International Journal of Epidemiology, 2016, 1727–1732

doi: 10.1093/ije/dyw263
Advance Access Publication Date: 8 December 2016
Commentary

Commentary

Epidemiology, the International Epidemiological

Association and the International Journal of
Epidemiology: a personal chronicle
Rodolfo Saracci*
*Corresponding author. Senior Visiting Scientist, International Agency for Research on Cancer, 150 Cours A.Thomas,
69008 Lyon (France). E-mail: [email protected]
Accepted 11 August 2016

Two main dates stand out in the more than 40-year life of Corresponding Club’ on a small informal basis to facilitate
the International Journal of Epidemiology (IJE): 1972, the international communication between physicians working
year the journal was established by the International in university departments mostly- but not only-of prevent-
Epidemiological Association (IEA) and 2000, the entry ive and social medicine. A first formal meeting of what had
into the new century coinciding, as it happened, with the already been transformed into the ‘International
first year in office of the Editors-in-Chief, George Davey Epidemiological Association’ (IEA) took place in London
Smith and Shah Ebrahim, both of whom will terminate in 1956.
their tenure at the end of 2016. It is convenient to take In 1957 a seminal book was published, Uses of
these dates as signposts for two periods in the IJE history. Epidemiology, by Jerry Morris.1,2 On the premise, true
today as 60 years ago, that ‘the epidemiological method is
the only way of asking some questions in medicine, one
The first 30 years way of asking others, and no way at all to ask many’,
IEA launches IJE Morris nonetheless portrayed a wide scope for epidemiolo-
gical research, ranging from completing the clinical picture
The population approach to health and disease was con-
through aetiological studies to preventive and therapeutic
fronted at the emergence from the World War II with rad-
trials and health services evaluations. The year 1960 saw
ically new challenges. In industrialized countries, where
the publication of Epidemiologic Methods, by Brian
scientific research was essentially located, non-
MacMahon, Thomas Pugh and Johannes Ipsen,3 the first
communicable diseases of obscure aetiology were gaining
systematic introduction to the methods of epidemiological
the lead in mortality and morbidity, overtaking communic-
investigation especially for non-communicable diseases. It
able diseases, the traditional field of investigation of epi-
defined epidemiology as ‘the study of the distribution and
demiology. Two research fellows travelling in each other’s
determinants of disease prevalence in man’. In 1965 the
countries, John Pemberton of the UK, active in public
American Journal of Hygiene, founded in 1921, changed
health, and Harold Willard from the USA, primarily active
its name to the American Journal of Epidemiology, and the
as a clinician, had noted the handicapping lack of informa-
editor Neal Nathanson wrote:
tion on current research in epidemiology, particularly on
the enigmatic non-communicable diseases, among the vari- There is no journal in the English language which has the
ous medical schools and research institutions. To remedy word epidemiology in its title. Epidemiology is both a
this they established in 1954 an ‘International method and a substantive field, and epidemiologic papers

C The Author 2016; all rights reserved. Published by Oxford University Press on behalf of the International Epidemiological Association
V 1727
1728 International Journal of Epidemiology, 2016, Vol. 45, No. 6

