Intermittent Fasting and the Heart: How to Read the Headline About a Higher Risk of Death
An eating window under eight hours, more heart deaths: this news went around the world in 2024. Here you'll read what the data can support, where they reach their limits and what studies with random allocation measure.
A headline can frighten you in three seconds. Reading a study properly takes longer. I'm taking that time with you here, so that in the end you make the decision, not the breaking news.
Imagine it's 11 in the morning. You haven't eaten since seven last night, and now your porridge is steaming in front of you. You've been eating like this for months and you feel good about it. Then your phone buzzes on the table.
Breaking news: Intermittent fasting linked to 91 percent higher risk of dying from heart disease.
Your spoon stops in midair. Have you been harming yourself with your routine without noticing? Should you go back to eating breakfast at seven starting tomorrow?
I know this pattern very well. A news item collides with a habit that has felt good so far. And suddenly everything is in question. Intermittent fasting and the heart: until a moment ago, the two seemed to fit together well for you.
Before you put the spoon down, let's take a look together. Not to play the study down. It deserves to be taken seriously. But taking it seriously also means reading carefully what it was able to measure and what it was not.
What the study measured and what it did not
The number 91 sounds like almost twice the danger. That hits hard if you eat this way yourself. So let's look at where it comes from.
The news came from the American Heart Association's EPI|Lifestyle conference, held in Chicago in March 2024. What was presented there was an abstract, a one-page summary. An abstract is like the trailer for a movie. It shows the most exciting scenes, but not the whole movie. At that point, independent experts had not yet reviewed the work.
A team led by Zhong at Shanghai Jiao Tong University School of Medicine used data from the large US health survey NHANES covering the years 2003 to 2018. How long each person ate during the day was calculated from two 24-hour dietary recalls. This information was then linked to the US national death registry. The American Heart Association's press release spoke of a 91 percent higher risk of cardiovascular death. The comparison was between people with an eating window under eight hours and people who spread their eating over 12 to 16 hours.
Even then, the research team stressed that an association does not mean the eating window caused the deaths. In 2025, the analysis was published after peer review, in a journal for diabetes and metabolism. One detail shifted along the way. The comparison group now covers 12 to 14 hours instead of 12 to 16. That is why the new figure cannot be compared one to one with the old one.
A team led by Zhong analyzed 19,831 adults from NHANES, with a median follow-up of 8.1 years. People who ate within a window of less than eight hours on average across the two recall days had a hazard ratio of 2.35 for cardiovascular death compared with 12 to 14 hours, while no association was found with all-cause or cancer mortality. What this means for you: there is a signal, but the authors themselves write that it remains open whether the short eating window or unmeasured confounders are behind it.
Chen M, Xu L, Van Horn L, et al. Diabetes Metab Syndr. 2025;19(7):103278. DOI: 10.1016/j.dsx.2025.103278
What does hazard ratio mean? Imagine two groups walking the same long path. The hazard ratio tells you how much more often an event occurs per stretch of the path in one group than in the other. So a value of 2.35 does not mean that 2.35 out of 100 people die. It means that the risk per unit of time in this group was, statistically, well over twice as high as in the comparison group.
Just as important is the range of uncertainty, the confidence interval. Here it runs from 1.39 to 3.98. That's like a weather forecast predicting 14 to 40 degrees Celsius for tomorrow. The direction is clear, but the exact size remains very uncertain. And a relative figure doesn't tell you how many people are affected in absolute numbers.
It gets interesting when a second team analyzes the same data source differently.
A team at the University of Pittsburgh analyzed NHANES data from 33,052 adults, also using two 24-hour recalls and with a median follow-up of eight years. The result was a U-shaped curve: the lowest risk was at about 11 to 12 hours, 8 hours or less went along with at least 30 percent higher all-cause mortality, especially in older adults, and 15 hours or more with 25 percent higher. What this means for you: in these data, both ends of the scale stand out, very short and very long eating windows.
Mao Z, Grant H, Kritchevsky SB, et al. Aging Cell. 2025;24(11):e70230. DOI: 10.1111/acel.70230
Two teams, the same database, partly different results. Zhong's team found no association with all-cause mortality, while the Pittsburgh team did. This shows how strongly findings like these can depend on choices made in the analysis.
These data did not measure intermittent fasting. They measured the time span in which people ate on two individual days. Whether someone was fasting on purpose, forgot a meal or felt too ill to eat is not recorded in these recalls.
And now you know why there is a big step between two recall days and the term intermittent fasting.
Two remembered days as a window into many years
Do you remember what you ate the day before yesterday? And at exactly what time you took your last bite?
That is exactly what people in NHANES were asked. A 24-hour recall is an interview in which you describe from memory what you ate the previous day. The eating window was estimated from two such days. That single value then stood for many years of follow-up.
It's a bit like judging an entire photo album from two vacation photos. Maybe they were two especially busy days. Maybe a sick day, a night shift or a trip. The pattern of those two days doesn't have to be the pattern of the next ten years.
In 2015, a team led by Subar addressed the criticism of self-reported dietary data in the Journal of Nutrition. The conclusion: self-reported energy intake is not suitable as a measure of actual energy intake, but self-reports remain valuable if their measurement errors are carefully taken into account. What this means for you: an eating window based on two remembered days is a rough estimate, not a precision instrument.
Subar AF, Freedman LS, Tooze JA, et al. J Nutr. 2015;145(12):2639-45. DOI: 10.3945/jn.115.219634
There is a second point. A short eating window on two days is not the same as planned fasting over months. Someone who fasts deliberately according to a plan may eat very differently from someone who skips meals because of time pressure or loss of appetite. In the recalls, both look the same.
Measurement errors don't make a study worthless. They make it blurry. A blurry photo can show that someone is in the picture. Who it is, you can't tell.
And now you know why two remembered days can be a clue, but not proof.
Confounders: people who eat within a short window often differ in other ways too
Imagine two people who both eat only between 12 noon and 6 pm. She planned it that way on purpose, sleeps well and goes for a walk in the evening. He works nights, smokes, barely has time to eat at work and is under a lot of pressure. In the statistics, both end up in the same group.
Researchers call this confounding, and the hidden factors behind it confounders. The classic illustration: in summer, ice cream sales and sunburns rise together. The ice cream doesn't cause the sunburn. The sun is behind both.
Good studies statistically adjust for known confounders, such as age, smoking or income. Zhong's team also tested its finding in 8 subgroups and 14 sensitivity analyses. The association with cardiovascular death held up. That suggests a robust observation. But you can only adjust for what was measured, and only as precisely as it was measured. That is why observational studies always leave some residual uncertainty.
A team led by Sun and Bao, mostly from the University of Iowa, analyzed NHANES data from 24,011 adults aged 40 and over from the years 1999 to 2014, also based on 24-hour recalls. People who ate only one meal a day had a hazard ratio of 1.83 for cardiovascular death compared with three meals, and those who skipped breakfast had one of 1.40. What this means for you: skipped meals also show up as risk markers in such data, and here too there is no way to tell whether the eating pattern itself or the life circumstances behind it make the difference.
Sun Y, Rong S, Liu B, et al. J Acad Nutr Diet. 2023;123(3):417-426.e3. DOI: 10.1016/j.jand.2022.08.119
In case you're thinking of your 11 am breakfast right now: yes, skipped breakfast shows up in these data too. But here as well, a risk marker is not yet a cause.
A risk marker is like a smoke detector. It goes off when something is going on. Whether there's a fire or the toast just got too dark, only a closer look will tell you.
And now you know why the same time on the plate can mean something entirely different for two people.
Reverse causality: when the illness comes first
Think of the last flu that really knocked you out. How much did you eat on those days? Probably not much, and hardly spread over twelve hours.
This is the third stumbling block. An illness in its early stages can dampen appetite, often long before it has a name. Heart failure, cancer, depression or certain medications can lead people to eat less and within shorter time spans. In that case, the short eating window would not be a cause of later deaths, but an early sign of illness.
The classic illustration: there are far more sick people in a hospital than in a gym. Yet the hospital doesn't make people sick. The illness came first, the hospital stay came after.
In a 2017 editorial in the journal Circulation, Sattar from Glasgow and Preiss from Oxford asked whether reverse causality in cardiovascular epidemiology is more common than imagined. This question belongs with every observational study on eating behavior. Sensitivity analyses can make the problem smaller, but they cannot rule it out entirely.
Sattar N, Preiss D. Circulation. 2017;135(24):2369-2372. DOI: 10.1161/CIRCULATIONAHA.117.028307
The question is not only whether a short eating window might make you ill. The question is also whether being ill shortens the eating window. In a table, both directions look the same.
