Antibiotics can cure. They kill infectious bacteria and save lives. Type 2 diabetes is a chronic disease. It shortens lives.

Now there is disturbing evidence that the cure may be contributing to the disease—in other words, certain antibiotics may increase the risk of developing diabetes.

The connection is the ecosystem of bacteria in our gut that scientists call the microbiome. It affects digestion and immunity, and an unhealthy microbiome has been linked to diseases as diverse as obesity, certain cancers, inflammatory bowel disease, rheumatoid arthritis and…diabetes. Several studies have shown that type 2 diabetes, the kind that affects most people, is more common in people who have microbiomes with altered or low bacteria diversity. What we eat and drink changes the composition of the bacteria, and so can the medication we take…especially antibiotics.

Penicillin, the original wonder drug, saved soldiers from battlefield infections in World War II and later revolutionized medicine by curing once fatal infections. But antibiotics by their very nature disturb the microbiome by killing bacteria…including beneficial bacteria in the gut.

Now, the newest research finds an association between the repeated use of certain antibiotics and the diabetes epidemic that affects 30 million Americans…and counting.


In the latest study, researchers had access to nearly complete medical records of almost 10 million people living in the United Kingdom. The records included medical diagnoses, tests and procedures, prescription medications and lifestyle factors, including smoking and drinking history.

The research team identified 208,002 people who were diagnosed with diabetes (either type 1 or 2). Each case was matched with four controls…people of the same age and gender who did not have diabetes. In all, the study included more than one million men and women, with an average age of 60.

Looking deeper into the medical records of the participants, the researchers searched for prescriptions for several different antibiotics, including, yes, penicillin, still the most popular choice. They excluded antibiotics prescribed in the year before a diabetes diagnosis, since many of these patients may have had undiagnosed diabetes already. They adjusted statistically for many variables, including smoking, high cholesterol, obesity, heart disease, skin and respiratory infections, and previous blood sugar measurement. The results…

• In most cases, a single course of antibiotics was not associated with any increased risk for diabetes, compared with taking no antibiotics at all.

• The exception was a class of antibiotics called cephalosporins, broad-spectrum antibiotics often prescribed for strep throat and UTIs. Even taking a single course of these antibiotics was associated with a 9% increase in type 2 diabetes risk.

• For the antibiotics linked with type 2 diabetes, the more courses people took in any one year, the greater the risk. Taking two to five courses of penicillin in a single year raised diabetes risk 8%, for example, while taking more than five courses raised risk by 23%. Similarly, taking two to five courses of quinolones, prescribed for skin and respiratory infections as well as UTIs, raised diabetes risk 15%, while taking more than five courses raised risk 37%.

• Tetracyclines raised type 2 diabetes risk only in people who took them for five or more courses in a year.

• Nitroimidazoles, prescribed for vaginal infections as well as skin infections such as rosacea, were not associated with increased diabetes risk when taken at any frequency.

• Neither antiviral nor antifungal medications were linked with diabetes risk.

• While there was an increased risk for type 1 diabetes, an autoimmune condition, for some antibiotics, the results were inconclusive.


This study, while big and statistically powerful, doesn’t tell us whether using antibiotics actually causes diabetes. That’s because it’s observational. It looks back and draws connections. A prospective study would assign one group of people to take antibiotics whether they need them or not, and deny them to another group, and follow them for years to see who gets diabetes. For practical and ethical reasons, of course, that’s impossible.

So it’s possible that people who would go on to develop diabetes even years later are more prone to infections, and so would need more antibiotics. On the other side, prospective animal studies have shown that antibiotics promote the growth of bacteria that promote diabetes. Because diabetes is so common and such a damaging disease, researchers are looking for other ways to tease out whether and how antibiotics contribute to diabetes.

You and I don’t have to wait to do the right thing, though. We’ve overused these wonder drugs, both for human medicine and animal livestock, and many are losing their effectiveness due to rising antibiotic resistance, a scary prospect. Using antibiotics only when they are really needed not only protects your own health but helps keep these drugs effective when they are really needed.

By all means take an antibiotic if it’s the right treatment. But we already know many good reasons to avoid antibiotics if possible, and the truth is, they are often prescribed for health conditions for which they can’t possibly work. Antibiotics kill bacteria, so they won’t help with, say, the common cold, which is caused by a virus. Most sinus infections, even those caused by bacterial infections, don’t require antibiotics either.

To learn more about ways to use antibiotics safely and sparingly, see Bottom Line’s The Wrong Antibiotic Can Be Life Threatening,  Do You Really Need an Antibiotic?,  Dangerous Antibiotics, and Say No to Antibiotics for Traveler’s Diarrhea.

In many cases, doctors prescribe antibiotics when they’re not needed because a patient insists on it for almost any sort of infection or even suspected infection.

Don’t be that patient!