What Foods Should I Avoid for Weight Loss? Honest Evidence Review

The honest answer: seven food categories drive most of the weight-gain signal in long-term cohort studies. Ultra-processed snacks, sugar-sweetened beverages, processed meats, fried fast food, refined-grain refined-carb foods, hidden-sugar condiments and sauces, and alcohol-dense drinking patterns. The strongest evidence comes from two sources: the Mozaffarian 2011 NEJM food-by-food weight-change table^[1] tracking 120,877 US adults across three Harvard cohorts (Nurses' Health Study, NHS-II, Health Professionals Follow-up Study) over 20 years, and the Hall 2019 Cell Metabolism inpatient RCT^[2] — the cleanest causal test ever run, where 20 weight-stable adults crossed over between matched-macro ultra-processed vs unprocessed diets and ate +508 kcal/day on the UPF arm, gaining +0.9 kg in 14 days while losing -0.9 kg on the unprocessed arm. The reverse list — foods that consistently protect against weight gain in the same cohorts — is short: yogurt, nuts (in portion), whole fruit (not juice), vegetables, whole grains, fish. Magnitude check: even completely eliminating the worst single category produces a fraction of the magnitude of GLP-1 pharmacotherapy — Wegovy (STEP-1)^[8] produced −14.9% body weight at 68 weeks; Zepbound (SURMOUNT-1)^[9] produced −20.9% at 72 weeks. But food quality still matters enormously on a GLP-1 because the residual calories drive whether weight loss is fat vs lean mass.

TL;DR — the seven categories to avoid

• (1) Ultra-processed snacks — chips, cookies, crackers, snack cakes, breakfast cereals high in added sugar. The Hall 2019 RCT^[2] is the only inpatient causal trial; UPF arm ate +508 kcal/day spontaneously, gained +0.9 kg in 14 days.
• (2) Sugar-sweetened beverages (SSBs) — regular soda, sweetened iced tea, energy drinks, fruit-flavored drinks, sweetened coffee drinks. Malik 2013 AJCN meta-analysis^[3]: ~0.22 kg additional weight gain per daily 12-oz serving over time in prospective cohorts; cluster-RCTs in children confirm causal direction. Mozaffarian 2011^[1]: +1.00 lb per 4-yr period per daily serving.
• (3) Processed meats — bacon, sausage, hot dogs, deli meats, cured meats. Mozaffarian 2011^[1]: +0.93 lb per 4-yr period per daily serving. Pan 2011 AJCN^[5]: HR 1.51 for type 2 diabetes per daily 50-g serving of processed red meat across 442,101 person-years.
• (4) Fried fast food — french fries, fried chicken sandwiches, fried fish sandwiches, burgers with fried sides. Pereira 2005 CARDIA^[4]: >2 fast-food meals/wk vs <1/wk in 3,031 young adults tracked 15 years: +4.5 kg additional weight gain and a 104% increase in insulin-resistance. Mozaffarian 2011^[1]: french fries +3.35 lb per 4-yr period per daily serving — the single most weight-gaining food in the entire 20-year analysis.
• (5) Refined-grain refined-carb foods — white bread, pastries, croissants, sugary breakfast cereals, white-flour pasta, white rice in large portions. Mozaffarian 2011^[1]: refined grains +0.39 lb per 4-yr period per daily serving. Holt 1995 satiety index^[7]: croissants score 47, doughnuts 68, white bread 100 (reference) — among the least satiating foods measured.
• (6) Hidden-sugar condiments and sauces — barbecue sauce, ketchup, sweet chili sauce, teriyaki, honey-mustard, salad dressings, flavored yogurts, pasta sauces. These aren't in the Mozaffarian table by name, but they're a meaningful fraction of the “added sugar” channel in NHANES data. AHA-recommended cap: <25 g added sugar/day for women, <36 g/day for men.
• (7) Alcohol — especially beer, mixed drinks with sweet mixers, and heavy or binge drinking patterns. Traversy 2015 review^[6]: heavy + binge drinking consistently associated with weight gain; light-to-moderate drinking not consistently associated, but alcohol adds 7 kcal/g (between carb and fat) with no satiety compensation.

