Document resource
Introduction Well-known non-communicable disease (NCD) risk factors tend to cluster, forming groups with similar ‘risk factor phenotypes’. However, the characterisation of such clusters and the role of socioeconomic position (SEP) in modifying their associations with NCDs still remain unclear.Methods We utilised data on smoking, physical activity, alcohol consumption, body mass index, blood pressure, cholesterol and triglyceride from health examination surveys conducted in Norwegian adults from 1974 to 2019 (n=5 76 299; age 20–65 years; 64% participation; 52% women) linked to registry data on education, income and mortality (follow-up through 2020). We identified risk factor clusters using latent profile analyses and used flexible parametric survival models to estimate differences in mean age at NCD death across sex, different birth years and SEP (education and income levels).Results We identified four distinct clusters containing 44% (‘healthy’), 45% (‘suboptimal’), 4% (‘dyslipidaemics’) and 7% (‘hypertensives’) of the study sample, respectively. During a mean follow-up of 26 years, 61 241 died from an NCD.Compared with the cluster with the healthiest profile, other clusters had markedly shorter survival across sex and different birth cohorts. However, differences were also substantially modified by SEP: we estimated that individuals with primary education or low income died several years earlier than individuals with tertiary education or high income within each of the four risk factor clusters.Conclusions Our findings show how NCD risk factors cluster and how SEP modifies the association between this clustering and NCD mortality. The substantial influence of SEP highlights the need for public health strategies that focus not only on changing behavioural and metabolic risk factors but also on addressing socioeconomic disparities to effectively reduce the burden of NCDs.