“Showing” Gesture in Children with Autism Spectrum Disorder

Beyond Pointing: Development of the “Showing” Gesture in Children with Autism Spectrum Disorder

Caitlin Clements & Katarzyna Chawarska

YALE UNIVERSITY

While research on early indicators of autism spectrum disorder (ASD) has focused predominantly on infants’ inability to point, other social gestures such as showing may also help to distinguish ASD from non- autistic language delays in high-risk populations. Parents of infants at high-risk (HR) for ASD ( n = 28) and typically developing infants at low-risk (TD-LR; n = 12) reported their child’s gesture use at nine and twelve months. Infants were followed until 24 months, at which time 8 high-risk children showed symptoms of ASD (ASD-HR), 14 showed symptoms of language delay (LD-HR), and 6 were developing typically (TD-HR). Results indicate a lack of pointing in both ASD-HR and LD- HR groups at twelve months, but decreased frequency of showing only in the ASD-HR group. These results suggest that the showing gesture might be a better indicator of ASD than pointing among infant siblings of children with ASD. Implications for early identification are discussed.

INTRODUCTION early recognition and diagnosis must rely on indicators of nonverbal communication Autism Spectrum Disorder skills that typically emerge before twelve Autism Spectrum Disorder (ASD) is a group months of age. of neurodevelopmental disorders character- ized by profound deficits in communication, Nonverbal communication social functioning, and the presence of re- Nonverbal communication skills develop petitive and stereotyped behaviors or inter- before verbal language and therefore can ests (American Psychiatric Association, play an important role in identification of 2004). It affects 1 in 110 children in the US ASD during the first year of life (Baron- (CDC, 2009) and in most cases impairs the Cohen, Cox, Baird, Sweettenham, & Night- child’s ability to function independently ingale, 1996; Chawarska et al., 2007; Lord, (Howlin, 2005). Because there is currently 1995; Mundy, Sigman, Ungerer, & Sherman, no known cure, to date the best means of 1986; Paul, 2007; Stone, Ousley, Yoder, Ho- mitigating the negative effects of ASD are gan, & Hepburn, 1997; Wetherby, 1986). early identification. This way, children can Nonverbal communication behaviors such begin an early behavioral intervention pro- as pointing, waving goodbye, and smiling gram (Klin, Chawarska, Rubin, & Volkmar, convey a message to another person without 2004). language (Tager-Flusberg et al., 2005). Nonverbal communication is particularly ASD and nonverbal communication relevant to early ASD diagnosis because Parents and clinicians use early indicators, children use hand gestures such as reach- such as delayed speech and language, to ing, pointing, and giving to express their recognize ASD during the second year of life desires before they can do so with verbal (Chawarska et al., 2007). Though commu- language. In typical development, gestures nication is often impaired in ASD (American emerge around eight to ten months and are Psychiatric Association [ DSM-IV-TR ], 2004), used frequently until around 18 months, identification during the first year of life when their use begins to fade as verbal lan- cannot depend on verbal language alone. guage develops to replace them (Bates & Even in typical development, children do not Dick, 2002; Capone & McGregor, 2004). speak their first words until around twelve months of age and do not combine words Pointing into phrases until around 18 months (Tager- The vast majority of studies on nonverbal Flusberg, Paul, & Lord, 2005). Therefore, communication in ASD have focused on

The authors would like to give special thanks to Jessie Bradshaw, Suzanne Macari, and Fred Shic. The first author can be reached at [email protected] or at Massachusetts General Hospital, Simches Research Bldg., 6th floor, 185 Cambridge St., Boston, MA 02114.

47 THE YALE REVIEW OF UNDERGRADUATE RESEARCH IN PSYCHOLOGY

pointing gestures. Pointing emerges around 3.5 years, Baron-Cohen et al. (1996) found eight to ten months in typically developing that a deficit in protodeclarative pointing children (Bates & Dick 2004; Capone & was one of the three best predictors of au- McGregor, 2004), and accounts for the ma- tism. (Gaze monitoring and pretend play jority (60%) of gestures by twelve months also predicted autism.) An early deficit in (Kita, 2003). There are two types of point- protodeclarative pointing is consistent with ing: protodeclarative and protoimperative the social impairments that characterize au- (Paul, 2007; Tomasello, 2007). Protoimpera- tism, and other highly social gestures are tive points represent desire for an object, likely to be impaired early as well. e.g., pointing to a cookie. Protodeclarative points indicate the desire to share an experi- Other protodeclarative gestures: giving and ence with another person, e.g., a child showing pointing to an airplane overhead. Protode- Though pointing is a crucial indicator of so- clarative pointing requires joint attention, or cial functioning, other gestures such as the ability to share experiences with others showing and giving also reflect social skills by attracting or following their attention by and interest. Showing refers to a person’s looking or pointing (Paul, 2007; Tomasello, arm extending toward another person’s face 2007; Loveland & Landry, 1986). Joint atten- while holding an object (Wetherby, Yonclas, tion is an early emerging social cognitive & Btyan, 1989; Stone et al., 1997). Giving re- skill and is impaired in autism (Dawson et fers to placing an object in another person’s al., 2004; Loveland & Landry, 1986). hand or pushing an object at least halfway toward another person (Wetherby et al., Pointing in Autism Spectrum Disorders 1989; Stone et al., 1997). While many in- In autism, protodeclarative pointing is im- stances of giving serve protodeclarative pur- paired while protoimperative pointing re- poses (e.g., giving an interesting object to a mains relatively intact, especially in older parent), others serve protoimperative pur- children (Baron-Cohen, 1989; Camaioni, poses (e.g., handing a jar to a parent to get it Perucchini, Muratori, & Milone, 1997; Ca- opened) or occur in the context of sharing maioni, Perucchini, Muratori, Parrinini, & (e.g., giving a cracker or a toy to another Cesari, 2003; Itoh, 2000). Because protode- person). Both showing and giving emerge clarative pointing requires social interest, its around eight to ten months in typically de- deficit is considered one of the critical ASD veloping children, concurrently with point- risk indicators on a number of screening ing (Capone & McGregor, 2004; Bates & instruments, including the Checklist for Au- Dick, 2004). tism in Toddlers (CHAT; Baron -Cohen, Al- len, & Gillberg, 1992) and Modified Check- Giving and showing in Autism Spectrum list for Autism in Toddlers (M-CHAT; Rob- Disorder ins, Fein, Barton, & Green, 2001). Pointing Little is known about the early development deficits have proven to be an effective flag of giving and showing in children with ASD, for autism as measured by the longitudinal but studies of older children with ASD have stability of CHAT results. In a study that fol- found deficits in both types of gestures. lowed 16,000 children from 18 months to Children with autism at 52 months were less

