J R Soc Med 2003;96:384-388
doi:10.1258/jrsm.96.8.384
© 2003 Royal Society of Medicine
Association between course of study at university and cause-specific mortality
Peter McCarron PhD MFPHM
Mona Okasha PhD 1
James McEwen FFPHM FRCP 2
George Davey Smith MD FFPHM 1
Department of Epidemiology and Public Health, Queen's University
Belfast
1 Department of Social Medicine, University of Bristol
2 Department of Public Health, University of Glasgow, UK
Correspondence to: Dr Peter McCarron, Unit of Descriptive Epidemiology,
International Agency for Research on Cancer, 150 Cours Albert-Thomas, 69372
Lyon Cedex 08, France E-mail:
peter.mccarron{at}qub.ac.uk
 |
SUMMARY
|
|---|
Although socioeconomic position is clearly related to mortality
and one
measure of this is length of education, it is not known
whether the choice of
course at university determines future
health. We therefore investigated the
association between faculty
of study and all-cause and cause-specific
mortality in a prospective
follow-up of male students who underwent health
examinations
while attending Glasgow University from 1948 to 1968.
Among the 9887 (84%) alumni traced by means of the NHS Central Register,
8367 (85%) had full data on important potential confounding variables; 939 of
these men had died. Physiological variables differed little between students
from the various faculties. Medical students were most likely to come from
affluent social backgrounds and, after law students, were most likely to be
smokers. Compared with former medical students, former arts and law students
had excess all-cause and cardiovascular disease mortality, while science and
engineering alumni had similar risks. Former medical students had lower lung
cancer mortality than other alumni but higher mortality from alcohol-related
causes including accidents, suicide and violence.
The lower mortality risks observed among former medical and engineering
students may be due to their better employment prospects and healthier
lifestyle behaviours, although the high mortality from alcohol-related causes
among former medical students underscores the complexity of choice of health
behaviour. The findings point to the potential for disease prevention among
the large proportion of the population who now have third-level education.
|
INTRODUCTION
|
|---|
The inverse association between adult socioeconomic position
and mortality
is well
established
1-4
and is reflected in measures
including length of
education.
5,6
For example, members of the
medical profession have lower overall mortality
than that in
the general
population,
7 despite
higher rates of death from
suicide and accidents. There has been little work,
however,
on whether the chosen course of study at university might influence
later
mortality. Using data collected on former Glasgow University
students we
have demonstrated important associations between
several factors measured at
the beginning of adulthood and later
health.
8-12
Availability
of data on faculty of study allowed us to test the hypothesis
that
mortality is lower among former medical students than in students
who
pursued other degrees.
|
METHODS
|
|---|
Full details of the study methods are reported
elsewhere.
13
Briefly,
students attending Glasgow University between 1948 and 1968
were
invited to participate in a medical examination conducted
by the student
health department. Information collected included
blood pressure, height and
weight, sociodemographic data and
details of health behaviours. Faculty of
study was recorded
as arts, divinity, and law (including chartered
accountancy),
and science, engineering, and medicine (including dentistry
and
veterinary science). Scottish universities traditionally
admit students to
faculties rather than specific courses and
it was not possible to further
distinguish students in the law
and medical faculties. Participants were
traced through the
NHS Central Register, which provides details of date and
cause
of death. Deaths up to 30 June 2000 were eligible for inclusion.
Statistical analysis
For each faculty the age-adjusted means of height (cm), body mass index
(kg/m2), and systolic blood pressure (mmHg), and the proportions of
students who were non-smokers, did not consume alcohol, and whose fathers were
from social classes I and II were calculated. Analysis of variance and
2 tests were used to test for heterogeneity across faculty for
continuous and categorical variables, respectively. Cox proportional hazard
models were used to estimate the hazard ratio associated with being a former
medical student, compared with being a student enrolled in a different
faculty, for the following causes of death: all-causes; cardiovascular disease
(CVD); lung cancer; all other cancers; respiratory disease; suicide, accidents
and violence; and alcohol-related causes. Models were adjusted for the
following potential confounding variables: father's social class (I to IV/V);
body mass index (kg/m2); cigarettes smoked per day (none, 1-10,
11-20, 20+) and also for year of birth quintile to account for possible cohort
effects. For all-cause and CVD mortality we adjusted also for systolic blood
pressure. Because of the relatively small number of deaths among females, the
analyses are restricted to males. All analyses were performed by use of STATA
(version
7.0).14
|
RESULTS
|
|---|
11 755 male students—50% of the male student
population—participated
in the original examinations from 1948 to 1968.
