Health Information Decoded

In the 1960s, the Sugar Research Foundation paid three Harvard scientists to publish a review in the New England Journal of Medicine. The review systematically downplayed the connection between sugar and heart disease and redirected blame toward dietary fat. The payment was not disclosed. The review shaped nutritional policy for the next five decades. Millions of people avoided fat, replaced it with sugar-laden products, and got sicker. The funding was not uncovered until 2016, when a researcher at the University of California, San Francisco named Cristin Kearns found the industry's internal documents. For half a century, the foundational nutritional science that governments used to create dietary guidelines was shaped by an industry that knew its product caused harm and paid to redirect attention elsewhere. This is not an anecdote. It is a case study in how the health information ecosystem actually operates.

The Problem

Health recommendations flip-flop. Eggs are dangerous, then they are fine. Fat is the enemy, then sugar is the enemy. Butter is poison, then butter is back. Red wine extends life, then it does not. Every claim arrives with "studies show" attached. The studies contradict. The experts disagree publicly. The ordinary person trying to make reasonable choices about what to eat, what to avoid, and how to live is left paralyzed by contradictory information from sources that all claim authority.

This paralysis is not noise. It is signal. The contradictions are not random—they reveal the structure of corruption acting on the information system. When we decode that structure, we can begin to separate what is true from what is merely well-funded.

The Corruption Stack (Health Edition)

The same seven-layer corruption structure that distorts other institutional domains operates in health with particular force. Each layer compounds the others, creating an information environment where truth must survive a gauntlet of distortions before reaching anyone who might act on it.

Layer 1: Funding Bias. Who pays determines what gets studied. Patentable molecules attract research funding because they represent potential revenue. Natural interventions—sleep, sunlight, fasting, whole foods—generate minimal research investment because they cannot be patented and sold. The result is not that natural interventions are proven ineffective. The result is that they are studied less, and the absence of studies is then cited as absence of evidence. In other words, the funding doesn't just bias what we find. It biases what we look for.

Layer 2: Regulatory Capture. The FDA, the agency tasked with protecting the public from unsafe food and drugs, operates through a revolving door with the industries it regulates. FDA officials move to pharmaceutical and food industry positions. Industry executives move to FDA leadership roles. Career incentives within the agency reward not upsetting powerful interests, because upsetting powerful interests ends careers while enabling them opens doors to lucrative post-government employment.

Layer 3: Publication Bias. Scientific journals strongly prefer positive results—studies that found something over studies that found nothing. Negative trials (studies showing a drug or supplement had no effect) sit in file drawers, unpublished. Meta-analyses (studies of studies) then aggregate from a biased sample, because the null results that would moderate the conclusions were never published. The published literature systematically overstates effect sizes as a direct consequence.

Layer 4: Complexity Hiding. Human metabolism is a complex adaptive system with thousands of interacting variables. Studies typically test single variables in isolation: "Does X cause Y?" But X interacts with Z, A, B, and C, and the interaction depends on genetic background, gut microbiome, sleep quality, stress levels, and dozens of other factors that most studies do not measure. Reducing a complex system to isolated variable testing produces findings that are technically accurate within the study conditions but misleading when applied to real-world complexity.

Layer 5: Time Horizon Mismatch. Most clinical studies run for six weeks to two years. Most chronic diseases develop over decades. Short-term studies simply cannot detect long-term harms. A substance can produce impressive six-month results while causing damage that manifests twenty years later. The time horizon of the evidence does not match the time horizon of the harm, and that mismatch is structural, not accidental—long studies are expensive, and the entities funding research want results fast.

Layer 6: Economic Misalignment. Sick people are profitable. Healthy people are not. The healthcare system, as an economic entity, optimizes for treatment rather than prevention. Chronic disease management generates recurring revenue. Cures and prevention do not. This misalignment does not require conspiracy. It requires only that economic incentives shape what gets funded, promoted, and practiced—which they do, reliably, across every industry.

Layer 7: Information Warfare. Conflicting claims paralyze decision-making. Paralysis defaults to the status quo. In health, the status quo means continuing to consume products made by the industries generating the conflicting claims. The confusion itself serves incumbent interests. Tobacco companies pioneered this strategy—not by proving cigarettes were safe, but by creating enough doubt and contradictory studies that smokers felt unable to reach a clear conclusion. The technique has been widely adopted.

Evidence the Stack Is Operating

If errors in health information were random, their directionality would be random—sometimes favoring industry, sometimes opposing it, in roughly equal measure. But the errors consistently favor industry interests. The sugar industry funded research blaming fat—internal documents prove they knew the science was misleading. Tobacco companies funded doubt about the cancer link for decades while possessing internal research confirming it. Opioid manufacturers like Purdue Pharma promoted addiction-risk denialism through industry-funded studies. The USDA food pyramid was shaped by agricultural lobbying rather than nutritional science, as subsequent revisions implicitly acknowledged.