The evidence map — what we're drawing from

Most popular “foods to avoid for weight loss” lists are vibes — assembled from anecdote, single small-trial outliers, or wellness-industry framing. We anchor this list on two pieces of evidence that the nutrition field treats as load-bearing:

(1) The Mozaffarian 2011 NEJM food-by-food analysis^[1]. The largest and longest prospective cohort study ever assembled on diet-and-weight, pooling 120,877 US adults across the Nurses' Health Study, NHS-II, and Health Professionals Follow-up Study with up to 20 years of follow-up. Subjects' food intake was re-measured at 4-year intervals; the analysis regressed weight change against each food's per-serving-per-day increase. The output is the canonical per-food weight-change table. Net weight-gaining foods (per daily serving, per 4-year period): potato chips +1.69 lb, french fries +3.35 lb, sugar-sweetened beverages +1.00 lb, unprocessed red meat +0.95 lb, processed meats +0.93 lb, refined grains +0.39 lb, sweets and desserts +0.41 lb. Net weight-protective foods: yogurt −0.82 lb, nuts −0.57 lb, fruit −0.49 lb, whole grains −0.37 lb, vegetables −0.22 lb. This is observational data — it cannot prove causation by itself — but the consistency across three independent cohorts and the dose-response signal is what gives it weight.

(2) The Hall 2019 Cell Metabolism inpatient RCT^[2]. The only randomized controlled trial that has directly tested ultra-processed vs unprocessed diets at matched macronutrients. Twenty weight-stable adults were admitted to the NIH Clinical Center for 4 weeks; each subject spent 2 weeks on an ultra-processed diet and 2 weeks on an unprocessed diet (crossover design). Both arms were matched for calories presented, macronutrient ratio, sugar, sodium, and fiber. Subjects ate ad libitum — no calorie targets, just “eat until satisfied.” The UPF arm spontaneously ate +508 kcal/day more than the unprocessed arm and gained +0.9 kg in 14 days; the unprocessed arm lost −0.9 kg. This is the cleanest causal evidence available that ultra-processed foods drive over-eating independently of macros. The mechanism is partly faster eating rate, partly lower satiety per calorie, partly easier-to-swallow textures.

Category 1 — Ultra-processed snacks

The Hall 2019 inpatient RCT^[2] is the cleanest single piece of evidence in the entire weight-and-food literature. Ultra-processed foods (using the NOVA-4 classification: industrial formulations containing ingredients you wouldn't find in a home kitchen, like hydrogenated oils, emulsifiers, flavor compounds, and cosmetic additives) drove +508 kcal/day of additional spontaneous intake compared to a macronutrient-matched unprocessed diet. Subjects didn't know they were over-eating; they just ate faster, ate more bites, and reached fullness later.

The category extends well beyond “junk food.” It includes most packaged breakfast cereals, most commercially baked breads, most snack crackers, most frozen meals, most flavored yogurts, most protein bars, most plant-based meat alternatives, most fast-food entrees. The shared properties: industrial reformulation that makes the food easier and faster to consume than its whole-food counterpart, often with displaced fiber and increased palatability.

Practical rule: the more an ingredient list reads like a chemistry lab, the more cautious to be with portion. A bag of potato chips has 3-4 ingredients but the processing (slicing, frying, salting) makes it radically easier to over-consume than a baked potato. The same calories of baked potato + olive oil would fill most adults to a stop signal; the same calories of chips usually don't.

Category 2 — Sugar-sweetened beverages

Liquid sugar calories are the most well-documented weight-gaining channel in the entire nutrition literature. The Malik 2013 AJCN meta-analysis^[3] pooled 32 prospective studies and cluster-randomized trials: each daily 12-oz SSB serving was associated with ~0.22 kg of additional weight gain over follow-up in adult cohorts; cluster-RCTs in children directly confirmed the causal direction (substituting SSB for water reduced weight gain). Mozaffarian 2011^[1] independently found +1.00 lb per 4-yr period per daily serving in the Harvard cohorts.

