The Financial Times reports today the “good news” that as well as living longer we are staying healthier longer, sourcing the data to yesterday’s publication of the ONS publication Pension Trends.

The figures it cites are these: in 1981, 65 year-old men had 13 years of life expectancy ahead of them, which rose to 17.6 years by 2008. And while in 1981 they remained healthy for 9.9 of those years, by 2006 that had risen to 12.8 years.

For women, the corresponding figures for 65 year-olds in 1981 were 16.9 years of life expectancy, 11.9 of them in good health, rising by 2006 to 19.9 and 14.5 years respectively.

These aren’t new data, since they appeared in Pension Trends as long ago as June 2010. (The link on the ONS release calendar says this edition was published on 18 December 2008, but the actual document bears the date 25 June 2010 and includes 2009 data, so that’s when it must have been published. We’ll put the discrepancy down to another glitch on the ONS website.) Nor are the data featured by the FT actually the headline figures quoted in the latest Pension Trends, which says that a 65 year old man in 2008 has 9.9 years of healthy life ahead of him, and a 65 year-old woman 11.5 years – much lower than the FT figures.

What’s going on? As is so often the case, the rules have changed, making the time series a tricky source to interpret. In 2006 the method for calculating healthy life expectancy changed, to harmonise with EU regulations. Instead of asking people about their health on a three-point scale and counting as healthy those who said their health was “good” or “fairly good”, since 2006 a five-point scale has been used, and only those in the top two categories, “good” or “very good”, have been classified as healthy. The effect has been to raise the bar, so fewer people now qualify as being in good health.

The FT has restricted its comparison to the old basis of calculation, which is why it ends in 2006. That’s entirely appropriate, but it is not, as the paper claims, “new data”.

Nor is it necessarily especially good news. What the data mean is that (on the old calculation) men will spend 72.7 per cent of their retirement years in good health, as opposed to 76.2 per cent in 1981. They will have 4.8 years of health that isn’t good, against 3.1 years for those who were 65 in 1981. So for men the proportion of retirement years damaged by poor health has actually increased, along with the number of those years.

For women, the picture is different. For them, at least the proportion of retirement years in good health has increased, from 70.4 per cent to 72.9 per cent, even though the years spent in poor health has also increased, from 4.0 to 5.4 years.

If we look at the new method of calculation, no comparison with 1981 is possible, though ONS has “backtracked” the data to 2001 to try to provide some sort of time series. This suggests (Figure 3.3) that healthy life expectancy for men is rising more slowly than actual life expectancy, and for women at about the same rate. But the time series is too short to draw any real conclusions.

However the raw figures are much starker than the ones in the FT. In 2008, it says, men of 65 had 9.9 years of healthy life expectancy, and women of the same age 11.5 years. That is to be set against actual life expectancy of 17.6 years for men and 20.2 years for women.   

So, using the new tougher definition of good health, men face 7.7 years of poor health, and women 8.7 years. It all comes down to how you define good health.

More important is whether life expectancy is rising faster than healthy life expectancy, however defined, because that is a crucial factor in estimating future burdens for health and social care. The change in the rules, however well-intentioned or unavoidable as it might have been, makes it a little bit harder to work that out.