47,XXY, also known as Klinefelter syndrome, is the most frequently occurring sex chromosome aneuploidy with an incidence of approximately 1:600 live male births. It is medically characterized by hypogonadism, azoospermia, micropenis, and tall stature. The neurodevelopmental phenotype includes speech and motor delay, language-based learning disabilities, executive dysfunction, reduced expressive language, and increased risk for anxiety. Academic achievement for reading and writing has been shown to be largely impacted in school-aged males with 47,XXY. For example, an early study reported an increased risk for dyslexia due to challenges in processing speed, language, and memory despite intact intelligence quotients (IQ). Further, a more recent study revealed mastering school curricula often requires special education intervention. Math-related achievement, however, is a component of the 47,XXY neurodevelopmental profile that has been largely overlooked. There are a few studies that have indirectly demonstrated reduced arithmetic capabilities in 47,XXY compared to neurotypical controls, but none have delineated the specific mathematical skillsets that are impacted by a supernumerary X. Thus, we aimed to expand on this paucity of literature by comprehensively examining mathematical skills in school-aged males with 47,XXY. Further, though consistent in symptomatology, the phenotype is widely variable in severity depending on factors such as timing of diagnosis, family learning disability (FLD), and access to treatment. The potential profound benefits of Hormonal Replacement Therapy (HRT), or the administration of testosterone enanthate, on neurodevelopmental challenges in males with 47,XXY have been well-documented. In school-aged males specifically, HRT has been associated with increased expressive ability, working memory, and executive function. Therefore, we also explored the potential effect of HRT on mathematical achievement in our cohort. Our cohort consisted of 111 prenatally diagnosed school-aged males with 47,XXY (CA: 133 months). Those who were postnatally diagnosed, presented with mosaicism or other copy number variants (CNVs), or had a positive family history of learning disabilities were excluded from this study. An in-depth history was recorded in which HRT statuses were obtained. All participants were assessed by their respective pediatric endocrinologists who determined HRT administration. Study participants completed the KeyMath-3 and Wechsler Intelligence Scale for Children, 4th or 5th Edition (WISC-IV/V). For the purpose of this study, only the Arithmetic subtest was used. To investigate the potential effect of HRT on mathematical skills, participants were bifurcated into untreated (No-T, N=22) and treated (T, N=89) groups. A two-tailed t-test determined statistically significant group differences on each assessment, if any. Participants showed no significant differences on any demographic scales including maternal and paternal age (p=0.97 and p=0.19, respectively), birthweight (p=0.90), delivery method (p=0.79), race (p=0.55), or socioeconomic status (p=0.88). Significant effect of treatment was observed on several mathematical subtests of the KeyMath-3 including Numeration (p=0.005), Algebra (p=0.03), Geometry (p=0.03), Mental Computation and Estimation (p=0.005), Multiplication and Division (p=0.03), and Foundations of Problem Solving (p=0.05). The T group also demonstrated significantly increased scores on the Operations (M=115.28, p=0.004) and Total Test (M=116.38, p=0.01) areas compared to the No-T group (M=103.65 and M=106.62, respectively). Additionally, while not statistically significant, the T group exhibited higher average scores on the rest of the subtests and areas including Measurement (M=12.88), Data Analysis (M=12.40), Basic Concepts (M=117.03), Addition and Subtraction (M=12.55), Applied Problem Solving (M=12.21), and Applications (M=112.61) compared to the No-T group (M=11.95, M=11.55, M=110.90, M=11.11, M=11.18, and M=106.29, respectively). On the WISC-IV/V Arithmetic subtest, the T group (M=115.28, SD=2.72) did demonstrate a higher mean with less variability than the No-T group (M=115.28, SD=3.67), though not quite statistically significant (p=0.06). Our findings are the first to shed light on the mathematical profile of treated and untreated school-aged males with 47,XXY. Our observations of reduced scores in the untreated group corroborates a past study describing a proposed susceptibility to mathematics disability in this population. Other previous studies have documented the similar cognitive mechanisms shared between math disability and reading disability, as well as the increased risk of comorbidity. However, the neurodevelopmental profile for males with 47,XXY is usually restricted to only include language-based learning disability. Our findings suggest the phenotype may be expanded to include all learning disabilities, including those math-related.
Read full abstract