Mapping intersectional inequalities in biomarkers of healthy ageing and chronic disease in older English adults

Daniel Holman,Andrew Bell,Sarah Salway

doi:10.1038/s41598-020-69934-8

Abstract

Chronic diseases and their inequalities amongst older adults are a significant public health challenge. Prevention and treatment of chronic diseases will benefit from insight into which population groups show greatest risk. Biomarkers are indicators of the biological mechanisms underlying health and disease. We analysed disparities in a common set of biomarkers at the population level using English national data (n = 16,437). Blood-based biomarkers were HbA1c, total cholesterol and C-reactive protein. Non-blood biomarkers were systolic blood pressure, resting heart rate and body mass index. We employed an intersectionality perspective which is concerned with how socioeconomic, gender and ethnic disparities combine to lead to varied health outcomes. We find granular intersectional disparities, which vary by biomarker, with total cholesterol and HbA1c showing the greatest intersectional variation. These disparities were additive rather than multiplicative. Each intersectional subgroup has its own profile of biomarkers. Whilst the majority of variation in biomarkers is at the individual rather than intersectional level (i.e. intersections exhibit high heterogeneity), the average differences are potentially associated with important clinical outcomes. An intersectional perspective helps to shed light on how socio-demographic factors combine to result in differential risk for disease or potential for healthy ageing.

Highlights

Chronic diseases and their inequalities amongst older adults are a significant public health challenge
We focus on a set of six biomarkers which were available in both surveys and are commonly accepted as biomarkers of healthy and functional ageing and predictors of chronic disease morbidity and mortality[16]
Given we find no evidence for multiplicative effects, the MAIHDA models offer no advantages over conventional models for estimating mean differences in outcomes between intersectional g roups[36], so we proceed with conventional linear regression analysis and use marginal effects to predict intersectional disparities, and we use MAIHDA to ascertain the discriminatory accuracy of intersectional clustering

Summary

Introduction

Chronic diseases and their inequalities amongst older adults are a significant public health challenge. Prevention and treatment of chronic diseases, and attainment of the policy goal of healthy ageing, will benefit from insight into which population groups show greatest risk. Many of the current gaps in knowledge on healthy ageing and chronic disease inequalities centre around the complex interaction of sociocultural, political and biological factors[2]. These inequalities have often been investigated according to single or at most a limited number of categories of difference at a time, such as gender, ethnicity or socioeconomic position (SEP). Different intersections defined by combinations of social attributes such as gender, ethnicity and SEP, are potentially associated with different (though overlapping) health outcomes and are subject to different causal processes driving these outcomes. By bringing an intersectionality lens to the analysis of biomarker data, we offer a novel approach to filling this gap in knowledge

Objectives

Methods

Results

Discussion

Conclusion