Cited by 333Autism Spectrum Disorder ResearchChild Nutrition and Feeding IssuesLanguage Development and Disorders

The DOI matches its Crossref record; 44 of the 46 references it cites were found in a registry or an index; none of the 38 sources its claims cite is checked for support yet.

Evaluation of the Diagnostic Stability of the Early Autism Spectrum Disorder Phenotype in the General Population Starting at 12 Months

Open access

Abstract

Importance: Universal early screening for autism spectrum disorder (ASD) in primary care is becoming increasingly common and is believed to be a pivotal step toward early treatment. However, the diagnostic stability of ASD in large cohorts from the general population, particularly in those younger than 18 months, is unknown. Changes in the phenotypic expression of ASD across early development compared with toddlers with other delays are also unknown. Objectives: To examine the diagnostic stability of ASD in a large cohort of toddlers starting at 12 months of age and to compare this stability with that of toddlers with other disorders, such as developmental delay. Design, Setting, and Participants: In this prospective cohort study performed from January 1, 2006, to December 31, 2018, a total of 2241 toddlers were referred from the general population through a universal screening program in primary care or community referral. Eligible toddlers received their first diagnostic evaluation between 12 and 36 months of age and had at least 1 subsequent evaluation. Exposures: Diagnosis was denoted after each evaluation visit as ASD, ASD features, language delay, developmental delay, other developmental issue, typical sibling of an ASD proband, or typical development. Main Outcomes and Measures: Diagnostic stability coefficients were calculated within 2-month age bands, and logistic regression models were used to explore the associations of sex, age, diagnosis at first visit, and interval between first and last diagnosis with stability. Toddlers with a non-ASD diagnosis at their first visit diagnosed with ASD at their last were designated as having late-identified ASD. Results: Among the 1269 toddlers included in the study (918 [72.3%] male; median age at first evaluation, 17.6 months [interquartile range, 14.0-24.4 months]; median age at final evaluation, 36.2 months [interquartile range, 33.4-40.9 months]), the overall diagnostic stability for ASD was 0.84 (95% CI, 0.80-0.87), which was higher than any other diagnostic group. Only 7 toddlers (1.8%) initially considered to have ASD transitioned into a final diagnosis of typical development. Diagnostic stability of ASD within the youngest age band (12-13 months) was lowest at 0.50 (95% CI, 0.32-0.69) but increased to 0.79 by 14 months and 0.83 by 16 months (age bands of 12 vs 14 and 16 months; odds ratio, 4.25; 95% CI, 1.59-11.74). A total of 105 toddlers (23.8%) were not designated as having ASD at their first visit but were identified at a later visit. Conclusions and Relevance: The findings suggest that an ASD diagnosis becomes stable starting at 14 months of age and overall is more stable than other diagnostic categories, including language or developmental delay. After a toddler is identified as having ASD, there may be a low chance that he or she will test within typical levels at 3 years of age. This finding opens the opportunity to test the impact of very early-age treatment of ASD.

Overview

AI snapshot

Written from the paper’s full text.

TL;DR

In a general-population cohort of 1269 toddlers evaluated longitudinally from 12 to 36 months, an ASD diagnosis was more stable than other diagnoses, with overall stability of 0.84. Stability was weakest at 12–13 months (0.50) and improved by 14 months (0.79) and 16 months (0.83), while late-identified ASD occurred in 23.8% of toddlers ultimately diagnosed with ASD.

