Systematic review of nutrient profile models for front-of-pack nutrition labelling.

Front-of-pack nutrition labelling (FOPNL) systems, based on underlying nutrient profile models (NPMs) that classify foods according to their nutritional composition, are key policies for tackling diet-related non-communicable diseases. Yet, the growing number of NPMs poses a challenge for policymakers to select appropriate models for FOPNL policies. This systematic review identifies and characterizes the NPMs developed and/or used for FOPNL systems, from both peer-reviewed and grey literature. We identified 42 NPMs, of which 15 were developed for warning labels, 11 for nutrient-specific systems (including traffic-light labelling), 11 for health endorsements, 4 for summary scores and 1 for other systems. All NPMs limited the content of sugar, and most limited sodium and saturated fat. Convergent and criterion validity were demonstrated for nine and three NPMs, respectively. Considerable heterogeneity across NPMs suggests limited consensus on what constitutes the most appropriate NPM for FOPNL, reinforcing the need for greater regional guidance and alignment.