What is epidemiology? Changing definitions of epidemiology 1978-2017

Context Epidemiology is a discipline which has evolved with the changes taking place in society and the emergence of new diseases and new discipline related to epidemiology. With these evolutions, it is important to understand epidemiology and to analyse the evolution of content of definitions of epidemiology. Objectives The main objective of this paper was to identify new definitions of epidemiology available since 1978. Secondary objectives were to analyse the content of these definitions, to compare them with those used by Lilienfeld and to determine whether changes have taken place over the last forty years. Methods A review of grey literature and published literature was conducted to find the definitions of epidemiology written between 1978 and 2017. Results 102 definitions of epidemiology were retained. They helped to highlight 20 terms and concepts related to epidemiology. Most of them were already used in the definitions used by Lilienfeld. Five terms were present in more than 50% of definitions from the period 1978 to 2017: “population”, “study”, “disease”, “health” and “distribution”. Several developments have occurred: strengthening of the terms “control” and “health” already used, the concept of “disease” was less frequently encountered whereas the concepts “infectious diseases”, “mass phenomenon” are no longer used in definitions from 1978 to 2017. Conclusion This evolution of content of definition of epidemiology is absent from books on epidemiology. A thematic analysis of definitions of epidemiology could be conducted in order to improve our understanding of changes observed.


Introduction
Epidemiology is a recent discipline which has evolved with the changes taking place in society and the emergence of new diseases. This evolution has allowed epidemiology to remain a useful and relevant tool in bringing to light and understanding diseases and health events. Since its origins, more than a century ago, many definitions of epidemiology have been suggested.
In 1978, Lilienfeld published a seminal paper on definitions of epidemiology. For him, there was no consensus among epidemiologists as to the definition of epidemiology. The objective of his work was to provide a single and understandable definition of epidemiology that was suitable for all types of diseases and populations. Lilienfeld based his work on 23 existing definitions of epidemiology and proposed the following definition [1]: "Epidemiology is a method of reasoning about disease that deals with biological inference derived from observations of disease phenomena in population groups".

Analysis of the content of the new definitions
The content of new definitions of epidemiology was analyzed to identify the terms and concepts present in these definitions. Once a term or concept was identified in at least two definitions, it was added to the list of terms.

Analysis of the evolution of definition content between 1978 and 2017
In 1979, Evans [3] analyzed the content of the definitions used by Lilienfeld to build his definition of epidemiology. In summary, he identified 23 different terms and concepts. For each term and concept, the frequency of occurrence in the definitions was calculated. Then, they were grouped into eight categories: status of the person (disease, infectious disease, physiologic conditions, injuries, health); group affected (populations, community, mass phenomena, outbreak); distribution of disease; spread (spread, propagation, dynamics); incidence, occurrence; etiology (causes, determinant factors, circumstances of occurrence, ecology); understanding disease (natural history or nature, understanding the process); prevention and control.
To study the evolution of the content of definitions of epidemiology over time, the content of the definitions used by Lilienfeld was compared with new content from the period 1978-2017.
For each term and concept identified by Evans, the frequency of appearance in the definitions from the period 1978-2014 was estimated.
For each term and concept identified using the definitions from 1978 to 2017, the frequency of appearance in the definitions used by Lilienfeld was estimated.
Then for each term and concept, whatever the period, the frequency of appearance in the definitions used by Lilienfeld and definitions from 1978 to 2017 were compared.
All statistical comparisons were performed using bayes factors [12] with R software [13] (package BayesFactor). We have chosen bayes factor instead of p value, which is used in the context of null-hypothesis significance testing (NHST). Indeed, a nonsignificant p-value does not quantify evidence in favour of the null hypothesis. As stated by the American Statistical Association, "a relatively large p-value does not imply evidence in favor of the null hypothesis" [14].