Key findings

  • Overall diagnostic stability was 0.84 (95% CI, 0.80–0.87) for ASD and 0.79 (95% CI, 0.74–0.83) for typical development (TD); ASD vs TD stability did not differ significantly (OR, 0.86; 95% CI, 0.57–1.29). ASD was significantly more stable than ASD features (OR, 0.11; 95% CI, 0.03–0.32), developmental delay (DD; OR, 0.15; 95% CI, 0.09–0.25), language delay (LD; OR, 0.04; 95% CI, 0.03–0.06), and other diagnoses (OR, 0.16; 95% CI, 0.09–0.28).
  • ASD stability was weakest at 12–13 months (0.50; 95% CI, 0.32–0.69), increasing to 0.79 by 14 months and 0.83 by 16 months (12 vs 14/16 months OR, 4.25; 95% CI, 1.59–11.74). When toddlers with ASD features were counted as ASD, stability rose to 0.70 at 12 months, 0.85 at 14 months, and 0.94 at 16 months.
  • Of 400 toddlers initially designated as having ASD, 336 retained the diagnosis at final visit; only 7 (1.8%) transitioned to TD, and 5 of those 7 were initially evaluated at 12–13 months. The most common transition was from an initial diagnosis of LD or DD to ASD.
  • Among toddlers ultimately diagnosed with ASD, 66 (15.9%) were early-age persistent, 270 (61.2%) middle-age persistent, and 105 (23.8%) late-identified. Late-identified toddlers had better first-visit scores than the other ASD groups (Cohen d range, 0.44–2.35) but worse scores than TD toddlers (Cohen d range, 1.43–1.84); 39 late-identified toddlers (37%) fell within the ADOS toddler-module concern cutoff but were designated non-ASD by practitioner judgment.

Methods

  • A total of 2241 toddlers aged 12–36 months were referred to a University of California, San Diego autism expert evaluation center through the Get SET Early program or community referral; 1269 were longitudinally evaluated 2 or more times with an interval of 6 months or longer and were the focus. Approximately 75% came from Get SET Early and 25% from community referral. Median age at first evaluation was 17.6 months, median age at final evaluation was 36.2 months, and mean number of diagnostic visits was 2.7.
  • Licensed psychologists performed diagnostic and psychometric testing, including ADOS-2, Mullen Scales of Early Learning, and Vineland Adaptive Behavior Scales. Toddlers were designated as having ASD, ASD features, developmental delay, language delay, other issue, TD, or typical sibling of an ASD proband.
  • Stability coefficients were calculated as the proportion retaining the same diagnosis from first to last visit within 2-month age bands. Logistic regression included sex, age at first diagnosis, interval between first and last diagnosis, and diagnostic group at first visit; age was binned as younger than 14 months, 14–17.99 months, 18–23.99 months, and 24 months or older. Analyses were performed in R.

Limitations

  • The paper states that practitioners making final diagnoses were not masked to previous diagnoses because DSM-5 procedures and criteria require consideration of historical information regarding ASD symptoms; this could have biased psychologists toward increased diagnostic stability, although the relatively weak stability in other delay groups was cited as counterevidence.
  • The paper notes that lower ASD stability at 12 months likely reflects limitations in the diagnostic tools used at that age, including the ADOS-2 and DSM.
  • The study was not designed to determine whether early treatments affected the speed with which toddlers with language delay caught up by final diagnosis age.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1001/jamapediatrics.2019.0624