The mechanism is twofold: (a) liquid calories don't trigger the satiety response that solid food does, so intake doesn't reduce at the next meal to compensate; (b) the speed of glucose absorption from SSBs produces a large insulin response and a delayed return to hunger sooner than a meal-equivalent calorie load would.

Practical substitutions that the trial literature supports: water (zero-calorie, neutral), sparkling water (zero-calorie, similar oral experience), unsweetened tea or coffee (also neutral; appetite-suppressing effects are small and short-lived), and low-fat milk or unsweetened plant milk in moderation. Diet sodas / artificially sweetened beverages are weight-neutral to slightly favorable in substitution trials (e.g. Peters 2014 NIH-funded RCT) though long-term observational data is mixed.

Category 3 — Processed meats

The Mozaffarian 2011^[1] 20-year table places processed meats at +0.93 lb per 4-yr period per daily serving — comparable to SSBs in per-serving weight impact. Pan 2011 AJCN^[5] extends the story to type 2 diabetes risk: across 442,101 person-years in the same three Harvard cohorts, each daily 50-g serving of processed red meat was associated with HR 1.51 for incident T2D (95% CI 1.25-1.83).

The category includes bacon, sausage, hot dogs, salami, bologna, pepperoni, deli ham, prosciutto, jerky, and cured meats. The defining characteristic in the IARC classification is preservation by salting, curing, fermentation, smoking, or addition of nitrites — which is a separate signal from total fat or total calories.

Note that unprocessed red meat is also in the weight-gaining column in Mozaffarian 2011 (+0.95 lb per 4-yr period per daily serving), although the T2D and CV signal is substantially weaker than for processed meats. For weight-loss purposes the distinction matters less than the portion-size and frequency distinction: 4-oz portions a few times a week, prioritizing lean cuts, is a different pattern from 8-oz portions daily.

Category 4 — Fried fast food

Pereira 2005 CARDIA^[4] is the canonical prospective analysis: 3,031 young US adults (Black and white, 18-30 at baseline) tracked 15 years. Subjects who reported eating fast food >2 times/wk at baseline AND at follow-up gained an additional ~4.5 kg of weight and had a 104% larger increase in insulin resistance compared to subjects who ate fast food <1 time/wk at both timepoints. The dose-response held across race and baseline BMI.

Mozaffarian 2011^[1] places french fries specifically at +3.35 lb per 4-yr period per daily serving — the single most weight-gaining food in the entire 20-year three-cohort analysis. The mechanism is well-described: the frying medium adds 100-300 kcal per portion to the underlying potato; the salting and crisping accelerate eating rate; the convenience format encourages frequency.

The category extends to fried chicken sandwiches, fried fish sandwiches, fried side items, and combo meals with SSBs. The substitution that the trial evidence supports: grilled or roasted protein + non-fried sides + water or unsweetened drink. A grilled chicken sandwich + side salad + water is often 400-500 kcal total — a fried-combo equivalent is routinely 1,200-1,600 kcal.

Category 5 — Refined-grain refined-carb foods

Mozaffarian 2011^[1] places refined grains at +0.39 lb per 4-yr period per daily serving — smaller per-serving than the categories above, but the per-day consumption pattern is much higher (3-6 servings/day of bread, pasta, crackers, cereal for many US adults).

The Holt 1995 satiety index^[7] provides the mechanistic explanation: refined-carb foods score among the lowest on per-calorie satiety. White bread = 100 (reference baseline); croissants 47, cake 65, doughnuts 68, Mars bar 70. By contrast, oranges 202, apples 197, baked potato 323, fish 225. Per equivalent calorie, refined carbs leave you the least full of any food category.

