Ultraprocessed food consumption and the global burden of noncommunicable disease: a 45-country comparative risk assessment.

BACKGROUND

Ultraprocessed foods (UPFs) increasingly contribute to diets in high-income countries (HICs) and are rapidly displacing traditional food systems in low- and middle-income settings (LMICs). Recent meta-analyses have established dose-response associations between UPF consumption and multiple noncommunicable diseases (NCDs), yet the global disease burden attributable to these foods remains unquantified across diverse populations.

OBJECTIVES

This study aimed to estimate population attributable fractions (PAFs) and attributable burden, including disability-adjusted life years (DALYs) and deaths, for 8 NCD outcomes associated with UPF consumption across 47 country/regional estimates spanning all World Bank income groups, with attributable burden derived for the 45 estimates with matching WHO Global Health Estimates (GHE) cause data and type 2 diabetes (T2D) as the prespecified primary outcome.

METHODS

We conducted a comparative risk assessment using UPF consumption data (% energy from NOVA Group 4 foods) from 47 dietary surveys, dose-response relative risks per 10% energy increment from meta-analyses, and the 2021 WHO GHE. PAFs were computed with a theoretical minimum risk exposure level (TMREL) of 10% energy and Monte Carlo simulation (10,000 iterations) for uncertainty quantification.

RESULTS

A clear income gradient was observed: mean T2D PAFs were 20.2% in HICs (n = 32), 14.2% in upper-middle-income countries (n = 6; excluding Lebanon), 0.9% in LMICs (n = 6), and 2.7% in low-income countries (n = 2). UPF consumption ranged from 4.0% (Ethiopia) to 57.5% (United States). Country-specific T2D PAFs reached 41.4% [95% confidence interval (CI): 29.2%, 52.2%] in the United States, 40.8% (28.4%, 51.6%) in the United Kingdom, 36.2% (24.6%, 46.6%) in the Netherlands, and 35.5% (24.2%, 45.5%) in Norway. UPF consumption above the TMREL was associated with an estimated 3.4 million all-cause deaths (95% CI: 1.3, 6.0 million) and 5670 thousand (3266, 8348) T2D DALYs annually. Results were stable across TMREL sensitivity analyses (aggregate all-cause attributable deaths range 2469-6590 thousand for TMREL bracketing 0%-15%; primary 3362 thousand at TMREL = 10%).

CONCLUSIONS

UPF consumption above a 10% energy threshold is associated with a sizable NCD burden that varies markedly across countries and is concentrated in high-income settings. The income gradient points to an opportunity for early preventive food policy in countries undergoing the nutrition transition.