Cites

44 of 44 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1. Sample Characteristics Distribution of key features associated with the study cohort, including the age (in months) that toddlers received their first comprehensive diagnostic evaluation (A), the number of toddlers who received 2, 3, or 4 or more diagnostic evaluations (B), the age (in months) that toddlers received their last diagnostic evaluation, and the interval (in months) between a toddler’s first and last diagnostic evaluation (D).
Figure 1. Sample Characteristics Distribution of key features associated with the study cohort, including the age (in months) that toddlers received their first comprehensive diagnostic evaluation (A), the number of toddlers who received 2, 3, or 4 or more diagnostic evaluations (B), the age (in months) that toddlers received their last diagnostic evaluation, and the interval (in months) between a toddler’s first and last diagnostic evaluation (D).CC BY · PubMed Central
Figure 2. Transition Table and Diagnostic Heat Maps A, Summary of the proportion of toddlers from within the entire sample (N = 1269) who retained or moved to a different diagnostic group between their first and last diagnostic visits. Stability coefficients are denoted within each cell (coefficients are not adjusted for the age at first diagnosis; a high concordance with age-adjusted coefficients was observed) (eTable 4 in the Supplement gives coefficients adjusted for age at first diagnosis). To read the table, compare values across each row or vertically within each column. For example, of the 400 toddlers initially designated as having autism spectrum disorder (ASD), 336 retained this diagnosis at their last (final) diagnostic visit, yielding a diagnostic stability coefficient of 0.84, whereas 35 toddlers had ASD features but no longer met ASD criteria, 6 tested as developmentally delayed, 6 as language delayed, 10 had other developmental issues, and 7 were designated as typically developing with no residual symptoms. For transition tables with stability coefficients within 2-month age bands, see eFigure 4 in the Supplement. B, Changes in diagnosis across visits. Colors represent each diagnostic group and rings represent each diagnostic visit, with the smallest center ring representing the first visit. The left-most panel summarizes diagnostic changes for toddlers who received 2 diagnostic evaluations a mean of 15 months apart; the center represents toddlers who received 3 diagnostic evaluations a mean of 11 months apart; and the right-most panel represents toddlers who received 4 or more diagnostic evaluations a mean of 8 months apart. The heat map indicates that ASD was the most stable diagnostic category, and that toddlers initially suspected as having developmental delay (DD) or language delay (LD) frequently received a final diagnosis of ASD. TD indicates typical development.
Figure 2. Transition Table and Diagnostic Heat Maps A, Summary of the proportion of toddlers from within the entire sample (N = 1269) who retained or moved to a different diagnostic group between their first and last diagnostic visits. Stability coefficients are denoted within each cell (coefficients are not adjusted for the age at first diagnosis; a high concordance with age-adjusted coefficients was observed) (eTable 4 in the Supplement gives coefficients adjusted for age at first diagnosis). To read the table, compare values across each row or vertically within each column. For example, of the 400 toddlers initially designated as having autism spectrum disorder (ASD), 336 retained this diagnosis at their last (final) diagnostic visit, yielding a diagnostic stability coefficient of 0.84, whereas 35 toddlers had ASD features but no longer met ASD criteria, 6 tested as developmentally delayed, 6 as language delayed, 10 had other developmental issues, and 7 were designated as typically developing with no residual symptoms. For transition tables with stability coefficients within 2-month age bands, see eFigure 4 in the Supplement. B, Changes in diagnosis across visits. Colors represent each diagnostic group and rings represent each diagnostic visit, with the smallest center ring representing the first visit. The left-most panel summarizes diagnostic changes for toddlers who received 2 diagnostic evaluations a mean of 15 months apart; the center represents toddlers who received 3 diagnostic evaluations a mean of 11 months apart; and the right-most panel represents toddlers who received 4 or more diagnostic evaluations a mean of 8 months apart. The heat map indicates that ASD was the most stable diagnostic category, and that toddlers initially suspected as having developmental delay (DD) or language delay (LD) frequently received a final diagnosis of ASD. TD indicates typical development.CC BY · PubMed Central
