Autism, Autism Spectrum Disorder
Conditions
Keywords
autism, autism spectrum disorder, intensive intervention for toddlers with autism, PDD NOS, early steps study, early start denver model, early intervention for infants and toddlers with autism, toddler treatment, Sally Rogers, University of California, Davis, University of Michigan, MIND Institute
Brief summary
Goals of the current project: (1) Does the Early Start Denver Model experimental intervention for toddlers with autism reduce disability associated with autism significantly more than standard community interventions?; and (2) What environmental, child, and biological characteristics mediate and moderate intervention response and outcomes at age 4?
Detailed description
Thanks to the development of better diagnostic tools and a greater level of professional education, autism is being identified in two year olds and in even younger children, with such early diagnosis justified by the rationale that the earlier intervention begins, the better the outcomes may be. However, there are no published outcome data on intervention models or effectiveness for children who begin intervention by or before 24 months. Furthermore, some teaching procedures considered appropriate for older children, (e.g., 40 hours per week of adult-directed intervention, much repetitive practice while sitting at a table (Lovaas, 2002), 1987) are considered developmentally inappropriate for toddlers (Sandall, McLean, & Smith, 2000). Dawson and Rogers have implemented a feasibility study of a intervention designed for toddlers with autism using a randomized controlled design. The approach involves a relationship-based frame to accomplish developmentally based objectives using naturalistic application of applied behavior analytic principles. The approach fuses the Denver Model (Rogers, Hall, Osaki, Reaven, & Herbison, 2000) and Pivotal Response Training (Koegel, Koegel, & Carter, 1999), and is delivered 1:1 for 25 or more hours per week to 24 toddlers with autism for a two year period. The contrast group receives standard community based intervention. Preliminary results demonstrate large and significant group effects after only 12 months and considerable variability of intervention outcomes in both groups. All families will be referred to the appropriate community service programs, if they have not been referred previously.
Interventions
BEHAVIORALEarly Start Denver Model
Phase 1 of ESDM intervention: 12 weekly, 1 to 1.5 hr. sessions focused on teaching & coaching parents to use the ESDM in all natural caretaking routines & play periods with their child. Parents are taught & coached on 1 aspect of the ESDM each week in the clinic session, & then practice it at home daily in natural family routines & play. Phase 2: each child in the ESDM will receive 25 hrs. a week of ESDM intervention in their homes, 50 wks. a year, for 2 years. 20 hrs. weekly will be delivered by trained interventionists (ITs); 5 hrs. weekly will be delivered by parents. (ITs) will provide ten 2 hour teaching episodes involving play activities per week in the home. Parents will continue to deliver the ESDM in natural family routines & play activities. In addition, each child will receive additional services through public services, or other therapies that the parents may choose, for several more hrs. per week.
BEHAVIORALStandard community care
Treatment and interventions, chosen by families, meeting current standards of community intervention for toddlers with autism and ASD
Sponsors
University of Washington
University of Michigan
National Institutes of Health (NIH)
University of California, Los Angeles
National Institute of Mental Health (NIMH)
University of California, Davis
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
TREATMENT
Masking
SINGLE (Outcomes Assessor)
Eligibility
Sex/Gender
ALL
Age
12 Months to 24 Months
Healthy volunteers
No
Inclusion criteria
(1) 12-24 months of age at the time of telephone screening and living within 60 minutes of the university clinic; (2) ambulatory; (3) unimpaired hand use; (4) parent agreement to have a caregiver present during all home sessions; (5) attendance at all intake sessions; (6) permission to videotape evaluations and ESDM treatment; (7) English as one primary language of the parent; (8) unimpaired hearing and vision; (9) developmental quotient of 35 or higher on the Mullen Scales of Early Learning (MSEL); and (10) meets all of these ASD diagnostic criteria: (a) Autism Diagnostic Observation Scale for Toddlers (ADOS-T) cutoff score of 12 if child produces no words or 10 if child produces some words; (b) agreement by two experienced psychologists that ASD is present; and that child meets DSM-IV criteria for Autistic Disorder or Pervasive Developmental Disorder Not Otherwise Specified (PDDNOS). 5,7,8
Exclusion criteria
(1) serious parental substance abuse; (2) parental self-report of bipolar disorder or psychosis; (3) known genetic syndromes; (4) serious medical conditions (e.g., encephalitis, concussion, seizure disorder); (5) significant sensory impairment; (6) birth weight \<1600 grams and/or gestational age \< 34 weeks; (7) history of intraventricular hemorrhage; (8) known exposure to neurotoxins (including alcohol, drugs); (9) non-English-speaking parents; and for the ESDM-assigned group only, (10) current enrollment in an intensive, 1:1 delivered behavioral intervention of more than 10 hours per week. There was no other restrictions on additional community-based treatment.
