Medical Research Council

Social & Public Health Sciences Unit

Perspectives on Health Inequalities:

The Need for a Lifecourse approach

Patrick West

MRC Social & Public Health Sciences Unit

Occasional Paper
September 1998

MRC Social & Public Health Sciences Unit,

University of Glasgow,
4 Lilybank Gardens,
Glasgow, G 12 8RZ.
Tel: 0141 357 3949
http://www.msoc-mrc.gla.ac.uk/

MRC Social and Public Health Sciences Unit

Perspectives on Health Inequalities:

The Need for a Lifecourse approach
Table of Contents
Page

About the Author

About the paper

INTRODUCTION

THE NATURE OF HEALTH INEQUALITIES

EXPLANATIONS OF HEALTH INEQUALITIES

Artefact explanation

Natural/social selection

10

Cultural/behavioural explanation

11

Materialist or structuralist explanation

12

A CONTRAST; RELATIVE POVERTY OR A LIFECOURSE PERSPECTIVE

15

CONCLUSION

18

BIBLIOGRAPHY

19

Acknowledgements
The author wishes to thank Sally Macintyre and Helen Sweeting for comments on this
paper, and Ossi Rahkonen and Eero Lahelma for the opportunity of participating in the
Helsinki symposium and subsequent publication. He is financially supported by the
Medical Research Council of Great Britain

Page i - Occasional Paper September 1998

MRC Social and Public Health Sciences Unit

About the Author

Patrick West is a senior researcher at the Social and Public Health Sciences Unit and is
supported financially by the Medical Research Council of Great Britain. He has overall
responsibility for the Units Youth Programme. His research interests include young
people and health (particularly mental health), health behaviors, lifestyles, and family and
school influences on health as well as health inequalities both in relation to the whole
lifecourse and the specific stages of childhood and youth. He is currently on the editorial
boards of Health Education, Youth & Policy and the Journal of Youth Studies.

Page 1 - Occasional Paper September 1998

MRC Social and Public Health Sciences Unit

About the paper

Perspectives on Health Inequalities: The Need for a Lifecourse Approach
This paper was first presented with the same title at a symposium on health inequalities
which formed part of a postgraduate course in Medical Sociology in the Department of
Social Policy, University of Helsinki, Finland on September 1st, 1997.
It was
subsequently translated into Finnish and published as a chapter in a collection edited by
Ossi Rahkonen and Eero Lahelma.
The English version was first published by MRC Medical Sociology Unit, 6 Lilybank
Gardens, Glasgow G12 8RZ 1998.
Republished by the MRC Social & Public Health Sciences Unit, 4 Lilybank Gardens,
Glasgow G12 8RZ May 2000.
The Finnish translation was published as Patrick West (1998), Terveydentilan eriarvoisuus ja elmnkaarinkkulma, In: Ossi
Rahkonen ja Eero Lahelma (eds) Elmnkaari ja terveys, Gaudeamus Books, Helsinki,
pp. 23-47.

Page 2 - Occasional Paper September 1998

MRC Social and Public Health Sciences Unit

INTRODUCTION
At the time of writing, there is renewed interest in health inequalities in several countries
in Europe and elsewhere which extends beyond the academic community to the policy
arena. In Britain, the recently elected Labour Government has commissioned an
independent inquiry into health inequalities in England and Wales under the
chairmanship of Sir Donald Acheson, the former Chief Medical Officer, with the specific
remit to review the evidence and produce recommendations. In Germany, senior
academics are preparing a briefing paper to persuade their Government to do the same.
In Sweden, meetings are being held this Autumn with the aim of getting the issue onto
the political agenda, and similar moves are being made elsewhere. Even in countries as
dissimilar as the U.S.A. and Hungary, there is a growing concern about health
inequalities as evidence accumulates about the existence - and persistence - of
differences in health between people in different social classes or income levels in
advanced societies, many of them with well developed welfare states.
This flurry of activity suggests a degree of confidence among academics and policy
makers alike that we now not only have the necessary evidence about the existence of
health inequalities but that we understand their causes and are therefore in a position to
offer appropriate remedies. While applauding this at a general level, I want to introduce
a note of caution by arguing that we are still quite a long way from a comprehensive
understanding of the causes of health inequalities and that because of this there is a
danger we will offer too simple a solution to a complex problem. The health inequalities
debate, as it has become known, is not a neutral one. It inflames passions and divides
people of different political persuasion. The debate has been characterised by a sharp
polarisation of views about a range of issues such as the appropriateness of particular
ways of measuring health, the degree to which health inequalities are widening over
time, and particularly the explanations for health inequalities. Here, a sharp polarity has
been apparent between those who see a role for individual responsibility (via health
behaviours like smoking, for example) and those who dont; between those who see a
role for health related mobility and those who dont; and more recently between those
advocating a role for psychosocial mechanisms and those who dont. In some part,
perhaps in very large part, this polarisation of views is not a reflection of a balanced
assessment of the evidence but of a tendency to seek a single all-embracing explanation
of health inequalities and downplay the significance of other possibilities. It is what Sally
Macintyre, in a recent review (1997), has characterised as a hard rather than soft
version of the explanation of health inequalities.
In this paper, I want to look at these issues in more detail. In reviewing several
perspectives on health inequalities, including the recent and potentially very influential
thesis of Richard Wilkinson (1996), I will argue that there remains a tendency in the
health inequalities debate to go for the big explanations rather than concentrating on all
the evidence and seeing each bit as part of a very complex jigsaw. In this respect, what
has become known as the lifecourse perspective on health inequalities (Lundberg,
1993; Wadsworth, 1997, Power et al., 1997) seems to offer a way forward in that it forces
investigators to consider each stage of the lifecourse, from the embryo through to very
old age, as containing particular influences on health and biography which are
synergistically interrelated. In this model, health inequalities are the product of an
accumulation of (dis)advantages over the lifecourse, some of which may be outside the
individuals control, some not; some of which may derive from the social class an
individual is born into, some from influences related to social mobility; some of which
may take the form of physical effects of the environment, some of which may involve a
Page 3 - Occasional Paper September 1998

MRC Social and Public Health Sciences Unit

psychosocial mechanism. Not only does this perspective direct the investigator to stick
very closely to the evidence indeed, but it also warns against the advocacy of any one
all-embracing solution to the problem and suggests instead a multifaceted approach.