BEYOND POINTING CLEMENTS & CHAWARSKA 48 likely to give, take, and show objects in re- found that of 469 Danish people diagnosed sponse to adult cues compared to develop- with a developmental language delay as mentally delayed children (Leekam & Rams- children, 10 later carried a diagnosis of ASD den, 2006). At 49 months, children with in adulthood (Mouridsen & Hauschild, ASD showed similar differences in frequen- 2009). Deficits in social gestures may pro- cy of giving (DiLavore, Lord, & Rutter, vide an extra indicator in these ambiguous 1995). Children with ASD demonstrated situations. Wetherby and Prizant (2002) fewer “gestural joint attention skills” (de- found that a deficit in showing differentiat- fined as pointing and showing) at 45 ed children with ASD from both develop- months when compared to children with mentally delayed and typically developing mental retardation matched on language, children. Protodeclarative pointing, howev- mental age, or IQ (Mundy, Sigman, & Kasa- er, did not distinguish children with ASD ri, 1990). Wetherby, Watt, Morgan, and from developmentally delayed children. A Shumway (2007) found that a strong predic- similar problem was also noted in the Bar- tor of ASD between 24 and 36 months was on-Cohen et al. (1996) study on 18-month “inventory of gesture,” which assessed the olds. These results suggest that other signs children’s use of eight gestures and includ- beyond pointing are necessary for distin- ed showing, pointing and giving. In an ear- guishing between these groups. Showing lier prospective study, the authors found and giving, which depend heavily on social that limited showing was a red flag that dif- skills, may serve this purpose. ferentiated children with ASD from both developmentally delayed and typically develop- Diagnosis in high-risk populations ing children. Frequency of pointing, howev- Distinguishing ASD from similar non- er, did not differ between children with ASD autistic disorders is of particular concern and those with developmental delays. When when diagnosing children at high risk for Stone et al. (1997) measured giving and ASD. Children with an older sibling with showing within the same age range, a sig- ASD have increased genetic risk for develop- nificant difference was found for showing ing the disorder (approximately 20%; Rog- but not for giving between children with ers, 2009), and a substantial proportion of ASD and developmentally delayed and lan- the high-risk group (18-30%; Landa, Hol- guage-impaired children. These studies in- man, & Garrett-Mayer, 2007) do not meet dicate that multiple aspects of nonverbal diagnostic criteria for ASD but do exhibit communication may be impaired in ASD other symptoms such as linguistic disorders and suggest that nonverbal gestures beyond and impaired social, cognitive, and linguis- pointing may also serve as early indicators tic skills. The siblings without ASD often of the disorder. show early deficits similar to children with ASD, such as diminished pointing, social Importance of studying giving and showing smiling, receptive language, and overall use in diagnosis of gestures and words (Toth, Dawson, Melt- More data is needed to disambiguate ASD zoff, Greenson, & Fein, 2007). Since the from other symptomologies, especially population of infants at high risk is often when diagnosed early. A longitudinal study closely monitored by clinicians for signs of