Literature search
The search of published and grey literature revealed 102 definitions of epidemiology: 93 for human medicine, including subspecialties of epidemiology (n = 24) and 9 for veterinary medicine. We have selected 29 definitions of epidemiology from websites. The definitions of epidemiology were found in different media: articles, epidemiology online courses and excerpts from books. A total of 69 definitions from 1978 to 2017 were selected ( Table 1).
Analysis of the content of definitions from the period 1978 to 2017

Human medicine, general epidemiology
A total of 20 terms and concepts were identified in definitions from 1978 to 2017. Of the 20 terms and concepts, only three were not present in the definitions used by Lilienfeld: "knowledge" appeared in 1979 [3], "problems" appeared in 1991 [15] and "public health" appeared in 1999 [16] Table 2 shows the frequencies of appearance of terms in definitions from 1978 to 2017. The term "population" or "group" was found in 84% of the definitions (58/69). For 25% of definitions (17/69), it was stated that epidemiology applies to human beings.
Analysis of the evolution of definition content between 1978 and 2017 Table 3 shows a comparison of the content of definitions reviewed by Lilienfeld and those from 1978 to 2017 according to the concepts identified by Evans.
The term "health" was more frequent in definitions from 1978 to 2017 than in the definitions used by Lilienfeld (58% versus 4%, Bayes Factor(BF) = 18709.9).
The term "infectious disease" was less frequent in definitions from 1978 to 2017 than in the definitions used by Lilienfeld, with a BF estimated to 65.9 (0% versus 17%). "Mass phenomena/outbreak" and "infectious disease" can be related. The BF associated with "mass phenomena/outbreak, was estimated at 18, and this also reflects a decrease in the use of the term "infectious disease". The study of the distribution and determinants of health-related states or events in specified populations, and the application of this study to control of health problems

Timmreck TC [110]
2002 Epidemiology is an investigative method used to detect the cause or source of diseases, disorders, syndromes, conditions, or perils that cause pain, injury, illness, disability, or death in human populations or groups. . . . Epidemiology has been defined in several ways. One definition is the study of the nature, cause, control, and determinants of the frequency and distribution of disease, disability, and death in human populations.
Friedman G et al. [111] 2003 Epidemiology is the study of disease occurrence in human populations. The primary units of concern are groups of persons, not separate individuals.
(Continued ) At the beginning of the twenty-first century, epidemiology is a broad-based population science, drawing on many disciplines from biology and sociology to biostatistics and philosophy of science, which investigates the causes of human disease and methods for their control

Boslaugh S [53]
2008 Epidemiology is the study of frequency and determinants of morbidity and mortality in populations

Bhopal RS [17]
2008 Epidemiology is the science and craft that studies the pattern of disease (and health, though usually indirectly) in populations to help understand both their causes and the burden they impose. This information is applied to prevent, control or manage the problems under study

Porta M [42]
2008 The study of the occurrence and distribution of health-related states or events in specified populations, including the study of the determinants influencing such states, and the application of this knowledge to control health problems

Gordis L [114]
2009 Epidemiology is the study of how disease is distributed in populations and the factors that influence or determine this distribution Olsen J et al. [115] 2010 Epidemiology is defined by the object of research, "to identify determinants that change the occurrence of health phenomena in human populations."

Saracci R [21]
2010 "Epidemiology is the study of health and disease in populations". The population aspect is the distinctive trait of epidemiology, while health and disease are investigated at other levels as well.
Webb P et al. [35] 2010 Epidemiology . . . is about measuring health, identifying the causes of ill-health, and intervening to improve health. . . Perhaps epidemiology's most fundamental role is to provide a logic and structure for the analyses of health problems both great and small Carneiro I et al. [51] 2011 Epidemiology is the study of the distribution and determinants of health states or events in specified populations, and the application of this study to control of health problems (Adapted from Porta, 2008). Quoted reference: [116] Fos PJ [44] 2011 Epidemiology is the study of the distribution and determinants of health-related states and events in defined populations and the application of this study to the control of health problems (Adapted from Last, 2001) Quoted reference: [43] The Greek root of epidemiology and the two definitions have a common theme, namely, the people. The people are considered as a group, which is referred to as a population. This population-centered nature of epidemiology leads to one of the differences between public health services and medical services. . . . Epidemiology is the study of factors affecting the health and illness of populations.
Krickeberg K et al. [117] 2011 . . .epidemiology is about illness in populations, not in individual subjects. More precisely, it is concerned with the frequency of diseases in various parts of the population in which we are interested, the so-called "target population".

