Autism and Obesity: Assessing Antipsychotic-Induced Weight Gain and BMI Associated SNPs

Background:  A general assumption is that the observed elevated rate of obesity in autism (i.e., 40%) is caused by antipsychotic-induced weight gain (AIWG). It is not clear if obesity is co-occurring with autism or is related to AIWG.

Objectives: Our hypothesis is that the prevalence of known AIWG associated SNPs in obese and non-obese autistic subjects is comparable; thus, AIWG cannot be the only reason for the observed higher rate of obesity. To test this hypothesis, we evaluated the prevalence of AIWG associated SNPs in obese and non-obese autistic subjects.

Methods: We curated a list of about n=150 SNPs with at least one report of association with AIWG. We assessed phenotypic and genotyping data (three Illumina platforms) from Simons Simplex Collection (SSC). We analyzed genotyping data for n=115 AIWG SNPs from > 2000 probands from SSC. BMI data on probands was used to identify obese and non-obese subjects (> 1000 probands met the applied BMI cut off for each weight category). Similarly, we evaluated unaffected SSC siblings (> 2000) stratified to obese and non-obese groups.

Results:  AIWG SNPs analysis. (1) Only one out of 115 AIWG SNPs (rs7702361) achieved an FDR corrected significant p-value, comparing obese versus non-obese autistic subjects. (2) No difference was found for rs7702361, comparing obese versus non-obese unaffected siblings. (3) Current medication history data (i.e., Attention Deficit-ADD medication and Mood Stabilizers) were used to assess the potential medication effects on obesity in probands. Analyses were run separately for probands stratified by medication use (e.g., Mood Stabilizers “Yes” or “No”). The rs7702361 SNP showed an association with both Mood Stabilizers “Yes” or “No” subsets and none with the subjects on ADD medication; suggesting that the detected association of this SNP with obesity category in probands is more likely independent of drug exposure and may be related to the underlying condition (i.e., co-occurring autism and obesity). Comparing obese probands stratified by Mood Stabilizers use (i.e., “Yes” versus “No”) did not detect any differences for rs7702361, which further validates the above conclusion about the potential association of this SNP with the underlying co-occurring conditions not drug exposure. BMI SNPs analysis. Similarly, we evaluated genotyping data for n=74 BMI SNPs for SSC subjects. This analysis showed both shared (i.e., familial BMI/obesity risk factors) and distinct BMI associated SNPs in probands and siblings, which may indicate potential differences in the underlying mechanism of obesity in autistic and non-autistic family members. This observation further complements/strengthens our finding for the AIWG SNPs.

Conclusions:  Our results suggest that factors beyond drug side effect may contribute to the observed higher rate of obesity in autism. Furthermore, an association with rs7702361 (located in the RXFP3 gene, ch5p13.2) in autistic subject with obesity was found, which appears to be independent to anti-psychotic medication use. It is intriguing that recent reports highlight the importance of this gene and related networks in neuropsychiatric disease and eating disorders. Further investigations are warranted on the potential role of the RXFP3 gene in association with co-occurring autism and obesity.