49 THE YALE REVIEW OF UNDERGRADUATE RESEARCH IN PSYCHOLOGY

ASD, robust indicators are needed to differ- STUDY OVERVIEW entiate those who develop non-autistic symptomology from those who develop tra- The present study explores the development ditional ASD. Studies have shown that of pointing, showing, and giving at twelve pointing may not serve this purpose because months in children who were later diag- deficits in pointing are present both in chil- nosed with ASD. The children were at high dren with ASD and in those with develop- risk (approximately 20%; Rogers, 2009) for mental delays (Toth et al., 2007). Investiga- developing the disorder in comparison to tion of other protodeclarative gestures such the average American child (< 1%; CDC, as giving and showing may help disambigu- 2009) because they had an older sibling with ate the two groups (Leekam & Ramsden, an ASD. Studying the high-risk population 2006; Stone et al., 1997; Wetherby et al., from birth yields insights into the early de- 2007). velopment of ASD (Zwaigenbaum et al., 2009). This study assessed early gesture de- Gestures tend not to be studied individually velopment in infants who later showed Although gestures such as giving and show- symptoms of ASD or language delays at 24 ing may provide valuable diagnostic infor- months. The main objective was to investi- mation, they have yet to be carefully studied. gate the individual profiles of giving, show- Studies that do include measurements of ing and pointing, and to compare the pat- showing or giving usually analyze “gestur- terns of development in children with ASD ing” as a single construct or focus on point- to patterns among typically developing chil- ing and do not analyze giving or showing dren as well as those with language delays. independently. Charman, Drew, Baird, and We hypothesize the following: Baird (2003) investigated summary scores on the MacArthur Communication Devel- (1) ASD infants will generally gesture less opment Inventories (CDI) of “early” and frequently than language delayed (LD) and “late” gestures but did not analyze differ- typically developing (TD) infants. ences in scores on specific items that assess showing and giving independently. Mitchell (2) Pointing will be diminished in the ASD et al. (2006) replicated the same study with a and LD groups compared to the TD group. younger sample and more control groups, Showing and giving will be diminished in but again focused on “early” and “late” ges- the ASD group alone. tures instead of the individual development of giving or showing. These studies yielded (3) Across development from nine to twelve interesting results about global gesturing months, children who are later diagnosed but did not explore profiles of specific ges- with ASD will exhibit fewer gains in gestures tures such as giving and showing. Data on compared to children in the LD and TD the development of individual gestures groups. could inform the dialog on very early indicators and aid in distinguishing language disorders from ASD.

BEYOND POINTING CLEMENTS & CHAWARSKA 50

Method ADOS-T scale than the remaining diagnostic groups at 24 months (see Table 1). Participants

Forty participants were recruited from an ASD Inclusion Criteria ongoing prospective longitudinal study of Eight participants at risk for ASD received a early social-cognitive development. Twenty- provisional diagnosis of ASD at 24 months. eight of the infants had an older sibling with A team of two experienced clinicians evalu- ASD and therefore were at increased genetic ated each child over two days and assigned a risk for developing it. Gesture proficiency provisional diagnosis based on DSM-IV cri- was measured at nine and twelve months teria modified for children under the age of using the MacArthur Communication De- three (Chawarska & Volkmar, 2005) and velopment Inventories (CDI). Based on as- performance on the Autism Diagnostic Ob- sessments at 24 months, infants in the high- servation Schedule–Toddler Module (Lord et risk for ASD group were divided into three al., 2000), Vineland Adaptive Behaviors sub-groups: those diagnosed with ASD Schedule-Expanded (Sparrow, Balla, & Cic- (ASD-HR; n = 8); language and communica- chetti, 1984), and Mullen Scales of Early tion delay but not ASD (LD-HR; n = 14); and Learning (Mullen, 1995). infants developing typically (TD-HR; n = 8). A fourth group included typically developing Language Delay Inclusion Criteria children with no familial history of ASD and The LD group included fourteen partici- thus considered at low-risk for developing pants at risk for ASD who showed signifi- the disorder (TD-LR; n = 12). Males repre- cant communicative delays at 24 month as- sented 42% of the TD-LR group, compared sessments but did not receive a diagnosis of to 33% of the TD-HR group, 93% of the LD- ASD. Five participants received a clinical di- HR group, and 88% of the ASD group (see agnosis of a language-related disorder (ex- Table 1). Participants in the TD-LR, TD-HR, pressive language disorder, receptive lan- and LD-HR groups did not differ in level of guage disorder, developmental delay, and development as measured by the five scales two with mixed language disorder). In nine of the Mullen Scales of Early Learning (re- participants, expert clinicians identified im- ceptive language, expressive language, visu- paired communicative functioning and so- al reception, fine motor skills, and gross cial deficits in the subclinical range (Broader motor skills), with the exception of visual Autism Phenotype, BAP). reception (see Table 1). Participants in the ASD-HR and LD-HR groups also had com- Procedure parable scores on the developmental test Parents reported their infants’ gesture use (see Table 1), but infants with ASD had low- using the MacArthur Communication and er scores than the TD-HR group on expres- Development Inventories (CDI), which pro- sive language scales and lower scores than vides a comprehensive profile of a child’s the TD-LR group on expressive language, communication skills (Fenson et al., 1993). visual reception, and fine motor scales. As The present study used responses on the expected, children in ASD-HR group had Words and Gestures version in the subsec- significantly higher total scores on the tion “First Communicative Gestures.” This

51 THE YALE REVIEW OF UNDERGRADUATE RESEARCH IN PSYCHOLOGY Table 1. Group m eans (standard deviation) on Mullen Scales of Early Learning and ADOS-T tests at 24 months.