Rothman KJ [18]
2012 Often considered the core science of public health, epidemiology involves "the study of the distribution and determinants of disease frequency", or put even more simply "the study of the occurrence of illness Ward H et al. [22] 2012 Epidemiology (is) the study of the distribution and determinants of health and illness in populations.
(Continued ) Macera CA et al. [46] 2013 Epidemiology is the study of a scientific method of problem solving that helps "disease detectives" understand how people get sick and die, who gets sick and die, and how to avoid getting sick. The current definition of epidemiology is the study of the distribution (who has the problem) and determinants (things that influence the problem) of health-related conditions in human populations and the application of this method to the control of health problem. 2015 Epidemiology is "The study of the occurrence and distribution of health-related events, states, and processes in specified populations, including the study of the determinants influencing such processes, and the application of this knowledge to control relevant health problems." (Adapted from Lilienfeld, 1980). Quoted reference: [103] Wassertheil-Smoller S et al. [23] 2015 Epidemiology may be defined as the study of the distribution of health and disease in groups of people and the study of the factors that influence this distribution. Modern epidemiology also encompasses the evaluation of diagnostic and therapeutic modalities and the delivery of health-care services.

Bhopal RS [50]
2016 . . .epidemiology will be defined here as the science and practice that describes and explain disease patterns in populations. It uses this knowledge to prevent and control disease, and improve health. The central idea of epidemiology is that patterns of ill health and disease in population may be analyzed systematically to understand their causes and to improve health. The term "population" was more frequent in definitions from 1978 to 2017 than in the definitions used by Lilienfeld (84% versus 48%, Bayes (BF) = 63.1). Table 4 shows a comparison of the content of definitions used by Lilienfeld and those from 1978 to 2017 according to the terms and concepts we have identified.
For the term "population" and "health", the results shown in Table 3 and Table 4 are identical.
The term "problems" was probably more frequent in definitions from 1978 to 2017 than in the definitions used by Lilienfeld (29% versus 0%, BF = 43.8).
The concepts mentioned in the definitions of subspecialty of epidemiology are similar to those found for general epidemiology. Some authors did not provided general definition of epidemiology [54][55][56][57][58], nor specific definition to subspecialty [59][60][61].
Others took up the main concepts of epidemiology and adapted them to their subspecialty. In this respect, the definition of the epidemiology of obesity speaks for itself: it contains virtually all concepts: "causes and consequences"; "human populations"; "the distributions, patterns, and dynamics"; "the determinants"; "to prevent and control"; "health conditions". This definition also mentions a specificity of the subspecialty: "the development and validation of body composition measurement methods used in epidemiologic studies". Others referred to authoritative epidemiological literature or to WHO. For this case, the definitions of Last, 1995 and2001 [43, 62] were most often cited (6 times out of 24 definitions), WHO (1994), Rothman and Greenland, 2008 [24], Susser, 1973 [63] and Wikipedia being quoted once. Different subspecialties referred to the definition of Last: oncology [64], psychiatry [65], genetic epidemiology [66,67], reproduction [68].
Finally, others Other authors proposed a pure and simple application of the general definitions of epidemiology to their subspecialty [64,65]. Thus, Dos Santos after recalling Last's definition of epidemiology proposed a definition of cancer epidemiology and stated: "Therefore, this definition is as valid to cancer epidemiology as it is to epidemiology in general" [64]. Tsuang defined psychiatric epidemiology as follow "Psychiatric epidemiology is simply the epidemiology of psychiatric disorders-no more, no less [65]. Some authors provided definitions by highlighting differences. This is the case, for example, for genetic epidemiology. Austin stated "Genetic epidemiology differs from epidemiology by its explicit consideration of genetic factors and family resemblance" [58]. Veterinary medicine Table 6 shows the frequencies of appearance of terms in definitions of epidemiology from 1977 to 2015, for veterinary medicine. A total of 9 definitions were selected Again, the concepts mentioned in the definitions of veterinary epidemiology are similar to those found for human epidemiology, with several authors using definitions proposed in human medicine and quoting for example MacMahon and Plug [69].
One notable concept, specific to veterinary medicine, is productivity which is mentioned by two authors [70][71][72].

General epidemiology (websites)
The definitions of epidemiology are given in Table 7 and are classified according to the nature of the website: websites of international organizations, WHO, CDC, FAO (for veterinary medicine); national health institutes, academies of sciences . . .; public health schools; medical journals (BMJ); nonprofit organizations; Wikipedia; For-profit organizations; non-medical dictionaries online.
The distribution of concepts, by order of frequency, is the following: health (12/30); population (18/30); disease / illness (27/30); study (19/30); distribution (16/30). This distribution is almost that found with the definitions of epidemiology provided by books. What is epidemiology? Changing definitions of epidemiology 1978-2017  [18,24,26,135] The field of nutritional epidemiology has developed from interest in the concept that aspects of diet may influence the occurrence of human disease.
(Continued ) Genetic epidemiology. Is the study of the role of genes and environments on markers of health and disease risk in population.