Participants
were representative of the total male student body in terms
of
the proportion of males taking part and faculty of
study,
13 and the
high percentage of students from affluent backgrounds
is in line with that of
men attending university during the
period of the
surveys.
15 The 9887
(84.1%) men who have been
traced are closely similar to all participants in
terms of father's
social class, height and blood pressure. The sample was also
representative
in terms of faculties: 23.1% medicine (original 23.4%), 23.4%
arts
(24.0%), 1.5% divinity (0.9%), 16.2% engineering (15.6%), 10.5%
law
(10.0%) and 26% science (25.4%). After exclusion of those
aged over 30 years
at examination, those who had left the UK
at an undetermined date, and those
with missing data on confounding
variables, 8367 (84.6%) individuals were
available for inclusion
in the analyses. Their mean age at the time of
examination was
20.5 years (range 16-29 years) and the median follow-up was
41.3
years. Since age-adjusted results were similar for the full
cohort and
for the 8367 individuals with data on confounding
variables only the latter
are reported here.
Characteristics of traced students according to faculty are shown in
Table 1. Divinity students were
older than other students and had the lowest age-adjusted mean systolic blood
pressure. There was little difference in mean systolic blood pressure among
students from other faculties. Differences in mean height and BMI across all
faculties were small, although arts students were on average shorter than
other students. Medical students had the highest proportion of fathers in
social classes I and II, followed by law students; arts and divinity students
had the lowest proportion of fathers in these social classes. Medical students
were also most likely to have a father who was a doctor (7.8%), followed by
law (1.8%) and science (1.4%) students. Science students were most likely to
be non-smokers, with law students and then medical students having the largest
proportions of smokers. The question on alcohol consumption was incompletely
answered: students of the science and law faculties were the most forthcoming,
with 91.2% and 90.0% response rates, while students of the medical faculty
were the most reticent (84.9% answering). Of those who did respond, divinity
students were least likely to consume alcohol while only one-third of law
students said they did not imbibe.
There were 939 deaths among former students who had data on confounders
(Table 2). Compared with
medical faculty alumni, former arts and law students had excess all-cause and
cardiovascular mortality, divinity students had excess all-cause mortality,
while science and engineering alumni had similar risks. Former arts students
had over twice the lung cancer and respiratory disease mortality of medical
alumni but half the risk of death from accidents, suicide and violence. There
was some evidence that law and science students, also, had greater risk of
death from lung cancer than their medical faculty peers. The risk of
alcohol-related deaths was higher in former medical students than in alumni
from any other faculty. Controlling for potential confounders made little
difference to these findings.
|
DISCUSSION
|
|---|
We have shown that course of study at university is associated
with
mortality risk and that those who studied science subjects
had substantially
lower risk of mortality from several causes
than former arts students. Medical
alumni had low all-cause,
CVD and lung cancer mortality but were at greater
risk of death
from alcohol-related causes.
The advantages of this study include its large size, long follow-up and the
availability of data on several important confounding variables. However, the
small number of female deaths precluded determination of whether choice of
course is also relevant to female mortality risk. It is possible that there
was some misreporting of smoking and alcohol habits and we have no data on
lifetime smoking. However, we have previously shown that, in this cohort,
mortality from smoking-related disease was increased for individuals who
reported smoking in adolescence and early adult
life.16 Although
those who took part in the original surveys were broadly representative of the
total male student body of the time, they nevertheless represent only 50% of
that overall population. This would bias the results only if there were
systematic differences in the two groups. This is unlikely in view of the
similarity for variables which were available for participants and
non-participants, and also the young age at the time of the health
examinations, when a negligible number of students would have been expected to
have chronic illnesses. The trace rate was high and the very close
similarities between traced and untraced men constitute strong evidence that
the results we present are generalizable to the whole student community.
