The University of Toledo
The University of Toledo Digital Repository Master’s and Doctoral Projects
2007
Autism : early diagnosis and intervention Megan Leigh Ruhe The University of Toledo
Follow this and additional works at: http://utdr.utoledo.edu/graduate-projects Recommended Citation Ruhe, Megan Leigh, "Autism : early diagnosis and intervention" (2007). Master’s and Doctoral Projects. Paper 403. http://utdr.utoledo.edu/graduate-projects/403
This Scholarly Project is brought to you for free and open access by The University of Toledo Digital Repository. It has been accepted for inclusion in Master’s and Doctoral Projects by an authorized administrator of The University of Toledo Digital Repository. For more information, please see the repository's About page.
Autism: Early Diagnosis and Intervention
Megan Leigh Ruhe The University of Toledo 2007
ii Dedication To my parents, who always believe in me and continue to guide me towards a life welllived.
iii Acknowledgements My sincere thanks to Professor Karen Graham, whose professional and personal guidance has helped me become a better student and future physician assistant.
iv Table of Contents Introduction ......................................................................................................................................1 Risk Factors Associated with Autism ..............................................................................................4 Neurological Characteristics of Autism ...........................................................................................6 Early Diagnosis of Autism Using CHAT, M-CHAT, and STAT ....................................................8 Gait Disturbances and Motor Movement Abnormalities ...............................................................13 Validity in Diagnosing Autism in Children 2-3 Years ..................................................................15 Outcomes and Prognosis for Early Intervention Therapy..............................................................17 Conclusion .....................................................................................................................................22 References ......................................................................................................................................28 Abstract ..........................................................................................................................................32
1 What is Autism? Matthew, a three-year old boy who entered preschool four months ago, is brought in to your office by his mother who is concerned about his behavior around his peers. The problem was actually brought to her attention by the preschool teacher, who says Matthew often appears to be daydreaming and rarely plays with others even though they try to interact with him. He seems to get “stuck” on particular actions, such as picking up lint from the floor and collecting it, to the point where he cannot pay attention to other tasks or requests from the teacher. Although he answers your questions, Matthew rarely speaks more than two or three words sentences, and does not make good eye contact with you when you speak with him. This scenario represents a classic presentation of an autistic spectrum disorder that any number of health care providers may see. To qualify as an autistic spectrum disorder (ASD), three criteria must be met: (i) qualitative impairments in social communication such as delay or total lack of spoken language without nonverbal compensation, inability to initiate or maintain conversation, repetitive or stereotyped language, and lack of spontaneous make-believe or imitation play, and (ii) qualitative impairments in social interactions shown by impairment of nonverbal behaviors such as eye contact, facial expressions, and gestures, lack of peer relationships, and lack of social or emotional reciprocity. In addition, there are (iii) repetitive and stereotypic patterns of behavior, interests, and activities characterized by abnormal intensity of focus, strict routines and rituals, and repetitive motor movements (e.g. hand or finger flapping) (DSM-IV, 1994). These intricate and complex diagnostic criteria have come through many revisions since the autistic spectrum disorders, also known as pervasive developmental disorders (PDD), were first described by Kanner in 1943 (Filipek et al., 1999). Disorders in this spectrum are widely
2 varied, encompassing Asperger Syndrome, Rett Syndrome, Autistic Disorder, Childhood Disintegrative Disorder, and Atypical Autism (PDD-Not Otherwise Specified). The topic of autism, which encompasses the most severely debilitating degrees of the three criteria, has become a terrifying and devastatingly common possibility for parents. Various studies have shown that the prevalence of autism may be anywhere from 20 to 70 children per 10,000, with as many as 27.1 per 10,000 more diagnosed with other disorders of the spectrum (Filipek et al., 1999; Filipek, Steinberg-Epstein, & Book, 2006). This represents as many as 1 in 143 children having autism. The male gender dominates the diagnosis of autism, ranging from a 3:1 to 4:1 ratio compared with females (Filipek et al., 1999). Because of the extensive volume of research done on the various autistic spectrum disorders, only the diagnosis and early intervention of autism will be the focus of this paper. With those staggering numbers of prevalence in mind, it is obvious that health care providers of any kind will see and treat autistic patients on a regular basis. The importance of being able to recognize behavioral signs through observation and parental input, as well as seriously acknowledging parental concerns about their child cannot be emphasized enough. The average age at autism diagnosis usually occurs after age 3, often not until age 4 or 5 (Charman & Baird, 2002; Filipek et al., 1999). It has been noted in many studies that parents often notice abnormalities in their child by age 18 months, even though diagnosis is usually delayed for two to three years after symptoms are first noticed (Charman & Baird, 2002; Filipek et al., 1999; Landa & Garrett-Mayer, 2006). It is often the first interactions with peers, such as in preschool, which highlight the social communication and interaction deficiencies so classically present in autism. We know little of what is lost during this period of delayed diagnosis, and only an idea of what can be regained through intense treatment afterwards.
