Gadgets

A view to regressive autism through home movies. Is early development really normal?

Description
A view to regressive autism through home movies. Is early development really normal?
Categories
Published
of 5
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  Brief communication A view to regressive autism through homemovies. Is early development really normal? Maestro S, Muratori F, Cesari A, Pecini C, Apicella F, Stern D. A viewto regressive autism through home movies. Is early development reallynormal?Acta Psychiatr Scand 2006: 113: 68–72.    2005 Blackwell Munksgaard. Objective:  To describe early development of children with regressiveautism. Method:  Home movies from the first 18 months of life of three groupsof children with early onset autism (EOA), regressive autism (RA) andtypical development (TD) were rated through the Grid for Attention inInfants in three age ranges. Different  anova  and  post hoc t -tests wereconducted on frequencies of behaviours. Results:  Differently from TD, for both RA and EOA non-socialattention is higher than social attention across ages. While EOA ischaracterized by an early deficit in social attention, in RA socialattention increases until the first birthday when its decrease proceeds atthe same rate as the increase of non-social attention. Conclusion:  We hypothesize that the intense interest towards objectscan be the first sign of an atypical development also in RA;this distinguishing feature lead us to believe that regression, basedon the assumption of a previous normal development, is onlyapparent. S. Maestro, F. Muratori, A. Cesari,C. Pecini, F. Apicella, D. Stern Department of Developmental Neurosciences, ScientificInstitute Stella Maris, University of Pisa, Calambrone(Pisa), ItalyKey words: autism; retrospective study; regression;child developmentProf. F. Muratori, c/o IRCCS Stella Maris, Via deiGiacinti, 2-56018 Calambrone (Pisa), Italy.E-mail: f.muratori@inpe.unipi.itAccepted for publication October 10, 2005 Significant outcomes •  The first year of life may be considered a sensitive window to detect the appearance of autism inchildren. •  The prevalent interest to non-social stimuli can be considered a red flag during the first year of children with autism. •  Previous normal development in regressive autism may be only apparent. Limitations •  The small number of the sample do not allow strong recommendations. •  No available data to state if the early onset and regressive autism coincide with two different clinicaloutcome. •  The absence of a correlation analysis prevents us from exploring the reciprocal interference betweenabsence of competencies and signs of autism. Introduction Autism occurs early in life and is characterized bysocial, verbal and non-verbal impairments, andrigid repetitive behaviours (1). Autism is thoughtto be present right from birth, even if a minority of children displays a normal course during infancyfollowed by a regression during the second year of life. The phenomenon of reported regression is apotential problem for early detection of autism. Acta Psychiatr Scand 2006: 113: 68–72All rights reserved  DOI: 10.1111/j.1600-0447.2005.00695.x Copyright    2005 Blackwell Munksgaard  ACTA PSYCHIATRICASCANDINAVICA 68  Various studies have documented parental reportsof regression in 20–40% of cases (2, 3). For theseparents, the child seems to develop normally for atime and then loses skills. However, examination of these reports suggests that the question of loss of skills is a complex one with some parents reportinga failure to gain, rather than an actual loss. In fact,several accurate studies on the development of typical behaviours in autism shows subtle andqualitative abnormalities so that the reportedbehavioural regression, based on the assumptionof a previous normal development, may be parti-ally apparent (1). However, the issue of a morerigorous identification – of cases of regression – isof potential interest for various reasons includinggenetic anomalies associated with clock genes andthe recent concern that exposure to immunizationsmight cause autism to develop (4). Retrospectiveanalysis of home movies (HM) of children who arelater diagnosed with autism can be considered avalid and ecological tool to address this issue; infact HM not only allow us to obtain data andinformation that are not influenced by time orparent’s recall, but also to study the actual courseof more rigorously defined cases of regression (5). Aims of the