Constances is a "general purpose" population-based epidemiological cohort.
It is a nationally representative sample of
200,000 adults aged between 18 and 69 (inclusive).


Due to its population size, the quality and diversity of data, and its monitoring methods, Constances is a unique epidemiological research tool. Constances, is a research platform broadly accessible to the scientific community that can be compared to the largest international cohorts.

Constances is also a public health tool, designed to support the public health objectives of the French National Health Insurance Fund for Employees (CNAMTS) and of the national government, owing to the collection of highly diverse data from multiple sources on a representative sample.

Constances is an epidemiological surveillance tool, implemented through a partnership with the French institute for public health surveillance. Its data covers multiple domains, such as the epidemiological surveillance of occupational hazards.

The Constances project, managed through the participation of French local health insurance funds and health clinics, is a partnership between INSERM, Versailles Saint Quentin University (UVSQ), the French national health insurance fund (CNAMTS), the French national retirement pension fund (Cnav) and the support of the French Ministry of health (Directorate general for health). Constances has received French government funding for an 8-year period (Investment for the Future Program).



Researchers in 10 countries have begun a major new project studying the link between socioeconomic status and healthy ageing. The Lifepath project, coordinated by Professor Paolo Vineis from the School of Public Health at Imperial College London, involves 15 institutions in Europe, the US and Australia. It has been supported by a six million euro grant from the EU’s Horizon 2020 programme. It is mainly based on the analysis of data collected in population-based cohorts in several European countries. Constances is one of them

Many previous studies have shown that health outcomes in later life are strongly influenced by wealth and social status. Traditional risk factors such as diet and smoking account for a fraction of this link. Other factors such as psychosocial stress, working conditions and genetic influences may be important, but the mechanisms involved are not well understood. We know that it’s possible to remain healthy into old age, and this is much more likely among people with high socioeconomic status. The research consortium aims to better understand why, making healthy ageing a reality for everyone, not just the wealthiest parts of society. The consortium, which includes scientists, economists and policy experts, hope to amass evidence to inform efforts to reduce health inequality.


The four-year research programme will investigate links between socioeconomic factors, such as education, income and achievements, and age-related health outcomes, such as cancer, heart disease, cognitive impairments and frailty. It will also use molecular techniques to analyse blood samples from existing population studies to uncover biological changes associated with poverty.

Conducting a research project in Constances

French or foreign research teams wishing to make use of the Constances cohort infrastructure must submit an application. Projects may only use the available data or biological materials and/or collect additional data for a specific purpose. Cohort data access applications are submitted in the context of a permanent Call for Proposals (CFPs). Applicants are invited to draft a scientific protocol of their research project. The projects are then examined by the Constances International Scientific Committee and, where applicable, by its Ethics Review Board; authorisations are issued by the Institutional Steering Committee comprised of Constances partner organisations. Applications must be submitted to the French legal authority for personal data processing, before the required data are sent to the researcher in charge of the project.

To ensure that data confidentiality and security are maintained, the Constances team generates a specific study number for each cohort participant included in a research project. Moreover, a log of variable requests submitted for a given project is kept to ensure that no-one has access, even after submitting multiple variable requests, to the entire Constances database.



The goal of the Constances project is to implement a large epidemiological cohort aimed at contributing to the development of epidemiological research and to provide public health information. The purpose of this cohort, created in the context of a partnership between the French national health insurance fund for employees (CNAMTS) and the French national retirement pension fund (Cnav), is to constitute an infrastructure open to the research community. Constances is labelled National Biology and Health Research Infrastructure by the French government's Commissariat-General for Investment..

Constances is a "general purpose" research infrastructure designed to help analysing a broad range of scientific problems. Constances was also designed as a public health and surveillance tool, thanks to the particularly exhaustive nature of the system for collecting and monitoring a great variety of data from a large representative sample of the adult population covered by the General Social Security scheme.

Download Constances' protocol

Random sampling

Subjects considered as eligible due to their age and place of residence are drawn randomly by stratified sampling with unequal probabilities, over-representing individuals with a higher probability of non-volunteering according to age, gender and SES. Random sampling is performed by the Cnav from the French national inter-scheme registry of health insurance beneficiaries (RNIAM), linked to the National careers management system (SNGC).

Data flow

The Constances cohort constitutes a complex database with characteristics rendering it highly sensitive under the terms of the French data protection act, in particular due to the collection of personal data. Moreover, some data collected at the individual level come from linkage to national databases: National inter-scheme health insurance information system (SNIIRAM), causes of death information system (CépiDc-INSERM) and the Cnav (Annual Social Security declarations, Named quarterly data, absenteeism due to illness, Active solidarity income, maternity).
Very strict data collection, organisation, and management constraints are required. The procedures in place conform to legislative and regulatory texts intended to preserve high-level personal data confidentiality and security.

Health check and questionnaires

This invitation to participate to Constances implies to undergo a health examination in a health clinic. The randomly drawn individuals first receive at home an invitation letter presenting the project, along with a mail-back coupon enabling them to give their consent.

