Epidemiology would be out of breath...

Professor Arnaud Fontanet, a renowned epidemiologist... (1), which states: "After a prosperous period that could be called the glorious thirty years, from 1960 to 1990, epidemiology is beginning to run out of steam. Epidemiology provides methods to measure the health status of a population and to identify the causes of disease. It is the subtle science of risk estimation that played a key role in increasing life expectancy in the 20th century. While microbiology has made it possible to identify the agents responsible for infectious diseases - each infection has its own microbe - research is struggling with new multifactorial causes. So should we fear unannounced pandemics? What concrete contributions does this discipline still make to human health?

 

Ne live longer and healthier than generations before us. The increase in life expectancy in the 20th century, over 30 years in industrialized countries, is out of all proportion to the meagre gains of past centuries. In the United States, the decline in infectious diseases as a result of advances in hygiene saved 23 years between 1900 and 1950. The gain in life expectancy was more modest in the second half of the same century, estimated at seven years, and can be attributed equally to advances in medicine and public health. For example, it was epidemiological studies in the 1940s and 1950s that first linked lung cancer to smoking. Vaccination and antibiotics have consolidated the gains made against infectious diseases. However, as life expectancy has increased, new, slower-growing diseases have replaced infectious diseases as the leading cause of death. Myocardial infarction, stroke, cancer and soon neurodegenerative diseases are the new scourges of modern times.

 

Since the early 1990s, epidemiology has been the subject of sometimes heated discussion and debate among specialists: the windfall effect of certain studies, difficulties in adjusting and discriminating multiple factors in the case of certain chronic diseases, sophisticated statistical modeling that sometimes leaves the impression of a "black box", inability to take into account the social contexts underlying health behaviours and events, etc. The results of these discussions and debates have been very positive.

 

The advances in molecular genetics and the entry of epidemiology into the era of "big data" with access to human genome sequencing data and the arrival of huge biological data banks (UK Biobank, Million Veteran Study, etc.) have led to a renewal of the discipline and raised new hopes. Methodological approaches have continued to evolve since then. The adventure towards precision medicine and public health oriented towards prevention has therefore only just begun.

By relying on cohorts of hundreds of thousands of individuals followed for several decades, and sophisticated statistical analysis methods, epidemiology will identify the main cardiovascular risk factors, and the role of tobacco, alcohol, and several environmental exposures at the origin of many cancers. This will lead to recommendations on screening for high blood pressure, diabetes, and campaigns against tobacco and alcohol in particular, thus making a major contribution to the increase in life expectancy in the second half of the 20th century.

For Prof. Fontanet, epidemiology, through its history and the challenges it faces, remains a demanding discipline capable of evolving to further refine our understanding of the causes of disease, particularly new pandemics whose rate of emergence is estimated at five years.

 

Nevertheless, while microbiology had made it possible to identify the agents responsible for infectious diseases - each infection has its own microbe - Professor Fontanet believes that research is stumbling over the multifactorial causes of these non-communicable diseases. While experimental physiology and biology do help us understand the mechanisms associated with the development of cardiovascular diseases and cancers, the part of these diseases that is due to our genes, our behaviour or our environmental exposures remains an enigma.

Key behavioural risk factors for chronic disease have been identified, and the discipline is facing a more complex exploration of genetic or environmental determinants. Investigative and measurement tools are lacking, and the expected small effects are difficult to identify. We need only take the example of certain emerging countries where "data are collected but not analysed", according to one World survey of this 22 January 2019 on the poor management of epidemiological records, which are nonetheless valuable for public health.

 

In addition, new studies contradict previous ones and suggest that many of the published results correspond to chance associations related to random sampling fluctuations or even methodological biases inherent in the discipline. In this uncertain context, advances in genomics at the beginning of the 21st century are opening up new perspectives: analysis of the human genome in its entirety will be able to be integrated into predictive models of disease.