Giesecke J [137]
Infectious diseases 2017 Most standard definitions of epidemiology mention something like "the study of distribution and determinants of diseases in the population". . . My own practical view is that epidemiology is about putting people into groups.
. . .Epidemiology identifies such groups, ignoring the uniqueness of its members, and tries to discover whether this division of people into groups tells us something more than we could have learnt by just observing each person separately.
(Continued ) In a book compiled by Last the definition of epidemiology provided is "the study of the distribution and determinants of health-related states or events in specified populations, and the application of this study to control for health problems." Quoting Last, 1995 [62] . . . epidemiology involves more than just the study of disease states (e.g. respiratory illness), but also includes the study of events (e.g., injury) and behaviors and conditions associated with health (e.g., hand washing). In addition, note that epidemiology is concerned with healthrelated states or events that occur in populations, not a specific individual Environmental epidemiologists study the frequency and pattern of disease and health-related events and attempt to explain the environmental factors that influence these conditions. The study of why and how environmental factors affect peoples' health is the essence of environmental epidemiology. Environmental epidemiology emphasizes the idea that health is largely influenced by environmental factors, and that by identifying these factors and their modes of transmission, the public's health can be better protected.

2008
Environmental epidemiology is a subspecialty of epidemiology, the basic science of public health. Epidemiology is the study of the distribution of health and disease in the population, and of the determinants of this distribution. Environmental epidemiology studies the effects of environmental exposures on health and disease in the population. The subject matter of environmental epidemiology is environmental health: this, in principle, covers all factors external to the human body which may affect health. However, . . . environmental epidemiology does not try to cover all external factors which may conceivably influence population health, but focuses on physical, chemical and (noninfectious) biological factors in our everyday environment

Venables K [141]
Occupational epidemiology 2013 Occupational epidemiology is the study of the distribution and determinants of illness and injury related to the work environment. . . . Furthermore, harmful work exposures can be conceptualized as a "natural experiment" in the causation of illness and injury, because the workplace usually has much better-defined levels and timing of exposure than in the general community.

O'Campo P et al. [142]
Social epidemiology 2011 Epidemiology is the study of the patterns of health and illness in populations, while social epidemiology focuses on the social determinants that shape the risk and occurrence of poor health in these populations. Quoting Berkman and Kawachi, 2000 [143] and James, 2009 [144].

Berkman LF et al. [145]
Social epidemiology 2014 Epidemiology is the study of the distribution and determinants of states of health in population. Quoting Susser, 1973 [63]. We define social epidemiology as the branch of epidemiology that studies the social distribution and social determinants of states of health.
. . .We focus on specific social phenomena such as socioeconomic stratification, social networks, discrimination, workplace organization, and public policies rather than on specific disease outcomes.
The majority of websites offer definitions of epidemiology already known, but does not cite the authors.
Public health schools do not provided the source of their definition of epidemiology except one (University of Alabama) with Last, 2001 [43] and MacMahon, 1970 [69] being quoted.
Few websites mentioned the date of the last update (e.g. National Institute on Deafness and Other Communication Disorders).

Discussion
This study allowed us to identify, synthesize and analyse the definitions of epidemiology from the period 1978-2017. This evolution is almost absent from books on epidemiology and deserved to be presented. Even though many other fields of epidemiology have grown, only definitions of epidemiology were studied. A significant number of definitions of epidemiology were found despite the short period of study and the inclusion criteria. However, this number must be related to the increase in the number of publications in epidemiology.
Many of the books consulted and the definitions of epidemiology they contained summarized definitions that were already known [74] or definitions of other authors [75] or proposed mnemonics tools "the 3-D definition of epidemiology" [76,77]. The goal of epidemiology, very broadly speaking, is to understand the patterns of disease and health dynamics in populations as well as the causes of these patterns, and to use this understanding to mitigate and prevent disease, and to promote health. The goal of digital epidemiology is exactly the same. . .Digital epidemiology is epidemiology that uses data that was generated outside the public health system, i.e. with data that was not generated with the primary purpose of doing epidemiology.
Epidemiology, in general, is "the science of occurrence of diseases in human populations. Disease occurrence is In summary, 69 definitions of epidemiology were retained from 1978 to 2017. All of these came from epidemiology textbooks or published articles. Only one definition (WHO) was not associated with a date of appearance [32]. The fact that this definition was not cited by Lilienfeld was enough to keep it in this study.  [73] 1977 The study of the health status of populations" (Schwabe et al., 1977). Quoted reference: [73] Martin SW et al. [146] 1987 Epidemiology may be defined as the study of the patterns of disease that exist under field conditions. More specifically, epidemiology is the study of the frequency, distribution, and determinants of health and disease in populations.