BEYOND POINTING CLEMENTS & CHAWARSKA 52

sub-section contains 12 items and asks par- child performed the gestures “often,” ents to report how frequently their child us- “sometimes,” or “not yet.” For the purposes es particular gestures. Sample gestures as- of this study, responses were coded accord- sessed include showing, giving, pointing, ing to whether the child performed the ges- waving goodbye, and nodding. Parents read ture proficiently, indicated by frequent use a description of the gesture, such as “Ex- and a response of “often.” This coding tends arm to show you something he/she is scheme was employed because the gestures holding” and then checked whether the of interest were those used functionally, not

Table 2. Proportion of children in ASD -HR, LD -HR, TD -HR , and TD -LR group s demonstrating gesture “often” based on parent report. Gestures that occurred infrequently were excluded from analysis.

53 THE YALE REVIEW OF UNDERGRADUATE RESEARCH IN PSYCHOLOGY

Figure 1 . Average number of functional Figure 2. Proportion of children with ASD gestures (± 1 SE) in child’s repertoire and children with LD who often point or based on parent report for children at high show at 12 months. At twelve months, risk for ASD and diagnosed with ASD, LD, parents of children at risk for developing or typically-developing and typically- ASD reported how frequently their child developing children at low risk. Gestures exhibited showing or pointing. A larger that occurred infrequently (ie., exhibited proportion of children who were diag- often by fewer than 25% of all groups) nosed with a language delay at 24 were omitted. months often showed objects than children who were later diagnosed with ASD. There was less difference between groups in the proportion that frequently pointed at 9 months.

Figure 3 . Propo rtion of ASD -HR, LD -HR, Figure 4 . Propo rtion of ASD -HR, LD -HR, TD-HR, and TD-LR children who often TD-HR, and TD-LR children who did not reach to be picked up at 9 months. demonstrate functional use of any of the five gestures of interest (point, show, give, reach to be picked up, and waves goodbye) at 9 months and did develop the gesture by 12 months.

BEYOND POINTING CLEMENTS & CHAWARSKA 54

those still in development and used only HR, LD-HR, TD-HR, TD-LR) & 2 (time: 9 sometimes. This study examined partici- months, 12 months) ANOVA on the total pants’ scores on each individual gesture, number of gestures with repeated measures which is notable because previous studies on the last factor indicated a significant ef- combined all individual gestures into a sin- fect of time, F(1, 36) = 38.99, p < .001, and gle summary score (Charman et al., 2003; no effect of diagnosis F(3, 36) = 1.47, p < Mitchell et al., 2006). Each parent complet- .238, or diagnosis time interaction F(3, 36) ed the CDI as part of a larger battery of sur- = .24, p < .868. To test our specific hypothe- veys mailed 2 to 3 weeks prior to their sis regarding between-group differences, we child’s scheduled 9- and 12-month visits. followed up with a post-hoc test comparing Parents returned the surveys at the time of the four groups. Infants with ASD had sig- the visit, which was scheduled within a nificantly fewer gestures than infants in the month to the date of the chronological age TD-LR group t(36) = -2.09, p < .043, d = - (with two exceptions when the child could .70, but differences between other groups not visit until 2 and 3 months following were not significant (ASD-HR versus LD- their birthday). HR: p < .230; ASD-HR versus TD-HR: p < .363)). Results General Frequency of Gesturing Patterns in Pointing, Showing and Giving at Twelve Months Frequencies of Gestures At both nine and twelve months, seven of Pointing the twelve gestures were reported as occur- Children with autism pointed less frequently ring “not yet” or only “sometimes” in at than all other children at twelve months. In least 89.8% of children at nine and twelve comparison to the TD-LR group, the group months across all groups. Therefore, these with ASD pointed less than half as frequent- gestures were excluded from subsequent ly (58.3% and 25.0%, respectively), although analyses. At nine months, the only gesture the difference failed to reach significance, that occurred frequently enough by all chi-square (1, 20) = 2.16, p = .142. Children groups to allow for analysis was “extending in the ASD-HR and LD-HR groups pointed one’s arms to be picked up.” At twelve at similar frequencies (25.0% and 35.7%, months, infants demonstrated five gestures respectively), chi-square (1, 22) = .27, p = frequently enough for analysis: pointing, .604. showing, giving, waving goodbye, and extending one’s arms to be picked up. Showing In contrast to pointing, the frequency of Limited frequency of gestures in children showing differed significantly between the with ASD ASD-HR and LD-HR groups at twelve A visual inspection of Figure 1 suggests that months. A chi-square test indicated a mar- children with ASD generally gestured less ginally significant difference between than other children at both time points. A groups in the proportion of participants mixed models 4 (diagnostic group: ASD- demonstrating proficiency in the showing

