What Is the Best Treatment for Type 2 Diabetes?

There is no single best treatment for type 2 diabetes that works for everyone, because the disease itself varies enormously from person to person. What has changed dramatically in recent years is the goal of treatment: diabetes care has shifted from focusing narrowly on blood sugar numbers toward protecting the heart, kidneys, and other organs at the same time. Metformin remains the most common starting medication, but newer drug classes and a growing emphasis on early, aggressive combination therapy have reshaped what “best” looks like in practice.

Metformin Still Comes First for Most People

Metformin has been the default first-line drug for type 2 diabetes for decades, and it holds that position for good reason: it lowers blood sugar effectively, has a long safety track record, and is inexpensive. Its primary effect is reducing the amount of glucose your liver releases into the bloodstream.¹ Researchers have also found increasing evidence that metformin works partly through the gut, not just the liver, which may explain some of its broader metabolic effects.²

For someone newly diagnosed with mildly elevated blood sugar, metformin alone combined with lifestyle changes can be enough to bring things under control. It does not cause low blood sugar on its own, which is a meaningful advantage over some older drugs. The most common complaints are gastrointestinal: nausea, diarrhea, and stomach upset, which usually improve over a few weeks or with an extended-release formulation. For a large share of people with type 2 diabetes, though, metformin alone eventually stops being sufficient, and the question becomes what to add next.

GLP-1 Receptor Agonists Have Changed the Landscape

If one drug class has transformed type 2 diabetes treatment in the past decade, it is the GLP-1 receptor agonists. Names like semaglutide (Ozempic, Wegovy) and liraglutide (Victoza) have become widely known, partly because of their effectiveness for weight loss in addition to blood sugar control. These medications mimic a natural gut hormone that stimulates insulin release when blood sugar is high, slows stomach emptying, and reduces appetite. They lower HbA1c, promote weight loss, and have cardiovascular protective effects.³

GLP-1 receptor agonist therapy is now recommended early in treatment guidelines, not just as a fallback when metformin fails, owing to its combined benefits on blood sugar, weight, and cardiovascular outcomes.⁴ The newer dual agonist tirzepatide (Mounjaro) targets both the GLP-1 and GIP receptors, delivering even larger reductions in blood sugar and body weight in clinical trials. The main downsides are cost, the need for injection (though oral formulations are emerging), and gastrointestinal side effects like nausea that tend to fade as the body adjusts.

SGLT2 Inhibitors Protect the Heart and Kidneys

SGLT2 inhibitors, drugs like empagliflozin (Jardiance), dapagliflozin (Farxiga), and canagliflozin (Invokana), work by blocking glucose reabsorption in the kidneys so that excess sugar leaves the body through urine. That mechanism alone is relatively modest for blood sugar control, but the real story with these drugs is what they do beyond glucose. They reduce blood pressure, promote mild weight loss, and, most strikingly, protect against heart failure and kidney disease progression.⁵

The protective mechanisms go well beyond removing sugar from the blood. SGLT2 inhibitors lower pressure inside the kidney’s filtering units, reduce the oxygen demand of kidney tissue, act as mild diuretics that ease fluid overload on the heart, and stimulate the production of red blood cells, improving oxygen delivery throughout the body.⁵ These benefits are so consistent that SGLT2 inhibitors are now used in people with heart failure or chronic kidney disease even when they do not have diabetes at all.⁶ If you have type 2 diabetes and existing heart or kidney problems, an SGLT2 inhibitor is likely to be part of your regimen regardless of your blood sugar numbers.

The Shift from Blood Sugar to Organ Protection

For years, treating type 2 diabetes meant one thing: get the HbA1c number down. That glucocentric approach has given way to something broader. Multiple cardiovascular outcome trials with GLP-1 receptor agonists and SGLT2 inhibitors showed that these drugs significantly improved heart and kidney outcomes, and that benefit was not explained purely by their glucose-lowering power. This triggered what researchers describe as a paradigm shift: the primary focus of diabetes treatment moved from HbA1c targets to protecting the cardiovascular and renal systems.⁷

Major diabetes associations worldwide, including the Korean Diabetes Association and others, have formally adopted this organ-protection framework, and it is filtering into primary care practice.⁸ In practical terms, this means your doctor should be thinking about your heart and kidney risk profile when choosing medications, not just picking whichever drug lowers your HbA1c the most. Someone with early signs of kidney damage may be started on an SGLT2 inhibitor even if their blood sugar is only modestly elevated, while someone with a history of cardiovascular events may benefit most from a GLP-1 receptor agonist.

