ORIGINAL CONTRIBUTION
Medical conditions affect the outcome of early intervention in preschool children with autism spectrum disorders
reportedMats Anders Eriksson •Joakim Westerlund •A
˚sa Hedvall •Per A ˚mark •Christopher Gillberg •
Elisabeth Fernell
Received:1April 2012/Accepted:14July 2012/Published online:27July 2012ÓSpringer-Verlag 2012
Abstract The aim was to explore the frequency of genetic and other medical conditions,including epilepsy,in a population-based group of 208preschool children with early diagnosis of Autism spectrum disorders (ASD)and to relate outcome at a 2-year follow-up to the co-existing medical findings.They had all received early intervention.The Vineland Adaptive Behaviour Scales (VABS-II)composite score served as the primary outcome measure.In the total group,38/208children (18%)had a significant medical or genetic condition.Epilepsy was present in 6.3%at the first assessment and in 8.6%at follow-up and was associated with more severe intellectual impairment.A history of regression was reported i
n 22%.Children with any medical/genetic condition,including epilepsy,as well as children with a history of regression had significantly lower VABS-II scores at the 2-year follow-up.Children with a medical/genetic condition,including epilepsy,had been diagnosed with ASD at an earlier age than those without such conditions,and early age at diagnosis also
correlated negatively with adaptive functioning outcome.The results underscore the importance of considering medical/genetic aspects in all young children with ASD and the requirement to individualize and tailor interven-tions according to their specific needs.
Keywords Autism spectrum disorder (ASD)ÁMedical/genetic disorders ÁEpilepsy ÁEarly intervention ÁOutcome
Introduction
The aim of the present study was to explore the frequency of genetic and other medical conditions,including epi-lepsy,in a representative group of preschool children with early diagnosis of Autism Spectrum Disorders (ASD)and to relate outcome to co-existing medical findings.
According to several studies,early intervention has been reported to improve outcome in children with
ASD [14,22].However,these studies have demonstrated considerable variability in outcome with low initial IQ repeatedly dem-onstrated to contribute to a less favourable outcome [23].It is important to remember that in many of these studies children with co-existing genetic and medical disorders,such as epilepsy,have been excluded [12,14,23,48].Autism spectrum disorders are highly heritable neuro-developmental disorders affecting about 1%of the child-hood population [6,18].The etiological heterogeneity in ASD is extensive.In about 10–20%of all ASD cases,a genetic cause is identified [1,10].More than 100disease genes have been reported in individuals with ASD [9].In children with ASD conventional karyotyping identi-fies abnormalities in about 2%[42]and Fragile X syn-drome is identified in about 2%[38].With the use of whole genome array-based comparative genomic hybridization
M.A.Eriksson ÁP.A
˚mark Department of Women’s and Children’s Health,Karolinska Institutet,Stockholm,Sweden
M.A.Eriksson (&)
Department of Neuropaediatrics,Astrid Lindgren’s Children’s Hospital,Karolinska University Hospital,17176Stockholm,Sweden e-mail:iksson@ki.se
J.Westerlund
Department of Psychology,Stockholm University,Stockholm,Sweden
A
˚.Hedvall ÁC.Gillberg ÁE.Fernell Gillberg Neuropsychiatry Centre,Sahlgrenska Academy,Gothenburg University,Gothenburg,Sweden
Eur Child Adolesc Psychiatry (2013)22:23–33DOI 10.1007/s00787-012-0312-7
(array-CGH),sub-microscopic deletions and duplications (copy number variations;CNVs)have been increasingly identified in ASD as well as in intellectual disability(ID) [32]and epilepsy[30].In children with ASD,about7% have been reported to have causative CNVs[41,45].
Autistic features or ASD and/or epilepsy occur in sev-eral well-recognized single-gene disorders such as Tuber-ous sclerosis,Rett syndrome,Fragile X,Smith-Lemli-Opitz and Angelman syndrome[10].Although genetic susceptibility is of major importance,environmental and epigenetic factors may influence the risk of ASD.The (relatively small)discordance in monozygotic twins sug-gests the presence of non-heritable,pre-and perinatal risk factors for autism[26,39].