55 THE YALE REVIEW OF UNDERGRADUATE RESEARCH IN PSYCHOLOGY

gesture (ASD-HR: 37.5%; LD-HR: 78.6%; Giving chi-square (1, 22) = 3.71, p = .054). The fre- The frequency of giving did not differ signif- quency of showing was similar across the icantly between groups at twelve months. LD-HR, TD-LR (75.0%), and TD-HR Parents reported the frequency “often” for (66.7%) groups. 50.0% of children with ASD, 64.3% of LD- A binary logistic regression was con- HR children, 66.7% of TD-HR children, and ducted to determine which gestures (i.e., 50.0% of TD-LR children. pointing, showing, giving, waving goodbye, and extending arms to be picked up) at 12 Exploratory Analyses of Gesturing at Nine months best predicted that a twelve-month Months infant with some communicative delays would develop ASD instead of a language Nine-month old infants exhibited low fre- delay. We fitted a series of models to the da- quencies of most gestures. The only notable ta and began by including all five gestures result was the proportion of parents who with high frequency at twelve months as reported that their child “often” extended predictors of diagnostic outcome. The best his or her arms upward to signal a wish to fit model included only the showing and be picked up. In the typically developing pointing gestures, and showing was found groups, 58.3% of parents of low-risk infants to be a stronger predictor of diagnosis than and 50.0% of parents of high-risk infants pointing (showing: b = 2.30, p < .077; point- reported this frequency, as did 64.3% of ing: b = -.88, p < .522). The odds ratio, parents in the LD-HR group. A chi-square which indicates the likelihood of belonging analysis suggests a trend between these to a group, indicated that showing, not groups and the ASD-HR group, where the pointing, was a stronger predictor of diag- proportion was only 25.0%, chi-square (1, nosis (showing: e b = 10.00; pointing: e b = 40) = 3.03, p = .082. .42). The odds of belonging to the LD-HR group are 10 times greater than the odds of Group Differences in Acquisition of belonging to the ASD-HR group, given that Gestures the infant is proficient in showing. Thus, high-risk infants who engage in frequent The children with ASD exhibited fewer gains showing at 12 months are less likely to be in gestures between nine and twelve months diagnosed with ASD at 24 months than compared to the two typically developing high-risk infants who are not proficient in groups. Of the infants who did not exhibit the showing gesture at twelve months. any of the five gestures of interest (the most While the presence of either pointing or frequent gestures) at nine months, only 43% showing at 12 months increases the odds of the ASD-HR group developed pointing or that the infant belongs to the LD group, showing by twelve months, in contrast to those odds are much higher when the in- 80% of the TD-HR group and 78% of the fants use functional showing (10.00) as op- TD-LR group. Although these differences posed to functional pointing (.42). between groups are not significant, they il- lustrate a pattern of slower gesture acquisition in children with ASD.

BEYOND POINTING CLEMENTS & CHAWARSKA 56

DISCUSSION Children with ASD gesture less frequently The results of this study reaffirm previous Data suggest that children with ASD exhibit findings that frequency and repertoire of fewer gestures than typically developing gestures can provide important insights for children at low risk at both twelve and nine clinicians assessing children at risk for ASD months. This pattern replicates the results in their first year. The most notable finding of past studies that measured gestures with is the potential value of the showing gesture the CDI (Charman et al., 2003; Mitchell et as a predictor of ASD in high-risk children. al., 2006) and with other measures (Carpen- Due to the prospective nature of the study, ter, Pennington, & Rogers, 2002; DiLavore the number of infants included in the analy- et al., 1995; Leekam & Ramsden, 2006; sis was low, which contributed to the failure Mundy et al., 1990; Stone et al., 1997; of some observed effects to reach statistical Wetherby et al., 2007). ASD impairs com- significance. However, the study is ongoing munication, so it follows that early nonver- and we are currently analyzing data from a bal communicative forms such as gestures larger sample. Thus, the analyses presented should also be impaired. in this paper are necessarily preliminary. The results thus far indicate the following: Value of the showing gesture The frequency of showing might differenti- (1) Infants who are later diagnosed with ASD ate high-risk infants who later develop ASD may gesture less frequently than high-risk from those who will develop language de- infants who later develop language delays as lays or mild social difficulties. At twelve well as typically developing infants from months, both the ASD-HR and LD-HR both high- and low-risk samples. groups lacked proficiency in pointing. How- ever, only the ASD-HR group exhibited lim- (2) At twelve months, a lack of showing may ited frequency of showing . In fact, in this differentiate ASD-HR from LD-HR better study, if a parent reported at twelve months than a lack of pointing. that a child with communication difficulties and at risk for ASD often showed objects (3) At nine months, the only gesture and toys, the calculated odds that the child demonstrated by any group at a high fre- would be diagnosed with LD rather than quency was extending one’s arms upward to ASD were ten times greater at 24 months. be picked up, which was diminished in the Proficient pointing, however, did not ASD-HR group compared to all other strongly predict LD over ASD. This differ- groups. ence between pointing and showing could prove useful for clinicians during the diag- (4) Children with ASD may acquire fewer nostic process. Infants with increased ge- gestures between nine and twelve months netic risk for developing ASD, such as those compared to typically developing children. in this study, are often monitored closely for early signs of the disorder. The high-risk Each of these results will now be addressed infants in the LD group did not receive a di- in turn. agnosis of ASD by a team of expert clini-

