Chapter 11 — Fiber: The Nutrient That Does Everything Good and Almost Nobody Gets Enough Of

The Hook: The question that always disappoints

I get asked the same question at the end of every talk I give, in almost the same words.

"If you could only tell people one thing — what's the single most impactful change?"

And I've stopped softening the answer, because softening it was making it worse.

"Eat more fiber. Most of you are getting about half of what you should. Beans, mostly."

You can watch the room deflate. There is a specific quality to the silence — polite, slightly embarrassed, the sound of two hundred people who came for something else. Somebody usually asks a follow-up about intermittent fasting.

I want to take that reaction seriously rather than complain about it, because the deflation is diagnostic. It tells you exactly what's wrong with how nutrition information reaches people.

Here is what I'm actually offering when I say that.

Fiber intake has one of the strongest, most consistent, most dose-responsive bodies of evidence in all of nutrition — associated with lower cardiovascular disease, lower type 2 diabetes, lower colorectal cancer, lower all-cause mortality, better cholesterol, better glycemic control, better satiety, and better bowel function. It converges across cohorts, trials on intermediate outcomes, and mechanism (Chapter 2 §2.10's criterion, met about as well as nutrition ever manages).

And roughly one adult in twenty gets enough.

That is a gap between evidence and behaviour unmatched by anything else in this book. There is no supplement with a fraction of that evidence. There is no diet in Chapter 10 whose named mechanism comes close.

And nobody is selling it to you, because you cannot patent a bean, the margins on lentils are terrible, and "eat more fiber" has been the same advice since before you were born — which, per Chapter 1 §1.7, is precisely the property that makes something reliable and unmarketable at the same time.

Theo's fiber intake is 14 grams a day against a target of 38.

That gap is doing more to his health than every macronutrient argument in the last four chapters combined, and he had never once thought about it.

🏃 Fast Track: §11.4 (the evidence), §11.6 (how much, and where you actually are), and §11.8 (how to increase it without misery). Twenty minutes, and §11.8 is the one that determines whether any of this happens.

🔬 Deep Dive: §11.2 (the four properties — this replaces soluble/insoluble), §11.3 (mechanisms), and §11.10 (when fiber is the wrong advice) are where students and clinicians should spend time.

11.0 A note on why this chapter is structured backwards

Most chapters in this book begin with mechanism and end with what to do. This one front-loads the evidence and spends its longest section on how to increase your intake without feeling terrible — and that ordering is deliberate.

The reason is that fiber has an unusual failure profile. With most nutrition interventions, the problem is that people don't know what to do. With fiber, almost everyone already knows what to do. "Eat more fiber" is not information anybody lacks. It has been on cereal boxes since the 1980s.

The failure is entirely downstream of knowing, and it happens in two specific places:

Failure one: it never becomes a specific action. "Eat more fiber" is not a behaviour; it's a category. Nobody has ever executed it. "Put a tin of black beans into Tuesday's dinner" is a behaviour, and §11.8's checkpoint exists to force the translation.

Failure two: people do it too fast, feel awful, and conclude fiber doesn't suit them. This is extremely common, entirely preventable, and produces a durable false belief — the person now has personal evidence, rung 3 and highly persuasive to them, that a well-supported intervention is harmful in their case.

So the chapter is arranged around those two failures rather than around the biochemistry. If you're short of time, §11.8 matters more than §11.3.

11.1 What fiber actually is

Dietary fiber is carbohydrate that human enzymes cannot digest. That's the whole definition, and it's a definition by absence — which is why the category is so chemically heterogeneous.

It includes cellulose, hemicellulose, pectin, beta-glucan, inulin, fructans, gums, mucilages, resistant starch, and lignin (which isn't a carbohydrate at all but travels with them and is counted).

They have almost nothing in common chemically. What they share is that your small intestine can't break them down, so they arrive in the colon intact (Chapter 3 §3.6) — and what happens next is where the whole chapter lives.

The distinction you were taught, and why to drop it

Soluble versus insoluble. Soluble dissolves in water (oats, beans, apples, psyllium); insoluble doesn't (wheat bran, vegetable skins, nuts).

It's not wrong. It's just not the property that predicts what a fiber does — the same problem Chapter 7 had with simple-versus-complex, and it fails the same way. Two soluble fibers can behave completely differently; some insoluble fibers are highly fermentable and some aren't.

