A Longitudinal Cohort Study of Precocious Puberty and Autism Spectrum Disorder

Abstract

<b><i>Introduction:</i></b> Autism spectrum disorder (ASD) is defined by persistent deficits in communication, socialization, and stereotypic behaviors. It was previously hypothesized that hormone dysfunction is a frequent occurrence among children diagnosed with an ASD. <b><i>Objectives:</i></b> A hypothesis-testing epidemiological study examined the relationship between precocious puberty (PP) (a known disorder of childhood sex hormone dysfunction) and ASD diagnoses. <b><i>Methods:</i></b> The Independent Healthcare Research Database is composed of de-identified linked eligibility and claims health-care records prospectively generated from the Florida Medicaid system. A cohort of 101,736 children eligible for Florida Medicaid from 1990 to 2009 and continuously eligible with ≥10 outpatient office visits during the 120-month period following birth were examined using SAS and StatsDirect software. There were 1,593 children (15,738 person-years) in the ASD diagnosed cohort utilizing the Diagnostic and Statistical Manual of Mental Disorders, 4th revision criteria (the International Code for Disease, 9th revision [ICD-9] codes: 299.00 or 299.80) and 100,143 children (996,835 person-years) in the undiagnosed cohort. <b><i>Results:</i></b> The incidence rate of PP (ICD-9 code: 259.1) was examined using Cox proportional hazards ratio (HR) and frequency models. PP per 10,000 person-years in the ASD cohort (43.2) relative to the undiagnosed cohort (13.7) was significantly increased in frequency modeling (risk ratio = 3.15, <i>p</i> &#x3c; 0.0001) and Cox proportional HR modeling (adjusted HR = 4.64, <i>p</i> &#x3c; 0.0001). Further analyses revealed the incidence rate of PP diagnosed after 3 years of age was significantly increased (adjusted HR = 5.16, <i>p</i> &#x3c; 0.0001) in the ASD cohort relative to the undiagnosed cohort but not for the incidence rate of PP diagnosed before 3 years (adjusted HR = 1.57, <i>p</i> = 0.44). <b><i>Conclusion:</i></b> This hypothesis-testing study provides strong evidence of an increased incidence rate of PP among children diagnosed with an ASD.