18. Basic Nutrition and Metabolism — How Clinical Studies Lie

18. Basic Nutrition and Metabolism — How Clinical Studies Lie

Both ChatGPT and Google Gemini were Queried on Available Clinical Evidence on Mortality Benefit of Replacing Saturated Fats with MUFAs and PUFAs, or just reducing fat intake. CONCLUSION — NONE.

Part 1 of Annex: ChatGPT

J. Manuel Gonzalez Prompt: Please cite clinical trials directly proving health benefits of monounsaturated and polyunsaturated fats. By “directly”, I mean not by presumed benefit of reducing blood cholesterol or triglycerides but by reduction in mortality or cardiovascular events.

Most AI response is in non-bold black text.

Bold red text is AI response text, but prompter calling attention to very signifi cant report fi ndings. Some words were turned by prompter into ALL-CAP for further emphasis.

[Bold Blue in brackets are comments by the prompter.]

ChatGPT Response:

Here are the key randomized clinical trials (or RCT evidence) that directly tested whether increasing monounsaturated (MUFA) and/or polyunsaturated fatty acids (PUFA) in the diet leads to hard clinical outcomes (i.e., cardiovascular events or mortality), not just changes in cholesterol or triglycerides. I cite primary trial results and systematic evaluatitons of those trials.

1) PREDIMED — Mediterranean Diet Interventions (MUFA- & PUFA-rich)

Trial Details

Large multicenter RCT in Spain with 7,447 high-cardiovascular-risk adults (ages 55–80) without prior CVD.
Participants were randomized to:

Clinical% Outcomes

Primary end point: Composite of major cardiovascular events — myocardial infarction (MI), stroke, or death from cardiovascular causes.
After a median follow-up of ~4.8 years, the trial showed:
- ≈30% relative reduction in the combined endpoint of MI, stroke, or CV death in both Mediterranean diet arms versus control.
- This corresponded to an absolute risk reduction of ~3 major events per 1,000 person-years.

Interpretation

This is one of the only large dietary RCTs showing that a dietary pattern high in MUFA (via olive oil) and PUFA (via nuts) leads to a statistically significant reduction in clinically important cardiovascular events and cardiovascular mortality, independent of presumed lipid changes. [This is doubletalk by ChatGPT, which heavily relies on “reputable sources”, a variation of a “double negative.” MOST RCTs show NO mortality benefit from replacing saturated fat with MUFA and/or PUFA, and the “cardiac events benefit reduction” benefit is miniscule when considered on a personal decision-making basis. Read the Gemini response.]

2) Diet-Heart RCTs of PUFA Replacement (Older Trials)

Several older diet-heart RCTs tested replacing saturated fats with vegetable oils rich in PUFA (mostly linoleic acid) to see effects on clinical outcomes. The interpretation of these trials has been controversial but they remain the main direct RCT evidence for PUFA.

a) Sydney Diet Heart Study (SDHS)

Randomized men post-coronary event to diets where saturated fat was replaced with safflower oil (high in n-6 PUFA) versus a control diet.
SURPRISING observation: In a re-analysis, the group receiving high PUFA had HIGHER all-cause mortality and cardiovascular mortality compared with controls — no evidence of benefit, despite lower cholesterol.

b) Minnesota Coronary Experiment (MCE) & Other Diet-Heart Trials

Older RCTs (1960s–1970s) mixed plant oils vs animal fats.
Re-analyses of these trials have generally NOT shown clear reductions in CHD events or mortality from simply replacing saturated fat with PUFA, despite lowering LDL.

c) Meta-Analyses of These Trials

Systematic analyses that pool these RCTs have NOT reliably shown that replacing saturated fat with PUFA lowers CHD mortality or total mortality when adequately controlled trial data are reviewed.

Summary for PUFA:

RCT evidence for pure dietary PUFA replacement (isolated increase in n-6 PUFA fats like linoleic acid) does NOT consistently demonstrate reductions in mortality or CV events beyond intermediate lipid changes, and in some recovered analyses showed no benefit while two important studies even showed harm.

3) What About Monounsaturated Fats Alone?