Page 4 - Occasional Paper September 1998

MRC Social and Public Health Sciences Unit

THE NATURE OF HEALTH INEQUALITIES

Before considering the central question of explanations for health inequalities, let us first
examine what we know about the form and magnitude of health differentials in advanced
societies. Of necessity, this must be brief and several excellent reviews on this issue are
available (Macintyre, 1986, 1997).
Interest in health inequalities is not new. In Britain, as early as the first half of the 19th
Century, public health reformers documented striking differences in death rates between
the poor and the well-to-do. Edwin Chadwick (1842), for example, reported that in
Liverpool in 1840 the average age of death for the gentry was 35 while that for labourers
was 15 years. Nearly a century later, Stevenson (1928), the originator of the Registrar
Generals classification of social classes, showed a steep increase in the rate of infant
mortality from 48 per 1000 live births in social class I to 123 per 1000 in social class V,
and a corresponding increase in the rate of adult mortality from 81 per 1000 to 127 per
1000. At various intervals thereafter, similar class differences in mortality were reported
in Government publications. In general, however, the issue we now refer to as health
inequalities was largely invisible, and particularly so in the period after World War II up to
about the mid-seventies. A number of factors, including the increasing affluence of the
British population and the belief that the welfare state (exemplified by a National Health
Service free at the point of delivery) was eradicating problems of poverty, probably
contributed to the relative inattention and lack of interest in health inequalities during this
time.
This situation changed very dramatically with the publication of the Black Report
(Townsend et al) in 1981. The Report, which had been commissioned by the previous
Labour administration in response to an increasing concern that the NHS was not in fact
solving the problem of health inequalities, had three central aims: to review the evidence
about class differences in health, to consider possible causes, and provide policy
recommendations. Its importance cannot be overestimated, and in one way or another it
has set the agenda for research on health inequalities in Britain and elsewhere (not least
Finland) to this day. Among the many thousands of publications on health inequalities,
there is hardly one which does not make some reference to it, and it has become
something of a bible for generations of students interested in the issue.
The Black Report relied almost exclusively on mortality rates as the best available
indicator of the health (ibid., 55) of people in different social classes, an ironic indicator
of health but nevertheless one which because of its objective nature continues to be
favoured by many investigators to this day (Davey Smith et al, 1994; Wilkinson, 1996).
Using data from the 1970-72 Decennial Supplement on occupational mortality, they
showed that the crude death rates of both men and women aged 15-64 in social class V
were two and a half times those of men and women in social class I. Furthermore, using
standardised mortality ratios (SMRs - standardised for age), regular class gradients were
shown to exist for stillbirths, infant mortality (deaths in the first year), childhood mortality
(age 1-14 years) and adult mortality (15-64 years). An examination of causes of death
also revealed that health inequalities existed for most causes, notably accidents in
childhood and diseases of the respiratory, digestive and genito-urinary systems and
malignant neoplasms in adulthood. Although, as indicated, the Black Report relied
heavily on mortality statistics, some other data on morbidity from the British General
Household Survey (GHS) were also interrogated. This showed that in the total
population, rates of limiting longstanding illness increased by a factor of three from
people in the professional classes to those in the unskilled manual class. Taken together,

Page 5 - Occasional Paper September 1998

MRC Social and Public Health Sciences Unit

all the evidence led to the now famous conclusion that health inequalities were a
pervasive feature of the lifecourse.
Since the publication of the Black Report, much has been achieved by the burgeoning
research effort into the study of health inequalities. Until relatively recently, the
predominant focus has been descriptive rather than explanatory, the main findings of
which can be summarised as follows.
First, there is now evidence that health inequalities are a feature of all developed
societies. At the time of the Black Report, very little information existed on this issue and
it was widely believed that in some societies such as Sweden class differences in health
simply did not exist. We now know that this is not true both for Sweden and other
countries in western Europe (Illsley & Svensson, 1986), eastern Europe (Vagero &
Illsley, 1992), North America (Kunst et al.,1992) and Australasia (Pearce et al., 1993). In
Finland, this was first demonstrated by Valkonen et al. (1993). Most of the evidence on
this issue refers to mortality but in some countries, including Finland, similar measures of
morbidity as used in the British GHS have revealed evidence of health inequalities
(Lahelma et al., 1994).
Second, most but not all the evidence suggests that health inequalities have increased
over time in spite of the fact that almost everywhere (Hungary is a notable exception),
infant mortality and premature mortality have declined. In Britain, mortality differentials
between men in social class I and V widened between 1971 and 1981 (Davey Smith et
al., 1990), and continued to widen up to 1991 (Illsley & Baker, 1997). Very recent
evidence, based on a comparison of mortality rates in different geographical areas in
Britain also suggests that health inequalities have widened (Dorling et al., 1997). In
Finland, class differences in mortality among middle-aged and elderly men increased
between 1971 and 1981 (Valkonen et al., 1993). The conclusion that class differences in
health are widening has, however, been recently criticised by Illsley and Baker (1997)
who argue that it is only sustainable in comparisons between the top and the very bottom
of the social scale, the latter comprising an increasingly small proportion of the
population, the underclass.
Third, at the time of the Black Report very little evidence about variations in mortality (or
morbidity) by dimensions of class other than that based on occupation was available. In
the light of the changing nature of the class structure in contemporary society, and the
inadequacy of an occupational based classification scheme to classify women, the
retired and other groups excluded from the labour market, this has been an important
source of criticism about the magnitude (and even existence) of health inequalities.
There is now evidence from several countries that health differentials are apparent with
respect to alternative measures of social class such as housing tenure, car ownership,
educational level, income (Goldblatt, 1990) and area of deprivation (McLoone & Boddy,
1994). Of particular interest in this connection is income since the finding that health and
wealth are related is open to quite different interpretations.
Fourth, one of the most striking features of the observed relationship between social
class (and income) and health is that the relationship is stepwise or linear; that is, that
increases or decreases in health occur at every point in the class structure or income
level. Thus, the difference in mortality rates between classes at the top of the social
scale (e.g. I and II) is similar in magnitude to the difference between classes at the
bottom of the social scale (e.g. IV and V). This pattern has actually been evident for a
very long time, as Macintyre (1997) has noted, but the significance of the hierarchical
relationship has only relatively recently been seen as interesting, most notably in the
U.S. with the MRFIT study (Davey Smith et al., 1996) and in the U.K. with the Whitehall I

