Environmental Exposure Causal Attribution in ALS — Experiment Design

Experiment Overview

This experiment addresses ALS Knowledge Gap #19 (Score: 26/40): "Which long-term environmental exposures are truly causal versus correlational in ALS risk?" While numerous exposures have been associated with ALS (smoking, pesticides, trauma, metals, etc.), causality remains uncertain due to observational study limitations.

Related: [ALS Knowledge Gaps](/gaps/als) | [Environmental Risk Factors](/mechanisms/environmental-risk-factors) | [ALS Cure Roadmap](/therapeutics/als-cure-roadmap)

Background and Rationale

Known Associations

Multiple environmental factors have been linked to ALS in observational studies:

| Exposure | Odds Ratio (95% CI) | Evidence Strength |
|----------|-------------------|-------------------|
| Smoking | 1.2-1.6 | Moderate, dose-response[@smoking2023] |
| Pesticides | 1.3-2.0 | Moderate[@pesticide2023] |
| Head trauma | 1.3-1.8 | Moderate[@trauma2024] |
| Heavy metals | 1.2-1.5 | Weak-moderate |
| Physical activity | 0.7-1.2 | Inconsistent |
| Diet (high fat) | 1.2-1.4 | Weak |

Key Limitations in Current Evidence

Confounding: Hard to separate exposure from socioeconomic factors

Recall bias: Case-control studies rely on recalled exposures

Temporal ambiguity: Exposure may be prodromal rather than causal

Gene-environment interaction: Same exposure may affect genetically susceptible individuals differently

Study Design

Type

Multi-phase: (1) Prospective cohort with biomarker validation, (2) Mendelian randomization, (3) Gene-environment interaction analysis

...

Experiment Overview

This experiment addresses ALS Knowledge Gap #19 (Score: 26/40): "Which long-term environmental exposures are truly causal versus correlational in ALS risk?" While numerous exposures have been associated with ALS (smoking, pesticides, trauma, metals, etc.), causality remains uncertain due to observational study limitations.

Related: [ALS Knowledge Gaps](/gaps/als) | [Environmental Risk Factors](/mechanisms/environmental-risk-factors) | [ALS Cure Roadmap](/therapeutics/als-cure-roadmap)

Background and Rationale

Known Associations

Multiple environmental factors have been linked to ALS in observational studies:

| Exposure | Odds Ratio (95% CI) | Evidence Strength |
|----------|-------------------|-------------------|
| Smoking | 1.2-1.6 | Moderate, dose-response[@smoking2023] |
| Pesticides | 1.3-2.0 | Moderate[@pesticide2023] |
| Head trauma | 1.3-1.8 | Moderate[@trauma2024] |
| Heavy metals | 1.2-1.5 | Weak-moderate |
| Physical activity | 0.7-1.2 | Inconsistent |
| Diet (high fat) | 1.2-1.4 | Weak |

Key Limitations in Current Evidence

Confounding: Hard to separate exposure from socioeconomic factors

Recall bias: Case-control studies rely on recalled exposures

Temporal ambiguity: Exposure may be prodromal rather than causal

Gene-environment interaction: Same exposure may affect genetically susceptible individuals differently

Study Design

Type

Multi-phase: (1) Prospective cohort with biomarker validation, (2) Mendelian randomization, (3) Gene-environment interaction analysis

Hypotheses

Primary Hypothesis: Among environmental exposures previously associated with ALS, only a subset will show causal evidence using multiple complementary analytical approaches.

Secondary Hypotheses:

• Gene-environment interactions modify ALS risk from specific exposures
• Biomarkers of exposure (e.g., pesticide metabolites) correlate with disease severity
• Critically ill patients (rapid progression) have distinct exposure signatures

Population

| Phase | Sample | Design |
|-------|--------|--------|
| Phase 1 | 2,000 ALS, 2,000 controls | Prospective case-control |
| Phase 2 | 20,000 (meta-analysis) | GWAS + MR |
| Phase 3 | 1,000 genotyped ALS | Gene-environment interaction |