The category includes white bread, croissants, pastries, most commercial bakery items, sugary breakfast cereals, white-flour pasta in large portions, refined crackers, and white rice in large portions. The substitution that the cohort data supports: whole grains (weight-protective in Mozaffarian 2011 at −0.37 lb / 4 yr / daily serving). Whole-wheat bread, sourdough (lower glycemic index), brown rice, oats, quinoa, intact whole grains.

For the canonical sourdough and bread evidence walk-through see our sourdough bread and weight loss evidence review; for the rice question specifically see our is rice good for weight loss evidence article.

Category 6 — Hidden-sugar condiments and sauces

Added sugar in condiments and sauces is one of the most under-tracked channels of caloric intake. Barbecue sauce averages 14 g sugar per 2 tablespoons; ketchup ~8 g per 2 tbsp; teriyaki ~12 g; honey-mustard ~10 g; many commercial salad dressings 6-10 g; flavored yogurts 15-25 g per cup; pasta sauces 8-12 g per ½-cup. These amounts add up rapidly across a day — a single restaurant meal with a sweet glaze, ketchup, and a dressed salad can deliver 40-60 g of added sugar before dessert.

The AHA-recommended cap is <25 g added sugar/day for women and <36 g/day for men — corresponding roughly to the 6% / 9% of total calories targets in the 2020-2025 Dietary Guidelines for Americans. A single 12-oz can of regular soda is 39 g of added sugar (above both AHA caps in a single serving).

Practical substitutions: mustard (essentially zero added sugar) instead of ketchup or honey-mustard; olive oil + vinegar instead of commercial salad dressings; plain Greek yogurt + fresh fruit instead of flavored yogurt; tomato paste + herbs as a low-sugar pasta-sauce base.

Category 7 — Alcohol

Traversy 2015 Curr Obes Rep^[6] reviewed the cross-sectional and prospective evidence on alcohol and weight. The pattern: heavy drinking and binge drinking are consistently associated with weight gain across cohorts; light-to-moderate drinking is not consistently associated with weight in either direction in adult cohorts. Spirits appear more obesogenic than wine or beer in several analyses, though the data isn't fully consistent across populations.

The macronutrient math: alcohol contributes 7 kcal/g — between carbohydrate (4) and fat (9) — with essentially no satiety compensation. A 5-oz glass of wine is ~125 kcal; a 12-oz beer is ~150-200 kcal; a mixed drink with sweet mixer is often 250-400 kcal. Three drinks at dinner = 400-1,000 kcal of essentially non-satiating intake on top of the meal calories.

The second-order effect matters more than the calories for many drinkers: alcohol disinhibits eating decisions and commonly results in higher subsequent food intake. The post-drink late-night meal is a real signal in intake-diary data.

For GLP-1 patients specifically, the picture is different — covered in detail in our can you drink alcohol on a GLP-1 evidence review and the GLP-1 and alcohol use disorder deep-dive. Most GLP-1 patients report dramatically reduced alcohol tolerance on therapy.

The reverse list — foods to ADD instead

The Mozaffarian 2011 NEJM table^[1] identifies the food categories that were net weight-protective across 20 years and 120,877 adults. The reverse list is short:

