**1. Introduction**

Huntington's disease (HD) is an autosomal dominant, neurodegenerative disorder that causes cognitive, behavioral, and motor symptoms [1]. An abnormal number of repeats (≥36) of cytosine–adenine–guanine (CAG) within the huntingtin gene causes the mutation that leads to HD. The negative relationship between the number of CAG repeats that a person has and the age of motor onset (AMO) has been well-established for patients with adult-onset HD (AOHD) [2–6]. However, given the rarity of juvenile-onset HD (JOHD), the relationship between CAG repeat length and AMO is less clear. In general, it is believed that the correlation between CAG repeat length and AMO increases in higher CAG repeat lengths [7–9], but these reports are limited by small sample sizes. A previous review paper utilized data from 15 patients with JOHD who had a CAG repeat length of >60 to demonstrate that the association between AMO and CAG seemed to be higher at CAG lengths of 60 to 80. However, at CAG repeat lengths above approximately 80, the association between AMO and CAG became much weaker [10]. In order to further investigate this phenomenon in a larger group, the same authors gathered patient data from seven separate case studies or case series of patients with JOHD and found 26 patients with a CAG repeat length of >80. In this cohort, CAG repeat length accounted for only 26% of the variance in AMO [10]. A recent systematic review was performed and gathered information on more than 200 patients with reported JOHD [11]. This cohort of patients included a wide range of CAG repeats from 39 to 265. The authors reported a Pearson's correlation coefficient of −0.56 between CAG repeat length and the age of onset of clinical symptoms. In line with other studies, the strength of correlation seemed to weaken at longer CAG repeats [11]. Both of these large reviews are limited, however, by their reliance on a collection of data, rather than primary collection from a single cohort. Thus, the aim of the present study was to investigate the relationship between CAG repeat length and AMO amongs<sup>t</sup> patients with JOHD, with the goal of providing additional clarity to this current gap in knowledge. We also leveraged data from the Enroll-HD platform and the Predict-HD study to compare the results of patients with JOHD to those of patients with AOHD.

#### **2. Materials and Methods**

#### *2.1. Description of Data*

We analyzed data from a large sample of participants enrolled in the Kids-JOHD studies (*N* = 27). These studies recruited participants from across the United States who were between the ages of 6 and 25 at the time of their first study visit and had a confirmed diagnosis of JOHD. To be enrolled, all participants were required to have already had molecular confirmation of HD and a clinical diagnosis of JOHD before traveling to the University of Iowa for assessment. To confirm that the participant was manifesting significant motor symptoms (meaning that they would meet criteria for a clinical diagnosis of JOHD), we utilized a combination of (1) detailed parental report of motor symptoms; (2) clinical evaluation by a pediatric neurologist at the University of Iowa; and (3) a Unified Huntington's Disease Ratings Scale (UHDRS) Total Motor Score (TMS) [12] of at least 12 to consider a participant to be motor-manifest. Each subject was required to have all three to be considered motor-manifest. Although a TMS of 12 may be considered low, patients with JOHD often present with far greater hypokinetic symptoms than do those with AOHD, and the UHDRS may not be sufficiently sensitive to these as it was designed for AOHD. These criteria excluded 10 individuals enrolled in the Kids-JOHD study and represented subjects who were tested locally mainly based on family history or behavioral issues, rather than significant motor symptoms. Of note, the TMS derived from the UHDRS takes into account various motor symptoms beyond just chorea. At their baseline visit, the participants with JOHD demonstrated the highest scores in saccade initiation, finger tapping, and rigidity (Figure 1). Additionally, the summative chorea score was one of the areas where patients with JOHD had the lowest score. This is of particular importance to consider when confirming the diagnosis of JOHD, as symptoms other than chorea can often be present. Participants were also excluded if they had a history of brain surgery or significant head trauma. All participants and their guardians (if under 18) signed informed consent prior to enrolling in these studies, which were approved by the University of Iowa Institutional Review Board (IRB).

**Figure 1.** Mean scores on the various domains of the Unified Huntington's Disease Rating Scale (UHDRS). Specifically, ocular pursuit score is the mean of horizontal and vertical pursuit scores; saccade initiation is the mean of the horizontal and vertical initiation scores; saccade velocity is the mean of the horizontal and vertical velocity scores; finger tapping is the mean of the right and left hand scores; pronate/supinate is the mean of the right and left hand scores; rigidity is the mean of the scores in the right and left arms; dystonia score is the mean scores from the trunk, right and left upper extremities, and right and left lower extremities; chorea score is the mean of maximal chorea scores from the face, buccal-oral-lingual scores, trunk, right and left upper extremities, and right and left lower extremities. All other scores include only one component and represent the mean of that domain.

We also analyzed data from the Predict-HD study [13] and from the Enroll-HD platform [14]. Specifically, we included participants from both studies who received a motor diagnosis of HD during the study after the age of 21 and who had a CAG repeat length of less than 60. This was done to ensure that we were investigating only those participants with adult-onset HD (AOHD). Participant reporting of historical motor onset timing is included in the Enroll-HD platform, but this may be subject to significant recall bias. Participants must have had a motor exam conducted with a diagnostic confidence level (DCL) of less than four to be included in the analyses from Enroll-HD and Predict-HD. The age at which these participants had their first report of a diagnostic confidence level of four on the UHDRS [12] was considered the age of motor onset (AMO). There were 242 participants who received a motor diagnosis during the Predict-HD study and 782 participants in the Enroll-HD database.

#### *2.2. Statistical Analysis*

The primary analysis investigated the relationship between CAG repeat length and AMO amongs<sup>t</sup> the JOHD participants using simple regression models. We used natural cubic splines to transform the independent variable (CAG repeat length) to investigate nonlinear relationships for the primary analysis. RStudio was used for all analyses, with a *p*-value of <0.05 considered statistically significant.

#### *2.3. Institutional Ethical Approval*

The University of Iowa institutional review board initially approved the study on 10/13/2011 (IRB # 201109879). For participants younger than 18 years, parents or guardians provided written consent and children provided written assent. For participants who were 18 years or older, participants provided written consent.
