Cohort studycited by the Tayebat archivefull text checked★ key study A-UPF-03 in-depth review

Ultra-processed food and death in Spanish graduates

Association between consumption of ultra-processed foods and all cause mortality: SUN prospective cohort study.

Overview

Rico-Campà A et al.. BMJ 2019 · PMID 31142450 · DOI 10.1136/bmj.l1949

More than four servings a day was linked with 62% higher death risk.

Does the study support the specific claim? Yes · Methodological quality: moderate

Original publication

Plain-language summary

Study explained simply

What did the researchers want to find out?: Is UPF consumption associated with all-cause mortality?
How was the study done?: Prospective cohort of Spanish university graduates; FFQ every 2 years.
How many people, animals or samples were studied?: 19,899; 335 deaths; University graduates aged 20-91
What was actually found?: Highest vs lowest quarter HR 1.62 (1.13-2.33); P trend 0.005; per additional serving HR 1.18 (1.05-1.33).
How large was the observed effect?: HR 1.62 (1.13-2.33)
What do the results mean in everyday life?: Association with uncertainty.
What are the study's limitations?: Only 335 deaths, so estimates are imprecise; diet questionnaire not designed for processing levels; people eating most UPF also smoked more and had more illness at baseline; results for heart and cancer deaths were not significant.
What does it say about Elawady's specific claim?: Supports C-UPF-03.
Do the results apply to Germany and Europe?: High (Spain).
What can explicitly not be concluded from this study?: Causation.

Scientific in-depth review

1. Full reference

Authors: Rico-Campà A, Martínez-González MA, Alvarez-Alvarez I, Mendonça RD, de la Fuente-Arrillaga C, Gómez-Donoso C, Bes-Rastrollo M
Journal: BMJ
Year: 2019
Identifier: PMID 31142450 DOI 10.1136/bmj.l1949 · PMC6538973
Verification: verified – PubMed E-utilities efetch, 2026-10-25
Retracted: no
Corrections: none
Extent checked: full text
Conflicts of interest and funding: Funded by Instituto de Salud Carlos III, European Regional Development Fund (FEDER), Navarra Regional Government and University of Navarra; funders had no role. ICMJE: no competing interests declared.

2. Research question

Is UPF consumption associated with all-cause mortality?

3. Study design and population

Study design: Cohort study – Prospective dynamic cohort (SUN, Spanish university graduates; registered NCT02669602); baseline 136-item validated FFQ classified by NOVA (34 UPF items), diet updated at 10 years in a sensitivity model.
Population: Spanish university graduates aged 20-91 (mean age 37.6, SD 12.3); 12,113 women, 7,786 men; participants with prevalent disease were included and adjusted for
Country: Spain
Period: Recruitment December 1999 to March 2014; follow-up to February 2018 (median 10.4 years)
Sample size: 19,899 (of 22,279 recruited; 2,215 lost to follow-up, 165 energy outliers); 335 deaths in 200,432 person-years
Inclusion and exclusion criteria: Recruited before March 2014 (minimum 2 years follow-up); excluded energy intake below 1st or above 99th centile and those lost to follow-up (retention 90%).

4. Methods and endpoints

Exposure / intervention: Energy-adjusted UPF servings (quarters)
Comparator: Lowest quarter
Endpoints: All-cause mortality
Statistical methods: Cox models with age as time scale; energy-adjusted UPF servings in quarters; multivariable adjustment (sex, marital status, activity, smoking, pack-years, snacking, special diet, BMI, energy, alcohol, education, family history of CVD, baseline diabetes, hypertension, hypercholesterolaemia, CVD, cancer, depression), stratified by recruitment period, age deciles, sedentary index and TV viewing; propensity-score and repeated-measures models; E-value; splines; subgroup and sensitivity analyses; Simes correction.

5. Actual results

Results: All-cause mortality, highest (>4 servings/day) vs lowest (<2/day) quarter: HR 1.62 (1.13-2.33), P trend 0.005; per additional serving HR 1.18 (1.05-1.33). Propensity score model 1.89 (1.34-2.67); repeated diet measures 1.44 (1.01-2.05). Cardiovascular deaths (n=71): HR 2.16 (0.92-5.06), not significant. Cancer deaths (n=164): HR 1.22 (0.70-2.12), not significant. E-value 2.62 for point estimate, 1.51 for lower CI. Association became non-significant when excluding baseline hypertension or baseline depression; not attenuated by adjusting for saturated/trans fat, added sugars or sodium.
Absolute risks: 335 deaths in 200,432 person-years. Deaths by quarter (lowest to highest): 108/49,814, 74/50,322, 80/49,971, 73/50,323 person-years; crude rates are not higher in the top quarter (younger participants), so the HR depends on age adjustment. Absolute risk differences not reported.
Authors' conclusion: Higher consumption of ultra-processed foods (>4 servings daily) was independently associated with a 62% relatively increased hazard for all-cause mortality; each additional serving with an 18% higher hazard. Discouraging UPF consumption should be considered in health policy.