The field has largely moved on, and so should you.

11.2 The four properties that actually matter

Judge a fiber on these instead.

A single food usually has several. Oats are viscous and fermentable. Psyllium is viscous and bulk-forming and only modestly fermentable — which is why it's unusually well tolerated. Wheat bran is bulk-forming and barely fermentable, which is why it relieves constipation without much gas.

💡 Aha moment. This table explains something that confuses a lot of people: why "high fiber" products don't all do the same thing.

If you want to lower LDL, you want viscous fiber — oats, barley, psyllium, legumes. A high- fiber cereal bulked out with cellulose won't do it.

If you're constipated, you want bulk and water — wheat bran, or psyllium, plus fluid. Inulin will mostly give you wind.

If you want to feed your microbiome, you want fermentable fiber and variety — and gas is the byproduct of the thing working, not evidence that it isn't.

"Fiber" on a label tells you a grams number and nothing about which of these you're getting, which is why matching fiber type to purpose is the most under-taught practical skill in this area.

11.3 What fiber does, mechanism by mechanism

Seven mechanisms. Notice how few of them are the one everyone knows.

1. It lowers LDL cholesterol. Viscous fiber binds bile acids in the small intestine and carries them out. The liver, needing to replace them, pulls cholesterol out of circulation to make more. This is well established — well enough that regulators have authorized health claims for oat beta-glucan and psyllium and blood cholesterol.

📊 Diagram (described). Mechanism 1 is worth picturing properly, because it explains why viscous fiber specifically lowers cholesterol and other fibers don't.

Picture a loop. Your liver makes bile acids out of cholesterol and sends them into the duodenum, where they emulsify fat (Chapter 3 §3.4). Having done that job, they travel down the small intestine — and at the ileum, they're reabsorbed and shipped back to the liver to be used again. Round and round, several times per meal. It is an efficient recycling system: the great majority of bile acids are recovered.

Now drop viscous fiber into the tube. Psyllium, beta-glucan, and pectin form a gel that traps bile acids, preventing some of them from reaching the ileal reabsorption site. Those trapped bile acids continue into the colon and leave the body.

The loop is now leaking. The liver, needing to maintain its bile acid supply, has to make more — and bile acids are made from cholesterol. So the liver increases its uptake of cholesterol from the bloodstream to feed the loop. Blood LDL falls.

Two things follow. First, this is why viscosity is the property that matters — a non-gelling fiber doesn't trap anything, which is why wheat bran is excellent for constipation and does very little for cholesterol. Second, notice how mechanical the mechanism is. No metabolic magic; a gel in a tube interrupting a recycling loop. It's also, structurally, exactly how the bile acid sequestrant drug class works — psyllium is doing a gentler version of a pharmaceutical mechanism, for twenty cents.

2. It blunts glucose absorption. The gel slows gastric emptying and slows glucose reaching the intestinal surface, flattening the post-meal curve — mechanically, without any change to the carbohydrate itself.

3. It increases satiety. Volume, chewing time, slower gastric emptying (Chapter 3 §3.3), and fermentation products that stimulate GLP-1 and PYY hours later (§3.7). High-fiber diets reliably reduce spontaneous energy intake, which is why fiber does so much work in every successful pattern in Chapter 10.

4. It feeds your colonocytes. Fermentation produces butyrate, the preferred fuel of the cells lining your colon (Chapter 3 §3.6). Your gut lining is partly fed by bacteria, from material you couldn't digest.

5. It increases stool weight and speeds transit. The mechanism everyone knows, and the least interesting one — though it matters clinically, and the reduction in colorectal cancer risk may be partly about reducing the contact time between the colon wall and its contents.

6. It shapes the microbiome. Fiber is the primary food supply of your colonic bacteria, and dietary fiber diversity is one of the better predictors of microbial diversity — which Chapter 27 will treat with appropriate caution.

7. It reduces the energy density of everything around it. Fiber adds bulk and water without adding usable energy. A high-fiber meal is physically larger per calorie, which is most of what "filling" means.

🔍 Why this works. Energy density is worth understanding properly, because it's the quiet mechanism behind most of what fiber does for appetite — and it's more mechanical than people expect.