• Yogurt: −0.82 lb per 4-yr period per daily serving. The single most weight-protective food in the table. Greek yogurt and plain yogurt deliver protein density (15-25 g per cup) and probiotic content in a low-calorie format. See our cottage cheese weight loss evidence review for the dairy-protein-density companion.
• Nuts: −0.57 lb per 4-yr period per daily serving — in portion. Despite being calorie-dense (~160-200 kcal per oz), nuts are weight-protective in prospective cohorts, likely because of high satiety, slow eating rate, and the fact that ~10-15% of nut fat is not absorbed.
• Whole fruit: −0.49 lb per 4-yr period per daily serving. Whole fruit — not juice — was weight-protective across all three cohorts. See our fruits for weight loss evidence hub for the per-fruit breakdown (apples, bananas, berries, grapes, pineapple, watermelon).
• Whole grains: −0.37 lb per 4-yr period per daily serving. Oats, brown rice, quinoa, intact whole-grain bread. The fiber and slower glycemic response distinguishes them from the refined-grain weight-gainers.
• Vegetables: −0.22 lb per 4-yr period per daily serving. The smallest per-serving signal, but a dose-response benefit. Volume-dense / calorie-thin — the canonical caloric-density-reducing food category.
• Fish: ranked highest on the Holt 1995 satiety index^[7] (white fish 225 vs white bread 100 reference). High-protein, low-volume, well-tolerated on GLP-1s. See our salmon for weight loss evidence review for the per-species nutrient and Mozaffarian 2006 cardiometabolic breakdown.
• Water — in place of any SSB. The single highest-leverage substitution in the entire intake pattern. Replacing one 12-oz daily SSB with water maps to ~0.22 kg less weight gain over the average follow-up window in Malik 2013^[3].

The shared property: volume + protein per calorie. The weight-protective foods deliver either volume (vegetables, fruit), protein density (yogurt, fish, nuts), or fiber and slow-glycemic carbohydrate (whole grains). They occupy the top half of the Holt 1995 satiety index^[7]; the weight-gaining foods occupy the bottom half.

Foods to avoid vs Wegovy / Zepbound — the honest magnitude framing

It's important to be clear about what diet quality can and cannot do at a population scale, especially in the GLP-1 era.

Diet-quality optimization, on its own, does not reach the magnitude of GLP-1 pharmacotherapy. The Wilding 2021 STEP-1 trial^[8] reported a 14.9% body-weight reduction at 68 weeks on semaglutide 2.4 mg weekly. The Jastreboff 2022 SURMOUNT-1 trial^[9] reported a 20.9% reduction at 72 weeks on tirzepatide 15 mg weekly. For a 220-lb starting weight, those are −33 lb and −46 lb respectively. Even completely eliminating the worst single food category from a typical American diet (e.g. daily fries, daily SSB, daily processed meat) maps to roughly −1 to −3.5 lb per 4-year period in the Mozaffarian 2011 regression coefficients^[1] — an order of magnitude smaller.

This is not an argument against avoiding the seven categories. It is an argument against the wellness-industry framing that food choices alone produce GLP-1-magnitude outcomes for the >42% of US adults with BMI ≥ 30. They don't. The clinical reality:

• For someone maintaining or slowly gaining weight, dietary-pattern improvement can be the primary lever. Eliminating SSBs, fast food, and processed-meat-heavy patterns can stop weight gain and drive modest loss over years.
• For someone with clinical obesity (BMI ≥ 30 or ≥ 27 with comorbidities), the trial data consistently shows that diet-and-exercise interventions produce 3-7% body weight loss at 1 year (Look AHEAD, DPP). That's real but typically not clinically sufficient for the cardiometabolic risk reduction the patient needs. GLP-1 pharmacotherapy or bariatric surgery deliver 15-30% at trial endpoint — a different magnitude entirely.
• For someone already on a GLP-1, diet quality matters in a different direction: it drives whether the lost weight is fat vs lean mass. The SURMOUNT-1 DXA substudy documented 25-39% of weight lost on tirzepatide is lean mass. Adequate protein and whole-food density during the deficit is what determines that ratio. See our semaglutide and muscle mass deep-dive for the lean-mass preservation protocol.

Common bad takes

The popular framing around “foods to avoid” contains several persistent misconceptions worth addressing:

“Just count calories — food quality doesn't matter.” Mathematically true in a metabolic ward; behaviorally false outside one. The Hall 2019 inpatient RCT^[2] tested exactly this and found that at matched macros, subjects spontaneously ate +508 kcal/day more on the ultra-processed arm. Food choice changes spontaneous intake even when calories aren't being deliberately tracked. The macros-only framing misses the satiety, eating-rate, and reward-system effects that drive compliance.

