David Keith Cundiff- Volunteer Collaborator with the Institute of Health Metrics and Evaluation, San Anselmo, CA, United States
Introduction
The EAT-Lancet Commission Report is a sweeping global effort to define a “planetary health diet” (PHD) that nourishes humanity while remaining within ecological limits (1). Anchored in Stockholm, the initiative aspires to unite human and environmental health through a coherent nutritional framework.
While the vision is laudable, the scientific footing of its dietary targets—particularly those concerning animal-source foods—remains uncertain. The Commission relies heavily on the Theoretical Minimum Risk Exposure Level (TMREL) framework developed by the Institute for Health Metrics and Evaluation (IHME) (2). TMREL estimates the intake level of each food associated with the lowest disease risk and forms the backbone of the Global Burden of Disease (GBD) studies and, by extension, the PHD.
The idea of the TMREL predates the Institute of Health metrics and evaluation, which was founded in 2007. The concept of using a counter factual minimum risk level predates the IHME but the IHME formalized the TMREL term within the GBD's comparative risk assessment framework. Therefore, while the core idea is older, its formal use and publication by IHME as the TMREL within the detailed GBD study reports became standard practice in the mid-to-late 2010s.
TMREL values are derived primarily from meta-analyses of observational studies, largely conducted in high-income nations. Extrapolating these findings globally risks neglecting the nutritional realities of low- and middle-income populations still facing under nutrition and micronutrient deficiencies (3).
The problem with TMREL
The wide fluctuations in TMREL estimates—ranging from zero to over 20 g/d—underscore the instability of current modeling approaches. Theoretical constructions may produce elegant visualizations but remain detached from real-world diversity. Ecological data, while imperfect, reflect how populations actually thrive or falter under varying dietary patterns.
Moreover, IHME's own GBD updates reveal striking instability in TMREL estimates for red meat: GBD TMREL.
Year | Red meat (g/d)
2019 | 22.5 | “Threshold for harm” (not TMREL)
2020 | 0 | Any intake deemed harmful
2022 | 0 (95% UI: 0–200) | Extremely wide uncertainty
Such volatility suggests model instability rather than empirical consistency. The EAT-Lancet Commission, whose Planetary Health Dietary recommendations generally mirror these IHME TMREL estimates—red meat 15 g/d (range 0–28) and processed meat 0 g/d—rests on an uncertain foundation.
Methods
In 2023, statistician Chunyi Wu, PhD, and I compared the EAT-Lancet Planetary Health Diet (PHD) with Global Burden of Disease (GBD)–derived mortality and dietary data (4). The full GBD data analysis methodology underlying the tables and results in this report is detailed in that publication. All associated Data and Statistical Analysis System (SAS) code is publicly available on Mendeley depository (5).
We obtained the IHME GBD data as volunteer collaborators in 2018, receiving the raw GBD 2017 dataset in R format. Readers may download the analysis database (termed: wtedCVDRfsCov2017), along with (1) SAS code to import the data, (2) SAS code to format the database for analyses, and (3) multiple SAS programs for evaluating associations between health outcomes—such as non-communicable disease (NCD) early deaths (ages 15–69)—and approximately 20 food variables, physical activity, smoking, air pollution, and other risk factors.
Readers are encouraged to become IHME volunteer collaborators and obtain GBD 2023 raw data to further extend the analyses that we have done.
Results
A GBD data–driven alternative
The dietary intakes and premature NCD mortality (ages 15–69) across 1,000 cohorts (1 million people per cohort) representing the world's longest-lived 1 billion people are presented in Table 1. Table 2 has 1,014 cohorts with NCD deaths/100k/year m/f ages 15–69 > 1,070 and Sociodemographic index (SDI) > 0.67.
Table 1. GBD data from 1990–2017 representing the world's longest living people.
Table 2. Cohorts of people with Socio-demographic index (SDI) > 0.67 and are not in Table 1.
Compared to Table 1, Table 2 cohorts consumed 37% more red and processed meat, only 22% of the fish intake, and 57% of the legume consumption. Additionally, the average BMI of Table 2 cohorts was 1.24 units higher than that of Table 1 cohorts.
The “sweet spot” for meat and longevity
Synthesizing these findings, the optimal longevity range in Table 1 corresponds to roughly 34.5 g/d of combined red and processed meat—approximately 1.7 times the current meat consumption global percapita g/d average. This suggests that humanity may benefit from about 70% more, not less, animal-source food, particularly among the lower SDI of three-fourths of the human population (Table 3).
Table 3. About 5.8 billion people from developing countries, SDI < 0.67, and not in Table 1.
US data are revealing. Only Alaska, Colorado, Hawaii, Washington, and Wyoming (Totaling 26 million US residents out of 336 million) were included in Table 1. The rest were in Table 2. Washington, DC had the highest US mortality rate—1,870.5 NCD deaths m/f/100k/year. Mississippi was second—1,500 NCD deaths m/f/100k/year. Average meat consumption for US residents was 68.1 g/d, compared with 47.2 g/d mean meat intake in Table 2 and 34.5 g/d in Table 1.