People eat a fairly consistent weight of food, not a consistent number of calories. Across many studies, when the energy density of a diet is manipulated while palatability and variety are held constant, people tend to eat a similar mass and therefore take in different amounts of energy — often without noticing.

Now look at what fiber does to mass per calorie:

Food	Energy density
Oil	~9 kcal/g
Crisps, biscuits, most snack foods	~4.5–5.5 kcal/g
Bread	~2.5 kcal/g
Cooked lentils	~1.1 kcal/g
Cooked vegetables	~0.3–0.5 kcal/g
Broth-based soup	~0.3 kcal/g

A cup of lentils and a small handful of crisps deliver similar energy. One weighs about 200 grams; the other weighs about 40. If you eat by mass — and people largely do — the difference is enormous and requires no willpower at all.

This is why fiber's satiety effect isn't really about a hormone. It's about the fact that fiber and its associated water take up space that calories would otherwise occupy. Chapter 33 will argue that environment beats willpower; energy density is the same argument, running inside the food.

🔬 Claim → Evidence → Verdict

The claim: "Fiber is just roughage — indigestible bulk that scrubs your intestines out. It has no nutritional value."

Where it comes from: The old model, and it isn't stupid — fiber genuinely is indigestible by human enzymes, and its laxative effect is real and obvious.

What the evidence actually shows: Six of the seven mechanisms above have nothing to do with scrubbing. Fiber changes bile acid recycling, glucose kinetics, satiety hormone release, colonic epithelial fuel supply, microbial ecology, and energy density. It also isn't calorie-free — colonic fermentation returns roughly 2 kcal/g as short-chain fatty acids (Chapter 3 §3.9). The "scrub" model isn't just incomplete; it's the least important thing fiber does.

📉 Evidence quality: Well-established mechanisms across multiple lines.

Verdict: ❌ Not supported. You aren't scrubbing. You're farming.

11.4 The evidence

This is the part that justifies the chapter's opening claim, so let me be specific about what exists.

The 2019 Lancet systematic review and meta-analysis by Reynolds, Mann and colleagues pooled observational studies and clinical trials on dietary fiber and whole grains. Its findings, in brief:

• Higher fiber intake was associated with lower all-cause mortality, coronary heart disease incidence and mortality, stroke, type 2 diabetes, and colorectal cancer.
• There was a clear dose-response relationship — more fiber, better outcomes, across the observed range.
• The greatest reductions were seen at intakes of roughly 25–29 g/day, and the curves suggested benefit continuing above that.
• Trial data supported reductions in body weight, blood pressure, and total cholesterol.

Why this is unusually strong evidence by nutrition standards, in Chapter 2 §2.10's terms:

The honest caveats, because this book states them even when it likes the finding:

• Healthy-user bias applies. People eating 35 g of fiber differ from people eating 12 g in many ways. This is Chapter 2 §2.2 and it doesn't stop applying because the conclusion is convenient.
• Fiber travels with everything. High-fiber diets are high in vegetables, legumes, whole grains, nuts and fruit — and low in ultra-processed food. Disentangling fiber from its food matrix may not even be a meaningful thing to attempt, and §11.9 is where that matters.
• There is no long-term randomized trial of fiber intake with hard endpoints, and there won't be.

So the honest summary: this is about as good as nutrition evidence gets, and it is still observational at its core. That's why the practical recommendation is "eat more high-fiber foods" rather than "take fiber."

And what the numbers look like in absolute terms

Chapter 2 §2.6 established that relative risk without a baseline is uninterpretable, and this book should apply that to findings it likes as rigorously as to findings it doesn't.

So: the meta-analytic estimates for higher versus lower fiber intake sit in the region of 15–30% relative reductions across several outcomes. What does that mean in absolute terms?

Take colorectal cancer. Lifetime risk in a typical Western population is roughly 5% — about 5 in 100 people.

Baseline lifetime risk            ≈ 5.0%
With ~20% relative reduction      ≈ 4.0%
Absolute difference               ≈ 1.0 percentage point
                                  ≈ 1 fewer case per 100 people, lifetime

One fewer case per hundred people. That is a real effect, and it is not a transformation — and it is roughly the same magnitude, in the opposite direction, as the processed meat figure from Chapter 2.

Three honest observations about that number.

First: it is genuinely modest per person and large per population. One in a hundred, applied to a country, is an enormous number of people — which is why public health cares about it and why it can still be rational for an individual to shrug.

