A Healthy Old Age?

Author: Andrew McCulloch

In my last article we had a look at how you calculate the length of time that you might be expected to have left to live given your current age, or life expectancy, using age-specific mortality rates. Life expectancy has significant appeal as a measure of population health because it summarises the age-specific mortality rates in a population in a way that is meaningful to most people. For example, in England and Wales in 2008 life-expectancy at age 65 for a man in a professional job was 19 years but for a man in a routine job it was only 15 years. It is not necessary for anyone to understand statistics in depth for them to grasp that men with different types of jobs can expect to live for different lengths of time.

Figure 1: Social class and life expectancy in England and Wales at age 65 from 1982/86 to 2002/06. Source: Pension Trends 2012. Chapter 3: Life expectancy and healthy ageing.

While life expectancy is a useful concept the use of mortality rates as the basis for its calculation is a potential weakness. Health consists of more than not being dead and mortality rates do not give a complete description of the health of a population. In particular, in countries where life expectancy is high most people will die from degenerative diseases and rising life expectancy has to be weighed against the length of time in old age that people spend in poor health. Whether rising life-expectancy is accompanied by a delay in the onset of disability and dependence in old-age is also significant for many policy issues, including the changing demand for health care and whether, and by how many years, the state pension age should be increased.

Because of the importance of distinguishing between different levels of health in old age a range of measures have been developed which take into account levels of health in addition to mortality. The simplest such measure is healthy life expectancy which divides the expected remaining length of life at any age into years lived in good health and years lived in poor health. The optimistic scenario for the evolution of population health is that rising life expectancy will be accompanied by a fall in rates of old-age disability (the compression of morbidity thesis).The pessimistic scenario, on the other hand, is that rising life expectancy will be accompanied by an increase in rates of old-age disability (the expansion of morbidity thesis), as illnesses that would have been fatal in the past no longer result in death but rather in rising levels of impairment and dependence in old age. Healthy life expectancy can be used to try to distinguish between these different scenarios. For example, if the compression of morbidity thesis is correct we might expect that health expectancy would increase at a faster rate than life expectancy.

The simplest method used to adjust life expectancy to take account of different health states was developed in the late 1960s and early 1970s by Sullivan. The aptly named Sullivan method follows the same methodology as the calculation of life-expectancy but simply multiplies the person-years that we would expect a cohort to live through over an age interval by the prevalence of ill-health or disability in the population at that age. For example, if we expected a cohort to live for a total of 100,000 person-years between ages 80 and 85 years and the prevalence of good health at these ages was 80 percent, the cohort would only contribute 80,000 person-years, rather than the full 100,000 person years lived, to the calculation of healthy life expectancy.

The prevalence of good health at different ages used in the calculation of healthy life expectancy is usually taken from cross-sectional survey data. For example, calculation of health life expectancy in the UK has used estimates of the proportion of the population at different ages who are in either good or fairly good health derived from responses to the following question in the General Household Survey: Over the last 12 months would you say your health has on the whole been good, fairly good, or not good?

Figure 2: life expectancy and healthy life expectancy for males and females 2001-2008. Source: Pension Trends 2012. Chapter 3: Life expectancy and healthy ageing.

The figure above shows how life expectancy and healthy life expectancy have changed in Great Britain over the period from 2001 to 2008. As we have seen life expectancy has been rising over the last 10 years, however, at least over this period, healthy life expectancy appears to have been rising at a slower rate than life expectancy. The gap between life expectancy and healthy life expectancy (or the length of time lived in less than good health) was 8.2 years for women and 6.5 years for men in 2001, while in 2008 it was 8.7 years for women and 7.7 years for men. This suggests that over the last ten years there has been a modest rise in the length of time that is lived in poor health in old age, particularly for men.