3 The significance of 18 months of age has recently become the focal point in establishing whether children at high risk of having autism can be correctly identified through screening and observation before that age. The purpose of this early recognition lies within the overall outcome of autism, as it is a lifelong, debilitating disease with no cure. Age cohort studies show that there is no decline in the number of those classified as having autism, indicating that it is a constant disability with little improvement (Newschaffer, Falb, & Gurney, 2005). Only 12% of autistic adults can be described as functioning at a high level of independence, while the majority, 57%, qualify as having a “Poor” or “Very Poor” outcome in their ability to be selfsupporting (Howlin, Goode, Hutton, & Rutter, 2004). The severity of this disease and the impact it has on future functioning within society presents an urgent need for effective early intervention. The research that has been done regarding early intervention, albeit on autistic children who are older than two and a half years old, has shown incredible improvements in IQ, language skills, developmental rates, and ability to function as a “normal” child in public school (Rogers, 1998). The possibilities for advanced treatment options, ranging from refinements in specific behavioral therapies to starting therapy at an earlier age of diagnosis, are numerous.
4 Risk Factors Associated with Autism A definitive cause of autism has yet to be identified; however, it is the responsibility of any health care provider to be aware of risk factors that may increase the possibility of a positive diagnosis. Although clinicians may hesitate in being overly cautious in diagnosing autism due to parental distress, it has been shown that the positive benefits of this increased awareness outweigh any negatives (Filipek et al., 1999). Any parent who hears the diagnosis of autism will most likely be greatly concerned with what caused this disorder in their child. Even though researchers can point to possible origins of autism, this enigmatic disease still defies complete understanding. In the past, it was thought that autism resulted from the psychogenic influence of a cold, indifferent mother upon her infant, but this has been solidly refuted since then. Another recent controversial cause implicated the measles, mumps, and rubella vaccine, although epidemiological evidence indicates that this also has no role as a risk factor for autism (Baird, Cass, & Slonims, 2003; Szatmari, 2003). What is known is that although autism is a behaviorally defined disorder, it is recognized to have many biological origins. These include complications such as the use of anticonvulsants during pregnancy, untreated phenylketonuria, postnatal infections such as encephalitis, and prenatal infections such as rubella (Baird et al., 2003). Identical twins with a concordance rate of autism at 82% have shown that genetics play an important role in the development of autism (Trottier, Srivastava, & Walker, 1999). The general population prevalence of autism ranges from 0.04% to 0.1%, but the risk for a sibling increases to 2-6%, about 50 times higher than population rates (Baird et al., 2003; Trottier et al., 1999). In particular, chromosomes 2q, 7q, 15, 16p, and 19p have been identified as having gene components involved in autism (Baird et al., 2003). The fact that so many genes are suspected to
5 be involved, along with the uncertainty of what other cytogenic gene activations are still yet to be discovered, make each diagnosis of autism uniquely different and especially complicated to concretely define.