Second: the outcomes stack. Fiber isn't offering you one 1-percentage-point reduction. It's offering a modest reduction in colorectal cancer and cardiovascular disease and type 2 diabetes and all-cause mortality, plus non-mortal benefits — cholesterol, glycemic control, satiety, bowel function — that you'll actually notice. Very few single interventions appear across that many endpoints, and the stacking is the real argument, not any individual figure.

Third: the cost is negative. Most interventions ask you to trade something. This one costs about $2 a week and improves how full you feel. The relevant comparison isn't "1% vs. nothing" — it's "1% across five endpoints, for two dollars, versus $2,244 a year for nine supplements with a fraction of the evidence." That comparison is Walt's, and it isn't close.

11.5 The gap

Now the number that should be more famous than it is.

The Adequate Intake is set at 14 g per 1,000 kcal consumed — which is a more useful way to hold it, because it scales with how much you eat.

Roughly 5% of adults in the US meet the recommendation. Comparable figures apply across most industrialized countries.

Sit with that. We are talking about a nutrient with one of the strongest evidence bases in the field, where ninety-five percent of the population falls short, and where the average shortfall is around half the target.

There is no other nutrient where the gap between what we know and what people do is this wide.

Our four:

🔄 Check your understanding. Walt has type 2 diabetes with an A1c of 7.4% and spends $2,244 a year on nine supplements, including berberine for glucose control. His fiber intake is 14 g against a target of 30. What's the argument you'd make to him, and why is it hard to make?

Answer

The argument: raising fiber from 14 g to 30 g would be expected to improve his post-meal glucose excursions (mechanism 2), his LDL (mechanism 1), and his satiety and therefore his weight (mechanism 3) — with a body of evidence vastly stronger than anything supporting berberine, at a cost of perhaps **$3 a week in dried beans and oats** against $187 a month.

Why it's hard to make: because it's boring, and because it isn't an addition to his identity as someone taking charge of his health. The nine bottles feel like action. A tin of beans doesn't. He has been sold a story in which optimizing is a purchase, and I am offering him a chore.

This is Chapter 10 §10.10's identity problem applied to a single nutrient — the intervention with the best evidence-to-cost ratio in his entire life offers him nothing to be.

(And note: berberine does have a real glucose-lowering effect. The problem is unregulated dosing stacked on metformin without monitoring — Chapter 16. It isn't that his supplement does nothing; it's that the free thing does more, more safely.)

11.6 So how much, and from where

Target: roughly 25–38 g/day depending on sex and size, or 14 g per 1,000 kcal.

Here's where it actually comes from. This is the table to keep.

(Approximate; check USDA FoodData Central.)

Two things jump out.

Legumes dominate. One cup of lentils is 15 g — nearly two-thirds of a woman's daily target in a single, cheap, shelf-stable ingredient. Nothing else on this table comes close per serving or per dollar. This is why "beans, mostly" is the honest answer to the one-thing question.

And the entire animal-food column is zero. Which is not an argument against eating them — it is the explanation for why a diet composed largely of meat, eggs, dairy and oil (Chapter 10 §10.6) contains no fiber at all, and why that's the specific thing to worry about with it.

11.7 What 38 grams actually looks like

Targets are abstract. Here are two days that hit them, so you can see the shape.

Day A — omnivore, ~38 g

Day B — plant-forward, ~40 g, under $4

Notice what neither day requires: a supplement, a specialty product, an unusual ingredient, or a skill. Day B costs under four dollars.

And notice how a single legume serving reshapes the day. Remove the lentils from Day B and you drop from 47 g to 32 g. One ingredient, one meal, fifteen grams — which is why §11.8's advice reduces to almost one sentence.

🧩 Productive struggle. Five minutes before reading on.

Camila Ortiz eats roughly 10 g of fiber a day. Her constraints: three twelve-hour night shifts a week, main meal at 3 a.m. from a hospital vending machine or break room, cooks nothing on shift days, household budget $180/week for four, and a four-year-old who eats nine foods.

Get her to 25 g. Write the actual changes. Then check yours against mine.

What I'd do

The constraint that dominates is that three days a week she has no kitchen. So any plan built around cooking fails 43% of the time, and a plan that fails 43% of the time is not a plan.

