Is Dementia the Next Big Longevity Event?

For many countries with ageing populations, dementia presents major challenges: from the provision and costs of social care to the burden of appraising novel medical technologies targeting the dementia-care market. Insurers and reinsurers are not immune from such challenges; improving dementia outcomes could affect mortality rates in our elderly populations, affecting annuities and changing the balance of premiums and payments on protection products. 

In England, it is estimated that about 6.5% of persons aged 65 and over have dementia (4.3% with recorded diagnoses) at the start of 2026¹. Both incidence and prevalence increase sharply with age; in the UK, prevalence at ages 65-69 (75-79) is approximately one tenth (one third) of that for ages 85-89. A similar relative relationship is observed globally, with prevalence at ages 60-69 (70-79) approximately 12% (34%) of that for ages 80-89².   

For insurers writing business exposed to longevity risk, the increasing importance of dementia as a cause of death has resulted in an increasing focus on treatments for dementia as a potential source of future mortality improvements. This, combined with a vast research pipeline for novel therapeutics for the most prevalent types of dementia, provides ample reason for insurers to thoroughly investigate the potential impacts of changes in dementia-related mortality as part of their broader longevity risk framework. Indeed, in a WTW survey, novel dementia drugs were amongst the most commonly modelled scenarios for ‘new information risk’. 

Treating dementia 

The variety of diseases contributing to dementia prevalence (often co-occurring) and the poorly understood causes of disease progression present a challenge for researchers developing novel treatments. There remain significant questions over the causes of even the most common cause of dementia, Alzheimer’s disease, and the appropriate therapeutic targets for treatment; side-effects of treatments can also be significant. 

However, these challenges are not without their technological match. Researchers are taking advantage of our rapidly evolving machine learning capabilities to explore predictors of Alzheimer’s disease and identify potential therapeutic targets, as well as to streamline the identification of novel candidate treatments³. Artificial intelligence may also be set to streamline clinical trials for Alzheimer’s disease medications; it is already helping to pinpoint groups of patients who may be more likely than others to benefit from specific treatments⁴. 

Diagnosing dementia 

NHS England estimates that around a third of dementia cases in England remain undiagnosed⁵. Attempts to seek help are often delayed because of patients’ denial, the stigma attached to the condition, fear of loss of autonomy and the lack of any curative treatment. Barriers to diagnosis may also originate with the medical system (lacking time, resources and availability of support services), medical professionals (with difficulty in communicating a diagnosis or having little to offer patients post-diagnosis) and carers (who may question the value of a diagnosis in the absence of a cure and seek to preserve the patient’s independence or fear the impact of a diagnosis on their own lives)⁶.  

Screening for the main causes of dementia is not generally available; even in light of recent tentative steps forward in the treatment of Alzheimer’s disease, a review of the benefits and harms of population screening recently suggested that such a programme was not yet justified⁷. 

This situation could change quite rapidly as a result of progress in the treatment of key causes of dementia. The ability to significantly slow, halt or reverse disease progression would incentivise earlier diagnosis and remove or reduce key barriers to diagnosis (e.g. lack of treatment, stigma).  

Preventing dementia 

The biggest risk factor for dementia is age; the disease can take many years to develop, so individuals living longer gives more time for dementia to arise. 

Cerebrovascular disease is a strong risk factor for vascular dementia; a history of stroke doubles the risk of incident dementia in the older population⁸, and this increased risk is not explained by pre-stroke cognitive decline or demographic risk factors, nor by cardiovascular risk factors, which themselves are an additional risk factor. 

About one-third of those with mild cognitive impairment (MCI) progress to dementia within three years (with similar proportions either staying the same or spontaneously improving, the latter suggesting that their symptoms were caused by depression, anxiety or other similar conditions)⁹. 

Inherited risk (through genes) is a less material risk factor for dementia. Familial genes — which cause dementia with certainty if passed down from parent to child — are very rare for most types of dementia. 

One estimate suggests that around 30% of Alzheimer’s disease cases worldwide may be attributable to modifiable risk factors, such as physical inactivity, midlife obesity, smoking and depression⁶. A 2024 report from The Lancet estimated that 45% of all worldwide dementia cases could be prevented or delayed by addressing 14 modifiable risk factors at different stages of life⁷.  

It is clear, then, what sort of policies and behavioural changes might be required in order for dementia prevalence to fall (or rise at a reduced rate): those which discourage or disincentivise the relevant modifiable risk factors. However, policies to change behaviours, control population obesity or encourage increased physical activity have had mixed results.  

An improvement in diagnosis would, in isolation, have limited impact; unless the disease process can be slowed or halted, knowing that a patient has the disease is of limited value. Combined with an effective treatment (particularly targeting the early stages of disease), improved diagnosis could increase the impact in the medium-term. On the other hand, preventative measures (reducing exposure to modifiable risk factors) would have a much longer-term impact.  

Conclusion 

If improvements in dementia mortality are to be the “next big longevity event”, then these improvements should come primarily from novel treatments, supplemented by improved diagnosis (spending on which will be more justifiable once a viable treatment is available). 

There are still significant challenges facing those developing treatments: both identifying the most effective therapeutic targets and identifying and trialling safe and effective drugs to exploit those targets. Current trials (at least later-stage trials of drugs for Alzheimer’s disease⁸) include a number of drugs which are similar in nature (monoclonal antibodies) to the first generation of drugs, which demonstrated at best a very limited degree of disease-modification⁹. We might expect slight improvements in these drugs over time, but a breakthrough drug seems less likely. 

Even with advances due to uses of machine learning in both the study of diseases and drug development^3,4, a significant advance in treatment seems distant and access to such a treatment could be further delayed depending on national policies concerning the required level of cost-effectiveness of medications before they can be prescribed.  

Nevertheless, insurers and reinsurers will want to keep abreast of the outcomes of the range of current and future dementia drug trials; early indications of better-than-anticipated benefits from drugs undergoing phase 2 trials could be a useful signal that dementia-related mortality improvements might accelerate, giving time for a gradual adjustment of pricing and reserving assumptions.