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Language Age Equivalent | 24 months | Composite language measure consisting of an average of the Expressive Language and the Receptive Language age equivalent scores from the Mullen Scales of Early Learning.The Mullen Scales of Early Learning (MSEL) is a developmental test with five subscales: gross motor, visual reception, fine motor, receptive language, and expressive language. The lower the score on this scale, the more immature the ability; the higher the score, the more mature the ability. Measurements were taken at baseline, 6 months, 12 months, and 24 months and a hierarchical longitudinal growth curve approach used to calculate overall rate of growth during the 24 month period. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Overall Developmental Quotient (DQ) | 24 months | Overall DQ was calculated by averaging the Time 1 (baseline) age equivalence scores of the two language and two nonverbal subtests from the Mullen Scales of Early Learning (MSEL), dividing by child age in months, and multiplying by 100 to create a quotient score, in order to capture the full range of variability of the sample, since many children fell below the basal standard score. The Mullen Scales of Early Learning (MSEL) is a developmental test with five subscales: gross motor, visual reception, fine motor, receptive language, and expressive language. Gross motor score was not used for the calculations. The lower the score on this scale, the more immature the ability; the higher the score, the more mature the ability. Measurements were taken at baseline, 6 months, 12 months, and 24 months and a hierarchical longitudinal growth curve approach used to calculate overall rate of growth during the 24 month period. |
| Adaptive Behavior Age Equivalent Scores | 24 months | Adaptive behavior age equivalent was characterized by averaging the means of the age equivalents in months of the four domain scores from the Vineland Adaptive Behavior Scales - Second edition (VABS) because the manual does not provide developmental ages corresponding to total scores that could be used to construct quotient scores. Data were provided by parents who were not naïve to group assignment. The Vineland Adaptive Behavior Scales 2 provide a standardized measure of adaptive behavior in four domains: motor, language, social, and activities of daily living. Information is gathered from parents via parent questionnaire. The lower the score on each domain score and the overall score, the more immature the ability; the higher the score, the more mature the ability. Measurements were taken at baseline, 6 months, 12 months, and 24 months and a hierarchical longitudinal growth curve approach used to calculate overall rate of growth during the 24 month period. |
| Autism Severity | 24 months | Autism severity was calculated using the Calibrated Severity Scores, derived by using tables in publications by Gotham, Pickles, & Lord( 2012) (Esler, Bal, Guthrie, Weismer, and Lord, 2015). We identified the severity score listed in the CSS table that was associated with a subject's ADOS-2 total score for the ADOS module that was administered to each subject at each of four time points - at entry into the project, and 6 months, 12 months, and 24 months after enrollment. The ADOS Calibrated Severity Score scale range is 1-10, with lower scores representing milder and less numerous symptoms and higher scores representing more severe and more numerous systems. Scores 1-3 represent few to no ASD symptoms, scores of 4-5 represent mild to moderate symptoms and concerns related to ASD, and scores of 6-10 represent moderate to severe symptom severity. |
Countries
United States
Participant flow
Recruitment details
Children were recruited from pediatricians, developmental disability settings, and university website postings using IRB-approved materials.