study The main aim of this study was to describe theearly developmental pathways of social and non-social competencies in regressive cases. Theauthors hypothesize a different initial pathologicalprocess in these cases compared with cases wheresigns of autism are present from the first month of life. Material and methods We studied HM from the first 18 months of life of three groups of children matched for gender andsocioeconomic status. The children of the first twogroups (autism sample) were drawn from a pool of children arrived from multiple community sourcesto the Division of Developmental Neuroscience, asuburban public academic hospital providing careto patients of all socioeconomic levels. Inclusioncriteria were diagnosis of autism and the presenceof HM including sequences for each child from allthe three analysed periods of time. All childrenreceived a diagnosis of autism through the AutismDiagnostic Interview-Revised when they werebetween 4 and 12 years old (mean: 6 years; SD:2.3 years). On this interview, all cases with broaderautism spectrum or Pervasive DevelopmentalDisorders Not Otherwise Specified (PDD-NOS)were automatically excluded from the autismdiagnosis. The absence of an identified genetic ormetabolic disorder and of a severe sensory ormotor impairment were verified for all children.The autism sample was composed of an earlyonset group (EOA) and a regressive group (RA)on the basis of the presence or absence of autismsigns in the first year of life verified directly on theHM through the rating of the Behavioural Sum-marized Evaluation Scale [see Ref. (5) for a fulldescription of this instrument and of its applicationto identify regressive cases]. Both the EOA (13males and 2 females) and the RA (12 males and 3females) were composed of 15 children. HM of thecontrol group belonged to 15 children (12 malesand 3 females) with typical development (TD). Thewhole sample (autism and control group) wascomposed of Italian children belonging to middle/upper-middle class families according to theHollingshead criteria.To analyse the HM, we separated recordings of the first 18 months of life in three periods: 0–6(T1), 6–12 (T2) and 12–18 (T3). The films of thethree groups were mixed and rated by blindobservers, with a paper and pencil method,through the Grid for the Assessment of Attentionin Infants. For a full description of this Grid andfor further details about the HM selection proce-dures, see Ref. (6). No overlap between the item of this Grid and the item of Behavioural SummarizedEvaluation (BSE) was present. The eight items of the Grid were grouped into two developmentalareas: the area of social attention (looking atpeople, orienting towards social stimuli, smiling atpeople, vocalizing in response to social stimuli) andthe area of non-social attention (looking at objects,orienting towards non-social stimuli, smiling atobjects, vocalizing in response to non-social stim-uli). We used an interval-scoring method andcomputed the frequencies for each analysed beha-viour; because of the differences in length of eachHM, frequencies were converted to a ratio of thenumber of behaviours to time. Proportion scoreshave been similarly used in other video studies (7).After a complete description of the study to thefamilies, written informed consent was obtained. Statistics A three-factor  anova  [time (3)  ·  area (2)  ·  group(3)] was used to describe differences between socialand non-social area according to the group and thetime. A two-factor  anova  [time (3)  ·  area (2)] wasused separately for each of the three groups toverify intragroup differences in time and area. Atwo-factor  anova  [time (3)  ·  groups (3)] was usedto compare, separately for each time, the three Regressive autism 69  groups on the basis of the different scores obtainedfor social and non-social area.  Post hoc t -tests(Bonferroni) were conducted within and betweengroups. Statistical analyses were conducted with spss  12.0  (SPSS, Chicago, IL, USA). Results The three-factor  anova  showed significant effects of time [ F  (2,54)  ¼  16.02,  P  < 0.001] and group[ F  (2,27)  ¼  7.1,  P  < 0.005] regardless of the area.The following significant interactions were found:i) time  ·  group [ F  (4,54)  ¼  4.2,  P  < 0.005]; ii)area  ·  group [ F  (2,27)  ¼  13.8,  P  < 0.001];iii) time  ·  area [ F  (1,54)  ¼  4.6,  P  < 0.05]; iv)time  ·  group  ·  area [ F  (4,27)  ¼  5.5,  P  < 0.005].Results and significant effects from the two-factor  anova  [time (3)  ·  area (2)] conducted sepa-rately for each group are reported in Figs 1–3.  