“All carbs are bad / cut carbs to lose weight.” Refined-grain refined-carb foods are weight-gaining in Mozaffarian 2011 (+0.39 lb / 4 yr / daily serving). Whole grains are weight-protective (−0.37 lb / 4 yr / daily serving) in the same dataset. The carbohydrate category is too broad to be useful as a single avoidance rule. The relevant distinction is processing level and fiber content.

“Natural sugar / honey / coconut sugar / agave is fine; only refined sugar is the problem.” The added-sugar evidence base (AHA guidance, Mozaffarian 2011 SSB data, Malik 2013 SSB meta) treats all caloric sweeteners equivalently. The body metabolizes sucrose, high-fructose corn syrup, honey, agave, and coconut sugar similarly — the modest differences in glycemic index don't produce meaningful weight or cardiometabolic differences at typical intake levels. Whole fruit is different (intact fiber + slower absorption); juice and honey are not.

“Fat makes you fat / avoid all fat.” The dietary fat ↔ body fat link the 1980s pushed has not held up. Nuts (−0.57 lb / 4 yr / daily serving) and fish (most satiating food on the Holt 1995 index) are weight-protective despite being relatively calorie-dense. Olive oil and avocados are similarly weight-neutral or favorable in trial data. The fat category to limit is the one bundled with refined carbs and ultra-processing (fried fast food, pastries, snacks) — not whole-food fat sources.

“Eating after 8 pm causes weight gain.” The timing-of-eating literature is mixed but the late- eating effect, when present, is small relative to total intake. A 600-kcal late dinner doesn't become 700 kcal because of the clock. The behavioral correlation (late eating tracks with snacking-heavy days and alcohol- accompanied meals) is what produces most of the observed signal.

Practical use on a GLP-1

For patients on semaglutide, tirzepatide, or another GLP-1 receptor agonist, the “foods to avoid” list takes on additional weight for two reasons:

(1) The same categories that drive weight gain also drive GI side effects. The Wharton 2022 clinical practice review for managing GLP-1 GI side effects documents the same offenders: high-fat fried meals, large portions, sugary drinks, alcohol, and very rich processed foods consistently produce the worst post-meal nausea on GLP-1 therapy. The slowed gastric emptying that defines GLP-1 pharmacology amplifies the GI response to ultra-processed and high-fat foods. See our full GLP-1 diet protein and tolerance guide for the meal-timing and tolerance pattern.

(2) The reduced total intake on a GLP-1 makes nutrient density of the remaining calories critical. When a patient drops from 2,400 kcal/day to 1,400 kcal/day on therapy, those 1,400 calories have to deliver enough protein (1.6-2.2 g/kg body weight), fiber (25-35 g), and micronutrients to prevent lean-mass loss, constipation, and deficiency. A 1,400-kcal day built around ultra-processed snacks and SSBs will deliver almost none of that. A 1,400-kcal day built around fish, eggs, yogurt, vegetables, whole grains, fruit, and nuts will hit all three targets comfortably. The diet quality decision essentially determines whether the GLP-1 produces healthy weight loss or accelerated lean-mass loss.

Practical GLP-1 substitution patterns from the seven categories above:

• SSBs → sparkling water or unsweetened tea. Highest-leverage single substitution.
• Chips/cookies → Greek yogurt + fruit, or nuts in portion. Both deliver protein density and satiety far above the snack-category baseline.
• Processed deli meats → rotisserie chicken, canned tuna, hard-boiled eggs, or canned salmon. Same convenience, much better protein-to-calorie ratio. See our tuna for weight loss evidence review and eggs for weight loss evidence review.
• Fried fast food → grilled chicken or fish + side salad + water. Same restaurant, very different macronutrient profile.
• White bread/refined pasta → whole-grain versions, in portion. The refined-vs-whole distinction is more important than total carbohydrate.
• Sweet sauces → mustard, vinegar, olive oil, herbs, spices. Zero-added-sugar flavor options that don't accumulate across meals.
• Alcohol → less of it, and avoid the sweet-mixer drinks specifically. Most GLP-1 patients report tolerance falls dramatically on therapy anyway.