Page 6 - Occasional Paper September 1998

MRC Social and Public Health Sciences Unit

and II studies of civil servants (Rose & Marmot, 1981: Marmot et al., 1997). The pattern
is surprising given that most assumptions about the relationship between class or income
and health would commonsensically suggest a change in shape beyond a certain point
or threshold at which basic needs were met. That this does not happen poses an
interesting problem for the explanation of health inequalities.
Fifth, the Black Report, and many commentators since (Whitehead, 1987), assumed that
health inequalities were an invariant feature of the lifecourse. On the basis of the
evidence available to them, this was in fact a reasonable assumption. Further work, with
which I have been particularly associated, has thrown doubt on the validity of that
assumption as it pertains to the situation of youth, a stage in the lifecourse between
childhood and adulthood I have characterised as being one of relative equality rather
than inequality (West, 1988, 1997). This pattern, which has also been found in other
countries, including Finland (Rahkonen & Lahelma, 1992), suggests that in youth other
influences associated with the school, the peer group and youth culture may cut across
the influence of social class to promote equalisation, at least in respect of some
measures of health (West, 1997). That this is no more than a temporary process is
indicated by the way health inequalities (re)emerge in early adulthood. This occurs both
in respect of mortality and morbidity as indicated in our own West of Scotland Twenty-07
study (Macintyre et al., 1989) on a wide range of health measures, including the GHS
measure of longstanding illness (West et al.,1990). This pattern, which has been
referred to by Vagero (1991) as an important anomaly, also poses some interesting
problems of interpretation - how does a pattern of relative equality get translated into one
of health inequalities in such a relatively short time?
Sixth, while the Black Report depended heavily on mortality statistics as evidence of
health inequalities, increasingly research on the issue has used indicators of morbidity,
most usually of the type included in the British GHS but also extending to measures of
general physical health, mental health, symptoms and physical measures such as height,
weight, blood pressure and respiratory function. As a general rule, class gradients in
relation to morbidity are not as steep as are observed for mortality (Macintyre, 1997).
They also vary between measures, being more marked for measures of health status
such as chronic illness than for health state such as physical symptoms. In the Twenty07 study, in two adult cohorts aged 35 and 55 respectively, class gradients were found
for longstanding, and limiting longstanding, illness, self-rated general health, mental
health, height and weight, and respiratory function, but not blood pressure (Ford et al.,
1994). Other studies have found blood pressure to be class-related, but in general the
gradient is not very steep (Cox et al., 1987). In the British 1958 birth cohort (NCDS),
class patterning of health in young adulthood (age 23) varied considerably between
measures, with some (e.g. malaise) exhibiting a health inequalities pattern, others (e.g.
limiting longstanding illness) showing no relationship, and others (e.g. allergies) a
reverse class gradient (Power et al., 1991). Similar findings have been reported from the
later sweep of that study (age 33) with, in addition and against expectations, very little
evidence that class gradients had widened over the ten years since the earlier sweep
(Power et al., 1997). It is a salutory reminder to researchers in the health inequalites
field that on some health measures only a small proportion of the variance is explained
by social class or another measure like income (Der et al., 1997).
In the years since the Black Report, we have learned a great deal about the social
patterning of health, much of which is now broadly agreed though there remain some
areas of controversy and others which pose problems of interpretation. There is broad
consensus that health inequalities are a feature of all developed societies, that they have
probably increased over time, that they are observable with a range of measures of
social class, and that the observed relationship between class and health conforms to a

Page 7 - Occasional Paper September 1998

MRC Social and Public Health Sciences Unit

linear rather than threshold model. There is increasing recognition that the apparently
anomalous situation in youth is a real one though some dispute this, principally on
methodological grounds (Blane et al., 1994). This finding, that health inequalities are not
an invariant feature of the lifecourse, together with the finding that health inequalities are
not invariably found with every health measure, is important and poses problems for a
single all-embracing explanation of health inequalities. I shall return to these issues
later.

Page 8 - Occasional Paper September 1998

MRC Social and Public Health Sciences Unit

EXPLANATIONS OF HEALTH INEQUALITIES

If there is an emerging consensus about many descriptive features of health inequalities,
this is not (yet) the case in respect of the explanation of the phenomenon. The history of
this aspect of the health inequalities debate since the Black Report is one characterised
by controversy in which various individuals and research groups have engaged in
vociferous argument, often holding or being characterised as holding highly polarised
positions. In large part, this has been a peculiarly British preoccupation, other countries
having much more civilised debates, but it is important to remember that the publication
of the Black Report happened within a year of the election of the Thatcher Government,
undeniably the most ideologically driven administration of recent times, and this set an
agenda for a polarisation of views about almost everything. It is also the case that the
authors of the Black Report in many respects contributed to this polarisation in the way
they constructed a typology of explanations of health inequalities which either confused
certain issues or rested on a particular view of the mechanisms involved.
The Black Report identified four candidate explanations of health inequalities which to
this day continue to dominate the debate. These were the artefact explanation, natural
or social selection, the cultural/behavioural explanation and materialist or structural
explanation, the latter by far the most favoured by the authors. I will consider each one
of them in turn.

Artefact explanation
The artefact explanation of health inequalities rests on the idea that the relationship
between class and health is artificial rather than real. It is a measurement phenomenon
which arises either through the (inadequate) measurement of social class and/or health,
or in the measurement of the relationship between the two. An early problem of the latter
kind was referred to in Britain as the numerator/denominator bias, a possibility which
arose in the process of matching occupational data available from death certificates (the
numerator) with population data about social class derived from the census (the
denominator), the implication being that working class rates might be artificially elevated
or middle class rates reduced (it could of course have worked the other way round to
conceal the real magnitude of class differentials). With the availability of linked
(individual data on class and mortality) datasets in Sweden, Finland and elsewhere, and
the advent of the Longitudinal Study (LS) in Britain (Goldblatt, 1990), the
numerator/denominator bias has been shown to be false. More complex problems of
measurement of both social class and health remain. In respect of the former, while
studies using alternative measures of class such as housing tenure or income
demonstrate that at any one time health inequalities are not simply the product of an
occupational based classification system, there are problems of comparison over time
occasioned by the changing nature of the class system (Illsley, 1997). With respect to
health, there are also potential problems of reporting bias in relation to measures of
morbidity, one of the reasons why some investigators (e.g. Wilkinson, 1996) prefer to use
mortality. The usual assumption is that because of the greater volume of ill-health in the
working classes, what is defined as illness is more severe than among the middle
classes, a difference which results in an underestimation of the true extent of health
inequalities (Blane et al., 1994). While important, this is only one view about how illness
is identified (and reported), another suggesting that through middle class stoicism quite
the reverse could occur (Williams, 1990). More generally, the question of the
appropriateness of health measures has been raised as a possible explanation for the

Page 9 - Occasional Paper September 1998

MRC Social and Public Health Sciences Unit

finding of a pattern of relative equality in youth, an issue which certainly merits further
investigation (West, 1997).
There is no doubt that there remain several problems in relation to the measurement of
health inequalities. The cumulative evidence, however, is overwhelming that health
inequalities are real rather than artefactual. This is one explanation advanced by Black
which by almost universal agreement can be rejected.