appear in a wide variety of medical journals. This is as it lucidly developed the theoretical and practical aspects of
should be; nevertheless, it is believed that a journal this distinction (which echoes the one between idiographic
devoted to epidemiology will fill a distinct need.4 and nomothetic research in the philosophy of science) in
It was a growing need that led in 1972 to the establish- the context of clinical trials. It clearly has vast cur-
ment by the IEA Council of the ‘yellow cover’ (unchanged rent implications for outcome research and what, in
until today) IJE. Driving this decision was the vision and techno-commercial jargon, may be labelled as value-based
energy of Walter Holland, who became the first IJE health care.
Editor. Introducing the journal he wrote: Of long-lasting interest is a paper by Lester Breslow9
For many years the term ‘epidemiology’ has referred to
which explores ways of giving concrete substance to the
quite specifically, the study of epidemics, their prophy- World Health Organization’s well-known definition of
laxis and control. During recent years, however, the health.
definition has broadened to include not only the study Perhaps easy to overlook is an interesting 13-line letter
of acute infectious diseases but also of non- to the Editor10 in which Richard Doll, after expressing the
communicable diseases, both physical and mental. hope that IJE becomes a world leader, recommends that it
Although this definition may not yet be universally ac- reverts from the practice of numbered references to ‘the
cepted, it will be used in this Journal.5 standard British technique as in the BMJ’, with references
This faithfully reflected the status of the art and lexicon at indicated by the author(s) name(s). He continues:
the time, with the addition, in Holland’s words: With numbers it is disturbing to have to go continu-
ously to the end of the article to find out needed names
Epidemiological techniques are now being used increas-
ingly in the study of health services and in the attempt . . . [because] ‘to judge the significance of a reference,
one so often needs to know who the authorities referred
to produce a more rational deployment of available
medical resources. The Journal will provide an oppor- to were’ (emphasis added).
tunity for an international comparison of the various
approaches to these problems.
IJE develops successfully
Although in the first issue of the journal seven out of 13 A systematic reading and reviewing of papers published by
articles concerned communicable diseases in the whole of the IJE from 1972 to 2015 would allow one to build a
Volume 1 (1972) they represented fifth (12 out of 60) of fully-fledged picture of the evolution of epidemiology, the
the total articles, a clear reflection of the changing em- journal and their relationship.
phasis of epidemiological research. Six Chief Editors have successfully run and developed
Some articles in Volume 1 dealt with topics that have the journal during its 45 years of life: W. W. Holland
had far-reaching implications. A paper by Donald (1972–77), A. E. Bennett (1978-81), C. du V. Florey
Henderson,6 the mastermind of the WHO Smallpox (1982-90), P.O.D.Pharoah (1991-99) and G. Davey Smith
Eradication Programme, stressed how the principal on- and S. Ebrahim (2000 to the present).
going progress in the declining incidence of the disease The first three decades saw a steady annual increase of
occurred because of a better understanding of the epidemi- 4% in the number of published papers and pages: from 58
ology of smallpox and the application of surveillance pro- papers and 400 pages in 1972 to close to 200 papers and
cedures to interrupt transmission two principles of 1200 pages in 1999 and 2000. The proportion of papers
permanent validity as evidenced by the recent Ebola and devoted to communicable disease tended to decrease to
Zika outbreaks. He predicted that complete world eradica- 15% or less (for instance there were only four such papers
tion was in sight and in fact the last natural case of small- in 1982) until the mid 1980s, a time in which in the eco-
pox, outside laboratory accidents, was recorded 5 years nomically developed countries they seemed to be mostly a
later in 1977 in Somalia, for the first time putting an end residual of the past. Yet in 1981, the first cases were re-
to a disease. ported of what would be soon recognized as a new disease,
A second short paper by Daniel Schwartz7 highlighted AIDS, induced by infection with the HIV virus. The first
in one specific area (evaluation of preventive measures in epidemiological paper on AIDS, by R. Detels et al.11 ap-
high-risk subjects), the fundamental distinction between peared in the journal in 1987. It marked the return of com-
explanatory and pragmatic intervention studies. The first municable diseases to the front of epidemiological
seeks generalizable answers to questions of treatment effi- research: throughout the late 1980s and the 1990s, the pro-
cacy, the second addresses questions of treatment effective- portion of papers devoted to communicable diseases
ness under naturally occurring field conditions, variable in shifted back to what it had been in the first year of the
space and time. Daniel Schwartz and Joseph Lellouch8 had journal, about a quarter of the total. A second important
International Journal of Epidemiology, 2016, Vol. 45, No. 6 1729

