Toddlers with Autism Developmental Perspectives

LWW/IYC AS188-01 May 30, 2003 11:17 Char Count= 0

Infants and Young Children Vol. 16, No. 3, pp. 201–214 c 2003 Lippincott Williams & Wilkins, Inc. Toddlers With Autism Developmental Perspectives

Linda R. Watson, EdD; Grace T. Baranek, PhD; Pamela C. DiLavore, PhD

Recent research has greatly expanded our knowledge about the early development of children with autism and related disorders. Familiarity with this literature will improve the ability of professionals to appropriately diagnose and intervene with young children with autism. This article reviews the literature pertaining to the development of children with autism under the age of 3 years. We ex- amine findings on affective development, sensory processing and attention, praxis and imitation, communication, play, and motor features and stereotyped behaviors, and discuss the interrelationships among these different aspects of development. Screening and diagnostic tools with specific applicability to young children with autism are reviewed as well. Key words: affect, assessment, autism, communication, play, sensory processing, toddlers

ECENT research has greatly expanded EVIDENCE REGARDING EARLY Rour knowledge of the early development DEVELOPMENT of children with autism and related disorders. An increased understanding of how autism The first attempts to study the early de- is manifested in very young children holds velopment of children with autism relied on promise for improving our ability to diagnose parents’ retrospective reports of their chil- children at a younger age and to offer appro- dren’s early development (eg, Hoshino et al., priate interventions that address the needs 1982; Ornitz, Guthrie, & Farley, 1977). This of very young children with autism and their approach provided intriguing insights into the families. The purpose of this article is to sum- early development of children with autism marize the current evidence base related to prior to their diagnoses. Concerns about the development of children with autism un- this research methodology, however, include der the age of 3 years, and to discuss assess- questions about the accuracy of parental re- ment instruments that have particular applica- call of early development after a period of bility to screening for and diagnosing autism several years has elapsed, and the possible in young children. influences of the subsequent knowledge of a child’s diagnosis on that recall. Over the past decade, newer research using different methodologies has helped to confirm much From the Divisions of Speech and Hearing Sciences of the information from parent recollections, (Dr Watson) and Occupational Sciences (Dr and to significantly elaborate upon that in- Baranek), Department of Allied Health Sciences, and the TEACCH Division (Dr DiLavore), Department of formation. This research includes prospective Psychiatry, Medical School, University of North studies of children referred to a diagnostic Carolina at Chapel Hill. clinic at age 2 because of concerns about pos- We thank Vesna Costello for her helpful assistance in sible autism (eg, Gillberg et al., 1990; Lord, preparation of the manuscript. 1995), studies related to the development of Corresponding author: Linda R. Watson, EdD, Depart- screening instruments for autism in young ment of Allied Health Sciences, CB # 7190, Medical children (eg, Baron-Cohen et al., 1996; Stone, School, Wing D, University of North Carolina at Chapel Hill, NC 27599-7120 (e-mail: Linda [email protected]. Lee, et al., 1999), developmental studies edu). of toddlers with diagnosed or suspected 201 LWW/IYC AS188-01 May 30, 2003 11:17 Char Count= 0

202 INFANTS AND YOUNG CHILDREN/JULY–SEPTEMBER 2003

autism (eg, Charman et al., 1997; Stone, shorter durations of looking at people, rela- Ousley, Hepburn, Hogan, & Brown, 1999; tive to children with typical development and Swettenham et al., 1998), and retrospective those with developmental delays. In research studies of home videotapes made of infants with the same sample of children, Charman and young children who were later diagnosed et al. (1997) demonstrated that children with with autism (eg, Adrien et al., 1993; Baranek, autism were unlikely to look at and show fa- 1999; Osterling & Dawson, 1994). cial concern for an experimenter expressing We organize our discussion of the de- distress, but children in the other 2 groups of- velopment of young children with autism ten did so. by the developmental domains of affective development, sensory processing and atten- Sensory processing, attention, tion, praxis and imitation, communication, and self-regulation play, and motor features and stereotyped be- There is some controversy regarding the ex- haviors. We recognize, however, that develop- tent to which unusual sensory symptoms are ment is integrated in complex ways across do- manifest in young children with autism, and mains, and that the concepts of researchers more controversy with respect to the nature investigating one domain of development of these difficulties. Phenomenological narra- overlap with those of researchers investigat- tives from persons with autism recalling their ing other domains. Our discussion addresses childhood experiences (eg, Grandin, 1997), some of these interrelationships. as well as retrospective reports from parents of children with autism and/or reports from Affective development their clinical charts (eg, Dahlgren & Gillberg, Limitations in facial expression are listed as 1989; Greenspan & Weider, 1997; Hoshino a possible symptom of autism under the cat- et al., 1982; Ornitz et al., 1977) attest to nu- egory of “nonverbal social behaviors” in the merous symptoms reflective of sensory pro- Diagnostic and Statistical Manual of Men- cessing and/or attentional disturbances dur- tal Disorders, 4th Edition (DSM-IV, Amer- ing the first 3 years of life. Among these ican Psychiatric Association, 1994). In typ- symptoms are hypersensitivities to sound, ical development a broad range of facial aversion to social touch, avoidance of cer- expressions is in place by 2–3 months of age, tain food textures, lack of response to pain, and smiling in response to the smiles of oth- poor orientation to visual stimuli, and over- ers is observed reliably by 6 months. In gen- focused attention or preoccupations with eral, fewer facial expressions are seen in tod- various sensory features of objects (eg, dlers with autism, and fewer are directed to watching things spin, licking objects). These other people. In addition, these children are unusual behaviors may range from mild to se- more likely to show ambiguous facial expres- vere, perhaps moderated in part by the child’s sions. For example, retrospective parent re- coping/self-regulatory abilities and environ- ports have been more likely to describe in- mental supports, and may interfere with par- fants and toddlers with autism as “having an ticipation in a variety of daily activities (eg, expressionless face” (Hoshino et al., 1982) picky eater, temper tantrums in noisy environ- than to apply this description to children with ments, sleeping problems). other developmental disabilities or children It is unclear whether or not sensory fea- who are typically developing. Retrospective tures described in young children with autism video research has indicated a lack of social are completely unique to this diagnosis. S. J. smiling in infants with autism (Adrien et al., Rogers, S. Hepburn, and E. Wehner (unpub- 1993). lished data), using a parent report instrument, Swettenham et al. (1998) demonstrated reported higher levels of sensory symptoms that 20-month-old infants with autism had in 2–4-year-olds with autism as compared deficits in social orienting, as noted by to children with Down syndrome, but not LWW/IYC AS188-01 May 30, 2003 11:17 Char Count= 0