57 THE YALE REVIEW OF UNDERGRADUATE RESEARCH IN PSYCHOLOGY

cians at 24 months, and early behaviors that Giving distinguish them from the high-risk infants Contrary to our hypothesis, the giving ges- who were diagnosed with ASD could help ture did not uniquely characterize the ASD clinicians diagnose earlier and more effec- group. In fact, similar proportions of chil- tively. dren across all groups demonstrated profi- In addition, if confirmed in the final ciency in giving. If parents report data that analysis, our finding on decreased frequency accurately reflects child behavior, it is re- of showing in infants with ASD could con- markable that infants with ASD do not ex- tribute to public discussion about early hibit deficits in giving, as previous research signs. Public education about autism tradi- has yielded contradictory results on the im- tionally emphasizes pointing as the hall- pairment of giving later in childhood (im- mark gesture that is absent in autism. How- paired: DiLavore et al., 1995; Mundy et al., ever, ASD impairs pointing in a complex 1986; not impaired: Stone et al., 1997). Our way that is difficult to communicate in a few results may reflect the versatility of the giv- bullet points on a waiting room pamphlet. ing gesture; children may give to share (pro- Protoimperative pointing remains intact in todeclarative) or to request something, such ASD, and only protodeclarative pointing, or as help with opening a toy (protoimpera- pointing to an object or event of shared in- tive). terest, is impaired (Baron-Cohen, 1989; Camaioni et al., 1997; Camaioni et al., 2003; Reaching to be picked up Itoh, 2000). Parents who are aware of the At nine months, few infants demonstrated commonly accepted “first signs” and who many gestures proficiently, with the excep- often witness their child point protoimpera- tion of extending their arms upward to be tively may falsely discount ASD as an expla- picked up. This result mirrors Kanner’s nation for their child’s social difficulties. (1943) original remarks about children with The showing gesture may offer a clearer al- autism, whom he claimed could be identi- ternative to pointing as a warning sign for fied in part by their failure to anticipate be- ASD, and the two gestures could comple- ing picked up by their parents. It was the ment each other in educational materials on only gesture demonstrated frequently by early signs. Unlike pointing, showing is more than a quarter of any group at nine used only to initiate a shared social experi- months, and therefore, as Kanner indicated, ence and not to indicate a need. A lack of it could be used to distinguish children with showing may be easier for some parents to autism from typically developing children. identify than a lack of pointing, which refers only to protodeclarative pointing and ex- Children with ASD develop fewer gestures cludes protoimperative pointing. In con- between nine and twelve months trast, pointing may be a more salient gesture The data suggest that children with ASD because it is marked by a very specific hand may develop fewer gestures between nine posture. Publicity about the early signs of and twelve months compared to typically ASD may be more effective if both pointing developing children. Previous studies indi- and showing are used. While the former is cate that gestures are impaired in 18-month more salient, the latter is more accurate. old infants who later develop ASD (Baron-

BEYOND POINTING CLEMENTS & CHAWARSKA 58

Cohen et al., 1996; Mitchell et al., 2006), but idate parent reports. In addition, during di- little is understood about the development agnostic evaluation, clinicians could assess of gestures in ASD during this time period. the type, frequency, and communicative function of gestures used, which would help Summary and Future Directions elucidate the ambiguities of gesture classifi- cation by parents . However, clinicians lack The data suggest that children with ASD ex- the extensive perspective of parents, whose hibit fewer gestures overall at nine and reports still contribute to the discussion of twelve months. In addition, it suggests that early signs of ASD. Parents, not trained ob- high-risk infants whose parents report that servers or clinicians, daily witness their they often show objects or toys at twelve child play and communicate in a naturalistic months are less likely to be diagnosed with setting and are in the best position to identi- ASD, but might show less severe impair- fy ASD early. Therefore, the publicly adver- ments involving language developing and tised “early signs of ASD” should be easy for mild social difficulties at 24 months. At nine parents to utilize, and therefore should be months, children with ASD extended their based on research about which signs were arms to be picked up half as frequently as actually observed by parents of children with typically developing children and children ASD. with language delays, regardless of risk sta- One implication of these results con- tus. Data also suggest that children with cerns future studies of gesture in ASD. ASD and LD may acquire fewer gestures be- While studies often combine all gestures tween nine and twelve months than their and report a composite score (Charman et typically developing peers. al., 2003; Leekam, 2006; Mitchell et al., Further investigation with a larger sam- 2006; Mundy et al., 1990; Wetherby et al., ple is needed to explore whether these 2007), the results of this study indicate that trends reflect true differences between the data on individual gestures should not groups. The larger subject pool from which be overlooked. Trends differed between spe- these participants were recruited is matur- cific gestures such as pointing, showing, ing, and the data is currently being analyzed and giving, and a summary score would for more children from this unique popula- have ignored these differences. Examining tion. Further research is also needed to explore the reliability of parent report data from the CDI. Parents were not trained and Parents may struggle to interpret descriptions of may vary in their interpretation of the CDI gestures on the CDI. For example, the CDI de- categories of frequency of “not yet,” “some- scribes the reaching gesture as “Requests something by extending arm and opening and closing times,” and “often.” We attempted to com- hand.” Few parents reported proficiency in this pensate with a coding scheme that com- item at nine or twelve months, although reaching is bined “not yet” and “sometimes” responses, known to emerge between ten and twelve months which both indicate relatively low frequen- (Capone & McGregor, 2004). These results sug- cies of the behavior; however, reliability is- gest that the questionnaire may not accurately sues likely persist. Clinician assessments of capture infant gesturing through parent report. children using the CDI could be used to val-

59 THE YALE REVIEW OF UNDERGRADUATE RESEARCH IN PSYCHOLOGY

the CDI data on gestures in a new way yielded a potential new warning sign for ASD.

