Article review

Therealdeal00

Goin-Kochel26Myers2C2005.pdf

Home >Psychology homework help >Article review

Versus Regressive Onset

of Autism Spectrum Disorders: Parents' Beliefs About Causes

Robin P. Goin-Kochel and Barbara J. Myers

Recent studies have validated the phenomenon of autistic re

gression, but little is known about how regressive and congeni

tal onsets of the disorder influence parents' thinking about

autism and its etiology. Parents (N = 327) of children with

autism spectrum disorders completed an online questionnaire

about their children's development. Approximately half of the

sample indicated that their children showed signs of autism

from birth or shortly thereafter; the remaining half described

their children as developing autism in the regressive fashion.

Parents' beliefs about the causes of autism varied widely, but

two of the most popularly cited explanations were "genetics"

and "external mechanisms" (e.g., vaccinations, environmental

toxins). When analyzed relative to type of autism onset, parents

more often advocated genetics as the cause for autism when

their children exhibited the congenital type and external mech

anisms when their children presented with the regressive type.

M uch work has been accomplished in the identifica

tion of autistic characteristics during the first few

years of life, with many researchers noting differ

ences in the ways that such symptoms emerge in young chil

dren (sec Goin & Myers, 2004a, for a complete review). We

now recognize two paths of autistic emergence: congenital and

regressive, with timing being the primarv distinction between

the two. Generally speaking, congenital refers to the appear

ance of autistic symptoms from birth or shortly thereafter; chil

dren who are described as developing autism in this fashion

typically present with developmental delays and anomalous

behaviors from the outset. Alternately, regressive refers to the

emergence of autistic svmptoms following a period of tvpieal

development; these children tend to be described as develop

ing normally up until a certain age, at which point they begin

to lose previously acquired skills and fail to progress at their

former pace. Type of autism onset is not, at this time, associ

ated with the diagnostic process, as criteria from the Diagnos

tic and Statistical Manual of Mental Disorders ( 4th ed., text

revision; DSM- IV-TR; American Psychiatric Association,

2000) simply requires that behaviors within the autism spec

trum be present before age 3, which occurs in both the con-

genital and regressive trajectories. It appears, however, that

either a later emergence or a later detection of autistic charac

teristics is descriptive of its development in a substantial num

ber of cases (DcGiacomo & Fombonnc, 1998), with one third

(Kurita, 1985) to more than one half (Young, Brewer, & Pat

tison, 2003) of samples claiming a regressive onset.

The regressive form of autism has been mentioned in the

autism literature since the early 1950s, when most investiga

tors focused their cfl<:irts on discovering sociocnvironmcntal

triggers for the regressive onset. In one of the earliest studies

to clearly delineate the two developmental trajectories, Harper

and Williams ( 1975) categorized 13 I children diagnosed with

autism into either a natal group (n = 37), in which onset was

described from birth (i.e., congenital), or an acquired group

(n = 94), in which onset was described as following a period

of normal development (i.e., regressive). Data for the classifi

cations were based on parent questionnaires and interviews,

the latter of which included queries into parents' beliefs about

the precipitating causes of their children's autism. The authors

concluded that the natal and acquired types of autism pre

sented along a continuum, with those in the natal group (who

also showed more evidence of neurological impairment) ex

periencing difficulty interacting in the social world since birth

and those in the acquired group only experiencing such diffi

culties frillowing a physical or psychological trauma. The au

thors, however, also suggested that the children in the acquired

group likely possessed a neurophysiological \'ulnerability to

ward de\'cloping autism under certain environmental concli

tions, which is in line with some modern theories on autism's

etiology. During the past few vears, we have witnessed a surge of in

terest in the regressive emergence of autism. Regressive emer

gence has been difficult to characterize, though, because it

unfolds in different ways among children who de,,clop autism

in this fashion. In their work on maternal perceptions of de

velopmental regression in children with autism, Davidovitch,

Glick, Holtzman, Tirosh, and Safir (2000) intcn·icwcd 39

mothers about their children's development of gross-motor,

fine-motor, social, expressivc~language, receptive-language,

and nonverbal language skills. Out of40 children, 19 (47.5%)

FOCUS ON AUTISM AND OTHER DEVELOPMENTAL DISABILITIES

----- VOLUME 20, NUMBER 3, FALL 2005 ----

PAGES 169-179

Congenital

Congenital Versus Regressive Onset of Autism Spectrum Disorders: Parents' Beliefs About Causes Goin-Kochel, Robin P;Myers, Barbara J Focus on Autism and Other Developmental Disabilities; Fall 2005; 20, 3; PsycINFO pg. 169

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

ON AUTISM AND OTHER DEVELOPMENTAL DISABILITIES 170

were reported to experience regression in all areas but motor skills at an average age of 24 months; of the 19 children who regressed, 11 regressed before 24 months ( considered early re gressors) and 8 regressed after 24 months ( considered late re gressors). In a case study of a young boy diagnosed with autism, Bernabei and Camaioni (2001) analyzed data from the family's home videos, parent interviews, cognitive and lin guistic evaluations, and autism-diagnostic scales to generate a rich picture of the child's development since birth. The child's mother indicated that he appeared to be developing typically until about l 0 months, and videotape analyses confirmed her retrospective reports of a gradual decline across skills. The authors observed a marked degeneration of the child's com municative, social, and functional~play skills around the be ginning of his second year. They further noted that such regression was most prominent between the ages of 30 and 38 months, as the child perfrirmed at a mental age ofl3.5 months on the Uzgiris-Hunt (1975) scnsorimotor scales at age 2 and at a mental age of l 0.2 months a little more than a year later. In another examination of regressive tendencies, Amorosa and Notcrdaeme (2002) reviewed the medical files ofl0l children diagnosed with autism. Three fourths of the children in their sample had been described as "different" from others very early in life; 21 (20.8%) had exhibited developmental regres sion in language, and 14 (13.9%) had a regression in social behavior.

Several investigators have examined the phenomenon of autistic regression through questionnaires designed to elicit parents' retrospective reports about their children's develop ment. For example, Williams and Ozonoff (2001) employed the Early Development Questionnaire to categorize children with autism as experiencing either a congenital onset (n = 29); a clearly regressive onset ( n = 23 ), in which the children showed both social skill and communication losses; or an un clcarly regressive onset ( n = 8 ), in which the children showed developmental skill losses that were qualitatively different from those in the clearly regressive group. The authors learned that about half of the children described as experiencing regression did so following a seemingly normal development, whereas the remaining half had exhibited delays prior to regression that had not been previously detected. Werner and Munson (2001) conducted a similar study using their Early Development In terview (EDI) with parents of children with autism (n = 30), developmental delay ( n = 31 ), and typical development ( n = 39); however, they fi.1rthcr analyzed home \·idcotapes of the children to judge the accuracy of parents' retrospective de scriptions. They found the EDI to successfully discriminate children expressing continuous autistic development from those who manifested a sudden onset. Moreover, observations of families' home movies validated these findings; children said to h,wc dc\'eloped autism in the regressive manner did not sho\\' atypical behaviors during the videos taken when the chil dren were 11 to 13 months old, but they did show atypical be havior in the videos taken when the children were 24 months

old. More recently, Young ct al. (2003) collected question naire data from parents of 81 children diagnosed with autism to ascertain which characteristics initially caused the parents' concern and at what ages. The researchers learned that 48 ( 59 .3%) parents in their sample indicated that their children experienced developmental regression. In a similar study, Coonrod and Stone (2004) queried parents of toddlers with either developmental delays ( n = 22) or autism ( n = 22) about their earliest concerns regarding their children's development. Parents of children with autism reported first concerns when their children were, on average, 17 .8 months old, with a range of 6 to 30 months-a 2-ycar span of time that illustrates great variability in either the emergence or detection of autistic char acteristics.