Several prenatal and perinatal risk factors have been associated with ASD.Intrauterine congenital central ner-vous system viral infections and exposure to various agents such as sodium valproate may increase the risk for cogni-tive impairment,including ASD[15,33,37].Epidemio-logical studies have not yet found any association between alcohol consumption and increased risk for ASD[21],but in two Swedish studies,9–13%of children whose mothers abused alcohol in pregnancy had ASD[4,27].Regarding perinatal factors,several studies have shown increased risk for cognitive problems,including ASD in extremely pre-term and low birth weight children[13,24,29].
The prevalence of epilepsy in childhood in general is 0.5–0.7%.Epilepsy,much like ASD,can be viewed as a spectrum disorder with multiple aetiologies and highly variable clinical presentation[25].Cognitive deficits are prevalent in children with epilepsy with approximately 25%having mental retardation[7].The prevalence esti-mates of epilepsy in children with ASD vary from5to 46%[43,46].The rates of epilepsy in ASD increase with age and the association between ASD and epilepsy is closely linked to the presence and degree of cognitive impairment.In a meta-analysis by Amiet et al.[3],a pooled frequency of epilepsy was21.4%in patients with autism. The rate of epilepsy in subjects without cognitive impair-ment was8%,clearly a higher prevalence than in the normal population.In children with intellectual disability and active epilepsy co-existing ASD is found in37%[44]. It is well established that early onset
epilepsy increases the risk of ASD.Saemundsen[40]found that14%of children with epilepsy onset in thefirst year of life developed aut-ism.In children with infantile spasms,the risk for ASD was46%and in children with more severe structural brain disorder(hypoxic ischemic encephalopathy,cortical dys-plasia,tuberous sclerosis)69%were found to have ASD. The author’s conclusion was that the risk of developing autism is probably mostly predicted by the origin of the seizures and,usually,not a consequence of an epileptic encephalopathy.
Developmental regression with loss of acquired skills is reported in children with ASD at rates ranging from15to 40%depending on the definition of regression and sample of children[28,31].The most consistent documented finding in such cases is loss of verbal skills,but this is often accompanied or preceded by loss of social skills.The onset of regression is often reported to be between the ages of15 and24months.Previous studies have not found epilepsy, seizures or subclinical EEG abnormalities to be more prevalent in children with‘‘autistic regression’’compared to non-regressive autism[5,19].Children with autistic regression tend to have poorer cognitive outcome com-pared to children without regression[8].In a subset of children with regressive autism,early development may be delayed and some children have social deficits before the regression is identified[35].To our knowledge,no previ-ous study has explored the influence of co-occurring medical conditions or a history of developmental regres-sion in relation to early intervention.
Method
Participants
The study included208children,20–54-month-old,when referred to a specialized habilitation centre with a diagnosis of ASD.They were drawn from a population-based group of313children(born2002–2006)with ASD diagnoses in Stockholm County,a region with approximately28,000 births per year.Of these313children,288had been referred for early intervention to a specialized habilitation centre for preschool children with ASD.Twenty-five children with more severe medical conditions and syn-dromes in combination with ASD were referred to general habilitation centres due to more complex needs.At these centres,no specific ASD interventions were provided.This group was not included in the outcome study.
Of the288children,24had been referred to the centre prior to study start and could not be included.Of the remaining264children,37families declined participation, 15families could not communicate in either Swedish or English,two children moved abroad and another two were referred back to general habilitation centres because they had complex medical needs.This means that208children(176 boys and32girls)were included in this prospective two-year follow-up study(Fig.1).No exclusion of children w
ith ASD and identified medical conditions/or significant ID was done once type of intervention had been initiated.However,it is clear that the study group did not include some of the most medically‘‘complex’’cases with ASD in the population.
Of the total group,64(31%)had been referred before the age of36months,83(40%)between37-48months and61(29%)between the ages of49–54months.
Of the208children,198participated in the two-year follow-up.Complete cognitive data was obtained in196, and data regarding type of ASD diagnosis was obtained in197[17].Autistic disorder/childhood autism was diagnosed in105(53%),atypical autism in58(29%), Asperger syndrome in13(7%),and autistic traits in21 (10%)at follow-up after2years.Four of those with autistic traits(2%of the total group)had ID with autistic traits and17(8%)had attention/language impairment with autistic traits.Of the196children,95(48%)had ID (IQ\70),50(26%)had low normal/borderline intel-lectual function(IQ70–84),and51(26%)average or above average intellectual level(IQ C85).In12children with incomplete cognitive assessment at the follow-up, the cognitive level was classified according the level defined at referral(5children with ID,5with develop-mental delay/borderline IQ and2with average intellec-tual level).Measures
At referral to the habilitation centre,Time1(T1),data were collected from each child’s maternity and obstetric care units,the Child Health Care centre(CHC)and all hospital and outpatient clinics attended by the child.Both at T1and at the two-year follow-up,Time2(T2)the paediatrician/ child psychiatrist in the research team examined all chil-dren.Details regarding interview protocols,including cognitive assessments and questionnaires at T1and T2have been reported in previous publications[16,17].
