Acarbose Added 22 Percent to Male Mouse Lifespan. The Five-Year Human Trial Was Measuring the Heart.
Most compounds in the longevity conversation have the same problem. The mouse data look good and the drug does not exist as an approved product, so nobody can run the human trial without building a company first. Acarbose has the opposite problem. It has been sold in pharmacies since the 1990s under names like Precose and Glucobay, it costs almost nothing as a generic, its safety profile has been mapped in tens of thousands of patients, and it produced one of the largest male lifespan gains the National Institute on Aging has ever reported. Two large randomized trials of acarbose have already been completed, one with 1,429 participants and one with 6,522. Neither of them measured aging. They measured progression to type 2 diabetes and major adverse cardiovascular events, because those are the questions a diabetes drug gets funded to answer. So the field is left with a strange asymmetry: a drug any physician can already prescribe, an animal result strong enough to survive the hardest replication standard in the discipline, and an empty column where the human aging evidence would go. This is an educational review of what that evidence shows and what it does not.
Key Takeaways: What the Acarbose Evidence Actually Shows
Acarbose is an alpha-glucosidase inhibitor that blocks enzymes in the small intestine from breaking complex carbohydrates into absorbable sugars, which flattens the glucose rise after a meal and pushes undigested starch further down the digestive tract. The National Institute on Aging Interventions Testing Program reported in 2014 that acarbose at 1000 parts per million in the diet, started at 4 months of age, raised median lifespan in genetically heterogeneous UM-HET3 mice by about 22 percent in males and about 5 percent in females, with gains in the 90th percentile of survival in both sexes. Harrison and colleagues reported in Aging Cell in 2019 that a dose-response study at 400, 1000 and 2500 parts per million produced male median lifespan gains of about 16 to 17 percent at the two higher doses, female gains of about 4 to 5 percent, and no significant separation between the three doses, indicating the response saturates at the low end. Strong and colleagues reported that starting acarbose at 1000 parts per million at 16 months of age, roughly late middle age for a mouse, produced only about a 6 percent median lifespan gain in males, though the 90th percentile of survival still rose by about 12 percent. Harrison and colleagues reported in Aging Cell in 2022 that rapamycin combined with acarbose from 9 months of age raised male median lifespan by about 34 percent and female median lifespan by about 28 percent, exceeding what the same dose of rapamycin alone had produced in earlier cohorts of male mice. The 2019 dose-response report also found that acarbose reduced lung tumors in male mice, reduced liver degeneration in both sexes, reduced glomerulosclerosis in females, blunted the blood glucose response to refeeding in males, and improved rotarod performance in aging females. Smith and colleagues reported in BMC Microbiology in 2019 that acarbose shifted gut bacterial communities and that fecal concentrations of acetate, butyrate and propionate together predicted mouse longevity even after controlling for sex, study site and acarbose treatment itself. The same body of work found that acarbose raises fasting FGF21 in the opposite direction from what dietary restriction produces, and that lifespan extension occurred without meaningful weight loss, meaning acarbose is not acting as a calorie restriction mimetic. The STOP-NIDDM trial, published by Chiasson and colleagues in The Lancet in 2002, randomized 1,429 people with impaired glucose tolerance to acarbose 100 milligrams three times daily and found that 32 percent progressed to type 2 diabetes against 42 percent on placebo, a relative risk of 0.75. The Acarbose Cardiovascular Evaluation trial, published by Holman and colleagues in The Lancet Diabetes and Endocrinology in 2017, randomized 6,522 people in China with coronary heart disease and impaired glucose tolerance to 50 milligrams three times daily for a median of five years, found no reduction in major adverse cardiovascular events with a hazard ratio of 0.98, and reduced new-onset diabetes by about 18 percent. Gastrointestinal effects are the practical ceiling on acarbose in humans, with flatulence reported in roughly 74 percent of treated patients against 29 percent on placebo in United States registration trials, diarrhea in about 31 percent against 12 percent, and treatment discontinuation for gastrointestinal reasons in about 13 percent of the acarbose arm in STOP-NIDDM. No trial has ever tested acarbose against a human aging, frailty or survival endpoint, and no trial has tested it in metabolically healthy adults for that purpose.
What Is Acarbose and How Does It Work?
Acarbose is an oral drug that slows carbohydrate digestion by competitively inhibiting alpha-glucosidase enzymes in the brush border of the small intestine, along with pancreatic alpha-amylase. Those enzymes are what convert starch and complex sugars into glucose molecules small enough to cross the intestinal wall. Blocking them does not reduce how many calories a person eats and does not block absorption permanently. It delays and flattens the process, so the sharp glucose peak that normally follows a starch-heavy meal becomes a lower, longer curve. Acarbose is taken with the first bite of a meal, because it has to be present in the gut at the same time as the food it is meant to act on. The dose in human diabetes practice is typically 25 to 100 milligrams three times daily, titrated upward slowly. There is a second consequence that turns out to matter more to the longevity story than the glucose curve does. Carbohydrate that escapes digestion in the small intestine does not disappear. It travels to the colon, where resident bacteria ferment it into hydrogen, carbon dioxide and short-chain fatty acids, principally acetate, butyrate and propionate. That fermentation is the reason acarbose causes flatulence, and it may also be the reason it extends lifespan in mice. The drug is, in effect, a way of feeding the colon a large amount of starch on purpose.