6 Neurological Characteristics of Autism The various aspects of autism indicate several levels of involvement within the brain, particularly the cerebellum, temporal lobe, and frontal lobe. The cerebellum has been shown to participate in affect, social interaction, learning, motivation, and in processing and modulating sensory and motor information. Imaging studies have consistently shown that this area of the brain is hypoplastic in autistics, but at the same time has an increased volume (Nayate et al., 2005; Trottier et al., 1999). Purkinje cells, which communicate through the cerebellar nuclei, are fewer in number in autistic brains, indicating that some dysfunction in brain communication is directly related to mechanical deficits. Acquired cerebellar lesions in rats demonstrated autisticlike symptoms such as repetitive motions and inappropriate social communication (Bobee, et al., 2000). Children with acquired lesions in the cerebellum and autistic children were similarly impaired in activities of rapid attention shifting. This implies that attention dysfunction related to the cerebellum limits the ability to take in and express more complex communication styles (Nayate et al., 2005). The cerebellum is also related to coordinated activities such as gait and motor movements, two emerging areas of study in the greater understanding of autism. These will be discussed in a later section. One specific marker of autism is an increase in head circumference and cortical lobe enlargement, common in approximately 25% of autistics (Trottier et al., 1999). This macrocephaly is related to abnormal brain growth during the post-natal period. One unique factor unique associated with macrocephaly is hindrance in the use of single words, an important step towards the development of language (Lainhart, et al., 2006). A more extensive study of this increase in head circumference found that when compared to a control sample of non-autistic subjects with a 3.0% occurrence of macrocephaly, children with autism had a 17.5% rate of
7 macrocephaly (Lainhart et al., 2006). The implications of this finding are significant, but the usefulness of identifying macrocephaly in a normal clinical setting has not been demonstrated. Macrocephaly in autistics does however support the idea of abnormal neuropathological mechanisms causing the causing the classic developmental delays noted in autism. MRI studies have reliably found greater amounts of white matter in autistic brains, as well as structural changes within the limbic system. Functional MRI studies during cognitive activities also show that there are unusual areas of activation in and around the amygdala as a reaction to social stimuli (Szatmari, 2003). These abnormalities, along with dysfunctional communication between cortical and subcortical coordination of attention, may all possibly contribute to the impairments in social communication and interaction found in autism (Trottier et al., 1999). Rogers et al. (2003) also found that autistic children have a highly increased sensitivity to sensory experiences, especially tactile, visual, and auditory stimulation. Further exploration into the association between elevated sensory stimulation and social communication and mental retardation impairments in autistic children showed there was no association between the severity of symptoms. This indicates that although specific areas of the brain, such as the limbic system, have unusual activity in autism, there is no single association profile that can explain the severity of the varying characteristics found in the disorder.
8 Early Diagnosis of Autism Using CHAT, M-CHAT, and STAT Most parents of children with autism notice abnormalities in their child’s development by the age of 18 months. Common characteristics noticed by parents of children with autism are listed in Table I. Shortening the time between onset of symptoms and initiation of intervention is associated with greater developmental gains (Rogers, 1998). Therefore, the use of a standardized screening tool with high specificity and sensitivity is crucial. Most screening tools have disadvantages that make them inappropriate for standardized use in the general public. Shortcomings include being designed for children of an older age, no validity when used with young children, taking too long to administer, no use of parental input, or screening for developmental delays that are not present before age 3 (Robins, Fein, Barton, & Green, 2001). Arising from these negatives is the Checklist for Autism in Toddlers (CHAT), develop by Baron-Cohen in 1992 for assessing children at age 18 months. The CHAT uses a brief nine question form filled out by parents in combination with a short five question section of observations completed by the health care provider. This checklist measures developmental milestones in three main components that are indicative of autism: protodeclarative pointing (pointing at an object of interest), gaze monitoring, and pretend play. The theory behind the CHAT is valid, but several studies have shown that although its specificity is above 97%, its sensitivity for positively identifying autism is only around 35% (Baird et al., 2003; Robins et al., 2001; Scambler, Rogers, & Wehner, 2001). CHAT is also less sensitive to milder forms of the ASD, such as Asperger, because of less prominent developmental delays. This makes it inappropriate to use as a general screening for all children. However, children who are identified by CHAT as being high risk for autism or an ASD at initial screening are shown to have an extremely stable diagnosis again at follow up at 42 months (Charman & Baird, 2002).
