Why Can Humans Digest Starch but Not Cellulose?

Both starch and cellulose are built from the same raw material, glucose, but they link those glucose units together with different chemical bonds, and that single difference is the reason your body handles them so differently. Human saliva and pancreatic juice contain amylase enzymes that snap the bonds in starch with ease, releasing glucose for energy. Those same enzymes are physically unable to grab onto the bonds in cellulose, so plant fiber passes through you mostly intact. The distinction comes down to molecular geometry so slight that it sounds trivial, yet it shapes everything from how your gut works to how entire ecosystems of herbivores have evolved.

The Same Building Block, Two Different Bonds

Glucose is a six-carbon sugar that can arrange itself in two mirror-image shapes. When glucose units snap together into a long chain, the orientation of the bond between each pair matters enormously. In starch, the bond angles downward from each glucose ring in what chemists call an alpha linkage. In cellulose, the bond juts outward in a beta linkage. That flip changes everything about the chain’s behavior.

Starch chains (particularly the form called amylose) are relatively flexible and tend to coil into loose helices in water, with imperfect internal hydrogen bonding that leaves the molecule open and accessible to enzymes.¹ Cellulose chains, by contrast, are intrinsically stiff. Their beta linkages allow every other glucose ring to flip, creating a flat, ribbon-like strand. The hydroxyl groups along that ribbon form a regular, tight pattern of hydrogen bonds both within and between chains, locking thousands of cellulose strands into rigid, crystalline sheets. These sheets stack together into fibers tough enough to hold up a tree trunk. Simulations confirm that even when additional electronic forces are modeled, cellulose barely changes shape, while amylose readily folds into more compact forms.¹

Think of starch as a loosely coiled garden hose that you can easily cut at any point, and cellulose as a bundle of steel cables welded side by side. The raw material is the same, but the architecture is not.

How Human Amylase Breaks Down Starch

Digestion of starch begins the moment food touches your tongue. Salivary amylase starts chopping the long starch chains into shorter fragments while you chew. Once the food reaches your small intestine, your pancreas releases a much larger dose of alpha-amylase that finishes the job, cutting starch into maltose and other small sugars that intestinal enzymes then split into individual glucose molecules for absorption.

Human pancreatic alpha-amylase has an active site that spans about five glucose-binding positions, and it works most efficiently when it can seat a chain across all of them.² The enzyme uses a specific amino acid (Asp197) as its chemical “knife,” and swapping out that residue drops catalytic power by roughly a millionfold.² Another surprising detail: the enzyme depends on a chloride ion sitting near its active site to work properly. Replace that chloride with different ions and activity changes dramatically; the azide ion, for example, can boost the enzyme’s performance nearly fivefold above normal chloride levels.³

The critical point is that amylase’s active site is shaped to grip alpha-linked glucose chains. It simply cannot accommodate the geometry of a beta linkage. The enzyme and the bond are like a key and the wrong lock. And humans produce no endogenous cellulase, the class of enzyme needed to break beta-1,4 bonds, anywhere along the digestive tract.

Why No Animal Makes Quick Work of Cellulose Alone

It is tempting to think of cellulose digestion as something that “other animals can do and we can’t.” The reality is more nuanced. Very few animals produce their own cellulase enzymes, and even those that do typically rely on microbial partners to get meaningful energy from cellulose.

Ruminants like cows and sheep are the textbook example. They do not digest cellulose themselves. Instead, their multi-chambered stomachs house dense communities of bacteria, protozoa, and fungi that ferment cellulose on the animal’s behalf.⁴ The microbes break down the plant cell walls and produce short-chain fatty acids that the cow absorbs for energy. This is not a fast process; a cow may spend eight hours a day chewing and rechewing its cud to give those microbes enough time and surface area to work.

Horses take a different approach. As hindgut fermenters, they digest what they can with their own enzymes in the stomach and small intestine, then pass the leftovers into an enlarged cecum and colon where microbial fermentation occurs. More than half of a horse’s maintenance energy comes from that microbial back end.⁵ The tradeoff is efficiency: horses extract less energy from the same forage than ruminants do, which is why horses need to eat more.

