Article Text

Download PDFPDF
P139 Differentials in key health risk factors between five post-war cohorts of working age, England
  1. Madhavi Bajekal
  1. Applied Health Research, University College London, London, UK


Background Cohort life expectancies are projected to increase (ONS), based on extrapolations of long-run historical trends in rates of mortality improvement (RMI). However, the assumptions underlying this constant increase in RMI are rarely explicitly linked to evidence on improvements in known drivers of mortality such as lifestyle, ill-health and morbidity rates between cohorts. For policy-makers and annuity providers, it is important to know the likely health profile of today’s 45 year olds when they reach retirement age to better understand any limitations in the extrapolation of mortality trends.

Data We constructed a cohort series, pooling data from the nationally representative annual cross-sectional Health Survey for England 1994–2019 samples, for five decennial cohorts - those born in the 1940s, 1950s, 1960s, 1970s, and 1980s - aged between 25–60 years.

We first plotted age-by-cohort sex-specific estimates of behavioural risk factors (smoking, drinking, diet, physical activity); bio-markers (BMI, blood pressure, HbA1C, cholesterol); morbidity (limiting illness, prescribed medications); and general health. Using the 1960s cohort as the reference, we then calculated age-adjusted rate ratios (RR) to quantify the overall magnitude of the difference between cohorts (with 95% CIs), for all 11 indicators.

Results Age-adjusted rate ratios for successively younger cohorts were lower for all four risky behaviours. Eg, relative to the reference 1960s cohort, RR for smoking prevalence was 25% (1.26, 1.18- 1.33) higher and 25% (0.76, 0.72- 0.80) lower for the 1940s and 1980s cohorts, respectively. In contrast, RRs for younger cohorts were higher for obesity, waist-hip ratio, diabetes and numbers of prescribed medicines; but rates did not differ significantly across cohorts for subjective factors such as self-rated good health or reported limiting illness. Systolic blood pressure and cholesterol levels were lower for younger cohorts; however, results for most blood analytes were mixed partly due to volatility due to smaller sample sizes (reduced frequency of, and consent to, collection).

Conclusion This is ongoing analysis. Our first stage analysis points to younger cohorts having a better health profile than older cohorts. We are currently working to extend the analysis to estimate inequalities in individual and clustered risks by cohort.

  • cohort profiles
  • health risk factors
  • inequalities

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.