Socio-economic deprivation and excess winter mortality and emergency hospital admissions in the South Yorkshire Coalfields Health Action Zone, UK

Abstract

The aims of this study were to describe the pattern of excess winter mortality and emergency hospital admissions in the South Yorkshire Coalfields Health Action Zone, and to examine the relationship between excess winter mortality and emergency hospital admissions and socio-economic deprivation at the enumeration district level. We analysed monthly deaths from 1981 to 1999 and monthly emergency hospital admissions from 1990 to 1999 for cardiovascular disease, respiratory disease and all other causes of death for people aged 45 years and above. We used the enumeration district level Townsend socio-economic deprivation score to categorize enumeration districts by quintile. Excess winter mortality ratios (observed/expected) for females and males, respectively, were 1.70 and 1.58 for respiratory disease, 1.25 and 1.20 for cardiovascular disease, and 1.09 and 1.07 for all other causes of death. The excess winter hospital admission ratio for respiratory disease was 1.80 for females and 1.58 for males. No excess was evident for the other two groups of conditions. We found no significant increase in excess winter mortality ratios with increasing socio-economic deprivation. There was also no significant increase in the excess winter respiratory admission ratio with increasing deprivation. With regard to age, we found significant increases with increasing age in the excess winter mortality ratios for cardiovascular disease (P<0.0001) and for all other diseases (P<0.001), and also in the excess winter hospital admission ratio for respiratory disease (P<0.0001). With regard to sex, the excess ratios were lower in men than in women for both respiratory mortality (P<0.05) and respiratory hospital admissions (P<0.0001). We also observed that excess winter mortality ratios decreased significantly over the 18-year period for cardiovascular disease (P<0.05) and for all other diseases (P<0.05). Our results suggest that measures to reduce excess winter mortality should be implemented on a population-wide basis and not limited to socio-economically deprived areas. There may also be a case for tailoring interventions to specifically meet the needs of older people.

Introduction

Mortality is significantly higher during the winter months, and this pattern has been seen all over Europe.1., 2., 3. Excess winter mortality has been evident in Britain since the 1840s, and McDowell showed how the national crude quarterly mortality rate for January–March varied with the annual rate.¹ Using the whole year as 100, the January–March quarter increased gradually from 110 for 1841–1850, peaked at 128 during 1921–1930, and stayed high until the 1960s and 1970s when it started to decline. By 1980, excess winter mortality had fallen to 116. International mortality statistics documenting the period 1976–1984 showed that the excess winter mortality index (number of excess deaths in winter expressed as a percentage of the autumn to summer average) of 21% in England and Wales was considerably higher than in other countries.⁴ The index was 11% in France and Belgium, whilst in Finland, where it is much colder, it was even lower at 8%.⁴ Curwen estimated that winter accounted for 20,000 extra deaths each year in Britain, 55% of which were from circulatory disease and a further 33% from respiratory disease.⁴

Cardiovascular and respiratory disease also account for a large proportion of winter emergency hospital admissions, contributing to the recurrent shortage of beds during winter.5., 6., 7. Douglas studied the seasonality of deaths and hospital admissions in Scotland from 1974 to 1988.⁷ A peak in the number of deaths and admissions occurring during winter months and a trough in the summer that fitted a cosine wave pattern was observed. Coronary heart disease and respiratory disease showed different patterns for deaths and admissions. Coronary heart disease deaths produced a pattern of deep regular seasonal waves, whilst admissions lacked significant seasonality. Both respiratory deaths and admissions had high amplitude for seasonality.

Cold outdoor and indoor temperatures are thought to be important factors affecting excess winter mortality, but there are conflicting views about their relative importance. Many studies have suggested that excess winter mortality could be reduced by wearing warm clothing and being physically active outdoors, in combination with maintaining a warm indoor temperature.8., 9., 10. Keatinge et al. suggested that outdoor temperature exposure was more important than indoor temperature.11., 12. However, other studies have implied that the use of central heating may be effective in reducing excess winter mortality.13., 14. McKee suggested that the use of low-cost geothermal energy in houses in Iceland, where the average outdoor temperature is below freezing, could explain why it has one of the lowest excess winter mortality indices in Europe.¹⁴

There are several biological mechanisms that could potentially account for excess winter mortality. Exposure to low temperature stresses the circulatory system causing an increase in blood pressure, blood viscosity and fibrinogen concentration.15., 16., 17. An increase in plasma fibrinogen also occurs as a result of an acute response to respiratory infection.¹⁸ The effects of fibrinogen increase may be counteracted by vitamin C, and it has been suggested that a high intake of vitamin C in winter may be protective for respiratory and cardiovascular disease by reducing fibrinogen concentration.¹⁹ The increased incidence of influenza and other respiratory infections contribute to respiratory disease-related winter mortality.12., 20. Cold temperatures may also cause immunosuppression that decreases resistance to infections and assists the survival of bacteria in droplets.²¹

Surveys have found that socio-economically deprived people are more likely to have a poor diet, eating less fresh fruit and vegetables.22., 23. Poor people are also less likely to have a car, which could increase outdoor exposure to cold.²⁴ The lack of access to a car suggests increased use of public transport, where crowded conditions may increase the transmission of airborne infections.¹² Likewise, excess respiratory mortality may be related to overcrowding in houses.¹⁰ As many of the factors that would explain excess winter mortality and hospital admissions vary with socio-economic deprivation, it might be expected that excess winter mortality and hospital admissions would be particularly problematic amongst socio-economically deprived people. McDowell, using data from 1959 to 1963 and 1970 to 1972, found that all social classes suffered from excess winter mortality but this was higher in semiskilled and unskilled working classes.¹ However, he noticed that the differential had diminished between the two time periods. From a longitudinal study based on a 1% sample from the 1971 population census and followed until death, Curwen compared people with no access to a car and living in rented accommodation with people living in owner-occupied accommodation with cars.⁴ He found a slightly greater excess winter mortality of 23% in the more deprived group compared with 18% in the less deprived group. Bernard and Smith found that emergency admissions increased with socio-economic deprivation, but this was not specific to winter months.²⁵

The main aim of this study was to investigate the association between socio-economic deprivation and excess winter mortality and emergency hospital admissions in the South Yorkshire Coalfields Health Action Zone (HAZ), parts of which have substantial socio-economic deprivation related to the closure of the coal mines. First, we planned to examine the pattern of excess winter mortality and admissions, and then to see if there was a pattern of increasing excess winter mortality and admissions with increasing socio-economic deprivation at the small area geographical level. A subsidiary aim was to describe patterns of excess winter mortality by age and sex.