In addition to data from records,parents were inter-viewed regarding the child’s pre-and perinatal history, early development,and any genetic,neurologic or other clinically significant medical condition,including epilepsy.
Regression was defined as loss of more thanfive spoken words used communicatively in children more
than
15months of age.In children younger than15months, regression was determined when there was a clear indica-tion of loss of social interest and contact.Parents were interviewed(in close temporal connection to thefirst diagnostic assessment)regarding a history of regression in the child and this information was compared to available data in CHC and medical records.Consistency was required between parental information and the notes in the record’s from CHC.
Participating families were recommended genetic test-ing using array-based Comparative Genomic Hybridization (array-CGH)(thesefindings will be described in detail in a forthcoming paper)and testing for Fragile X.Conventional chromosomal karyotyping had been performed in many children.In girls with a clinical suspicion of Rett syndrome mutation for MECP2were analysed.
Medical and genetic conditions were defined as:(1)a significant intrauterine harmful exposure,(2)substantial prematurity(gestational age less than29weeks),(3)an identified genetic disorder including significant genomic imbalances identified with array CGH,(4)abnormal brain MRIfindings and(5)a clinical diagnosis of epilepsy.
At T2,the parents of198children were interviewed a second time to collect information about any new
medical condition that had affected the child during the observation time and208hospital and outpatient medical records were once again reviewed.At that time,a new cognitive assessment was carried out to establish the child’s Devel-opmental Quotient/Intelligence Quotient(DQ). Intervention
All208children received intervention based on princi-ples of Applied Behaviour Analysis(ABA).The study design was naturalistic.There was no randomization to treatment type.Parental preference was the most decisive factor.The presence of medical/genetic condition did not affect choice of intervention type.In many children,a medical condition had not yet been identified.One group (n=93)received‘‘intensive’’and the remainder (n=105)‘‘non-intensive’’targeted intervention.The different levels of interventions were defined as:Inten-sive intervention based on ,early intensive behavioral intervention(EIBI),given at the preschool and by the parents at home with assistance from the center,with the intention to treat(a)15h or(b)25–30h per week or30–40h per week.Non-intensive targeted intervention based on ABA,consisting of different tar-geted types of training(toilet training,speech and lan-guage training,training of compliance or other specific training that the child was deemed to need).Details of these interventions have been reported in a previous paper[17].
Outcome measure
The primary outcome measure was change in Vineland Adaptive Behaviour Scale(VABS-II)composite scores between T1and T2.VABS II is a valid and often used outcome measure of adaptive skills in children with ASD [14,20].A total of192children had complete VABS-II scores both at T1and T2.Numbers of children in the outcome analyses reflect the number of individuals that could be assessed according to the studied variable and to the Vineland change score.Outcome was measured by raters who were blind to the type of intervention given. Statistical analyses
Data analysis were conducted with the IBM Statistical Package for the Social Sciences(SPSS)Version19.In order to examine the influence of potential dichotomous risk der)on various dichotomous a medical/genetic condition),odds ratios(OR)and their95% confidence intervals(95%CI)were calculated by logistic regression analysis.Mean age at referral for intervention for children with versus for children without medical/genetic conditions was compared with a t test.
Vineland Adaptive Behaviour Scales change scores were related to(1)any medical/genetic condition including epilepsy,(2)epilepsy‘‘only’’,(3)regression and(4)age at referral,and for each of these four factors also to the level of intensity of intervention.For each factor,a2(T1/ T2)92(level of intensity of intervention)92(value on current ession Yes/No)mixed Analysis of Variances(ANOVA)wa
s computed,using eta-squared(g2) as a measure of effect size.
An alpha level of0.05was used for all statistical analyses.
Ethics
The Regional Ethics Committee in Stockholm approved the study.