Shift days — build a portable box, made on a non-shift day: - Tinned chickpeas or beans, drained, with whatever dressing she likes — 1 cup, 12 g - Wholemeal bread or oatcakes — 4 g - An apple or pear — 4–6 g - A handful of almonds — 3.5 g

That's 23–25 g in a box she assembles in four minutes, needs no fridge for a shift, costs about $1.80, and replaces a vending machine.

Non-shift days — two changes to what the family already eats: - Porridge instead of toast for her breakfast — +6 g - A tin of beans into whatever Erik is already making (chilli, soup, pasta sauce) — +6–8 g per serving, and per Chapter 7's Case Study 1 this is additive, so nobody has to be persuaded of anything and Alma can ignore it.

Total: roughly 25–28 g on both kinds of day. Cost: under $4 a week for the whole household.

What I would NOT do: tell her to meal-prep on Sundays (she's sleeping), buy fresh vegetables that die during an overrun shift (Dominic's problem, Chapter 10 Case Study 1), or address her fiber before addressing the fact that she's eating from a vending machine at 3 a.m., which is a logistics problem wearing a nutrition costume.

If your answer involved cooking on shift days, go back and look at where the constraint actually is. That's the exercise.

11.8 The ramp: how to increase fiber without misery

This section matters more than everything above it, because the way most people fail at fiber is by succeeding at it too fast.

Go from 14 g to 38 g overnight and you will be bloated, uncomfortable, and windy — and you will conclude, entirely reasonably, that fiber doesn't agree with you. That conclusion is wrong and it is extremely common, and the sequence is preventable.

Why it happens: your colonic bacteria are a population adapted to your current substrate supply. Triple that supply abruptly and you get rapid fermentation, gas, and osmotic water shifts before the microbial community has adjusted. Given a few weeks, the community shifts and tolerance improves substantially.

The protocol

Theo's ramp, as an example: 14 g → 19 g (week 1: oats replacing toast) → 24 g (week 3: half a cup of beans added to two dinners) → 30 g (week 5: fruit as the afternoon snack, wholemeal bread) → 38 g (week 8: full legume servings, more vegetables at lunch).

Eight weeks, five changes, no supplement. He reported "a rough week three" and nothing after that.

🍽️ On your plate. If you do exactly one thing from this chapter: add one serving of legumes to one meal, three times a week. Tinned beans cost about sixty cents a can and require no cooking skill — drain, rinse, add to whatever you were already making.

That single change is roughly +7 g/day averaged, costs under $2 a week, requires no new recipe, and is invisible to everyone else at the table. It is the highest evidence-to-effort ratio intervention in this book, and it takes eleven seconds.

11.9 Do fiber supplements work?

A fair question, given how much easier a spoonful is than a lifestyle.

🔬 Claim → Evidence → Verdict

The claim: "A fiber supplement gives you the same benefits as high-fiber foods, without having to change what you eat."

Where it comes from: Reasonable inference from mechanism, plus genuinely good evidence for specific supplements on specific outcomes.

What the evidence actually shows: It splits cleanly, and the split is instructive.

Where supplements genuinely work — mostly psyllium: psyllium has good trial evidence for lowering LDL cholesterol, improving glycemic control, and — unusually — normalizing stool in both constipation and diarrhea, because it's viscous and water-holding without being highly fermentable. It's cheap, well tolerated, and it does what it says.

Where they don't: the whole-outcome evidence in §11.4 — mortality, cardiovascular events, colorectal cancer — comes from fiber-rich diets, not from supplements. Those diets deliver fiber alongside potassium, magnesium, folate, polyphenols, resistant starch, protein and a low energy density, and displace something else. A supplement delivers isolated fiber and displaces nothing. This is the whole-food fallacy from Chapter 2's beta-carotene case (Case Study 1), arriving again: observed in food → inferred to a compound → tested as a pill.

Isolated fibers also differ enormously. Inulin is highly fermentable and produces significant gas. Methylcellulose is bulk-forming and non-fermentable. "Fiber supplement" is not one product.

📉 Evidence quality: Good trial evidence for psyllium on specific intermediate outcomes. No supplement evidence for the hard outcomes in §11.4.

Verdict: 🟡 Unclear / it depends. Psyllium is genuinely useful for LDL, glycemia, and bowel regularity, and is a reasonable addition. As a substitute for high-fiber foods, unsupported — you'd be buying one mechanism out of seven and none of the food matrix.