9 Using CHAT as a base, the Modified-CHAT was developed as an extension using the parental portion of CHAT along with an addition 14 questions (Robins et al., 2001). This screening test uses parent reports as the sole source of information and is given at 24 months of age instead of the 18 month marker used by CHAT. This is significant because approximately 15-30% of autistic children will have a period of severe regression of speech skills and social interaction between the ages of 15-24 months (Charman & Baird, 2002). The logic behind this suggests that M-CHAT will be more sensitive than CHAT because it tests children after this regression would be expected to occur (Robins et al., 2001). The sensitivity of M-CHAT is shown to be 87-97%, and the specificity is 95-99%, indicating that it is a more efficient screening tool for autism (Robins, Fein, Barton, & Green, 2001). Because it has only recently been developed, further conclusive studies are needed. The results of M-CHAT should be carefully interpreted, as it has mainly be researched in clinical samples or population-based samples. It also raises the question of whether the 6-month time period between screening times of the two tests has any negative impact upon the delay in initiating early intervention therapy. It is important to remember that using the CHAT and M-CHAT can only be the first step in identifying high-risk children. If suspicion is aroused by these screenings or any behavioral signs, referral to a specialist in developmental disorders is recommended. At this point a more intense and thorough set of diagnostic tests can be performed. Robins et al.(2001) warn that although parental reporting as used in M-CHAT can be accurate, clinicians are still responsible for observation of any delays, as parents may have poor judgment about their child’s behavior. Filipek et al.(1999) found that referral did not occur until an average of 40 months of age, well after the presentation of developmental signs. There are other, more physical observations such
10 as movement and motor skills that can be gathered by a clinician that can be more thoroughly discussed. Because autism is primarily a behavioral diagnosis, the importance of emphasizing behavioral abnormalities is crucial in its identification. Both CHAT and M-CHAT mainly use written checklists to solicit specific identifiers of autism. These are considered to be Level 1 screenings in that they are used for the general population to identify at-risk toddlers. A large component of each is parental reported activity, which allows for the possibility of reporter bias and also failure to recognize abnormal behaviors. Inherently, no screening test for autism can completely avoid drawbacks and inconsistencies. To take the diagnosis of autism to a more definite level in children younger than 36 months, Stone et al. (2004) recently developed a Level 2 screening test, the Screening Tool for Autism in Two-Year-Olds (STAT). This test is to be used in a more select population of children in early intervention programs or evaluation clinics who have already been identified with developmental delays and global impairments and are at greater risk for autism. STAT is the only Level 2 screening test to include interactive measures. Twelve activities exploring early social-communication skills are directly observed by the administrator. This is greater in comparison even to the M-CHAT’s five observed interactive behaviors. The specific interactive behaviors focus on the child’s cooperative play, requesting, directing attention, and motor imitation. Results from a study of 41 children showed that STAT has a sensitivity of 0.92 for identifying autistic children and a specificity of 0.85 (Stone et al., 2004). The age range of the subjects was 24-35 months, providing an advantage of STAT to be used in children over a 12-month age range. This is compared to a specific age of 18 months and 24 months required to administer the CHAT and M-CHAT, respectively. STAT was also shown to
11 be consistent in interobserver agreement, making its administration more reliable and less subject to operator failure (Stone et al., 2004). However, it is important to remember that STAT is designed to specifically identify autism out of a wide range of already present developmental delays, not out of the general population. (Stone et al. 2004) The drawback to all screening tests, no matter what level, is that they must be given by a trained professional and are time-consuming to administer. This problem of time is especially relevant in the Level 1 screenings for a general population. It seems to be more prudent to administer a screening test to a child suspected to display autistic characteristics, whether they are reported by parent, teacher, or clinician, rather than screen every single child who enters a medical office. The importance of being able to recognize behavioral delays cannot be emphasized enough; however, resources to educate any and all persons involved in a child’s care are