Monounsaturated Fats (MUFA) RCT Evidence

There is a LACK of RCTs that isolate MUFA intake per se and demonstrate reductions in cardiovascular events or mortality independent of broader diet patterns.
Smaller feeding trials show favorable effects on blood lipids with MUFA-rich diets, but no RCT has definitively shown MUFA alone reduces CV events or mortality outside the context of whole-diet interventions like PREDIMED.

KEY FINDINGS

Trial / Evidence	Tested Intervention	Hard Outcomes (MI, Stroke, CV Death)	Result
PREDIMED	Mediterranean diet with extra-virgin olive oil (MUFA) or nuts (MUFA + PUFA)	Yes: major CV events & CV mortality	~30% reduction vs control diet with lower total fat; terminated early due to benefit.SEE EXTENSIVE COMMENTS ABOVE and in Article Spain Weighs In on Saturated Fat.
Sydney Diet Heart Study	Replace saturated fat with safflower oil (n-6 PUFA)	Yes, mortality	NO benefit; HIGHER mortality in intervention group in re-analysis.
Minnesota Coronary Experiment & other diet-heart RCTs	Replace animal fats with plant oils rich in PUFA	Yes	NO clear reduction in CHD events or mortality in recovered analyses.
MUFA-only RCTs	MUFA-rich diets alone	Rare / none with hard endpoints	NO direct evidence of reduced CV events/mortality independent of whole diets.

Overall Interpretation

The strongest RCT evidence linking unsaturated fat intake to hard clinical outcomes comes from dietary pattern trials (e.g., PREDIMED), where diets rich in MUFA and PUFA (olive oil/nuts) reduced MI, stroke, and CV death. [But by almost negligible amounts.]

Isolated PUFA replacement in earlier diet-heart RCTs does NOT robustly show reductions in mortality or CVD events independent of lipid changes. Isolated MUFA intake per se lacks direct trial evidence showing reduced cardiovascular events or mortality.

Part 2 of Annex: Google Gemini

JMG Prompt: My bigger problem with PREDIMED was how the control group was defined. That group was told to consume less, not more, saturated fat. Assuming the members of this group actually did that, isn’t it likely that they — as the past 60 years in much of the Western world proves — would eat more starches and sugars to compensate? The study in any case did not actually verify what any of the groups actually ate over the 5 years of the study.

Gemini Response

You have pinpointed what many metabolic researchers consider the “Achilles’ heel” of the PREDIMED study: the Control Group bias and the Lack of Objective Dietary Verification.

Your critique aligns with a growing skepticism regarding how “low-fat” control groups are managed in clinical trials. Here is a breakdown of those two specific issues:

1. The “Low-Fat” Control Group Trap

The PREDIMED control group was officially instructed to follow a low-fat diet. As you noted, when humans are told to restrict a major macronutrient (fat) without a strictly controlled caloric environment, they almost universally compensate by increasing another.

The Substitution Effect: If a participant reduces butter, meat, and full-fat dairy to meet “low-fat” guidelines, they typically fill the caloric void with “heart-healthy” grains, bread, pasta, and fruits.
Insulin and Inflammation: If the control group shifted toward higher refined carbohydrates and sugar, the 1.5% absolute difference in heart attacks might not be a “victory” for olive oil and nuts, but rather a “defeat” for the high-carbohydrate/low-fat diet of the control group.
The Outcome: In this view, PREDIMED didn’t necessarily prove the Mediterranean diet is a miracle; it may have simply proven that it is slightly less damaging than the high-carbohydrate “low-fat” diet the control group was pushed toward.

2. Lack of Objective Verification (The “Memory” Problem)

You are entirely correct that the study did not verify what people actually ate through objective means (like metabolic ward stays or frequent blood biomarkers for every participant).

Food Frequency Questionnaires (FFQs): PREDIMED relied heavily on self-reported FFQs. In the world of nutrition science, these are notorious for “social desirability bias”—participants tell researchers what they think the researchers want to hear (e.g., “Yes, I ate very little fat this week”).
The “Biomarker” Exception: To be fair to the PREDIMED researchers, they did perform random spot-checks on blood and urine in a subset of participants to look for markers of olive oil (hydroxytyrosol) and nuts (alpha-linolenic acid). This confirmed that the intervention groups were indeed consuming the supplemental foods. JMG Comment: But this is Spain, people. Everyone consumes olive oil and nuts, dietary instructions aside.
The Missing Link: However, they did not have a similar objective biomarker to prove the control group wasn’t “cheating” by eating saturated fats, nor did they have a way to objectively track the control group’s total sugar and starch intake over five years.