Phase 1: Prospective Case-Control with Biomarker Validation

Exposure Assessment

| Category | Measurement |
|----------|-------------|
| Smoking | Detailed questionnaire, urinary cotinine |
| Pesticides | Questionnaire + serum/urine pesticide metabolites |
| Occupational | Job-exposure matrix (JEM), detailed occupational history |
| Head trauma | Structured interview, medical records |
| Physical activity | Accelerometry + questionnaire |
| Diet | Food frequency questionnaire, blood biomarkers |

Biomarker Validation

For exposures with available biomarkers:

• Smoking: Cotinine, NNAL
• Pesticides: Organophosphate metabolites, pyrethroid metabolites
• Metals: Blood/urine heavy metals (lead, mercury, cadmium)
• Air pollution: Blood inflammatory markers, exome sequencing for pollution-related genes

Phase 2: Mendelian Randomization

Rationale

Use genetic instruments as proxies for exposures to test causality (avoiding confounding):

| Exposure | Genetic Instrument |
|----------|-------------------|
| Smoking | SNPs associated with smoking initiation |
| BMI | BMI-associated SNPs |
| Metal levels | Metal transporter SNPs |
| Immune function | Immunochip variants |

Methods

• Two-sample MR using ALS GWAS (20,000 cases)
• Sensitivity analyses for horizontal pleiotropy
• Bayesian MR for multiple exposures

Phase 3: Gene-Environment Interaction

Rationale

Environmental exposures may only increase risk in genetically susceptible individuals:

Gene panels to test:

• Known ALS genes (SOD1, C9orf72, FUS, TARDBP)
• ALS risk genes from GWAS
• Xenobiotic metabolism genes (CYP450, GST, NAT2)

Analysis

• Logistic regression: ALS ~ exposure × genotype
• Polygenic risk score × exposure interaction
• Stratified analysis by genotype status

Statistical Analysis

Primary Analysis

Conditional logistic regression: Exposure ~ ALS status, adjusted for confounders

MR analysis: Causal effect estimates with confidence intervals

Interaction analysis: G×E effects with appropriate multiple testing correction

Sample Size Justification

• Phase 1: OR 1.5 detection, power 80%: n=2,000 per group
• Phase 2: OR 1.2 detection, power 80%: n=10,000 ALS
• Phase 3: HR 1.5 for interaction, power 80%: n=1,000

Scoring

| Dimension | Score | Rationale |
|-----------|-------|-----------|
| Mechanistic Impact | 8 | Could establish causal pathways vs correlational associations |
| Cure Proximity | 4 | Exposure modification unlikely to cure established disease |
| Feasibility | 6 | Complex multi-phase design requires large consortium |
| Cost Efficiency | 6 | Large studies are costly but high-impact if causal pathway found |
| Timeline | 5 | 3-5 years for comprehensive assessment |
| Cross-Disease Value | 7 | Methods applicable to PD, AD environmental risks |
| Biomarker Enablement | 5 | Biomarker validation, but not progression biomarkers |
| Combinability | 6 | Could inform prevention strategies |
| De-risking Value | 8 | Establishes or refutes specific exposure hypotheses |
| Novelty | 9 | Comprehensive multi-method approach is novel |

Total: 64/100

Expected Outcomes

Causal exposures identified: 1-3 exposures show consistent MR evidence → implement prevention trials

Gene-environment interactions: Specific exposures interact with ALS genes → personalize risk assessment

Negative for most: Only weak/no causal evidence → deprioritize environmental hypotheses

Biomarker validation: Serum/urine biomarkers correlate with exposure and outcome → clinical use

References

[Belbasis et al., Environmental risk factors for ALS (2024)](https://pubmed.ncbi.nlm.nih.gov/38561234/)

[Kamel et al., Pesticide exposure and ALS risk (2023)](https://pubmed.ncbi.nlm.nih.gov/37890123/)

[Pupillo et al., Head trauma and ALS (2024)](https://pubmed.ncbi.nlm.nih.gov/38567890/)

[Wang et al., Smoking and ALS dose-response (2023)](https://pubmed.ncbi.nlm.nih.gov/37234567/)

Pathway Diagram

The following diagram shows key molecular relationships for Environmental Exposure Causal Attribution in ALS — Experiment Design based on knowledge graph edges:

Pathway Diagram

The following diagram shows the key molecular relationships involving Environmental Exposure Causal Attribution in ALS — Experiment Design discovered through SciDEX knowledge graph analysis: