Geography Reference
In-Depth Information
Box 31.1 continued
Figure 31.2
Associations between life expectancy at
birth (LEB) and deprivation category for the District
Health Authorities (DHAs) of England. For each graph,
the horizontal axis plots the 105 DHAs according to a
seven-category ranking of Jarman deprivation scores
(category 1=least deprived); (A) LEB and deprivation
category for males (triangular symbols, lower line trace)
and females (circular symbols, upper line trace), 1992-
4; (B) gender (female—male=) difference in LEB and
deprivation category, 1992-4; (C) average annual
change in LEB and deprivation category for males
(triangular symbols, upper line trace) and females
(circular symbols, lower line trace). Average annual
change is expressed as a percentage and has been
computed for the period between 1984-6 and 1992-4.
and negative correlation coefficients for both males
(
r
=-0.77;
p
<0.01 in a two-tailed test) and females (
r
=-
0.56;
p
<0.01 in a two-tailed test).
2
Gender differentials in life expectancy are directly
associated with the level of deprivation in an area
Figure 31.2B plots the (female—male=) gender
difference in the LEB. Again, there is considerable
variation within a given deprivation category. However,
the overall tendency is for gender differentials to widen
as the level of deprivation in an area increases (
r
=0.77;
p
<0.01 in a two-tailed test).
3
Relatively affluent areas made the most pronounced
gains in life expectancy during the late 1980s and
early 1990s
For each deprivation category and DHA,
Figure 31.2C plots the percentage change (computed
as an annual average) in LEB between the periods
1984-6 and 1992-4. For males (upper line trace) and
females (lower line trace) in all districts, the LEB for
1992-4 was higher than 1984-6, marking a systematic
gain in life expectancy during the observation period.
But, Figure 31.2C shows that the level of improvement
was inversely associated with the level of deprivation
for both males (
r
=-0.47;
p
<0.01 in a two-tailed test)
and females (
r
=-0.30;
p
<0.01 in a two-tailed test). By
implication, the gap in life expectancy between
relatively affluent and relatively deprived areas
widened during the late 1980s and early 1990s.
These findings are consistent with a growing literature
on the spatial association of mortality differentials and
socioeconomic deprivation in the United Kingdom (see,
for example, Eames
et al
. 1993; Morris
et al
. 1996; Watt
and Ecob 1992; Wilkinson 1987). However, a lively
debate surrounds the nature and persistence of the
association (DHSS 1980; Blane 1985; Illsley and Le
Grande 1993; Macintyre 1998), and this impinges on the
policy implications of the research. On the one hand, the
association could be an artefact of the available data,
while the problems inherent in ecological studies (the
modifiable areal unit problem and the so-called
ecological fallacy) cannot be ignored. Alternatively,
spatial associations may reflect patterns of social mobility
as they relate to health status, while Illsley and Le
Grande (1993) argue that health-related behaviour and
ethnicity (rather than deprivation
per se
) underpin
regional variations in mortality.
Clearly, much research and, more importantly, well-
informed interpretation of research findings is still
required to elucidate the nature of the association
between ill-health, longevity and the socio-economic
environment. As described in the main text, it is precisely
these issues that research initiatives such as the ESRC
Health Variations Programme seek to address. Not only
has work of the type outlined by Soni Raliegh and Kiri
served as a catalyst for such investigations, but it also
provides a baseline for monitoring the success of future
policy initiatives.
Source:
Data from Soni Raleigh and Kiri 1997: Tables 1,
p. 651 and Table 2, pp. 654-5.