Natural/social selection
Any assessment of the role of selection in relation to health inequalities has to begin with
a distinction between natural and social selection which the Black Report coupled
together, thereby obscuring a fundamental difference in the processes involved and
inadvertently creating the conditions for polarised views thereafter. For the authors, this
explanation of health inequalities involved the idea that health was a cause of social
class position rather than a consequence, and that fitter individuals (both in relation to
health and other characteristics) are selected into higher classes while the weakest and
most frail drift down the social scale to occupy the lowest position and reap the most
meagre rewards (op. cit., 105). In this view, both health and class position are the
outcome of innate characteristics or genetic predispositions, and it is therefore a
profoundly Social Darwinist view which is advanced wherein processes of social
selection play no part other than presumably reflecting natural selection. With the
occasional exception (Himsworth, 1987), there is almost nobody who would subscribe to
this view but within the inequalities debate there has been an unfortunate tendency to
assume that those who see a role for social selection (e.g. Illsley, 1986) are closet
natural selectionists (Wilkinson, 1986).
Reconceptualised within a sociological perspective (West, 1991), the idea of health
selection proceeds from the unremarkable postulate that health - and health related
characteristics - can have consequences for social life, including success or failure in the
labour market and class structure. In this view, for profoundly social reasons, those
people in better health are more likely to experience upward social mobility while those in
worse health are more likely to experience downward mobility, the combined effect of
which is to contribute to a picture of health inequalities. It is postulated that this can
happen between generations (inter-generational mobility), as when a disabled young
person for example moves down the social scale relative to his/her class of origin, or
within a generation (intra-generational mobility) as when a chronically sick adult changes
from a higher to lower status occupation. There are several processes by which this
might occur, including family characteristics which impact both on health and education,
thereby affecting subsequent inter-generational mobility chances (Vagero & Illsley,
1995), and discrimination by employers against those with health or health related
problems (West, 1991). Finally, a distinction exists between what has been called direct
health selection in which health itself (consciously or unconsciously) is related to social
mobility, and indirect health selection in which some other attribute marking health
potential is similarly involved with the mobility process (Wilkinson, 1987; West, 1991).
An example of the latter is height, which is associated both with health and social
mobility (Macintyre & West, 1991), though the problems of sorting out the causal process
are considerable.
A good deal of work has been undertaken to examine the health selection hypothesis,
most usually to rule it out as an explanation of health inequalities. Most, but not all,
studies find some evidence of direct health selection. In the 1946 British birth cohort,
children with serious illness were more likely than those without to be downwardly mobile
(Wadsworth, 1987). In the 1958 cohort, that was true among those with malaise at age
Page 10 - Occasional Paper September 1998

MRC Social and Public Health Sciences Unit

16, much less so among those with poor general health (Power et al., 1991). In the
Swedish Project Metropolitan study, much greater evidence of selection was found in
relation to mental health (Timms, 1995). By contrast, in a recent analysis of the British
GHS, almost no evidence of downward mobility was found for those with poor self-rated
health or limiting longstanding illness (Rahkonen et al., 1997a). The question which has
assumed more importance in recent years, however, is not whether health related
mobility occurs but how important it is as an explanation of health inequalities. The
general conclusion is that it plays a small but significant role (Blane et al., 1993), though
some recent commentators, perhaps rather unwisely since the analysis rests on one
health measure only, describe its influence as negligible (Power et al., 1996).
There has in addition been some interest in the process referred to as indirect health
selection which I earlier indicated might be much more significant than direct health
selection (West, 1991). Its potential significance derives from the way particular
characteristics of individuals which may mark underlying health status might impact both
on social mobility and on subsequent class inequalities in health. Among several
characteristics considered, height has received the most attention and there is abundant
evidence from several studies that height is related to mobility, with shorter people being
more likely to be downwardly mobile, taller people more likely to be upwardly mobile
(Macintyre, 1988). The problem of its relationship to health inequalities, however, is
knowing how height translates into health since it could simply be that stature
predisposes to subsequent health risk rather than containing differential health risk.
Other attributes which have received some attention are health behaviours such as
smoking, and more generally young peoples lifestyles, which are also linked in a regular
way to social mobility (Glendinning et al., 1995) but similar problems of interpretation are
encountered here. At the present time, the jury is out on the issue of indirect health
selection, and it may well be that the main problem is one of lack of conceptual clarity.

Cultural/behavioural explanation
In contrast to health selection, the third explanation identified in the Black Report views
class as antecedent to health with cultural/behavioural factors as the mechanism by
which health inequalities are produced. This is therefore one type of social causation
explanation, but in rather the same way that the linking of natural and social selection
caused misunderstanding, the coupling together of behavioural and cultural explanations
has, in my view, obscured the potential importance of culture as an influence on health
inequalities (Sweeting & West, 1995). For the authors of the Black Report, although they
allude to the role of ideas and values, the definition of culture is reduced to a set of
specific behaviours which have consequences for health. These behaviours, which are
well known and typically comprise smoking, drinking, diet and exercise (Blaxter, 1990),
were regarded as cultural because of the common understanding that they are, or
should be, within the control of individuals. In this respect, they were reflecting the
message of the previous British Labour administrations white paper Prevention:
Everybodys Business (1976), a message which continues to dominate health education
and health promotion in Britain and elsewhere.
There is no disputing the aetiological significance of behaviours like smoking or diet for
subsequent morbidity and mortality, nor that these behaviours are socially patterned. In
Britain, for example, smoking among adults continues to be much more prevalent among
working class compared with middle class people, in some measure because of the
greater capacity of the latter to quit the habit (Amos & Hillhouse, 1993). Successive
studies of diet, too, have revealed a pattern of poorer nutrition among both children and
adults from lower social class backgrounds (RUHBC, 1997). The important question,
however, concerns the extent to which such behaviours constitute a complete
Page 11 - Occasional Paper September 1998