Non-random directionality is the signature of non-innocent error. When mistakes consistently favor the party funding the research, the pattern is not bad luck. It is structural corruption.

What Survives the Corruption Filter

Not everything is corrupted. Some health claims survive every layer of the stack because they share specific features that make them resistant to distortion. Look for claims that meet these criteria:

No profitable stakeholder. Nobody gets rich if the claim is widely believed. Sleep is good for you. No one can patent sleep. Movement is good for you. No one owns a trademark on walking. When there is no financial incentive to promote a finding, the finding is far more likely to reflect reality.

Multiple independent evidence paths. Different research methods, different disciplines, different funding sources all reach the same conclusion. Single studies are weak. Convergence across independent paths is very difficult to manufacture.

Evolutionary coherence. The claim fits what humans did for millions of years. Our biology was shaped by specific conditions. Practices and substances consistent with those conditions carry lower uncertainty than novel introductions.

Mechanism clarity. A plausible biological pathway connects the claimed cause to the claimed effect. We can explain how it works, not just that it correlates.

Long-term population data. Traditional populations that practiced what the claim recommends show corresponding health outcomes. This is real-world evidence accumulated over centuries, outside the reach of modern funding bias.

High Confidence

Processed food is worse than whole food. Evolutionary fit (humans ate whole foods for millions of years), population data (traditional diets outperform processed diets in every comparison), mechanism clear (ultra-processing destroys micronutrients and adds inflammatory compounds), consistent replication across studies and populations. No industry profits from this claim—in fact, the processed food industry actively contests it.

Added sugar at modern consumption levels is harmful. Multiple independent research lines converge. Mechanism is clear at the metabolic level (insulin resistance, liver fat accumulation, inflammatory cascades). Natural experiment confirms: populations that adopt Western sugar consumption develop Western metabolic diseases. The sugar industry's fifty-year campaign to suppress this finding is itself evidence of its truth.

Movement is good for health. Universal across studies. No funding bias (no one profits from recommending walking). Evolutionary fit (humans evolved as endurance movers). Mechanism clear across cardiovascular, metabolic, neurological, and musculoskeletal systems.

Sleep deprivation is harmful. No funding bias. Universal findings across every population studied. Mechanism clear at molecular, cellular, and systems levels. Maximum confidence.

Chronic stress is harmful. Cortisol and HPA axis mechanisms well-understood. Universal findings. Replication across cultures and methodologies. No industry profits from claiming stress is harmful.

Medium Confidence

Seed oils are suspect. Novel in evolutionary terms—industrial extraction has existed for roughly a century. Oxidation chemistry is concerning (polyunsaturated fats oxidize readily under heat). But there are fewer independent evidence paths than the online discourse suggests, and the corruption cuts both ways (industry defends them, but the anti-seed-oil movement also has commercial interests).

Specific supplements. Some have strong evidence for specific conditions (vitamin D in deficient populations, magnesium for those with documented low intake). Many are overhyped by an industry that profits from their sale. The supplement industry is itself a corruption vector, not a corruption-free zone.

Low Confidence (Likely Corrupted)

Anything with a history of flip-flopping recommendations. Anything where pharmaceutical or food companies have strong financial stakes in the conclusion. Anything that contradicts evolutionary heuristics without presenting a compelling mechanism for why our biology would respond differently than our evolutionary history predicts.

A Practical Framework

Default to ancestral heuristics. Whole foods, regular movement, adequate sleep, sunlight exposure, social connection. These practices have millions of years of supporting evidence and zero industry lobbying behind them.

Follow the money. For every health claim, ask: who funded this study? Who profits if this recommendation is widely adopted? The answer does not automatically invalidate the finding, but it determines how much scrutiny the finding deserves.

Demand convergence. Require mechanism, replication, and population data before treating any claim as established. A single study is a data point, not a conclusion.

Trust careful self-experimentation. N-of-1 experiments with proper controls (systematic tracking, one variable at a time, adequate time between changes) generate data that applies to us specifically—something population-level studies cannot do.

Accept irreducible uncertainty. Some things we cannot currently know. Sitting with uncertainty is more honest—and ultimately healthier—than adopting false certainty from a corrupted information source.

How This Was Decoded

Applied the DECODER corruption stack framework to the health information domain. Pattern recognition across documented cases: sugar industry manipulation (Kearns et al., 2016), tobacco science (Proctor, Golden Holocaust, 2011), opioid crisis (Keefe, Empire of Pain, 2021), FDA regulatory capture (documented revolving door patterns). Four independent inference paths converge on systematic, directional bias in health information. Counterpath tested: random complexity alone does not explain the consistent industry-favoring directionality of information errors. Applied incentive divergence, Goodhart's Law, and convergent confidence principles from the DECODER framework.