Reviewing the previously cited studies, it becomes clear that a considerable percentage of children with autism devel oped the condition following a period of seemingly typical de velopment (i.e., in the regressive fashion). The numbers range from 27% (Werner & Munson, 2001) to almost 60% (Young ct al., 2003). Findings such as these imply potential difkrcnces in the course that autism can take and could indicate separate causal mechanisms. To our knowledge, De Long ( 1999) is the only researcher, thus far, to propose explanations for these dis tinct onsets. He hypothesized two forms of autism, one re sulting from bilateral brain damage early in life ( congenital onset) and one not stemming from any form of neurological or biological damage but rather representing the early mani festation of affective psychopathology ( regressive onset). He indicated that this latter, idiopathic form of autism "often has a distinct onset with regression in the second year oflife, even tual higher function and some development of language, spe cial skills or islands of normal function, prominent affrctin· symptoms, and a better prognosis" (p. 912). DcLong's theo ries were originally developed in support of his research and clinical observations regarding a preponderance of affective disorders in the family members of children who developed autism regressively; however, our work friund that children born to parents with affective disorders, as \\'ell as those born to parents with any psychological malady, were more ofren de scribed as exhibiting congenital autism versus the regressive type ( Goin & Myers, 20046 ). Because many pathologies arc now understood to have a strong genetic component, includ ing affective disorders and autism, these findings lend support to the notion that the congenital onset may have strong ge netic underpinnings.

The presence of such distinct developmental trajectories obviously prompts researchers to consider various etiologies for autism, but it also colors the way parents view and cope with the disorder, depending on whether their child experi enced a congenital or regressive onset. Da\·idovitch ct al. (2000) found that mothers whose children developed autism regres sively harbored more guilt about their children's delays as compared to mothers whose children had always exhibited autistic symptoms. Those in the former group, who had wit-

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

20, NUMBER 3, FALL 2005 ---------------------

ncsscd a period of typical development in their children, often

\\'ondcrcd whether they had done something or allowed some

thing to happen that caused their children's autism. Da\·ido

\'itch ct al. also noted that the mothers of children with

rcgrcssi\'c onset more readily described specific events that

thcv deemed responsible for their children's autism; however,

the researchers did not include a list of those e\·ents. In related

work, Gray ( 1995) learned that parents' explanations for

autism in their children were quite varied, the three most pop

ular being a birth-related trauma, congenital damage ( caused

lw any number of factors that resulted in damage to the child's

central ncr\'Crns system), and heredity. Watts, Le Blanc, and

Daggett ( 2003) found a similar \'aricty of responses during

their assessment of mothers' belief; about causes of autism,

highlighting genetics, allergies, and chance as those most fre

quently reported. Similarlv, Lingam ct al. (2003) disco\'crcd

that 44 parents ( 42%) of children \\'ith regressive autism spec

iticd triggers for their children's autism. The most popular ex

planation was a sociocnvironmcntal change, such as the birth

of a sibling, followed by vaccination, \'iral or bacterial infec

tions, seizures, surgery, and other causes.

The literature on parents' beliefs about causes of autism is

scant but not without value. Assessing parents' ideas about the

causes of autism is important for understanding the rationale

behind parents' behaviors, as such beliefs will guide the choices

that they make on behalf of their children (Booth, 1999;

McGillicuddy-Dc Lisi & Sigel, 1995). Gi\'en the range of treat

ment options available to parents of children with ASD, it

seems logical to \'Cnturc that, when deciding how to manage

autism in their children (with some still hoping f()f a cure), par

ents would take into account what they deem to be the cause

of autism (Reimers, vVackcr, Derby, & Cooper, 1995 ). Morc-

0\'cr, as Sonuga- Barke and Balding ( 1993) discovered, par

ents' attributions for childhood psychological disturbances can

be very different from what researchers ( and perhaps other

helping professionals) understand the causes to be. Potentially

discrepant beliefs as to what "caused" and \\'hat will "elimi

nate" autistic behaviors may create challenges for parents and

professionals who work as a team to plan therapeutic and ed

ucation goals for affrctcd children.

To date, most of the \\'ork on autistic regression has fo

cused on \'alidating this phenomenon as a distinct type of on

set. vVc aim to contribute to this literature by (a) pro\'iding

rich, dcscripti\'c data on regression from a large sample of par

ents and (b) extending the understanding of this phenomenon

to the autism spectrum, not just autistic disorder. There arc no

reports, hm\'C\Tr, linking the tnx of onset children experience

to their parents' beliefs about causes for autism spectrum dis

orders (ASD). Therefore, our primary goal was to understand

how parents understood the etiology of ASD in the contexts

of rcgrcssi\'C and congenital onsets. vVc hypothesized that par

ents would more often ad\'CJCatc a genetic etiology when their

children exhibited congenital ASD and an environmental trig

ger when their children exhibited rcgrcssi\'C ASD.

171

Method

Participants

Parents of children with ASD participated as part of a larger

investigation on families' experiences with ASD. The sample

contained 327 participants, including 294 mothers (89.9%),

26 fathers (8.0%), 4 stepmothers (1.2%), and 1 grandmother

(0.3%). Two participants did not indicate their relationship to

the child. The participants' average age was 38.1 years (SD=

7 .1, range = 24 to 72 ), and their mean education level \\'as

15.4 years (SD= 2.5, range = 9 to 26), cqui\'alent to the

third year of college. In terms of race, 293 were White

(89.6%), 10 were Asian (3.1%), 9 were Hispanic/Latino

(2.8%), 5 were Black (1.5%), 2 were Biracial/Mixed (0.6%), 2

were Native American (0.6%), 2 were Middle Eastern (0.6%),

l was Australian/New Zcalandcr (0.3%), 2 reported as Other

(0.6%), and 1 did not indicate a race .. Most of the participants

were married during the time of the study (n = 271, 82.9%).

Regarding families' locations, 283 ( 86.5%) resided in the

United States or Canada, 20 ( 6.1 %) in England or Ireland, and

12 (3.6%) in Australia or New Zealand (the remainder did not

report their locale).

