Attia’s July 22, 2026 Premium Article, “Should you choose organic food?” is worth a read by those who are entertaining this question. The article is authoritative, well written, clear, and provides a good summary of most of the relevant science.
Beyond that, I recommend this article as an instructive example of how even a good article – one that has the overall appearance of objectivity – can bring a subtle bias to complex topics such as this and in so doing may lead readers toward conclusions that favor one perspective over another in a way unsupported by the best research. I’m not suggesting that Attia set out to write a biased article – from what I know of him, he did not and much of the article runs straight down the middle. To the contrary, Attia’s interjection of bias is common in scientific articles whenever they fail to apply appropriate safeguards.
If I had to summarize Attia’s bias in this article it would be that he applied standards of evidence differentially when unpacking what we know about standard and organic foods. A summary of his asymmetrical framework might be:
-
For evidence favoring organic food he required affirmative evidence to progress all the way from altered food composition, to altered human biomarkers, to demonstrated clinical outcomes.
-
For evidence favoring conventional food his standard generally stepped down to regulatory toxicology, mechanistic plausibility, animal data, absence of a strong epidemiologic signal, or the fact that exposures fall below regulatory thresholds (which sets up a circular argument).
I’ll touch briefly on a few of moving standards in greater detail. The most important slippage occurs between these propositions:
- Organic diets have not been shown to improve clinical outcomes.
- Organic diets do not meaningfully improve clinical outcomes.
The first generalization is well supported. The second requires either sufficiently powered negative trials, strong natural experiments, or persuasive evidence that the relevant exposure differences are biologically inconsequential, none of which are present in the body of knowledge on this topic. Yet, the article repeatedly acknowledges that outcome evidence is absent and then converts that absence into positive reassurance. Phrases like "choosing organic has not been shown to improve health” gradually becomes: “there is little reason to believe conventional exposure poses meaningful risk.”
While demanding direct evidence of clinical outcomes before crediting organic food with benefit, Attia accepts several lower-level forms of evidence as reassuring for conventional food:
- No-effect levels derived from animals (generally short term)
- Uncertainty factors used by regulators (potentially circular)
- Residues below legal tolerances (potentially circular)
- Lack of a large cancer signal in occupational cohorts
- Mechanistic arguments (rejected on the other side)
- Livestock feeding records
- Absence of an identified biomarker
- Absence of consistent epidemiologic associations.
These forms of evidence are relevant and potentially important, but they are not equivalent to demonstrating long-term safety in consumers. One need only live in the US for a decade to see the truth of this claim play out.
Overall, the article is a good read. It unpacks many issues of potential importance. This is the conclusion I would substitute for Attia’s as better supported by current science.
Organic production reliably reduces exposure to many synthetic pesticide residues and avoids routine antibiotic use in certified livestock. It may also alter selected phytochemical, contaminant, ecological, and resistance-related outcomes, although effects vary greatly by commodity and farming practice. There is currently insufficient evidence that choosing organic produces measurable improvements in clinical outcomes for the average adult. Evidence is likewise insufficient to exclude modest, cumulative, developmental, mixture-related, or population-level benefits. The strongest rationale for organic purchasing is exposure reduction and precaution, especially during pregnancy and early childhood and providing support for particular production practices.
This summary will not satisfy many organic food advocates – it does not satisfy me. Nonetheless, I think it summarizes the current state of evidence. I do think I may have left some evidence on the table with respect to the pointers of harm from some “safe” traditional food products. Perhaps someone can pick that up.