language and social difficulties. A significant weakness of some studies (Papers 7 and 9) was failing to identify how behaviours were assessed at all, whereas other studies (Papers 6, 12, 13, 14 and 16), often of lesser quality (as assessed by the QATSDD) than the above-mentioned studies, referred to clinician and parent observation, impression and reports, which resulted in less valid conclusions and less robust study designs.
Implications of Symptoms of ASD
Seven of the 16 studies (Papers 3, 5, 6, 9, 12, 13 and 16) reported initial misdiagnosis (Table 2) including misdiagnoses of ASD, Attention Deficit Hyperactivity Disorder (ADHD), acquired language disorder and intellectual disability (reported as ‘mental retardation’). Such misdiagnoses were primarily observed in single-case reports and illustrated the phenomenological overlap between the behavioural phenotype of these disorders and MPS III.
Seven of the 16 studies (Papers 2, 4, 5, 6, 7, 9 and 10) noted that a possible initial focus on problematic behaviour and developmental delay resulted in late diagnoses of MPS III. The larger sample sizes in five of these seven studies (Papers 2, 4, 5 and 6) suggest that this can be taken as a relatively reliable and consistent finding across presentations of MPS III. Three of the seven studies (Papers 5, 10 and 15) noted that sub-types MPS IIIB and IIIC were prone to late diagnosis owing to slow progression and attenuated phenotypes.