↓ Skip to main content
NCW
  1. Health/

The Starch That Feeds Your Bacteria Instead of You

·1154 words·6 mins· loading · loading · ·
Gut Health Nutrition Metabolic Health
Neo W.
Author
Neo W.
Writing about things that intrigue me.
Table of Contents

The Starch That Feeds Your Bacteria Instead of You
#

In a randomised crossover trial, 40 g a day of resistant starch for eight weeks produced a mean weight loss of 2.8 kg — with no calorie restriction, against an energy-matched control starch. The mechanism wasn’t willpower. It was a change in which bacteria were living in the participants’ colons.

The fat that actually matters
#

Visceral fat sits behind the abdominal wall, packed around the liver, pancreas and intestines. Unlike the subcutaneous fat you can pinch, it’s metabolically active — continuously releasing inflammatory signalling into portal circulation. It carries substantially higher risk of cardiovascular disease, type 2 diabetes, several cancers and cognitive decline.

The uncomfortable part is that you can’t see it. Someone can look acceptable in a t-shirt and carry a dangerous amount — the “thin outside, fat inside” pattern. Appearance tells you close to nothing here.

What resistant starch does differently
#

Ordinary starch — rice, pasta, potato — is broken down in the small intestine and absorbed as glucose, producing the usual rise in blood sugar and insulin.

Resistant starch resists that digestion. It arrives in the colon intact, where bacteria ferment it into short-chain fatty acids, principally butyrate and propionate. Those travel via the gut–liver axis and act as signalling molecules rather than as calories.

So the same gram of starch either becomes glucose in your bloodstream or becomes bacterial fuel that changes how your liver behaves. The routing is the entire intervention.

What the trials actually show
#

Here the evidence needs stating carefully, because this area attracts inflated claims.

The strongest result is Li et al. (2024, Nature Metabolism): 37 participants with overweight or obesity, randomised placebo-controlled crossover, 40 g/day of resistant starch for eight weeks against an energy-matched control starch, with a four-week washout between phases. Outcome: mean weight loss of 2.8 kg and improved insulin resistance.

The mechanism they identified is the interesting part. Resistant starch reshaped the gut microbiota, which altered the bile acid profile, restored intestinal barrier integrity, reduced inflammation and inhibited lipid absorption. One species — Bifidobacterium adolescentis — was strongly associated with the improvement, and supplementing it protected mice from diet-induced obesity.

On visceral fat specifically, the relevant work is an earlier crossover trial in which participants took 40 g of high-amylose RS2 against energy-matched control starch; visceral and subcutaneous fat areas were both significantly reduced.

On liver fat, Ni et al. (2023, Cell Metabolism) randomised people with fatty liver disease to 40 g/day for four months and found a 9.08% absolute reduction in intrahepatic triglyceride, and 5.89% after adjusting for weight loss — meaning most of the effect was independent of weight.

What I’d drop. Claims circulating alongside this — a “300% elimination of visceral fat,” a specific participant losing “100 cm² in two months,” a “6× greater liver-fat reduction” — are not figures I can locate in the published trials, and “300% elimination” isn’t a coherent quantity in the first place. The real results are strong enough without them: 2.8 kg without dieting, measurable visceral and liver fat reduction, largely weight-independent.

The GLP-1 comparison needs calibrating too
#

Resistant starch does raise GLP-1, the hormone that semaglutide and tirzepatide mimic. That’s real and it plausibly contributes to the appetite effects.

It is not, however, comparable in magnitude to those drugs, which produce supraphysiological receptor activation and correspondingly larger effects. A fermentable fibre nudging endogenous GLP-1 and a weekly injection are different interventions that happen to touch the same pathway. Treat “raises GLP-1 naturally” as a mechanism, not as a substitute.

Making it, and taking it
#

RS3 is the practical type, and you produce it in your own fridge through retrogradation — cook a starchy food, then cool it. Rice, potatoes, oats and pasta all work; refrigerate for 12–24 hours and some digestible starch recrystallises into a resistant form. Gentle reheating retains most of it.

RS1 is locked inside cell walls — whole grains, beans, lentils. RS2 occurs in raw green bananas and raw potato and is destroyed by cooking.