Toddlers With Autism 203

compared to children with fragile X syn- dren with respect to the order of difficulty of drome. Analyses of home movies taken of in- motor imitation, but that children with autism fants who were later diagnosed with autism overall show less proficiency. Rogers and (Adrien et al., 1992; Baranek, 1999) doc- Pennington (1991) hypothesized that infants ument specific patterns of subtle sensory- who later exhibit autism would not have the attentional symptoms that differentiate clin- ability to engage in neonatal imitation because ical groups during the first year of life, of an innate impairment of this important so- often prior to the time that parents are cial ability, but there is no research evidence cognizant of the child’s diagnosis. Interest- available related to neonatal imitation in in- ingly, infants later diagnosed with autism ap- fants who are later diagnosed with autism. In pear to be differentiated from infants with early intervention programs for children with other developmental disabilities more on the autism derived from a variety of perspectives, basis of their hyporesponsiveness to both imitation is considered a pivotal skill. The social and nonsocial stimuli than by hyperre- literature confirms that the development of sponsiveness. Some symptoms intensify dur- skills in other areas, such as play and language, ing the second year while new symptoms is closely tied to the child’s ability to imitate. also emerge (Adrien et al., 1993), suggest- Stone, Ousley, and Littleford (1997) found that ing the importance of a developmental per- for 2-year-olds with autism, the imitation of spective to understanding these unusual be- actions on objects is more closely tied to the haviors. Some researchers (Baranek, 1999; children’s concurrent play skills, and imitation Gillberg et al., 1990) hypothesize that the in- of body movements is more closely tied to the tegrity of early sensory processing and atten- children’s expressive language development tion regulation abilities is critical for the devel- both concurrently and predictively. opment of later-emerging cognitive and social Successful imitation is dependent upon skills (eg, joint attention). Thus, assessment both comprehension of the social-communi- of disruptions in these basic developmental cative interaction as well as the sensorimotor processes during the infancy period may pro- demands of the task. Several researchers have vide a window to earlier identification and suggested that young children with autism interventions. may have difficulty forming internal (so- matosensory) representations of visually Praxis and imitation modeled actions (eg, Hughes & Russell, 1993; Imitation skills are very important in early Smith & Bryson, 1998; Stone, Lee, et al., diagnosis and intervention with children with 1999). Thus, it is possible that aspects of sen- autism. In typical development, children have sory processing and integration impede the a range of imitation skills by the age of successful performance of some imitated as 12 months, including the ability to imitate well as nonimitated motor acts by these unfamiliar movements and sounds, and to children. imitate actions for interactive play. Imitation deficits are evident in the youngest children Communication with confirmed diagnoses of autism. Charman et al. (1997) have reported that 20-month- Comprehension old children with autism show less profi- Although language and communication im- ciency in procedural imitation of actions on pairments are one of the criterial areas for objects than do children who are typically the diagnosis of autism, comprehension is not developing or those with other developmen- specifically mentioned in the diagnostic crite- tal delays. Research by Stone, Ousley, and ria. Nevertheless, relative to their nonverbal Littleford (1997) with 2-year-olds with autism cognitive abilities, the comprehension abili- suggests that children with autism have the ties of children with autism are significantly same developmental patterns as other chil- worse than those of their counterparts with LWW/IYC AS188-01 May 30, 2003 11:17 Char Count= 0

204 INFANTS AND YOUNG CHILDREN/JULY–SEPTEMBER 2003

specific language impairment or low intelli- ents of children with other developmental de- gence (Fein et al., 1996). Comprehension is a lays (Lord, 1995). key concern for toddlers with autism. Com- prehension problems not only impede the Communicative means child’s progress in social interaction and ex- Diagnostic criteria in DSM-IV describe the pressive communication, but also often are impairment of nonverbal as well as verbal implicated in the development of challenging communication skills in autism. Thus, these behaviors. disorders not only impact communication Chapman (1978), in seminal work, pro- through verbal language. Eye contact is an posed that nonverbal “comprehension strate- important consideration for diagnosis not so gies” are used by typical 9–12-month old in- much in terms of absolute quantity of eye con- fants to decode and respond appropriately tact (although that may also be reduced in to language input, and also constitute the young children with autism; eg, Swettenham foundation on which more sophisticated lan- et al., 1998), but more importantly in terms guage comprehension skills are built. Impor- of its use in regulating communication (eg, tant strategies include looking at what adults Stone, Ousley, Yoder, et al., 1997). For exam- look at, acting on objects noticed, and imi- ple, parents of 2-year-olds with autism are less tating ongoing activities. Subsequent research likely than parents of 2-year-olds with other has at least partially supported Chapman’s developmental disabilities to report the co- model for early comprehension development. ordination of gaze with other behaviors in For example, individual differences in gaze- requesting (Lord, 1995). Related to other pre- following proficiency at 6 and 16 months of cursors of verbal communication, parents ret- age have been found to correlate with recep- rospectively report less babbling and gestur- tive vocabulary at 12 and 20 months of age, ing during early development in children with respectively (Morales, Mundy, & Rojas, 1998; autism compared to typically developing chil- Mundy & Gomes, 1998). dren (Ornitz et al., 1977); however, retro- As discussed elsewhere in this article, tod- spective reports of these behaviors do not dlers with autism are impaired in these key as- discriminate children with autism from chil- pects of nonverbal behavior that typically sup- dren with mental retardation (Hoshino et al., port the mapping of language onto real world 1982). Another observation regarding com- objects and relations. In this light, the communicative means is pertinent to early di- prehension deficits of individuals with autism agnosis: if a child’s primary communicative are not surprising. Considering one of the means entail manipulating another person’s earliest comprehension skills, the reduced re- body (eg, placing another person’s hand on sponsiveness of children with autism to their a doorknob to request to go outside), this own names at the age of 12 months discrimi- points to a diagnosis of autism (DiLavore nates children with autism from children who et al., 1995; Stone, Ousley, Yoder, et al., 1997). are typically developing or developmentally delayed but without autism (Baranek, 1999; Communicative intention Osterling & Dawson, 1994). This continues Consideration of communication intention to be a variable that discriminates between highlights the inseparable nature of social preschoolers with autism who have no or and communicative development. Many few verbal skills and children with other de- communicative behaviors in young children velopmental disabilities (DiLavore, Lord, & are essentially social in nature. In descriptions Rutter, 1995). Problems comprehending com- of the social-communicative development of munication extend beyond difficulties in un- typical infants, the period prior to 6 months derstanding verbal language. Parents of 2-year- has sometimes been referred to as a period olds with autism are less likely to report their of “primary intersubjectivity” (Trevarthen, children comprehend gestures than are par- 1979). During this time, infants engage in LWW/IYC AS188-01 May 30, 2003 11:17 Char Count= 0