Discussion
It should be noted that ecological associations cannot infer individual-level causality. While empirical GBD data show that the sweet spot of combined processed meat and red meat consumption is roughly 34 g per day, individual level causality Is not inferred and would present an ecological fallacy (e.g., “more meat = better survival or less meat equals better survival”).
LDL-c mmol/L levels of Tables 1 and 2 were very similar and far higher than in Table 3. LDL-c showed limited sensitivity at ecological scale, indicating possible scale-mismatch when informing global dietary TMREL targets. Patterns and relationships observed in nature can change dramatically depending on the scale at which you look at them.
Potential confounders inherent to ecological comparisons—such as food system availability, healthcare access, infectious disease burden, environmental exposures, and total energy intake—are competing explanations for mortality differences, particularly across SDI strata.
In all three Tables, mean SBP mm Hg differed little while NCDs m/f/100k/year increased over 60% in Table 2 and Table 3, possibly indicating that intakes of foods were the primary influences on longevity and SBP secondary in determining NCD early deaths. Meanwhile CVD deaths in Table 2 (585.70 CVD deaths/100K/year) and Table 3 (567.35 CVD deaths/100K/year) far surpassed 274.65 CVD deaths/100K/year in Table 1. This implies that food profiles, particularly meat, contribute strongly, perhaps by different mechanisms, to early CVD mortality. Along the same line, Table 2 shows that early cancer and CVD mortality were both high and associated with high red and processed meat.
These results parallel the findings of You et al. (6) who analyzed 175 populations using FAO and WHO data and found that higher total meat intake correlated with longer life expectancy, even after adjusting for caloric intake, urbanization, and education. Together, these data suggest that moderate meat consumption contributes to optimal health outcomes, particularly where nutrient deficiencies persist as in Table 3. Based on these data, people with diets fitting in with Table 2 data might consider reducing or eliminating meat consumption while increasing fish and lentils. Conversely, people with low meat intake in Table 3 could try to markedly increase overall intake of animal-based as well as plant-based food.
The dietary contrast between high-SDI cohorts (Table 2) and the longest-lived cohorts (Table 1) —especially the much lower fish and legume intake in Table 2 despite higher meat exposure—deserves deeper mechanistic discussion to support the hypothesis of a composite dietary profile effect rather than a single food driver.
Notably, the average metabolic parameters in Tables 1, 2 were not significantly different except for BMI and LDL-c in Table 3: BMI kg/M2 (23.76 BMI kg/M2 m/f Table 1 vs. 25.0 BMI kg/M2 m/f Table 2 vs. 20.86 in Table 3), LDL cholesterol mmol/L (LDL cholesterol: 2.74 mmol/L Table 1 vs. 2.84 mmol/L Table 2 vs. 2.20 in Table 3), systolic blood pressure mm HG (SBP: 134.1 mm Hg Table 1 vs. 134.7 mm Hg Table 2 and 133.75 mm Hg in Table 3), and type 2 diabetes deaths (12.59/100K/year in Table 1 vs. 13.61/100K/year in Table 2 vs. 19.02/100K/year in Table 3).
To monitor intake of food to decrease BMI and/or to decrease risk of cancer and CVD, apps that may help in accurately following food intake include Levels (7) and Cronometer (8).
The ecological impact of agriculture in creating a healthful diet
To craft credible global dietary policy, models must integrate data from diverse regions and income levels. For much of humanity, animal-source foods remain vital for nutrient adequacy, child development, and disease resilience. A planetary diet that disregards this biological and social reality risks deepening, rather than narrowing, global health inequities.
The consensus across major international bodies and peer-reviewed literature is that the expansion of conventional, high-impact animal agriculture (especially confined animal feeding operations (CAFOs)) is a leading global driver of negative environmental change (9).
For example, livestock in African countries contribute to about 10% of enteric methane emissions from dairy cattle worldwide despite producing only 3.9% of the world's milk. Livestock in Sub-Saharan Africa (SSA) also cause extensive land degradation with 48% of rangelands in SSA degraded due to overgrazing. Strategies for sustainable intensification of livestock such as improving quality of feed, wholistic grazing as promoted by Alan Savory (10), land rehabilitation, introduction of improved forages and silvopastoral systems, and improvement of herd genetics can reduce both total emissions and emission intensity while improving productivity (11).
According to the Food and Agriculture Organization of the UN (12), the underlying causes of climate change, such as reliance on conventional agriculture grown with chemical fertilizers, herbicides, and pesticides for 98% of food for humanity, must be addressed by shifting from conventional agriculture to growing food with organic, regenerative methods. Poor farmers in developing countries especially benefit from farming biointensively with hand tools, growing their fertilizers, saving seeds, close spacing of plants to save on water, weeding, fossil fuels, and land (13). If scaled globally with communities living in ecovillages with few or no fossil fuel requiring vehicles, biointensive farming could dramatically reduce greenhouse gas emissions.