BEYOND POINTING CLEMENTS & CHAWARSKA 60

REFERENCES

American Psychiatric Association. (1994). Diagnostic Centers for Disease Control and Prevention. (2009). and statistical manual of mental disorders (revised 4th Prevalence of autism spectrum disorders—Autism ed.). Washington, DC: Author. and Developmental Disabilities Monitoring Net- work, United States, 2006. Morbidity and Mortality Baron-Cohen, S. (1989). Perceptual role-taking and Weekly Report , Dec. 18, 2009. protodeclarative pointing in autism. British Journal of Developmental Psychology , 7, 113-127. Chawarska, K., Paul, R., Klin, A., Hannigen, S., Dichtel, L.E., & Volkmar, F. (2007). Parental Baron-Cohen, S., Cox, A., Baird, G., Swettenham, J., recognition of developmental problems in tod- & Nightingale, N. (1996). Psychological markers dlers with autism spectrum disorders. Journal of in the detection of autism in infancy in a large Autism and Developmental Disorders, 37 , 62-72. population. British Journal of Psychiatry, 168 , 158- 163. Chawarska, K., & Volkmar, F. (2005). Autism in infants and toddlers. In F. Volkmar, Klin, A., & Paul, Baron-Cohen, S., Allen, J., & Gillberg, C. (1992). Can R. (Eds.), Handbook of Autism and Pervasive Develop- autism be detected at 18 months? The needle, the mental Disorders . New York: Wiley. haystack, and the CHAT. British Journal of Psychia- try,161, 839-843. Charman, T., Drew, A., Baird, C., & Baird, G. (2003). Measuring early language development in pre- Bates, E., & Dick, F. (2002). Language, gesture, and school children with autism spectrum disorder us- the developing brain. Developmental Psychobiology, ing the MacArthur Communicative Development 40 , 293-310. Inventory (Infant Form). Journal of Child Language , 30 , 213-236. Capone, N., & McGregor, K. K. (2004). Gesture development: A review for clinical and research prac- Dawson, G., Toth, K., Abbott, R., Osterling, J., Mun- tices. Journal of Speech, Language, and Hearing son, J., Estes, A., & Liaw J. (2004). Early social at- Research, 47, 173-186 tention impairments in autism: Social orienting, joint attention, and attention to distress. Develop- Camaioni, L., Perucchini, P., Muratori, F., Parrinini, mental Psychology, 40 , 271-283. B., & Cesari, A. (2003). The communicative use of pointing in autism: Developmental profile and fac- DiLavore, P. C., Lord, C., & Rutter, M. (1995). The tors related to change. European Psychiatry, 18 , 6-12. pre-linguistic autism diagnostic observation schedule. Journal of Autism and Developmental Disor- Camaioni, L., Perucchini, P., Muratori, F., & Milone, ders , 25, 355-379. A. (1997). A longitudinal examination of the communicative gestures deficit in young children Fenson, L., Dale, P. S., Reznick, J. S., Thal, D., Bates, with autism. Journal of Autism and Developmental Dis- E., Hartung, J. P., Pethick, S., & Reilly, J. S. (1993). orders, 27 , 715-725. The MacArthur Communicative Development Inventories: User’s guide and technical manua l. San Diego, CA: Carpenter, M., Pennington, B. R., & Rogers, S. J. Singular Publishing Group. (2002). Interrelations among social-cognitive skills in young children with autism. Journal of Au- Howlin, P. (2005). Outcomes in autism spectrum tism and Developmental Disorders, 32 , 91-106. disorders. In F. Volkmar, Paul, R., Klin, A., & Co- hen, D., (Eds.), Handbook of Autism and Pervasive De- velopmental Disorders (3rd ed.; vol. A, pp. 201-222). New York: Wiley.

61 THE YALE REVIEW OF UNDERGRADUATE RESEARCH IN PSYCHOLOGY

Itoh, H. (2000). Development of pointing behavior in Luyster, R., Lopez, C., & Lord, C. (2007). Character- young autistic children and children with other de- izing communicative development in children re- velopmental disorders. Japanese Journal of Child and ferred for Autism Spectrum Disorders using the Adolescent Psychiatry, 41 , 57-70. MacArthur-Bates Communicative Development Inventory (CDI). Journal of Child Language, 34 , 623- Kanner, L. (1943). Autistic disturbances of affective 654. contact. Nervous Child, 2 , 217-250. Mitchell, S., Brian, J., Zwaigenbaum, L., Roberts, W., Kita, S. (2003). Pointing: Where language, culture, and Szatmari, P., & Smith, I., & Bryson, S. (2006). Ear- cognition meet . Mahwah, NJ: Lawrence Erlbaum. ly language and communication development of infants later diagnosed with autism spectrum dis- Klin, A., Chawarska, K., Rubin, E., & Volkmar, F. order. Developmental and Behavioral Pediatrics, 27 , (2004). Clinical assessment of young children at S69-S78. risk for autism. In R. DelCarmen-Wiggins, & Carter, A. (Eds.), Handbook of Infant, Toddler, and Pre- Mouridsen, S. E., & Hauschild, K-M. (2009). A longi- school Mental Health Assessment (pp. 311-336). New tudinal study of autism spectrum disorders in in- York: Oxford University Press. dividuals diagnosed with a developmental language disorder as children. Child: Care, Health and Landa, R., Holman, K. C., & Garrett-Mayer, E. Development, 35 , 691-697. (2007). Social and communication development in toddlers with early and later diagnosis of autism Mullen, E. (1995). Mullen Scales of Early Learning (AGS spectrum disorders. Archives of General Psychiatry, ed.). Circle Pines, MN: American Guidance Ser- 64 , 853-864. vice.

