Public Health shock: Scots match English for healthy life expectancy

What would it take to narrow the gap in life expectancy between Scotland and England? A revolution in Scottish eating habits, most experts might say.

But the European Union has found a simpler way. Its method of calculating healthy life expectancy at birth shows that Scotland’s women already boast the highest figures in the UK – 63.1 years, against England’s 62.9.
Scotland’s men still lag a little behind England’s (61.2 vs 61.6 years) but those that reach the age of 65 can expect as many healthy years ahead as those in England - 9.9 years. Scottish women at 65 can expect another 11.3 years of healthy life, comfortably the highest figure in the UK.
Yet at the same time, traditionally-measured healthy life expectancy at birth remains far higher in England than in Scotland: 77.5 vs 74.8 years for men, 81.7 vs 79.7 years for women. What’s going on?
It certainly isn’t any change in diet. Plenty of studies have shown that the Scots do not invairiably eat healthily. In 2004, Glasgow-based health service researchers David S Morrison and Mark Petticrew reported in The Lancet that the deep-fried Mars bar was no urban myth, but a popular and readily available fast-food in that city. They also discovered another delicacy, Scotland's reluctant acknowledgement of the Mediterranean Diet - deep-fried pizza.
In 2008, an even more alarming Glaswegian take-away known as the “Munchy Box” was revealed in Scotland on Sunday. On sale for £4-5, the boxes were reported to include large helpings of pizza, doner meat, chips and pakora, amounting to 87 per cent of the recommended daily adult calorie intake, 123 per cent of the recommended daily adult intake of saturated fat and more than double the recommended daily intake of salt.
British Heart Foundation statistics further suggest that Scotland has the worst record for deaths from coronary heart disease in the UK, with Glasgow second only to Birmingham for the highest number of deaths from CHD in 2004-2006 (see Table 1.7).
So how is it that the EU method for measuring healthy life expectancy shows Scotland in such an encouraging light? The figures were released on 24 February by the Office of National Statistics, and cover healthy life expectancy (HLE) at birth and at 65 for the UK as a whole and its constituent countries for 2005-07.   
The tables include estimates for HLE calculated conventionally and using the EU’s Statistics on Income and Living Conditions questionnaire. The immediate and obvious effect is a downward jolt of about 10 per cent for healthy life estimates across the UK as a whole. This one-off adjustment was anticipated. But the new methodology, which is set to become the national standard, has also resulted in some curious and unexpected changes to the comparisons between the countries of the UK.
Under the old HLE measure, English women averaged about six months to a year more healthy life than Scottish women. Under the new system, Scottish women have overtaken their English counterparts. Scottish men had been lagging about 2 years behind English men, but under the new regime the gap has narrowed to about five months.
The explanation can be found (by the persistent) in a 2009 paper referenced as a footnote in the ONS bulletin, nesting within layers of web links: “Review of sources and methods to monitor healthy life expectancy' published in Health Statistics Quarterly 26”.
It all turns on the meaning of the word "fair". 
With both the old and the new question, answers given by respondents in surveys are ultimately grouped (“dichotomized”) into two categories: “Good” and “Not Good” health. But as the Box below shows, with the new EU method the answer “Fairly Good” disappears. Of the five rather than three potential answers, only “Very good” and “Good” now go into the “Good” health category. The three remaining answers, including “fair”, count as “Not Good”.
The authors succinctly summarise the difference: “[the EU definition] is conceptually less inclusive than the current definition.”
So there you have it: those who might have described their health as “fairly good” must now choose between “good” and “fair”. If they choose the latter, they are put into the “not good” health category. And when the Good/Not Good data is fed into the complex HLE calculation, known as the Sullivan method, Scottish HLEs improve.
Does this mean the old healthy life estimates were inaccurate?
“Not at all,” says Dr Mike Smith, co-author of the ONS paper. “The new EU-HLE measure has no bearing on the original measure - they should be considered as separate entities.
“It is, however, worth noting that there may be slight issues relating to question order and exposure (the original and EU general health questions were asked in parallel in the General Household Survey between 2005 and 2007), although our extensive modelling suggested that these had no significant impact upon the way in which people reported their health.”
A fair answer, if you ask me.