Shorter lives for UK women

British women’s life expectancy “ranks alongside some of the poorest countries in Europe” reported the Daily Telegraph yesterday, suggesting that on average, UK females do not live as long as their counterparts from 25 EU countries.

Their claims were based on research that was actually investigating how long 50-year-olds in each country will live without being affected by disability. In that respect, researchers found that both males and females from the UK live free from disability for significantly longer than the average EU citizen. The study also shows that the UK fared significantly better than Estonia in terms of both life expectancy and years in good health, in contrast to recent reports in other newspapers that Estonian healthcare was rated higher than that in the UK.

As the researchers acknowledge, the study itself has some limitations, meaning its results for whole populations may not be accurately applied to individuals. This study should only be taken as preliminary research into the factors that could be linked to healthy ageing.

Where did the story come from?

This study was conducted by Dr Carol Jagger and colleagues from the University of Leicester, INED in Paris, the Institute of Public Health in Belgium, University Medical Centre Rotterdam in the Netherlands and the French Institute of Health and Medical Research in Montpellier, France. The work was funded by the EU Public Health Programme and published in the peer-reviewed medical journal, the Lancet .

What kind of scientific study was this?

This was a large ecological study to explore differences in life expectancy and health during aging for the people in 25 European countries. The researchers wanted to look at a measure of quality of life as a means of estimating the health of a nation, as opposed to simply relying on life expectancy.

To do this researchers used a measure known as ‘healthy life years’ (or HLY), which is the number of further years that a person of a certain age will live free from ‘disability’, as defined by the researchers. For this study researchers looked at life expectancy and HLY from age 50.

Data came from a general population survey, called the Statistics of Income and Living Conditions (SILC) survey, which was initiated by the EU and adopted by the European countries as a way to collect information of this nature. The original purpose of the SILC survey was to investigate possible reasons for HLY differences between countries.

The researchers used disability data from the SILC survey from each country in 2005 to construct an index of ‘healthy life years’. In these surveys, disability had been defined as a long-term (greater than six months) limitation in activity and rated in severity as ‘none’, ‘limited but not severely’ and ‘severely limited’ health. People reporting disability of any severity were not counted in the HLY tally.

They also collected data on life expectancy, GDP, poverty risk for people aged older than 65 years, inequality of income distribution, expenditure on elderly care, unemployment rate, employment rate, age of exit from labour force and level of education. Most of this data had been collected in the respective countries in 2005.

What were the results of the study?

The researchers calculated the average life expectancy for 50-year-old males and females across all the countries as of 2005. This was 28.6 years for men and 33.5 years for women, though there was a lot of variability between countries. Life expectancy beyond 50 years was above average for UK males at 29.46 years, while for UK women it was slightly below average compared to the rest of Europe, at 32.69 years.

The researchers also calculated the average number of healthy life years 50-year-olds could expect to live across all countries. Men could expect to live 17.3 disability-free years and women could expect to live 18.1 disability-free years. The figures for the UK were significantly greater than the across-Europe average, namely 19.74 years for men and 20.78 years for women. The 10 newly joined EU countries performed worse than the established 15 countries.

Other factors that were associated with the differences in healthy life year values for men and women included GDPs and expenditure on elderly care. For men only, long-term unemployment rate, life-long learning and low educational attainment were associated with HLY values (positively or negatively). When the researchers repeated their investigation into the potential associated factors within just the 15 established EU countries, they found that none of the factors they included were associated with HLY values.

What interpretations did the researchers draw from these results?

The researchers conclude that their study has shown a large variation in ‘remaining years spent free of activity limitations’ in men and women aged 50 years across countries in Europe in 2005.
They say that given that a major target for Europe is that the employment rate for older adults (aged 55 to 64 years) should reach 55% by 2010, HLYs (as an indicator of disability) could be used to assess whether such targets are realistic.

What does the NHS Knowledge Service make of this study?

Ecological studies such as this suffer from some weaknesses that should impact on how these results are interpreted.

Firstly, given that the data used in this study is cross-sectional, there is no way to explore the ‘temporal’ link (i.e. time) between associated factors, e.g. unemployment rate, care of the elderly, education and the outcome. It is not possible to know whether these factors are a ‘cause’ of poorer healthy life years.

Secondly, the researchers relied on population-level data to explore these factors, rather than on data from individuals. As it is impossible to extrapolate findings from populations back to individuals, the study cannot prove that the people with lower healthy life years were the same ones with poorer education, healthcare etc.

To make an assumption like this (that what is happening at a population level is also happening at an individual level) is known as the ‘ecological fallacy’, a common weakness of these types of studies. The researchers acknowledge both these and other problems with their research method.

The measurement of HLY values, although more harmonised now following the adoption of the SILC surveys, is still not perfect. Each country would have carried out their surveys slightly differently and bias may be introduced because of this.

People who live in institutions were not included in the SILC surveys and the assumption was made in this study that their health was the same as that of people who were not institutionalised. This is unlikely to be an accurate assumption and a different health profile of this group may have biased the results, though the researchers say that this is unlikely to affect their conclusions.

The researchers themselves say that more data is needed (preferably from individuals) to confirm the associations seen in this study.