[Skip to content]

Sign up for our daily newsletter
The Actuary The magazine of the Institute & Faculty of Actuaries

Life expectancy in Scotland

uring the past century, life expectancy (LE) has increased by nearly 20 years in most developed countries. But are we living longer healthy lives or spending more time in poor health? The answer to this question may be found by studying healthy life expectancy (HLE), which measures years lived in good health. HLE estimates have been published for most European countries since the 1980s, but were not available for Scotland until Sutton et al (2004) published estimates at the national level, NHS board level, local authority level, and area deprivation level. These led to many open questions, such as why HLE in Scotland might be different from HLE in England and the rest of Europe, and why there are differences within Scotland. The Faculty of Actuaries’ project, with researchers Angus Macdonald, Matt Sutton, and Jennifer Straughn, aims to make these comparisons and explore some of these questions.

Sullivan’s method
The usual way to estimate HLE is by Sullivan’s method, which actuaries may recognise as Manchester Unity with 0% interest. The probabilities of surviving each year in the usual formula for LE are weighted by the proportion of people judged to be in good health at that age. For Great Britain, these proportions are generally based on two questions in the General Household Survey (GHS): one about self-assessed general health (SAH) and the other about a diagnosed limiting longstanding illness (LLI). HLE based on the former seems to be higher. Furthermore, while women can expect to live longer than men, they also spend a higher proportion of their total LE in poor health.
Sullivan’s method is very simple and hence popular, but it is crude; other frameworks such as multi-state models ought to lead to more objective estimates. However, the longitudinal data that better models would need are scarce, so Sullivan’s method continues to dominate.
Sutton revealed that on a national level, using the GHS, women and men born in 2000 could expect to live 79.5% and 80.4% respectively of their total LE in good health based on LLI, or 85.5% and 89.1% based on SAH. However, using the Scottish Household Survey (SHoS), HLE (LLI) falls sharply to 73.1% (females) and 74.4% (males), while HLE (SAH) does not change much. The SHoS contains more information about health than the GHS, but the discrepancy makes us wary of comparing HLE estimates from different sources. This is particularly true when comparing HLE in different countries since different surveys are normally used, and cultural differences may influence perceptions of good health. These factors will matter when comparing HLE in Scotland with HLE in England and western Europe.

Variation by geography
Of equal interest is the large variation of HLE within Scotland, for example at NHS board level. Using the SHoS, Sutton showed that HLE (LLI) for women ranged from 53.3 years in Lanarkshire to 65.0 years in Orkney, while HLE (SAH) ranged from 63.3 years in greater Glasgow to 71.6 years in Shetland and Orkney. For men, HLE (LLI) ranged from 49.9 years in Greater Glasgow to 61.3 years in Orkney, and HLE (SAH) from 60.3 years in greater Glasgow to 71.1 years in Shetland. Similar patterns were found at the local authority level. The reasons are not understood.
The NHS in Scotland has not been subjected to quite the same upheavals as has the NHS in England. In particular, it has for many years collected health data in a uniform way, allowing individuals’ life histories to be followed and linked to their survey responses. This growing volume of data, maintained by the Information Services Division of NHS Scotland, is a magnificent resource and is unique in the UK. We are using part of this data, linking hospital episodes and other events in 1981 to 2004 with Scottish Health Survey responses, to try to get a more objective view of the rather subjective HLE estimates based on Sullivan’s method. This will be the final part of the project and it represents an opportunity to explore what may be the causes of HLE in the Scottish population.