Of the children described by the participants, 267 were

male (81.7%) and 58 were frmale (17.7%). Gender was not

identified for two of the children. The children's a\'cragc cur

rent age was 8.5 years (SD= 4.7, range= 2 to 36; 95.1 % were

under 18 years), and 283 were described as White (86.5%),

21 as Biracial/Mixed (6.4%), 7 as Hispanic/Latino (2.1%),

5 as Black ( 1.5%), 3 as Asian (0.9%), 2 as Australian/New Zea

lander (0.6%), 2 as Middle Eastern (0.6%), and 4 as Other

(1.2%). The majority of the children, 212 (64.8%), had been

diagnosed with autism, 67 (20.5%) with Aspcrgcr syndrome,

47 (14.4%) with Pervasive Developmental Disorder-Not Oth

erwise Specified (PDD-NOS), and 1 with Childhood Disinte

grative Disorder ( CDD ). Regarding the diagnoses, 153 ( 46.8%)

had been diagnosed by a specialist doctor ( e.g., neurologist,

dc\Tlopmental pediatrician), 73 (22.3%) by a psychologist, 40

( 12.2%) by a psychiatrist, 32 (9 .8%) by multidisciplinary teams

of professionals, 5 ( 1.5%) bv a family doctor or primarv-carc

physician, and 19 ( 5 .8%) by another professional, typically

someone affiliated with the school system ( e.g., teacher, oc

cupational therapist; sec Note). Five participants did not spec

if}' the diagnosing professional. More descriptive information

on children by diagnosis, gender, and age is pro\'idcd in Ta

ble I.

Instrument

Data were collected through an onlinc questionnaire con

cerning parents' perceptions of the development of ASD in

their children. The questionnaire covered a variety of topics,

including families' diagnostic experiences, early svmptoms of

children's ASD, and opinions of therapeutic ct1icacy, to name

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

FOCU S ON AUTISM AND OTHER DEVELOPMENTAL DISABILITIES 172

TABLE 1 Ag e, Gender, and Diagnosis of Ch ild ren in Study

ASD Diagnoses

Autism• Asperger syndrome 6 PDD-NOS<

Age

Current

At diagnosis

7.8

3.4

Male

4.9

1.7

8 .6

3.9

Female

4 .6

2.9

10.2

6.9

Male

3.6

3.1

Female

11.5

8.7

4.6

4 .2

7.4

4.3

Male

4.0

3.9

Female

9 .1

6.3

4.4

2.5

Male

12.1

2.8

Note. PDD-NOS = Pervasive Developmental Disorder-Not Otherwise Specified ; CDD = Childhood Disintegrative Disorder. •n = 212; 6 n = 67; <n = 47; dn = 1.

a kw. Our goa l was to collect descriptive data on a range of topi cs from a very large sa mple, thus the nature of qu estio ns asked was largel y guided by the available literature on respec tive topics. For some topics, it made se nse to se ek straightfor ward , quantitative information with closed-ended questions; for others, particularly those that had not been well researched (e. g., parents' beliefs abo ut causes of ASD ), we o pted for open-ended respo nses that wo uld inform us abo ut the phe nom ena of interest. Because of p;irametcrs set forth by our Internet -based, infiirmation- collec tion techn o logy, the final questio nnaire containe d 255 va riables ( the maximum allotted to us at that time), which includ ed demographic information on participants and their children . ft was estim ated that com pleti o n of the questionnaire took bcnvccn 30 and 60 mi n utes , depending on how much information parents wished to share in their open -e nded responses.

During the process of revising and streamlining our ques tionnaire, we also so ug ht the input o f t\vo mo thers of children with AS D to ensure that all the questions were appropriately word ed and easy to understand and that we were not exclud in g viable answe r choices. Alth o ug h the questio nnaire soug ht in formation on a variety of topics, for the current study, we fo cuse d on responses to the following open-ended questions:

1. Do yo u kel that you r child has always had characteristics of autism, or did the y develop after a certain point or age?

2. Do you ever notice your child making devel o pmental im provements and then regressing , apparen tl y "forgettin g" new skills? If so, please describe.

3. What is you r personal theory of w hat causes autism , at least in your own child?

Because the phenomenon of regression h ad not been con sistcntl~, characterized among previo us studies o n this topic, we adop ted a q ualitati ve approach to understand ing regression that allowed parents to tell us how t heir children developed and what regression meant to them. As alluded to previ o usly, we chose thi s same approach to un ders tandin g parents' ca usal attribution s frlr AS D .

A Web site was developed to host the questionnaire. The site includ ed descriptive infrirm ation about the study and in for med consent. T his method ology was ~1dvantageous in that we could adve rtise the stud y inexpensively to a wider audience than could be reached by postal mail. Furtherm ore, the data were ente red and submitted electronically by the participants, thus eliminating the potential for data-entry errors on the part of the research team .

Procedure

The study was adve rtised throu gh organizations that offer sup port and referral services to parents of ch ildren with ASD in the United States and 7 other English-speaki ng countries. More than 220 o rganizatio ns ( e.g. , chapter affi lia tes of the Autism Soc iety of Ame1·ica, the Nati ona l Auti stic Society) were con tacte d, and an ann ounce ment about the study appeared in the newsletters, Web sites, and electronic mailin g li sts of coopera tive grou ps. When potential participants located our vVcb site, they had the opportunity to read throu gh infrlrnuti on ab o ut the study and decide whether to participate. All participants used a computer to complete their questionnaires and sub mitted their responses electro nically. Data from submitted q uestionn aires were stored in a database fo r later ana lysis.

Qualitative Analysis

Responses to the three o pen -e nded questions were qu alita tively analyzed usin g Strauss an d Corbin's ( 1990 ) three-level coding process. Ac cording to this method, the raw data are first reviewed for the emergence of themes. These themes are then classified into distinct categories, which arc organ ized and reorgani zed to most appropriately repre sent the data and ad dress the research questions. Last, cate gories arc finalized and translated into codes . The first author ass umed responsibility for identifyi ng th e themes and then classit)•ing the themes into codes, periodically discussing and sharin g results of this p ro cess with the second author. W hen the final codes were agreed o n by both authors , we read and coded all parti cip ant re-

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

20, NUMBER 3, FALL 2005 ----------------------

sponscs independently; then we jointly reviewed and finalized

all codings of the data; any discrepancies were resolved through discussion. As a reliability check, two parents of children with autism and two teachers at a local school frir preschoolers with

autism read the same randomly selected 20% of the raw data to assess the accuracy of assigned codes. One parent and one teacher each had questions about how to classii)' different seg ments of data, which were resolved afi:cr reviewing the coding instructions with the first author. At no time was there a dis crepancy between the reliability checkers' and researchers' in terpretations of parent responses that warranted recoding of other data.