Results
Medical/genetic conditions,including epilepsy,
in preschool children with ASD
At T1,a medical/genetic condition with probable association with the child’s ASD was identified in22children(11%),
and at T2another16children had been found to have an associated medical/genetic condition(38/208)(18%).
At T1,7of the208children had a medical disorder/ genetic syndrome identified.Of these,three boys had
tuberous sclerosis(all with epilepsy),one boy had Fragile X syndrome(141/142FMR1-examined children tested negative for Fragile X),one boy had pyridoxine dependency(with a verified mutation in the ALDH7A1-gene),one boy had duplication17p13.3,and one girl had deletion22q11syndrome.Another nine children had medical conditions with less clear(but probable)asso-ciation with the child’s ASD.Of these nine,three had abnormal magnetic resonance imagingfindings indicating cortical dysplasia/migrational disorder,two had been born preterm(gestational week24and28),one adopted child met criteria for foetal alcohol spectrum disorder, one child had a history of prenatal intrauterine exposure to sodium valproate(mother treated for epilepsy),one boy had congenital hydrocephalus with ventriculoperito-neal shunt,and one boy was clinically diagnosed with Pierre Robin syndrome.Epilepsy without aetiology was found in another six children(all male)at T1.
At T2,medical/genetic conditions could be analysed in the total group of208children.Five children developed epilepsy between T1and T2.Two of the three girls clini-cally suspected of having Rett syndrome were‘‘confirmed’’by genetic testing for MECP2mutation.In another nine children,analysed with array-CGH,a significant genomic imbalance was found.Details on thesefindings and asso-ciated developmental profiles in children with abnormal CGH-array results will be presented in a separate article. The preliminary diagnostic yield was7.4%(12/162).
ID was found in63%(24/38)of those with medical/ genetic/epilepsy disorders as compared with45%(76/170) in those without such disorders(OR=2.12,95% CI=1.03–4.38,p=0.042).
Medical/genetic conditions including epilepsy were more often,but not significantly,found in children with the combination of ASD and ID(24/102)(24%)as compared with those who had ASD without ID(14/106)(13%) (OR=2.095%CI=0.98–4.17,p=0.057).
Of the32girls,10(31%)had an identified medical/ genetic condition.The corresponding rate in boys was 28/176=16%(OR=2.40,95%CI=1.03–5.62, p=0.043).
Any medical/genetic condition including epilepsy
and age at referral
Children with medical/genetic conditions,including epi-lepsy,had been referred for intervention at significantly earlier ages compared to those without such conditions (M=38.3,SD=10.1vs.M=41.4,SD=8.3; t206=1.97,p=.05).
Epilepsy
At T1,13/208(6%)children had been diagnosed with epilepsy and at T2(ages between 4.5and 6.5years) 18/208(9%)had epilepsy.Excluding those children whose epilepsy was part of a known medical/genetic syndrome,the epilepsy prevalence was5%(‘‘idiopathic epilepsy’’).
Epilepsy onset varied from thefirst days of life up to the age of6years.Four children(4boys)had infantile spasms(West syndrome),of whom two had tuberous sclerosis,one had a CNS-malformation,and one was ‘‘idiopathic’’.A definite aetiology was identified in7/18 cases with epilepsy;three children had tuberous sclerosis, two had CNS malformations,one boy had B6depen-dency,and one girl had Rett syndrome.Of the11chil-dren without an identified medical disorder/genetic syndrome and with normal magnetic resonance imaging, 6could be classified as having epilepsy with generalized seizures,2had focal seizures and3had unclassifiable seizures.Epilepsy was diagnosed in4/32girls(13%)and in14/176boys(8%),(OR=1.65,95%CI= 0.51–5.39,n.s.).In children with the combination of ASD and ID(IQ\70)13/102(13%)had epilepsy whereas children with ASD and borderline,average or above average IQ(IQ[70),5/106(5%)had epilepsy (OR=2.95,95%CI=1.01–8.60;p=0.047).
Clinical characteristics of children with epilepsy are presented in Table1.
Regression and clinical characteristics
A history of regression was described both in the child’s CHC records and at parental interview in46/208children (22%).The most consistentfinding was loss of expres-sive language skills.A few children were also reported to have deterioration in social interaction skills and some behavioural changes such as hyperactivity or extreme passivity.Regression was reported in5/32girls(16%) and41/176boys(23%),(OR=0.61,95% CI=0.22–1.68,n.s.).
