Is the association between self-rated health and underlying biomarker levels modified by age, gender, and household income? Evidence from Understanding Society – the UK Household Longitudinal Study

The goal of this study was to evaluate how self-rated health (SRH) and objective measures of health (biomarkers) are associated, and if this association varies by gender, age, and socioeconomic position (measured by household income). Data come from the UK Household Longitudinal Study nurse visit (2010–2012), including a representative sample of adults in Great Britain (N = 15 687 maximum sample). SRH was assessed by the question “In general, would you say your health is excellent, very good, good, fair, or poor?” and dichotomized into good or poor. Indices were created for four biomarker categories based on the aspects of health they are likely to reflect, including visible weigh-related, fitness, fatigue, and disease risk biomarkers. Logistic regression models were run with SRH as the outcome and each biomarker index as a predictor, adjusting by gender, age, and income. Further, interaction terms between each biomarker index and gender, age, and income (independently) were added to test for effect modification. All biomarker indices were associated with SRH in expected directions, with the fitness index most strongly predicting SRH. Gender, age, or income modified the associations between SRH and all biomarker indices to different extents. The association between the visible weight-related biomarker index (including body mass/fat variables) and SRH was stronger for women than men and for those in higher income groups than lower income groups. Income also modified the association between SRH and the fitness biomarker index, whereas age modified the association between SRH and the fatigue biomarker index. When using SRH to investigate health inequalities, researchers and policy makers should be clear that different social groups may systematically consider different dimensions of health when reporting their SRH.