How Much Lifespan Did Acarbose Add in Mice?
Male median lifespan rose by about 22 percent and female median lifespan by about 5 percent in the first Interventions Testing Program report. That 2014 result used 1000 parts per million of acarbose in the diet from 4 months of age in genetically heterogeneous UM-HET3 mice, and the sex gap was the finding that made the compound famous. A dose-response follow-up published in Aging Cell in 2019 tested 400, 1000 and 2500 parts per million at three sites. Every dose produced a significant gain in both sexes, males gained more than females at every dose, and the two higher doses produced male median gains of about 16 to 17 percent against female gains of about 4 to 5 percent. There were no significant differences among the three doses, which is a useful and slightly deflating detail. If tripling the dose does nothing extra, the response is already saturated at the bottom of the tested range, and the interesting variable is not how much acarbose you give. The Interventions Testing Program design is what gives these numbers their weight. Every candidate compound is tested at the same time in three geographically separate laboratories, in a genetically heterogeneous mouse stock rather than a single inbred strain, under an analysis plan specified before the data arrive. Results that survive that design are the closest thing aging research has to a replicated finding, and most candidates entering the program do not survive it.
Twenty-Two Percent, and Then Five
Median lifespan gains reported by the National Institute on Aging Interventions Testing Program for acarbose, split by dietary dose, sex and the age at which feeding began. The first report in 2014 used 1000 parts per million from 4 months of age and raised male median lifespan by about 22 percent against roughly 5 percent in females. A later dose-response study run at 400, 1000 and 2500 parts per million produced male median gains of about 16 to 17 percent at the two higher doses with no significant separation between doses, and female gains of about 4 to 5 percent. Starting the same 1000 parts per million at 16 months of age produced roughly 6 percent in males.
Does Acarbose Still Work If You Start in Middle Age?
Far less well, and this is the number most summaries of acarbose leave out. Strong and colleagues tested acarbose at the same 1000 parts per million starting at 16 months of age rather than 4 months, which corresponds roughly to a person in late middle age rather than a person in their twenties. Male median lifespan rose by about 6 percent, not 22. The 90th percentile of survival still rose by about 12 percent in males, so something real is happening at the far end of the distribution, but the headline median effect largely evaporates. This matters more than the original result does for anyone thinking about translation, because no human being is going to start a longevity drug at the equivalent of 4 months of age. The comparison to make here is with 17-alpha-estradiol, which came through the same program and held roughly 19 percent in males even when treatment began at 16 months. Late-start durability is a property that separates candidates, and acarbose does worse on it than its headline number implies. The combination data point in a more interesting direction. Harrison and colleagues reported in Aging Cell in 2022 that rapamycin plus acarbose from 9 months of age produced about a 34 percent male median lifespan gain and about a 28 percent female gain, exceeding what the same 14 parts per million dose of rapamycin alone had produced in earlier male cohorts. Started at 16 months, the same combination gave about 14 percent in both sexes. Two drugs that work through different mechanisms appear to add rather than overlap, and the late-start penalty still applies to both.
Where Acarbose Sits, and What Happens When You Start Late
Approximate male median lifespan gains from the Interventions Testing Program, which runs every candidate in genetically heterogeneous UM-HET3 mice at three independent laboratories under a pre-specified analysis plan. Rapamycin combined with acarbose from 9 months of age produced the largest gain the program has reported in males at about 34 percent. Acarbose alone from 4 months sits close to rapamycin alone. The two amber bars are the same interventions started at 16 months instead, and the drop is the most practically important number on this chart, because no human takes a longevity drug from the equivalent of early adulthood.
Why Does Acarbose Work Better in Male Mice?
