For those actuaries in the UK’s longevity and mortality modelling industry, 2025 has provided plenty of activity to keep them busy. Greg Winterton spoke to James Hadley, Senior Consultant at Barnett Waddingham, to learn more about what is front and centre of the conversation in his world as the year draws to a close.
GW: James, let’s start with the obvious. What has been the big news story in mortality modelling circles in the UK this year, and why?
JH: One of the big developments this year has been the eagerly anticipated release of CMI_2024 by the CMI, which feeds into the projections of mortality used in valuations of long-term liabilities.
The Model includes two important new features. The first is the fitted overlay. Recent versions of the CMI Model have reflected the extraordinary mortality during the pandemic by giving reduced weight or no weight to data for years since 2020. CMI_2024 instead reflects the pandemic through an “overlay” term, which starts in 2020 and decays exponentially, halving in each successive year.
The second is the addition of multiple period terms. Previous versions of the Model have included one period term, which helped determine mortality trends over time at all ages. CMI_2024 includes five period terms, allowing it to better reflect trends at different ages, namely worsening mortality at working ages contrasted with improvements at older ages.
At the same time, GLP-1 agonists have continued to pique the interest of practitioners, with prescriptions soaring and their likely effects on mortality trends starting to be better understood.
GW: There has been significant coverage of GLP-1 agonists this year, just like in 2024. They seem to still be the talk of the town. How is the mortality industry currently factoring in the effects of these drugs when the general understanding of their medium to long-term impact is embryonic?
JH: A major development this year has been the licensing of some of these medicines for weight loss treatment on the NHS, with further programme expansions expected over the next few years. However, most prescriptions are still accessed privately, disproportionately benefiting individuals in higher socioeconomic groups, who are also more likely to hold life insurance products.
Swiss Re recently projected a 3.2% fall in all-cause mortality in the UK over the next 20 years due to these medicines. Insurers are beginning to factor this into their longer-term mortality modelling, particularly by considering the socioeconomic composition of their portfolios and their current claims experience. This may involve adjusting assumptions and exploring scenarios to reflect different accessibility and health outcomes.
GW: When you look at cause-of-death trends (e.g. cardiovascular, cancer, neurodegenerative diseases), what surprises you most right now, and what implications might this have for longevity expectations?
JH: While major causes like circulatory diseases and cancer continue to dominate mortality statistics, what’s most surprising is the sharp rise in less prevalent causes, particularly mental and behavioural disorders, and endocrine, nutritional and metabolic diseases. For some working-age groups, mortality rates from these causes have increased by over 50% between 2013 and 2023, which is a considerable and unexpected shift.
The Institute and Faculty of Actuaries’ Mortality Research Steering Committee published a “drivers of mortality” dataset last year. This shows a marked increase in comorbidities, likely driven in part by these emerging causes, although also affected by changes in reporting practices. The growing complexity of causes of death profiles suggests that longevity expectations may need to be revised for certain groups. Insurers will need to closely monitor these trends and adapt their modelling to reflect the evolving risks.
GW: From an insurer or pension scheme perspective, where do you see the biggest modelling blind spots — areas where risk may be under- or over-estimated?
JH: I think one of the biggest blind spots in current mortality modelling is the limited consideration of economic factors. The Global Financial Crisis and the ensuing recession, Brexit, and the COVID-19 pandemic have all impacted public finances and, in turn, the ability to invest in health and social care. We have seen sustained impacts on population mortality, and I expect there could be longer-term consequences as well.
Between 2000 and 2011, five-year average mortality improvements were above 2% a year. Since then, we entered a “new normal” of much slower improvements of around 0.5% a year for the years ending 2015-2019. Emerging post-pandemic analysis suggests this trend is persisting, and improvements are worse for working ages and those in lower socioeconomic groups. If models fail to account for these diverging trends, they risk underestimating liabilities, particularly for portfolios with greater exposure to younger or less affluent populations.
GW: As we approach the end of 2025, what is top of mind in your world as we look to 2026?
JH: As we approach 2026, one of the key areas we’re watching is the evolving pattern of circulatory disease mortality. In my article, I noted a step-change in mortality for some working-age groups between 2019 and 2020, which persisted through 2023. Interestingly, recent all-ages Public Health England data shows a return to 2019 levels in 2024.
What’s top of mind now is understanding whether this feature is seen for both working-age individuals and pensioners. We’re also monitoring how other causes of death, particularly those linked to socioeconomic status, continue to shift. These trends could have large implications for longevity modelling, especially in portfolios with exposure to more diverse demographic profiles.
James Hadley is a Senior Consultant at Barnett Waddingham
