Key takeaways
- In the federally funded Look AHEAD trial of 5,145 adults with type 2 diabetes, an intensive lifestyle program didn’t lower heart attacks or strokes, but it improved weight, blood sugar, mobility and mood12.
- In the company-funded SELECT trial, weekly semaglutide cut heart attacks, strokes and heart deaths from 8.0% to 6.5% over about 3.3 years, or 15 fewer per 1,000 people by our math4.
- Benefits run on different clocks: in pooled statin trials, about 2.5 years passed before 1 in every 100 people treated avoided a major heart event7.
- The American Diabetes Association’s 2026 standards call for starting diabetes medicine at diagnosis, and, when weight loss or lifestyle changes lower a person’s needs, for cutting back first on the drugs most likely to cause side effects or low blood sugar8.
- In a Mayo Clinic trial, a decision aid showing patients their own risk raised their knowledge and cut their uncertainty10.
In June 2013, researchers reported why a federally funded study of weight loss and heart health had been halted years ahead of schedule. The Look AHEAD trial had followed 5,145 adults with type 2 diabetes and overweight or obesity at 16 U.S. centers1.
Half joined an intensive lifestyle program: weekly coaching at first, a lower-calorie eating plan and at least 175 minutes of activity a week. The other half got diabetes support and education classes a few times a year2.
The plan allowed up to 13.5 years of follow-up. After a median of 9.6, the trial’s safety board recommended stopping. The reason was futility: heart attacks, strokes and heart deaths weren’t dropping13.
In November 2023, a drug trial reported a different result. In SELECT, weekly shots of the drug semaglutide cut heart attacks, strokes and heart deaths by 20% in people with heart disease and excess weight46.
Read as headlines, that’s a loss for lifestyle and a win for a drug. Read as numbers, the story is closer. The two kinds of evidence come in different currencies, and this is how to convert them, with a sheet to bring to your next visit.
A decade of coaching didn’t prevent heart attacks. It did plenty else
After one year, people had lost 8.6% of their body weight, versus 0.7% in the comparison group. By the end, the gap had narrowed to 6.0% versus 3.5%1.
Their blood sugar improved more. So did their fitness and, at first, every heart risk factor except LDL cholesterol1. They were less likely to need insulin in the first year and more likely to reach partial remission of diabetes in the first four years3.
Early on, the coached group also reported fewer bladder leaks, less sleep apnea and less depression. People felt better day to day and got around more easily2.
But the main target didn’t move. That was a combined count of heart attacks, strokes, heart deaths and hospital stays for chest pain. There were 403 such events in the lifestyle group and 418 in the comparison group, or 1.83 versus 1.92 per 100 people per year1.
By our math, that’s about 0.9 fewer events per 1,000 people a year, a gap small enough to be chance1. The investigators wondered whether more weight loss, or another eating pattern such as a Mediterranean diet, would have done better3.
Hertzel Gerstein, who wrote an editorial on the trial, still saw a firm place for lifestyle in diabetes care. Eating and activity changes, he argued, can safely trim pounds and prescriptions, and now and then put diabetes into remission2.
A weekly shot cut heart events by 20%, and per 1,000 people, that’s 15
SELECT enrolled 17,604 adults 45 and older in 41 countries. All had heart disease and a body mass index of 27 or higher, and none had diabetes46. About three in four had already had a heart attack5.
Half got a weekly injection of semaglutide, and half got a placebo shot4. Everyone kept their usual heart care, such as cholesterol and blood-thinning medicines6. Most, 88%, were already taking a statin5.
Over an average of 39.8 months, 8.0% of the placebo group had a heart attack, stroke or heart death, versus 6.5% on semaglutide4. That’s the 20% relative cut.
By our math, it’s 15 fewer events for every 1,000 people over about 3.3 years. Put another way, about 67 people would take the drug that long for one to avoid an event.
Deaths from any cause were 4.3% versus 5.2%. Heart deaths alone, 2.5% versus 3.0%, didn’t differ by a statistically clear margin5. At two years, people on the drug had lost 9.4% of their weight, versus 0.9% on placebo5.
The downsides showed up, too. Twice as many people stopped the drug because of side effects, 16.6% versus 8.2%4. The American College of Cardiology’s trial summary says those stops were mainly over stomach and bowel trouble, and it lists stomach and bowel complications in 10.0% versus 2.0%5.
Gallbladder problems ran 2.8% versus 2.3%6. Serious adverse events overall were a bit less common on the drug, 33.4% versus 36.4%5.
Novo Nordisk, which makes semaglutide, paid for the trial4. Its lead author, Michael Lincoff of Cleveland Clinic, has received consulting fees from the company6. Look AHEAD was sponsored by the National Institutes of Health3.
