Socioeconomic differences in purchases of more vs. less healthy foods and beverages: Analysis of over 25,000 British households in 2010

Rachel Pechey,Susan A Jebb,Michael P Kelly,Eva Almiron-Roig,Susana Conde,Ryota Nakamura,Ian Shemilt,Marc Suhrcke,Theresa M Marteau

doi:10.1016/j.socscimed.2013.05.012

Abstract

Socioeconomic inequalities in diet-related health outcomes are well-recognised, but are not fully explained by observational studies of consumption. We provide a novel analysis to identify purchasing patterns more precisely, based on data for take-home food and beverage purchases from 25,674 British households in 2010. To examine socioeconomic differences (measured by occupation), we conducted regression analyses on the proportion of energy purchased from (a) each of 43 food or beverage categories and (b) major nutrients. Results showed numerous small category-level socioeconomic differences. Aggregation of the categories showed lower SES groups generally purchased a greater proportion of energy from less healthy foods and beverages than those in higher SES groups (65% and 60%, respectively), while higher SES groups purchased a greater proportion of energy from healthier food and beverages (28% vs. 24%). At the nutrient-level, socioeconomic differences were less marked, although higher SES was associated with purchasing greater proportions of fibre, protein and total sugars, and smaller proportions of sodium. The observed pattern of purchasing across SES groups contributes to the explanation of observed health differences between groups and highlights targets for interventions to reduce health inequalities.

Highlights

Recent focus in the UK and elsewhere on the social determinants of health inequalities (Commission on Social Determinants of Health, 2008; Marmot, 2010) has raised the question of how social, economic and political environments influence health outcomes such as obesity and heart disease at the individual- and population-levels (Galea, Riddle & Kaplan, 2010; Kelly, 2010a)
Household purchasing differed by SES group for 28 of our 43 categories
Purchasing by lower SES groups was characterised by proportionally less energy from healthier food categories (4 percentage points difference between groups A&B and D&E) and proportionally more energy from less healthy food categories (5 percentage points difference between groups A&B and D&E)

Summary

Introduction

Recent focus in the UK and elsewhere on the social determinants of health inequalities (Commission on Social Determinants of Health, 2008; Marmot, 2010) has raised the question of how social, economic and political environments influence health outcomes such as obesity and heart disease at the individual- and population-levels (Galea, Riddle & Kaplan, 2010; Kelly, 2010a). While it is often argued that there are social class-based patterns in dietary behaviours, beyond social disparities in the consumption of fruits and vegetables (De Irala-Estévez et al, 2000; Diez-Roux et al, 1999; Galobardes, Morabia, & Bernstein, 2001; Giskes, Avendano, Brug, & Kunst, 2010), relatively little evidence has accumulated to support this claim from representative surveys of food consumption Such studies are vulnerable to bias due to misreporting and measurement error (Carriquiry, 2003; Poslusna, Ruprich, de Vries, Jakubikova & van’t Veer, 2009; Rennie, Coward & Jebb, 2007). Are any observed purchasing patterns consistent with the extant empirical evidence for differences in health outcomes by SES?

Methods

Results

Discussion