Why Starting with Combination Therapy May Be Smarter

The traditional approach to type 2 diabetes has been stepwise: start metformin, wait, and add a second drug only when the first one fails. A growing body of evidence suggests that waiting is the wrong strategy for many people. A meta-analysis found that starting with two drugs from the outset lowered HbA1c by about half a percentage point more than stepwise escalation and nearly doubled the time before treatment failure, from roughly 36 months to roughly 62 months.⁹

Critically, early combination therapy preserved the insulin-producing capacity of the pancreas for about three and a half years, compared with a steady decline under stepwise treatment. The VERIFY trial, a five-year randomized study of metformin plus vildagliptin versus metformin alone, confirmed this: patients who started on combination therapy had a significantly longer time before their treatment stopped working, with a halved risk of initial treatment failure.¹⁰ The implication is that early combination may actually modify the course of the disease rather than just managing symptoms, which makes the old “start low, go slow” philosophy look increasingly outdated for people diagnosed with moderately high blood sugar.

Sulfonylureas and DPP-4 Inhibitors as Second-Tier Options

Sulfonylureas like glipizide and glyburide have been around for decades and are very cheap, which keeps them in wide use globally. They work by stimulating the pancreas to release more insulin, regardless of blood sugar level. That blunt mechanism is their main drawback: they cause hypoglycemia more often than newer alternatives. A meta-analysis comparing sulfonylureas to DPP-4 inhibitors (drugs like sitagliptin and linagliptin that boost insulin in a glucose-dependent way) found that when combined with metformin, DPP-4 inhibitors were associated with far fewer hypoglycemic episodes, fewer cardiovascular events, and fewer adverse events overall.¹¹

A separate study found that sulfonylureas, compared with DPP-4 inhibitors, carried roughly double the risk of severe hypoglycemia and a higher risk of cardiovascular events and death.¹² Current international consensus gives preference to DPP-4 inhibitors over sulfonylureas except when cost is the overriding concern. DPP-4 inhibitors do not cause weight gain or hypoglycemia, though they lack the dramatic cardiovascular benefits seen with GLP-1 receptor agonists and SGLT2 inhibitors. They occupy a useful middle ground for people who need modest additional blood sugar lowering without injectable therapy.

When Insulin Becomes Necessary

Type 2 diabetes is a progressive disease. Over time, the insulin-producing beta cells in the pancreas wear out, and eventually many people need insulin injections to keep their blood sugar in a safe range.¹³ This is not a failure on anyone’s part. It reflects the underlying biology of the disease.

Insulin therapy usually begins with a once-daily long-acting injection (like insulin glargine or degludec) added to existing oral medications. The dose is adjusted gradually based on fasting blood sugar readings. Some people eventually need mealtime (rapid-acting) insulin as well, though the newer non-insulin drugs have delayed or reduced that need for many patients. The biggest practical challenges with insulin are weight gain, hypoglycemia risk, and the daily routine of injections and monitoring. But when blood sugar is very high or when other medications are no longer enough, insulin remains the most powerful tool available.

Diet, Exercise, and the Possibility of Remission

Lifestyle changes are not just a polite prelude to “real” treatment. For people diagnosed relatively early, intensive dietary interventions can produce actual remission, meaning blood sugar returns to normal levels without medication. Two large randomized trials of very low-calorie diets found that the approach led to sustained remission in roughly half of participants at one year, and about a third were still in remission at two years.¹⁴ The key mechanism appears to be rapid weight loss, which reduces fat deposits in the liver and pancreas and allows beta cells to recover some function.

Exercise complements dietary changes by directly improving how well your cells respond to insulin. Physical training has been shown to ameliorate insulin action in insulin-resistant individuals through complex molecular improvements at the cellular level.¹⁵ A combination of aerobic and resistance exercise is generally recommended, though any increase in physical activity helps. Remission is most achievable in people who are diagnosed recently, still have meaningful beta-cell function, and can achieve significant weight loss. For someone who has had diabetes for 15 years and is on multiple medications, remission through diet alone is unlikely, but lifestyle improvements still reduce cardiovascular risk and medication needs.

Metabolic Surgery for Severe Cases

Bariatric surgery (now often called metabolic surgery when performed specifically for diabetes) produces the most dramatic and durable improvements in blood sugar of any intervention. A study following patients for 12 years found that the surgical group maintained about a 1.1 percentage point greater HbA1c reduction compared with medical management alone, and roughly 13% were in diabetes remission at 12 years compared with zero percent in the medical-management group.¹⁶

A meta-analysis pooling longer-term cohort data showed even more striking figures: surgery was associated with a nearly six-fold increase in diabetes remission rates, a major reduction in both small-vessel and large-vessel complications, and substantially lower mortality compared with non-surgical treatment over at least five years of follow-up.¹⁷ Surgery is generally considered for people with a BMI over 35 who have not achieved adequate blood sugar control with medications, though some guidelines have lowered the threshold. It is not a casual decision: it is major surgery with lifelong nutritional consequences, but for the right patient, the metabolic benefits are unmatched by any drug.