Participants by arm
| Arm | Count |
|---|---|
| Early Start Denver Model Phase 1 of ESDM intervention: 12 weekly, 1 to 1.5 hr. sessions focused on teaching & coaching parents to use the ESDM in all natural caretaking routines & play periods with their child. Parents are taught & coached on 1 aspect of the ESDM each week in the clinic session, & then practice it at home daily in natural family routines & play.
Phase 2: each child in the ESDM will receive 25 hrs. a week of ESDM intervention in their homes, 50 wks. a year, for 2 years. 20 hrs. weekly will be delivered by trained interventionists (ITs); 5 hrs. weekly will be delivered by parents. (ITs) will provide ten 2 hour teaching episodes involving play activities per week in the home. Parents will continue to deliver the ESDM in natural family routines & play activities. In addition, each child will receive additional services through public services, or other therapies that the parents may choose, for several more hrs. per week. | 55 |
| Standard Care Available in the Community Standard community care: Treatment and interventions, chosen by families, meeting current standards of community intervention for toddlers with autism and ASD | 63 |
| Total | 118 |
Withdrawals & dropouts
| Period | Reason | FG000 | FG001 |
|---|---|---|---|
| Overall Study | Lost to Follow-up | 3 | 0 |
| Overall Study | Withdrawal by Subject | 7 | 27 |
Baseline characteristics
| Characteristic | Total | Early Start Denver Model | Standard Care Available in the Community |
|---|---|---|---|
| Age, Continuous | 20.64 months STANDARD_DEVIATION 3.29 | 20.58 months STANDARD_DEVIATION 3.37 | 20.70 months STANDARD_DEVIATION 3.21 |
| Ethnicity (NIH/OMB) Hispanic or Latino | 16 Participants | 6 Participants | 10 Participants |
| Ethnicity (NIH/OMB) Not Hispanic or Latino | 94 Participants | 45 Participants | 49 Participants |
| Ethnicity (NIH/OMB) Unknown or Not Reported | 8 Participants | 4 Participants | 4 Participants |
| Race (NIH/OMB) American Indian or Alaska Native | 0 Participants | 0 Participants | 0 Participants |
| Race (NIH/OMB) Asian | 7 Participants | 5 Participants | 2 Participants |
| Race (NIH/OMB) Black or African American | 7 Participants | 2 Participants | 5 Participants |
| Race (NIH/OMB) More than one race | 17 Participants | 8 Participants | 9 Participants |
| Race (NIH/OMB) Native Hawaiian or Other Pacific Islander | 0 Participants | 0 Participants | 0 Participants |
| Race (NIH/OMB) Unknown or Not Reported | 7 Participants | 3 Participants | 4 Participants |
| Race (NIH/OMB) White | 80 Participants | 37 Participants | 43 Participants |
| Region of Enrollment United States | 118 participants | 55 participants | 63 participants |
| Sex: Female, Male Female | 26 Participants | 14 Participants | 12 Participants |
| Sex: Female, Male Male | 92 Participants | 41 Participants | 51 Participants |
Adverse events
| Event type | EG000 affected / at risk | EG001 affected / at risk |
|---|---|---|
| deaths Total, all-cause mortality | 0 / 55 | 0 / 63 |
| other Total, other adverse events | 0 / 55 | 0 / 63 |
| serious Total, serious adverse events | 0 / 55 | 0 / 63 |
Outcome results
Primary
Language Age Equivalent
Composite language measure consisting of an average of the Expressive Language and the Receptive Language age equivalent scores from the Mullen Scales of Early Learning.The Mullen Scales of Early Learning (MSEL) is a developmental test with five subscales: gross motor, visual reception, fine motor, receptive language, and expressive language. The lower the score on this scale, the more immature the ability; the higher the score, the more mature the ability. Measurements were taken at baseline, 6 months, 12 months, and 24 months and a hierarchical longitudinal growth curve approach used to calculate overall rate of growth during the 24 month period.
