In Part 1 of this series, we explored how the Daubert standard makes it difficult for plaintiffs to prove nutrition-related claims in court. But this tension between correlation and causation reaches beyond the courtroom. In the U.S., it shapes how federal agencies evaluate scientific evidence and limits how aggressively they can regulate food products like ultra-processed foods (UPFs).

A Critical Distinction

Public health research often shows that certain behaviors or exposures are correlated with specific health outcomes. For example, observational studies might reveal that people who consume more UPFs tend to experience worse health over time. But that is correlation—a statistical relationship—not proof that UPFs cause those outcomes. Many other factors, including income, education, and lifestyle choices, could be influencing both diet and health.

Causation, by contrast, means one thing directly leads to another. Proving this requires more than statistical association; it often demands experimental evidence, like a randomized controlled trial, to rule out alternative explanations. In legal and regulatory contexts, this distinction is critical. Policies that restrict products or mandate disclosures must be defensible not just scientifically but procedurally. Regulators must demonstrate that actions are based on credible evidence—not assumptions.

How This Plays Out in U.S. Food Regulation

The U.S. Food and Drug Administration (FDA) faces this challenge directly. While the agency can act on emerging risks, its actions must hold up under judicial review and public scrutiny. That creates a high bar for action when the evidence is mostly correlational.

Take the FDA’s 2024 update to the definition of “healthy.” The rule tightens criteria for nutrients like sodium and added sugar but does not incorporate food processing levels, despite calls to address UPFs more directly (FDA, 2024). One reason: there is no officially recognized legal or scientific definition of what qualifies as an ultra-processed food. While systems like NOVA are widely used in research, they lack consensus across disciplines and are not formally adopted by U.S. agencies (Monteiro et al., 2019).

Without a clear, enforceable definition, the FDA risks regulatory overreach if it imposes restrictions based on a concept that lacks legal clarity. At the same time, avoiding the issue altogether leaves a gap in how emerging food categories are handled.

Why Regulatory Inaction Is So Persistent

This evidentiary tension often results in what experts call “regulatory inertia.” Agencies are hesitant to act without the kind of causal proof that courts and stakeholders demand. Even strong patterns in population data may not be enough to justify broad new rules. And absent consensus on definitions, regulators struggle to categorize products consistently or enforce standards reliably.

The result is a system in which science progresses faster than regulation. Researchers may find compelling reasons to suspect harm or risk based on statistical associations, but agencies are constrained in what they can do with that information. In the case of food policy, this has left regulators focused on nutrients and ingredients—where definitions and thresholds are clear—rather than food categories or manufacturing techniques that are more complex to define and measure.

Bridging the Evidentiary Divide

U.S. regulators are caught in a bind. Acting too early on correlational evidence invites criticism and legal challenge. Acting too late risks public health consequences and erodes public trust. A more structured and transparent framework for weighing evidence—short of proving causation—can help resolve that tension. The goal is not to lower scientific standards, but to align them with the realities of policymaking. Only then can food regulation respond effectively to complex, evolving risks.

References

U.S. Food and Drug Administration (FDA). 2024. Food labeling: Nutrient content claims; definition of term “healthy.” Federal Register. 89(249):93629–93659.

 Monteiro, C. A., Cannon, G., Moubarac, J.-C., Levy, R. B., Louzada, M. L., & Jaime, P. C. (2019). Ultra-processed foods: What they are and how to identify them. Public Health Nutrition, 22(5), 936–941.