Reaching 40 g from food alone is impractical — you’d need something like eight cups of cold rice. Realistically that means either accepting a lower food-based intake, or supplementing: unmodified raw potato starch is cheapest and well studied, green banana flour is gentler, and high-amylose maize starch is the research-grade form used in most trials.

Four practical rules:

  • Ramp up over about four weeks — roughly 5 g, then 10–15 g, then 30 g, then 40 g. Jumping straight in produces genuinely unpleasant bloating and is the main reason people quit.
  • Never mix it into hot liquid. Heat reverses the structure and turns it back into ordinary starch.
  • Take it on an empty stomach or well before your largest meal, rather than buried inside a high-carbohydrate one.
  • Expect variable response. Baseline microbiome composition determines how much you get from it, which is exactly what the Bifidobacterium finding implies. Pairing it with a broadly Mediterranean pattern and resistance training is the sensible stack.

Summary
#

Resistant starch escapes digestion and is fermented in the colon into butyrate and propionate, which act on the gut–liver axis rather than as calories. A randomised crossover trial at 40 g/day for eight weeks produced 2.8 kg of weight loss without calorie restriction and improved insulin resistance, via measurable shifts in gut bacteria; separate trials show reductions in visceral fat and in liver fat that persist after adjusting for weight. It raises GLP-1, though not on the scale of GLP-1 drugs. You can generate the useful type at home by cooking and cooling starchy food.

What to actually do:

  1. Batch-cook rice or potatoes and refrigerate overnight. Eat cold or gently reheated — that’s RS3 for free.
  2. If supplementing, ramp over four weeks to 40 g. Skipping the ramp is why people abandon it.
  3. Keep it away from heat — no stirring it into hot drinks or soup.
  4. Take it on an empty stomach or 30 minutes before your largest meal.
  5. Add beans, lentils and whole grains for RS1 alongside.
  6. Pair with resistance training, which independently targets visceral fat.
  7. Judge it on waist and bloods, not the scale — the liver-fat effect was largely weight-independent.

Sources & further reading
#

  • Li H. et al., Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota, Nature Metabolism (2024) — Nature Metabolism / PMC
  • Metabolic phenotypes and the gut microbiota in response to dietary resistant starch type 2 in normal-weight subjects: a randomized crossover trial, Scientific Reports (2019) — Nature / PMC
  • Ni Y. et al., Resistant starch decreases intrahepatic triglycerides in patients with NAFLD via gut microbiome alterations, Cell Metabolism (2023) — Cell Metabolism
  • Effects of resistant starch on metabolic markers and gut microbiota in women with metabolic syndrome risk factors: a randomized, double-blind, pilot study — PMC
  • News-Medical, Resistant starch diet proves a game changer for weight loss and diabetes control — News-Medical

Related

Creatine Is a Multiplier, Not a Source

·1127 words·6 mins· loading · loading
Strength Training Nutrition Biomarkers
Creatine Is a Multiplier, Not a Source # Across 22 randomised trials in older adults, creatine plus resistance training added about 1.37 kg of lean tissue over training alone. That’s a real effect and a small one — and it’s zero if you aren’t lifting anything. What it actually is # Creatine is a compound your body already makes, mostly in the liver, and tops up from meat and fish. Plants contain virtually none, so vegetarians and vegans run lower stores and often notice supplementation more. The supplement itself is synthesised, so it’s vegan-suitable.

Price Moves When Someone Stops Waiting

·856 words·5 mins· loading · loading
Market Cycles Investment Risk Valuation
Price doesn’t move because of news. It moves because somebody decided they’d waited long enough and accepted a worse deal. Every chart you’ve ever looked at is a record of who ran out of patience first. The order book is where price actually happens # Forget the chart for a moment and look at the thing underneath it.

Becoming Ideal Is Still Waiting

·558 words·3 mins· loading · loading
Confidence & Self-Concept Relationships Discipline
Becoming Ideal Is Still Waiting # 45% of men aged 18–25 have never approached a woman in person. 62% of Gen Z men have stepped out of dating altogether. Underneath that is a strategy that feels like effort but isn’t: become impressive enough that the thing you want arrives on its own. The passive fantasy # Jack Reacher is the clean example. Hugely capable, obviously masculine, and romantically completely passive — the women in those books make the first move, and the second, and the third.