Scientists are learning why ultra-processed foods are bad

Original Article ↗ Hacker News Discussion ↗

Definitions and terminology

  • Strong disagreement on what “ultra‑processed” actually means.
  • Some argue the term is vague, sensationalist, and conflates harmless processing (washing, milling, cooking) with industrial formulations.
  • Others defend a practical rule: foods that couldn’t realistically be made in a normal kitchen or rely on industrial ingredients/additives (HFCS, hydrogenated oils, gums, flavorings, etc.).
  • NOVA classification is often referenced but criticized as “woolly” and overbroad.

Evidence and studies

  • Highlighted RCT (NIH, ~20 people inpatient): ultra‑processed vs minimally processed diets, matched for macros and calorie density; people on UPF diets ate ~500 kcal more/day and gained weight; on minimally processed, they lost weight.
  • Ongoing similar NIH study with ~36 subjects mentioned; some scientists question whether such small samples can yield general conclusions.
  • Commenters stress that many nutrition studies are non‑replicable; estimates like “40–60% can’t be replicated” are cited.

Proposed mechanisms

  • Main robust finding: UPFs drive overeating via hyper‑palatability, variety, and higher “calories per bite”.
  • Other hypothesized factors:
    • Reduced fiber and disrupted food structure, altering digestion and satiety.
    • Easier, faster eating (less chewing).
    • Additives/emulsifiers, seed oils, contaminants, and packaging chemicals with possible chronic effects (unclear/contested).
    • Higher sugar and refined carbs leading to rapid glucose spikes.

Critiques of NOVA / category problems

  • NOVA lumps together very different items (e.g., protein powder and chips; sugary yogurt and plain yogurt).
  • Some studies show certain UPFs (sugary drinks, processed meats) correlate with worse health, while others (some breads, cereals, yogurts) correlate with better outcomes, which undermines blanket demonization.
  • Several argue the useful axes are calorie density, protein/fiber content, and “engineered hyper‑palatability,” not “processing” per se.

Sugar, carbs, and fiber

  • Big debate on whether sugar is “necessary nutrient” vs something to treat like alcohol.
  • Agreement that added sugars are excessive and ubiquitous; disagreement on carbohydrate necessity and keto/zero‑carb diets.
  • Multiple comments emphasize fiber and intact food matrices as central: whole foods with fiber blunt calorie absorption and spikes; juices and refined flours do not.

Societal and structural factors

  • Portion sizes, constant snack availability, delivery apps, sedentary lifestyles, stress, and urban design (walkability) are seen as major co‑drivers of obesity.
  • Japan cited as heavily processed yet lean; explanations include smaller portions, more walking, social norms.

Heuristics and personal rules

  • Common personal rules: “If it couldn’t be made outside a factory, don’t eat it”; “5 or fewer recognizable ingredients”; “eat food, not too much, mostly plants”; avoid factory foods when possible.
  • Others warn against purity politics and note some highly processed items (e.g., protein isolates) may be net beneficial in context.

Trust, regulation, and industry influence

  • Deep skepticism about regulators (e.g., GRAS system), health organizations, and industry‑funded research.
  • Calls for better ingredient transparency, public databases, and regulations targeting calorie density, additives, and misleading health branding.