limited. A broader scope of characteristics outside of behavior are gaining greater importance in the role they play in identifying autism. These include early motor movements and gait abnormalities, and may be recognized by parents by the age of 12 months, leading to an ability to explore a suspected diagnosis of autism further with the specific screening tests. In October of 2007, the American Academy of Pediatrics (AAP) announced its official recommendations for screening all children by primary care providers at 18 and 24 months with a standardized autism screening tool (Johnson et al., 2007). Providing an algorithm for providers, the AAP emphasizes the importance of a provider questioning about a child’s behavioral and social development, as well as being aware of the characteristics of autism and performing surveillance at every appointment. The AAP also acknowledges that parents are usually the first to notice abnormalities in their child's development and recommends screening for autism when there is parental concern. If the initial screening is not positive for autism or a
12 spectrum disorder but a child has risk factors of ASD such as a diagnosed sibling or parental concerns, a follow-up exam in one month is still required of the provider. Parents should also be provided with information on autism spectrum disorders so as to be more aware and discerning of traits they may appreciate in their child. If a child under three years does test positive for an ASD, a referral to an early intervention or early childhood education service should be made immediately for a more indepth analysis and final diagnosis (Johnson et al., 2007). Early intervention services are government funded programs designed to provide education services to children with special needs or developmental disabilities until their third birthday. Many communities also have various providers who are capable of diagnosing a child older than three with autism, including child neurologists, child psychologists, and developmental pediatricians. Testing should also be performed for comorbidities of ASD, which include epilepsy, tuberous sclerosis, sleep disorders, and psychiatric disorders such as schizophrenia. The AAP specifically advises against the “wait and see” model of care, because this can not only delay diagnosis, but cause a breakdown in the relationship and trust between a provider and patient (Johnson et al., 2007). These new recommendations are a positive change in the early identification of autism and autism spectrum disorders as a mandatory part of the assessment of every child's growth and development.
13 Gait Disturbances and Motor Movement Abnormalities Because of neuroimaging studies, it is known that autistic brains have an increased size and greater neuronal dysfunction, particularly in the cerebellum, than the normal brain (Nayate et al., 2005). Studying the motor movements of autistic children has become an area of increased focus, in part due to the early parental concerns by age 12 months that their child is not developing normally. Using retrospective video studies and normal movement milestones, Teitelbaum et al. (1998), methodically analyzed the infantile movements of 17 children later diagnosed with autism in comparison with 15 non-autistic children. A normal child is expected to be able to turn from back to stomach by three months of age using a “corkscrew” motion, first leading with their pelvis and then by six months leading with their head. In autistic infants, it was identified in three of the videos studied that the infants could not right themselves at all, or used an arched-back technique that moved the body in one gross movement instead of the classic corkscrew motion (Teitelbaum et al., 1998). This type of difficulty in righting was also gathered by Teitelbaum et al. (1998) in non-video studies, observing that some autistic infants could not right themselves at all, or used completely different characteristics of righting that were never used by non-autistic children. The ability to sit evenly and balanced is apparent in normal infants around six months of age; however, some of the autistic children in these retrospective videos were unable to maintain stability in a symmetrical manner or use inherent reflexes to remain upright. Often they were shown to fall when reaching for a toy instead of balancing upon the sitting bones as an infant normally would. The autistic infants were sometimes able to sit without movement for a few minutes, but eventually the instability of weight distribution and posture not seen in non-autistic infants caused them to fall over. Crawling, which also requires a balance between all four limbs,