These trials were never a contest. Look AHEAD studied people with diabetes, and SELECT studied people with heart disease but no diabetes146. Neither tested lifestyle against the drug.
Benefits arrive on different clocks
Time to benefit is how long a treatment takes to start paying off.
Statins offer a well-studied example. In 2021, researchers led by Lindsey Yourman pooled 8 trials with 65,383 adults who mostly didn’t have heart disease7. They estimated how long it took for statins to prevent one major heart event:
- about 0.8 years for every 500 people treated;
- about 1.3 years for every 200 treated;
- about 2.5 years for every 100 treated7.
The authors concluded that statins may help adults 50 to 75 who expect to live at least 2.5 more years7.
Lifestyle runs on its own clock. In Look AHEAD, weight and insulin needs fell within a year, but no heart benefit appeared13. Diabetes prevention is a different story: our article on prevention versus treatment covers the trial where a lifestyle program beat a drug.
Then ask whether you can stop. You can quit a lifestyle program any day, but its gains can fade, as Look AHEAD’s weight loss did1. You can stop a medicine with your prescriber; in SELECT, about 1 in 6 stopped semaglutide over side effects4.
When lifestyle first makes sense, and when guidelines say don’t wait
Wanting to try food and movement before a prescription is a fair value, not stubbornness. It fits the natural-health instinct to do the least that works. Guidelines leave room for it in some situations and draw firm lines in others.
Cholesterol. The 2026 cholesterol guideline comes from the American College of Cardiology (ACC), the American Heart Association and nine other groups. According to a September 2026 feature in the ACC’s Cardiology magazine, it names groups who qualify for a statin without further risk calculation: people with diabetes or an LDL of 190 mg/dL or higher, now joined by people with chronic kidney disease or HIV9.
The same feature adds a nuance. People with room to improve their diet and an LDL only slightly above goal are likely to get there with food changes. People already eating well whose LDL is high, or far from goal, will need more than one medicine9. Our aspirin and statin article covers goals after a heart attack.
Diabetes. The American Diabetes Association’s Standards of Care 2026, published online in December 2025, say diabetes medicine should start at diagnosis, without delay, unless there’s a medical reason not to8. The same section says healthy habits, self-management education and life circumstances belong in the plan8.
The standards also draw firm lines8:
- With artery disease, such as a past heart attack or stroke, or a high risk of it, the plan should include a medicine proven to lower heart events, such as a GLP-1 drug or an SGLT2 inhibitor, whatever the A1C.
- With symptoms of high blood sugar, an A1C above 10% or a blood sugar of 300 mg/dL or higher, insulin should be considered.
The standards cut the other way, too. They say changes to treatment, including stepping it down, shouldn’t be delayed for people who aren’t meeting their goals. When weight loss or lifestyle changes lower a person’s needs, the text advises reducing first the drugs most likely to cause side effects, low blood sugar or burden without heart, kidney or metabolic benefit8. That’s a decision for you and your prescriber, not a reason to stop anything on your own.
Most people end up combining, and that’s often the point. In SELECT, the drug worked on top of standard heart care6. Our article on combining healthy habits with proven treatment shows how to keep both safe.
Decision aids put the numbers on the table
Decision aids are cards, booklets or websites that lay out options in plain numbers. A Mayo Clinic team tested one called Statin Choice in 98 people with type 2 diabetes seen by 21 specialists10. It showed each person’s 10-year heart risk, how much a statin would lower it and the drug’s downsides10.
Compared with people given the clinic’s usual cholesterol pamphlet, those using it scored 2.4 points higher on a 9-point knowledge test and 10.6 points lower on a 100-point scale of decisional conflict, the uncertainty people feel about a choice10. Three months later, 2 of 33 statin users in the decision-aid group had missed a dose in the past week, versus 6 of 29 in the pamphlet group10.
Not every trial looks that good. Mayo Clinic researchers also built Diabetes Medication Choice, seven cards comparing drugs on blood sugar, weight, side effects, cost and daily routine11. In a Greek trial of 204 adults, the cards made no measurable difference in comfort with the decision, knowledge, adherence or A1C11. The ADA notes that decision aids have improved A1C mainly when used face to face8.
Sponsors matter here, too. Two of the Greek trial’s authors came from the Mayo unit that built the cards. Its funding came through a European diabetes foundation program backed by drug makers AstraZeneca and BMS11.
In Canada, the Ottawa Hospital Research Institute’s decision aids group has studied these tools since 1995. Its framework, updated in 2020 from more than 250 studies with over 50,000 patients, found that decision support beat usual care12.
The group’s online A to Z Inventory lists existing aids. If none fits your choice, its Ottawa Personal Decision Guide is a blank worksheet13. Our article on patient autonomy covers the 2024 Cochrane review of these tools.