And now you know why researchers looking at numbers like these always ask first: what came first?
What studies with random allocation show
Now let's turn the perspective around. What happens when your eating window is not chosen by you, but by drawing lots?
That is exactly what randomized controlled trials, or RCTs, do. Chance distributes smokers, night shift workers, stressed people and sick people evenly across both groups. As with a well-shuffled deck of cards, each group gets similar cards. Differences at the end can therefore much more readily be traced back to the eating window. The price: these studies run for weeks to months, not decades. They measure intermediate markers such as weight, blood pressure or blood sugar, not deaths.
Why would an eating window change anything at all? A review by de Cabo and Mattson in the New England Journal of Medicine describes the metabolic switch. After many hours without food, the body can switch from sugar to fatty acids and ketones as fuel. It's like a hybrid car switching from its gasoline engine to its electric motor. Mechanistically this is plausible, but it has not been shown for hard endpoints in humans.
de Cabo R, Mattson MP. N Engl J Med. 2019;381(26):2541-2551. DOI: 10.1056/NEJMra1905136
When the window falls early in the day
A team led by Sutton at the Pennington Biomedical Research Center had men with prediabetes eat within six hours for five weeks, with dinner before 3 pm, and compared this in alternation with a 12-hour window while their weight was kept constant. Even though their weight stayed stable, insulin sensitivity, the response of the insulin-producing beta cells, blood pressure and oxidative stress improved. What this means for you: the timing of your meals might play a role even without weight loss, although the study was small and short.
Sutton EF, Beyl R, Early KS, et al. Cell Metab. 2018;27(6):1212-1221.e3. DOI: 10.1016/j.cmet.2018.04.010
A second study combined the early window with weight loss.
At the University of Alabama at Birmingham, 90 adults with obesity ate for 14 weeks either from 7 am to 3 pm or over at least 12 hours, both groups with calorie reduction. The group with the early window lost 2.3 kilograms more, their diastolic blood pressure was 4 mmHg lower, although the range of uncertainty extended to zero, and their mood improved. What this means for you: an early window can give weight loss a little support, and according to the authors the effect corresponded to about 214 fewer kilocalories per day.
Jamshed H, Steger FL, Bryan DR, et al. JAMA Intern Med. 2022;182(9):953-962. DOI: 10.1001/jamainternmed.2022.3050
When the risk is already elevated
Especially interesting for your heart are people who already have several risk factors.
At the Salk Institute and the University of California San Diego, 108 adults with metabolic syndrome finished a study, and at the start their eating window averaged 14.19 hours. Three months with a personalized window of 8 to 10 hours lowered the long-term blood sugar marker HbA1c by 0.10 percentage points more than usual care alone, and no major side effects occurred. What this means for you: a moderate window was quite manageable even with elevated risk, and the authors themselves describe the benefit as modest.
Manoogian ENC, Wilkinson MJ, O'Neal M, et al. Ann Intern Med. 2024;177(11):1462-1470. DOI: 10.7326/M24-0859
Honest null results
A fair assessment also includes studies in which the eating window brought no additional benefit.
At Nanfang Hospital in Guangzhou, 139 people with obesity reduced their calories for twelve months, and one group additionally ate only between 8 am and 4 pm. The eating window group lost 8.0 kilograms, the comparison group 6.3 kilograms, and the difference was not statistically significant. What this means for you: in this study, the time window brought no measurable added benefit over simple calorie reduction.
Liu D, Huang Y, Huang C, et al. N Engl J Med. 2022;386(16):1495-1504. DOI: 10.1056/NEJMoa2114833
In the TREAT trial at the University of California San Francisco, 116 adults with overweight or obesity ate for twelve weeks either on a 16:8 schedule between 12 noon and 8 pm or with three set meals spread over the day. The difference in weight was only 0.26 kilograms and not significant, but in a subgroup of 50 people examined in person, a measure of muscle mass in the arms and legs fell more in the fasting group. What this means for you: a narrow window on its own doesn't automatically bring more, and your muscles deserve attention along the way.
Lowe DA, Wu N, Rohdin-Bibby L, et al. JAMA Intern Med. 2020;180(11):1491-1499. DOI: 10.1001/jamainternmed.2020.4153
How you can protect your muscles while losing weight is covered in the article on preserving muscle during weight loss.
What the big summaries say
Individual studies vary. Meta-analyses pool many studies, much as a choir carries many individual voices.
A team from Beijing Hospital pooled ten randomized trials with 694 participants. With an eating window, systolic blood pressure was on average 4.15 mmHg lower, diastolic pressure did not change significantly, and in the analysis the reduction was linked to weight loss. What this means for you: part of the blood pressure effect might simply work through weight loss.
Wang W, Wei R, Pan Q, Guo L. Nutr Metab (Lond). 2022;19(1):77. DOI: 10.1186/s12986-022-00711-2
An international team led by Semnani-Azad, with the participation of Charité Berlin, combined 99 randomized trials with 6,582 adults in the BMJ. All forms of intermittent fasting and of classic calorie reduction lowered weight compared with unrestricted eating, the differences between them were mostly small, and compared with whole fasting days, total cholesterol, LDL and non-HDL cholesterol were slightly higher with an eating window. What this means for you: intermittent fasting is one of several options with similar benefit, and the authors consider longer studies necessary.
Semnani-Azad Z, Khan TA, Chiavaroli L, et al. BMJ. 2025;389:e082007. DOI: 10.1136/bmj-2024-082007
The observational data ask who dies over the years. The RCTs ask what changes in the body over weeks to months. Both questions are legitimate. The second has been studied with random allocation, the first so far only through observation.
And now you know why reassuring RCTs and a worrying cohort do not disprove each other.
Intermittent fasting and the heart: how both pictures fit together
Maybe you feel torn right now. One analysis warns, other studies sound reassuring. Who should you believe?
You don't have to pick a side. Both pictures can be true at the same time, because they answer different questions. Here is how I see it, sorted by strength of evidence.
- Supported by RCTs and meta-analysesOver weeks to months, a limited eating window can slightly improve weight, systolic blood pressure and blood sugar. The advantage over simple calorie reduction is usually small or absent.
- Observed, but not shown to be causalIn NHANES data, very short eating windows, and in one analysis very long ones too, go along with higher mortality. Whether this is due to the eating window, to life circumstances or to illnesses that were already present is an open question. Correlation is the right word here, not causation.
- Mechanistically plausible, evidence in humans still thinAn early window might fit better with your internal clock. Very narrow windows might mean too little protein and energy for some people and so cost muscle mass.
- What I observe clinicallyPeople who stick rigidly to very narrow windows often don't get enough protein. That is an observation from clinical practice, not a study.
From the perspective of psychoneuroimmunology, it's worth looking at all four systems. In the metabolism, a longer break from eating can improve insulin sensitivity, as the Pennington Center study suggests. In the hormonal system, an internal clock is ticking: in the morning, the body often handles sugar better than late in the evening, which is mechanistically plausible. In the nervous system, long breaks from eating combined with lack of sleep might register as an additional stressor, which so far is more of a hypothesis. For the immune system, oxidative stress is a possible driver of silent inflammation, and that is exactly the value that fell in the same study. Which of these lenses weighs most for you is individual.
The study doesn't say: stop fasting. It says: take a closer look at who eats within a short window, for how long and why. You can accept that invitation for yourself without falling into panic.
This is about more than a time on the clock. It's about being able to read headlines without letting them take away your joy in eating. Knowledge is not a burden here. Knowledge is peace of mind.
And now you know why a warning and an all-clear can sometimes both contain something true.
Three levers for your everyday life
Back to your breakfast table at 11 am. What do you actually do now?
- Choose the middle rather than the extreme. An eating window of about 10 to 12 hours can be a calm starting point. In the Pittsburgh data, the lowest risk was at about 11 to 12 hours, and in RCTs, 8 to 10 hours also showed favorable intermediate markers. A narrower window is an option, not a must. What a gentle start can look like is described in the beginner's guide.
- Set your window on the early side and fill it wisely. Early windows did well in initial studies, more on this in the article about the early eating window. Within your window, make sure you get enough protein, vegetables and movement, so your muscles don't end up fasting too. How much protein may make sense is covered in the article on protein needs.
- Measure instead of guessing, and talk to your doctor. Blood pressure, waist circumference, HbA1c and blood lipids before you start and after eight to twelve weeks show you whether your pattern suits your body. If you have heart disease, take glucose-lowering medication or have a history of an eating disorder, your plan should be agreed with a doctor beforehand.
