a slice of pizza on a white plate

Does Cutting Processed Foods Improve Health?

A small trial in adults 65+ cut ultra-processed food to about 13% of calories: they ate less, lost about 4 kg, and markers improved — but the cause is unclear.

Older adults who cut ultra-processed food to about 13% of calories ate less and lost about 4 kg — but the trial can’t separate the diet change from the weight loss

In a small controlled feeding trial, adults 65+ who cut ultra-processed food from about half their calories to about 13% ate a few hundred fewer calories a day and lost about 4 kg — gains that faded a year later.

What this means

In a small controlled feeding trial, adults 65 and older who cut ultra-processed food from about half of their calories to about 13% — with no instruction to eat less — ate a few hundred fewer calories a day, lost about 4 kg (8.8 lb), and saw insulin-resistance, cholesterol and inflammation markers improve. The boundary: these were secondary outcomes in a trial with no high-ultra-processed control arm, and because participants also ate less and lost weight, the study cannot say whether cutting ultra-processed food itself drove the changes. What remains uncertain is whether the improvements would hold if the diet were kept up — a year after the trial, as ultra-processed intake climbed back, most of them faded.

The question, and a press claim the paper doesn’t support

Does cutting ultra-processed food improve older adults’ metabolic health on its own — or only because they end up eating less? A controlled feeding trial published in Clinical Nutrition found improvements on several markers, but it cannot separate the two. The press framing that the diet improved how the body “regulates appetite and metabolism” is not supported by the paper: it measured no appetite, hunger, satiety or craving outcome. FGF21 and leptin, the two hormones it tracked, are nutrient-sensing hormones, not appetite measures.

What the trial did

PRODMED2 (NCT05581953) was an 18-week randomized, double-blind crossover feeding trial that provided all food. Participants began on their usual diet, about half of it ultra-processed (UPF), then — in random order — spent two 8-week stretches on Dietary Guidelines-aligned menus that cut UPF to about 13% of calories, one built on minimally processed pork and one on lentils, separated by a washout. The trial’s primary endpoints were insulin, ferritin, homocysteine and chair-stand performance. Everything reported here is secondary or exploratory; the trial was not powered for these measures, was not designed as a UPF-reduction intervention, and had no high-UPF control arm — each person’s own habitual diet was the comparison.

Who took part

The 36 adults analyzed were 65 and older, living in the rural Upper Midwest (26 women, 10 men; mean age 71.7; mean BMI about 28, the overweight range). All were White and about 75% were college-educated. Earlier controlled trials in adults found that freely eating UPF raises intake and weight; this one asked the reverse question.

What changed — and what didn’t

With no instruction to eat less, participants ate about 333 (pork) and 437 (lentil) fewer kilocalories a day and lost 3.8 kg (8.4 lb) and 4.4 kg (9.7 lb) of body weight plus 2.6–2.9 kg (5.7–6.4 lb) of fat mass. Insulin resistance fell (HOMA-IR 2.24 to 1.71 and 1.73); total, LDL and non-HDL cholesterol and ApoB improved; inflammation markers fell; FGF21 rose and leptin fell. Triglycerides and adiponectin did not change, and pork and lentils performed identically.

The gains faded

About a year after the trial — a follow-up added after the fact, with 32 participants — UPF intake had climbed back to about 44% of calories. Weight and most markers drifted back toward baseline, and fasting glucose ended above where it began (100.3 vs 93.6 mg/dL).

What the companion paper adds

The trial’s registered primary endpoints have since been published. Across both low-UPF phases, fasting insulin fell and grip strength and chair-rise performance were maintained rather than improved. The measured between-diet differences were small: HDL cholesterol was higher after the minimally processed pork phase (P=0.034), and homocysteine was lower after the lentil phase (P=0.009). Those authors describe the weight loss as unintentional and say it “may have contributed” to the metabolic improvements — the same caveat this brief gives. Blood pressure was registered as a secondary outcome; no result for it is reported.

What this can’t say

UPF share, fiber, food form and calories all changed together while participants lost weight, so the trial cannot attribute the improvements to cutting UPF itself rather than eating less overall; the authors concede it was not designed to pinpoint the pathways. It tested 36 people from one narrow population. A secondary analysis of a separate controlled UPF trial in lean young adults (Testing ultra-processed food and fullness-hormone responses) found a near-total UPF diet left satiety-hormone responses unchanged — a bounded null in a markedly different group. The trial was funded by the National Pork Checkoff and USDA NIFA. The authors declare no competing interests; the headline finding — minimally processed pork matching lentils — happens to favor the funder’s product.