Results

The first open-ended question was asked to generate partici pants' beliefs about the onset ofASD in their children. In most cases, participants indicated clear beliefs that it was either al ways present (congenital) or it developed following a period of seemingly typical dc\'elopment (regressive). Some partici pants were unsure of when the characteristics of ASD first began, and a few responses were unclear to the point that their opinions on this issue could not be determined. The coding scheme for participants' responses to this question is provided in Table 2.

The second open-ended question inquired about the re gressive nature of ASD in participants' children. Again, most individuals stated plainly that this did or did not occur; how ever, in many cases in which regression was reported, partici pants included descriptions of specific types of regression. In some instances, regression was described as a one-time event that coincided with the onset of autistic characteristics and no subsequent regressive bouts were reported. Other participants stated that their children regressed periodically but only in one area of development, usually language, academics, or toileting. Still others suggested either that their children had experi enced regressive episodes prior to a therapeutic program ( e.g., Applied Behavior Analysis, Picture Exchange System) or that regressive incidences were mitigated through constant rein forcement or maintenance of desired skills. Finally, many par ticipants described regression as a regular event that had global effects on their children, presenting as deterioration across a variety of skills and behaviors ( e.g., language, toileting, eating, compliance). In another way of thinking about regression, some parents responded to this question with a description of newly acquired oddities in their children's behavior as opposed to a loss of skills. These new behaviors were typical of those shown by individuals with ASD, and parents ,·iewed these new symptoms as infantile and problematic. Thus, the phenome non of developmental regression meant somewhat different, albeit related, things to different parents. Interestingly, it was not always viewed from a negative slant, although it undoubt edly represented setbacks for children. Several individuals noted that their children tended to regress just prior to mak-

173

ing a significant developmental imprm·ement; thus for some,

it was a sign that something positive was about to happen, as one parent eloquently stated:

A developmental breakthrough is ofrcn followed lw a period of regression, but I would not describe it as "forgetting new skills." I sec it more as a pendulum, which, having been moved in one direction (impro,·ement), naturallv swings back in the other di rection (regression). The regression, then, is as much a sign of progress as the irnpro,·emcnt. In addition, we han: ofren obscnni periods of difficulty/regression immediately preceding a notice able impron:rnent in our daughter's abilities. The pendulum swings both ways, but it is the movement that is important, not just the direction.

The coding scheme for participants' responses to this second question is pro,,ided in Table 2.

Codes to these two questions were combined to create a new variable indicating (a) whether ASD was \'iewed as ha\'ing a congenital or regressive onset ( al\\'ays or did not always have characteristics) and (b) whether regression, associated \\'ith onset or as a regular phenomenon, was seen as a part of the child's experience with ASD (sec Table 3 ). B,, far, the most fre quent responses were that children had not always presented with autistic characteristics and had experienced developmen tal regression ( n = 155, 47.4% ), that children alwavs had symp toms of ASD and never experienced regression ( n = 73, 22.3%), or that children always had symptoms ofASD and ex perienced regression (n = 72, 22.0%). Descriptions from 27 parents (8.3%) were unclear and could not be classified into one of these three main groups (another 2 descriptions origi nally marked unclear were not used in subsequent analysis).

Parents of children who did not always present with autis tic characteristics but had experienced developmental regres sion (coded as "NAY" in Table 3) clearly described regression as the hallmark of ASD in their children. All children in this group were described as first experiencing typical development followed by one or recurring episodes of skill regression; hence, they became the regression group. Parents of children who always had S\'mptoms ofASD and experienced regression ( coded as "AY'' in Table 3) described their children as always exhibiting atypical, autistic-like characteristics; these parents also described recurring episodes of skill regression, but they did not idcntif,r an episode as the initial demarcation between typical and atypical behavior, as did parents in the first group. Consequently, the two groups of participants whose children "always had ASD-ne,,er regressed" (AN) and "always had ASD-experienced regression" (AY) were combined to form the congenital group (n = 145, 44.3%). All the parents in this group clearly believed that autistic-like characteristics had al ways been present in their children.

Given the group arrangement described above, approxi mately half of the respondents believed ASD was always pres ent in their children and the remaining half believed that ASD developed afrer a certain age or event in the child's life. fam ilies were equally likely to report a congenital or regressive

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

174

Que sti on

Do you feel that your child has always had characteristics of autism, or did they develop after a cer tain point or age?

Do you ever notice your child making developmental im provements and then regressing, apparently "forget ting" new skills?

FOCU S ON AUTISM AND OTHER DEVELO PME NTAL DISABILITIES

TABLE 2 Initial Qualitative Coding Scheme for Questions on Autistic Regression

Code and definition

Always: Believed that autis

tic characteristics were present from birth or very early on in life .

Not always: Believed that

autistic characteristics only or largely developed following a certain age or event .

Not sure: Participant was unsure of when autistic

characteristics began .

Unclear: Coders were un able to determine participant's response.

Yes: The child has regressed or does periodically regress.

No: The child has not exhib

ited regression.

Unclear: Coders were un able to determine partici pant's response.

Quoted exa mpl es

" I think that my son always had autism; in looking back at family movies, pic

tures, etc ., he seemed to have the characteristics of autism from a very

young age." "Yes. Very early, she was extremely hyperactive at age 6 months. I noticed a

clumsiness about the way she moved. She would hyperfocus on certain

things like babies or horses and became very demanding about being

around them."

"No . My son walked, crawled, etc ., all on time. He had good eye contact,

played with toys, etc. He began to develop words: mama, paw paw, bot

tle, bye bye, stop. Between 15 and 24 months, this all changed, beginning

with no more speech development to no speech at all by age 2 [years].

Between 18 and 24 months, we also lost eye contact, appropriate play,

etc ." "I believe in my heart he was not born with autism. My son's development

was 'right on'. I had two older children and was familiar with developmen

tal milestones. Shortly after the age of 1 [year], my son regressed more

and more as the months passed ."

"I'm not sure. He seemed to be normal up to 6 months. I didn't notice any

thing dramatic. It seemed just to be a case of late language development

in his first year." "I can't truly be sure . He was always quirky, but I don't know when I really no

ticed the quirkiness. He was a very calm baby and slept very well. He was

very active in utero."

"Other people used to ask me if she was hard of hearing when she was 18

months old." "I had no idea until the Early Intervention services final report, then I put the

pieces together after visiting Web sites."

"Improvement in behavior and social skills would appear and then regress.

We are constantly teaching and re-teaching social and behavior skills ."

"Yes, especially with language . . . he may just start saying new words and

then it's like he forgets the words or how to say things."

"My son never regressed. He just never spoke ." "We haven't noticed that at all, fortunately."

"Th is one is difficult one to say. Because he thought what he was doing was

normal and proceeded in his life that way. But now realizing, perhaps it

was too late for him, he must struggle with this every day. We are trying to

get him help but he is getting older."

"He has toileting problems and occasionally gets lazy about asking for things .

