#167 - Gary Taubes: Bad science and challenging the conventional wisdom of obesity
Gary Taubes, an investigative science and health journalist, discusses his journey from physics to journalism, developing skepticism for "pathologic science." He shares stories behind his books on cold fusion and the W/Z bosons, emphasizing rigorous background analysis. He then details his pivot to public health, nutrition, and obesity, challenging conventional wisdom.
Deep Dive Analysis
19 Topic Outline
Gary Taubes' Background and Entry into Science Journalism
Developing Skepticism and Early Experiences with 'Bad Science'
Boxing Experiences and Understanding Behavioral Risk
Learning the Craft of Science Writing and Identifying Great Authors
Investigating 'Bad Science': The Nobel Dreams Story and Carlo Rubbia
Theoretical vs. Experimental Physicists and the Scientific Process
Pathological Science: Unconscious Bias and Premature Commitments
The Cold Fusion Debacle: A Case Study in Scientific Self-Deception
Transition to Public Health and Nutrition Journalism
Critiquing Epidemiology and the Power Line Cancer Controversy
Gary's First Foray into Nutrition: The Salt Controversy
Investigating Dietary Fat and the Origins of the Low-Fat Dogma
The New York Times Magazine Article and the Obesity Epidemic Hypothesis
The Carbohydrate-Insulin Model vs. Energy Balance Hypothesis
History of Obesity Research: From Hormonal Dysregulation to Caloric Imbalance
Mutant Mice Experiments and Competing Obesity Theories
Challenges in Addressing Obesity and Diabetes Epidemics
Difficulty of Good Nutrition Studies and Reconciling Findings
Progress and Future of Dietary Interventions for Obesity and Diabetes
8 Key Concepts
Pathological Science
This refers to research tainted by unconscious bias or subjective effects, where scientists fool themselves into believing a result based on premature evidence. It's characterized by public commitment to a result before it's ironclad, and a tendency to collect supporting evidence while ignoring or inadequately studying background noise or alternative explanations.
Signal-to-Noise Problem
A fundamental challenge in science where researchers must distinguish a true discovery (signal) from background variations or experimental flaws (noise). It requires predicting and understanding the noise almost perfectly, making room for 'unknown unknowns' that could be fooling the experimenters.
Scientific Method (Feynman's Principle)
Richard Feynman's principle emphasizes that good science requires bending over backward to be honest about what is known and unknown, and rigorously testing hypotheses by actively trying to prove oneself wrong. The easiest person to fool is oneself, so the initial assumption should be that one's equipment or interpretation is flawed.
Correlation vs. Causation
The critical distinction that an observed association between two phenomena does not inherently imply one causes the other. Establishing causality requires rigorous experimental testing and ruling out all alternative explanations, which is often difficult or impossible in observational fields like epidemiology.
Energy Balance Hypothesis (Obesity)
The conventional wisdom that obesity is primarily caused by taking in more calories than one expends. Gary Taubes argues this is a 'meaningless statement' and a flawed paradigm that overlooks the hormonal regulation of fat storage and fuel partitioning.
Carbohydrate-Insulin Model (Obesity)
An alternative hypothesis suggesting that chronic diseases associated with modern diets, including obesity, are primarily driven by the carbohydrate content of the diet, particularly through its impact on insulin, which regulates fat storage and fuel partitioning.
Hyperphagic Obesity
An observation from animal models where lesioning the ventromedial hypothalamus (VMH) in rats causes them to become extremely hungry and obese. This was conventionally interpreted as overeating causing obesity, but an alternative view suggests the lesion dysregulates fat accumulation, leading to hunger as a *response* to energy being shunted into fat tissue.
Intermediary Metabolism
The science of what the body does with proteins, fats, and carbohydrates after they are eaten, and how the body continues to make fuel available as necessary. This field, which became well-established by the mid-1960s, was largely ignored by obesity researchers who focused on energy balance.
7 Questions Answered
Learning to write involves constant and relentless criticism, extensive rewriting, and reading widely to understand what good writing looks and feels like. The process requires continuous revision until the work meets a high standard.
Pathological science describes research where scientists unknowingly fool themselves into believing a result, often by prematurely committing to a hypothesis and then selectively seeking evidence to support it. It differs from fraud, which involves knowingly manipulating or fabricating data.
Establishing causality in epidemiology is challenging because true randomization is often impossible, and numerous confounding variables (e.g., socioeconomic status) can explain observed associations. This makes it hard to rule out all alternative explanations and isolate a true causal link.
Early clinical trials comparing low-carbohydrate, high-fat diets (like Atkins) to low-fat, calorie-restricted diets showed that participants on the high-fat diet not only lost more weight but also improved heart disease risk factors, directly contradicting prevailing hypotheses.
The shift occurred largely due to early 20th-century research, particularly Lewis Newburgh's 1930s experiments, which concluded that obese individuals got fat because they ate too much. This view, reinforced by later work on hyperphagic obesity in rats, overshadowed earlier hormonal hypotheses.