Posthoc  comparisons [with Bonferroni correction forthree comparisons for each time ( a  ¼  0.016)]within groups showed that: i) EOA had highernon-social competencies at each time with astatistical significance only at T3; ii) RA showedhigher social than non-social competencies at T1and a progressive increase of non-social compe-tencies reaching statistical significance at T3; iii)TD showed significantly higher social competen-cies at each time.The time (3)  ·  group (3)  anova  conducted onsocial area showed significant effects of time[ F  (2,54)  ¼  7.6,  P  < 0.005], of group [ F  (2,27)  ¼ 11.4,  P  < 0.001], and of time  ·  group [ F  (4,54)  ¼ 10.1,  P  < 0.001]. These interactions were furtherexplored with  post hoc t -tests with Bonferronicorrection.  Post hoc  comparisons between groups(Table 1) showed: i) EOA was significantly differ-ent from RA at T1 and at T2, but not at T3; ii)EOA was significantly different from TD at T1 andat T3; iii) RA was significantly different from TDat T2 and at T3.The time (3)  ·  group (3)  anova  conducted onthe non-social area showed significant effectsof time [ F  (2,54)  ¼  14.8,  P  < 0.001] and of group[ F  (1,27)  ¼  8.7,  P  < 0.001], but not of time  · group.  Post hoc  comparisons between groups(Table 1) showed: i) EOA and RA were neversignificantly different; ii) EOA and TD weredifferent at T3; iii) RA and TD were different atT2 and at T3. 00.20.40.60.811.21.41.6T1 T2 T3 Early-onset children        F     r     e     q     u     e     n     c       i     e     s Social Non-social Fig. 1.  Developmental pathways of social and non-socialattention in early onset autism. Competencies increase withtime regardless of area [ F  (2,18)  ¼  23.7,  P  < 0.001]. Non-social attention is always higher with a significant effect of area[ F  (1,9)  ¼  18.6,  P  < 0.005] regardless of age range, but with asignificant effect of time  ·  area [ F  (2,18)  ¼  7.8,  P  < 0.005]. At  post hoc  comparisons, the difference between areas approachessignificance at T2 ( t  ¼  ) 2.12,  P  ¼  0.06) and is statisticallysignificant at T3 ( t  ¼  ) 4.9,  P  < 0.001). 00.20.40.60.811.21.41.6T1 T2 T3 Typical children        F     r     e     q     u     e     n     c       i     e     s Social Non-social Fig. 3.  Developmental pathways of social and non-socialattention in typical development. Competencies increase withtime regardless of area [ F  (2,18)  ¼  11,34,  P  < 0.001]. Socialattention is always higher with a significant effect of area[ F  (1,9)  ¼  50.8, P  < 0.001] regardless of age range, and with asignificant effect of time  ·  area [ F  (2,18)  ¼  6.5, P  < 0.005]. At  post hoc  comparisons differences between areas are statisticallysignificant both at T1 ( t  ¼  3.13,  P  < 0.05), at T2 ( t  ¼  3.4, P  < 0.01) and at T3 ( t  ¼  5.5,  P  < 0.001). 00.20.40.60.811.21.41.6T1 T2 T3 Regressive children        F     r     e     q     u     e     n     c       i     e     s Social Non-social Fig. 2.  Developmental pathways of social and non-socialattention in RA. Competencies changes with time regardlessof area [ F  (2,18)  ¼  3.6,  P  < 0.05]. A significant effect of time  ·  area is present [ F  (2,18)  ¼  4.2, P  < 0.05] because of theincrease of both performances from T1 to T2 and of thedecrease of social attention from T2 to T3. At  post hoc  com-parisons, the difference between areas is statistically significantat T3 ( t  ¼  ) 7.6,  P  < 0.001). Maestro et al. 70  Discussion The availability of HM of infants, made beforeparents suspected anything unusual about theirchildren’s development, can allow to overcome thelack of studies involving direct observation of infants with autism because of the fact that autismis not yet diagnosed in infancy. HM are of particular importance for the description of RAfor two reasons: first, they allow to identify morerigorously cases of regression; secondly, they allowthe inspection of the developmental trajectoriesduring the period before the loss of skill. In theend, they allow to verify if the initial pathologicalprocess in autism (8) is characterized by differentpathways on the basis of the different age of symptoms appearance. According to previous lit-erature (2–4) we expected in RA, differently fromcases with early symptoms, normal developmentbefore the appearance of typical autistic features at12–18 months of life.Indeed, the developmental pathway of social