Toddlers With Autism 205

mutual imitation, vocal turntaking, and other strumental purposes (eg, requesting, reject- reciprocal social interactions with caregivers ing, and protesting; eg, Stone, Ousley, Yoder, (Meltzoff & Moore, 1998; Reddy, Hay, Murray, et al., 1997). & Trevarthan, 1997). By 8–12 months, infants progress to “secondary intersubjectivity,” Play in which there is a focus outside of the Children with autism often display uneven face-to-face interactions between infants and play skills and demonstrate unique differences their partners (Trevarthan & Hubley, 1978). as compared to children with other develop- During this period, infants begin to direct mental disorders (Stone & Lemanek, 1990; gestures and nonspeech vocalizations to Wing, 1981). Given that play skills are crit- others to communicate a range of intentions, ical to successful social participation for all including seeking attention to self or “show- young children, assessment and intervention ing off”;directing attention to objects, events, practices for young children with autism need and other people; requesting objects, acti- to specifically address this important develop- ons, and information; greeting; giving; mental domain. protesting and rejecting; responding and acknowledging; and informing (Bates, 1976). Social play A hallmark of this period is the ability of the Many young children with autism are ca- infant to coordinate attention to an object or pable of engaging in some social play; how- event and to another person within a single ever, the play is often characterized by less communicative act, otherwise known as joint proximity to peers, reduced levels of social attention. initiations, fewer responses to social over- In contrast, children with autism demon- tures, and more solitary activities than the strate a more limited range of preverbal com- play of children with other developmental dis- municative intentions. The DSM-IV criteria orders (McGee, Feldman, & Morrier, 1997; specifically list a limited range of commu- Sigman et al., 1999; Stone & Lemanek, 1990). nicative intents as a possible characteristic of Specifically, Sigman and colleagues found that children with autism. Studies indicate that regardless of the child’s level of function- children with autism are especially at risk for ing, children with autism were less socially lacking communication with more social in- engaged with peers. Play skills, in addition tent, such as showing off or directing a part- to nonverbal communication abilities, pre- ner’s attention to outside objects, people, or dicted the number of initiations the child with events (Mundy, 1995; Osterling & Dawson, autism produced, and the extent to which the 1994). A lack of protodeclarative pointing is child with autism engaged with peers. one of the key items for identifying 18-month- Some researchers hypothesize that inability old children at risk for autism using the Check- to understand complex human interactions, list for Autism in Toddlers (Baron-Cohen et al., share experiences, and take another’s per- 1996). Lord (1995) also reported that parents spective (ie, have a “theory of mind”) may of 2-year-olds with autism are less likely to contribute to difficulties in the peer engage- report referential pointing by their children ment process (Baron-Cohen, Leslie, & Frith, than are parents of children with other disabil- 1985; Frith, 1996; Frith & Happe, 1994). ities. Thus, limitations in the ability of young Successful play interactions require orienting children with autism to engage in joint at- to salient stimuli, sharing of attention, and tention have proven to be highly reliable in coordination of gaze between people and helping to distinguish them not only from objects—the skills that typical children ac- children developing typically, but also from quire easily during the first 2 years of life, but children with other types of developmental young children with autism often lack (eg, disabilities. Much of the communication ini- Baranek, 1999; Dawson, Meltzoff, Osterling, tiated by these children appears to have in- Rinaldi, & Brown, 1998; Swettenham et al., LWW/IYC AS188-01 May 30, 2003 11:17 Char Count= 0