Nobody knows, and the honesty of that answer is worth stating plainly because acarbose is frequently presented as a men's health compound on the strength of a sex difference the field cannot explain. The pattern is real and it is consistent across every acarbose cohort the Interventions Testing Program has run, with males gaining three to five times what females gain at every dose tested. The pattern is also not unique to acarbose. Rapamycin, canagliflozin and 17-alpha-estradiol all produced larger male effects in the same program, and for 17-alpha-estradiol the female effect was absent entirely. What separates acarbose from 17-alpha-estradiol is that the mechanism behind the sex difference has been pinned down for the estrogen and not for acarbose. Castrating male mice abolishes most of the 17-alpha-estradiol response, which points directly at testicular hormones. No equivalent experiment has resolved the acarbose case. Some of the male-specific findings are suggestive rather than explanatory. The 2019 report found that acarbose blunted the blood glucose response to refeeding in males but not females, and reduced lung tumors in males while improving rotarod performance in aging females, which suggests the drug is doing partly different things in the two sexes rather than doing the same thing with different intensity. For a man reading this, the practical implication is narrow. A sex difference in mice with no established mechanism is a reason to pay attention to the male data, not a reason to assume the mechanism carries into human males.
Is the Mechanism the Glucose Curve or the Gut Bacteria?
The evidence currently favors the gut, or at least says the glucose curve is not the whole story. The obvious hypothesis when a drug that flattens post-meal glucose extends lifespan is that the glucose flattening is the mechanism, and there is real support for it, including reduced glucose response to refeeding in treated male mice. But two findings complicate that reading. Acarbose extends mouse lifespan without meaningful weight loss and without the metabolic signature of calorie restriction, and fasting FGF21 moves in the opposite direction from what dietary restriction produces. Whatever acarbose is doing, it is not mimicking eating less. Smith and colleagues reported in BMC Microbiology in 2019 that acarbose treatment substantially restructured the gut bacterial community and increased fermentation products in the colon, which is the expected downstream consequence of delivering undigested starch to bacteria that can use it. The striking part of that paper is the correlation. Fecal concentrations of acetate, butyrate and propionate together predicted mouse longevity even after statistically controlling for sex, study site and acarbose treatment itself, meaning the short-chain fatty acid signal carried information about lifespan beyond simply marking which mice got the drug. That is a correlation rather than a demonstrated causal chain, and the authors present it as such. No experiment has yet shown that blocking short-chain fatty acid production removes the lifespan benefit, which is the experiment that would settle it. So the honest position is that acarbose changes at least three things simultaneously, the post-meal glucose curve, the microbial community in the colon and the fermentation products it generates, and the field has not established which one carries the effect.
Glucose, Gut, or Something Nobody Has Named Yet
A schematic map of the mechanistic evidence behind acarbose and lifespan. Teal marks findings that are well supported, amber marks findings that are real but awkward for the simple story, and grey marks questions the field has not answered. Acarbose reliably blunts the post-meal glucose rise, and Smith and colleagues reported in BMC Microbiology in 2019 that fecal acetate, butyrate and propionate concentrations together predicted mouse longevity even after controlling for sex, site and acarbose treatment itself. The awkward finding is fasting FGF21, which moves in the opposite direction to what dietary restriction produces, meaning acarbose is not simply mimicking a low-calorie diet.
What Happened When Acarbose Was Tested in Humans?
Two large randomized trials were run, both in people with impaired glucose regulation, and neither one asked an aging question. STOP-NIDDM, published in The Lancet in 2002, randomized 1,429 people with impaired glucose tolerance to acarbose 100 milligrams three times daily or placebo. Progression to type 2 diabetes occurred in 32 percent of the acarbose group against 42 percent of the placebo group, a relative risk of 0.75. A secondary analysis published in JAMA the following year reported a large reduction in cardiovascular events, but that analysis rested on a small absolute number of events and drew sustained criticism, which is the context that makes the second trial important. The Acarbose Cardiovascular Evaluation trial, led by Rury Holman at the University of Oxford Diabetes Trials Unit and published in 2017, was designed to test that cardiovascular claim properly. It randomized 6,522 people in China with established coronary heart disease and impaired glucose tolerance to acarbose 50 milligrams three times daily or placebo and followed them for a median of five years across 176 hospitals. It missed its primary endpoint outright, with a hazard ratio of 0.98 for major adverse cardiovascular events. It did reduce new-onset diabetes by about 18 percent, with a number needed to treat of 41 over five years. Read together, the human record is consistent and unglamorous. Acarbose reliably slows the progression from impaired glucose tolerance to diabetes, and it does not appear to prevent heart attacks in people who already have coronary disease. Nothing in either trial speaks to lifespan, frailty, muscle preservation or biological aging, because neither trial was built to.
What the Human Trials Were Actually Measuring
Relative risk reductions from the two large randomized trials of acarbose in people with impaired glucose regulation. STOP-NIDDM randomized 1,429 people with impaired glucose tolerance to acarbose 100 milligrams three times daily and found that 32 percent of the acarbose group progressed to diabetes against 42 percent on placebo, a relative risk of 0.75. The Acarbose Cardiovascular Evaluation trial randomized 6,522 people in China with coronary heart disease and impaired glucose tolerance to 50 milligrams three times daily for a median of five years, missed its cardiovascular primary endpoint outright with a hazard ratio of 0.98, and reduced new-onset diabetes by about 18 percent. Neither trial measured survival to old age, frailty, or any aging endpoint.