14 is impaired in autistic infants who showed impairment in bilateral movements, balance issues, and the correct pattern of hand to knee coordination needed to crawl (Teitelbaum et al., 1998). The last aspect of movement studied by Teitelbaum et al. (1998), is the crucial coordination needed for walking, which requires symmetrical movements and appropriate weight shifting for forward motion to occur. In some of the autistic children in the video studies, the infantile pattern of walking moving only the thigh and allowing the lower leg to passively follow remained present even up to age three (Teitelbaum 1998). Inappropriate delaying shifting of the weight, resulting in a “goose-walk,” was also noted in a five-year-old (Teitelbaum 1998). Because of the cerebellar involvement, particularly in the vermis, gait analysis in young and old autistics has also emerged in research. It is thought that the cerebellum is responsible for coordinating the upper and lower brain centers, balance, orienting movements, and positioning to produce locomotion. The basal ganglia is also involved in changing posture to facilitate weight shifting in movement (Nayate 2005). A study of autistic adult movements showed signs of cerebellar dysfunction, including unstable gait, mild clumsiness, irregularly timed stepping pattern, and ataxia (Hallett et al., 1993). Vilensky et al. (1981), also found that autistic children used decreased stride length and an almost parkinsonian gait, indicating that underlying striatal dysfunction played a role in autistic movements. In the previous study by Teitelbaum et al. (1998), nearly all the parents felt that something was not right with their child, but were often placated by the pediatrician’s assurance that it was nothing to be concerned over. This is a reminder of the continually confounding problem of autism; because every child develops with variation, and because autism is different in every single child, it becomes extraordinarily difficult to separate the affected from the normally developing.
15 Validity in Diagnosing Autism in Children 2-3 Years The impetus for clinicians being able to diagnose autism as early as possible lies with the potential for greater gains made with early intervention. Despite several questionnaires and checklists having been formulated to identify the characteristics of autism in children as young as 18 months, the question of how secure practitioners can be in their assessments remains. Crais et. al (2006) ask “How early can we go?” when diagnosing autism. With the risk of misdiagnosis looming in the back of every clinician’s mind, as well as the prospect of greater gains with the earliest possible correct diagnosis and intervention, autism must be treated with a delicate balance. Several studies using various diagnostic tests have shown that autism can for the most part be accurately diagnosed in children as young as 18 months. As reported earlier, the CHAT screening test diagnosed all but one child accurately with autism between the ages of 18 and 42 months (Charman & Baird, 2002). Gillberg et al. (1990) correctly diagnosed all but one child under the age of three with autism. Using the Autism Diagnostic Interview-Revised, Lord (1995) found that out of thirty children tested, only one was diagnosed with autism at age two and not at age three. In this same study, only two children were found to have autism at age three and not at age two. A study by Stone et al. (1999) used basic diagnostic criteria to determine whether children under age 3 could be accurately diagnosed by different raters with an autistic spectrum disorder, and more specifically with autism. It was found that when diagnosing ASD versus non-spectrum disorders, there was an agreement rate of 80% between inter-raters. However, agreement between inter-raters on a diagnosis of autism versus other autism spectrum disorders was only 42% (Stone et al., 1999). The initial diagnoses for the majority of children did remain stable after one year, though (Stone et al., 1999). This emphasizes again that although a
16 diagnosis of an ASD is accurate in children under age three, the difficultly in obtaining a stable, specific diagnosis of a spectrum disorder is more likely to be subject to a variable observer or clinician.
17 Outcomes and Prognosis for Early Intervention Therapy When a child is diagnosed with autism, a multitude of challenges are present concerning their care and the difficulties involved in coordinating school officials, health care providers, therapists, and the family in providing the most effective environment for improving overall outcome. While intense behavioral therapy is financially costly, emotional, and time-consuming, the prognosis of children receiving intervention is much more positive than those not included. (Rogers, 1998) It has been clearly shown that the majority of autistics do not function independently in society, relying on their parents, institutions, hospitals, and sheltered communities to provide them with care and support (Howlin et al., 2004). One indication of how well future functioning will be is the childhood performance and verbal IQ. Children with a verbal IQ >70 or a performance IQ >70 are more likely to have better outcomes in terms of independence, employment, and advanced schooling. In one study, all but one out of 29 children with verbal IQ