Figure 3. Diagnostic Stability Plots by Age at First Diagnosis A, Plots show diagnostic stability per group across 2-month age intervals based on the age at first diagnostic evaluation. Age intervals with missing data points reflect an absence of toddlers who received their first diagnostic evaluation at that age. B-spline logistic regression line is shown; bands represent 95% CIs for the fit line. Overall stability was highest in toddlers initially designated as having autism spectrum disorder (ASD) or typical development as illustrated by the relatively tight CI bands, and the largely consistent stability coefficients within each age band. B, Diagnostic stability coefficients in the 4 age bins used in the logistic regression model across diagnostic groups. The lines represent 95% CIs. Coefficients were estimated based on within group logistic regression models. eFigure 5 and eFigure 6 in the Supplement give complementary analyses.
Figure 3. Diagnostic Stability Plots by Age at First Diagnosis A, Plots show diagnostic stability per group across 2-month age intervals based on the age at first diagnostic evaluation. Age intervals with missing data points reflect an absence of toddlers who received their first diagnostic evaluation at that age. B-spline logistic regression line is shown; bands represent 95% CIs for the fit line. Overall stability was highest in toddlers initially designated as having autism spectrum disorder (ASD) or typical development as illustrated by the relatively tight CI bands, and the largely consistent stability coefficients within each age band. B, Diagnostic stability coefficients in the 4 age bins used in the logistic regression model across diagnostic groups. The lines represent 95% CIs. Coefficients were estimated based on within group logistic regression models. eFigure 5 and eFigure 6 in the Supplement give complementary analyses.CC BY · PubMed Central
Figure 4. Comparison of Clinical Features in Toddlers With Autism Spectrum Disorder (ASD) Stratified by Identification Age Violin plots show differences in Autism Diagnostic Observation Schedule (ADOS) total scores (A), Mullen Expressive Language T scores (B), Vineland Adaptive Behavior Composite scores (C), and Mullen Receptive Language T scores (D) at the first diagnostic evaluation between toddlers with ASD identified at 12 to 18 months of age (early-age persistent ASD diagnosis), toddlers with ASD identified after 18 months (middle-age persistent ASD diagnosis), or toddlers not identified as having ASD at their first diagnostic visit (late-identified ASD). Black dots represent the mean. The width of the shape represents patient density, and the length illustrates the range of the scores. Data from 270 toddlers with typical development (TD) identified at their first diagnosis visit and retaining that diagnosis at their last visit are shown for comparison. Note that scores from the late-identified group were significantly different from toddlers with TD across all clinical domains, suggesting that symptoms were already present at the first diagnostic visit in a large fraction of late-identified ASD cases. Also note that 39 toddlers in the late-identified group (37%) did fall within the range of concern on the Autism Diagnostic Observation Schedule toddler module (cutoff score for concern using the few to no words algorithm = 10), however, were designated as non-ASD based on practitioner judgment, underscoring the challenges in differential diagnoses particularly at the youngest ages. Effect sizes are reported as Cohen d (95% CI). eFigure 3 in the Supplement gives an expanded figure that includes all diagnostic groups.
Figure 4. Comparison of Clinical Features in Toddlers With Autism Spectrum Disorder (ASD) Stratified by Identification Age Violin plots show differences in Autism Diagnostic Observation Schedule (ADOS) total scores (A), Mullen Expressive Language T scores (B), Vineland Adaptive Behavior Composite scores (C), and Mullen Receptive Language T scores (D) at the first diagnostic evaluation between toddlers with ASD identified at 12 to 18 months of age (early-age persistent ASD diagnosis), toddlers with ASD identified after 18 months (middle-age persistent ASD diagnosis), or toddlers not identified as having ASD at their first diagnostic visit (late-identified ASD). Black dots represent the mean. The width of the shape represents patient density, and the length illustrates the range of the scores. Data from 270 toddlers with typical development (TD) identified at their first diagnosis visit and retaining that diagnosis at their last visit are shown for comparison. Note that scores from the late-identified group were significantly different from toddlers with TD across all clinical domains, suggesting that symptoms were already present at the first diagnostic visit in a large fraction of late-identified ASD cases. Also note that 39 toddlers in the late-identified group (37%) did fall within the range of concern on the Autism Diagnostic Observation Schedule toddler module (cutoff score for concern using the few to no words algorithm = 10), however, were designated as non-ASD based on practitioner judgment, underscoring the challenges in differential diagnoses particularly at the youngest ages. Effect sizes are reported as Cohen d (95% CI). eFigure 3 in the Supplement gives an expanded figure that includes all diagnostic groups.CC BY · PubMed Central