Termites are a fascinating exception. Some species do produce their own cellulase enzymes in their midgut tissue, rather than outsourcing the job entirely to gut microbes.⁶ In fact, about three-quarters of termite species lack the flagellate protists long thought to be responsible for cellulose breakdown; they rely instead on endogenous enzymes secreted from their own midgut cells.⁶ One species, Reticulitermes speratus, expresses two endoglucanases and a beta-glycosidase that together handle all stages of cellulose breakdown.⁷ Even in higher termites that lack symbiotic flagellates, cellulase activity is concentrated in the midgut, secreted by the termite’s own tissue.⁸ These are among the rare animals that genuinely make their own cellulose-cutting tools, and they evolved that ability over hundreds of millions of years of eating wood.

Humans never faced that selection pressure. Our primate ancestors were fruit eaters and later omnivores; starch from tubers, seeds, and grains became a primary carbohydrate source, and the evolutionary investment went into better starch digestion, not cellulose digestion.

The AMY1 Story and Starch in Human Evolution

If you want evidence that starch digestion mattered to human survival, look at the gene for salivary amylase, AMY1. Most people carry multiple copies of this gene, and the number varies widely across populations. People from populations with traditionally high-starch diets tend to carry more AMY1 copies, and more copies mean more salivary amylase protein.⁹

For a long time, researchers assumed this gene duplication happened recently, driven by the invention of agriculture roughly ten thousand years ago. But more detailed genetic analysis tells a deeper story. Haplotypes carrying three AMY1 copies appear across all continents and make up about 70% of observed haplotypes, suggesting they originated before modern humans left Africa.¹⁰ Analysis of archaic hominin genomes, including Neanderthals and Denisovans, found additional AMY1 copies already present, pushing the initial duplication back as far as 800,000 years ago.¹⁰ In other words, the genetic groundwork for efficient starch digestion was laid long before farming existed. Agriculture then accelerated the process, increasing the frequency of higher-copy haplotypes in populations eating grain-heavy diets.

Researchers have also argued that cooked starch, as a readily available source of glucose, played a direct role in fueling the metabolic demands of our unusually large brains. The human brain consumes roughly a quarter of the body’s glucose at rest, and preformed glucose from starchy tubers and roots, especially once cooking made them more digestible, would have been a reliable energy supply for early hominins.¹¹ This is a compelling complement to the more widely known “meat made us human” narrative: meat provided protein and fat, but cooked starch may have provided the glucose.

What Actually Happens to Cellulose in Your Gut

Just because your own enzymes cannot break cellulose does not mean it vanishes into a void. Cellulose is a major component of what we call dietary fiber, and it does real, useful work as it travels through you. Insoluble fiber like cellulose adds bulk to stool and speeds transit through the intestines.¹² This mechanical effect is one reason high-fiber diets are associated with better bowel regularity. Studies in animals confirm that adding cellulose-rich fiber to the diet significantly increases fecal bulk and shortens the time food takes to travel through the gut.¹³

But the story has a microbial chapter, too. Your colon hosts a vast community of bacteria that ferment carbohydrates your own enzymes missed, producing short-chain fatty acids like acetate, propionate, and butyrate.¹⁴ These fatty acids serve as fuel for the cells lining your colon, help regulate inflammation, and influence metabolism well beyond the gut. Both resistant starch (starch that escapes digestion in the small intestine) and nonstarch polysaccharides like cellulose contribute to this fermentation.¹⁵

So while you do not “digest” cellulose in the traditional sense of breaking it down into glucose and absorbing it, your gut bacteria can partially ferment it, and the products of that fermentation feed your colon and affect your health. The distinction between “indigestible” and “useless” is an important one that popular nutrition writing often blurs.