andnon-social attention, in the first year of life, showeddifferent patterns in RA compared with EOA. Thislatter group shows an early specific deficit, only inthe attention to social stimuli with a progressivedevelopmental gap between social and non-socialattention so that these children appear rapidlymore attracted by objects compared with typicalchildren. However, RA do not display this earlyreduction of social attention and the developmentof social and non-social attention grow up until thefirst birthday when the decrease of social attentionproceed at the same rate of the increase of non-social attention, postponing after the first birthdaythe developmental gap observed in EOA from thebeginning. Differently from TD, for both RA andEOA the non-social attention is always higher thansocial attention; moreover, during the third semes-ter of life, both groups have a converging featurecharacterized by a significant impairment of socialattention and an excessive interest in non-socialstimuli.These different pathways provide other explana-tions to our previous study on attentional skillsduring the first year of life (6). There we found adecrease of differences between autism and typicalinfants in the second semester of life; now wehypothesize that the finding was due to the relativeincrease of social attention in RA. On this basis,RA might be considered a different population of infants as far as the initial pathological process isconsidered. We suggest that the increase in bothsocial and non-social attention showed by RAduring the second semester of life could be thebehavioural marker of these cases and we couldsuggest that it is the expression of altered regula-tory systems. The reduced pruning and apoptosishypothesized by Courchesne et al. (9) on brainovergrowth in the first year of life in autism couldbe the neurobiological mechanism of this regula-tory disorder which results in the decrease of socialattention from 12 to 18 months, as a consequenceof an overcrowded system.On the contrary, EOA is characterized by apersistent early deficit in social attention acrossages.Theearlysymptomaticpattern(5)maysuggesta different organization because of the early inter-relationbetweenpresenceofsymptomsand absenceof expected behaviours. Differently, RA is charac-terized by high rates in social and non-socialattention during the first year of life which couldreduce the first emerging autism signs also after thefirst birthday and may mask the first deviations.However, both autism groups have the distin-guishing feature of a higher non-social attentionacross the first 18 months of life with no differencebetween the two groups. This can result in anabnormal social information from the beginning, inan altered development of experience-expectantneurodevelopmental processes, and in the disrup-tionofbrainconnections.Thisdistinguishedfeaturemay also explain the deficit of joint attentiondescribedbymanyauthorsasofseminalimportancefor autism (8). Our new data suggest that jointattention deficit is two-fold: in fact, beside the lower Table 1.  Post hoc   comparison between groups (df  ¼  18) for social and non-social attentionEOA RAT1 T2 T3 T1 T2 T3Social attentionRA  t   ¼  ) 4.8 ( P   < 0.001)  t   ¼  ) 3.1 ( P   < 0.01) ns – – –TD  t   ¼  ) 3.8 ( P   < 0.01) ns  t   ¼  ) 4.8 ( P   < 0.001) ns  t   ¼  2.4 ( P   < 0.05)  t   ¼  ) 4.5 ( P   < 0.001)Non-social attentionRA ns ns ns – – –TD ns ns  t   ¼  3.12 ( P   < 0.01) ns  t   ¼  2.7 ( P   < 0.05)  t   ¼  3.5 ( P  <0.005)EOA, early onset autism; RA, regressive autism; TD, typical development. Regressive autism 71  social attention of EOA compared with TD, in thewhole autism group there is a clear and permanentprevalence of non-social attention, as opposed toTD whose social attention is always higher com-pared with non-social attention. This pattern char-acterized by the predominance of interest to objectsseems peculiar of RA also when social attentiongrows up during the second semester of life. Fromthis point of view, both EOA and RA are charac-terizedbyanalteredcompetitionbetweensocialandnon-social attention which probably means analtered organization of the brain area for face andobjectprocessing,asdescribedinolderpatientswithautism through functional Magnetic ResonanceImaging (fRMI) (10).Two final questions remain unanswered, namelywhat is the actual autistic process and at what ageis it detectable? This study confirms the hypothesisthat autism is characterized by a specific pattern of positive