Inequalities in behavioural risk factor prevalence between five post-war generational cohorts of working age, England

Background and objectivesLong-term trends in health risk factor prevalence, and in inequalities, are often summarised using age-standardised point prevalence to allow for age distribution changes over time and between groups. Policies to effectively promote health require the decomposition of social change into its age-period-cohort components. As a first step, we provide a descriptive age-cohort analysis to identify patterns of generational change in key behavioural risk factors between and within post-war cohorts of working age in England. DataCross-sectional Health Survey for England data for participants aged 25-60 years was pooled between 1994-2019 (n=153,172) to construct five decennial cohorts (1940s, 1950s, 1960s, 1970s, 1980s). Socioeconomic status was classified by neighbourhood deprivation quintiles using the Index of Multiple Deprivation. Five behavioural risk factors were analysed: cigarette smoking (current-and heavy-smoking); alcohol consumption (frequency and heavy drinking); obesity; meeting recommended levels of physical activity and fruit and vegetables consumption. MethodsSex-specific analyses were conducted. Log-binomial regression models quantified the magnitude and direction of change in (i) prevalence ratios (PRs) between cohorts adjusting for age and deprivation and (ii) deprivation-specific PRs within- and between-cohorts to examine changes in absolute and relative inequalities between generations. ResultsIn more recent cohorts, decreases in prevalence, independent of age and deprivation, were observed for current smoking, frequency of alcohol consumption and heavy drinking, resulting in decreasing absolute inequalities. However, obesity levels, particularly among females, reached their highest levels in the youngest 1980s cohort. Relative inequalities in current smoking (most-versus least-deprived quintiles) peaked in the 1950s cohort (Males: PR 2.79 (95% CI: 2.57-3.04); Females: PR 2.81; 95% CI: 2.60-3.05)), decreased in the 1960s cohort (M: PR 2.00 (95% CI: 1.70-2.34); F: PR 2.58 (95% CI: 2.40-2.77)), and remained stable thereafter. Inequalities in heavy smoking persisted over time among current smokers. Higher obesity levels in the most-versus least-deprived quintiles were generally persistent across all five cohorts, albeit with some suggestion of widening inequalities in the younger-versus older-cohorts in females (1940s cohort: PR 1.55 (95% CI: 1.40-1.72); 1960s cohort: PR 1.87 (95% CI: 1.73-2.01)). This pattern was also observed for mean body mass index (BMI). For heavy drinking, relative inequality remained stable. Relative inequalities in fruit and vegetable consumption were lower in more recent cohorts. Physical activity levels were similar across cohorts, with little evidence of inequalities. ConclusionOur analysis of generational change reveals credible signals of behavioural risk factor changes in levels and in inequalities over successive post-war cohorts of working age.