Mortality data in the UK has continued to show a slowdown in the trend towards longer life expectancy. Suspicions that the early signs were a blip have been dashed. We are still living longer than before, but the rate at which average lifespans are improving is slowing down. Attention is turning to what has caused this trend.
Quick recap: This post reported on heavier than expected mortality in England and Wales in 2015, and that after a period of gradual mortality improvements, there had been almost no improvement after 2011. It wasn’t possible to say what had caused this, or whether it was a short-term blip. However, it would be a very big deal indeed for future generations to start having shorter average lifespans than the current generation.
So, the message was to still plan on average lifespans continuing the trend of getting longer, even if not as fast as previously thought, and to keep a close eye on mortality trend data.
What’s new? More data has emerged which has confirmed that the slowdown in the rate of improvement of mortality since 2011 is a trend not a blip.
Specifically, the standardised mortality rate** for men aged 50-89 in England & Wales has shown the following average annual rates of improvement*:
| Years | Average annual improvement in standardised mortality rate, men ages 50-89, England & Wales |
| 1961-1975 | 0.7% |
| 1975-2000 | 1.8% |
| 2000-2011 | 3.1% |
| 2011-2016 | 0.7% |
Source: CMI calculations, based on ONS data, from Palin (2017)
Mortality improvement is still above zero, so it is still true to say that men in this age group in England & Wales are living longer over time. This has been the case for decades, indeed centuries. It’s also generally true for populations in nearly all countries.
However, a mortality improvement rate of 0.7 per cent a year shows up badly against the very high rates of improvement in the first decade of 21st century.
Why the slowdown? Mortality trends are a result of many complex sub-trends, including changes in the population base or changes in external causes. For example, there has been a super-healthy “golden cohort” born around 1930, of which there are, naturally, a diminishing number in the population we are considering. Their super-healthy mortality improvement has enhanced mortality improvements at older ages in the UK and likely in other countries too.
The main diseases have also changed over time, from cardiovascular in the early second half of the last century to dementia and respiratory now. These diseases mediate how environmental and age-related factors can affect the mortality rate.
The researchers referenced in this post make a case for “austerity” as a leading cause of the slowdown in mortality improvement. Austerity is used as a shorthand for lower supply than demand in health and social care. This could bite particularly on the current main causes of death, contrary to how medical science enabled the reduction in mortality from cardiovascular causes in previous decades.
Research will no doubt continue to attempt to isolate the key factors. Some, but not all, countries are seeing similar trends, which suggests a structural or environmental element rather than a biological one. It is also worth bearing in mind that mortality has a multitude of risk factors, including genes, early-life conditions, ongoing behavioural and environmental impacts as well as chance. There is unlikely to be a single neat explanation.
See Jon Palin (2017) Mortality improvements in decline, The Actuary, August 2017.
Also presentations at the Institute and Faculty of Actuaries’ Mortality and Longevity Seminar 2017, and Longevity Bulletin 10: Is the Tide Turning?
** A standardised mortality rate adjusts the raw data to allow for the changing age profile of the population within the age 50-89 range, so it measures only the trend in the average mortality rate of that age group. Mortality improvement is a good thing. It means the average probability of death in a group (the “mortality rate”) is lower than it was last year. A trend of mortality improvement means average lifespans are getting longer.