It's difficult because researchers often enter a field where a paradigm (e.g., energy balance in obesity) is so intuitively obvious and deeply ingrained by mentors that it's never questioned. Overturning such a fundamental belief system challenges the reputation and work of many established scientists.
The carbohydrate-insulin model has led to the widespread acceptance and recommendation of low-carbohydrate, high-fat diets (ketogenic diets) for conditions like type 2 diabetes and weight control. These diets are now extensively studied for various health conditions, and the previous assumption that they were deadly has been largely refuted.
15 Actionable Insights
1. Prioritize Disproving Your Hypothesis
Rigorously test your hypotheses by actively trying to prove yourself wrong, as self-deception is easy in science. This involves seeking out all possible alternative explanations and flaws in your own work before making definitive claims.
2. Demand Independent Replication
For any scientific claim, especially a significant discovery, demand independent replication by multiple labs. This helps distinguish true signals from background noise or experimental flaws and ensures reliability.
3. Design Robust Experimental Controls
When conducting experiments, ensure you design robust controls that account for all possible confounding factors and potential ways your equipment or methods could fool you. This meticulous background analysis is crucial for accurate results.
4. Guard Against Confirmation Bias
Actively guard against confirmation bias by paying less attention to evidence that confirms your beliefs and more to evidence that challenges them. In leadership roles, foster an environment that welcomes skepticism and dissent to avoid groupthink.
5. Acknowledge Unknown Unknowns
Recognize that in any complex scientific endeavor, there will always be ‘unknown unknowns’ that can influence results. Employ high statistical thresholds and comprehensive background analysis to account for potential unperceived variables.
6. Be Radically Honest About Limitations
When reporting or evaluating scientific research, emphasize and thoroughly discuss all known limitations and potential ways the researchers could have fooled themselves. Ideally, the limitations section should be as extensive as the results.
7. Question Intuitively Obvious Dogma
Be skeptical of scientific paradigms that are accepted as ‘intuitively obvious’ or have become dogma without rigorous experimental testing, especially if their application is failing to solve the problem they address.
8. Consider Alternative Obesity Hypotheses
If struggling with weight or blood sugar control, consider the ‘carbohydrate-insulin model’ of obesity, which posits that obesity is primarily a hormonal regulatory disorder driven by carbohydrate intake and insulin, rather than simply an energy balance problem caused by overeating.
9. Evaluate Dietary Advice Critically
Question long-standing dietary advice (e.g., low-fat, low-salt) if the supporting evidence is based on weak epidemiology or lacks robust randomized controlled trials. Be open to new evidence that refutes established dietary hypotheses.
10. Experiment with Low-Carb/Keto Diets
If you struggle with weight or blood sugar control, consider trying a low-carbohydrate or ketogenic diet. These diets are now widely accepted as safe and effective for many conditions, including type 2 diabetes, and are supported by a growing body of research.
11. Demand Real-World Proof
Maintain a high degree of skepticism for extraordinary scientific claims until they demonstrate practical, real-world applications or widespread, undeniable evidence. This ensures that claims are not just theoretical but have tangible impact.
12. Seek Relentless Criticism
To improve skills like writing or scientific thinking, actively seek out and appreciate constant, relentless, and critical feedback on your work. This rigorous review process is essential for personal and professional growth.
13. Practice Iterative Writing
Adopt a writing process that starts with a ‘vomit out’ first draft to get all ideas down, followed by continuous and rigorous rewriting and editing. This iterative approach helps refine prose until it is clear, compelling, and meets high standards.
14. Acknowledge Personal Limitations
Recognize and acknowledge your personal limitations, especially in high-risk activities, as a form of ‘hubris protection.’ This self-awareness can guide decisions to avoid activities that pose significant threats to long-term health.
15. Evaluate Research Funding Mechanisms
Critically evaluate research funding mechanisms, as current systems often favor ’normal science’ that fits existing paradigms over funding for paradigm shifts or the questioning of fundamental assumptions.
7 Key Quotes
You must not fool yourself and you're the easiest person to fool.
Richard Feynman (quoted by Gary Taubes)
When somebody gets a negative result, when they're unable to determine how the shroud was created, does that mean that the shroud was created by supernatural means that are beyond the ability of the equipment to detect, or does it mean that the equipment that they used are simply inadequate for doing the job?
Gary Taubes
Nothing makes your girlfriend look more attractive than your best friend trying to steal her from you.
Chase Peterson (quoted by Gary Taubes)
I am never going to be able to change my mind and accept... I'm going to be crawling around the outside of that perimeter fence with you know a beard and pieces of dunkin donuts in the beard and a tattered copy of bad science looking for where the damn thing is plugged in because I know it's a con.
Gary Taubes
I think the problem here is that when you entered the field there was a paradigm, a way of thinking about obesity that seemed so intuitively obvious... that you never questioned it.
Gary Taubes
The conventional wisdom fails. It's clear it fails because we have obesity and diabetes epidemics that haven't been stemmed in any way. We have to question our assumptions.
Gary Taubes
This is almost as impressive as the Apollo 11 program from an engineering perspective, like the way they made this thing, the crunch, the taste, the lingering flavor, it's unbelievable.
Gary Taubes