feature appeared at the end of the 1980s: the publication World Congress of Epidemiology) and the appointment
of supplements, initiated by C. du V. Florey in 1988 and of the new IJE Editors-in-Chief.
sustained by P. O. D. Pharoah. Eleven supplements were
published between 1988 and 1997, on topics ranging from
international trends in coronary heart disease mortality12
The IEA XV International Scientific Meeting:
to effects on the fetus and early child development of alco-
Florence, 31 August to 4 September 1999
hol consumption by the pregnant mother,13 to infant and The philosophy that inspired the XV International
child mortality in Africa.14 Scientific Meeting ‘Epidemiology for sustainable health’, is
It was the IJE that published in 1985 one of those rare outlined in the presentation I wrote at the time:
papers that leave a mark on the history of ideas: ‘Sick indi- The world’s health and its sustainability in time and in
viduals and sick populations’ by Geoffrey Rose,15 later ex- different countries which epidemiologists are called to
panded into a book The Strategy of Preventive Medicine.16 investigate are the results of forces biological, social,
Even in our ‘omics’ era, it remains as an inescapable term economic, scientific, technical and medical, the full ap-
of reference against which to test the value of old and new preciation of which demands a multiplicity of view-
approaches. If epidemiology and the IJE were flourishing, points, often located outside the competence of the epi-
biology was advancing in giant steps. In 1975, the demiological profession.
‘Southern blot’ technique had for the first time allowed the The structure of the meeting reflected this philosophy,
measurement of gene variants, including in human DNA with plenary lectures by distinguished scientists active in
specimens; and in 1983, the PCR (polymerase chain reac- fields bordering epidemiology: demography, economics,
tion) rendered the measurement more simple and speedier ethics, molecular biology, genetics, nutrition, global envir-
for large-scale use. ‘Something’ brand new was definitely onment and climatic changes, work and societal structures,
approaching not only biomedicine but epidemiology as as well as sessions devoted to key epidemiological themes.
well. John Cairns, an outstanding cell and molecular biolo- With a total of 1200 participants including fellowships for
gist, visited several times the International Agency for attendees from developing countries, 300 oral communica-
Research on Cancer in Lyon, where I was in charge of tions, 500 posters and an $8000 surplus in the account bal-
Analytical Epidemiology. On one occasion I told him: ance, the meeting proved a complete success.
I understand too little of these developments but I am
certain they will very soon change epidemiological re-
The IJE Editor-in-Chief appointment
search. Would you agree to plan and direct a course on
‘Molecular biology for epidemiologists’? I can act as a The appointment of new editors for the IJE was decided at
sounding board.17 an IEA Council session during the Florence meeting. The
John Cairns agreed and in 1986 a 2-week course, the position, most effectively filled for a decade by P.O., had
first ever to combine modern basic biology with epidemi- been advertised and six applications received, all of a qual-
ology, proved highly attractive to epidemiologists, many of ity deserving full consideration. A review committee, com-
whom soon after started laboratory-based field projects, posed of the President, President-Elect, Secretary, a
particularly in the domains of cancer and nutritional Regional Councillor and the current IJE Editor-in-Chief,
epidemiology. set up six evaluation criteria: status as epidemiologist, edi-
torial experience, international experience, support sour-
ces, availability of time for the IJE and work programme.
Changing century and Editor-in-Chief The scoring of these criteria was discussed at length within
After 3 years as President-Elect, I took office in 1996 as the review committee and within the Council, to which the
President of the IEA for the triennium 1996-99. What I selection of the Editor-in-Chief belonged. One of the six
write reflects what was done at the time: what was not applicants was actually a duo: George Davey Smith and
done, including what should have been done, by defin- Shah Ebrahim. Although Davey Smith and Ebrahim
ition never existed! Eleven meetings took place during ranked first in the scoring, the IEA Constitution stipulates
that period across the IEA Regions. There was a steady ‘Editor-in-Chief’ in the singular. This formal obstacle com-
improvement in the budgetary balance, John Last super- bined with their obvious and intensive engagement in re-
vised the fourth edition of the Dictionary of search, which could limit their time available to the IJE,
Epidemiology18 and the census of IEA membership was became a matter of extended debate. For me, the fact that
updated. However, in my perception the two most im- they were fully active researchers, and their ages (early to
portant and closely related events of that era were the late 40s) that would allow mid- to long-term developments
XV International Scientific Meeting (now renamed for the IJE, were key reasons for supporting their
1730 International Journal of Epidemiology, 2016, Vol. 45, No. 6

