|| Checking for direct PDF access through Ovid
The rapid rise in the number of people using food banks since 2010 has raised concerns about whether food insecurity is rising in the UK. Investigations into this problem have focused on the characteristics of food bank users, precluding identification of risk factors for food insecurity in the wider population. Using data from the 2016 Food and You Survey (FYS), the objective of this study was to identify socio-demographic characteristics associated with food insecurity in the UK and how risk has changed since food insecurity was last measured in the 2003–2005 Low Income Diet and Nutrition Survey (LIDNS).Data from the FYS, a nationally representative survey of adults aged 16+across England, Wales, and Northern Ireland (n=3001) were accessed from the UK Data Service. Food insecurity was measured using the USDA Food Security Module and coded into marginal, moderate, and severe levels. The prevalence of each level was examined across age, gender, work status, disability, income quartile, education, marital status, and family characteristics. The odds of increasing severity of food insecurity were calculated using a generalised ordered logistic regression model. Data from FYS were merged with the 2003–2005 LIDNS. Using Coarsened Exact Matching, the probability of low-income households being food insecure was compared between survey years, and across household characteristics.In 2016, 7.74% (95% CI 6.55% to 9.12%) of adults were marginally food insecure, 10.2% (95% CI 8.83% to 11.8%) were moderately food insecure, and 2.72% (95% CI 2.07% to 3.58%) were severely food insecure. Unemployment and disability were associated with higher odds of any food insecurity and higher odds of increasing severity of food insecurity compared to households with work and without a disability, respectively (OR for unemployed versus employed: 4.17, 95% CI 1.36 to 12.8 for severe food insecurity versus all other levels; OR for disability versus none: 6.06, 95% CI 2.92 to 12.6 for severe food insecurity vs all other levels). After matching respondent characteristics in the FYS with respondents from the LIDNS, the probability of low-income adults being food insecure rose from 27% (95% CI 25% to 31%) in 2003–2005 to 45% (95% CI 42% to 50%) in 2016. The rise was significantly greater for adults with a disability compared to those without a disability.Household characteristics associated with food insecurity reflect groups most likely to be in deep poverty in the UK. Rising food bank use in the UK may reflect rising food insecurity among low-income adults over the past 14 years, particularly for people with a disability.