🧾 Cost check: psyllium husk runs roughly $0.15–$0.30 per 7 g serving. Branded fiber gummies — which often deliver 3–4 g per serving, sometimes with added sugar — run $0.50–$1.00 for less fiber. A cup of dried lentils delivers 15 g for about twenty cents. The gummies are the worst value in the category by a wide margin, and they exist because they solve a compliance problem rather than a nutritional one.

11.10 When fiber is the wrong advice

This chapter has been enthusiastic. Here's the boundary, because "eat more fiber" is genuinely wrong for some people some of the time, and getting this wrong causes real harm.

⚠️ When to see a professional. If increasing fiber sensibly and gradually causes significant pain, persistent bloating, or a change in bowel habit that doesn't settle within a few weeks, stop and see a physician rather than pushing through. And the standing rule from Chapter 3: persistent digestive symptoms deserve a diagnosis, not an elimination diet — and celiac testing requires gluten in the diet to be valid.

Red flags requiring prompt assessment: blood in stool · unexplained weight loss · symptoms waking you at night · iron-deficiency anemia without obvious cause · new persistent change in bowel habit, particularly over age 50.

🔬 Claim → Evidence → Verdict

The claim: "Fiber causes bloating and gas, which means it's irritating your gut. Cutting fiber makes people feel better — proof that we're not meant to eat it."

Where it comes from: A real experience. People who cut fiber sharply — on carnivore, on low-residue, on very low-carbohydrate diets — frequently report less bloating, and they're reporting accurately.

What the evidence actually shows: Less fermentation means less gas. That's mechanism, not pathology — it's the same logic as concluding that a quiet factory is a better factory. Gas is the byproduct of colonic bacteria doing the thing that produces butyrate and feeds your gut lining (§11.3). The transient discomfort of a rapid increase (§11.8) is an adaptation problem with a known solution, and most people tolerate a gradual ramp well.

What's genuinely true in it: for people with IBS, fermentable fiber really can drive real symptoms, and for them the type of fiber matters more than the amount (§11.10). That's a specific clinical situation, not a general principle — and even there, the answer is usually different fiber rather than no fiber.

📉 Evidence quality: The symptom report is accurate; the inference from it isn't supported, and the outcome evidence in §11.4 runs strongly the other way.

Verdict: 🟠 Probably false as a general claim. Feeling less bloated is not the same as being healthier, and this is one of the clearest cases in the book of a real short-term sensation being read as a long-term signal.

🔄 Check your understanding. Three people ask for fiber advice. Match each to the right fiber type and say why.

(a) A 58-year-old with LDL of 155 who wants a dietary lever before considering medication. (b) A 34-year-old with chronic constipation, drinks little water, currently eats no fiber. (c) A 41-year-old with diagnosed IBS whose symptoms flare after beans and onions.

Answer

(a) Viscous fiber. Oats or barley (beta-glucan), psyllium, legumes, pectin-rich fruit. This is mechanism 1 — bile acid trapping — and it's the one with an authorized health claim behind it. A high-fiber cereal bulked with cellulose would do nothing here. ⚠️ And be clear with them that dietary change moves LDL modestly; at 155 this is a conversation about total risk with a physician (Chapter 9 §9.10), not a diet-versus-statin choice.

(b) Bulk and water — and the water is not optional. Wheat bran or psyllium, with a genuine increase in fluid. Adding bulk-forming fiber to someone who drinks little water can worsen constipation, which is the opposite of the intent and a common clinical own-goal. Start low, go gradually (§11.8).

(c) This one is the trap. The instinct is to reduce fiber, and that's often wrong. Beans and onions are high in fermentable fiber — galacto-oligosaccharides and fructans, the FODMAPs from Chapter 7 §7.11. The answer is usually different fiber, not less fiber: psyllium is viscous and water-holding but only modestly fermentable, and is often well tolerated in IBS. A supervised temporary low-FODMAP approach with structured reintroduction may help identify their specific triggers (Chapter 28) — temporary and supervised being the operative words, because permanent broad elimination trades one problem for a worse one.

The general lesson: "eat more fiber" is the right advice for the population and an incomplete instruction for a person. Match the property to the purpose (§11.2).

11.11 Why nobody is selling you this

Return to the deflating room.