There are several possible explanations for the findings. Early life
factors, including childhood socioeconomic position, may be important.
Individuals with affluent childhood circumstances are known to have lower
mortality than those from more deprived
backgrounds.17 In
the current study students in the medical faculty had the highest proportion
of fathers in social classes I and II followed by those in law and
engineering, while divinity and arts students were more likely to have
experienced socioeconomic deprivation in childhood. To some extent the
distribution of height reflects this, with taller students in engineering,
medicine, and law, although divinity students were also tall. However,
controlling for confounding variables, including father's social class, had
little effect on the findings (father's social class is a poor marker of
earlylife socioeconomic circumstances).
The possibility that intergenerational factors have a role should also be
considered. In a study examining the substantial excess of premature deaths
among Scottish men with patrilineal Irish descent, this higher mortality could
not be explained by established risk factors, and it was suggested that
intergenerational transmission of mortality risk may be
important.18 In the
current study we could find no evidence of intergenerational effects of a
medical career, since men whose fathers were doctors did not have lower
mortality than those whose fathers had other occupations (data not shown).
Lower mortality from all causes and from CVD among students of
science—and more specifically among medical and engineering
students—may also reflect the importance of factors such as employment
and income, since for medical and engineering graduates full, permanent, and
well-paid employment may be more likely than for former arts students.
Although it could be hypothesized that former law faculty students should
likewise be at low mortality risk as a result of enhanced employment
prospects, these graduates are less likely than medical and engineering
students to pursue the subject in later life. We have previously shown that
socioeconomic position over the lifecourse is associated with cause-specific
mortality.19 The
findings from the current study suggest that, even among individuals who would
have had largely affluent childhood social circumstances and who went to
university at a time when the majority of the population left school before 16
years, adult employment has effects on later health additional to those due to
early life exposures.
Health behaviours are also likely to be influential. Blaxter, using data
from the Health and Lifestyle Survey, argued that harmful behavioural habits
such as smoking have a greater impact on health in the non-manual than in the
manual social
classes.20 However,
formal testing of this hypothesis demonstrated that the health impact of
smoking is similar across socioeconomic
groups.21 It is
plausible that the low mortality risk we observed among former medical
students could be due to continuation or adoption of positive health
behaviours in later life. Although students in the medical faculty smoked more
than all other students with the exception of law faculty students, the
research by Doll and
Peto22 indicates
that, among students who smoked, medical students may have been particularly
likely to quit and therefore experience the greatest health benefit. If the
high prevalence of smoking among law faculty students persisted more after
university, this might account for their excess mortality risk compared with
medical alumni. The finding that former medical students had the highest
mortality from alcohol-related causes illustrates the complexity of health
behaviour. We can only speculate as to whether their lower overall mortality
is in part a reflection of the positive effects of alcohol.
Further work is needed to improve understanding of the mechanisms for the
associations reported here. Nevertheless, with up to 40% of 18-year-olds
currently attending higher
education,23 and
with arts and social science students being the most likely to indulge in
cigarette
smoking,24 our
results suggest that successful strategies to stop this habit would contribute
greatly to disease prevention. For the present we can conclude that, during
the period of over 40 years after university entrance, doctors are at lower
risk of death than their peers—some compensation, perhaps, for the
reported unhappiness in their
profession.25
|
Acknowledgments
|
|---|
This work was funded in part by Chest Heart and Stroke (Scotland),
Stroke
Association, NHS Management Executive, Cardiovascular
Disease and Stroke
Research and Development Initiative.
|
REFERENCES
|
|---|
- Department of Health and Social Security. Inequalities
in Health: Report of a Working Group (the Black Report). London:
DHSS, 1980
- Whitehead M. The health divide. In: Townsend P, ed.
Inequalities in Health: The Black Report and the Health
Divide. Harmondsworth: Penguin, 1992
- Drever F, Whitehead M, eds. Health Inequalities.
