Exposure examples¶
Six examples show how the same residential history can support different research questions. The study ideas below are proposals. The linked papers describe published findings, including null results and unresolved mechanisms. Use the data source guide for access links and practical limits.
Lead¶
Published evidence. A case-control study of US veterans found higher blood lead among ALS cases than controls. Blood sampling after disease onset complicates interpretation because illness and bone turnover can influence circulating lead. A separate, smaller study associated higher bone lead with longer survival among people with ALS. These findings concern different outcomes and do not establish that lead is protective. R1, R2
Proposed examination. Combine dated residences with documented lead sources, service-line information and site-specific environmental measurements where available. Analyze these sources separately before considering a combined index. A current service-line inventory cannot, by itself, reconstruct the pipe material or water concentration decades earlier.
What would strengthen it. Add occupational and hobby histories, household water sampling and appropriately timed blood lead measurements. Retain information about bone turnover when relevant. These data help evaluate whether the geographic estimate corresponds to exposure in the person.
Mechanistic follow-up. If an association is replicated, test lead-related cellular injury in an experimental system at justified concentrations. A reduction in a laboratory injury marker would be a lead for further research; it would not establish a treatment effect in ALS.
Pesticides and herbicides¶
Published evidence. The Agricultural Health Study examined specific pesticides using participant-reported use and ALS mortality. Its compound-specific estimates were imprecise and were not statistically significant. The broader literature and the cohort component did not give identical results. This supports chemical-specific investigation without treating every pesticide as an established ALS cause. R3
Proposed examination. Select an active ingredient or justified class, specify the exposure window, and link addresses and work locations to dated application records when available. Use USGS estimates for regional context, respecting their limited reliability at small geographic scales. A crop map alone does not show whether a particular pesticide was applied.
What would strengthen it. Collect product names, application tasks, dates, frequency, protective equipment and occupational history. A job-exposure matrix can provide a standardized estimate, but it still needs uncertainty recorded. Match any biomarker assay and collection time to the compound's kinetics.
Mechanistic follow-up. Select assays based on the specific chemical. An insecticide, herbicide and fungicide need not share a mechanism. Prespecify any gene-exposure interaction and account for multiple testing; candidate genes are hypotheses, not a validated susceptibility panel.
Air pollution¶
Published evidence. A University of Michigan ALS cohort linked modeled residential pollution to repeated ALSFRS-R assessments and reported associations between some pollutants and faster decline. A large UK Biobank study found no association between the evaluated pollutants and incident ALS. Progression and incidence are different endpoints; neither study establishes that pollution reduction changes ALS outcomes. R4, R5
Proposed examination. Link every residential period to pollutant-specific estimates at a supported time scale. For progression, examine repeated ALSFRS-R values with an exposure-by-time term in a longitudinal model. For survival, define the time origin, account for delayed entry where needed and use only exposure information available before each outcome time.
What would strengthen it. Add smoking, occupation, indoor sources, socioeconomic context, residential moves and access to care. Examine correlated pollutants and geographic clustering. Monitor observations can help evaluate a model, but the nearest monitor may poorly represent a rural residence.
Mechanistic follow-up. Repeated metabolomic or proteomic measurements could test prespecified biological responses. Such responses may be affected by ALS itself, medications or nutrition. Their presence does not identify which pollutant produced them.
Solvents and industrial sites¶
Published evidence. A small ALS case-control study linked geocoded residences to modeled hazardous air pollutants and reported an association with an aromatic-solvent group. Its estimates were uncertain, and the grouping did not isolate one causal chemical. Evidence from another neurodegenerative disease should retain that disease label when informing ALS hypotheses. R6
Proposed examination. Begin with a named chemical and route. Link dated industrial releases or contaminated-site records to the residence period. Use an appropriate dispersion model, documented groundwater plume or measured contamination when available. Simple distance to a facility is a screening variable; it does not describe what reached a person.
What would strengthen it. Add facility operating dates, release medium, remediation history, water-source information and job tasks. Groundwater, drinking water and vapor intrusion are different pathways. A Superfund listing identifies a site requiring investigation, not a single exposure shared by all nearby residents.
Mechanistic follow-up. Measure candidate compounds or suitable metabolites where feasible, then test specific chemicals or justified mixtures experimentally. Keep the original chemical identities visible when comparing responses across exposures.
Cyanobacterial blooms, cyanotoxins and BMAA¶
Published evidence. A spatial study in northern New England examined satellite-derived lake phycocyanin and ALS geography. Its findings were hypothesis-generating and depended on model choices. This type of map estimates bloom-related environmental conditions, not an individual's toxin dose. R7
Proposed examination. Link residence periods and reported recreation sites to water bodies with usable satellite observations. Define the bloom metric and report observation coverage. Use only years supported by the selected sensor and product; do not assume a continuous 20-year daily record. A missing or cloud-obscured observation is not a bloom-free day.
What would strengthen it. Collect contact dates and activities, drinking-water provider and source, treatment information, and diet where it supports a plausible pathway. Connect public-water users to the relevant supply rather than assuming they drink from the nearest lake. Use toxin-specific sampling to distinguish an environmental bloom from a measured chemical exposure. EPA explains that satellites cannot detect toxins. Data sources
Mechanistic follow-up. Keep the bloom event, cyanobacterial organisms, microcystins and BMAA separate. The proposed incorporation of BMAA into proteins is disputed: one cell study supported it, while another did not detect it after protein purification. Both belong in the evidence record. Neither the mechanism nor a residential bloom association establishes a benefit from an ALS treatment. R8, R9
Occupational formaldehyde¶
Published evidence. A large prospective study reported a formaldehyde signal that was stronger in analyses with exposure-duration information. A cohort of exposed garment workers did not support an association with ALS mortality. These results justify attention to how exposure and occupation were measured. R10, R11
Proposed examination. Start with dated jobs and tasks, then use a suitable job-exposure matrix or industrial hygiene records. A home-address air layer can provide background context, but cannot substitute for workplace measurements of task-related exposure.
What would strengthen it. Record task frequency, ventilation, protective equipment, worksite location, calendar era and co-exposures. A job title alone should not assign every worker the same dose. Keep disease occurrence and mortality endpoints distinct when comparing studies.
Mechanistic follow-up. Test a prespecified cellular response using concentrations and exposure durations supported by occupational measurements. Laboratory effects would still need validation in human studies before supporting an ALS-specific prevention or treatment claim.
From an exposure hypothesis to an intervention¶
The examples can guide exposure validation, biological sampling and experimental design. To inform drug development, a proposed mechanism should survive replication, experiments that perturb the pathway, and testing of whether an intervention changes meaningful disease outcomes. An adverse outcome pathway developed for Parkinson disease or another outcome cannot be relabeled as an ALS pathway without evidence for the connecting steps.
Similarly, an observational association with ALS progression is not evidence that reducing that exposure slows ALS. An intervention study would need to test that question. General environmental health protections and disease-specific efficacy are separate evidence questions.