Bottom line

• The strongest evidence on “foods to avoid” comes from two sources: Mozaffarian 2011 NEJM^[1] (120,877 US adults, 20-year food-by-food weight-change table) and Hall 2019 Cell Metab^[2] (NIH inpatient ultra-processed crossover RCT, +508 kcal/day spontaneously on UPF arm).
• Seven categories consistently rank as net weight-gainers: ultra-processed snacks, sugar-sweetened beverages (Malik 2013^[3]), processed meats (Pan 2011^[5]), fried fast food (Pereira 2005 CARDIA^[4]), refined-grain carbs, hidden-sugar condiments, and alcohol (Traversy 2015^[6]).
• The reverse list is short: yogurt, nuts, whole fruit, whole grains, vegetables, fish, water. These were the net weight-protective foods in the same Mozaffarian 2011 table and they top the Holt 1995 satiety index^[7].
• The shared property of the weight-protective list: volume + protein per calorie. The shared property of the weight-gaining list: low satiety per calorie, driven by processing, liquid format, or refined-carb density.
• Magnitude matters. Eliminating the worst single category maps to roughly −1 to −3.5 lb per 4-year period in Mozaffarian 2011^[1]coefficients. Wegovy (STEP-1^[8]) produces −14.9% body weight at 68 weeks; Zepbound (SURMOUNT-1^[9]) produces −20.9% at 72 weeks. Diet quality is not at GLP-1 magnitude on its own.
• On a GLP-1, diet quality changes role. It no longer drives the magnitude of weight loss (the drug does that); it drives whether the lost weight is fat vs lean mass, and it determines whether the residual calories deliver enough protein, fiber, and micronutrients to prevent side effects and lean-mass loss.

Related research and tools

• What to eat on a GLP-1: the protein, hydration, and fiber guide — the companion piece on the foods to ADD (protein-first eating, 1.6-2.2 g/kg targets, fiber 25-35 g, water 2-3 L)
• Semaglutide and muscle mass deep-dive — why diet quality matters specifically on a GLP-1 (lean-mass loss as the dominant adverse effect)
• Fruits for weight loss evidence hub — the per-fruit weight-change data and the whole-fruit-vs-juice distinction
• Salmon for weight loss evidence review — the top of the Holt 1995 satiety index and the canonical high-protein whole-food replacement for processed meats
• Sourdough bread and weight loss evidence review — whole-grain vs refined-grain bread choices and glycemic index by bread type
• Is rice good for weight loss — brown vs white rice + portion-size practicalities
• Eggs for weight loss evidence review — the canonical high-satiety breakfast replacement for refined-grain cereals and pastries
• Can you drink alcohol on a GLP-1 — the GLP-1-specific tolerance and intake guidance
• How to calculate macros for weight loss guide — turning the avoidance list into a daily macro target
• GLP-1 protein calculator — calculate your daily protein target
• Versión en español: ¿Qué alimentos debo evitar para bajar de peso? — Spanish translation of this evidence review (forthcoming)

Important disclaimer. This article is educational and does not constitute medical or nutrition advice. Individual response to specific foods varies, and patients with diabetes, kidney disease, gout, food allergies, or other clinical conditions should consult their care team before making major dietary changes. Pregnant women, women planning pregnancy, and young children should follow FDA guidance on fish consumption and the 2020-2025 Dietary Guidelines for Americans. Patients managing alcohol use disorder should follow clinical guidance and may need to abstain entirely. PMIDs were independently verified against the PubMed E-utilities API on 2026-05-17.

Last verified: 2026-05-17. Next review: every 12 months, or sooner if major new evidence on ultra-processed food, sugar-sweetened beverages, processed meat, or fast-food consumption is published.