A history of regression was more common in children with a clinical diagnosis of autistic disorder(33/105,31%)as compared with children with atypical autism/PDD-NOS(7/ 58,12%,OR=3.34,95%CI=1.37–8.14,p=0.008).
In children with the combination of ASD and ID, regression was reported in34%(35/102)versus10%(11/ 106)in children with ASD without ID(OR=4.5195% CI=2.14–9.51,p\0.001).
Regression and relation to epilepsy(including febrile seizures)
Epilepsy was no more common in children with regression (3/46)than in those without(15/162)(OR=0.68,95% CI=.19–2.47,n.s.).
Febrile seizures had occurred in16/208children(8%). Four of these children had epilepsy at follow-up.F
ebrile seizures were not significantly more common in children with ASD without a history of regression(15/162)com-pared to children with regression(1/46)(OR=4.59,95% CI=0.59–35.73,p=0.145).Outcome in relation to medical/genetic conditions, excluding epilepsy,and to intensity of intervention
As shown in Fig.2,children without any medical/genetic condition had a positive VABS-II change score at T2 (compared to T1)whereas children with a medical/genetic condition(excluding epilepsy)had a negative change score.A29292mixed ANOVA with Medical/genetic condition(yes,no)and Intervention intensity(intensive or non-intensive ABA)as between-subjects factors,Time(T1, T2)as a within-subjects factor and Vineland composite score as the dependent variable showed however that the
Table1Some medical characteristics in children with ASD and epilepsy
Sex Onset Initial seizure
semiology EEG Imaging Aetiology AED at Time2Seizure
free,
at T2
Male1day Tonic,apnoea,
eye deviation
Multifocal bilateral N(MRI)B-6dependency B-6?folic acid? Male4days Focal clonic Focal right hemisphere N(MRI)Unknown None? Female3weeks Head/eye deviation Focal right central?(MRI)CNS
malformation
RUF,ZNS
Male2months Focal?infantile
spasms?F.S.Focal left
frontal[hypsarrythmia
?(MRI)Tuberous
sclerosis
VPA
Male6months Infantile spasms Multifocal/hypsarrythmia?(MRI)Tuberous
sclerosis
VPA?
Male7months Infantile
spasms?F.S.
Multifocal N(CT)Unknown VPA,CLB,PB,LCM
Male11months GTCS Focal left hemisphere?(MRI)Tuberous
sclerosis
OXC,LEV?
Male14months Breath holding
spells/GTCS
Normal N(MRI)Unknown LEV?
Male18months Absence?tonic Generalized
sharp-slow wave
N(MRI)Unknown VPA,LEV,CLB
Male18months Myoclonic
?GTCS?F.S Focal bilateral frontal spike/
polyspike
N(MRI)Unknown LEV,LTG,CLB
Male19months Infantile spasms Multifocal/hypsarrythmia?(MRI)CNS
malformation RUF,ESM,ketogenic diet
Male3years Absence Focal right centrotemporal N(CT)Unknown None? Male4years GTCS Generalized3Hz spike
and wave
N(CT)Unknown LTG?
Male4years Absence Generalized sharp wave Not
done
Unknown VPA?
Female5years Complex focal Bilateral central poly
spike during sleep
N(MRI)Rett syndrome VPA,LTG?
Female5years Absence Generalized2.5–3Hz spike
and wave
N(MRI)Unknown VPA
Female6years Complex
focal?F.S.Generalized spike
and slow wave
N(MRI)Deletion(2)
p16.3
VPA,LEV,NZP
Male6years None Bilateral frontal
during sleep,CSWS
N(CT)Unknown LTG?
Initial seizure semiology refers to presenting seizure type/epilepsy syndrome
EEG refers to result of most pronounced pathology at the time of epilepsy diagnosis
F.S.History of febrile seizures,AED Antiepileptic drugs at intervention time2,RUF Rufinamide,ZNS zonisamide,VPA valproate,CLB clobazam,PB Phenobarbital,LCM lacosamide,OXC oxcarbazepine,LEV levetiracetam,ESM etosuximide,LTG lamotrigine,NTZ nitrazepam
版权声明：本站内容均来自互联网，仅供演示用，请勿用于商业和其他非法用途。如果侵犯了您的权益请与我们联系QQ:729038198，我们将在24小时内删除。
发表评论