MRC Social and Public Health Sciences Unit

explanation of health inequalities as some people seem to assume. The research

evidence strongly suggests they do not. For example, in the Whitehall I study of civil
servants, controlling for the risk factors of smoking, cholesterol and blood pressure did
not account for the fourfold increase in CHD mortality between administrative and other
grades, though they did explain about 25% of the difference (Rose & Marmot, 1981).
Recent evidence from a large community study in the Netherlands, which used a wider
range of health behaviours, found they had a larger explanatory role in relation to several
measures of morbidity, accounting for about 40% of the variance (Stronks, 1997). In this
study, the investigator attempted to separate an influence of health behaviours from that
associated with material circumstances, concluding that each had a part to play. This is
an interesting attempt to distinguish a voluntary component in behaviours like smoking
from an adaptive component in which the behaviour is conceptualised as a modifier of
stress associated with poverty (Graham, 1994). In this latter view, health behaviours are
not viewed as cultural expressions at all but as a particular type of materialist
explanation.
The tendency to regard health behaviours as being synonymous with cultural influences
is, I believe, short-sighted. It is certainly an impoverished view of culture, and fails to
give recognition to a wide range of factors either embedded in class cultures or
differentially characterising individuals or families within the same social class which
could, via social mobility, impact on health inequalities. One example of this refers to the
value placed on health and education respectively which, as Vagero and Illsley (1995)
suggest, could co-exist as components of a class culture to influence both morbidity and
achievement, and hence health inequalities. Another refers to the role of the family in
influencing health (Sweeting & West, 1995). In the Twenty-07 study, young people from
families characterised by conflict were more likely to report malaise and physical
symptoms and to do less well in the labour market. Inasmuch as this is class related,
these findings point up a way in which family culture may impact on health either
directly, via a psychosocial mechanism like self-esteem, or indirectly, by exposing young
people differentially to subsequent health risk. The potential significance of family life for
longer term health problems is also strongly suggested in Lundbergs (1993) findings
from the Swedish Level of Living survey where, controlling for earlier material factors, the
experience of family conflict and breakup were associated with poorer physical and
mental health in adulthood.

Materialist or structuralist explanation

The fourth explanation of health inequalities advanced in the Black Report is that referred
to as materialist or structuralist, by which the authors meant a whole range of factors
which are determined by the class or income structure and which have an impact on
health and wellbeing. As Macintyre (1997) notes, despite the authors attempt to clarify
the central components of the explanation, there remains a confusion between what she
terms the hard version, in which material conditions themselves (e.g. physical aspects
of the environment) are seen as the key determinants of health, and the soft version in
which social as well as economic capital (e.g. education) are regarded as important, the
latter incoporating psychosocial mechanisms. The inclusion of social capital within a
materialist explanation owes much to a Marxist view of the world in which class culture is
conceptualised as the superstructure of the material class base. A hard view of this
soft version would rule out a role for culture altogether since all values (and health
behaviours) are products of the underlying class structure.
Until recently, much of the effort of research workers concerned to develop the
materialist explanation has been directed rather more to disproving other explanations of
health inequalities (notably health selection) than in identifying mechanisms by which the
Page 12 - Occasional Paper September 1998

MRC Social and Public Health Sciences Unit

relationship between income or car ownership and health might actually work. Indeed, it
sometimes seems that the finding of a minor role for selection or health behaviours is
judged an adequate demonstration of a major role for materialist factors. There are,
however, four main areas which have been the subject of some research and which refer
to the intrauterine environment, physical features of the external environment,
psychosocial influences, and opportunity structures respectively.
The first of these was not originally formulated within a health inequalities perspective at
all, but has since been incorporated as one type of materialist explanation, sometimes to
the exclusion of all others. The research concerned has been developed by Barker and
colleagues (1992) who, in an ingenious series of studies using both aggregate and
individual level data, have revealed relationships between infant health, notably
birthweight, and much later morbidity and mortality in adult life. The principal finding
concerns the link between low birthweight and high blood pressure in adulthood which in
turn leads to increased CHD and stroke mortality. The mechanism by which this is held
to occur is an essentially biological one involving a relationship between intrauterine
growth and the programming of body functions like blood pressure such that when
growth is impaired the body sets a higher level of blood pressure. Since birthweight is
class related, it follows that such a process could have consequences for the subsequent
development of health inequalities, and since foetal growth is itself related to maternal
nutrition, the underlying cause is seen to be poor maternal nutrition associated with
poverty. The thesis has attracted a good deal of critical attention, particularly in respect
of the possibility that the relationship between birthweight and later health is spurious,
though in a recent Swedish study which controlled for a number of factors including
health behaviours in adulthood an independent effect of birthweight on blood pressure
was found (Koupilova et al., 1997). Despite the sometimes very exaggerated claims to
explain health inequalities, the relationship between birthweight and later blood pressure
is in fact quite small.
The second area which has received attention refers to a rather longer tradition of
research focusing on a range of physical hazards in the environment including pollution,
work settings and housing conditions. In general, though the effects of environmental
pollution and the physical hazards associated with work are acknowledged as potential
risks to health, their effect on health inequalities is not nowadays regarded as particularly
important. More importance has been given to housing conditions. In one study (Martin
et al., 1987), which is often referred to, a relationship between damp housing and
respiratory problems in children, though not adults, was found. The finding in respect of
children has not, however, been consistently replicated (Strachan et al., 1995), and it
may be that here too the size of the effect has been exaggerated.
A third area which has assumed particular importance in recent years refers to
psychosocial processes linking environmental stressors to physical and mental health.
Research into this issue has been heavily concentrated on the work situation, a common
finding being that workers in lower status occupations characterised by low levels of
control over work, a fast work pace, and low social support have higher CHD morbidity
and mortality (Karasek et al., 1988; Marmot et al., 1997). The plausibility of a
psychosocial mechanism underlying this relationship has been considerably increased
with the demonstration of changes in cortisone and fibrinogen levels in blood chemistry,
both of which are indicators of stress. In principle, the underlying processes presumed to
link stressors in the work situation to workers health might apply to other situations like
the domestic setting or even to society itself.
The fourth area of research takes an entirely different approach to those which in one
way or another are looking at the direct effects of the socio-economic environment on

Page 13 - Occasional Paper September 1998

MRC Social and Public Health Sciences Unit

health. Here the focus is on opportunity structures both for the promotion of good health
and for life chances more generally, both of which are differentially distributed in the
class or income structure. In respect of the former, attention is directed to the provision
of facilities and availability of goods and services in particular areas. Macintyre et al.
(1993), for example, have shown how people living in poorer areas have fewer
recreational and sports facilities than those living in better-off areas and that in addition
there is a price disincentive, particularly in the poorer area, to eat healthily (Sooman et
al., 1993). In respect of the latter, attention is directed to educational and labour market
opportunities which could involve several different kinds of effect via knowledge (e.g.
health education), social identity (e.g. self-esteem) or differential exposure to risk (e.g the
likelihood of unemployment). Because the relationship between social structure and
health in this model is essentially indirect, it focuses attention on the processes
connecting them as they unfold over time.
It is quite evident, even with such a sketchy outline, that there are several different ways
in which a materialist explanation of health could work. It is similarly apparent that these
are likely to differ between different environments such as the work or domestic situation,
and that they are likely to impact at different stages in the lifecourse, possibly also on
different dimensions of health. It is also the case that the opportunity model outlined
above in many respects overlaps with a broader cultural perspective on health
inequalities. This diversity of possible causes, and the blurring of the distinction between
material and cultural factors, is not what the authors of the Black Report had in mind
when advancing the idea of a materialist explanation. Although they acknowledged a
diversity of material influences, implicit in their understanding, and articulated more
starkly in the debate since that time, is the idea that each of these influences is caused
by a single underlying feature of society, the class structure itself or different income
levels. In more familiar terms, this means poverty, the eradication of which via a
redistribution of income is seen as the solution to the problem. The materialist
explanation, as a big explanation demanding big remedies, runs the risk of
oversimplifying the issues in a way that a lifecourse perspective does not.