The persons having consented to participate in Constances are invited by letter to come to their health clinic, specifying the date and location of the examination. They also receive two self-questionnaires to be filled in at home: a health and lifestyle questionnaire, and a professional calendar tracing their full job history.

In addition to the self-questionnaire completed at home, subjects undergo a health examination used to collect health-related data: clinical examination, blood analysis, blood pressure, weight, height and waist-to-hip ratio, electrocardiogram and spirometry, sight and hearing examination. This examination is standardised by means of Standard Operating Procedures (SOPs) and permanent quality control is ensured in collaboration with the ClinSearch company and the Asqualab and Eurocell associations.
Additional questionnaires, to be completed on-site (whole-life occupational exposures, women’s health), are collected at the health clinic. The informed consent signature validates the collection of these data and authorises their use for research purposes.
Participants aged 45 years and over are invited to undergo a cognitive and functional check-up. This examination is performed by a neuropsychologist and includes a series of tests. The documents pertaining to this check-up (protocol, data collection form), along with related videos, are available on request.
A yearly postal or Web-based self-questionnaire is used for active follow-up at the subjects' homes, and an invitation to come to the health clinic once every 5 years is scheduled for all cohort volunteers.
Moreover, health and socio-professional data, along with causes of death, are regularly retrieved from the national health insurance, retirement pension fund and mortality databases.

Confidentiality and ethics

The Constances cohort involves long-term longitudinal monitoring of volunteer participants. It cannot therefore be strictly anonymous as the various data concerning the individuals must be linked together throughout the follow-up and the data from different sources must be paired for each subject. Moreover, it must be possible to contact the participants (sending of questionnaires, newsletters, etc.) and to invite them to come in to their Health Clinic. It is thus essential to ensure that highly secure personal data collection and transfer procedures are followed to guarantee absolute confidentiality.
The underlying principle of these procedures is based on the assignment of multiple non-identifying numbers for each data stream, along with the implementation of independent transfer circuits between the various sources. A "trusted third party", accredited by the French national data protection commission (CNIL), acts as a conduit between the various streams and guarantees the confidentiality of all identifying data.
The data collection and management system architecture is based on processes intended to ensure that it is impossible for any person to gain (direct or indirect) access to identifiable data.
The system relies on file mapping each volunteer’s identity to a permanent anonymous number. This file contains no other information and is stored on an independent computer. It is drawn up and managed by an organisation acting as a "trusted third party", bound to medical and statistical secrecy and independent of the team managing the Constances project. The entire Constances system was authorised by the French national data protection commission (CNIL) and the INSERM Institutional Review Board.

Ethical aspects associated with the participation of multiple partners are crucial to a project of this scope. Strict rules of ethics must be enforced with the cohort participants, along with the teams and various Constances partners. The main document defining the rules of ethics that apply to epidemiological activities is "Epidemiology ethics and best practices guidelines by the association of French-speaking epidemiologists" (Adelf).
Participation in the Constances cohort relies upon the freely consented volunteering of all participants who accept the invitation to participate. Upon enrolment, participants are given detailed information brochure, specifying the benefits of the project, but also its constraints in terms of data collection, management and use. Information feedback to participants is organised with a view to ensuring better adherence of the volunteers of the Constances cohort, throughout the follow-up period. This is also an ethical obligation with respect to participants and of code of conduct with respect to physicians and health clinic employees actively involved in the project.
The Constances cohort was granted the "Label of general interest and statistical quality" by the French national council for statistical information (CNIS).


A biobank, developed through the collection of biological samples in the health clinics, will be associated with the Constances cohort. An in-depth technical analysis was conducted for the creation of a large-scale collection, including several hundred thousand distinct samples, combined with a database of clinical, paraclinical, biological, behavioural and environmental information. The aim is to develop the largest possible collection of biological samples for the future (blood: serum, plasma, urine) and to enable the storage of more specialist samples for specific research purposes.

Diagnostic validation platform

Investigating physicians validate certain identified health events (cancer, cardiovascular or neurodegenerative diseases, etc.). The purpose of these investigations is to collect data or documents (examination or hospitalisation reports, etc.) used to validate the health events and to collect detailed information essential for research.



Marie ZINS is a physician, epidemiologist, and lecturer at the Simone Veil Health Sciences School of the Versailles Saint Quentin University (UVSQ). She heads the INSERM - Versailles Saint Quentin "Population-based Epidemiological Cohorts" Unit (UMS 011).
Her research focuses on the scientific and technical development of population-based cohorts (Gazel and Constances cohorts), social epidemiology (social and occupational determinations of alcohol consumption, effects of retirement on behaviour and health) and on cohort-related methodological aspects (selection effects, longitudinal data analysis, data validation).
Marie Zins is the Constances cohort PI.