Table 1. Total number and percent distribution of published papers by selected year and type

Paper type 1972 1982 1992 2002 2012 2015

Original 68 (38) 63 (55) 74 (154) 34 (104) 38 (95) 38 (101)

Comment 14 (8) 8 (7) 1 (3) 32 (97) 24 (60) 18 (48)
Letter 4 (2) 6 (5) 10 (20) 7 (20) 8 (20) 10 (27)
Profiles 0 0 0 0 17 (42) 20 (53)
Other 14 (8) 23 (20) 15 (31) 27 (81) 13 (32) 14 (37)
Total 100 (56) 100 (87) 100 (208) 100 (302) 100 (249) 100 (267)

Figures in brackets are numbers of each paper type.

Original: original article
Comment: commentary, editorial, editor’s choice, reflections, reviews
Letter: letter of comment or short report letter
Profiles: profile of cohort, data resources or HDSS (Health and Demographic Surveillance System) components
Other: all other types

candidature. In addition, their research appeared to be The annual meeting of all Editors and editorial staff has
driven by similar motivations opening epidemiology to ad- been an essential component of a collaboration in which
vances from neighbouring fields of research that had all Editors felt closely involved in shaping the IJE journey
inspired the Florence meeting. Eventually the Council guided by George and Shah. In our era of telephone and
members converged and George Davey Smith and Shah videoconferences, there is no substitute for face-to-face
Ebrahim were appointed as Editors-in-Chief. meetings to create a spirit that motivates busy people to de-
vote time and effort to a common purpose.
Table 1 shows how the proportion of papers by cat-
Into the new century: 2000 to 2016 egory has changed since 2002, increasing room being given
Table 1 shows the total number of papers published by IJE to editorials, commentaries and reprints of important
for selected years and the distribution by broad types. papers that have otherwise fallen into oblivion. These were
After a steady increase during the first 30 years, the num- instrumental to the way the journal contents were treated
ber has levelled off since the turn of the century at 250-300 and deployed. A 2001 editorial19 set the tone: ‘In the case
papers per year. However, in the same period the number of the latest millennial transition, the coincidence of the
of pages has increased substantially from about 1200 to calendar shift with the announcement of the human gen-
about 2000 pages. The IJE impact factor in 1995 was 1.32 ome provides what we believe will be seen as a more than
and the journal ranked 20th out of 60 within the group symbolic change in the way (wo)man views wo(man)’.
Public Health, Occupational and Environmental Health. This anthropological, long-term perspective leads to
By the year 2000 the impact factor had risen to 1.89, with themes being presented in their historical unfolding like, to
the journal ranked 22nd out of 87, and in 2014 the two in- paraphrase Winslow (20), ‘chapters in the history of ideas’.
dicators peaked respectively at 9.17 and 2nd out of 162. In For example, the recurrent theme of early life influences on
1995 about 67% of submitted papers were rejected, a pro- adult health was introduced in a reprint of a 1966 paper by
portion which has reached 85-90% in recent years. Dubos et al.21 on ‘Biological Freudianism’, which in turn
If these crude indicators tell a story of remarkable suc- bridges to the panoply of conceptual antecedents in
cess, the substance of it lies in the approach to IJE develop- Freud’s ideas. A similar historical approach has been fol-
ment and its contents by Davey Smith and Ebrahim, lowed for three other main content themes: methods in epi-
‘George & Shah’. Their approach has in fact been to gather demiology, particularly meta-analysis,22,23 socioeconomic
a gradually expanding group of Associate Editors (later inequalities in health24,25 and Mendelian randomization,26
called simply Editors) in a friendly and personalized way, the new method proposed by the editors in their capacity
while establishing an efficient editorial office in Bristol. as original researchers as one possible way of tackling the
Assistant Editor Barbara Coatesworth was until 2013 the thorny problem of confounding in observational studies.
pillar of this. Efficient and highly reliable, she was suc- Epigenetics has expanded rapidly in the past decade as has
ceeded by Edwina Thorn, who developed the role and es- its potential for epidemiology27 : which was dealt with by
tablished a vital presence for the IJE in the social media several papers in the first issue of 2012 and again taken up
and is now ably supported by Audrey Hayes. Since 2011 this year in a stimulating historical perspective that in-
editorial strength has been added by the appointment of an cludes an unpublished 1889 paper by Francis Galton28 on
Academic Editor, Jane Ferrie. the possibility of hereditary transmission of acquired traits.
International Journal of Epidemiology, 2016, Vol. 45, No. 6 1731

