Sickle Cell Disease and Autism Spectrum Disorders

Background: Over 60 known medical disorders have been associated with increased risk of autism spectrum disorders (ASD) but none has been shown to be protective, either genetically or postnatally. Sickle cell disease (SCD) is a common heritable hemoglobinopathy that primarily affects African Americans (AA) with onset during the first year of life and associated with neurological complications, including stroke and cognitive decline, but has not been reported in association with ASD. SCD may provide postnatal physiologic stress that has a cellular protective effect during brain development despite the risk for vascular injury. Clinical improvements during fever in ASD also may result from activation of cellular stress responses (CSR). Objectives: This study examined the prevalence of SCD in AA with ASD, intellectual disability (ID) and cerebral palsy (CP) compared to the US population. We hypothesized that SCD might decrease the risk or ameliorate ASD severity and therefore the SCD prevalence among children with ASD would be lower than expected. Methods: Data on ASD, ID, and CP were obtained from the Autism and Developmental Disabilities Monitoring (ADDM) Network. ADDM identifies children with ASD, ID, and CP through systematic review of records from multiple educational and health sources. The current analysis utilized data on 8-year-old children from fourteen regions of the US over three surveillance years: 2000, 2002, and 2004. The total number of AA in ADDM included 1216 children with ASD, 1243 with ID, and 241 with CP. Presence of SCD in ADDM was identified by SCD diagnostic billing code or notation of SCD diagnosis in record. Comparison data on SCD in the general population were derived from the US National Newborn Screening Report - 2000, as the total number with SCD among all AA newborns screened nationwide. Results: Four children with SCD were found among 1216 AA with ASD (3.29/1000). SCD observed in the ASD group was not significantly higher than in the general AA population (Relative Risk =1.68; 95% Confidence Limits (CL) =0.63, 4.48). However SCD in CP and ID was 8.49 (95%CL: 3.21, 22.48) and 2.47 (95%CL: 1.11, 5.50) times, respectively, greater than expected. Clinical phenotypes in the 4 AA children with ASD and SCD showed mild cognitive impairment. Conclusions: Compared with the national AA population, SCD in ASD is not significantly more prevalent, whereas SCD in CP and ID is significantly more prevalent. Data from the small number of children in this pilot study do not support our hypothesis of decreased risk of ASD in SCD; however, SCD may ameliorate its cognitive severity. The higher frequency of SCD in children with CP is likely attributable to strokes. Behavioral changes and ID may also occur in some children with SCD due to early cerebrovascular disease. Benefits of CSR may be masked by the presence of vascular damage. Future studies should clarify the severity of SCD and its relationship to autism and protection or neurological injury due to SCD. A very large sample size will be required for sufficient statistical power to define the role of SCD in ASD.