EVALUATING THE IMPACT OF BIOLOGICAL SEX ON ADHD PRESENTATION, PREVALENCE, AND GENETIC RISK

Attention-deficit hyperactivity disorder (ADHD) is a neurodevelopmental disorder that impairs executive functioning, vigilance-attention, and motivation. Due to this, individuals with ADHD are at higher risk of addiction, poor academic and professional outcomes, and social deficits. Heterogeneity in presenting symptoms is well-established and may result in a delayed or missed diagnosis. The prevalence of ADHD is reported from two to seven times higher for males than females. The prevalence of ADHD appears consistent in childhood and adulthood. Only half of those diagnosed in childhood report persisting symptoms, implying many are first diagnoses as adults and those first diagnosed in adulthood tend towards a different symptom profile. Such sex and age trends may reflect protective effects of “female sex”, children “growing out of” ADHD, or adults experiencing a different clinical entity. However, others argue that the high heritability of ADHD (0.6-0.85) and similar genetic risk in females suggests that these trends may be due to a heterogenous expression of symptoms in response to the environment (e.g., modulated by the female or adult experience). We use Vanderbilt University Medical Center's (VUMC) biobank (n=3,285,882 electronic health records (EHR) and 119,750 genotyped samples) to analyze ADHD prevalence and genetic architecture. We observed the ADHD-associated ICD codes (n=38,419) were less frequent in EHR-recorded females relative to males (n=14,395 vs. 24,024) and the median age at first diagnosis was substantially older (21.72 years, IQR=20.96 vs.15.05 years, IQR=9.1). Among subset of European ancestry patient genotyped in VUMC (n=69,397), we observed an ADHD polygenic risk scores (PRS) was significant independent predictor of diagnosis, with stronger effects on females (males, beta= 13.47, p=0.03; females, beta=16.90, p=7.4e-7), and female cases having higher average PRS than male cases (p=0.04). In a sex-specific phenome-wide association study (PheWAS), the ADHD PRS was associated with similar phenotypes regardless of sex, including substance/tobacco use, other psychiatric disorders, obesity, diabetes mellitus, and respiratory problems. Our findings that female patients with ADHD appear to have higher genetic liability for the condition despite lower rates of diagnosis are consistent with previous studies. Additionally, ADHD PRS did not demonstrate differential comorbidity structures based on sex in VUMC. One explanation for this is that established genetic proxies of disease inadequately reflect the nuances of particular behaviors of ADHD subtypes, including but not limited to exhibition of externalizing hyperactive subtype (ADHD-H) opposed to internalizing inattentive subtype (ADHD-I), which is reported more frequently in females. Therefore, obtaining clinical diagnoses in females may require symptom manifestations that are largely overlapping with their male counterparts. Additional work in various EHR resources may shed additional light on the complexities and nuances of sex-specific trends in ADHD-associated comorbid phenotypes.