Side by side with these principal themes, each issue of the results of their scientific research will remain marginal
journal includes empirical papers on a variety of other in impacting on health.
topics, debates, symposia, photo essays and profiles. The Tomorrow’s epidemiology will go where young epi-
latter, edited by D. Lawlor and later J. Ferrie, provide de- demiologists choose to take it; the future is theirs. But
tailed descriptions of existing cohorts, databases and we can leave them Geoffrey Rose’s essential message
health and demographic surveillance site data available to that scientifically ‘an epidemiologist must have dirty
epidemiologists as potential research resources. hands and a clean mind’,31 and socially ‘we are all re-
sponsible to all for all’.16 That is the essence and the
worth of our profession.
A hope
Conflict of interest: None declared.
One concluding hope for the future of epidemiology
that the IJE will reflect and be able to influence is the
link, as fundamental and obvious on paper as it is frail Acknowledgment
and irregular in the real world, between epidemiology Audrey Hayes kindly abstracted the data summarized in Table 1.
and public health. The first component of a strong link
is epidemiological research, scientifically at the front References
line and at the same time closely relevant to major
1. Morris JN. Uses of Epidemiology. Edinburgh, UK: Livingstone,
public health issues. Very often this implies prompting
1957.
or harnessing, or both, technological advances. For ex- 2. Morris JN. Uses of epidemiology. Br Med J 1955:395–401.
ample, the firm identification of the HPV viruses as the Reprinted Int J Epidemiol 2007;36:1165–72.
causative agents of cervical cancer, a major scourge in 3. MacMahon B, Pugh TF, Ipsen J. Epidemiologic Methods.
developing countries, became possible thanks to the Boston, MA: Little Brown and Co., 1960.
technological development of viral DNA testing meth- 4. Nathanson N. Change in name. Am J Epidemiol 1965;81:1.
ods.29 This is a clear-cut success but, in an era of fast 5. Holland WW. Editorial. Int J Epidemiol 1972;1:3.
6. Henderson DA. Epidemiology in the global eradication of small-
and vast technological developments, offering tools
pox. Int J Epidemiol 1972;1:25–30.
from personal monitors to gene editing and ‘aug- 7. Schwartz D. Evaluation of preventive measures in high risk sub-
mented’ body functions (with trans-humans in the dis- jects. Int J Epidemiol 1972;1:51–53.
tance), the research priorities for epidemiology relevant 8. Schwartz D, Lellouch J. Explanatory and pragmatic attitudes in
to public health become much more problematic. Not therapeutic trials. J Chronic Dis 1967;20:637–48.
every technology-driven project can nor should be 9. Breslow L. A quantitative approach to the World Health
done. If public health is to be what the name implies, Organization definition of health: physical, mental and social
well-being. Int J Epidemiol 1972;1:347–55.
health for all, it must include overcoming large inequal-
10. Doll R. Letter to the Editor: References – alphabetical or numer-
ities between and within countries, and proposed re-
ical? Int J Epidemiol 1972;1:410.
search projects need to be carefully selected on the 11. Detels R, Visscher BR, Fahey J, Sever J, Gravell M, Madden DL
basis of how well they address this overarching goal. et al. Predictors of clinical AIDS in young homosexual men in a
This leads directly to the second component of a strong high-risk area. Int J Epidemiol 1987;16:271–76.
link between epidemiology and public health, the in- 12. Various Authors. Trends and determinants of coronary heart dis-
volvement of epidemiologists in decision-making proc- ease mortality: international comparisons. Int J Epidemiol
esses. I have brought with me from my clinical 1989;18(Suppl 1):1–232.
13. Various Authors. Euromac. European Concerted Action: mater-
academic years the firm conviction that a patient is
nal alcohol consumption and its relation to the outcome of preg-
first a patient and second a subject or in today’s lan-
nancy and child development at 18 months. Int J Epidemiol
guage a participant of research. In a similar way, I be- 1992;21(Suppl 1):1–87.
lieve that an epidemiologist should have for the 14. Various Authors. Monitoring child survival programmes in
population, the subject of research, the same active Africa: the Africa Child Survival Initiative. Int J Epidemiol
concern materializing in actions beneficial to the health 1993;22(Suppl 1):1–80.
of all people that a clinician researcher must have for 15. Rose G. Sick individuals and sick populations. Int J Epidemiol
the individual patient. The purist view, that epidemiolo- 1985;14:32–38.
16. Rose G. The Strategy of Preventive Medicine. Oxford, UK:
gists do research and leave to others the decisional
Oxford University Press, 1992.
processes on health, is both logically inconsistent (we 17. Saracci R, Wild C. IARC: the First Fifty Years. Lyon, France:
are always asking for funds in the name of health) and International Agency for Research on Cancer, 2015.
ethically debatable.30 To the extent that epidemiologists 18. Last J. A Dictionary of Epidemiology. 4th edn. Oxford, UK:
remain marginal to the decision-making process, the Oxford University Press, 2001.
1732 International Journal of Epidemiology, 2016, Vol. 45, No. 6

