Background:
The EAT-Lancet Commission proposed the Planetary Health Diet (PHD), which emphasizes plant-based foods and reduced intake of red meat and added sugars to improve human and planetary health. Several PHD adherence indexes exist, but their heterogeneity limits comparability across different dietary assessment tools and study populations. In addition, no study has yet investigated the association between PHD indexes and gastric cancer (CG).
Objectives:
Within the StoP Project, a global consortium of epidemiologic studies on GC, we evaluated the applicability and comparability of four PHD adherence indexes, EAT-Lancet Index (EATL), World Index for Sustainability and Health (WISH), PHD Indexes by Cacau (PHDI-C) and Bui (PHDI-B), and their association with GC risk.
Methods:
We included ten studies from Europe, Asia, and Latin America, with data from validated food frequency questionnaires (FFQs) and total energy intake (TEI). Dietary data were harmonized, converted to daily intakes (g/day), standardized for TEI, adjusted for seasonality and cooking processes, and disaggregated into constituent ingredients for mixed dishes.
PHD adherence was assessed using the four indexes. We summarized adherence distributions across studies and evaluated agreement between indexes through random-effects meta-analyses of study-specific Spearman (ρ) coefficients and quartile classification. A two-stage approach was applied to study the association with GC: study-specific odds ratios (ORs) and 95% confidence intervals (CIs) were estimated using multivariable logistic regression models for 10% increase in adherence score and across quartiles. Models were adjusted for age, sex, BMI, socioeconomic status, smoking, alcohol, and TEI. Estimates were then pooled using random-effects meta-analysis. Subgroup and sensitivity analyses were performed to assess heterogeneity and robustness.
Results:
The analysis included 12,804 participants (3,748 GC cases and 9,056 controls). Cases were more frequently males (62.1% vs 53.6%); median age was 64.0 vs 63.0 years. Median BMI was 24.8 and 25.9 kg/m², respectively, and low socioeconomic status was more frequent among cases (59.8% vs 47.0%). Never smokers were 44.7% and 47.6%, and never alcohol consumers 22.7% and 26.8%, respectively. Among controls, mean adherence varied across indexes and studies, with median values ranging from 35% to 60% of the maximum achievable score. Correlations between indexes were high to strong (ρ=0.6–0.8), while quartile agreement was moderate (~50%). Higher PHD adherence was associated with lower GC risk across all indices: EATL (OR=0.92; 0.85–0.99; I²=0.0%), PHDI-C (OR=0.87; 0.81–0.94; I²=0.0%), PHDI-B (OR=0.81; 0.72–0.91; I²=40.6%), and WISH (OR=0.86; 0.79–0.93; I²=0.0%). Comparing the highest versus lowest adherence quartiles, ORs were 0.58 (0.45–0.74) for PHDI-B, 0.63 (0.48–0.82) for PHDI-C, 0.34 (0.26–0.45) for EATL, and 0.56 (0.45–0.70) for WISH, with linear trends across quartiles. Associations were consistent across strata and persisted after adjustment for Helicobacter pylori infection. Results were robust in one-stage and sensitivity analyses.
Conclusions/Implications for practice or policy:
Despite methodological heterogeneity in their construction and application, PHD adherence indexes are applicable to international epidemiologic data and provide a consistent ranking of individuals. Greater adherence to sustainable dietary patterns may protect from GC. These results highlight the potential role of dietary changes in cancer prevention, particularly in reducing the burden of GC.

Pooled odds ratios (ORs) and 95% confidence intervals (CIs) for gastric cancer per 10% increase in adherence and by quartiles of the Planetary Health Diet, using two-stage and one-stage models across four dietary indexes