206 INFANTS AND YOUNG CHILDREN/JULY–SEPTEMBER 2003

1998). Difficulties with social orienting and play may have limitations for diagnostic use joint attention may significantly reduce lev- prior to 12 months of age when higher levels els of engagement in social play; such specific of functional play (presymbolic pretend func- deficits also correlate with measures of functions) tend to develop in typical children. By tional and symbolic play with objects later in 18 months children with autism demonstrate childhood (eg, Stella, 2001). clear deficits in early pretend skills (ie, functional play such as feeding a doll with Object play a spoon) relative to typical children (Baron- In typical development, meaningful inter- Cohen et al., 1996; Baron-Cohen, Allen, & actions with objects begin early in life and Gillberg, 1992). By 2–3 years of age, children progressively become more diverse, complex, with autism show significantly fewer func- and refined throughout development. Several tional play acts than do children with other authors describe the importance of infant ex- developmental disorders (Stone, Coonrod, ploratory play as a means to become famil- & Ousley, 2000), indicating not only a quan- iar with objects or situations that eventually titative, but also a qualitative difference in leads to the production of complex play acts the development of play skills. Preschool (eg, Belsky & Most, 1981; Doctoroff, 1996). and school-aged children with autism often Exploratory play emerges by 4 months of persist in lower developmental levels of play age and includes behaviors such as simple (eg, Libby, Powell, Messer, & Jordan, 1998), repetitive object manipulations. Later evolv- demonstrate fewer functional acts with toys, ing types of play include relational play such use fewer toys, and engage in more repetitive as putting blocks in a cup (emerging at 10– play (Stone, Lemanek, Fishel, Fernandez, 18 months); functional play such as giving a & Altemeier 1990). Some of the lack of doll a pacifier (emerging at 12–18 months); creativity with object play has been theoreti- and symbolic play such as using a block as a cally linked to executive dysfunction, which pretend airplane, or feeding a doll using imag- makes it harder for children with autism to inary utensils or food (emerging between 18 flexibly switch from familiar routines to alter- and 30 months). native, novel play actions (eg, Ozonoff, 1997; Because children with autism are often not Pennington & Ozonoff, 1996; Rogers, diagnosed prior to 2 years of age, very little is Bennetto, McEvoy, & Pennington, 1996). known about play development prior to this point except through retrospective research. Baranek (1999) analyzed home movies of Motor features and infants at 9–12 months of age who were later stereotyped/repetitive behaviors diagnosed with autism. She found that these Motor development plays an important infants engaged in more mouthing of objects role in learning—young children depend than did typical children or children with upon motoric abilities to explore the en- other developmental delays. However, many vironment, play with toys, or engage in typical children still mouthed objects to some social-communicative interactions. Consider- extent, making it difficult to use this item as ing strengths and weaknesses within this de- a specific behavioral marker for autism. In ad- velopmental domain is therefore critical for dition, the level of object play (eg, repetitive/ the purposes of assessment and intervention inappropriate vs functional/appropriate play) for young children with autism. The major- with objects did not discriminate between ity of children with autism are often de- groups—perhaps due to the fact that few scribed as (a) achieving basic motor mile- children in any group had developed object stones (eg, head control, sitting, crawling, play skills beyond the earlier sensorimotor walking) on time (eg, Ohta, Nagai, Hara, & (eg, exploratory/relational) stages. Thus, Sasaki, 1987), (b) having relative strengths in using traditional developmental markers of motor as compared to social-communicative LWW/IYC AS188-01 May 30, 2003 11:17 Char Count= 0

Toddlers With Autism 207

skills, and (c) having motor skills that exceed iors (eg, lining up objects) are among the hall- those of children with other developmen- mark features of autism, converging evidence tal disorders with comparable mental ages (eg, Baranek, 1999; Lord, 1995; Volkmar, (Stone, Ousley, Hepburn, Hogan, & Brown, Cohen, & Paul, 1986) demonstrates the rela- 1999). Nevertheless, up to 28% of children tively late-developing nature of these phenom- with autism were documented to have lags in ena. That is, repetitive behaviors are so com- gross motor development in one prospective mon in typical development that it becomes study that screened children during the first difficult to discriminate patterns that are spe- 18 months of life (Johnson, Siddons, Frith, & cific to children with autism until after the Morton, 1992). Also, parental reports of early third birthday when these behaviors may per- motor delays in their young children with sist or perhaps even intensify. Furthermore, autism are noted to increase with the increas- stereotyped behaviors may serve numerous ing age of their child (Ohta et al., 1987; Ornitz functions—some are easily elicited even in et al., 1977). typically developing young children under Teitelbaum and colleagues (Teitelbaum, very arousing or stressful conditions (eg, frus- Teitelbaum, Nye, Fryman, & Maurer, 1998) tration). For young children with autism, dif- conducted a movement analysis of home ficulties with social skills, communication, or movies of children with autism as compared sensory processing may appear to accentuate to those with typical development and found or maintain these behaviors. that postural differences were present in the first months of life. Baranek (1999), using an analysis of home movies at 9–12 months ASSESSMENT INSTRUMENTS FOR EARLY of age, found that posturing was evident DIAGNOSIS AND SCREENING in some infants with autism, but appeared symptomatic of neuromaturational delays in Approximately 25% of children show general and was not necessarily specific to some developmental concerns, but fewer autism. Children with autism often have con- than 30% of primary care practices conduct comitant mental retardation; thus, motor de- standardized screening tests (Filipek et al., lays in this population may be associated 1999). Thus, the American Academy of more with their level of mental retardation Neurology has developed practice guidelines than with their autism. However, there is (Filipek et al., 2000) that emphasize the also accumulating evidence that some chil- need for flexible, continual developmental dren with autism without cognitive delays surveillance at every well-child visit. These may exhibit unusual postures, clumsiness, and guidelines are available on the Web site http:// motor planning problems, particularly as the www . aan . com / public / practiceguidelines / complexity of tasks increases later in child- autism.pdf. Developmental surveillance in- hood (eg, DeMyer, Barton, & Norton, 1972; cludes obtaining information from parents Ghaziuddin, Butler, Tsai, & Ghaziuddin, 1994; regarding any developmental concerns they Jones & Prior, 1985; Smith & Bryson, 1998). have about their infant, probing for age- It is also important to note that motor per- appropriate skills in each developmental formance for children with autism, like that domain, and directly observing the child. of typical children, appears to be dependent Since most standardized developmental upon the context of purposeful, goal-directed assessments do not include items that specifi- tasks, such that children produce less profi- cally target symptoms associated with autism, cient skills in carrying out decontextualized a specific screening for autism is also rec- tasks than contextualized ones (eg, Stone, ommended in cases where “red flags” are Ousley, & Littleford, 1997). present. Red flags include items such as no Although repetitive/stereotyped (eg, arm babbling, pointing, or gestures by 12 months flapping, toe walking) and ritualistic behav- of age, no single words by 16 months of age, LWW/IYC AS188-01 May 30, 2003 11:17 Char Count= 0

