A global longitudinal study found that ecological links between food security research publications and real-world agrifood-system indicators are highly selective rather than uniform across global economies. Analyzing the scientific response to food security in public health, researchers evaluated whether annual research volume correlates with actual societal conditions across World Bank income groups between 2000 and 2025. The investigators emphasized that raw publication output does not directly reflect research implementation or concrete societal benefits.
To evaluate these relationships over time, the authors constructed a comprehensive analytical framework combining bibliometric records with structural data. The research team assembled a corpus of 61,158 Scopus-indexed publications and integrated them with 24 indicators spanning productive output, economic measures, resource use, and workforce metrics. This secondary-data framework organized observations across income-group-year units over the 26-year study window, allowing the team to inspect contemporaneous ecological associations across distinct national economic tiers.
The investigation applied multiple statistical techniques to test these relationships while accounting for potential confounding trends and multiple testing. Analytical approaches included exploratory income-stratified bivariate regressions, mixed-effects models incorporating centered calendar years and income-group random effects, and exploratory publication-moderator interaction models. To control for multiplicity and reduce false-positive discoveries, the authors applied Holm adjustments to the income-stratified analyses and Benjamini-Hochberg false discovery rate corrections to both the mixed-effects models and interaction evaluations.
Initial bivariate regressions suggested widespread statistical significance that largely dissipated after rigorous multiplicity control. Among 96 income-stratified associations evaluated, 66 demonstrated nominal p-values below 0.05. However, following the Holm adjustment, only 11 associations maintained statistical significance below the adjusted threshold. The authors noted that this sharp reduction demonstrates how much initial statistical evidence was not retained once multiple testing was formally addressed, underscoring the instability of unadjusted ecological correlations.
When examining the 24 mixed-effects models across the dataset, only seven maintained statistically significant false discovery rate-adjusted p-values below 0.05. These robust associations were confined to specific production sectors and structural economic measures, specifically beef, chicken, and pig meat production, aquaculture, fish and seafood production, overall agricultural output, and agriculture’s share of gross domestic product. Across other productive metrics, resource utilization, economic indicators, and workforce parameters, associations differed substantially in both direction and statistical support across differing income settings.
Contextual conditions further altered how scientific output related to agrifood conditions across the globe. Through moderator screening, the researchers identified 294 false discovery rate-adjusted interactions out of 989 examined outcome-moderator combinations. These interactions revealed that publication-outcome relationships were frequently attenuated, amplified, or reversed in direction across varying contextual environments.
Based on these findings, the authors concluded that scientific activity cannot be evaluated independently of the specific outcome under study or the context where knowledge is generated. Ecological correspondence persisted only for a restricted group of productive and structural agrifood metrics, while broader associations weakened after adjustment or varied across economic strata. Consequently, tracking raw scientific output alone offers an incomplete and potentially misleading proxy for real-world agrifood-system progress.
Subject of Research: Global ecological associations between food security scientific publications and agrifood-system indicators
Article Title: Scientific response to food security in public health: a global longitudinal secondary-data study of governance and population-level indicators
Article References: Vergara-Rios, J. C., Barceló-Castellanos, E. M., Barceló-Martinez, E., Visconti-Lopez, F. J., Acosta-Monterrosa, A. A., & Lozada-Martinez, I. D. (2026). Scientific response to food security in public health: a global longitudinal secondary-data study of governance and population-level indicators. BMC Public Health. https://doi.org/10.1186/s12889-026-30428-3
Image Credits: AI Generated
DOI: 10.1186/s12889-026-30428-3
Keywords: food security, public health, agrifood systems, scientific publication output, World Bank income groups, secondary data analysis, longitudinal study, mixed-effects models, bivariate regression, multiplicity adjustment, Holm adjustment, false discovery rate, meat production, aquaculture, agricultural output, gross domestic product, moderator analysis, ecological association
News Source: Ophelia Keating. (November 18, 2026). Food Security Research Correlates Selectively With Global Agrifood Indicators. Scienmag.