and negative signs as from 18 months of life. However, our study shows that this corepattern can be reached through two differentpathways: one consisting in an early lack of social attention associated with higher non-socialattention and early autism signs, and the otherconsisting in an early and increasing prevalence of interest for non-social stimuli without a clearreduction of social attention and without thepresence of positive signs of autism during thefirst year of life. Therefore, excessive non-socialattention might be included as one of the item toinvestigate in screening tools for autism.The impairment of social behaviours in the firstsemester of life may be considered a risk to developan EOA and this remains true during the first18 months of life. Besides, the period between 12and 18 months represents a second sensitivewindow to detect the emerging decrease of socialattention and the appearance of first autism signsin RA. However, also for these children, signs of autism may be preceded during the first year of lifeby the strong interest in non-social stimuli which issimilar to the one in EOA. This abnormal preval-ence of interest for objects in RA lead us to believethat developmental regression, based on theassumption of a previous normal development, isonly apparent. We hypothesize that the partic-ularly intense interest to non-social stimuli can bethe first sign of an atypical development, and couldbe considered as a red flag in the first year of life.For all these reasons, we propose to reserve theterm of regression only for the cases of autismemerging after the second year of life, which can beclassified as disintegrative disorder.Our study might present some limitations. First,the small number of HM analysed, i.e. because of the small percentage of regressive cases. Secondly,although both EOA and RA display a differentinitial pathological process, a long-term outcometo verify if they are two different clinical popula-tion or if differences tend to decrease overtime, isnot provided. Thirdly, the absence of a correla-tional analysis prevent us from exploring thereciprocal interference between the absence of expected behaviours and positive symptoms of autism. Future research on a larger sample couldexplore this issue. With respect to further studies, itwould also be interesting to know if there is adifference between videotaped children and chil-dren who were never videotaped.Despite these limitations, we signal that this is,to our knowledge, the first study describingregression through direct observation of infantsin the HM recorded by parents before diagnosis; infact, up to now developmental regression wasexplored only through maternal perception whichraised questions about parental sensitivity toearlier, subtle deviations (2). Acknowledgement This study was supported by grant RC 6/04 from NationalInstitute of Health. References 1.  Sigman M ,  Dynamo A ,  Gratier M ,  Rozga A.  Early detectionof core deficits in autism. Ment Retard Dev Disabil ResRev 2004; 10 :221–233.2.  Rogers SJ.  Developmental regression in autism spectrumdisorders. Ment Retard Dev Disabil Res Rev 2004; 10 :139– 143.3.  Kobayashi R ,  Murata T.  Setback phenomenon in autismand long-term prognosis. Acta Psychiatr Scand1998; 97 :99–108.4.  Siperstein R ,  Volkmar F.  Brief report: parental reporting of regression in children with PDD. J Autism Dev Disord2004; 34 :731–734.5.  Maestro S ,  Muratori F ,  Cesari A , et al. Course of autismsignsinthefirstyearoflife.Psychopathology2005; 38 :26–31.6.  Maestro S ,  Muratori F ,  Cavallaro MC , et al. How youngchildren treat objects and people: an empirical study of thefirst year of life in autism. Child Psychiatry Hum Dev2005; 35 :383–396.7.  Baranek GT.  Autism during infancy: a retrospective videoanalysis of sensory-motor and social behaviors at 9– 12 months of age. J Autism Dev Disord 1999; 29 :213–224.8.  Mundy P ,  Crowson M.  Joint attention and early socialcommunication: implication for research on interventionwith autism. J Autism Dev Disord 1997; 27 :653–676.9.  Courchesne E ,  Carper R ,  Akshoomoff N.  Evidence of brainovergrowth in the first year of life in autism. JAMA2003; 290 :337–344.10.  Schultz RT ,  Gauthier I ,  Klin A , et al. Abnormal ventraltemporal cortical activity during face discriminationamong individuals with autism and Asperger syndrome.Arch Gen Psychiatry 2000; 57 :331–340. Maestro et al. 72
Search
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks