Intervention Maps · Obesity · Eating patterns
Eating patterns
This page is about named eating patterns studied for weight. It is not a contest for the one best obesity diet. Much of the difference you see at six months fades by a year. Formula or very-low-calorie diets are a different protocol on this same page, not a diabetes-remission program, and not the same decision as joining a weight-management program.
See also: Behavioral weight-management programs · Exercise.
Bottom line
Eating patterns stay on the map as a comparison of weight and risk factors, not as the one best obesity diet. Low-carbohydrate and low-fat look similar at six months. Named-diet differences are small, and weight loss diminished at 12 months. Vegetarian programs are on the map. Unstudied vegan-only patterns should not inherit Mediterranean evidence. Formula diets are a different protocol here, with about 4 kg extra at one year versus a program alone, smaller later. They are not a diabetes-remission program.
What it is
A dietary pattern is a whole way of eating, not a single food slogan. Low-carbohydrate, low-fat, Atkins, DASH, Zone, Mediterranean, vegetarian, and related patterns appear in the same comparative reviews for weight and risk factors. Very-low-calorie or formula diets are a supervised protocol discussed here because people actually face that choice.
What taking part usually involves
Shopping, cooking, and eating across months, not a one-day cleanse. Some people work with counseling or structured menus. Pattern advice is not the same as joining a multicomponent program, even when both change what is on the plate. A formula-diet protocol is usually a medically supervised period of meal replacement, then food reintroduction, with support. This page does not photocopy a trial menu.
What it is commonly confused with
“Atkins won at six months, so it's the obesity diet.” “If differences shrink at a year, diets don't work.” “Mediterranean's remaining risk-factor signal means it prevents heart attacks.” “Plant-based means one vegan plate as strong as Mediterranean.” “A formula diet is the same as the diabetes-remission program on the other map.” “Joining a program is just picking a named diet.”
What the evidence shows
A network of 121 trials in 21,942 people compared named programs and macronutrient patterns. Versus usual diet at six months, low-carbohydrate and low-fat were similar (about 4.63 kg and 4.37 kg). Named-diet examples included Atkins at about 5.5 kg, Zone at 4.1 kg, and DASH at 3.6 kg. Gaps between diets were smaller than diet versus no diet. Weight loss diminished at 12 months. Risk-factor benefits largely disappeared except for Mediterranean eating.
Those remaining risk-factor findings are still risk-factor findings. They are not heart-attack, stroke, or death rankings.
Vegetarian and Ornish-class programs are in that comparison. That is honesty about what was studied. It is not a license to treat an unstudied vegan-only plate as proven because Mediterranean eating was studied.
Formula or very-low-calorie diets plus a behavioral program added about 3.9 kg extra loss at one year versus a behavioral program alone in a working review. The extra loss was smaller later, and the longest follow-up did not clearly show a lasting difference. Stopping early was similar in both groups. That describes a different protocol on this page. It is not a type 2 diabetes remission program.
Joining a structured weight-management program is a different decision. It lives on the programs page. If a program uses meal replacements, that is a program component. It does not replace this formula-diet discussion, and this discussion does not replace that page.
Who was studied, expectations, and limits
Population and context studied: adults with overweight or obesity in named-diet trials focused on weight and risk factors, plus a working comparison of very-low-calorie diets added to a behavioral program.
That is the research setting, not a personal meal prescription.
Realistic expectation: studied patterns can produce several kilograms at six months, with small differences between named diets, and less remaining difference at a year. A supervised formula-diet protocol can add some extra loss at one year. Limits: ranks are not event prevention; fade is part of the claim; vegan-only plates were not shown to inherit Mediterranean evidence; not a program page; not a diabetes-remission page.
Evidence consulted Open sources
Key sources used to prepare this answer. This is not a citation for every sentence.
- Ge et al. Named dietary programmes for weight and risk-factor reduction (BMJ, 2020) (C1)
- AACE. Algorithm for adults with obesity / adiposity-based chronic disease, 2025 (A1)
- Pedersen et al. Pharmacotherapy for obesity management in adults (CMAJ, 2025) (A2)
- NICE NG246. Overweight and obesity management (A3)
- Parretti et al. Efficacy of very-low-calorie diets (Obes Rev, 2016)
Parretti 2016 is labeled working evidence for the formula-diet section, not a library C source.
← Back to Obesity · find this again later from “What are you trying to figure out?”