208 INFANTS AND YOUNG CHILDREN/JULY–SEPTEMBER 2003

no 2-word spontaneous phrases at 24 months Level II tools are those that are targeted pri- of age, or loss of language or social skills at marily for early childhood settings or diag- any age. Children who have such concerns nostic settings where children are referred should be immediately screened for autism with known developmental delays or con- and/or referred for a more specialized evalua- cerns. These tools assist the early intervention (eg, audiological assessment, diagnostic tionist, therapist, or health professional in dis- evaluation) depending on the nature of their criminating autism from other developmental difficulties. delays and suggest the need for further specialized evaluation for autism when needed. Screening for young children at risk Level III tools are specialized instruments that for autism spectrum disorders discriminate among different types of autism The purpose of screening is to identify spectrum disorders. These tools are usually those children who are at risk for autism and utilized by clinics specializing in diagnosis of will need to undergo more focused diagnostic children with autism. procedures. There has been much interest in The majority of early screening tools uti- the development of screening tools that can lize a parent questionnaire format because it reliably identify young children with autism is very easy and quick to use. However, it is (eg, Filipek et al., 2000). Good screening tools important to note that supplemental observa- are those that are age-effective, appropriately tions are critical in most cases, and that pro- inclusive, and easy to use. They also need to fessionals skilled at identifying symptoms of reflect good sensitivity (ie, identify all chil- autism may pick up on subtle symptoms that dren who are at risk without “false nega- do not evidence on questionnaire data. Like- tives” [missed children]) and specificity (ie, wise, observational assessments have limita- identify only the children at risk for the con- tions in that they assess the child in only one dition and minimizing “false positives” [mis- context, and thus, supplemental parent re- diagnosing]). Other important psychometric ports help to identify behaviors that are low concepts include the positive predictive value in frequency or to obtain information about of a tool (ie, the percentage of positive test the occurrence of behaviors across several results that are “true positives”), and its neg- naturalistic contexts. The content of screen- ative predictive value (ie, what percentage ing tools varies considerably and is very de- of negative test results are “true negatives”). pendent upon the age for which the tool is A variety of specialized tools are available to targeted, because the specific symptoms of screen for and diagnose autism spectrum dis- autism are known to change somewhat across orders in young children. These tools vary in development. Some tools (eg, Siegel, 1998) in- terms of the ages for which they are targeted, clude both “negative” symptoms (absence of the format they use, the intended level of language or social behaviors that are expected screening (population targeted), and content. for the child’s age) and “positive” symptoms Table 1 summarizes screening and diagnostic (presence of unusual behaviors for the child’s tools for young children with autism. We have age). Others (eg, Baron-Cohen et al., 1992, included available information about the psy- 1996; Berument et al., 1999; Lord et al., 1999; chometric properties of each tool; however, Stone et al., 2000) rely more heavily on the in most cases the development and research negative symptoms, particularly in areas as- on these tools are continuing. sociated with joint attention that are known Level I tools are those that are targeted for a to be effective discriminators of young chil- general population screening and lend them- dren with autism. One reason why the posi- selves to use in primary care settings such as tive symptoms are often not included in tools the pediatrician’s office. They help to iden- aimed at children below 3 years of age is tify children who need to be referred for a that research has indicated that many repeti- more specialized developmental assessment. tive and stereotyped behaviors are not salient LWW/IYC AS188-01 May 30, 2003 11:17 Char Count= 0

Toddlers With Autism 209 ) = continues ( .80. .98 = = = .87 and specificity = .87 (high); specificity is .38 (low); specificity .83 (high) and specificity = = = .99. Positive predictive value reported to be moderate. Three-stage approach allows for application to a variety of settings (primary care, developmental clinics, or specialty clinics). Empirical work on this clinically derived tool continues; not yet published (high). Positive predictive value .29. With second administration, positive predictive value improves (.75). Has been used effectivelylarge-scale in prospective studies in the United Kingdom. Recommend for use by experienced health professionals .86 (high) for the validationEmpirically sample. derived tool; can beby used trained early intervention professionals or health care workers. Replication studies with larger samples are warranted; tool is unpublished high sensitivity = Large-scale prospective study needed; absolute sensitivity and specificity are likely to be lower. Follow-up screening and supplemental observations are warranted Sensitivity Sensitivity Preliminary findings (full scale) indicate Sensitivity (typical behavior (atypical behaviors present) negative as well as sensory patterns,temperament, attention, and other developmental markers. Divided into 3 stages, each targeting a different level of screening. 71 itemspositive tap both and absent) symptoms observation items. Measures social interest, social play, joint attention, pretend (functional) play, and protodeclarative pointing 9 items from the CHAT, asadditional well items as such as social referencing and comprehension context. Items are grouped into 3 categories: play (2 items), imitation (4 items), communication/joint attention (2 items) Content includes core autistic features Nine yes/no questions and 5 child Examiner interacts with child in a play for each level) Observational Assessment (15 min) (20 min) of Administration † I (& II) 24 mo Parent Questionnaire (10 min) 23 yes/no questions that include the I, II, III 18 mo through 5 y Parent Questionnaire (5–10 min I 18 mo Parent Questionnaire and II 24 through 35 mo Observational Assessment Screening tools for autism Barton, & Green (2001) (1992, 1996) M-CHAT Robins, Fein, and references screening Targeted age format & time Content description Sensitivity & specificity PDD-ST Siegel (1998) Screening tool* Level CHAT Baron-Cohen et al. STAT Stone et al. (2000) Table 1. LWW/IYC AS188-01 May 30, 2003 11:17 Char Count= 0