The Agency for Healthcare Research and Quality’s SHARE approach has five steps: seek your participation, help you explore and compare options, assess your values and preferences, reach a decision, and evaluate it later14. Medicare’s coverage rule for lung cancer screening, in the version in effect since February 2022, requires a shared decision-making visit, using at least one decision aid, before a first low-dose CT scan15.
The side-by-side comparison sheet
This is our own planning tool. Make one column per option, including “keep current care and recheck.” Then ask your clinician the question in each row.
| Row | Question to ask |
|---|---|
| Benefit | Out of 1,000 people like me, how many does this help, and with what? |
| Time to benefit | How long until it helps, and how long did the trial run? |
| Harms | Out of 1,000, how many get side effects, and how many quit because of them? |
| Cost and effort | What will it cost me in money, time and daily routine? |
| Can I stop it? | What happens to the benefit if I stop? |
| How we’ll know | Which number will we check, on what date, and what result would change the plan? |
A worked example. This person is hypothetical. Picture a 62-year-old man who had a heart attack, carries extra weight, takes a statin and doesn’t have diabetes. His cardiologist mentions semaglutide. He’d rather start with a structured lifestyle program. No trial has tested these options against each other. The lifestyle column comes from Look AHEAD, which enrolled people with diabetes, and the drug column from SELECT, so treat the numbers as rough guides, not a head-to-head score.
| Row | Intensive lifestyle program | Semaglutide, weekly | Current care and recheck |
|---|---|---|---|
| Benefit | No clear drop in heart events (in people with diabetes); better mobility, mood and sleep apnea12 | About 15 fewer heart events per 1,000 over about 3.3 years, by our math4 | In SELECT’s placebo group, 80 per 1,000 had an event in that time, by our math4 |
| Time to benefit | Weight loss within a year1 | Counted over about 3.3 years4 | Not applicable |
| Harms | Mostly effort; diet changes can mean medicines need adjusting8 | About 84 more per 1,000 quitting over side effects, mostly stomach and bowel trouble, by our math45 | His current risk continues |
| Cost and effort | Weekly sessions for six months, then fewer; at least 175 minutes of activity a week2 | Weekly shot; price and coverage vary by plan | Regular visits |
| Can I stop it? | Yes, but gains can fade1 | Yes, with his prescriber; about 1 in 6 stopped in SELECT4 | Not applicable |
Neither column wins every row. The drug has the clearer heart benefit for someone like him, with more side effects. The program brings benefits he’d likely feel sooner. Which matters more is his call, and the SHARE steps exist to bring that question into the room14.
A later article on keeping weight off covers drug costs and coverage. If dieting, tracking or weighing has been hard for you, the National Alliance for Eating Disorders runs a weekday helpline staffed by licensed therapists at 1-866-662-1235. If you’re in crisis, call or text 988.
For sudden symptoms such as chest pain, none of this applies. Our article on symptoms that mean call 911 lists them.
What the evidence says
| Claim | Evidence rating | What it rests on |
|---|---|---|
| An intensive lifestyle program improves weight, A1C and fitness in type 2 diabetes | Look AHEAD, 5,145 adults1 | |
| The same program prevents heart attacks and strokes in type 2 diabetes | (no benefit found) | No difference over a median 9.6 years1 |
| It eases sleep apnea, depression and incontinence and improves mobility | Early findings from one trial2 | |
| Semaglutide lowers heart events in people with heart disease and excess weight, without diabetes | One large company-funded trial46 | |
| Semaglutide prevents a first heart event in people without heart disease | SELECT didn’t enroll them6 | |
| Statins prevent first major heart events within a few years | 8 trials; about 2.5 years to spare 1 in 1007 | |
| Decision aids improve knowledge and lower decisional conflict | A randomized trial, plus the 250-plus studies behind the Ottawa framework1012 | |
| Decision aids improve adherence or A1C | Mixed trial results81011 |
Questions readers ask
Can I try lifestyle changes before starting a medicine?
For cholesterol, often, if your LDL is only slightly above goal and your diet has room to improve9. An LDL of 190 or higher is different9. For type 2 diabetes, the ADA’s 2026 standards call for medicine from diagnosis, alongside healthy habits8.
If I lose weight, can I come off my diabetes medicine?
Sometimes medicines can be reduced. When weight loss or lifestyle changes lower your needs, the ADA’s 2026 standards advise cutting back first on the drugs most likely to cause side effects or low blood sugar8. Make any change with your prescriber, never on your own.
Did Look AHEAD prove lifestyle doesn’t matter for the heart?
No. It showed that one intensive program didn’t lower heart events in people with type 2 diabetes over about a decade, while it improved many other things12. The investigators noted that more statin use in the comparison group may have narrowed the gap3.