Marcel Goldberg is a physician-epidemiologist, lecturer at the Simone Veil Health Sciences Research Unit of the UVSQ. For over twenty years, he has run an Inserm research unit focusing on the epidemiology of occupational and social health determinants and was science adviser to the Occupational health department of the French health watch institute. He is a member of the Public health council and chairman of its "Information systems for public health" workgroup.
His research work focuses mainly on epidemiological studies of occupational hazards and of social determinants of health and ageing. He has created various epidemiological observation systems, in particular the Gazel cohort that has been monitored for over 25 years and for which he is co-manager with Marie Zins. He was also responsible for the Constances cohort.
Marcel Goldberg is co-manager, with Marie Zins and Lisa Berkman, of the Constances cohort.


Lisa Berkman is an epidemiologist whose research focuses mainly upon social health inequalities, social networks and health, occupational hazards ,and ageing.
Director of the Harvard Center for Population and Development Studies, she was chair of Department of Health and Social Behavior at the Harvard T.H. Chan Public Health School from 1995 to 2008. She is also the former head of the Division of Chronic Disease Epidemiology at Yale University.
In France, she is a guest professor at UVSQ and has been actively involved in the Gazel cohort since 1994.
She is co-PI of the Constances cohort, with Marie Zins and Marcel Goldberg.


Matthieu Carton is a medical doctor specialising in public health.
His main research topics pertain to occupational health determinants (occupational exposure to carcinogens, asbestos, upper aerodigestive tract cancer, sinonasal cancer, musculoskeletal disorders, difficulty of work).
As principal investigator of the Spirale post-professional programme for monitoring workers exposed to asbestos and wood dust, he is in charge of "occupational diseases" aspects of the Constances cohort.


Sébastien Czernichow is a nutritionist and professor at the UVSQ Health Sciences School. He is head of the Nutrition department, specialising in obesity medicine at the Ambroise Paré University Hospital in Boulogne-Billancourt.
His personal research work focuses on clinical or population epidemiology of obesity-related complications, with particular emphasis on metabolic and cardiovascular complications. He is involved in both the Gazel and Constances cohorts.


Alexis DESCATHA is a physician and professor at the UVSQ Health Sciences school. He specialises in occupational diseases at the Public Hospitals of Paris, University Hospitals of the western Greater Paris area.
His research activities focus on the professional environment, and on health and socioprofessional disability. He has particular interest in the epidemiology of musculoskeletal disorders.


Joseph Henny was, until 2009, director of the clinical biology laboratory at the Vandoeuvre lès Nancy Centre for Preventive Medicine. During this tie, he was also a lecturer at Nancy University. He is in charge of implementing the biobank of the Constances cohort. He is also in charge of scientific coordination for the network of health medical clinics biology laboratories.
His main scientific interest concerns the effects of biological variations on laboratory examinations. Moreover, he has contributed to the development of the concept of reference values and associated topics at the French Society for Clinical Biology (SFBC). More recently, he was actively involved in numerous studies aimed at gaining a better understanding of the components of pre-analytical and biological variation and at improving laboratory examination standardisation.


Annette Leclerc is Emeritus Director of Research at INSERM. Her main areas of research are musculoskeletal disorders, social health inequalities, and statistical methodology in data analysis.
She is involved in numerous scientific appraisal activities and is associate editor for Occupational and Environmental Medicine. She is also a member of the French CCTIRS (healthcare research data processing advisory board).

Suscribe to the newsletter (in French)

Constances' Social Security health clinics

Constances' Social Security health clinics

Contact us

Cohorte Constances

UMS 011 Cohortes épidémiologiques en population - Inserm UVSQ - Hôpital Paul Brousse - Bâtiment 15-16 - 16 Avenue Paul Vaillant Couturier - 94807 Villejuif Cedex France

Vu l’avis favorable du Conseil National de l’Information Statistique, cette enquête est reconnue d’intérêt général et de qualité statistique sans avoir de caractère obligatoire. Visa n°2015A707AU du Ministre des affaires sociales et de la santé, du Ministre de l’enseignement supérieur et de la recherche, et du Ministre des finances et des comptes publics, du Ministre de l’économie, de l’industrie et du numérique, valable pour l’année 2015. En application de la loi n°51-711 du 7 juin 1951 modifiée, les réponses à ce questionnaire sont protégées par le secret statistique et destinées à l’UMS 011 Inserm-UVSQ, « Cohortes épidémiologiques en populations ». La loi n° 78-17 du 6 janvier 1978 modifiée relative à l’informatique, aux fichiers et aux libertés, s’applique aux réponses faites à la présente enquête. Elle garantit aux personnes concernées un droit d’accès et de rectification pour les données les concernant. Ce droit peut être exercé auprès du Dr Marie Zins et du Pr Marcel Goldberg, UMS 011 Inserm-UVSQ, « Cohortes épidémiologiques en population », Hôpital Paul Brousse, Bât. 15/16, 16 avenue Paul Vaillant-Couturier, 94807 Villejuif Cedex.