He points when he is capable oftalking."

onset when their children had ~rntism, were more likely to re

port a congenital onset when their children had Aspcrger syn

drome, and were more likel v to report a re gressive onset when

their children had a PDD-NOS diagnosis, x2(2 , 299) = 8.235, p < .02. Frequencies of type of onset per ASD diagnosis arc presented in Table 4.

Answe rs varied con siderably to th<.: third opcn-uxkd ques

tion co ncerning parents ' beliefs about the causes of ASD in

their children. The majority of respo nses indicated a belief in

a gcn<.:ric cause, external triggers ( e .g., vaccinations , environ mental toxins), or a combination of thl'.sl'. two. In many in

stances, participants illu strated uncertainty in their responses,

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

VO LUME 20, NUMBER 3, FALL 2005

TABLE 3 Congenital vs. Regressive Onset Code s (N = 327)

Code Description n

AN Always had ASD-Never experienced regression 73

AU • Always had ASD-Unclear regarding regression 3

AY Always had ASD-Experienced regression 72

NAN D id not always have ASD-Has not experienced regression 2

NAU • Did not always have ASD-Unclear regarding regression 2

NAY Did not always have ASD-Experienced regression 155

NSU • Not sure about ASD onset-Unclear regarding regress ion

NSY Not sure about ASD onset-Expe rien ce d regressio n 2

UN • Unclear regarding ASD onset-Has not experienced regression 8

uu• Unclear regarding ASD onset-Unclear regarding regression

UY• Unclear regarding ASD onset-Experienced regression 8

Note. Bolded items indicate those data used in subsequent analyses. • Unclear refers to the respondent's answers either being inconsistent or not clearly answering the question.

TABLE 4 Frequencies of Congenital vs. Regressive Onsets per Autism Spectrum Diagnosis

Diagnosis

Autism

Asperger syndrome

PDD-NOS

Congenital

Type of Onset

% n

45 .2 109

63.9 22

38.5 24

Note. PDD-NOS = Pervasive Developmenta l Disord e r-Not Otherwise Specified.

Regressive

54 .8

36.1

61 .5

175

22.3

0.9

22.0

0.6

47 .4

0.3

0.6

2.4

0.3

2.4

x2 (2) p

8.235 .016

often using suc h terms JS "maybe" or "possibly," suggestin g that t hey had ideas about ca uses but that they were no t fore gone conclusio ns. Table 5 prese nts the resul ts of this questi on.

sive ASD more ofrc n believed it was caused by so me cxtcrnJI trigger, x2( 1, 128) = 53 .994, p < .001. Res ults arc presented in Table 6.

A ch i-square an alys is was performed to assess whether

there \verc difterenccs in participants' beliefs about causes of ASD ( external versus genetic ) relative to the type of ASD d e velopm enr ( co nge nital versus re gress ive) witnessed in their children . To ob tain the clearest re sults, only those responses from the newl y created regressive and congen ital groups were included in t hi s ana lysis. In th e same vein, we incorporated on ly those responses from the cau sa l belief categories of purely external and pu rel y ge netic bdids, wh ich were the two most co mm on ly n.:ported by parents and co ll ectively enco mpassed almost 40% of all re spo nses to this query. Pa rticipants who in dicated that their ch ildren exh ibited congenital ASD were sig ni ficantly more likely to belie ve ASD was a gene tic disorder, whereas those who described their ch il dren as having regrcs-

Discussion

The cu rre nt study examined parents' perceptions of auti stic

deve lopment in their children , specifical ly the congenital ver sus regressive onset of ASD and how this was associated with their beliefs about causes for ASD. In :ipproxirnately half of the cases, au tistic emergence was desc ribed as regressive, :rnd in the ot her halt~ co ngenital-a ratio that is in line with type-of onset incidences repo rted in prior work (e .g., vV illi ams & Ozonoft~ 2001; Yo ung ct al. , 2003 ) but that demonstrates thi s within a larger sam ple. Re garding bclids about causes ofASD, parents oftercd a range of exp lanations, which is comm on

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

FOCUS O N AUTI SM AND OTHE R DEVELOP MENTAL DISABILITIES

176

TABLE 5 Qualitative Coding Scheme for Beliefs About Causes of Autism and Frequencies of Response per Category (N = 323)

Final category

Genetic (n = 82, 25.1%)

External (n = 55, 16.8%)

Genetic + External (n = 46, 14.1%)

Biological (n = 25, 7.6%)

Genetic + Biological (n = 12, 3.7%)

External + Biological (n = 7, 2.1%)

Other (n = 12, 3.7%)

Multiple (n = 28, 8.6%)

Don't know (n = 38, 11.6%)

Unclear/Unsure (n = 18, 5.5%)

Description

Implicates genes, hered ity, or family history of autism or mental health disorders.

Implicates vaccines, immuniza tions, shots, medications, environmental toxins, or other external or environ mental triggers.

Implicates a combination of genetic and external factors.

Implicates physiological or neu rological factors, including immune deficiencies, meta bolic issues, chemical imbal ances, physical illnesses or all ergies, medical conditions, and bra in deve lopment.

Implicates a combina tion of genetics and biological factors.

Implicates a combination of external and biological factors.

Implicates factors that include social influences, maternal illness and distress or med ications, prenatal or birth difficulties, newborn medical or early childhood trauma, God, destiny, or chance.

Implicates a combination of fac tors, either several (mo re than 3) or so me comb inatio n of "other" fact ors with those in the previously described groups.

Does not have a personal theory of causation.

Vacillates between causes, or authors were unable to determine code based on response.

Quoted examples

"Genetics. Her paternal grandmother has similar symptoms but has never been

diagnosed. " "I believe that we have a genetic predisposition on both sides of the family. I have

three family members who are bipolar in my immediate family (mother and sib

lings), and my father also su ffers from depression . My husband has a father who

is socially odd but has learned to navigate the world in a mostly rote way."

"He was so normal beforehand [that] I am fairly certain it was caused from his immunizations."

"He was born in Toms River, NJ, right next door to Brick Town, which has a high incident of autism . I think it is the environment where we lived in central Jersey.

We lived in between [two chem ical plants] ."

"Chi ldren are genetically pred isposed and all the autistic symptoms are aggra

vated by mercury in vaccines and other environmental injuries." "Genetic predisposition acted upon by some environmental source . I don't think

vaccines caused J's autism, but I think they played a role . . . . I th in k the genetic

markers had to be there first, otherwise every kid would end up autistic."

"Immune insult. Both of my children with autism have highly elevated natural killer cells, as if their bodies are still fighting an 'infection'."

" He developed croup at 3 months [of age] and had to be hospitalized in [the intensive care unit] on a ventilator for 4 days because his airway swelled shut . I

believe that this illness triggered something, because up until then, he smi led and cooed normally."

" I believe that it is a metabolic issue and that equally important, there is a genetic

component." "I think people are genetically pred isposed, and then something, probably prena

tal, triggers the changes in bra in structure that characterize autism."

