Identifying Replicable Subgroups in Neurodevelopmental Conditions Using Resting-State Functional Magnetic Resonance Imaging Data
Bibliographic record
Abstract
Importance: Neurodevelopmental conditions, such as autism spectrum disorder (ASD), attention-deficit/hyperactivity disorder (ADHD), and obsessive-compulsive disorder (OCD), have highly heterogeneous and overlapping phenotypes and neurobiology. Data-driven approaches are beginning to identify homogeneous transdiagnostic subgroups of children; however, findings have yet to be replicated in independently collected data sets, a necessity for translation into clinical settings. Objective: To identify subgroups of children with and without neurodevelopmental conditions with shared functional brain characteristics using data from 2 large, independent data sets. Design, Setting, and Participants: This case-control study used data from the Province of Ontario Neurodevelopmental (POND) network (study recruitment began June 2012 and is ongoing; data were extracted April 2021) and the Healthy Brain Network (HBN; study recruitment began May 2015 and is ongoing; data were extracted November 2020). POND and HBN data are collected from institutions across Ontario and New York, respectively. Participants who had diagnoses of ASD, ADHD, and OCD or were typically developing (TD); were aged between 5 and 19 years; and successfully completed the resting-state and anatomical neuroimaging protocol were included in the current study. Main Outcomes and Measures: The analyses consisted of a data-driven clustering procedure on measures derived from each participant's resting-state functional connectome, performed independently on each data set. Differences between each pair of leaves in the resulting clustering decision trees in the demographic and clinical characteristics were tested. Results: Overall, 551 children and adolescents were included from each data set. POND included 164 participants with ADHD; 217 with ASD; 60 with OCD; and 110 with TD (median [IQR] age, 11.87 [9.51-14.76] years; 393 [71.2%] male participants; 20 [3.6%] Black, 28 [5.1%] Latino, and 299 [54.2%] White participants) and HBN included 374 participants with ADHD; 66 with ASD; 11 with OCD; and 100 with TD (median [IQR] age, 11.50 [9.22-14.20] years; 390 [70.8%] male participants; 82 [14.9%] Black, 57 [10.3%] Hispanic, and 257 [46.6%] White participants). In both data sets, subgroups with similar biology that differed significantly in intelligence as well as hyperactivity and impulsivity problems were identified, yet these groups showed no consistent alignment with current diagnostic categories. For example, there was a significant difference in Strengths and Weaknesses ADHD Symptoms and Normal Behavior Hyperactivity/Impulsivity subscale (SWAN-HI) between 2 subgroups in the POND data (C and D), with subgroup D having increased hyperactivity and impulsivity traits compared with subgroup C (median [IQR], 2.50 [0.00-7.00] vs 1.00 [0.00-5.00]; U = 1.19 × 104; P = .01; η2 = 0.02). A significant difference in SWAN-HI scores between subgroups g and d in the HBN data was also observed (median [IQR], 1.00 [0.00-4.00] vs 0.00 [0.00-2.00]; corrected P = .02). There were no differences in the proportion of each diagnosis between the subgroups in either data set. Conclusions and Relevance: The findings of this study suggest that homogeneity in the neurobiology of neurodevelopmental conditions transcends diagnostic boundaries and is instead associated with behavioral characteristics. This work takes an important step toward translating neurobiological subgroups into clinical settings by being the first to replicate our findings in independently collected data sets.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.006 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.002 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.004 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".