Foods to Limit or Avoid

What you remove from your diet matters more than what you add. Ultra-processed food, sugary drinks, and processed meat carry the clearest harm signals in modern nutrition science — and they're the easiest changes to make.

The single biggest dietary lever in modern populations is often what you remove, not what you add. Ultra-processed food, sugar-sweetened beverages, processed meat, and excessive alcohol carry the clearest harm signals in modern nutrition science.

Ultra-processed food (UPF)

The strongest single dose-response signal in modern nutrition epidemiology — and the topic has its own dedicated article.

The headline numbers: each 10% rise in UPF as a share of energy intake associates with roughly 10% higher all-cause mortality, with adverse signals across cardiometabolic, cancer, and mental-health outcomes. The 2025 UCL randomised feeding trial in Nature Medicine showed the harm persists even when fat, protein, carbohydrate, salt, fibre, and produce intake are precisely matched — meaning processing itself, not just nutrient quality, is doing the damage.

→ Full article: Ultra-processed food — NOVA classification, mortality and disease evidence, the UCL trial, gut and inflammation mechanisms, biological aging effects, and practical strategies for spotting UPF on a label.

Quick rules of thumb:

Aim for <10–20% of calories from UPF (typical Western adult: 50–60%).
Cook from whole or minimally processed ingredients as the default.
Run the "kitchen test" on ingredient lists: if a substance can't be bought at a supermarket for home cooking, the product is ultra-processed.

Sugar-sweetened beverages

Strongest single-food-category mortality signal among beverages.

Each daily 12-oz serving of sugar-sweetened beverage associates with ~7% higher all-cause mortality and ~10% higher CVD mortality.^[1]
Driver of T2D, NAFLD, dental caries, weight gain.
Includes sodas, sweetened iced tea, energy drinks, sweetened coffee drinks, fruit "juices" with added sugar, sports drinks (in non-athletes).

Diet sodas and other non-sugar sweeteners. Less harmful than the full-sugar versions in the short term but not biologically inert, and the long-term cohort data on stroke, dementia, and cognitive decline is meaningful. Sugar alcohols (erythritol, xylitol) — common in keto and "sugar-free" products — carry their own cardiovascular signal. Plain water, unsweetened tea, sparkling water, and coffee are better defaults.

→ Full article: Sweeteners — biological-aging effects of added sugar, non-nutritive sweeteners, sugar alcohols, monk fruit and stevia, rare sugars (allulose, tagatose), and where honey fits.

Processed meat

IARC Group 1 carcinogen (same category as tobacco, asbestos — though absolute risk is much smaller).

Definition: meat preserved by smoking, curing, salting, or with added preservatives. Examples: bacon, sausages, hot dogs, ham, salami, deli meats.
Each 50 g/day serving associates with a ~16–18% increase in colorectal cancer risk.^[2]00444-1/fulltext)
Also linked to higher CVD mortality (multiple cohort studies).

Practical: treat processed meat as occasional indulgence, not regular diet item. There is no "safe" level supported by evidence; risk rises monotonically.

Red meat (unprocessed)

IARC Group 2A (probably carcinogenic; weaker evidence than processed meat).

Each 100 g/day of unprocessed red meat associates with ~17% increase in colorectal cancer risk.
The CVD signal is more debated and partially confounded by saturated fat content.
The NutriRECS controversy (2019 Annals of Internal Medicine) argued the evidence is weaker than dietary guidelines suggest. The quality of evidence is indeed observational; but the dose-response is consistent across many cohorts.

Practical: 1–3 servings/week is a reasonable upper bound; substitute with fish, poultry, legumes, eggs.

The cheese question (from the user's earlier query)

The MIND diet recommends <1 serving cheese/week. The basis:

Cheese is high in saturated fat, which raises LDL cholesterol — the classic mechanism.
However, the whole-food cheese matrix may attenuate this effect — multiple cohort studies show neutral or slightly favorable cardiovascular associations for cheese consumption (de Goede et al. 2015, Drouin-Chartier 2020). Possibly due to fermentation, calcium, vitamin K2, and the food matrix itself.
The MIND-specific concern is more about brain outcomes specifically — Morris et al.'s original MIND study found higher cheese intake correlated with worse cognitive trajectory, possibly via saturated fat and BBB inflammation.

Reasonable interpretation:

Modest cheese intake (1–3 servings/week) is compatible with Mediterranean and most longevity-oriented diets.
The MIND <1/week guidance is brain-health-specific and based on observational data; a hard rule against cheese is not strongly evidence-based.
Hard, fermented cheeses (parmesan, gouda, gruyère) > processed cheese products.
Cheese in volume daily, especially highly processed cheeses with added fats and salt, is genuinely worth limiting.

Industrial seed oils — the contested debate

The "seed oils are toxic" claim has become popular online. The actual evidence:

What's true:

Industrial deep-frying produces oxidized lipids that are pro-inflammatory.
Repeatedly heated oils are clearly harmful (commercial fryers running hours daily).
Refined oils are calorie-dense and easy to over-consume in processed foods.

What's overstated:

Linoleic acid (omega-6) at typical dietary levels is not "inflammatory" — RCTs and meta-analyses (Marklund 2019) show neutral or favorable cardiovascular associations.
Cohort data (Wang et al. 2016, Hooper et al. Cochrane) consistently show replacing saturated fat with polyunsaturated fat from vegetable oils reduces CVD mortality.

Practical synthesis:

Olive oil (especially EVOO) — best evidence; preferred default cooking fat.
Avocado oil, nut oils — fine; good for higher-heat cooking.
Canola, sunflower, safflower — neutral; not "toxic" in normal dietary use.
Avoid oils used for repeated commercial deep-frying (typical fast-food chains).

The headline isn't "avoid all seed oils." It's "avoid ultra-processed foods that contain seed oils alongside refined sugar, refined flour, and additives." The UPF context is what's harmful.

Trans fats

Worth its own line: industrial trans fats are unambiguously harmful.

Mostly eliminated from food supply in EU and US since 2018–2020 regulations.
Still present in some processed foods globally; check labels in non-regulated regions.
Naturally occurring trans fats in dairy/meat (vaccenic acid) are not the same and not harmful at normal intakes.

Excessive sodium

Population-level sodium excess (especially from processed food) drives hypertension.
DASH-Sodium trial: low-sodium DASH dropped systolic BP ~11 mmHg in hypertensives.
WHO: <2 g sodium/day (5 g salt). US average: ~3.4 g/day.
70%+ of dietary sodium comes from packaged/restaurant foods, not the salt shaker.

Alcohol

Covered in detail in Alcohol. The summary: minimum-risk dose is at or near zero, particularly for cancer.

What's not on this list

Eggs — earlier dietary cholesterol concerns largely resolved; whole eggs at moderate intake (5–7/week) are fine for most adults.
Saturated fat in moderation from whole foods (yogurt, cheese, eggs, occasional butter) — not the same as saturated fat from processed foods. The "saturated fat causes heart disease" framing has been moderated by 2010s+ evidence.
Coffee, tea — generally favorable.
Wine specifically — same as other alcohol; no special health benefit despite resveratrol marketing.
Gluten / dairy / legumes / nightshades — for non-allergic, non-celiac individuals, no evidence of harm.

The 80/20 framework

You don't need perfection. The goal is the pattern:

80%+ of intake from whole or minimally processed foods (vegetables, legumes, fish, nuts, fruit, whole grains, olive oil)
<20% from less-ideal categories (UPF, red meat, sweets, alcohol)

This captures most of the benefit available from dietary change. Going to 95/5 yields diminishing returns and often costs adherence.