"I think that his d igestive system may have been immature and that he may have

had casein allergies which contributed to the heavy metal build up from his vac cines . However, he experienced a major regression immediately following his

third [diphtheria-pertuss is-tetanus (DPT) and measles-mumps-rubella (MM R) vaccinations]."

"My family has a history of immune problem s. I believe the MMR shot overloaded

his immune system and ·caused damage. "

" Tr auma at birth." " I th ink for my child it may be the contrast of having two caregivers with very dif

ferent child rearing manners. Being treated like a king and not having to have to communicate vs. a more strict approach. Perhaps too many videos and not enough constructive interaction while my husband and I are working."

"I believe that my son may have been exposed to something toxic during my pregnancy, as his cous in born the same year also has autistic tendencies and

was born in the same town. The pregnancy was difficult and he was born w ith a trigonocep haly and experienced pressure both in womb and after surgery t o his

skull particu larly in the frontal area." "A cu lmin ation of things. In my child's case I believe it was vaccin e, genetics, and

maybe other factors ."

" Have no clue." " I honestly do not know."

"Lord knows. Maybe vaccinations, maybe gene combo, maybe too much tuna ." " Either born with it or caused by MMR vaccine . I can't decide ."

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

VOLUME 20, NUMBER 3, FALL 2005 ---------------------- 177

TABLE 6 Frequencies of Beliefs About Genetic and External Causes of Autism per Onset

Belief Regarding Causes

Onset

Congenital •

Regre ssiveb

Total

•n - 60. b n - 68.

External

6.7

70.6

among mainstream Western populations (Gray, 1995). vVith few exceptions, all fell into what could be categorized as bio medical or physical causes ( e.g., genetics, illness) as opposed to sociocultural ones ( e.g., God, luck; Danscco, 1997). As pre dicted, parents who reported a congenital onset of ASD in their children tended to attribute the disorder to a genetic cause, whereas those who reported a regressive onset attrib uted ASD to some external mechanism.

This latter finding can be explained through a causc-and effcct schema of ASD in which the alteration of typical devel opment is seen as resulting from some immediate insult to the child's system versus some underlying or dormant mechanism. For parents whose children experienced regression, it makes sense to think that something happened to alter the child's course of otherwise typical development. On the other hand, parents whose children have always expressed atypical behav iors or developmental delays ( congenital group) have little room to implicate some environmental force as markedly af~ fi:cting his or her development; thus, they tend to attribute cause to pre birth factors ( e.g., genetic, prenatal).

It is significant to note the great variability of parents' re ports of their children's development ofASD and the parents' ideas on autism's etiology. Only 40% of parents gave a purely genetic or purely environmental explanation for ASD. Fre quent answers (26'){, of parents) were of the type "this, plus this, or maybe this," suggesting that parents adopted a com plex, multifactorial view of causation, including a view that ac cepted simply not knowing. Perhaps more striking, however, is the fact that, across categories, 136 parents ( 41.6%) felt that an external mechanism had somethinJJ to do with the child's ASD, whether it was deemed the single cause or working in conjunction with other factors. The bulk of families advocat ing this external-trigger cause-either alone or in combination with genetic, biological, or other factors-implicated vaccines or immunizations, but relatively few implicated other envi ronmental toxins ( evidenced in the sample quotations in Ta ble 5 ). This finding relates to the work ofLingam et al. (2003 ), who discovered that parents of children with autism exhibited recall bias in the years fcillowing the research of Wakefield and colleagues, who found evidence ofa link between the measles-

Genetic

93.3

29.4

x 2(1J p

53.994 < .001

mumps-rubella (MMR.) vaccine and autism. After 1997, when the report was officially released, parents in Lingam ct al. 's sample were more likely to describe bouts ofrcgrcssion in their children than they had previously and were more likely to ad vocate the MMR vaccine as causing their children's autism. The Wakefield study has since been refuted, but the debate over the potential causal link between vaccines and ASD is still underway, with several works supporting such a connection ( e.g., Bradstreet, Geier, Kartzinel, Adams, & Geier, 2003; Geier & Geier, 2003a; Geier & Geier, 2003b; Singh & Jensen, 2003; Singh, Lin, Newell, & Nelson, 2002) and several argu ing against the connection ( e.g., DcStcfano & Chen, 2001; Fombonnc & Chakrabarti, 200 I; Madsen ct al., 2002; Parker, Schwartz, Todd, & Pickering, 2004; Taylor ct al., 2002 ). Some evidence, however, supports a genetic propensity toward autism among affected children who specifically experienced language regression (Lainhart ct al., 2002); thus, if environ mental factors have something to do with the regressive onset of ASD, it is likely that they interact with genes to create the condition. for many parents in the current study, though, and under the cause-and-effect schema ofASD previously described, the external-trigger theory carries much weight, especially when a comparable alternative explanation has not yet been offered to take its place.

Limitations

Because the questionnaire was posted on the Internet, it was only accessible to those individuals who had both computer and Internet availability. At the time of the 2000 U.S. census, 67% and 53% of homes with a school-age child ( 6 to 17 years old) had computer and Internet access, respectively. These fig ures vary across race, however, with White and Asian families reporting more access than Black and Hispanic families (Ncw burgcr, 2001 ). Thus, the medium through which the ques tionnaire was presented may have been exclusionary, which could help explain the disproportionate number of White participants; we do not suggest that the racial and economic samples who responded to this questionnaire arc an epidemi ological representation of children with ASD.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

ON AUTISM AND OTHER DEVELOPMENTAL DISABILITIES

178

In support of this mcthodolog\', a recent analysis evaluated

Internet samples compared to traditional samples in published

research and found the two to be comparable with respect to

gender, socioeconomic status, geographic region, and age.

rurthermore, the analysis concluded that Web-based ques

tionnain: results generalize across presentation fixmats, do not

appear to be tainted by false data or repeat responders, and are

consistent \\'ith results from traditional methods ( Gosling,

Vaz ire, Sri\'asta\'a, & John, 2004). The researchers also noted

that this new methodology makes it possible to access select

samples and the hard-to-reach. This advantage is an important

one for reaching families atkctcd by disabilities, especially

when the numbers in a gi\'cn proximit\' arc small, as is some

times the case with autism. Furthermore, our study did not address a related and in

teresting question: 'Where do parents get their inf;xmation

about ASD? It is possible that different sources influence their thinking about causation in different ways, but we have no ev

idence of this. The data, although collected, were not orga

nized for analyses in the present study. In a recent descriptive

report using a larger sample ( N = 498)-including data from parents in the current stud\'-\\'C learned that parents of chil

dren \\'ith ASD access an a\·eragc of se\'en different sources of

information about autism ( range = 0-15 sources), with 88%

citing books, 86% citing \Vcb pages or electronic mailing lists,

72'){, citing other parents of children \\'ith ASD, 69% citing

newsletters of autism-support organizations, and 44% citing

scientific journals (Mackintosh, Myers, & Goin-Kochel, 2005 ).