Leekam, S. R., & Ramsden, C. A. H. (2006). Dyadic Mundy, P., Sigman, M., & Kasari, C. (1990). A longi- orienting and joint attention in preschool children tudinal study of joint attention and language de- with autism. Journal of Autism and Developmental Dis- velopment in autistic children. Journal of Autism and orders, 36 , 185-197. Developmental Disorders, 20 , 115-128.

Loveland, K., & Landry, S. (1986). Joint attention and Mundy, P., Sigman, M. D., Ungerer, J., & Sherman, language in autism and developmental language T. (1986). Defining the social deficits of autism: delay. Journal of Autism and Developmental Disorders, The contribution of non-verbal communication 16, 335-349. measures. Journal of Child Psychology and Psychiatry, 27 , 657-669. Lord, C., Risi, S., & Pickles, A. (2004). Trajectory of language development in autism spectrum disor- Paul, R. (2007). Communication and its development ders. In M. Rice & Warren, S., (Eds.), Developmental in autism spectrum disorders. In F. Volkmar (Ed.), Language Disorders: From Phenotypes to Etiologies (pp. Autism and Pervasive Developmental Disorders (2nd 7-29). Mahwah, NJ: Erlbaum. ed.). Cambridge: Cambridge University Press.

Lord, C., Risi, S., Lambrecht, L., Cook, E.H., Le- Robins, D. L., Fein, D., Barton, M. L., & Green, J. A. venthal, B.L., & DiLavore, P.C., Pickles, A., & Rut- (2001). The Modified Checklist for Autism in Tod- ter, M. (2000). The Autism Diagnostic Observation dlers: An initial study investigating the early detec- Schedule–Generic: A standard measure of social tion of autism and pervasive developmental disor- and communication deficits associated with the ders. Journal of Autism and Developmental Disorders, 31 , spectrum of autism. Journal of Autism and Develop- 131-144. mental Disorders, 30 , 205-223. Roger, S. J. (2009). What are infant siblings teaching us about autism in infancy? Autism Research, 2 , 125- 137.

BEYOND POINTING CLEMENTS & CHAWARSKA 62

Sparrow, S.S., Balla, D.A., & Cicchetti, D.V. (1984). Wetherby, A. M. (1986). Ontogeny of communicative Vineland Adaptive Behavior Scales: Interview edition, sur- functions in autism. Journal of Autism and Develop- vey form manual . Circle Pines, MN: American Guid- mental Disorders, 16 , 295-316. ance Service. Wetherby, A. M., & Prizant, B. M. (2002). CSBS DP Stone, W. L., Ousley, O. Y, Yoder, P. J., Hogan, K. L., Manual: Communication and Symbolic Behavior Scales & Hepburn, S. L. (1997). Nonverbal communica- Developmental Profile (1st Normed Ed.). Baltimore: tion in two- and three-year-old children in autism. Paul H. Brookes Publishing. Journal of Autism and Developmental Disorders, 27 , 677- 696. Wetherby, A. M., Watt, N., Morgan, L., & Shumway, S. (2007). Social communication profiles of chil- Tager-Flusber, H., Paul, R., & Lord, C. (2005). Lan- dren with autism spectrum disorders late in the guage and communication in autism. In F. R. second year of life. Journal of Autism and Developmen- Volkmar, Paul, R., Klin, A., & Cohen, D.J. (Eds.), tal Disorders, 37 , 960-975. Handbook of Autism and Pervasive Developmental Disor- ders (pp. 335-364). New York: Wiley. Wetherby, A. M., Yonclas, D. G., & Bryan, A. A. (1989). Communicative profiles of preschool chil- Tomasello, M. (2007). If they’re so good at grammar, dren with handicaps: Implications for early identi- then why don’t they talk? Hints from apes’ and fication. Journal of Speech and Hearing Disorders, 54, humans’ use of gestures. Language, Learning, and 148-158. Development, 3 , 133-156. Zwaigenbaum, L., Bryson, S., Lord, C., Rogers, S. J., Toth, K., Dawson, G., Meltzoff, A.N., Greenson, J., & Carter, A., Carver, L., Chawarska, K., Constantino, Fein, D. (2007). Early social, imitation, play, and J., & Dawson, G., Dobkins, K., Fein, D., Iverson, language abilities of young non-autistic siblings of J., Klin, A., Landa, R., Messinger, D., Ozonoff, S., children with autism. Journal of Autism and Develop- Sigman, M., Stone, W., Tager-Flusberg, H., & mental Disorders, 37 , 145-157. Yirmiya, N. (2009). Clinical assessment and man- agement of toddlers with suspected ASD: Insights Volkmar, F., Klin, A., Paul, R., & Cohen, D. (2005). from studies of high-risk infants. Pediatrics , 123 , Handbook of autism and pervasive developmental disor- 1383-1391. ders. New York: Wiley.

63 THE YALE REVIEW OF UNDERGRADUATE RESEARCH IN PSYCHOLOGY