210 INFANTS AND YOUNG CHILDREN/JULY–SEPTEMBER 2003 Pervasive = = .80 = .95; rventionists or health = r use by professionals in .97; specificity ental disorder, not otherwise k for autism from the general = .96 (high). Specificity = (high) for differentiating autism vs other diagnoses (mental retardation excluded), but specificity drops to .67 when discriminating autism from mental retardation. Useful as ascreener brief for children above 4 yearsage. of The ASQ is soonunder to a be new published name (Social Communication Questionnaire) .94; Module 2: Sensitivity specificity .87 for discriminating autism/ASD from other nonspectrum diagnoses. This tool does notadequate provide opportunity to observe the full range of stereotyped/restricted behaviors; the addition of these stereotyped behaviors lowered overall sensitivity and specificity in validation studies. Other observational or parent-report measures are suggested to augment assessment of stereotyped behaviors Module 1: Sensitivity Sensitivity Screening Tool for Autism in Two Year Olds; PDD-ST = Autism Diagnostic Observation Schedule. = presses for communication, play, socialization, and stereotyped behaviors/restricted interests. Autism or autism spectrum disorder are diagnosed on the basismeeting of cut-off scores in communication and social areas Diagnostic Interview-Revised. Items include 3 traditional categories used for diagnostic purposes: reciprocal social interaction, language and communication, and repetitive and stereotyped behaviors Semistructured activities provide assessment (30–45 min depending on module selected) Modified Checklist for Autism in Toddlers; STAT Observational diagnostic Parent Questionnaire (15 min) 40-items derived from the Autism = Autism Screening Questionnaire; ADOS = adulthood y; over 6 y of Administration † ) III 18 mo through II Two versions: 4–6 Continued ( Checklist for Autism in Toddlers; M-CHAT = DiLavore, & Risi (1999) Lord, Pickles, & Bailey (1999) and references screening Targeted age format & time Content description Sensitivity & specificity ADOS Lord, Rutter, Screening tool* Level ASQ Berument, Rutter, Level I: A general population screening—for use by primary care specialists (eg, health professionals; pediatricians) to screen for children at ris CHAT Developmental Disorders Screening Test; ASQ Table 1. ∗ † population. Level II: A specialized screening for children known to have developmental concerns and suspected of having autism—for use by early inte care practitioners working in settings for children with known developmental delays. Level III: An autism specific screening or diagnostic tool—fo autism specialty clinics to determine thespecified). particular type and/or severity of autism spectrum disorder (eg, autistic disorder vs pervasive developm LWW/IYC AS188-01 May 30, 2003 11:17 Char Count= 0

Toddlers With Autism 211

prior to 3 years of age and these may intensify dependencies and effects among various do- during the preschool years (Lord, 1995). mains of development we have discussed. For example, orienting and responsiveness to stimuli in the environment, imitation, and CONCLUSIONS joint attention are likely pivotal skills for more self-regulated and efficient learning of play, Although we have considerably more language, and other social behaviors, and to learn about the common patterns and should be areas for intensive early interven- individual differences in development of tion. Giving early attention to the adaptations infants and toddlers with autism spectrum and strategies that improve the child’s ability disorders, recent research has greatly broad- to comprehend communication will reduce ened our knowledge in this area. A number frustration for both the child and caregivers, of instruments can be used to assist clinicians and will serve to prevent the development of in screening or diagnosing young children maladaptive behaviors related to communica- suspected of having disorders in the autism tive failures. In considering expressive com- spectrum. The evidence to date suggests that munication skills for this age group, we want we can reliably diagnose autism in some chil- to facilitate the development of adaptive ver- dren as young as 18 months of age, but that bal and nonverbal means of communication we will fail to recognize the disorder in many rather than focusing solely on the develop- toddlers who will later show clear symptoms ment of verbal communication. In addition, of autism spectrum disorders. Although we our intervention efforts will put priority on currently have the knowledge and tools to expanding the frequency of intentional com- identify many toddlers with autism spectrum munication expressed by the child, followed disorders who can benefit from early inter- by expanding the range of functions to in- vention services, the feasibility of diagnosing clude more social communicative functions. children below the age of 18 months has not Play is also a critical component of the early in- yet been systematically addressed. We also tervention curriculum for this population, to need to continue to advocate for and advance provide contexts for increased social aware- the research needed to develop more sensi- ness and interactions with adults and peers, tive screening and diagnostic instruments for and to promote more consistent levels of en- very young children with autism spectrum gagement with objects in the environment disorders. Large-scale prospective research and a symbolic level of processing. Of course, studies are needed to address methodological the needs of young children with autism and limitations inherent in previous studies, and their families must be addressed on an individ- to provide a more accurate representation ualized basis, but applying our current knowl- of the specificity, sensitivity, and predictive edge of development in infants and toddlers power of these assessments. with autism will greatly enhance our abil- Implications for early intervention include ity to implement effective early intervention recognizing the sequential and transactional programs.

REFERENCES

Adrien, J. L., Lenoir, P.,Martineau, J., Perrot, A., Hameury, from family home movies: Evaluation and compari- L., Larmande, C., et al. (1993). Blind ratings of early son between 1st and 2nd year of life using I.B.S.E. symptoms of autism based upon family home movies. Scale. Acta Paedopsychiatrica: International Journal Journal of the American Academy of Child and Ado- of Child and Adolescent Psychiatry, 55, 71–75. lescent Psychiatry, 32, 617–626. American Psychiatric Association. (1994). Diagnostic Adrien, J. L., Barthelemy, C., Perrot, A., Roux, S., Lenoir, and statistical manual of mental disorders (4th ed.). P., Hameury, L., et al. (1992). Early symptoms in autism Washington, DC: American Psychiatric Association. LWW/IYC AS188-01 May 30, 2003 11:17 Char Count= 0