Because most participants indicated that they accessed sc\'eral

different sources of information, it would have been impossi

ble, gi\'cn the current data, to relate their beliefs about ASD ctiolog\' to a single type of source. Nc\'crthcless, it is possible

that these data arc skewed toward attribution of\'accines as the

impetus for ASD because of the large percentage of families

reporting that they get their inti:xmation about ASD from \Vcb pages or electronic mailing lists, where promulgation of

such information is commonplace. hnallv, we only targeted families who were connected to

autism organizations. It is possible that the perceptions and ex

periences of families not affiliated with such support networks

are diftcrcnt from those represented in this \York. Morcm·cr,

because of our indirect mode of ad\'ertisemcnt and the fact

that we did not collect identif)•ing or contact information on

participants (a measure taken to encourage participation), it

was not possible to follm\· up with participants for \'erification

of some responses. \Ve mitigated this problem through inclu

sion ofrcliabilitv checkers, but it may be helpful in future stud

ies to ha\'e access to participants to resolve any questions

during the coding process. This could ha\'e sol\'cd the prob

lem of the small percentage of unclear, and thus unusable, an

swers.

Future Directions

This project sheds more light on the phenomenon of autistic

regression among children with ASD. \Ve now understand that

parents ha\·c a number of ways of\'iewing regression. Accord

ing to some, regression in their children happens only once, as

the demarcation between typical and atypical de\'elopment;

other parents report, however, that regression happens re

peatedly throughout their children's dc\'elopment. Still other

children with ASD arc described by their parents as never ex

periencing the phenomenon of regression. We also know that

when regression does occur, some parents state that it only af

frets one skill area, whereas others describe it as having severe,

global eftects across skills. Furthermore, some families have

frmnd ways to mitigate their children's regressions. Much

about this trajectory, however, remains unknown. The most

obvious question is why some children seem to always have

characteristics of autism yet others do not present characteris

tics until their second year. Docs this imply diftcrent causal

mechanisms) To our knowledge, De Long's ( 1999) theory is

the only one thus far proposed to explain the discrepant emer

gence of autism, and such theories require testing. Additional

work is also necessary to understand whether there are quali

tati\'e diftcrences in the presentation of autism characteristics,

severity of autism, comorbid diagnoses, and therapeutic out

comes between those children manifrsting congenital and re

gressi\'e onsets of the disorder. Our findings suggest that the type of autism onset that

children experience influences their parents' beliefs about au

tism's etiology. It is fi.irthcr concei\'able that this phenomenon

has a rippling effect, affrcting choices that parents make on be

half of their children ( e.g., therapies, education), other ways

that they think about or cope with the disability ( e.g., hopes

for a cure versus accepting things as they are), and which

autism-related eftorts they advocate. These are just a frw among sc\'cral topics in the arena of autism onset that warrant future

study.

ABOUT THE AUTHORS

Robin P. Goin-Kochel, PhD, is a postdoctoml fcllmP at the Vii~qinia

Institute fiir Psychiatric and RehaJ>ioral Genetics. Her current i11-

terests include autism spectrum disorders and cmnorbid asthma and

intemalizii7cq/extemalizinH svmptoms. Barbara J. Myers, l'hD, is an associate professor in the Departmellt of PsycholwfY at Vii~qinia Com

mmnFcalth Uniwrsi~1' and conducts research mt childreiz at riskfbr de

Pclopmrntal problems and childrcn of incarcerated mothers. Address:

Robin P Goin-Kochel, Vii~qinia Institute fin· Aychiatric and Behavioral

Genetics, 800 East Lc7cqh St., Richmond, VA 23298; e-mail: 1~qoin@Pett

.edu

AUTHORS' NOTES

I. This article was based 011 a doctoral di.1:rertati011 conducted by the

first author.

2. We are lfrtitefitl jiJr the cooperatio11 of the families who contributed

to this project and the assistance of the Arndemic Technologv staff

at Vii;,_qinia Commomvealth Universi~v Jbr Jacilitatin,_q the online

research.

NOTE

Dialf/lOses are all by parcnt report. Validatimt of' dialf1zoses throuHh cx

ami1zati01z of the children 11 1as not possible.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

20, NUMBER 3, FALL 2005

REFERENCES

Americ.111 Psychiatric Association. (2000). DiaJ1110stic aiid statistical 11ta1mal of 111c11tal disorders ( 4th ed., text re,.). \Vashington, DC: Author.

Amorosa, H ., & \!otcrdaeme, M. (2002 ). Earll- childhood autism: Age at onset and earh- regression. 7,citschri(t .fill' Kinder- mtd Jit._qmdpsvc/Jiatric mtd Aychothcrnpie, 30( 3 ), 211-220.

Bernabei, P., & C1111aioni, L. (200 I). Devclopmenul pro tile and re gression in .1 child with autism. Autism, S( 3 ), 287-297.

Booth, ( :. I,. ( 1999 ). Belicts about soci.11 skills among mothers of preschoolers \\'ith special needs. Earlv Education and J)cpc/opmmt, 10( 4 ), 455--474.

Bradstreet, B., c;eicr, D. A., Karuincl, J. J., Adams, J.B., & Geier, M. R. (2003). A. case-control stud\' ofrnercur\' burden in children with autism spectrum disorders. Jo11r11al of' Amcricmt Physicimts a11d S1t1//C01ts, 8( 3), 76-79.

Coonrod, E. E., & Sterne, \V. L. (2004 ). Earh' concerns of parents of children \\'ith autistic .111d nonautistic disorders. Infimts mtd YrmitJl Childrrn, 17(3), 258-268.

D.111scco, E. R. ( 1997). Parental beliefs on childhood disabilit\': Insights on culture, child de,-clopmcnt, and inten-cntion. hztcr natio11al Journal of' Disability l)cpclop111rnt and Edumtirm, 44( I), 41-52.

D.n·id<l\'itch, M., c;lick, L., Holtzman, G., Tirosh, E., & Safir, M. P. (2000). DCYcloprncntal regression in autism: Maternal perception. Journal of'A 11tis111 mtd DcPclopmrntal Disorders, 30(2 ), 113-119.

De Giacomo, A., & 1-'ombonne, E. ( 1998 ). Parental recognition of dc,-cloprncntal abnormalities in autism. E1tropem1 Child-=:-· Ado!cs crnt Psychiatry, 7, 131-136.

DeLong, c;. R. ( 1999). Autism: ;:-,,;ew data suggest a ne\\' l11vorhesis.

Ncurolitq_v, 52( 5 ), 911-917. DcStefano, F, & Chen, R. T. (200 I). Autism and measlcs-mumps

rubella nccination: Contro,·ers,, laid to rest) CNS DrztJTS, JS( 11 ), 831-837.