Verification

Claims the paper makes

37 Own result · 27 Not checked yet · 6 No basis given · 5 Own method · 5 Own interpretation

Verify all 38 references — free account

Show all 50 claims

  • Probably not a substantial claim. 23
  • It may state two separate claims as one. 9
30 of 30
  • The paper relies on this without support.

    The 1269 toddlers longitudinally evaluated are representative of the general population of toddlers.

    “A total of 2241 toddlers 12 to 36 months of age were referred for a diagnostic evaluation to an autism expert evaluation center created at University of California, San Diego.”
    “A total of 1269 of the 2241 toddlers were longitudinally evaluated 2 or more times and were the focus of this study.”
  • The paper relies on this without support.

    The final diagnostic evaluation at the last visit provides a valid reference standard for true ASD diagnostic status.

    “A toddler was designated as having 1 of the following: ASD, ASD features, developmental delay, language delay (LD), other issue, TD, or typical sibling of an ASD proband.”
    “Stability coefficients were first calculated within 2-month age bands by determining the proportion of toddlers with a particular diagnosis at their first diagnostic visit who retained that same diagnosis at their last visit.”
  • The paper relies on this without support.

    The lack of masking of psychologists to previous diagnoses did not materially bias the diagnostic stability estimates.

    “Psychologists were not masked to previous diagnoses during longitudinal test visits.”
  • The paper relies on this without support.

    The human brain has an enormous capacity to resculpt and remodel, particularly during the first postnatal years.

    “The human brain has an enormous capacity to resculpt and remodel, particularly during the first postnatal years.”
  • The paper relies on this without support.

    DSM-5 procedures and criteria require consideration of historical information regarding ASD symptoms.

    “This lack of masking was because DSM-5 procedures and criteria require consideration of historical information regarding ASD symptoms.”
5 of 5

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

2 of 2

Citations

Cites

Show all 46 references

Research works printed but not linked: 1

  1. Sparrow, 2005, Vineland-II Scales of Adaptive Behavior: Survey Form Manual

Cited by

Sorted by citations

Items 301–333 of 333

Metadata

Authors
Pierce, K., Gazestani, V. H., Bacon, E., Barnes, C. C., Cha, D., Nalabolu, S., Lopez, L., Moore, A., Pence-Stophaeros, S., & Courchesne, E.
Affiliations
  • University of California San Diego (US)
  • Department of Neurosciences, University of California, San Diego, La Jolla
Corresponding author
Pierce, Karen
Venue
JAMA Pediatrics
Published
2019-06-01
Volume
173
Issue
6
Pages
578
ISSN
  • 1072-4710 (ISSN-L)
  • 2168-6203
OpenAlex
W2941892929
Cited by
333
References
46
Citations by year
2026: 27 · 2025: 39 · 2024: 43 · 2023: 49 · 2022: 65 · 2021: 66 · 2020: 32 · 2019: 11
FWCI
25.27
Type
journal article
Language
en
Open access
Licence
cc-by
Locations
  • Archives of Pediatrics and Adolescent Medicine (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
Funders
  • National Institutes of Health (5r01mh110558-05)
  • National Institute of Mental Health (R01MH104446, R01-MH110558, R01MH036840, P50MH081755)
  • National Institute on Deafness and Other Communication Disorders (R01DC016385)
Topics
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Autism Spectrum Disorder Research
  • Health Sciences › Medicine › Psychiatry and Mental health › Child Nutrition and Feeding Issues
  • Social Sciences › Psychology › Developmental and Educational Psychology › Language Development and Disorders
Keywords
autism spectrum disorder, diagnostic stability, toddlers, primary care, developmental delay, language delay, sex differences, age differences, typical development
MeSH terms
Autism Spectrum Disorder, Developmental Disabilities, Child, Preschool, Female, Follow-Up Studies, Humans, Infant, Male, Mass Screening, Phenotype, Primary Health Care, Prospective Studies, Psychiatric Status Rating Scales, Referral and Consultation, Logistic Models, Early Diagnosis

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

A language model drew the claims from the paper on 2026-10-07, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-07 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review