Page 14 - Occasional Paper September 1998

MRC Social and Public Health Sciences Unit

A CONTRAST; RELATIVE POVERTY OR A LIFECOURSE

PERSPECTIVE
In a long tradition of big explanations, none has come bigger in recent times than the
thesis developed by Richard Wilkinson (1997). It is an explanation rooted in a materialist
perspective which paradoxically elevates psychosocial processes rather than material
(physical) factors as the key mechanism underlying health inequalities in contemporary
societies. It is a thesis which has achieved a very high profile on the scientific, political
and media agenda, and therefore merits some attention.
Wilkinsons thesis begins with the identification of a paradox. This is that while the health
of poorer societies, as indicated by life expectancy, closely mirrors their economic
development, beyond the epidemiological transition - a point at which the diseases of
affluence (e.g. heart disease or stroke) replace those of poverty (e.g. infections) as the
major killers - a nations health is not determined by economic growth per se but rather
by the degree of differentiation in living standards experienced by its population. In
developed societies, life expectancy is higher among those which are more egalitarian
and lower in countries like Britain which have greater inequalities in income. In the
contemporary context, he argues, it is not absolute poverty which kills but relative
poverty, as indicated in particular by the linear rather than threshold relationship income
has with mortality. Thus, within this hierarchy, it is not merely the poorest who are
deprived, and least healthy, but each and every one of us in relation to the position we
occupy vis-a-vis another (richer or poorer) social group. In inegalitarian societies, the
resultant stress this generates not merely compromises the immune system, leading to
higher mortality rates, but it also, he claims, has catastrophic effects on a range of other
behaviours such as crime, drug use and behaviour disorders in children. The solution,
he believes, is to rediscover, via a redistribution of income, a more cohesive society with
lower levels of chronic stress and better health.
In many respects, Wilkinsons thesis is extremely persuasive in its claim to provide a new
psychosocial explanation for health inequalities. The problem, as with other single
explanations, is that it explicitly rules out any role for other mechanisms such as
selection, health behaviours and even absolute poverty, and in addition is remarkably
weak in adducing evidence in support of two central tenets of the argument. The first
refers to the almost complete absence of any data on mental health which, given the
importance ascribed to the effects of chronic stress, would be expected to occupy a fairly
central position. Indeed, it is perplexing that although overall mortality rates are
correlated with the degree of income differentiation in societies, this does not apply to
suicide! The second refers to the similar absence of data on social cohesion which is
inferred either from historical case studies or apparent symptoms of disintegration such
as drug use rather than from the perceptions of people themselves both in relation to the
kind of society they live in and the position they occupy in the hierarchy. This is perhaps
a harsh criticism since such data is not widely available, but this has to be balanced
against the tendency of any grand theorist to be selective in the use of evidence. The
problem fundamentally is that at almost every point in the thesis there is an alternative
explanation. For instance, the linear relationship of income and health fits a health
behaviour explanation of health inequalities just as much as it does a psychosocial one.
It is also possible, as Marmot et al. (1997) have suggested, that different mechanisms
could operate at the top and bottom of the social scale, the former being more
susceptible to psychosocial processes, the latter being more affected by the effects of

Page 15 - Occasional Paper September 1998

MRC Social and Public Health Sciences Unit

absolute poverty.
Despite these criticisms, Wilkinsons achievement in locating
psychosocial mechanisms in the health inequalities debate is considerable.
One of the other criticisms of the thesis that psychosocial mechanisms are the key to
understanding variations in mortality between societies is that it implies that mortality
rates would respond rapidly to social change. There is indeed some evidence for this, as
the east European experience shows, but in general this is not the case, most countries
showing a steady improvement over time. This seems to suggest that, whatever factors
are responsible, they work over the lifetime of individuals. Indeed, if we pause to
consider the evidence, this is compatible with a range of research findings testifying to
the importance of early life factors for later morbidity and mortality. Furthermore, much
of what we know both about the shape of class patterning of health at different ages and
the factors which influence health at any one stage suggests that a profitable approach
to understanding health inequalities is to think in terms of an accumulation of risk over
the lifecourse.
It is observations like this that have given rise in recent years to what is becoming known
as the lifecourse perspective on health inequalities, a perspective held in particular by
those working with longitudinal datasets (Power et al., 1991; Wadsworth, 1997) or with
retrospective data about early economic and social conditions (Lundberg, 1993;
Rahkonen etal., 1997b). It is an approach which is not tied to any one of Blacks
explanations and would, theoretically at least, admit a role for health behaviours and
broader features of culture, material and psychosocial factors, and health selection.
Furthermore, rather than seeing these as competing factors, the emphasis in the
lifecourse perspective is to conceptualise influences as mutually reinforcing and linked
together in a chain of events and experiences which impact on health. The relationship
with social class is also seen in dynamic terms and would necessitate consideration of
health risks associated both with class of background and achieved adult class as well
as the relationship between them. Ultimately, the task is to identify particular types of
healthy and unhealthy careers or trajectories which offer the potential for intervention at
particular key points.
The lifecourse perspective is not yet sufficiently well developed to constitute a theory, but
it is possible to direct attention to a range of particularly important influences occurring at
particular stages. The first of these, following Barkers work, refers to the earliest period
of life and would focus as much on the consequences of low birthweight (e.g.
subsequent infant feeding practices) as on its causes. In respect of early childhood, the
evidence strongly suggests that the physical effects of the domestic environment (e.g.
damp housing) are of particular importance. In later childhood, the relationship between
values about health and education, as they constitute part of the socialisation of the
child, might be particularly important. A related issue in childhood and adolescence is
the way illness impacts on educational achievement. The adolescent period is also of
potentially great importance as it is at this time that health behaviours such as smoking
are developed. The research evidence bearing on this issue suggests that, though this
period may be characterised by relative equality in health, the consequences of poor
educational achievement and deviant lifestyles for subsequent health inequalities are
profound (Power et al., 1991). One such consequence in later youth is the experience of
unemployment which not only impacts on health but also on the likelihood of future
unemployment in adulthood (Montgomery, 1996). In adulthood, and perhaps particularly
among adult workers, the possibility of psychosocial effects from particular work
environments is encountered. Finally, as we proceed towards retirement and beyond, a
range of factors associated with changes in roles and social support become more and
more important. Each of these factors constitutes an influence on health in its own right,
but the particular strength of the lifecourse perspective resides in the way they may be

