*2.1. Participants*

This sample consists of children and young adults (ages 5–26) who participated in the Kids-HD or Kids-JOHD studies at the University of Iowa. Details of the Kids-HD program can be found in the 2019 publication on brain development in the Kids-HD sample [12]. From the children/young adults at risk from the Kids-HD study, we utilized the Gene-Non-Expanded (GNE) as controls.

Recruitment for the Kids-JOHD study was done through the Center of Excellence at the University of Iowa and national Huntington's Disease Society of America events. To be eligible, participants required a genetic confirmation of an expanded CAG repeat. If the participant was older than 21 at the time of assessment, they were required to have had a clinical diagnosis prior to the age of 21.

All participants were recruited from across the United States and travelled to the University of Iowa to complete the study. The study followed an accelerated longitudinal design (ALD), where some individuals were assessed once, while others were assessed on multiple occasions with variable lengths of follow-up [12]. This design is less a ffected by attrition, which is especially important in this population due to disease progression. It also allows us to study the disease over a larger age range, and thus a greater number of participants. However, each age range may not be equally represented [13]. Participants completed multiple visits if they were willing and able to return for follow-up.

Both study protocols were approved by the Institutional Review Board (IRB) at the University of Iowa and were conducted in accordance with the Declaration of Helsinki. Parents or legally authorized representatives provided written consent for participants under the age of 18 or those who were unable to provide consent due to disease progression. Participants 18 years of age or older with the capacity to consent provided written informed consent for participation. This project was initially approved by the University of Iowa Institutional Review Board (IRB Number: 201109879) on 9 January 2012, and most recently approved on 15 May 2020.

#### *2.2. Genetic Analysis*

A DNA sample of either blood or saliva was obtained from participants in the Kids-HD study. The presence or absence of the mutant CAG expansion was determined using PCR analysis by the University of Iowa Molecular Diagnostic Laboratory. This analysis was done for research purposes

only, and results were not disclosed to anyone including the participants, participants' families, or the clinical research study sta ff [12]. All the participants in the Kids-JOHD study had to have molecular confirmation prior to enrollment and came with medical record documentation of the gene expansion.

#### *2.3. Motor Rating*

Like AOHD, the diagnosis of JOHD requires the presence of significant motor abnormality. A few of the participants in the JOHD study had been tested locally, ye<sup>t</sup> examination by the neurologist locally showed no significant, or only subtle, motor findings. Our aim was to examine a homogenous group of motor-manifest patients. To that end, all participants were assessed by a trained motor examiner using the Unified Huntington's Disease Rating Scale (UHDRS). A total motor score (TMS) was obtained by summing the core UHDRS items. To be included in the analysis, the JOHD participants were required to have a TMS of greater than 18. The rationale for the relatively high cut-o ff for the TMS is because the UHDRS is sensitive to developmental motor changes such that normal developing younger children will show higher scores than older children. Therefore, even in a large cohort of children at risk, but who did not inherit the gene expansion (Gene Non-Expanded or GNE), the UHDRS can be as high as 17, as shown by our previous analysis of the Kids-HD cohort [14]. In the current cohort, the highest TMS was 17 for a GNE participant. This was used as a guide for the JOHD group where the cut-o ff was determined to be 18 (with lowest TMS in the JOHD group being 19).

#### *2.4. Behavioral Measures*

Behavior Rating Inventory of Executive Function (BRIEF). While some standardized tests are designed to measure specific executive functioning skills in an individual, the BRIEF is a questionnaire designed to assess executive function behaviors [14]. Items are coded as "never", "sometimes", or "often" being a problem. The total Global Executive Composite score is divided into the Behavioral Regulation Index (BRI; Inhibit, Shift, Emotional Control) and the Metacognition Index (MI; Initiate, Working Memory, Plan/Organize, Organization of Materials, Monitor). Higher scores indicate more problematic behaviors.

If the participant was younger than 18 years old, the informant completed the BRIEF-Parent form; however, if the participant was 18 years old or older, the informant completed the BRIEF-Adult (BRIEF-A) Informant form. Informants were individuals who accompanied the participant to the research appointment. For participants younger than 18 years of age, the informant was a parent or legal guardian. For participants 18 years of age or older, the informant was a parent, guardian, spouse/partner, or friend. Additionally, participants age 18 and older completed the BRIEF-A Self Report Form.

With regard to the Pediatric Behavior Scale-30 (PBS-30), the PBS is a parent/informant-report measure designed to assess broad domains of functioning; it was derived from the full 165-item version of the PBS [15]. The PBS-30 includes 30 Likert-scale items ("almost never or not at all" to "very often or very much") that describe di fferent behaviors, where higher scores reflect more problems. Four scales are calculated: Aggression/Opposition, Hyperactivity/Inattention, Depression/Anxiety, and Physical Health. The same form was used for all ages in the current study.

Supplementary Tables S1 and S2 summarize the number of observations available for each BRIEF and PBS subscale. Supplementary Table S3 summarizes the di fferences in scores between informant and self-reports for the BRIEF-A.

#### *2.5. Statistical Analyses*

Raw scores from the two scales were analyzed across the two groups via mixed linear regression models with individual BRIEF and PBS subscales as the outcome variables. Group, age, and sex were included as main e ffects in all regression models, and participant ID and family ID were included as random e ffects to account for non-independency of the observations. Including family ID in the model controlled for the e ffects of having more than one person from a family. Adding the participant ID

controls meant that the correlation between repeat visits from a single individual was controlled for in the model. Sex by group interactions were entered into the model and subsequently removed if not significant.

The impact of CAG repeat length was examined in the JOHD group for measures that were significantly different between groups using mixed linear regression models to predict behavioral and executive functioning outcomes by CAG repeat expansion length.

Differences in BRIEF-A informant and self-reported scores were analyzed in patients 18 and older by calculating the difference in informant-reported and self-reported raw scores for each BRIEF-A subscale outcome measure. Difference scores were predicted with mixed linear effects models including main effects of group, age, and sex and controlling random effects of family ID. The difference score represents a difference in perspective on behavioral/cognitive problems between the participant and their parent. Thus, difference scores close to 0 represent agreemen<sup>t</sup> between self and proxy assessments, while large differences represent incongruent perspectives. All models were corrected for multiple comparisons using the False Discovery Rate (FDR) method. All analyses were completed using RStudio version 1.2.5042 (RStudio, PBC, Boston, MA, USA).