Continuous Glucose Monitoring Is Not Just for Insulin Users

Continuous glucose monitors, small sensors worn on the skin that track blood sugar in real time, were originally developed for people on insulin. Their use has expanded, and there is now solid evidence that they help people with type 2 diabetes who are not on insulin at all. A meta-analysis of eight randomized trials found that CGM use in non-insulin-treated patients was associated with a meaningful reduction in HbA1c and a significant increase in the percentage of time spent in the target blood sugar range.¹⁸

The value of CGM for non-insulin users likely comes from behavioral feedback: when you can see your blood sugar spike after a particular meal or drop during a walk, you learn which habits matter in a way that periodic fingerstick checks cannot replicate. The same meta-analysis noted benefits in cost-effectiveness and reduced use of healthcare resources. Insurance coverage for CGM in non-insulin-treated patients remains inconsistent in many countries, but the clinical case for broader access is strengthening.

Liver Disease as a Hidden Treatment Target

Metabolic dysfunction-associated steatotic liver disease (what used to be called fatty liver disease) is extremely common in people with type 2 diabetes, and it often goes undiagnosed. The relationship is bidirectional: insulin resistance drives fat into the liver, and a fatty liver worsens insulin resistance. This has made liver health an increasingly important consideration when choosing diabetes medications.

GLP-1 receptor agonists appear to have a direct benefit here. A meta-analysis found that GLP-1 receptor agonist treatment significantly reduced liver fat, improved liver enzyme levels, and in patients with biopsy-confirmed liver inflammation, improved the key features of liver damage: fat accumulation, inflammation, and ballooning of liver cells. Tirzepatide showed particularly robust evidence, and semaglutide had the clearest effect on liver stiffness, a marker of scarring.¹⁹ A separate study using real-world data found that GLP-1 receptor agonist users had a lower rate of liver fibrosis progression compared with those on DPP-4 inhibitors.²⁰ For someone with both type 2 diabetes and fatty liver disease, a GLP-1 receptor agonist is increasingly looking like a two-for-one treatment.

Who Gets the Best Treatments and Who Does Not

One of the most uncomfortable realities of type 2 diabetes care is that access to the most effective newer medications is deeply unequal. In the United States, Black and Hispanic patients, those with lower incomes, those with less education, and those on public insurance or without insurance are consistently less likely to be prescribed GLP-1 receptor agonists and SGLT2 inhibitors compared with White, higher-income, privately insured patients.²¹ Per-prescription spending on these drugs has risen by nearly half, making cost an even steeper barrier.

A systematic review confirmed the pattern: patients with lower socioeconomic status had roughly 27% lower odds of being prescribed these newer therapies, and those on Medicaid, Medicare, or Medicare Advantage had significantly reduced odds compared to privately insured patients.²² Lower income at the ZIP-code level was independently associated with lower initiation rates, even after accounting for other factors.²³ This is not a minor gap in convenience; these are drugs with proven ability to prevent heart attacks, kidney failure, and death. The populations least likely to receive them are the populations at highest risk for those complications.

Precision Medicine and the Future of Treatment Matching

One reason the “best treatment” question has no universal answer is that type 2 diabetes is probably not a single disease. A clustering analysis using six clinical variables broke diabetes into five distinct subgroups, each with different risk profiles and likely different optimal treatments. One cluster, characterized by severe insulin resistance, had the highest risk of kidney disease and fatty liver disease. Another, marked by insulin deficiency, had the highest risk of eye damage.²⁴ If you are in the insulin-resistant cluster, an SGLT2 inhibitor that protects your kidneys may be more valuable than a drug that primarily boosts insulin secretion. If you are in the insulin-deficient cluster, early insulin therapy might prevent retinopathy more effectively.

Clinicians do not yet routinely classify patients into these subgroups, but the research is moving in that direction. Future models will likely integrate genetic markers, metabolic profiles, and possibly gut microbiome data to match individuals with specific therapies. Metformin itself, for instance, has been shown in a randomized trial to alter the gut microbiome, increasing certain bacterial species and changing the production of short-chain fatty acids, which may partially explain its metabolic effects beyond simple glucose lowering.²⁵ Understanding which patients respond best to which drug based on their individual biology, rather than following the same algorithm for everyone, is the long-term promise of precision diabetology. We are not there yet, but the tools are emerging faster than most people realize.