A number of factors determine an individual’s likely life expectancy: Past medical history, smoking, gender, age, and most importantly, socio-economic status. The link between the latter and life expectancy is well documented with a broad range of data to support it.
A 2019 study by Drexel University, for example, found the average life expectancy for the lowest socio-economic groups in the US to be nine to 12 years below that of people the same age in the highest socio-economic cohort.
But according to Sacha Dhamani, chair of the Mortality and Morbidity Research Steering Committee at the UK’s Institute and Faculty of Actuaries, as people age the importance of how much social contact they have increases – whereas the impact of socio-economic variation decreases.
“Typically, socio–economic factors are very powerful between the ages of 60 to 75 in terms of predicting life expectancy, but social contact becomes increasingly important between the ages of 75 and 85; and after that mental wellbeing, social interaction, and engagement are much more important than is commonly recognised.”
Dhamani says that this is also true for older people who suffer from some form of physical impairment.
“At which point, how much purpose there is in their life is what matters most. And that will drive their level of social engagement, but if they’re physically restricted this becomes very hard.”
Rahul Nawander, an Ontario-based medical director at Fasano Associates, which operates in the life, health, and life settlements sectors, agrees with Dhamani on the negative impact of low levels of social contact on life expectancy. But he says there are also other factors at play in the same age group that is most impaired by a lack of human interaction.
“It’s true from a research perspective there is an increased risk of mortality linked to loneliness, and social isolation, particularly in relation to elderly lives. But is it magnified to an extent that it trumps the other risk factors? That might not be entirely plausible.”
According to Nawander, biological factors also become important at a similar age to when loneliness becomes a risk factor, with mild cognitive impairment typically starting around 75: the age at which the grey matter in a person’s brain typically starts shrinking.
“It’s a natural process of ageing. Some people still live much longer, such as into their 90s before experiencing cognitive decline, but roughly 16 out of 100 lives will experience this form of brain atrophy, at which point the level of support becomes very important.”
The critical issue is that while loneliness is a simple concept to explain it is a difficult one to measure, and Dhamani says a lack of publicly available data on the issue makes it hard for UK life insurers to incorporate it into their underwriting.
This may change. In the UK, people typically annuitise at around 65, however, as part of a general trend towards later retirement, Dhamani says the age at which people annuitise is expected to rise, meaning the industry will need to account for the impact of loneliness on life expectancy.
“While it’s incredibly difficult to measure loneliness it’s something the industry will have to deal with if the annuitisation age moves closer to 75 as we expect. In which case the current medical underwriting approach will become less relevant and actuaries will need to find a way to proxy the impact of loneliness on life expectancy,” says Dhamani.
Currently, Dhamani says the UK life insurance sector is not explicitly pricing for loneliness when underwriting annuities.
“Because of the lack of data and the current annuity observation point being at retirement, this issue is not a critical one for the industry, because everyone is taking the same approach and therefore there isn’t a selection risk between companies,” he says.
The approach is slightly different in the US life settlement industry. John Lynch, Director of Actuarial and Underwriting Services at Longevity Holdings, said that the firm accounted for loneliness in its Fasano Associates business, but he declined to elaborate citing its proprietary nature.
He also said that Longevity Holdings’ other product line, TwentyFirst Services, does not explicitly debit for loneliness in its book of business; it does, however, look at linked factors including mild or major depression, and issues such as anxiety.
While there is uncertainty over the exact impact of loneliness on life expectancy, Lynch says that there is clear evidence for the negative effect of a spouse dying.
According to a study of Medicare data, roughly 50% of men die within nine years after the passing of a partner, a number which falls to about 30% for women.
But even with the clear level of data on the impact of a partner’s death there are still limits to its use as a proxy.
“Without a doubt there’s a loneliness factor after a spouse dying, but it’s more of a flat extra type risk factor than a debit that would endure forever. It’s a higher risk factor for the next two or three years. After that mortality levels revert to normal,” said Lynch.
It is unclear exactly why widowed women have longer life expectancy than men, but Nawander says that research suggests that after the age of 65 women expend more energy on expanding their social networks, whereas men are more likely to turn to alcohol.
“That could perhaps be the reason women live longer,” Nawander says.
Even with the clear link between the death of a spouse and life expectancy, Dhamani says there are still issues using marriage as a proxy for loneliness because there are a number of factors involved which are difficult to isolate.
“You might be the grumpiest person in the world. But if you’ve got a wife and children who make sure you’re okay, you’re going to get a mortality benefit from that, even if you are not talking to anyone else.”
Other proxy data for loneliness has its own complexities. Facial recognition software could be a good indicator; how often someone smiles is in theory a good predictor of happiness, however, there are already high profile issues relating to racial bias in new technology, as well as the difficulty of relying on as-yet-unproven innovations.
According to the UK’s Campaign to End Loneliness, the number of people who said they experienced loneliness increased by 6% during the Covid pandemic, a trend which shows no sign of retreating. But Dhamani points to the success of some retirement communities in North America in dealing with this issue and says this shows there is hope for the future.
“If British people had a more positive attitude about care, and recognised it’s not just there at the end of their lives and instead create systems that build upon those seen in the US, it’s possible for people to live healthier and happier lives at the end of their age range. It could also mean that some of these people are able to stay in work – and benefit from the social engagement that brings for longer.”