Toma B [147, 148] 1999
The study of the health status of populations.

Slater MR [149]
2003 What is epidemiology? Most simply, it is the study of diseases and health in populations. Diseases include not only the classic infectious diseases and disease outbreaks but also chronic diseases like arthritis, cancer, and renal failure. Also included are injuries and exposures to environmental contaminants. Factors that maintain good health and quality of life are also within the purview of epidemiology. In this context, questions about the causes of animals being relinquished to animal shelters or about animal well-being in long-term kennel settings may also be addressed by epidemiologists.

Smith RD [147] 2005
Over the years there have been many definitions of epidemiology. Some definitions follow: 1. "The study of the distribution and determinants of disease frequency in man" (MacMahon and Pugh, 1970).

"
The study of the patterns of disease" (Halpin, 1975).

Pfeiffer D [71]
2009 . . .the science of veterinary epidemiology deals with the investigation of these determinants of disease distribution in animal populations. Productivity and welfare of animals may also be outcomes of interest since disease will usually impact on both, and they may indeed often be what stakeholders will have as their primary focus.

Dohoo I et al. [70] 2009
Epidemiology is largely concerned with disease prevention and therefore, with the "succession of events which result in the exposure of specific types of individual to specific types of environment" (ie exposures) (MacMahon & Pugh, 1970). Thus, epidemiologists strive to identify these exposures and evaluate their associations with various outcomes of interest (eg health, welfare, productivity) so as to improve the lives of animals and their keepers. . . . associations which are likely to be causal in nature and which, once identified, we can take advantage of to improve the health, welfare and productivity of animals and the quality and safety of foods derived from them. Quoted reference: [69] Thrusfield M [72] 2013 Epidemiology is the study of disease in populations and of factors that determine its occurrence; the key word being populations. Veterinary epidemiology additionally includes investigation and assessment of others health-related events, notably productivity.

Sergeant et al. [151]
2015 Epidemiology is the study of patterns and causes of disease in populations.
Understanding these issues will in turn contribute to identification of options for control and prevention of diseases. As its simplest, epidemiology is about supporting better decision making to ensure appropriate response or preventative measure for population health.
https://doi.org/10.1371/journal.pone.0208442.t006 Several operational definitions of epidemiology have been suggested. The simplest is that epidemiology is the study of the occurrence of disease or other health-related characteristics in human and in animal populations.
Epidemiologists study not only the frequency of disease, but whether the frequency differs across groups of people; i.e., they study the cause-effect relationship between exposure and illness. Diseases do not occur at random; they have causes-quite often man-made causes-which are avoidable. . . Many definitions have been proposed; here are two that capture the underlying principles and the public health spirit of epidemiology: "Epidemiology is the study of the distribution and determinants of healthrelated states or events in specified populations, and the application of this study to the control of health problems." (2). "Epidemiology is the study of the distribution and determinants of disease frequency in man." (3).

(Continued )
(2) Last JM, 1995. Epidemiology is the study of the distribution and determinants of disease frequency in human populations. Epidemiology focuses on making comparisons in order to establish cause-effect relationships, evaluate information, and make good decisions that will improve outcomes. The definition of epidemiology is "the study of disease in populations and of factors that determine its occurrence over time." The purpose is to describe and identify opportunities for intervention. Epidemiology is concerned with the distribution and determinants of health and disease, morbidity, injury, disability, and mortality in populations. For veterinary epidemiology, this intervention is to enhance not only health but also productivity. The evolution of the content of definitions of epidemiology was studied on the basis of terms and concepts identified by Evans [3] and those identified by us.