So maybe today you simply keep eating your porridge. With one difference: you now know what's behind the breaking news, and you can decide for yourself. If you'd like to do more than read and get started right away: below this article you'll find the option to book an appointment.
And now you know why you don't need breaking news to take good care of your heart.
This article is for information only and does not replace personal medical advice.
Related reading from other areas
Frequently asked questions about intermittent fasting and the heart
Is intermittent fasting bad for your heart?
That cannot be concluded from the data so far. The widely cited NHANES analysis found a statistical association between an eating window under eight hours and more cardiovascular deaths. But an association is not yet a cause. Over weeks to months, randomized trials tend to show favorable or neutral effects on weight, blood pressure and blood sugar. Long-term studies with hard endpoints such as heart attack or death are still lacking.
What exactly did the 2024 study examine?
It linked eating habits from a large US survey with later deaths. A team led by Zhong at Shanghai Jiao Tong University School of Medicine used NHANES data from the years 2003 to 2018 for this. The eating window was estimated from two 24-hour dietary recalls. The work was presented in March 2024 as a conference abstract at the American Heart Association, at that time without peer review.
Has the study since been published after peer review?
Yes, the peer-reviewed version appeared in 2025. It analyzed 19,831 adults, with an eating window of 12 to 14 hours as the comparison group instead of the earlier 12 to 16 hours. For an eating window under eight hours, the hazard ratio for cardiovascular death was 2.35. No association was found with all-cause or cancer mortality. The authors leave open whether the short eating window itself or unmeasured confounders are behind it.
What does a 91 percent higher risk mean?
This is a relative figure from the press release on the conference abstract. It compares two groups with each other and says nothing about how many people are affected in absolute numbers. A relative risk can sound large even when the baseline risk is small. In addition, the figure describes a statistical association, not proof that the eating window caused the deaths.
Why can't an observational study prove cause and effect?
Because people chose their eating window themselves. People who eat within a short window may also differ in sleep, work, smoking, stress or health. Such confounders can only be adjusted for statistically as well as they were measured. Only a study with random allocation distributes these differences evenly across both groups.
What is reverse causality?
Reverse causality means that it is not the behavior that makes you ill, but an emerging illness that changes the behavior. Someone who has heart disease or cancer, or who feels unwell, may have less appetite and eat within shorter time spans. In that case, the short eating window would be a consequence and not a cause. A 2017 editorial in Circulation asked whether this pattern is more common in cardiovascular research than imagined.
What do randomized trials show about the eating window?
Over weeks to months, they tend to show favorable or neutral effects. A meta-analysis of ten trials found systolic blood pressure to be 4.15 mmHg lower on average, which was linked to weight loss. A network meta-analysis in the BMJ covering 99 trials found similar benefits for intermittent fasting as for classic calorie reduction. Heart attacks or deaths were not measured as endpoints in these trials.
How long should my eating window be?
There is no single number that fits everyone. In an analysis from the University of Pittsburgh, the lowest risk was at an eating window of about 11 to 12 hours. Randomized trials mostly tested 6 to 10 hours, often with favorable intermediate markers. A window of about 10 to 12 hours can be a calm middle path that you tailor to yourself together with your doctor.
Can I do intermittent fasting if I have heart disease?
You should discuss this with a doctor beforehand. With heart disease, blood pressure medication or glucose-lowering drugs, a change in your eating pattern can affect the right dosage. Low blood sugar and circulation problems should be avoided. A slow start with measurements before and after can give you more confidence.
Is an early eating window better than a late one?
Initial studies suggest that an early window might have advantages. At the Pennington Biomedical Research Center, insulin sensitivity and blood pressure improved in men with prediabetes even though their weight remained stable. At the University of Alabama at Birmingham, people with a window from 7 am to 3 pm lost more weight than those eating over 12 hours or more. The two studies ran for only five and 14 weeks respectively, and long-term data are still lacking.
Sources
All journal articles were cross-checked via PubMed, the conference abstract via Crossref.
