Study record
Levine et al. dietary protein observational cohort (n=6381)
All-cause mortality
Human — Adults aged 50-65, prospective observational cohort, n=6381 Tier 4 · Prospective human observational verification: verified
1. Species / populationAdults aged 50-65, prospective observational cohort, n=6381
2. Exposure, route, scheduleSelf-reported high dietary protein intake, observational
3. Comparator & durationLow dietary protein intake; prospective follow-up
4. Endpoint & numeric result~75% higher all-cause mortality (age 50-65), age-dependent, attenuated for plant protein
5. What this study did NOT establishObservational, not causal; not generalizable across ages or protein sources; no RCT
Full record
| Intervention(s) | Protein Restriction |
| Species | Human |
| Population | Adults aged 50-65, prospective observational cohort, n=6381 |
| Sample size | 6381 adults verification: verified_2026-07-11 |
| Exposure | self-reported dietary protein intake (high vs low) — Observational exposure, not an assigned intervention verification: verified_2026-07-11 |
| Comparator | Low dietary protein intake |
| Duration | Prospective observational follow-up |
| Endpoint | All-cause mortality |
| Evidence tier | Tier 4 · Prospective human observational |
| What was NOT established | Observational design does not establish causation. Effect was specific to the 50-65 age band and attenuated for plant protein, so it cannot be generalized as a single fixed effect across ages or protein sources. No randomized human trial of protein-restriction dose vs. mortality exists in this dataset. |
Result values (as reported in sourcing brief)
| mortality association | ~75% higher all-cause mortality with high protein intake in the 50-65 age band |
| age dependence | Age-dependent; attenuated for plant-derived protein |
| verification | verification: verified_2026-07-11 |
Source citation
Verification status: verified This record was checked against its primary publication on 2026-07-11.Identifiers: PMID verified; DOI verified; PMC not reported; trial registry ID not applicable. Record-level verification and identifier-level status are separate: a record can be fully verified against its primary publication while an identifier that does not apply to the study type — a trial registry ID for an animal study, for instance — is marked not applicable. No identifier is fabricated, and none is left in a “pending” state.
| trial name | Levine et al. observational cohort |
| pmid | 24606898 verified |
| doi | 10.1016/j.cmet.2014.02.006 verified |
| journal | Cell Metabolism 2014;19:407-417 |
| trial registry id | not applicable not applicable |
| authors | Levine ME, Suarez JA, Brandhorst S, et al. |
| year | 2014 |
| verified date | 2026-07-11 |
| verified note | NHANES III (~6381 adults >=50); ages 50-65 high protein assoc. +75% all-cause mortality, ~4x cancer mortality; attenuated for plant protein; opposite/attenuated in >65; +5x diabetes mortality all ages — MATCHES. OBSERVASJON, aldersavhengig. |
| verification status | verified |
Funding / conflict of interest
not compiled in this proof-of-concept verification: verified_2026-07-11
Version 0.2.0-mvp · literature search date 2026-07-11 · editorial owner Nabus Research