Productivity is taken into account
The definitions from 1978 to 2017 helped to highlight 20 terms and concepts related to epidemiology. A great majority of concepts were either identified by Evans [3] or present in definitions used by Lilienfeld [1]. Among the terms identified, only three ("problems", "knowledge" and "public health") were not present in Evan's work. Five terms were present in more than 50% of definitions from 1978 to 2017: "population", "study", "disease", "health" and "distribution". These five terms can summarize epidemiology as being the study of the distribution of disease and health in the population. The term "study" may not seem specific enough for epidemiology. But this term is linked to the concept of "science". Epidemiology was considered a "science" during both periods. In fact, the term "science" was used in more than 15% of definitions for the two periods. Several developments have occurred. (i) the emergence of the term "problem" in the definitions of epidemiology; (ii) strengthening of the terms "control" and "health" already used; (iii) other terms and concepts ("infectious diseases", "mass phenomenon") identified by Evans [3] were no longer used in definitions from 1978 to 2017. But these terms were rare in the definitions used by Lilienfeld: 4/23 (17%), 1/23 (4%), 3/23 (13%) and 3/23 (13%), respectively.
The prevalence and distribution of diseases, infectious or not, also influence epidemiology. Thus, in low-income countries, new infectious diseases have emerged such as HIV infection. Nevertheless, the prevalence of these infectious diseases has declined over time. In low-income countries, infectious diseases are still present [78], but there is an increase in the incidence of chronic diseases [79]. Epidemiological definitions have evolved in this direction, with a decrease in the presence of infectious disease terms.
Indeed, for definitions from 1978 to 2017, only one definition associated the terms "epidemiology" and "epidemic" [80]. Currently, the term "epidemic" is associated with both infectious disease and the growing prevalence and incidence of disease states in the population, for example, obesity (obesity epidemic) [81]. Thus, the concept of "disease" used in epidemiology is no longer limited to infectious diseases.
Epidemiology is currently associated with the study of disease, but also more broadly, with the study of health phenomena. In 23% (16/69) of the definitions retrieved, the concepts of "disease" and the concept of "health" were related. The term "disease" has a rather negative connotation, while the term "health" has a positive connotation. They reflect two opposing views of the same phenomenon. Some definitions suggested this phenomenon was dynamic and not static and evolved in the population, in space and in time [3,33,38,82,83].
In the definitions from 1978 to 2017, the link between epidemiology and public health was rare [16,18,20,44,84]. However, these definitions describe epidemiology as an essential part of public health. Other developments, nonetheless, were related to public health, like the expansion of the field of epidemiology to areas such as health promotion [18] and the use of statistical tools [3,30,32,52].
Finally, 69 definitions (general epidemiology) were analyzed. This is greater than the number of definitions used by Lilienfeld. The fact that the definitions retained came from different sources, both grey and published literature, suggests that the definitions retained are representative of the available definitions of epidemiology, thus limiting the risk of selection bias. The evolution of the content of definitions of epidemiology was studied by using the concepts identified by Evans [3] and those identified by us. The comparison with the concepts cited by Evans allowed an objective analysis of the evolution of the content of definitions of epidemiology from 1978 to 2017. The statistical analysis revealed the differences observed between the two sets of definitions of epidemiology, even if the definitions of epidemiology from 1978 to 2017 were not exhaustively from this period.
The analysis of the evolution of the definitions of human epidemiology by subspecialty and veterinary epidemiology was not possible, due to the low number of definitions.
The number of subspecialties of epidemiology increases from year to year and almost all fields of medicine are concerned ( [85]). As early as 2007, Lillienfeld already raised the question of the general epidemiologist [86]. As for general epidemiology, they are several definitions for each subspecialty. Austin [58], quoting Khoury's book published in 1993 [87], mentions that there were at least eight different definitions of genetic epidemiology at that time.
This increase in the number of definitions of epidemiological subspecialties can be linked to several factors including technological advances. The significant development of data analysis in epidemiology (epidemiological methods) could be related to the emergence of informatics. Molecular and genetic epidemiology is also undoubtedly linked to major advances in molecular and genomic biology [88]. Advances in genome sequencing techniques have led to a better understanding of the genetic determinants of disease occurrence. They have therefore promoted the emergence of genetic epidemiology as a subspecialty of epidemiology [89], with a specific definition [90].
The prevalence of some cancers has also changed over time. Epidemiology has made it possible to identify some risk factors related to the environment and lifestyles. Nevertheless, the etiology of many types of cancer is still poorly understood. Indeed, it is generally accepted that most cancers result from the combined effects of environmental and genetic factors and that only a few cancers are only "of genetic origin" [91]. The integration of molecular techniques into epidemiological studies has also led to the emergence of molecular epidemiology [92], with a definition proposed by Schulte et al. in 2012 [57].