What Are the Side Effects, and Who Actually Stays on It?
Flatulence is the defining side effect and it is not a minor footnote. United States placebo-controlled registration trials of acarbose at 50 to 300 milligrams three times daily reported flatulence in about 74 percent of treated patients against 29 percent on placebo, diarrhea in about 31 percent against 12 percent, and abdominal pain in about 19 percent against 9 percent. In STOP-NIDDM, flatulence occurred in 68 percent of the acarbose group against 27 percent on placebo, and about 13 percent of participants discontinued treatment for gastrointestinal reasons. Roughly a quarter of the acarbose group left the trial early overall. These numbers are the direct and predictable consequence of the mechanism, because starch fermented by colonic bacteria produces gas, and the same fermentation that may drive the lifespan effect in mice is what produces the symptom. There are two mitigating details. Registration data indicate that the frequency and intensity of flatulence tend to decrease over time, and that abdominal pain and diarrhea tend to return toward pretreatment levels during a year of use. Standard practice is to start at 25 milligrams with the first bite of one meal per day and titrate slowly upward over weeks, which is the reason dose escalation is written into the labeling. Acarbose also carries a documented signal for elevated liver enzymes at higher doses, which is why liver monitoring appears in prescribing guidance, and it can cause hypoglycemia when combined with insulin or sulfonylureas. A person on acarbose who becomes hypoglycemic needs glucose specifically rather than sucrose, because the drug is actively blocking the enzyme that would break table sugar down.
How Should a Man Read the Acarbose Evidence Right Now?
As the most accessible geroscience candidate in existence attached to an aging evidence base that does not exist yet, which is a frustrating combination rather than a green light. The case for taking acarbose seriously is unusually strong on paper. It passed the Interventions Testing Program with one of the largest male median lifespan gains the program has reported. It passed at three doses across three laboratories. It reduced lung tumors and liver degeneration in treated mice, so the lifespan extension came with pathology improvements rather than simply prolonged decline. It combines additively with rapamycin. It is an approved drug with a decades-long human safety record, a known dose, a known side effect profile and a generic price, which removes every practical barrier that blocks compounds like 17-alpha-estradiol. The case for restraint is equally concrete. The 22 percent figure that circulates in longevity discussions comes from mice started at 4 months of age, and the same dose started at 16 months produced about 6 percent, which is the version of the experiment that resembles a human decision. The female gain was about a fifth of the male gain and nobody can explain why, which means the mechanism is not understood well enough to predict what it does in a human male. The two large human trials of acarbose measured diabetes progression and cardiovascular events, found a reliable effect on the first and none on the second, and said nothing about aging. Three quarters of people who take it get flatulence, and roughly one in eight stop for that reason, which is a serious adherence problem for a drug that would need to be taken with every meal for decades to matter. A crossover study of acarbose in older men with prediabetes has been registered and completed without published results, and until data like that appear the human aging column stays empty. The reasonable posture is to treat acarbose as a strong hypothesis rather than a protocol, to note that it is prescribable today for its actual indication, and to have any conversation about it with a physician who can look at your glucose data and your liver enzymes. This article is educational and is not medical advice, and acarbose is a prescription drug that interacts meaningfully with other diabetes medications.
The Acarbose Record on One Axis
Every strand of the acarbose case arranged as a single schematic map. Teal marks what the evidence supports, amber marks findings that are real but narrower or more conditional than the headline suggests, slate marks results that limit who the drug could help, and dark grey marks claims the evidence does not carry. Acarbose is the inverse of most longevity compounds. It is an approved, cheap, well-characterized human drug with an unusually strong animal lifespan result, and nobody has ever run the trial that would connect the two.
Frequently Asked Questions
Have more questions? Contact us or explore our full peptide intelligence platform.
Related Articles
Does Telomere Length Determine Lifespan? Short Telomeres Track With Earlier Death. The Genetics Say Longer Ones Raise Cancer Risk.
Does telomere length determine lifespan? Short telomeres track with earlier death, yet genetic studies show longer ones raise the risk of several cancers.
Low-Dose Aspirin Added 8 Percent to Male Mouse Lifespan. In 19,114 Healthy Older Adults, It Added No Healthy Years and Deaths Ran Higher.
Does low-dose aspirin help longevity? It extended male mouse lifespan 8%, but the ASPREE trial in 19,114 healthy older adults found no gain and more bleeding.
Is NAD+ a Peptide? No. It Is a Coenzyme Built From Two Nucleotides, and the Peptide Label Comes From the Clinic, Not the Chemistry
Is NAD+ a peptide? No. NAD+ is a coenzyme built from two nucleotides. What NAD+ is used for, what IV NAD+ studies found, and how it differs from SS-31.