Acarbose

What acarbose does in the gut

Starch is a long chain of sugar units. To get into the blood, it has to be broken down in the small intestine; alpha-glucosidase takes care of that. Acarbose inhibits this enzyme. The starch is broken down more slowly and less completely, and the sugar appears in the blood spread out more flatly instead of as a steep peak.

That is the whole mechanism, and it has been understood for decades. What matters is its consequence: the undigested starch does not disappear, it moves on into the large intestine and is fermented there by bacteria. There are good indications that part of the effect runs via the gut microbiome and not only via blood sugar: in treated mice, more short-chain fatty acids were produced at all three sites, and these fermentation products helped predict lifespan. In humans the data are inconsistent: in a four-week crossover study with 40 evaluable people with prediabetes, acarbose shifted the gut flora markedly, among other things in favor of lactobacilli. In a two-week crossover study with 15 people with type 2 diabetes, by contrast, the composition stayed the same apart from small shifts. So this is not known for certain.

Why the mouse finding from the ITP carries more weight

The Interventions Testing Program, ITP for short, is funded by the US National Institute on Aging and is considered the gold standard of the field. The reason is the design: each substance is tested simultaneously at three different sites, with genetically heterogeneous mice, and both sexes must be included.

In longevity research, almost everything founders on the same cliff — one lab finds an effect, another does not. Mouse strains differ, diets differ, housing differs. The ITP eliminates this by testing the same substance in parallel. What gets through there is more robust than a single lab finding.

The number that is almost always left out

In male mice, acarbose extended median lifespan by 22 percent and maximum lifespan by 11 percent. That is one of the strongest findings the program has ever had. In females it was 5 percent for median and 9 percent for maximum lifespan — that is, a fraction.

This is not a fluke. Several ITP hits work preferentially in males, and acarbose is one of them. Why is open; hormones, metabolic differences and the way the animals eat are being discussed. In practice this means: anyone who offers acarbose as a longevity agent and does not mention this number either does not know it or is concealing it.

Who the human data actually apply to

What is established in humans does not apply to everyone but to a specific group: people with impaired glucose tolerance. That is a real benefit for this group — and it is identified by a blood value, not by a feeling. Anyone who wants to know whether this applies to them does not need a peptide shop, but a long-term blood sugar value and a doctor’s appointment.

What the studies show

ACE — 6,522 adults in China, a mean of 5 years

The ACE trial, published in Lancet Diabetes and Endocrinology, included 6,522 adults with impaired glucose tolerance and established coronary heart disease. They received 50 milligrams of acarbose three times daily or placebo, for a mean of 5 years. The primary endpoint was cardiovascular: death, heart attack, stroke, hospital admission for unstable angina or heart failure. Result: no difference, hazard ratio 0.98 with a p-value of 0.73 — practically exactly zero. The second part was positive and is genuine: new-onset diabetes occurred in 13 percent on acarbose compared with 16 percent on placebo, a relative reduction of 18 percent.

Interventions Testing Program — three sites, both sexes

In the ITP, each substance is tested simultaneously at three sites in genetically heterogeneous mice, and males and females are analyzed separately. For acarbose, males showed a 22 percent longer median and an 11 percent longer maximum lifespan. In females the values were 5 and 9 percent. The gap between the sexes is thus larger than the entire effect in females.

What is missing in humans

The question that leads most people to think about acarbose has not been answered in humans. A study meant to show life extension would have to run for decades and include tens of thousands of people. It does not exist for acarbose, and it will not exist. So we have a strong mouse finding, a solid human study on a different question — and between them a gap that remains open.

The cardiovascular benefit is not established either: in the largest study, the primary endpoint came in at a hazard ratio of 0.98 and a p-value of 0.73. It is also unclear how the mouse finding comes about. The flatter blood sugar peak is the obvious assumption and probably incomplete, because part of the effect is likely to run via the gut microbiome. And the finding holds at this strength only for one sex of a different species.

Status, approval and legal

Acarbose is prescription-only and has been approved in Germany as an antidiabetic since 1990; besides Glucobay there are generics. There is no approval for aging or for life extension. The substance is very inexpensive and available in every pharmacy — but only on prescription, and for good reason. Anyone thinking about it because of the mouse finding has a very good topic for their next doctor’s appointment and a poor one for an online order.

Safety

The side effects are inelegant, but they explain themselves: when undigested starch is fermented in the large intestine, gas is produced. Bloating is the most common side effect and the most common reason people stop again. In the ACE trial, gastrointestinal complaints led to discontinuation or dose reduction in 7 percent of participants, compared with 5 percent on placebo. In a study with medical supervision this is manageable; it can be eased by starting slowly and keeping an eye on the amount of starch. It is the reason acarbose was never popular even as a diabetes drug, although it works. On its own, acarbose does not cause low blood sugar; if it occurs in combination with other diabetes drugs, according to the package leaflet only glucose (dextrose) helps, not table sugar, because its breakdown is of course slowed.

Sources

• Holman et al., Lancet Diabetes Endocrinol 2017 (ACE trial)
• Harrison et al., Aging Cell 2014 (acarbose and lifespan in the Interventions Testing Program)
• Garratt et al., Aging Cell 2017 (sex differences in the ITP hits)
• Smith et al., BMC Microbiol 2019 (acarbose, undigested starch and the gut microbiome)
• Harrison et al., Aging Cell 2019 (ITP, acarbose at three dose levels)
• Chiasson et al., Lancet 2002 (STOP-NIDDM)
• Zhang et al., Diabetes Ther 2017 (acarbose and gut flora in prediabetes, crossover over 4 weeks)
• Dalsgaard et al., Endocr Connect 2024 (2 weeks of acarbose, gut flora in type 2 diabetes)