Income and individual deprivation as predictors of health over time

Poverty, often defined as a lack of resources to achieve a living standard that is deemed acceptable by society, may be assessed using level of income or a measure of individual deprivation. However, the relationship between low income and deprivation is complex-for example, not everyone who has low income is deprived (and vice versa). In addition, longitudinal studies show only a small relationship between short-term changes in income and health but an alternative measure of poverty, such as deprivation, may have a stronger association with health over time. We aim to compare low income and individual deprivation as predictors of self-rated health (SRH), using longitudinal survey data, to test the hypothesis that different measures of poverty may have different associations with health. We used three waves from the longitudinal Survey of Family, Income and Employment and fixed-effect linear regression models to compare low income (< 50% median income at each wave) and deprivation (reporting three or more items from the New Zealand individual deprivation index) as predictors of SRH (coded 1-5; SD 1.1-1.2). We also compared the impact of duration of low income and deprivation on SRH using mixed linear models. In the fixed-effect models, moving into deprivation between waves was associated with a larger decline in SRH compared to moving into low income, which persisted in models including both low income and deprivation. Similar findings were observed for duration of low income and deprivation in mixed models. Moving into high levels of individual deprivation is a stronger predictor of changes in SRH than moving into low income. When investigating the association of hardship poverty with health, using alternative measures, in addition to income, is advisable.