An excellent example is provided by analysis carried out on "excess winter deaths" across Europe. Every country in the world displays seasonal patterns of mortality whereby more deaths occur in winter than at other times of year. However the extent of this excess varies between countries even within Europe. Intuitively one might have expected the excess to be greater in countries where winter temperatures are more extreme, yet this is not so. Healy (2003) used data from 14 European countries to demonstrate that in 1988-97, the relative Excess Winter Deaths Index (EWDI) was greatest for Portugal, where the mean winter temperature was highest. Conversely Finland with the lowest mean winter temperature showed the lowest EWDI. Data on mortality were available from the United Nations Statistics Databank and the World Bank, as well as some macro-economic indicators, but Healy also availed himself of the European Community Household Panel survey on socioeconomic indicators and housing conditions. This revealed that high EWDI was associated with lower expenditure on public health per head of population, as well as income poverty, inequality, deprivation, and fuel poverty. Furthermore, several indicators of residential thermal standards appeared to carry influence, whereby countries where houses had better insulation experienced lower EWDI.
A similar study was reported in 2014 by Fowler et al, partly as an update of Healy’s work, this time on 31 countries across Europe for the years 2002-11. The same geographic pattern still seemed to be present, with southern European countries faring worse in terms of winter deaths. However a few countries such as Greece, Spain and Ireland demonstrated a reduction in their EWDI. It is possible that Healy’s study had highlighted the need for improvement in those countries. All 27 countries who by that time were members of the European Union were included in analysis, and use was made of the Eurostat database.
In view of the projected increases in global temperature in coming decades, it might be hoped that the problem of excess deaths in winter will gradually disappear from Europe. Yet the greater susceptibility of warmer European countries to winter deaths compared with colder countries suggests such an assumption may be mistaken. It will be important for carefully collected routine data to be analysed, to investigate any changes in the patterns previously seen in Europe.
My colleagues and I were led to consider whether relatively low temperatures were more threatening to older people than absolute temperature level, and whether this might hold for individuals, as well as at a national level as highlighted by Healy’s and Fowler et al’s studies. We carried out analyses of two European cohort studies, of around 10,000 people aged 60 or over, followed over 10 years. Using daily temperature data for the localities of where these participants lived, we investigated weather patterns experienced by those who suffered major heart attacks and strokes. There was some evidence that cold spells (cold in relation to the month of the year) increased people’s risk over a 3-4 day period. We hope to replicate this finding in other datasets.
Reflecting on the data used by Healy and Fowler et al, it is noticeable that most (though not all) came from EU countries. Some of the data in Healy’s study was held by the United Nations or World Bank. Yet the Eurostat database was a major contributor to these enlightening analyses. Eurostat was established as long ago as 1953, initially to meet the requirements of the Coal and Steel Community. Over the years its task has broadened, and when accessed on 29 June 2016 displayed detailed comparative data on many domains including aspects of health.
It would be deeply disappointing as well as surprising if the UK were in future to withhold such valuable information, or conversely if such pan-European data were to become unavailable to UK-based researchers. This would seem unlikely, as Eurostat seems to draw upon data from EFTA nations as well as the EU, and advertises its data as freely available. It behoves the UK research community to continue to use these valuable data in a collaborative way with EU-based partners, and also to encourage continuing provision of UK data so that our EU-colleagues (both academics and policymakers) may benefit from this common enterprise.
This blog is by Professor Richard Morris, from the University of Bristol's School of Social and Community Medicine. Richard's research focuses around statistics applied to epidemiology, primary care and public health research.
Read other blogs in the Brexit series:
- Brexit, trust and the future of global environmental governance- The EU, Brexit and nature conservation law- Brexit: A climactic decision?- Why is populism popular? A psychologist explains