Page 16 - Occasional Paper September 1998

MRC Social and Public Health Sciences Unit

dynamically interrelated over time. It also suggests that it is extremely unlikely that
health inequalities are the product of a single underlying mechanism but rather that
several different mechanisms (or pathways) are implicated for different dimensions of
health (Power et al., 1997).
The lifecourse perspective has several advantages over other perspectives on health
inequalities, not least because it admits influences which in Blacks terms belong to
different and potentially competing domains. It also seems to me to be much more likely
to be solidly evidence-based than the bigger all-inclusive explanations if for no other
reason than it demands attention to detailed social processes. Even so, it is unlikely to
be the final word on the issue for just as there is evidence which doesnt fit the big
explanations, so there are findings which dont fit the lifecourse perspective. One of
these is that, on the basis of two longitudinal studies (the Twenty-07 study and the
NCDS), class gradients do not appear to widen with age. Any explanation of health
inequalities which rests on the notion of an accumulation of risk over time would not
expect this and it is a puzzle as to why it doesnt occur.

Page 17 - Occasional Paper September 1998

MRC Social and Public Health Sciences Unit

CONCLUSION
There is no doubt that health inequalities are a persistent feature of our societies, an
inequity which demands that something be done. Given the passion the issue arouses, it
is not surprising therefore that the debate has been characterised, and continues to be
characterised, by the advocacy of simple, all or nothing, solutions. The Black Report,
despite its immense contribution in putting the issue on the scientific and policy agenda,
did not in my opinion help matters in polarising explanations along the axes of selection
vs causation or individual responsibility vs material influences. This polarises solutions
just as much as it does explanations in a way that seems to me to be counter-productive.
Thus, it may be just as misguided to advocate the redistribution of income as the
solution, as either the absolute or relative poverty model would suggest, as it is to
advocate that health inequalities would disappear if working class people smoked less or
ate more healthy diets. The potential effectiveness of income redistribution as the
remedy for health inequalities, in my view, rests either on the dubious assumption that
there is a single underlying psychosocial mechanism at work or that improvements in
material standards would translate, via health behaviours or changes in values, into
better health. Both are problematic.
The alternative view stems directly from a lifecourse perspective in which particular sets
of influences at particular stages demand particular solutions. These might include
nutritional policies for pregnant women, educational policies to improve standards in
schools, social policies to reduce the problem of disengagement among young people,
economic policies to reduce unemployment, anti-discrimination policies to prevent
discrimination against disabled people, and work-place policies to reduce levels of
stress. At the present time, this multifaceted approach to the problem seems to be the
one favoured by the British Government. It will offend those who see this as piecemeal
attempts to tackle a big problem, but it is almost certainly closer to the evidence for all
that.

Page 18 - Occasional Paper September 1998

MRC Social and Public Health Sciences Unit

BIBLIOGRAPHY
Amos, A. and Hillouse, A. (1991) Tobacco Use in Scotland. Edinburgh: ASH Scotland.
Barker, D. (1992) The Fetal and Infant Origins of Adult Disease. BMJ Publications,
London.
Blane, D., Davey Smith, G. and Bartley, M. (1993) Social selection: what does it
contribute to social class differences in health? Sociology of Health and Illness, 15 (1),
1-15.
Blane, D., Bartley, M., Davey Smith, G., Filakti, H., Bethune, A. and Harding, S. (1994)
Social patterning of medical mortality in youth and early adulthood. Social Science &
Medicine 39, 361-366.
Blaxter, M. (1990) Health and Lifestyles. Tavistock/Routledge, London.
Chadwick, E. (1842) Report of an Enquiry into the Sanitary Conditions of the Labouring
Population of Great Britain. Poor Law Commission, London.
Cox, B. D., Blaxter, M., Buckie, A. L. J. et al. (1987) The Health and Lifestyle Survey.
Health Promotion Research Trust, London.
Davey Smith, G., Bartley, M. and Blane, D. (1990) The Black Report on socio-economic
inequalities 10 years on. British Medical Journal, 301, 373-377.
Davey Smith, G., Blane D. and Bartley, M. (1994) Explanations for socio-economic
differentials in mortality: evidence from Britain and elsewhere. European Journal of
Public Health, 4, 131-144.
Department of Health and Social Security (1976) Prevention and Health: Everybodys
Business. HMSO, London.
Der, G., Macintyre, S., Ford, G., Hunt, K. & West, P (1997) The relationship of household
income to a range of health measures in three age cohorts in the West of Scotland.
Society for Social Medicine Conference, York.
Dorling, D. (1997) Death in Britain: How local mortality rates have changed: 1950s1990s. Joseph Rowntree Foundation, York.
Ford, G., Ecob, R., Hunt, K., Macintyre, S. and West, P. (1994) Patterns of class
inequality throughout the lifespan: class gradients at 15, 35 and 55 in the West of
Scotland. Social Science & Medicine, 39, 1037-1050.
Glendinning, A., Hendry, L. and Shucksmith, J. (1995) Lifestyle, health and social class
in adolescence. Social Science & Medicine, 41, 235-248.
Goldblatt, P. (1990) Longitudinal Study: Mortality and Social Organisation, HMSO,
London.

Page 19 - Occasional Paper September 1998

MRC Social and Public Health Sciences Unit

Goldblatt, P. (1990) Mortality and alternative social classifications. In Longitudinal Study:

Mortality and Social Organisation, ed. P. Goldblatt, pp. 163-192. HMSO, London.
Graham, H. (1994) Gender and class as dimensions of smoking behaviour in Britain.
Insights from a survey of mothers. Social Science & Medicine, 38, 691-698.
Himsworth, H. (1984) Epidemiology, genetics and sociology.
Science, 16, 159-176.

Journal of Biosocial

Illsley, R. (1986) Occupational class, selection and the production of inequalities in