Decennial Supplement, ONS Series DS No. 15. London: Stationery
Office, 1997
- Shaw M, Dorling D, Gordon D, Davey Smith G. The Widening
Gap. Bristol: Policy Press, 1999
- Kunst AE, Mackenbach JP. The size of mortality differences
associated with educational level in nine industrialized countries.
Am J Publ Health1994; 84:932
-7[Abstract/Free Full Text]
- Davey Smith G, Hart C, Hole D, et al. Education and
occupational social class: which is the more important indicator of mortality
risk? J Epidemiol Commun Health1998; 52:153
-60[Abstract]
- Drever F. Occupational Health 10, Decennial
Supplements, ONS Series DS No. 10. London: HMSO,1995
- McCarron P, Okasha M, McEwen J, Davey Smith G. Blood pressure in
early life and cardiovascular disease mortality. Arch Intern
Med 2002;162:610
-11[Free Full Text]
- McCarron P, Okasha M, McEwen J, Davey Smith G. Height in young
adulthood and risk of death from cardiorespiratory disease: a prospective
study of male former students of Glasgow University, Scotland. Am J
Epidemiol 2002;155:683
-7[Abstract/Free Full Text]
- Okasha M, McCarron P, McEwen J, Davey Smith G. Height and cancer
mortality: results from the Glasgow University study cohort. Publ
Health 2000;114:451
-5
- Okasha M, McCarron P, McEwen J, Davey Smith G. Body mass index in
young adulthood and cancer mortality: a retrospective cohort study.
J Epidemiol Commun Health2002; 56:780
-4[Abstract/Free Full Text]
- Davey Smith G, McCarron P, Okasha M, McEwen J. Social circumstances
in childhood and cardiovascular disease mortality: prospective observational
study of Glasgow University students. J Epidemiol Commun
Health 2001;55:340
-1[Free Full Text]
- McCarron P, Davey Smith G, Okasha M, McEwen J. Life course exposure
and later disease: a follow-up study based on medical examinations carried out
in Glasgow University (1948-68). Publ Health1999; 113:265
-71
- Stata Statistical Software Release 7.0.
College Station, Texas: Stata Corporation, 2000
- Higher Education (Robbins Report). London:
HMSO, 1963
- McCarron P, Davey Smith G, Okasha M, McEwen J. Smoking in
adolescence and young adulthood and mortality in later life: prospective
observational study. J Epidemiol Commun Health2001; 55:334
-5[Free Full Text]
- Davey Smith G, Hart C, Blane D, Hole D. Adverse socioeconomic
conditions in childhood and cause specific adult mortality: prospective
observational study. BMJ1998; 316:1631
-5[Abstract/Free Full Text]
- Abbotts J, Williams R, Davey Smith G. Association of medical,
physiological, behavioural and socio-economic factors with elevated mortality
in men of Irish heritage in West Scotland. J Publ Health
Med 1999;21:46
-54[Abstract/Free Full Text]
- Davey Smith G, Hart C, Blane D, Gillis C, Hawthorne V. Lifetime
socioeconomic position and mortality: prospective observational study.
BMJ1997; 314:547
-52[Abstract/Free Full Text]
- Blaxter M. Health and Lifestyle. London:
Tavistock/Routledge, 1990
- Marang-van de Mheen PJ, Davey Smith G, Hart CL. The health impact
of smoking in manual and non-manual social class men and women: a test of the
Blaxter hypothesis. Soc Sci Med1999; 48:1851
-6
- Doll R, Peto R, Wheatley K, Gray R, Sutherland I. Mortality in
relation to smoking: 40 years' observations on male British doctors.
BMJ1994; 309:901
-11[Abstract/Free Full Text]
- Office for National Statistics. UK 2003—The
Official Yearbook of the United Kingdom of Great Britain and Northern
Ireland. London: Stationery Office, 2002
- Webb E, Ashton H, Kelly P, Kamali F. Patterns of alcohol
consumption, smoking and illicit drug use in British university students:
interfaculty comparisons. Drug Alcohol Depend1997; 47:145
-53[Medline]
- Smith R. Why are doctors so unhappy? There are probably many
causes, some of them deep. BMJ2001; 322:1073
-4[Free Full Text]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
TOP
SUMMARY
INTRODUCTION