212 INFANTS AND YOUNG CHILDREN/JULY–SEPTEMBER 2003

Baranek, G. T. (1999). Autism during infancy: A retrospec- Filipek, P. A., Accardo, P. J., Baranek, G. T., Cook, E. H., tive video analysis of sensory-motor and social behav- Jr., Dawson, G., Gordon, B., et al. (1999). The screen- iors at 9–12 months of age. Journal of Autism and ing and diagnosis of autistic spectrum disorders. Jour- Developmental Disorders, 29, 213–224. nal of Autism & Developmental Disorders, 29, 439– Baron-Cohen, S., Allen, J., & Gillberg, C. (1992). Can 484. autism be detected at 18 months? The needle, the Filipek, P. A., Accardo, P. J., Ashwal, S., Baranek, G. T., haystack, and the CHAT. British Journal of Psychia- Cook, E. H., Jr., Dawson, G., et al. (2000). Practice pa- try, 161, 839–843. rameter: Screening and diagnosis of autism. Report of Baron-Cohen, S., Cox, A., Baird, G., Swettenham, J., Night- the Quality Standards Subcommittee of the American engale, N., Morgan, K., et al. (1996). Psychological Academy of Neurology and the Child Neurology Soci- markers in the detection of autism in infancy in a large ety. Neurology, 55, 468–479. population. British Journal of Psychiatry, 168, 158– Frith, U. (1996). Cognitive explanation of autism. Acta 163. Paediatrica, 416(Suppl), 63–68. Baron Cohen, S., Leslie, A. M., & Frith, U. (1985). Does Frith, U., & Happe, F. (1994). Autism: Beyond “theory of the autistic child have a “theory of mind”? Cognition, mind.”Cognition, 50, 115–132. 21, 37–46. Ghaziuddin, M., Butler, E., Tsai, L., & Ghaziuddin, N. Bates, E. (1976). Language in context. New York: Aca- (1994). Is clumsiness a marker for Asperger syndrome? demic Press. Journal of Intellectual Disability Research, 38, 519– Belsky, J., & Most, R. K. (1981). From exploration to play: 527. A cross-sectional study of infant free play behavior. De- Gillberg, C., Ehlers, S., Schaumann, H., Jakobsson, G., velopmental Psychology, 17, 630–639. Dahlgren, S. O., Lindblom, R., et al. (1990). Autism un- Berument, S. K., Rutter, M., Lord, C., Pickles, A., & Bailey, der age 3 years: A clinical study of 28 cases referred A. (1999). Autism screening questionnaire: Diagnostic for autistic symptoms in infancy. Journal of Child Psy- validity. British Journal of Psychiatry, 175, 444–451. chology and Psychiatry and Allied Disciplines, 31, Chapman, R. (1978). Comprehension strategies in chil- 921–934. dren. In J. F. Kavanaugh & W. Strange (Eds.), Speech Grandin, T. (1997). A personal perspective on autism. and language in the laboratory, school and clinic In D. J. Cohen & F. R. Volkmar (Eds.), Handbook of (pp. 309–327). Cambridge, MA: MIT Press. autism and pervasive developmental disorders (2nd Charman, T., Swettenham, J., Baron-Cohen, S., Cox, A., ed., pp. 1032–1042). New York: Wiley. Baird, G., & Drew, A. (1997). Infants with autism: An Greenspan, S. I., & Wieder, S. (1997). Developmental pat- investigation of empathy, pretend play, joint attention, terns and outcomes in infants and children with dis- and imitation. Developmental Psychology, 33, 781– orders of relating and communicating: A chart review 789. of 200 cases of children with autistic spectrum disor- Dahlgren, S. O., & Gillberg, C. (1989). Symptoms in the ders. The Journal of Developmental and Learning first two years of life. A preliminary population study Disorders, 1, 87–141. of infantile autism. European Archives of Psychiatry Hoshino, Y., Kumashiro, H., Yashima, Y., Tachibana, R., & Neurological Sciences, 238, 169–174. Watanabe, M., & Furukawa, H. (1982). Early symptoms Dawson, G., Meltzoff, A. N., Osterling, J., Rinaldi, J., & of autistic children and its diagnostic significance. Fo- Brown, E. (1998). Children with autism fail to orient to lia Psychiatrica et Neurologica Japonica, 36(4), 367– naturally occurring social stimuli. Journal of Autism 374. and Developmental Disorders, 28, 479–485. Hughes, C., & Russell, J. (1993). Autistic children’s diffi- DeMyer, M. K., Barton, S., & Norton, J. A. (1972). A com- culty with mental disengagement from an object: Its parison of adaptive, verbal, and motor profiles of psy- implications for theories of autism. Developmental chotic and non-psychotic subnormal children. Jour- Psychology, 29, 498–510. nal of Autism and Childhood Schizophrenia, 2, 359– Johnson, M. H., Siddons, F., Frith, U., & Morton, J. (1992). 377. Can autism be predicted on the basis of infant screen- DiLavore, P.C., Lord, C., & Rutter, M. (1995). Pre-linguistic ing tests? Developmental Medicine & Child Neurol- autism diagnostic observation schedule. Journal of ogy, 34, 316–320. Autism and Developmental Disorders, 25(4), 355– Jones, V., & Prior, M. R. (1985). Motor imitation abilities 379. and neurological signs in autistic children. Journal of Doctoroff, S. (1996). Parents’ support of infants’ object Autism and Developmental Disorders, 15, 37–46. play: A review of the literature. Infant Toddler Inter- Libby, S., Powell, S., Messer, D., & Jordan, R. (1998). Spon- vention, 6, 153–166. taneous play in children with autism: A reappraisal. Fein, D., Dunn, M., Allen, D., Aram, D., Hall, M., Morris, Journal of Autism and Developmental Disorders, 28, R., et al. (1996). Language and neuropsychological 487–497. findings. In I. Rapin (Ed.), Preschool children with Lord, C. (1995). Follow-up of two-year-olds referred for inadequate communication (pp. 123–154). London: possible autism. Journal of Child Psychology and Psy- MacKeith. chiatry and Allied Disciplines, 36, 1365–1382. LWW/IYC AS188-01 May 30, 2003 11:17 Char Count= 0