Why trust a trial paid for by the drug’s maker?
Scrutiny is fair. SELECT was funded by Novo Nordisk, and its lead author has consulted for the company46. Weigh the absolute numbers and side effects together, and see our article on who paid for the study.
The bottom line
Lifestyle and medicine aren’t rivals, but their evidence keeps score in different ways. Put every option in the same terms: benefit per 1,000 people like you, time to benefit, harms, effort and whether you can stop.
Trying lifestyle first is a fair choice when your numbers allow it and you set a date to recheck. Guidelines say not to wait in some situations, and they also call for stepping medicines down when your needs fall. Bring the sheet, ask for the numbers, and decide together.
Related reading
About this article
This article was researched using original trial reports, American Diabetes Association and American College of Cardiology guidance, federal agency materials and peer-reviewed studies, all listed below. It was reviewed in October 2026. It is general education, not medical, legal or financial advice.
Sources
- Look AHEAD Research Group (Wing RR, Bolin P, Brancati FL, et al.). “Cardiovascular Effects of Intensive Lifestyle Intervention in Type 2 Diabetes.” New England Journal of Medicine, 2013;369(2):145-154 (UT Health San Antonio research record). scholars.uthscsa.edu
- American College of Cardiology. “Look AHEAD Trial Finds No Benefit in Intensive Lifestyle Intervention in Reducing CV Events.” ACC News, June 24, 2013. acc.org
- Husten L. “Large NIH Trial Finds No Cardiovascular Benefits for Weight Loss and Exercise in Type 2 Diabetics.” NEJM CardioExchange, June 24, 2013. blogs.nejm.org
- Lincoff AM, et al., for the SELECT Trial Investigators. “Semaglutide and Cardiovascular Outcomes in Obesity without Diabetes.” New England Journal of Medicine, 2023;389(24):2221-2232 (UCL Discovery record; funded by Novo Nordisk). discovery-pp.ucl.ac.uk
- American College of Cardiology. “Semaglutide Effects on Cardiovascular Outcomes in People With Overweight or Obesity – SELECT.” Clinical trial summary, presented November 11, 2023, updated September 12, 2024. acc.org
- Cleveland Clinic. “International Clinical Trial Finds That Semaglutide Reduced Cardiovascular Events by 20% in Adults With Overweight or Obesity Who Don’t Have Diabetes.” Cleveland Clinic Newsroom, November 11, 2023. newsroom.clevelandclinic.org
- Yourman LC, et al. “Evaluation of Time to Benefit of Statins for the Primary Prevention of Cardiovascular Events in Adults Aged 50 to 75 Years.” JAMA Internal Medicine, 2021;181(2):179-185 (Beth Israel Deaconess Medical Center research record). research.bidmc.org
- American Diabetes Association Professional Practice Committee. “9. Pharmacologic Approaches to Glycemic Treatment: Standards of Care in Diabetes—2026.” Diabetes Care, 49(Suppl 1):S183-S215, published online December 8, 2025. pmc.ncbi.nlm.nih.gov
- American College of Cardiology. “Feature | Getting to Goals: Applying the 2026 Dyslipidemia Guideline in High-Risk Patients.” Cardiology magazine, September 1, 2026. acc.org
- Weymiller AJ, Montori VM, Jones LA, et al. “Helping Patients With Type 2 Diabetes Mellitus Make Treatment Decisions: Statin Choice Randomized Trial.” Archives of Internal Medicine, 2007;167(10):1076-1082 (Mayo Clinic research record). mayoclinic.elsevierpure.com
- Karagiannis T, Liakos A, Branda ME, et al. “Use of the Diabetes Medication Choice Decision Aid in Patients With Type 2 Diabetes in Greece: A Cluster Randomised Trial.” BMJ Open, 2016;6:e012185. pubmed.ncbi.nlm.nih.gov
- Ottawa Hospital Research Institute, Patient Decision Aids Research Group. “History.” Accessed October 2026. decisionaid.ohri.ca
- Ottawa Hospital Research Institute, Patient Decision Aids Research Group. “Implementation Toolkit, Step 2: Find Patient Decision Aid(s).” Accessed October 2026. decisionaid.ohri.ca
- Agency for Healthcare Research and Quality. “The SHARE Approach.” Page created October 2024, last reviewed February 2026, ahrq.gov; and “The SHARE Approach” poster, AHRQ Pub. No. 14-0026-2-EF, April 2014 (copy hosted by Alliant Health), quality.allianthealth.org
- Centers for Medicare & Medicaid Services. “National Coverage Determination: Screening for Lung Cancer with Low Dose Computed Tomography (LDCT) (210.14).” Version effective February 10, 2022. cms.gov