Fiber has no industry. Consider what a nutrient needs in order to be marketed:

Zero for seven. Whereas the metabolism aisle (Chapter 5, Case Study 2), the protein bar (Chapter 8, Case Study 2), and the diet book (Chapter 10, Case Study 2) score highly on most of them.

💡 Aha moment. This is Chapter 1 §1.5's incentive analysis producing its sharpest single result: the strength of the evidence for a nutrition intervention is roughly inversely proportional to how much you have heard about it.

Not because good things are suppressed — nobody is hiding fiber. Because attention is allocated by marketing spend, and marketing spend follows margin, and margin follows patentability and novelty. Fiber has none of those, so it gets no attention, so a room of two hundred health-interested people deflates when you name the single best-evidenced change available to almost all of them.

The deflation isn't their fault either. They came having absorbed, correctly, that the interesting answers are elsewhere — because everything that reached them was selected for being interesting.

🧾 Cost check. Since this chapter's whole argument is that the best-evidenced intervention is also the cheapest, it's worth pricing properly. Cost per 10 g of fiber:

Source	Cost per 10 g fiber
Dried split peas	~$0.12
Dried lentils	~$0.14
Rolled oats	~$0.20
Tinned beans	~$0.40
Psyllium husk	~$0.25
Wholemeal bread	~$0.55
Frozen vegetables	~$0.70
Fresh raspberries	~$3.50
Fiber gummies	~$2.50–$3.00 (and often with added sugar)
"High fiber" snack bars	~$2.00–$3.50

Closing Theo's 24 g gap with dried lentils costs about 34 cents a day. Closing it with gummies costs about $7.

And set that against the rest of the book: Walt's supplements at $187/month, the metabolism aisle at $240/month** (Chapter 5), a protein bar habit at **$2.90 per 20 g (Chapter 8). The intervention with the strongest evidence in Part II costs less than every one of them by an order of magnitude, which is either encouraging or infuriating depending on how much you've already spent.

What we don't know

We don't know how much of fiber's benefit is fiber. High-fiber diets are high in a dozen other things and low in ultra-processed food. Whether the effect is the fiber, the food matrix, the displacement, or the kind of person who eats that way is genuinely unresolved — and the fact that supplements haven't reproduced the outcome benefits is a real, uncomfortable data point.

We also don't know the optimum. The dose-response curves suggest benefit continuing above 30 g, but the data thins at high intakes because so few people are there. Whether 50 g is better than 35 g, for whom, and with what tolerability cost, is not established.

And we don't know why individual tolerance varies so much — why one person handles 45 g comfortably and another struggles at 25. Microbiome composition is the obvious candidate and Chapter 27 explains why we can't yet act on it.

Spaced Review

Answer before reading on.

1. (Chapter 3) What happens to fiber in the colon, and what's the most important product?

Colonic bacteria ferment it, producing short-chain fatty acids — principally acetate, propionate, and butyrate, which is the preferred fuel of the cells lining your colon (§3.6). Your gut lining is partly fed by bacteria, from material you couldn't digest. Fiber also returns roughly 2 kcal/g by this route, so it isn't calorie-free.

2. (Chapter 10) Every successful dietary pattern in Chapter 10 shares certain features. Where does fiber sit in that list, and which pattern is the exception?

Fiber is high in all of them except the low-carbohydrate variants — and it's part of the reason they work, via satiety and reduced spontaneous intake. Carnivore is the extreme exception at zero, which §11.6's table makes concrete: the entire animal-food column contains no fiber.

3. (Chapter 2) The fiber evidence was called "about as good as nutrition gets." Which of Chapter 2 §2.10's criteria does it meet, and which does it not?

Meets: convergence across designs with non-overlapping weaknesses · clear dose-response · consistency across populations · plausible mechanism (seven of them) · boring and old. Does not meet: long-term randomized trials with hard endpoints — which don't exist and won't. Healthy-user bias still applies, and it doesn't stop applying because the conclusion is convenient.

Project Checkpoint: Your Fiber Audit and Ramp Plan

Component eleven, and for most readers the single highest-yield checkpoint in this book.

Step 1 — the number. From your three-day diary, total your fiber and average it. If your diary didn't capture fiber, use §11.6's table and estimate — precision doesn't matter here.