- Chen M, Zhong VW. Abstract P192: Association Between Time-Restricted Eating and All-Cause and Cause-Specific Mortality. Circulation. 2024;149(Suppl_1). Conference abstract, AHA EPI|Lifestyle Scientific Sessions 2024, not peer reviewed at the time of the news report. DOI: 10.1161/circ.149.suppl_1.p192 [Cohort, conference abstract]
- Chen M, Xu L, Van Horn L, Manson JE, Tucker KL, Du X, et al. Association of eating duration less than 8 h with all-cause, cardiovascular, and cancer mortality. Diabetes Metab Syndr. 2025;19(7):103278. DOI: 10.1016/j.dsx.2025.103278 [Cohort]
- Mao Z, Grant H, Kritchevsky SB, Newman AB, Farsijani S. Association of Eating Window With Mortality Among US Adults: Insights From a Nationally Representative Study. Aging Cell. 2025;24(11):e70230. DOI: 10.1111/acel.70230 [Cohort]
- Subar AF, Freedman LS, Tooze JA, Kirkpatrick SI, Boushey C, Neuhouser ML, et al. Addressing Current Criticism Regarding the Value of Self-Report Dietary Data. J Nutr. 2015;145(12):2639-45. DOI: 10.3945/jn.115.219634 [Review]
- Sun Y, Rong S, Liu B, Du Y, Wu Y, Chen L, et al. Meal Skipping and Shorter Meal Intervals Are Associated with Increased Risk of All-Cause and Cardiovascular Disease Mortality among US Adults. J Acad Nutr Diet. 2023;123(3):417-426.e3. DOI: 10.1016/j.jand.2022.08.119 [Cohort]
- Sattar N, Preiss D. Reverse Causality in Cardiovascular Epidemiological Research: More Common Than Imagined? Circulation. 2017;135(24):2369-2372. DOI: 10.1161/CIRCULATIONAHA.117.028307 [Editorial]
- de Cabo R, Mattson MP. Effects of Intermittent Fasting on Health, Aging, and Disease. N Engl J Med. 2019;381(26):2541-2551. DOI: 10.1056/NEJMra1905136 [Review]
- Sutton EF, Beyl R, Early KS, Cefalu WT, Ravussin E, Peterson CM. Early Time-Restricted Feeding Improves Insulin Sensitivity, Blood Pressure, and Oxidative Stress Even without Weight Loss in Men with Prediabetes. Cell Metab. 2018;27(6):1212-1221.e3. DOI: 10.1016/j.cmet.2018.04.010 [RCT, crossover]
- Jamshed H, Steger FL, Bryan DR, Richman JS, Warriner AH, Hanick CJ, et al. Effectiveness of Early Time-Restricted Eating for Weight Loss, Fat Loss, and Cardiometabolic Health in Adults With Obesity: A Randomized Clinical Trial. JAMA Intern Med. 2022;182(9):953-962. DOI: 10.1001/jamainternmed.2022.3050 [RCT]
- Manoogian ENC, Wilkinson MJ, O'Neal M, Laing K, Nguyen J, Van D, et al. Time-Restricted Eating in Adults With Metabolic Syndrome: A Randomized Controlled Trial. Ann Intern Med. 2024;177(11):1462-1470. DOI: 10.7326/M24-0859 [RCT]
- Liu D, Huang Y, Huang C, Yang S, Wei X, Zhang P, et al. Calorie Restriction with or without Time-Restricted Eating in Weight Loss. N Engl J Med. 2022;386(16):1495-1504. DOI: 10.1056/NEJMoa2114833 [RCT]
- Lowe DA, Wu N, Rohdin-Bibby L, Moore AH, Kelly N, Liu YE, et al. Effects of Time-Restricted Eating on Weight Loss and Other Metabolic Parameters in Women and Men With Overweight and Obesity: The TREAT Randomized Clinical Trial. JAMA Intern Med. 2020;180(11):1491-1499. DOI: 10.1001/jamainternmed.2020.4153 [RCT]
- Wang W, Wei R, Pan Q, Guo L. Beneficial effect of time-restricted eating on blood pressure: a systematic meta-analysis and meta-regression analysis. Nutr Metab (Lond). 2022;19(1):77. DOI: 10.1186/s12986-022-00711-2 [Meta-analysis]
- Semnani-Azad Z, Khan TA, Chiavaroli L, Chen V, Bhatt HA, Chen A, et al. Intermittent fasting strategies and their effects on body weight and other cardiometabolic risk factors: systematic review and network meta-analysis of randomised clinical trials. BMJ. 2025;389:e082007. DOI: 10.1136/bmj-2024-082007 [Meta-analysis]