Conclusion
The EAT-Lancet Commission has sparked an essential conversation linking diet, health, and planetary boundaries. Yet inspiration must be grounded in empirical coherence. The volatility of TMREL estimates and their limited alignment with global dietary data highlight the need for recalibration. IHME's own GBD 1990–2017 data—contrary to TMREL modeling—offers consistent, population-level evidence capable of identifying real-world nutritional “sweet spots”—e.g., red meat and processed meat in Table 1 vs. Tables 2, 3. Empirically derived, globally representative dietary thresholds can align nutritional guidance with both human vitality and planetary stewardship. Future planetary health research should embrace this empirical turn—toward nutritionally adequate, culturally diverse, and environmentally balanced diets that sustain both people and planet. The EAT-Lancet Commission has sparked an essential conversation linking diet, health, and planetary boundaries rally diverse, and environmentally balanced diets that sustain both people and planet.
Data availability statement
Publicly available datasets were analyzed in this study. This data can be found here: https://data.mendeley.com/datasets/64gv2ffx72/1.
Author contributions
DC: Formal analysis, Project administration, Conceptualization, Funding acquisition, Writing – review & editing, Methodology, Validation, Supervision, Writing – original draft, Software, Investigation, Visualization, Data curation, Resources.
Funding
The author(s) declared that financial support was not received for this work and/or its publication.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was used in the creation of this manuscript. I used Generative AI to find the Frontiers journal with the best fit for my paper, to search for the optimal references, and to critique my drafts.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher's note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
1. Rockström J, Thilsted SH, Willett WC, Gordon LJ, Herrero M, Hicks CC, et al. The EAT lancet commission on healthy, sustainable, and just food systems. Lancet. (2025) 406:1625–700. doi: 10.1016/S0140-6736(25)01201-2
2. Afshin A, Sur PJ, Fay KA, Cornaby L, Ferrara G, Salama JS, et al. Health effects of dietary risks in 195 countries, 1990–2017: a systematic analysis for the global burden of disease study 2017. Lancet. (2019) 393:1958–72. doi: 10.1016/S0140-6736(19)30041-8
3. Stanton AV. Unacceptable use of substandard metrics in policy decisions which mandate large reductions in animal-source foods. NPJ Sci Food. (2024) 8:10. doi: 10.1038/s41538-024-00249-y
4. Cundiff DK, Wu C. The EAT-lancet commission's planetary health diet compared with the institute for health metrics and evaluation global burden of disease ecological data analysis. Cureus. (2023) 15:e40061. doi: 10.7759/cureus.40061
5. Cundiff D, Wu, C. Mendeley Data: The Eat-Lancet Commission's Planetary Health Diet Compared with the Ihme Gbd Ecological Data Analysis: Data, Excel Files, and Sas Codes (2024). Available online at: https://data.mendeley.com/datasets/64gv2ffx72/1 (Accessed February 14, 2024).
6. You W, Henneberg R, Saniotis A, Ge Y, Henneberg M. Total meat intake is associated with life expectancy: a cross-sectional data analysis of 175 contemporary populations. Int J Gen Med. (2022) 15:1833–51. doi: 10.2147/IJGM.S333004
7. Inside Levels Articles. New York, NY: Levels (2025). Available online at: https://www.levels.com/blog/category/inside-levels (Accessed August 3, 2025).
9. FAO. Livestock's Long Shadow: Environmental Issues and Options. Rome: Food and Agriculture Organization of the United Nations (2006).
10. Savory A. Agriculture Will Be Regenerative, When Management Is Holistic: Soil4Climate (2025). Available online at: https://soil4climate.org/holistic-planned-grazing# (Accessed November 6, 2024).
11. Balehegn M, Kebreab E, Tolera A, Hunt S, Erickson P, Crane TA, et al. Livestock sustainability research in Africa with a focus on the environment. Animal Front. (2021) 11:47–56. doi: 10.1093/af/vfab034
12. FAO. Organic Agriculture, Environment and Food Security. Rome: Food and Agriculture Organization of the UN (2002). Available online at: https://www.fao.org/4/y4137e/y4137e00.htm#TopOfPage.
Keywords: biointensive agriculture, global burden of disease data, meat consumption “sweet spot”, non-communicable diseases (NCDs), planetary health diet
Citation: Cundiff DK (2026) Rethinking the planetary health diet: GBD data reveal a “sweet spot” for red and processed meat and longevity. Front. Nutr. 12:1736366. doi: 10.3389/fnut.2025.1736366
Received: 31 October 2025; Revised: 18 December 2025;
Accepted: 30 December 2025; Published: 09 February 2026.
Edited by:
Raul Zamora-Ros, Institut d'Investigacio Biomedica de Bellvitge (IDIBELL), SpainReviewed by:
Hongyang Gong, Chosun University, Republic of KoreaCopyright © 2026 Cundiff. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: David Keith Cundiff, ZGF2aWRrY3VuZGlmZkBnbWFpbC5jb20=