My average fiber: ______ g/day · My target (§11.5): ______ g/day · Gap: ______ g

Most readers find a gap of 10–20 g. Theo's was 24.

Step 2 — find your fiber. Where is the fiber you do eat coming from? List the top three sources. Then ask: how many servings of legumes did you eat in three days? For most people the answer is zero, and that single fact usually explains most of the gap.

Step 3 — build the ramp. Fill this in with real foods you would actually eat:

Rules: +5 g per week maximum · increase fluid alongside · spread across meals · start with better-tolerated sources · expect a rough week somewhere around week two or three, and don't stop.

Step 4 — the legume commitment. Separately from the ramp, commit to one specific thing:

I will add ______ (beans/lentils/chickpeas) to ______ (which meal) on ______ (which days).

Be specific. "Eat more legumes" fails; "tinned black beans into Tuesday and Thursday's dinner" survives.

Step 5 — check your fiber type against your purpose (§11.2). If you're targeting LDL, you want viscous — oats, barley, psyllium, legumes. If constipation, bulk and water. If it's general health, variety beats any single source.

Non-tracking alternative. Skip the grams entirely. Instead, count servings of high-fiber foods per day — legumes, whole grains, vegetables, fruit, nuts, seeds. Most people managing well land around 7–9 servings. Count yours for three days, then add one serving a week. Same outcome, no numbers.

Next checkpoint (Chapter 12): your alcohol audit — an honest weekly count, the calories, and your own risk/benefit call.

Chapter Summary

Fiber is carbohydrate human enzymes can't digest — a definition by absence, which is why the category is chemically heterogeneous.

Drop soluble/insoluble. Use four properties instead:

Seven mechanisms: LDL via bile acid binding · blunted glucose absorption · satiety · butyrate for colonocytes · transit and stool weight · microbiome · reduced energy density. Only one of them is "roughage."

The evidence (Reynolds/Mann Lancet 2019 and the wider literature): lower all-cause mortality, CHD, stroke, type 2 diabetes, colorectal cancer, with clear dose-response, greatest reductions around 25–29 g/day and benefit continuing above. Meets nearly every Chapter 2 §2.10 criterion — except long-term RCTs with hard endpoints, which don't exist.

The gap: targets are 38 g (men) / 25 g (women), or 14 g per 1,000 kcal. Typical intake is about half. Roughly 5% of adults meet it.

Where it comes from: legumes dominate. One cup of lentils = 15 g. The entire animal-food column is zero.

The ramp — the part that determines whether this happens: +5 g/week · more fluid · spread across meals · better-tolerated sources first · expect 2–4 weeks of adjustment · vary sources.

And the one-line version: add one serving of legumes to one meal, three times a week. +7 g/day averaged, under $2/week, eleven seconds.

This chapter's verdicts:

⚠️ When fiber is the wrong advice: IBS (type matters) · active IBD flare · stricture risk · gastroparesis · recent bowel surgery · slow-transit constipation · very young children.

And the absolute numbers, honestly (§11.4): roughly 1 fewer colorectal cancer case per 100 people, lifetime — modest per person, enormous per population. The real argument is that the outcomes stack (cardiovascular, diabetes, mortality, plus cholesterol, glycemia, satiety and bowel function you'll actually notice) and that the cost is about $2 a week. Very few single interventions appear across that many endpoints, and none of the others is free.

Cost per 10 g of fiber: dried lentils ~$0.14** · oats ~$0.20 · psyllium ~$0.25 · tinned beans ~$0.40 · **fiber gummies ~$2.50–$3.00. Closing Theo's 24 g gap costs 34 cents a day with lentils and $7 with gummies.

The one thing to remember: the strength of the evidence for a nutrition intervention is roughly inversely proportional to how much you have heard about it.

What's Next

Chapter 12 closes Part II with the least comfortable chapter in it.

Where did the J-curve come from, and why is it probably an artefact of who ends up in the "non-drinker" group? What did Mendelian randomization add — and why does it matter that it disagreed with the cohort data, which is exactly the signal Chapter 2 §2.10 told you to watch for? What does IARC Group 1 actually mean here?

And then the part most books skip: how to make your own decision as an adult, with real numbers, without being lectured at.

Theo's two beers a night are 190 kcal, 6% of his intake, and the largest single lever on the lipid panel from Chapter 9. That's not a moral problem. It's an arithmetic one, and he gets to decide.