Terms explained

  • Ultra-processed food (NOVA category 4): the top tier of the NOVA classification, covering industrially formulated products built largely from refined ingredients and additives rather than whole foods. Here, UPF supplied about 50% of calories at baseline and about 13% during the low-UPF diets.
  • Crossover feeding trial: a design in which every participant completes each diet being compared, in randomized order, with a rest period (washout) in between — so each person serves as their own comparison, as here.
  • Secondary outcome: a result the trial measured but was not designed or powered to answer with. Secondary and exploratory findings are leads to check, not firm conclusions.
  • HOMA-IR: a score calculated from fasting glucose and insulin, used as a surrogate for insulin resistance; higher values mean more resistance. Here it fell during both low-UPF diets.
  • FGF21 and leptin: two hormones involved in nutrient sensing and energy metabolism. FGF21 rose and leptin fell in this trial; neither is a direct measure of appetite, hunger or fullness.

Related coverage: Nutritious News also reports a null satiety-hormone finding from a controlled UPF trial in lean young adults (Testing ultra-processed food and fullness-hormone responses).

Keep reading on Nutritious News: in Is "Healthy" Ultra-Processed Food as Good?, a randomised crossover trial compared “healthful” ultra-processed and minimally processed diets and found processing mattered less than expected; Can Ultra-Processed Food Hurt Your Focus? weighs a new observational signal linking higher ultra-processed intake to attention scores.

Keep reading on Nutritious News: How many daily steps lead to good weight management? examines a 14-trial analysis where higher step counts tracked with better-maintained weight loss — an association, not a proven target.

Keep reading on Nutritious News: Does a Longevity Diet Work in Humans?.

Sources

  1. Primary source: Vaezi S, Freeling JL, de Vargas BO, Weidauer L, Shoemaker ME, Sanders WM, Dey M. “Impacts of minimally-processed omnivorous vs lacto-ovo-vegetarian diets on insulin sensitivity, lipid profile, and adiposity in older adults: Secondary findings from a randomized crossover feeding trial.” Clinical Nutrition, online-first 2025-10-29; print issue 55 (December 2025), pp 90–103. Open access (CC BY-NC-ND). DOI: 10.1016/j.clnu.2025.10.010 · PMID: 41202666 · Registered trial: NCT05581953. Full-text access.
  2. Trial registration: PRODMED2, ClinicalTrials.gov NCT05581953 (status completed; randomized crossover). Registered primary outcomes: fasting insulin, ferritin, homocysteine, and chair-stand performance (ClinicalTrials.gov NCT05581953).
  3. Press account being corrected on the appetite claim: “Eating less ultraprocessed food supports healthier aging, new research shows,” The Conversation, January 12, 2026 (syndicated by PBS NewsHour, January 17, 2026) — a first-person account by the trial’s principal investigator. https://theconversation.com/eating-less-ultraprocessed-food-supports-healthier-aging-new-research-shows-271986
  4. Cross-referenced coverage: van Bruinessen M, et al. “Satiety responsiveness following dietary exposure to ultra-processed foods differing in texture-derived eating rate: a secondary analysis of a randomized controlled crossover diet intervention.” European Journal of Nutrition 2026;65(7):261. DOI: 10.1007/s00394-026-04111-7 · PMID: 42782361.
  5. Context (established short-term finding in adults; mean age about 31 years): Hall KD, et al. “Ultra-Processed Diets Cause Excess Calorie Intake and Weight Gain: An Inpatient Randomized Controlled Trial of Ad Libitum Food Intake.” Cell Metabolism 2019;30(1):67–77.e3. DOI: 10.1016/j.cmet.2019.05.008 · PMID: 31105044. Cited for the established short-term finding only.
  6. Context (replication in a different population): Hamano S, et al. “Ultra-processed foods cause weight gain and increased energy intake associated with reduced chewing frequency: a randomized, open-label, crossover study.” Diabetes, Obesity and Metabolism 2024;26(11). DOI: 10.1111/dom.15922 · PMID: 39267249. Abstract-level access; cited for the established short-term finding only.
  7. Companion paper (same trial, registered primary outcomes): Vaezi S, de Vargas BO, Weidauer L, Freeling JL, Dey M. “Effects of Minimally Processed Red Meat within a Plant-Forward Diet on Biomarkers of Physical and Cognitive Aging: A Randomized Controlled Crossover Feeding Trial.” Current Developments in Nutrition 2026;10(1):107615 (epub 2025-12-10). DOI: 10.1016/j.cdnut.2025.107615 · PMID: 41551721.

Related Articles