19. Davey Smith G, Ebrahim S. Epidemiology: is it time to call it a 26. Davey Smith G, Ebrahim S. Mendelian randomization: can
day? Int J Epidemiol 2001;30:1–11. genetic epidemiology contribute to understanding environmental
20. Winslow CEA. The Conquest of Epidemic Disease. A Chapter in determinants of disease? Int J Epidemiol 2003;32:1–22.
the History of Ideas. Princeton, NJ: Princeton University Press, 27. Ebrahim S. Epigenetics: the next big thing. Int J Epidemiol
1943. 2012;41:1–3.
21. Dubos R, Savage D, Schaeder R. Biological Freudianism: lasting 28. Galton F. Feasible experiments on the possibility of transmitting
effects of early environmental influences. Pediatrics 1966;38: acquired habits by means of inheritance. Int J Epidemiol
789–800. 2016;45:13–15.
22. Egger M, Ebrahim S, Davey Smith G. Where now for meta-ana- 29. Munoz N, Bosch X, de Sanjosé S et al. Epidemiological classifica-
lysis? Int J Epidemiol 2002;31:1–5. tion of human papillomavirus types associated with cervical can-
23. Various authors. Special theme: systematic reviews and meta- cer. N Engl J Med 2003;348:518–27.
analysis. Int J Epidemiol 2002;31:59–153. 30. Saracci R. Epidemiology: a science for justice in health. Int J
24. Various authors. Special theme: socioeconomic position. Int J Epidemiol 2007;36:265–68.
Epidemiol 2003;32:346–429. 31. Rose G, quoted in Hall AJ. Ecological studies and debate
25. Various authors. Special theme: inequality. Int J Epidemiol on rotavirus vaccine and intussusception. Lancet 2001;35:
2004;33:262–327. 1197–98.