6. Strengths

Prospective design; 90% retention; median 10.4-year follow-up
Validated FFQ (but not designed for NOVA)
Broad confounder adjustment; propensity score, E-value, repeated-measures and many sensitivity analyses
Homogeneous educational level limits socioeconomic confounding
Public funding; no COI

7. Weaknesses

Few deaths (335), so imprecise estimates; cause-specific analyses underpowered and non-significant
FFQ not designed for NOVA; some UPF items missing; serving counts may misclassify
High-UPF group was younger with more smoking, snacking, sedentary behaviour, baseline hypertension, depression and CVD: residual confounding plausible
Participants with prevalent disease were included (adjusted), so reverse causation is possible; association lost significance when excluding baseline hypertension or depression
Non-representative graduate cohort
Baseline diet only in main model

8. Risk of bias

Orienting tool: ROBINS-E
Type of assessment: structured assessment based on the full text by a single reviewer (no formal double review)
Full formal assessment: no
Overall risk of bias: some concerns
Rationale: Structured single-reviewer judgement from the full text; a complete formal ROBINS-E assessment requires two independent reviewers and a protocol.

8. Risk of bias
Domain	Judgement	Rationale
Confounding	some concerns	Many confounders and propensity score; strong lifestyle clustering with UPF; E-value 2.62.
Exposure measurement	some concerns	Validated FFQ not designed for NOVA; servings as metric; misclassification likely non-differential.
Selection of participants	some concerns	Prevalent disease included; 10% lost to follow-up excluded.
Post-exposure interventions	some concerns	Diet changes over time; repeated-measures model attenuated HR to 1.44.
Missing data	low	Little missing covariate data; multiple imputation gave unchanged results.
Outcome measurement	low	All-cause mortality confirmed via next of kin and National Death Index.
Selection of reported result	low	Registered cohort; null cause-specific results also reported; multiple models shown.

Full-text review (25 October 2026): Headline numbers confirmed. Added: registration, exclusions, death counts per quarter (crude rates not higher in top quarter), non-significant CVD (HR 2.16, 0.92-5.06) and cancer (1.22, 0.70-2.12) mortality, propensity (1.89) and repeated-measures (1.44) models, E-value 2.62, loss of significance excluding baseline hypertension/depression, inclusion of prevalent disease, funding. Domain-level ROBINS-E (overall some_concerns, unchanged).

9. Assessment of the statistical interpretation

Imprecise top-quarter estimate based on 73 versus 108 deaths; robust across several models, but significance lost in two sensitivity analyses. Cardiovascular and cancer mortality were not significantly associated. Relative, not absolute, risk; crude death rates were lower in the high-UPF (younger) group.

10. Relevance to the specific claim

Supports C-UPF-03.

11. Independent contrary findings

None.

12. Transferability to Germany and Europe

High (Spain).

13. Overall methodological judgement

Supportive for all-cause mortality in a single, small-event cohort; does not by itself support cardiovascular or cancer mortality effects.

Multidimensional assessment
Multidimensional assessment
Authenticity of the source	verified
Methodological quality	moderate
Risk of bias	some concerns
Precision of results	low
Directness	direct
Reproducibility	replicated
Consistency with the overall evidence	consistent
Regional transferability	high
Adequacy of the interpretation (by the archive)	essentially correct but incomplete

14. Open questions

Causality.

15. Linked fact checks

Ultra-processed foods raise the risk of cardiovascular disease, type 2 diabetes, cancer and all-cause mortality; an umbrella review found 'convincing' evidence for cardiovascular mortality and type 2 diabetes. – Studies supporting the claim

Use by the Tayebat archive

Note: according to the archive, the studies were compiled by the archive operators. Whether Dr. Elawady himself cited this study is not established.

What the archive says about the study: Spanish SUN cohort of 19,899 — more than four daily servings associated with a 62% rise in all-cause mortality; each serving +18%.
A. What was actually studied?: UPF and mortality.
B. What did the authors find?: HR 1.62 and 1.18 per serving.
C. What conclusions do the authors themselves draw?: As stated.
D. What does the archive derive from it?: '62% higher risk of dying' presented without the wide CI.
E. Is this derivation scientifically justified?: Numbers correct; imprecision not mentioned; the study found no significant association with cardiovascular or cancer deaths separately.
Adequacy of the interpretation (by the archive): essentially correct but incomplete