More generally, access to new sources of massive data of multiple origins, the "big data", is becoming possible: electronic medical records, social networks, connected objects, sensors of atmospheric pollutants will contribute to the volume of information available to the researcher. Thanks to these enormous amounts of medical data concerning patients, their history and their environment, this new abundance of data is not without regulatory, ethical and analytical difficulties and constraints; and like many other disciplines, it will call upon artificial intelligence to make up for the inadequacies of existing analytical methods. New types of cohorts are being set up, such as the UK Biobank, with 500,000 participants recruited in 2006, all of whom have benefited from complete genome sequencing and the availability of their medical data. The expected outcomes are those of precision medicine (treatment adapted to the individual), extended to prevention, where everyone will receive prevention and screening advice based on their individual risk of being affected by this or that disease according to their genome. Many questions remain, however, in addition to the complexity of the analyses to be carried out, the first of which concerns the willingness of individuals to change their behaviour on the basis of the results of genetic predisposition analyses.

 

More fundamentally, these individualistic approaches move away from the universal recommendations advocated by public health. And the persistence of pronounced social disparities in terms of life expectancy shows that much remains to be done to address inequalities in terms of risk behaviours, environmental exposures, and access to care, without the need for sophisticated individualistic approaches.

 

As for infectious diseases, they were painfully brought to mind with the sudden onslaught of AIDS in the early 1980s. Bird flu, mad cow disease, SARS, chikungunya, Ebola, Zika... so many health crises caused by so-called emerging infectious agents, threatening the whole planet.

Most of these new diseases, mostly linked to viruses from the animal world, spread within a few weeks in a densely populated and increasingly mobile world. This very high level of connectivity, which is our vulnerability, is also our strength: once an alert is issued, we are able to respond quickly on the spot in a concerted manner, and share information, samples and data. In the long term, "big data" approaches should even enable us to share in real time the pathogen sequences carried out locally and thus reconstruct the chains of transmission for better control of epidemics.

 

These encouraging prospects should not make us forget that many of these infectious emergencies are now taking place in countries with dilapidated health structures, or where armed conflicts make intervention very difficult. Let us remember that it took four months from the first cases to the identification of the Ebola virus during the major epidemic of 2013-2014 in West Africa, and that today the response to the ongoing epidemic in the Democratic Republic of Congo is facing armed opposition from local militias. Controlling the risk of infection on a global scale will first require strengthening the surveillance and response capacities of the poorest countries.

 

Arnaud FONTANET, Professor of the CNAM and the Pasteur Institute, has been invited to occupy the Public Health Chair. of the Collège de France (year 2018-2019).  Inaugural lesson on Thursday, January 31, 2019 at 6:00 p.m. (Program).

 

(1) A former intern at the Paris Hospitals, Doctor of Medicine (University of Paris V) and Doctor of Public Health (Harvard University), Professor Fontanet specialises in the epidemiology of infectious and tropical diseases. His main research themes are viral hepatitis and emerging viruses, which has led him to lead several projects on behalf of the ANRS (National Agency for Research on AIDS and Viral Hepatitis) and the ANR (National Research Agency) : HEPNILE project on viral hepatitis in Egypt, study of microcephalies linked to the Zika virus in Africa and Asia, acute encephalitis in children in North Vietnam, the SARS epidemic, MERS, and many other conditions related either to emerging diseases or to neglected diseases (Buruli ulcer).

Arnaud FONTANET is also the director of the Pasteur-Cnam School of Public Health which he founded, and of the Pasteur Institute's Global Health Centre. Having pursued a career combining field action, research and teaching, he has designed many courses, including the MOOC "Concepts and Methods in Epidemiology" (France digital university). He also works in epidemiology, biostatistics and public health at the Institut Pasteur, the Cnam, Sciences-Po, Sorbonne-University, the Ecole Normale Supérieure, etc.

 

To go further : 

 

– New recommendations from Public Health France of 22 January 2019 on diet, physical activity and physical inactivity, based on expertise in epidemiology, prevention and health promotion.