Objective:

To evaluate the association of ALS with pesticide exposure and investigate gene-environment interaction.

Background:

ALS is a progressive neurodegenerative disease affecting motor neurons in the brain and spinal cord; symptoms include weakness and exaggerated reflexes. Patients generally die within two to five years of diagnosis. ALS likely has a multifactorial etiology, with both environmental exposures and genetic susceptibility playing a role. Exposures proposed to increase risk of ALS include pesticides, heavy metals, and electromagnetic fields.

Design:

The Agricultural Health Study (AHS) is a cohort of ~90,000 licensed pesticide applicators and their spouses in Iowa and North Carolina. Participants were first enrolled in 1993-97 by completing questionnaires, and two follow-up interviews have been conducted at five year intervals. These contacts collected information on demographics, lifestyle, medical history, and lifetime pesticide exposure. Based on self-report or death certificates, there are ~60 AHS participants who may have ALS. We will contact these individuals or their proxies to request (1) permission to contact their physicians to obtain medical records and (2) a saliva sample from living cases to provide DNA. We will also attempt to retrieve buccal cell samples previously donated by deceased cases. Physicians of participants who consent will be requested to provide medical records.

Outcome parameters:

A study neurologist will make final ALS diagnoses based on available information. We will assess exposure using existing AHS data. Genotyping will focus on genes implicated in ALS or susceptibility to xenobiotic exposures. We will not collect DNA from controls for this study. Instead we will use data from 384 controls from the FAME study, another nested case-control study within the AHS with exposure and genotype data.

Data analysis:

Pesticide exposure analyses will compare the valid cases to the remaining AHS cohort using Poisson or Cox proportional hazards regression; genotype analyses will compare the valid cases to the FAME controls using logistic regression. Assuming that we identify 40 actual ALS cases in the cohort, we will have 80% power (alpha=0.05) to detect relative risks of 2 to 5, depending on exposure and genotype prevalence. Thus this study has limited power. It is nevertheless worthwhile because it is the only study able to address the role of pesticides in ALS with detailed, prospectively collected exposure data. Further, additional cases will accrue as the AHS cohort ages, permitting future studies with greater power.