Affiliate disclosure
Book titles on this page link to Amazon. As an Amazon Associate, DataField.Dev earns from qualifying purchases — at no additional cost to you.
Chapter 10 — Further Reading
The trials — read these first
DIETFITS — Gardner and colleagues, JAMA, 2018. "Effect of Low-Fat vs Low-Carbohydrate Diet on 12-Month Weight Loss in Overweight Adults and the Association With Genotype Pattern or Insulin Secretion."
The single most important paper in this chapter, and worth reading in full rather than in summary. Note three things: the within-group spread (visible in the figures and far more striking than the group means), the pre-specified genotype and insulin analyses, and the fact that both arms improved diet quality substantially — this was healthy-low-fat versus healthy-low-carb, not junk versus junk. For: everyone. If you read one primary paper from this book, read this one.
POUNDS LOST — Sacks and colleagues, NEJM, 2009. Four diets, 811 adults, two years, no meaningful difference.
A TO Z — Gardner and colleagues, JAMA, 2007. Atkins, Zone, LEARN, Ornish.
Look AHEAD — the intensive lifestyle intervention trial in type 2 diabetes, NEJM and subsequent papers. Read the primary outcome result and the secondary outcomes, and read about why it was stopped early. It's a case study in a trial that "failed" and taught a great deal. For: anyone who wants to understand what weight loss does and doesn't deliver.
PREDIMED — NEJM, retracted and republished 2018. Covered in Chapter 9; the strongest long-term randomized evidence for any named pattern.
On the diets themselves
Christopher Gardner's public talks and writing. Gardner ran DIETFITS and A TO Z, and has spent a career running head-to-head diet comparisons. He is unusually good at describing what his own trials did not show, which is rarer than it should be. For: everyone. Search for his lectures; they're freely available and unusually honest.
For the low-carbohydrate case at its strongest: the research of Stephen Phinney and Jeff Volek, and the Virta Health studies on carbohydrate restriction in type 2 diabetes. The T2D evidence is the strongest part of the low-carb case and this chapter says so. Note commercial interests — a reason to read carefully, not to dismiss (Chapter 1 §1.5).
For the plant-based case at its strongest: the Adventist Health Studies, and the work of Dean Ornish and Caldwell Esselstyn on intensive lifestyle programmes in cardiovascular disease. Read them alongside Chapter 2 §2.2 — the cohort evidence carries substantial healthy-user bias, and the intensive-programme trials bundle diet with exercise, stress management, and group support.
For DASH: the DASH and DASH-Sodium trials (NEJM). Genuinely excellent evidence, and the
pattern most people have never heard of. NHLBI (nhlbi.nih.gov) publishes free DASH materials.
For: anyone with elevated blood pressure. This is your pattern.
On paleo and evolutionary claims
The anthropological literature on contemporary hunter-gatherers — search for work on the Hadza (Herman Pontzer and colleagues), the Tsimané, and Arctic populations. The variation between groups is the point, and it's much larger than the paleo premise allows.
On continued human evolution: search for lactase persistence and amylase gene copy number variation. Both are well-documented adaptations to agricultural foods, and both directly contradict the "we haven't adapted" premise. For: anyone who finds the evolutionary argument persuasive. It's the most checkable claim in the whole diet wars.
On the Blue Zones — and the argument about them
The Blue Zones work (Dan Buettner and colleagues) has been enormously influential. It has also been subject to serious recent methodological critique concerning the quality of birth and death records in some identified regions, and whether apparent extreme longevity clusters reflect biology or administrative artefact.
Read both. This is a live dispute about whether a widely-cited body of evidence rests on reliable data at all, and it's an unusually good test of whether you can hold a favourite finding loosely — including a finding this book has been broadly sympathetic to. For: everyone who has cited the Blue Zones. Which is a lot of people.
On adherence, which is what actually matters
Research on dietary adherence and dropout in trials. Search for "dietary adherence weight loss predictors" and for analyses of dropout in diet trials. The consistent finding — that adherence predicts outcome far better than assigned diet — is the empirical basis of §10.3.
The National Weight Control Registry (nwcr.ws). Self-selected maintainers, no control group, so
read with Chapter 2 in hand. It tells you what maintainers do, not what causes maintenance — but
what they do is instructive, and Chapter 24 handles it properly.
On behaviour change generally: BJ Fogg's Tiny Habits and James Clear's Atomic Habits are not nutrition books and are more useful for §10.12 than most nutrition books are. The relevant insight — make the behaviour small enough that a bad week doesn't break it — is Case Study 1's "floor." For: anyone whose problem is doing it rather than knowing it, which is most people.
The diet books — read at least two, adversarially
I'd genuinely recommend reading two bestsellers from opposing camps, with Case Study 2's nine-part structure in hand. Not to be persuaded, but to see the architecture.
Pick one low-carbohydrate and one plant-based. Score both. Notice that they have the same skeleton, that both are sincere, that both cite real studies, and that both are unable to concede anything.
That exercise does more for evidence literacy than another textbook chapter would, and it's the reason this page recommends books it disagrees with.
A note on what to be careful with
The failure mode specific to this chapter is a lazy relativism: "the trials show no difference, so it doesn't matter what you eat."
That is not what the evidence says, and it discards the floor. A diet at 0.7 g/kg protein costs you lean mass. A diet at 10 g of fiber underperforms on satiety, lipids, and glycemia. A diet built on refined carbohydrate makes overeating easy. Those are compositional facts and they don't dissolve because the named diets tie.
The tell for a source that's got this right: does it distinguish "among reasonable diets" from "any diet"? A source that concludes "eat whatever you like" has over-read the same evidence that a source concluding "my diet is uniquely superior" has under-read.