health. Quarterly Journal of Social Affairs, 2, 151-165.
Illsley, R. and Svensson, P. G. (1990) Health inequalities in Europe. Social Science &
Medicine, 31, 223-420.
Illsley, R. and Baker, D. (1997) Inequalities in Health: Adapting the theory to fit the facts.
Centre for the Analysis of Social Policy, University of Bath.
Karasek, R. A.,Theorell, T., Schwartz, J., Schnall, P., Pieper, C. and Michela, J. (1988)
Job characteristics in relation to the prevalence of myocardial infarction in the US HES
and HANES. Americal Journal of Public Health, 78, 910-18.
Kunst, A. and Mackenbach, J. (1994) International variations in the size of mortality
differences associated with occupational status. International Journal of Epidemiology,
23, 742-750.
Koupilova, I., Leon, D. A. and Vger, D. (1997) Can confounding by sociodemographic
and behavioural factors explain the association between size at birth and blood pressure
at age 50 in Sweden? Journal of Epidemiology & Community Health, 51, 14-18.
Lahelma, E., Manderbacka, K., Rahkonen, O. and Karisto, A. (1994) Comparisons of
inequalities in health: evidence from national surveys in Finland, Norway and Sweden.
Social Science & Medicine, 38, 517-524.
Lundberg,O. (1993) The impact of childhood living conditions on illness and mortality in
adulthood. Social Science & Medicine., 36, 1047-1993.
Macintyre, S. (1986) The pattering of health by social position in contemporary Britain:
directions for sociological research. Social Science & Medicine, 23, 393-415.
Macintyre, S. (1988) Social correlates of human height. Science Progress Oxford, 72,
493-510.
Macintyre, S. (1997) The Black Report and beyond: what are the issues? Social Science
& Medicine, 44, pp. 723-745.
Macintyre, S., Maciver, S and Sooman, A. (1993) Area, class and health: should we be
focusing in places or people. Journal of Social Policy, 22, 213-214.
Marmot, M., Ryff, C. D., Bumpasss, L. L., Shipley, M. and Marks, N. F. (1997) Social
inequalities in health: next questions and converging evidence. Social Science &
Medicine, 44, 901-910.

Page 20 - Occasional Paper September 1998

MRC Social and Public Health Sciences Unit

Martin, C. J., Platt, S. D. and Hunt, S. M. (1987) Housing conditions and ill health. British
Medical Journal, 294, 1125-1127.
McLoone, P. and Boddy, F. A. (1994) Deprivation and mortality in Scotland, 1981 and
1991. British Medical Journal 309, 1465-1470.
Montgomery, S. M. (1996) The Relationship of Unemployment with Health and Health
Behaviour in Young Men. PhD thesis. SSRU, City University, London.
Pearce, N., Marshall, S. and Borman, B. (1991) Undiminished social class mortality
differences in New Zealand men. New Zealand Medical Journal, 104, 153-156.
Power, C., Manor, O. and Fox, J. (1991) Health and Class: The Early Years. Chapman
and Hall, London.
Power, C., Matthews, S. and Manor, O. (1996) Inequalities in self rated health in the
1958 birth cohort: lifetime social circumstances or social mobility? British Medical
Journal, 313; 449-53.
Power, C., Hertzman, M. D., Matthews, S. and Manor, O. (1997) Social differences in
health life cycle effects between ages 23 to 33 in the 1958 birth cohort. American
Journal of Public Health, 87,1499-1503.
Rahkonen, O. and Lahelma, E. (1992) Gender, social class and illness among young
people. Social Science & Medicine, 34, 649-655.
Rahkonen, O., Arber, S. and Lahelma, E. (1997a) Health-related social mobility: a
comparison of currently employed men and women in Britain and Finland. Scandinavian
Journal of Social Medicine, 25, 83-92.
Rahkonen, O., Lahelma, E. and Huuhka, M. (1997b) Past or present? Childhood living
conditions and current socioeconomic status as determinants of adult health. Social
Science & Medicine, 44, 327-336.
Robertson, B., Uitenbroek, D., Hay, S. and Platt, S. (1997) The RUHBC-CATI Lifestyle
and Health Survey: Health Behaviours Among Adults in Glasgow 1988-1995. The
Research Unit in Health and Behavioural Change, The University of Edinburgh.
Rose, G. and Marmot, M. (1981) Social class and coronary heart disease. British Heart
Journal, 45, 13-19.
Sooman, A., Macintyre, S. and Anderson, A. (1993) Scotlands health: a more difficult
challenge for some. The price and availability of healthy foods in socially contrasting
localities in the West of Scotland. Health Bulletin, 51, 276-284.
Stevenson, T. H. C. (1928) The vital statistics of wealth and poverty. Journal of the
Royal Statistical Society, XLI, 207-230.
Strachan, D. P., Anderson, H. R., Limb, E. S., ONeill, A. and Wells, N. (1994) A national
survey of asthma prevalence, severity and treatment in Great Britain. Archives of
Disease in Childhood, 70, 174-178.
Stronks, K. (1997) Socio-economic Inequalities in Health: Individual Choice or Social
Circumstances? Ponsen & Looijen BV, Wageningen.
Page 21 - Occasional Paper September 1998

MRC Social and Public Health Sciences Unit

Sweeting, H. and West, P. (1995) Family life and health in adolescence: a role for culture
in the health inequalities debate? Social Science & Medicine, 40, 163-175.
Timms, D. W. G. (1995) Mental Health, Mental Illness and Family Background. Project
Metropolitan Research Report No 42, Department of Sociology, University of Stockholm,
Stockholm.
Townsend, P. and Davidson, N. (Eds) Inequalities in Health:
Penguin, Harmondsworth, 1982.

The Black Report,

Vger, D. and Illsley, R. (1992) Inequality, health and policy in East and West Europe.
International Journal of Health Sciences, 3, 225-239.
Vger, D. and Illsley, R. (1996) Explaining health inequalities, beyond Black and Barker:
a discussion of some issues emerging in the decade following the Black Report.
European Sociological Review, 11, 219-241.
Valkonen, T., Martelin, T., Rimpela, A., Notkola, V. and Savela, S. (1993) Socioeconomic mortality differences in Finland 1981-90. Statistics Finland, Population: 1.,
Helsinki.
Wadsworth, M. E. J. (1986) Serious illness in childhood and its association with later life
achievement. In Class and Health: Research and Longitudinal Data (Edited by Wilkinson
R. G.), pp. 50-74. Tavistock, London.
Wadsworth, M. E. J. (1997) Health inequalities in the life course perspective. Social
Science & Medicine, 44, 859-869.
West, P. (1988) Inequalities? Social class differentials in health in British youth. Social
Science & Medicine, 27, 291-296.
West, P. (1991) Rethinking the health selection explanation for health inequalities.
Social Science & Medicine, 32, 373-384.
West, P. (1997) Health inequalities in the early years: is there equalisation in youth?
Social Science & Medicine, 44, pp. 833-858.
West, P., Macintyre, S., Annandale, E. and Hunt, K. (1990) Social class and health in
youth: findings from The West of Scotland Twenty-07 Study. Social Science & Medicine,
30, 665-673.
Whitehead, M. (1987) The Health Divide. Health Education Council, London.
Wilkinson, R. G. (1986) Class and Health: Research and Longitudinal Data. Tavistock,
London.
Wilkinson, R. G. (1996) Unhealthy Societies The Afflictions of Inequality. Routledge,
London.
Williams, R. (1990) A Protestant Legacy: Attitudes to Death and Illness among Older
Aberdonians. Oxford University Press, Oxford.

Page 22 - Occasional Paper September 1998