Ecology and environment are a concern of human populations. The same applies to their impact on health. The emergence of environmental epidemiology is a particular reflection of this. The consideration of environment in the definition of epidemiology appeared in the 2000s, with the definition of Gerstman [30], which includes the term "modern" (with reference to old definitions). In previous years, only Cole, 1979 [82] refers to the environment in his definition of epidemiology.
The digital revolution, linked to the rapid and unprecedented increase in the availability of data from various digital sources, has created new opportunities for collecting and analyzing data produced outside the health system. This led Salathé et al. [93] to introduce the term "digital epidemiology" and to propose a definition in 2018 [94]. Some tools such as Google Trends are available and their use has led to mixed results [95].
The use of medical devices, whether implanted or not, has become crucial in many medical specialties for the diagnosis and treatment of certain diseases. Although their use is frequently beneficial to the patient, side effects may occur sometimes (e. g. infection on a vascular catheter). Monitoring these side effects and studying the risk factors for their occurrence is a necessity that has led to the individualization of the epidemiology of medical devices, with a specific definition [96,97]. The methodological aspects specific to this field of epidemiology (definition of "real" exposure; choice of short-or long-term assessment criteria; various biases) were reported by Jalbert et al. [98].
Faced with the numerous specialties and the numerous books about on the topics, we had to make a selection. We recall that the main objective was general epidemiology, the description of the definitions of subspecialties of epidemiology being a complement.
The concepts found in the definitions of subspecialties of human epidemiology and those of veterinary epidemiology are generally the same. Consideration of economic aspects (productivity) is a specific feature of veterinary epidemiology.
The concepts found in the definitions of human epidemiology available on the websites are generally the same as those contained in the books or articles. Nevertheless, bibliographic references are often unavailable and access of information may change over time, with the disappearance of pages [99] or the existence of dead links ("error 404". . .). The posting date is also rarely available.
Publications about the definitions of epidemiology exist. The fields and the aim are for example: introductory epidemiology textbooks [10]; description of definitions epidemiology over 50 years [9]; epidemiology and the methods needed for public health assessments [100]; evolution of epidemiological methods and concepts [11]. But our review examines the evolution of definitions of epidemiology, taking into account more than 100 definitions of epidemiology retrieved from books or articles, introductory texts or not, general epidemiology or subspecialties, and a non-temporal description and analysis of the definitions of veterinary medicine. We also describe 29 definitions of epidemiology retrieved from websites.
This study has several limitations. Only English definitions were retained. National definitions of epidemiology not translated into English were eliminated [101]. These definitions may have contained terms and concept related to epidemiology that were different from those in the definitions retained. But the probability that we have missed an important concept seems quite low. The terms and concepts identified depend on the content of the definitions selected. The definitions of epidemiology available in online dictionaries and other media were easy to find but the quality of the definitions was heterogeneous. To determine the quality of definitions, quality criteria need to be used. For the definitions from 1978 to 2017, the quality criteria may seems subjective. Moreover, no weights were assigned to definitions. The choice of weighting criteria may be subjective.
The purpose of the study was to identify concepts related to epidemiology and not to prioritize them. Given the variety of definitions, the fact that the method used to identify terms and concepts was easy to use and the fact that the majority of terms and concepts were present in more than 10% of definitions suggests that for the definitions retained all of the terms and concepts related to epidemiology were identified.
For future studies, we propose some weighting criteria such as inverse-variance weighting (meta-analysis. . .) which are widely accepted. We also propose that some of the following criteria be taken into account in the weighting grid: the number of concepts in the definition; the book type (collective dictionary as those of Last/ Porta; advanced texts; introductory texts).

Conclusions
In summary, this work led to a synthesis of different concepts related to epidemiology proposed during the period 1978-2017 and highlighted the evolution of the content of definitions of epidemiology over time. Most of the terms and concepts identified by us had already been used in the definitions of Lilienfeld while several terms and concepts identified by Evans were no longer used in definitions from 1978 to 2017. Increased usage of the terms "control" and "health" was found in definitions of epidemiology from 1978 to 2017. A thematic analysis of definitions of epidemiology could be conducted to complete this study, in order to improve our understanding of the changes observed.