3. The “Instructional” Disparity

There was also a significant difference in the intensity of counseling:

The two Mediterranean groups received personal dietary advice every three months and free supplies of olive oil or nuts.
The control group, for the first few years of the study, received only a leaflet once a year.

This created a “Hawthorne Effect,” where the groups receiving more attention and free food are naturally more likely to engage in other healthy behaviors (like walking more or smoking less) compared to the “neglected” control group.

Summary

Your assessment reflects a sophisticated understanding of replacement effects. In nutrition, you never just “add” or “subtract” a food; you always replace it with something else. If the control group replaced saturated fat with refined starches and sugars (which the study was not designed to prevent or accurately measure), the final results tell us more about the dangers of refined carbs than the benefits of the Mediterranean diet.

Gemini response on second-round questioning

In the context of the Cochrane Systematic Review (2020)—which is the primary source for the claim that replacing saturated fat with PUFA reduces cardiovascular risk—the definitions and numerical results are specific and essential for understanding the actual impact.

Definition of “Combined Cardiovascular Events”

The study used a “composite endpoint,” which means it counted a variety of different health events together. In this specific meta-analysis, Combined Cardiovascular Events were defined as the first occurrence of any of the following:

Cardiovascular death (death from a heart attack, stroke, or sudden cardiac death).
Cardiovascular morbidity, specifically:
- Non-fatal myocardial infarction (heart attack).
- Angina (chest pain requiring medical attention).
- Stroke.
- Atrial fibrillation.
- Peripheral vascular disease.
- Heart failure.
Unplanned interventions, such as emergency bypass surgery or angioplasty.

The Actual Numerical Results

The review included 15 randomized controlled trials with approximately 56,000 participants followed for an average of 4.3 to 5 years.

Outcome	Numerical Result (Risk Ratio)	Interpretation
Combined CV Events	RR 0.83 (95% CI: 0.70–0.98)	A 17% relative risk reduction.
Cardiovascular Mortality	RR 0.95 (95% CI: 0.80–1.12)	NO significant difference.
All-Cause Mortality	RR 0.96 (95% CI: 0.90–1.03)	NO significant difference.
Non-fatal Heart Attack	RR 0.97 (95% CI: 0.87–1.07)	NO significant difference.

The “Absolute” Perspective (NNT)

To your point about these improvements seeming “negligible” to a layman, the researchers calculated the Number Needed to Treat (NNT):

Primary Prevention: In people without existing heart disease, FIFTY-SIX people had to follow the diet for ~4 years for ONE person to avoid a “combined cardiovascular event.”
Total Event Rate: In the control groups, approximately 8.5% of people had an event (85 per 1,000). In the reduced saturated fat groups, this dropped to 7.0% (70 per 1,000).

This results in an Absolute Risk Reduction of 1.5% over 4–5 years. This aligns with your observation regarding PREDIMED: while the relative improvement (17–21%) sounds substantial in medical literature, the absolute chance of an individual avoiding an event is about 1 in 56 over a half-decade of dietary change.

Regarding the Sydney Diet Heart Study (RE: Point 3) You asked for the specific conduct of the study where “replacing saturated fat with safflower oil led to higher mortality.”

Conduct: Conducted from 1966 to 1973 with 458 men who had recently had a coronary event. The intervention group replaced animal fats with safflower oil and safflower oil margarine (high in omega-6 linoleic acid).
The “Recovery”: The original 1970s report was inconclusive. However, in 2013, researchers recovered the raw data and performed a modern analysis.
Numerical Results: They found the group eating the safflower oil had a higher risk of death from all causes (17.6% vs. 11.8%) and from cardiovascular disease (17.2% vs. 11.0%) compared to the group eating the “standard” high-saturated fat diet of the era. This study is often used by skeptics to argue that increasing omega-6 PUFAs specifically can be harmful.

← Previous

18/18

Nutrition and Metabolism, Plantation Bay Assembles the Latest Science

by J. Manuel González, based on critical investigative research, and the mathematical evaluation of clinical trials supporting current health and nutrition advice. For Mr. Gonzalez's full background, please see https://plantationbay.com/cred.