1-'ombonnc, E., & Chakrabarti, S. (2001 ). No c,·idence for a new variant of measlcs-mumps-rubclb-induced aurism. Pediatrics, 8( 4 ), 991-998.

Geier, D. A., & Geier, M. R. (2003a). An assessment of the impact of thimerosal on childhood ncurodCl'clopmcntal disorders. Pedi atric Rehabilitation, 6(2), 97-102.

Geier, 1'vl. R., & Geier, D. A. (2003b ). \!curode,cloprnental disor ders afrcr thimerosal-containing vaccines: A brief communication.

Expcrimrntal Hiolrtq_v a11d Jv!edici11e, 228( 6 ), 660-664.

Goin, R. P., & M\'crs, B. J. (2004a). Characteristics of infantile au tism: Mm·ing toward earlier detection. Fows mt Autism a11d Ot/Jcr Dcnlopmrntal J)isabilities, 19( 1 ), 5-12.

Coin, R. P., & M\'crs, B. J. (2004b, Mai·). Pare/It perceptions 011 the c01tl{e11ital PS. rc,1rrc.1:i·ipe 01tset of' autism a11d its rc!atio11 to part'lltal aff'cctiPc psvchopatholrtq_v. Poster session presented at the lnterna tio11,1l 1\1ccti11g for Autis111 Rcsc.1rch, S.1cr.1111cnto, (:A.

Gosling, S. D., Vazire, S., Sri,astava, S., & John, 0. (2004 ). Should we trust \Veb-bascd studiesl A comparati,·e analysis of six precon ceptions .1bout Internet questionnaires. '17Je Amerim11 Psvc/Jolwrist, 59(2), 93-104.

c;ra\', D. E. ( 1995 ). I ,ay conceptions of autism: Parents' cxplanator1· models. Medical A1tthropol1tq_1', 16, 99-118.

Harper, J ., & Willi.1ms, S. ( 1975 ). Age and t\'fK of onset .1s critical ,ariablcs in earh· infantile ,rntism. Journal of'Autism a11d C/Jildhood Schizophrrnia, S( 1 ), 25-36.

179

Kurit,1, H. ( 1985 ). Infantile autism \\'ith speech loss before the age

of 30 months. Journal of't/Jc A111crim11 Arndemy of'C/Jild and Ado lcscmt A1'chiatry, 24, 191-196.

Lainhart, J.E., Ozonoff; S., Coon, 1-l., Krasn\', L., Dinh, E., Nice, J., ct al. (2002). Autism, regression, and the bro,1der ,wtism phc

not\'pe. Amcrirn11 Journal of' Medical GC1tetics, 113( 3 ), 231-237. Lingam, R., Simmons, A., And1-c1\'s, 0;., Miller, E., Stm,c, J ., & Ta1'

lor, B. (2003 ). P1-c,alence of autism and parentalh- reported trig gers in a north cast London population. Arc/JiPcs of' Disease i11 Childhood, 88, 666-670.

Mackintosh, V. H., M1-crs, B. J., & (;oin-Kochcl, R. I'. (2005). Sources of i11fiir111atio11 a11d support used by parrnts of' child rm ll'ith autism .1pcctnm1 disorders. Manuscript submitted for publication.

Madsen, K. M., H,·iid, A., Vestergaard, M., Schendel, D., Wohlfahrt,

J., Thorsen, P., ct al. (2002 ). A population-based stud\' of measles, mumps, and rubcli.l ,·.1ccin,1tion and autism. The Nell' E11lfla1td Journal of Medicine, 347( 19), 1477-1482.

McGillicudch'-De I ,isi, A. V., & Sigel, I. E. ( 1995 ). Parent.1! belids. In M. H. Bornstein (Ed.), Ha11dbooli ofpmn1ti1tJJ: Vril. 3. Status a11d social conditions of parc1tti1tlf (pp. 333-358). i\falrn·ah, NJ: Erl b,1urn.

Ne,,·burger, E. C:. (2001 ). Home computers and httcmct use i11 t/Jc United States: Au __ qust 2000. Washington, DC: U.S. Department of Commerce, U.S. Census Bureau.

P.1rker, S. K., Schwartz, B., Todd, J., & Pickering, L K. (2004). Thimerosal-containing ,-.1ccines and autistic spectrum disorder: A critical 1-cview of published original data.l'cdititrics,114(3), 793- 804.

Reimers, T. M., Wacker, D. P., Derlw, K. M., & Cooper, L. J. ( 1995 ). Relation between parental attributions and acceptabilin· of beha,' ioral treatments for their child's beha,·ior problems. Bc/Jal'ioral Disordcn, 20( 3 ), 171-178.

Singh, V. K., & Jensen, R. L. (2003). Eb'ated bels of measles an tibodies in children \\'ith autism. Pediatric Nrnrolrtf!Y, 28( 4 ), 292- 294.

Singh, V. K., Lin, S. X., Nrncll, E., & ;:-,,;eJson, C. (2002 ). Abnormal measles-mumps-rubella .rntibodies and CNS autoimmunit,· in chil dren with autism. Joumal of Bio111cdirnl Scirnce, 9( 4 ), 3:'>9-364.

Sonuga-Barke, E. J., & Balding, J. ( 1993). British parents' belicts about the causes of three forms of childhood ps\'chological distur bance. Joumal ofA/mormal C/Jild Aycholrtfl.1', 21(4), 367-376.

Strauss, A. L, & Corbin, J. ( 1990 ). 1fosics of' qualitatil'c rcscarc/J: Grrnmdcd theory procedures and tcc/;11iq11cs. Thousand Oaks, C:A: S,1ge.

Twlor, B., Iv!iller, E., Lingam, R., Andrc\\'s, ;:-,,;., Simmons, A., & Stowe, J. (2002). Measles, mumps, ,rnd rubella ,·accination and bowel problems or de,·clopmental regression in children with

autism: Population stuck Briti.1J1 Afcdiml Joumal, 324, 393-396. Uzgiris, I. C., & Hunt, J. ( 197:i ). A.,sc . .:,mrnt i11 i11frmcv: Ordinal

scales of' p,ycholwrical dcPelopmrnt. Urb.rna: L1 ni,·ersitv of Illinois Press.

\V,llts, K., Le Bl.me, N., & D,1ggctt, J. ( 200.o, April). / low iiwt/Jcrs rc spmtd to a dia,JT1wsis of 1111tis111 in their c/Jild: Causal bclicj.i-, pcrccp ti011s of' disabilit1', and ps1'cholrtrrical 11'c!l-bei11p. Poster session presented at the biannual meeting of the Societ\' for Research in Child De,-clopmcnt, Tampa, FL.

\Verner, E. B., & Munson, J. A. (2001, April). Ri~m-cssimt in autism: A description and J1alidatio11 of t/1c phrno111c11011 usiitrr parc11t report tmd home l'ideo tapes. Poster session presented at the biannual meet ing of the Societ\' for Research in Child De,-clopmcnt, i\linnc,1po lis, MN.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.