The most recent Continuous Mortality Investigation (CMI) model, CMI_2024, published in the middle of last year, indicated that life expectancy among younger, working-age people in England and Wales had declined, while that of retirement age people had increased.
Life expectancy among the over 65s rose by three months for men and by two weeks for women but shortened by about five months for men and women aged around 35.
The latest data on excess mortality in England, published by the Office for Health Improvement and Disparities (OHID), however, shows that there were significantly fewer deaths not only among the older cohorts, but in the working-age population as well.
The most recent excess death figures released by OHID show a significant decline through the end of 2025 for those of working age. December, for instance, registered almost 4,500 fewer deaths than had been forecast across all age groups and sexes. There were 144 fewer deaths among the 25–49 cohort and 467 fewer for the 50-64 group.
Still, the overall picture from the most recent reports is one of continued improvement in mortality, said Amy Walker, Actuary and Client Delivery Lead for the UK at longevity data provider Club Vita. In particular, she suggests that the effects of Covid-19 are over.
“2025 is the first year that we’re seeing fairly clear evidence that mortality is returning to pre-pandemic levels,” Walker said.
“That said, it’s still too early to say what mortality will do going forward, particularly for under 65s. Keeping an eye on trends in 2026 will be vitally important.”
Excess deaths are calculated by comparing the actual number of people who die in any period with the mortality forecasts. It became the key statistic used by the government to measure the impact of the Covid-19 pandemic.
At the time, it was welcomed, but some observers worry that it has lost its utility.
“People have become over-attached to it. It’s still very useful if you do want to think through what’s driving this year’s deaths and what might then drive next year’s deaths – if the same drivers persist,” said Matthew Edwards, a former CMI chair and senior director at WTW.
“Back in 2020 it felt like a godsend to see through the mist; at the moment it’s now almost a form of mist itself. It can be extra confusing or a distraction.”
Clouding the issue is the multiplicity of measurements for excess deaths, which all have different methodologies.
As well as the OHID gauge, the Office for National Statistics (ONS) also presents excess deaths data. Both projections are calculated by extrapolating data from the past five years but the OHID’s monthly readings don’t take into account population growth, which is adjusted for the high death rates during Covid-19 and covers only England.
Another gauge, published weekly by the UK Health Security Agency (UKHSA), is less granular, treating the working-age population – from 16 to 64 – as a single cohort.
Throughout 2025 and into this year, the OHID, ONS and UKHSA have shown a decline in mortality. The CMI model updates annually, typically around April (it was published in June last year due to substantial changes to the model).
“There is not a unique way to calculate expectations, and this can vary by outlook and methodology,” said Walker.
The contradictions between excess death data and the most recent CMI may be due to the underlying statistical calculations. Models that build projections from historical data are prone to error because there is no guarantee that what happened in the past will happen again in the future.
“It’s depressing for an actuary who specialised in longevity modelling to have to say… all of my models and my colleagues’ models and the industry models [have] all been wrong, but by definition you’re trying to project the future, so you’d be wrong half the time either way,” said Edwards.
Establishing an accurate predictor of mortality is hampered by changing risks to life. The main causes of deaths over the years have changed as treatments for illness and disease have improved and policies have helped reduce deaths by accidents.
A dramatic reduction in smoking through the 1980s and 1990s helped cut deaths from related diseases, for instance. And advances in treatments for cardiovascular illnesses have slashed death rates from heart disease.
In addition, new jeopardies are emerging; suicide and other so-called ‘deaths of despair’ such as drug overdoses have risen in recent years as have deaths among the socially isolated.
Consequently, actuaries are watching the evolution of innovations that could extend lives.
Trials for new drugs, especially for cancers, obesity and dementia, are high on their list. One highly anticipated trial is that of multi-cancer early detection (MCED) tests. Even so, Edwards doubts that treatments under trial will have much impact on longevity.
“These always sound very promising, but there’s very little rocking the overall longevity boat,” he said.
“Looking at recent phase-three clinical trials, they are just looking very, very small beer in terms of actually changing anything.”
How the latest data on excess deaths affects the next CMI depends on how final data readings play out, but there is some expectation it will show a slightly improved outlook.
“Certainly, when we look back at the CMI_2024 model and compare it with how 2025 actually played out, mortality improved more than what was projected,” said Walker.
“Given the volatility we’ve seen over the past few years, it’s understandable to be cautious… but we need to wait to see whether those improvements will be sustained.”