Toddlers With Autism 213

Lord, C., Rutter, M., DiLavore, P. C., & Risi, S. (1999). The adolescents with autism. Child Development, 67, Autism Diagnostic Observation Schedule. Los Ange- 2060–2073. les: Western Psychological Services. Siegel, B. (1998, June 15–17). Early screening and diag- McGee, G. G., Feldman, R. S., & Morrier, M. J. (1997). nosis in autism spectrum disorders: The Pervasive Benchmarks of social treatment for children with Developmental Disorders Screening Test (PDDST). autism. Journal of Autism and Developmental Dis- Paper presented at the NIH State of the Science in orders, 27, 353–364. Autism Screening and Diagnosis Working Conference, Meltzoff, A. N., & Moore, M. K. (1998). Infant intersub- Bethesda, MD. jectivity: Broadening the dialogue to include imitation, Sigman, M., Ruskin, E., Arbeile, S., Corona, R., Dis- identity and intention. In S. Braten (Ed.), Intersubjec- sanayake, C., Espinosa, M., et al. (1999). Continuity tive communication and emotion in early ontogeny. and change in the social competence of children with Studies in emotion and social interaction (pp. 47– autism, Down syndrome, and developmental delays. 62). New York: Cambridge University Press. Monographs of the Society for Research in Child De- Morales, M., Mundy, P., & Rojas, J. (1998). Following velopment, 64, 1–114. the direction of gaze and language development in Smith, I. M., & Bryson, S. E. (1998). Gesture imitation in 6-month-olds. Infant Behavior and Development, 21, autism: I. Nonsymbolic postures and sequences. Cog- 373–377. nitive Neuropsychology, 15, 747–770. Mundy, P. (1995). Joint attention and social-emotional ap- Stella, J. (2001, April). Social orienting in children with proach behavior in children with autism. Develop- autism: Concurrent and predictive validity. Poster ment and Psychopathology, 7, 63–82. session presented at the meeting of the Society for Re- Mundy, P., & Gomes, A. (1998). Individual differences in search in Child Development biennial meeting, Min- joint attention skill development in the second year. neapolis, MN. Infant Behavior and Development, 21, 469–482. Stone, W. L., Coonrod, E. E., & Ousley, O. Y. (2000). Ohta, M., Nagai, Y., Hara, H., & Sasaki, M. (1987). Parental Screening Tool for Autism in Two-Year-Olds (STAT): perception of behavioral symptoms in Japanese autis- Development and preliminary data. Journal of Autism tic children. Journal of Autism and Developmental and Developmental Disorders, 30, 607–612. Disorders, 17, 549–563. Stone, W. L., Lee, E. B., Ashford, L., Brissie, J., Hepburn, Ornitz, E. M., Guthrie, D., & Farley, A. H. (1977). The early S. L., Coonrod, E. E., et al. (1999). Can autism be di- development of autistic children. Journal of Autism agnosed accurately in children under 3 years? Journal and Developmental Disorders, 7, 207–229. of Child Psychology and Psychiatry, 40, 219–226. Osterling, J., & Dawson, G. (1994). Early recognition of Stone, W. L., & Lemanek, K. L. (1990). Parental report children with autism: A study of first birthday home of social behaviors in autistic preschoolers. Journal videotapes. Journal of Autism and Development Dis- of Autism and Developmental Disorders, 20, 513– orders, 24, 247–257. 522. Ozonoff, S. (1997). Components of executive function in Stone, W.L., Lemanek, K. L., Fishel, P.T., Fernandez, M. C., autism and other disorders. In J. Russell (Ed.), Autism & Altemeier, W. A. (1990). Play and imitation skills in as an executive disorder (pp. 179–211). New York: the diagnosis of autism in young children. Pediatrics, Oxford University Press. 86, 267–272. Pennington, B. F., & Ozonoff, S. (1996). Executive func- Stone, W.L., Ousley, O. Y., Hepburn, S. L., Hogan, K. L., & tions and developmental psychopathology. Journal of Brown, C. S. (1999). Patterns of adaptive behavior in Child Psychology and Psychiatry and Allied Disci- very young children with autism. American Journal plines, 37, 51–87. on Mental Retardation, 104, 187–199. Reddy, V., Hay, D., Murray, L., & Trevarthen, C. (1997). Stone, W. L., Ousley, O. Y., Yoder, P. J., Hogan, K. L., & Communication in infancy: Mutual regulation of af- Hepburn, S. L. (1997). Nonverbal communication in fect and attention. In G. Bremmer, A. Slater, & G. two- and three-year-old children with autism. Journal Butterworth (Eds.), Infant development: Recent ad- of Autism and Developmental Disorders, 27, 677– vances (pp. 247–273). East Sussex, England: Psychol- 696. ogy Press. Stone, W. L., Ousley, O.Y., & Littleford, C. D. (1997). Robins, D. L., Fein, D., Barton, M. L., & Green, J. A. (2001). Motor imitation in young children with autism: What’s The Modified Checklist for Autism in Toddlers: An ini- the object? Journal of Abnormal Child Psychology, tial study investigating the early detection of autism 25, 475–485. and pervasive developmental disorders. Journal of Swettenham, J., Baron-Cohen, S., Charman, T., Cox, A., Autism and Developmental Disorders, 31, 131–144. Baird, G., Drew, A., et al. (1998). The frequency and Rogers, S. J., & Pennington, B. F. (1991). A theoretical ap- distribution of spontaneous attention shifts between proach to the deficits in infantile autism. Development social and nonsocial stimuli in autistic, typically de- and Psychopathology, 3, 137–162. veloping, and nonautistic developmentally delayed in- Rogers, S. J., Bennetto, L., McEvoy, R., & Pennington, B. F. fants. Journal of Child Psychology and Psychiatry (1996). Imitation and pantomime in high-functioning and Allied Disciplines, 39, 747–753. LWW/IYC AS188-01 May 30, 2003 11:17 Char Count= 0

214 INFANTS AND YOUNG CHILDREN/JULY–SEPTEMBER 2003

Teitelbaum, P.,Teitelbaum, O., Nye, J., Fryman, J., & Mau- jectivity: Confidence, confiding and acts of meaning rer, R. (1998). Movement analysis in infancy may be in the first year. In A. Lock (Ed.), Action, gesture and useful for early diagnosis of autism. Proceedings of symbol: The emergence of language (pp. 183–229). the National Academy of Sciences, 95(23), 13982– London: Academic Press. 13987. Volkmar, F. R., Cohen, D. J., & Paul, R. (1986). An evalu- Trevarthen, C. (1979). Communication and cooperation ation of the DSM-III criteria for infantile autism. Jour- in early infancy: A description of primary intersubjec- nal of the American Academy of Child Psychiatry, tivity. In M. Bullowa (Ed.), Before speech (pp. 321– 25, 190–197. 347). New York: Cambridge University Press. Wing, L. (1981). Asperger’s syndrome: A clinical account. Trevarthen, C., & Hubley, P. (1978). Secondary intersub- Psychological Medicine, 11, 115–129.

Current articles- http://depts.washington.edu/isei/iyc/iyc_comments.html

Back to lists of previous articles- http://depts.washington.edu/isei/iyc/iyc_previous.html