Who consumes ultra-processed food? A systematic review of sociodemographic determinants of ultra-processed food consumption from nationally representative samples

Samuel J Dicken; Sulmaaz Qamar; Rachel L Batterham

doi:10.1017/S0954422423000240

Who consumes ultra-processed food? A systematic review of sociodemographic determinants of ultra-processed food consumption from nationally representative samples

Nutr Res Rev. 2023 Oct 31:1-41. doi: 10.1017/S0954422423000240. Online ahead of print.

Authors

Samuel J Dicken¹, Sulmaaz Qamar^{1

2

3}, Rachel L Batterham^{1

2

3}

Affiliations

¹ Centre for Obesity Research, Department of Medicine, University College London (UCL), London WC1E 6JF, UK.
² Bariatric Centre for Weight Management and Metabolic Surgery, University College London Hospital (UCLH), London NW1 2BU, UK.
³ National Institute for Health Research, Biomedical Research Centre, University College London Hospital (UCLH), London W1T 7DN, UK.

PMID: 37905428
DOI: 10.1017/S0954422423000240

Abstract

Ultra-processed food (UPF) intake is associated with increased non-communicable disease risks. However, systematic reports on sociodemographic predictors of UPF intake are lacking. This review aimed to understand UPF consumption based on sociodemographic factors, using nationally representative cohorts. The systematic review was pre-registered (PROSPERO:CRD42022360199), following PRISMA guidelines. PubMed/MEDLINE searches (‘ultra-processed/ultraprocessed’ and ‘ultra-processing/ultraprocessing’) until 7 September 2022 retrieved 1131 results. Inclusion criteria included: observational, nationally representative adult samples, in English, in peer-reviewed journals, assessing the association between sociodemographics and individual-level UPF intake defined by the NOVA classification. Exclusion criteria included: not nationally representative, no assessment of sociodemographics and individual-level UPF intake defined by NOVA. Risk of bias was assessed using the Newcastle–Ottawa Scale (NOS). Fifty-five papers were included, spanning thirty-two countries. All thirteen sociodemographic variables identified were significantly associated with UPF intake in one or more studies. Significant differences in UPF intake were seen across age, race/ethnicity, rural/urbanisation, food insecurity, income and region, with up to 10–20% differences in UPF intake (% total energy). Higher UPF intakes were associated with younger age, urbanisation and being unmarried, single, separated or divorced. Education, income and socioeconomic status showed varying associations, depending on country. Multivariate analyses indicated that associations were independent of other sociodemographics. Household status and gender were generally not associated with UPF intake. NOS averaged 5·7/10. Several characteristics are independently associated with high UPF intake, indicating large sociodemographic variation in non-communicable disease risk. These findings highlight significant public health inequalities associated with UPF intake, and the urgent need for policy action to minimise social injustice-related health inequalities.

Keywords: diet; health inequality; sociodemographic determinants; ultra-processed.

Publication types

Review