Table 1. Characteristics of studies that associate exposure to agricultural pesticides with autism by home address — studies cited according to the year of publication.

Author	Study population	Exposure and outcome	Results
Roberts, et al. 21	Children born between 1996 and 1998, residing in the 19 countries in the Central Valley in California. Cases (n=465) and controls (n=6,975)	Home address of birth and time in pregnancy during which exposure to pesticides reported by data from the California Department of Pesticide Regulation. Children receiving services for autism or with an ASD diagnostic code.	The quantity of organochlorines for the cases suggested OR=6.1 for the ASD. When comparing the fourth quartile versus no exposure, the autism risk coefficient had associations during pregnancy with organophosphates (p=0.042), organochlorines (p=0.025), trifluralin (p=0.046), and bifenthrin (p=0.048). During the embryogenesis of the central nervous system, the results for organochlorines were attenuated with an increase in the distance from home to the application, from 250 m (p=0.001) to 1,000 m (p=0.006).
Shelton, et al. 22	Children aged 2–5 years, residing in California between 1997 and 2008 with an ASD (n=486), DD (n=168), and typical development (n=316).	Home address for 3 months preconception until born, linked to the use of the California Pesticide Use Report. Previous diagnosis of ASD: ADI-R+ADOS. Typical development and DD: SCQ	Children with ASD were 60% more likely to have organophosphates applied close to the home (aOR=1.60) purchased with GC and increased for exposures in the third trimester (OR=2.0). Cases with DD had aOR = 2.48 of having carbamate pesticides applied close to home during pregnancy. Associations decreased as the size of the beam increased. Exposure to pyrethroid insecticide applications before conception or during the third trimester was at increased risk for ASD and DD (ORs 1.7–2.3).
Von Ehrenstein et al. 23	Individuals of any age, resident at birth, and diagnosed in California counties. Recruited between 1998 and 2010. Cases (n=296) and controls (n=35,370).	Home address at birth related to data from application reports with evaluation 3 months before pregnancy, during pregnancy, and during the first year of life. Service records are collected through regional centers and categorized by the department, using DSM-IV ed.	Stronger associations between ASD risk and exposure during the development period were for chlorpyrifos (OR=1.15), diazinon (OR = 1.14), and avermectin (OR=1.14). For ASD with ID, exposures resulted in attenuated effect estimates before and during pregnancy, while probability ratios became more pronounced for exposures in the first year of life, for glyphosate (1.60), diazinon (1.45), malathion (1.29), and bifenthrin (1.33). Exposure in the three pre-pregnancy months had weaker associations with ASD. Exposure to any pesticide substance in the first year of life brought an increase in risk for ASD combined with ID by up to 50%.

ASD: autism spectrum disorders; DD: developmental delay; ID: intellectual disability; OR: Odds Ratio; aOR: adjusted Odds Ratio; GC: control group; GP: general population; DSM-IV: Diagnostic and Statistical Manual of Mental Disorders, 4th edition; ADI-R= Autism Diagnostic Interview – Revised; ADOS: Autism Diagnostic Observation Schedule; SCQ: Social Communication Questionnaire.