Time frame: 24 months
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| 1 Early Start Denver Model | Language Age Equivalent | 39.59 language age in months | Standard Deviation 15.31 |
| Standard Care Available in the Community | Language Age Equivalent | 36.09 language age in months | Standard Deviation 14.52 |
Comparison: We used a generalized linear mixed model (GLMM). Change-over-Time was modeled as a linear within-subject effect of time across all observed data for each participant. We fit splines to manage the differing lengths of time in parent-training versus direct treatment. In all analyses, site was included as a categorical covariate to account for potential differences. Effect sizes were reported following Cohen's recommendations as f2.p-value: 0.03Mixed Models Analysis
Secondary
Adaptive Behavior Age Equivalent Scores
Adaptive behavior age equivalent was characterized by averaging the means of the age equivalents in months of the four domain scores from the Vineland Adaptive Behavior Scales - Second edition (VABS) because the manual does not provide developmental ages corresponding to total scores that could be used to construct quotient scores. Data were provided by parents who were not naïve to group assignment. The Vineland Adaptive Behavior Scales 2 provide a standardized measure of adaptive behavior in four domains: motor, language, social, and activities of daily living. Information is gathered from parents via parent questionnaire. The lower the score on each domain score and the overall score, the more immature the ability; the higher the score, the more mature the ability. Measurements were taken at baseline, 6 months, 12 months, and 24 months and a hierarchical longitudinal growth curve approach used to calculate overall rate of growth during the 24 month period.
Time frame: 24 months
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| 1 Early Start Denver Model | Adaptive Behavior Age Equivalent Scores | 39.76 months | Standard Deviation 12.07 |
| Standard Care Available in the Community | Adaptive Behavior Age Equivalent Scores | 36.69 months | Standard Deviation 14.32 |
Secondary
Autism Severity
Autism severity was calculated using the Calibrated Severity Scores, derived by using tables in publications by Gotham, Pickles, & Lord( 2012) (Esler, Bal, Guthrie, Weismer, and Lord, 2015). We identified the severity score listed in the CSS table that was associated with a subject's ADOS-2 total score for the ADOS module that was administered to each subject at each of four time points - at entry into the project, and 6 months, 12 months, and 24 months after enrollment. The ADOS Calibrated Severity Score scale range is 1-10, with lower scores representing milder and less numerous symptoms and higher scores representing more severe and more numerous systems. Scores 1-3 represent few to no ASD symptoms, scores of 4-5 represent mild to moderate symptoms and concerns related to ASD, and scores of 6-10 represent moderate to severe symptom severity.
Time frame: 24 months
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| 1 Early Start Denver Model | Autism Severity | 6.69 units on a scale | Standard Deviation 1.96 |
| Standard Care Available in the Community | Autism Severity | 6.19 units on a scale | Standard Deviation 2.52 |
Secondary
Overall Developmental Quotient (DQ)
Overall DQ was calculated by averaging the Time 1 (baseline) age equivalence scores of the two language and two nonverbal subtests from the Mullen Scales of Early Learning (MSEL), dividing by child age in months, and multiplying by 100 to create a quotient score, in order to capture the full range of variability of the sample, since many children fell below the basal standard score. The Mullen Scales of Early Learning (MSEL) is a developmental test with five subscales: gross motor, visual reception, fine motor, receptive language, and expressive language. Gross motor score was not used for the calculations. The lower the score on this scale, the more immature the ability; the higher the score, the more mature the ability. Measurements were taken at baseline, 6 months, 12 months, and 24 months and a hierarchical longitudinal growth curve approach used to calculate overall rate of growth during the 24 month period.
Time frame: 24 months
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| 1 Early Start Denver Model | Overall Developmental Quotient (DQ) | 83.09 developmental quotient | Standard Error 26.12 |
| Standard Care Available in the Community | Overall Developmental Quotient (DQ) | 79.14 developmental quotient | Standard Error 25.58 |