Cellulose-Degrading Bacteria Living Inside You

Here is where things get genuinely surprising. For a long time, the assumption was that the human gut simply lacked the microbial machinery for serious cellulose breakdown. Recent research has upended that idea. Scientists have identified Ruminococcus species in the human gut that assemble functional cellulosome structures, multi-enzyme complexes specifically designed to degrade plant cell wall polysaccharides including cellulose.¹⁶ One of these species likely originated in the ruminant gut and jumped to humans, possibly during the domestication of livestock, after which it diversified and acquired genes from other human gut microbes.¹⁶

Earlier culture-based work had already found Ruminococcus-like bacteria capable of degrading microcrystalline cellulose in the colons of some human subjects, particularly those who produced methane during breath tests.¹⁷ These isolates turned out to be previously undescribed Ruminococcus species, phylogenetically related to R. callidus and R. flavefaciens, both known cellulolytic bacteria in ruminant guts.¹⁷

Structural prediction tools have expanded the picture further. Using AlphaFold to predict protein structures across whole microbial proteomes, researchers uncovered six additional Ruminococcus species in the human gut that encode cellulosome machinery invisible to older gene-sequencing methods.¹⁸ One species, R. difficilis, appears to carry an unusual assembly enriched in amylases and starch-binding proteins, which may enable it to tackle resistant starches that escape digestion in the upper gut.¹⁸

None of this means humans are secret cellulose digesters. The amounts of cellulose these bacteria can handle are modest, and the energy extracted is small compared with what a cow’s rumen achieves. But the discovery reframes the old “humans can’t digest cellulose” claim from an absolute to more of a spectrum. We do carry some of the microbial hardware, and how much of it any individual person has likely depends on their diet, their microbial history, and possibly even whether their ancestors kept livestock.

Cooking and the Starch Digestibility Gap

Raw starch is harder to digest than cooked starch, a fact that connects back to the same structural principles. In its native state, starch granules are semi-crystalline, with tightly packed regions that resist enzyme access. Cooking gelatinizes starch: heat and water break apart the crystalline structure, swelling the granules and exposing the alpha-linked chains to amylase attack. This is why a raw potato is unpleasant and hard to digest, while a baked potato is soft and yields its glucose readily.

When cooked starch cools, some of it re-crystallizes into what is called resistant starch, a form that again resists digestion by amylase and behaves more like fiber. This is the reason cold pasta and day-old rice have a slightly different nutritional profile than freshly cooked versions. That resistant starch travels to the colon, where bacteria ferment it into short-chain fatty acids just as they do with cellulose.¹⁵

The importance of cooking in the evolutionary picture is hard to overstate. Researchers have argued that the control of fire and the practice of cooking starchy foods dramatically increased the glucose available to early human tissues, particularly the brain, red blood cells, and developing fetuses.¹¹ Cooking did not change the bond chemistry that distinguishes starch from cellulose, but it made starch’s already-digestible bonds far more accessible, amplifying the energy payoff of a starchy meal.

Cellulose Outside the Plate

Given that humans cannot metabolize cellulose for energy, it may seem odd that the substance shows up in a range of consumer products. Cellulose is the most abundant organic polymer on Earth, and its very inertness in the human body makes it useful. Microcrystalline cellulose, produced by treating plant fiber with acid to break it into tiny crystalline fragments, is widely used as a filler and binder in pharmaceutical tablets.¹⁹ It holds pills together, helps them dissolve at the right rate, and passes harmlessly through the gut if any remains undigested. You will also find cellulose-derived additives in shredded cheese (to prevent clumping), ice cream (as a stabilizer), and even some low-calorie foods, where its bulk and indigestibility are the point.

On the industrial side, breaking cellulose down into fermentable sugars is one of the central challenges of biofuel research. Because the beta-1,4 bonds are so resistant to hydrolysis, producing ethanol from agricultural waste like wheat straw or corn stalks requires aggressive pretreatment, whether chemical, thermal, enzymatic, or some combination.²⁰ Researchers have been engineering fungal cellulases, particularly from the mold Trichoderma reesei, and combining them with commercial enzyme cocktails to improve glucose release from cellulose substrates. The same beta linkage that makes cellulose useless to your digestive system is what makes lignocellulosic biofuels so expensive and difficult to produce at scale. If cellulose were as easy to break down as starch, plant waste would already be a dominant fuel source.

There is an appealing symmetry in the fact that the same molecular feature that shaped human evolution, our ability to exploit starch and our inability to exploit cellulose, now drives two very different industries. The pharmaceutical industry values cellulose precisely because it resists digestion. The biofuel industry is trying everything it can to overcome that resistance. Both are working with the consequences of one flipped chemical bond.