Over the past decades, longer life expectancy and aging populations have increased the burden of chronic diseases. Closely related to this is the issue of multimorbidity (defined as having two or more diseases).
More recently, the significance of multimorbidity was highlighted by the COVID-19 pandemic: beyond the infectious dimension, it had a major impact on all chronic pathologies, affecting in a potentially heavy and long-term manner the burden of disease. Close estimation of burden and multimorbidity is important to better target public policy actions that will promote healthy behaviours and environments throughout the course of life.
The wide range of approaches used to measure multimorbidity compromises comparability across countries and analysis of trends across time. Furthermore, the impact of multimorbidity on mortality, activity limitations and quality of life (perceived health) has rarely been studied. In order to better understand the determinants of multimorbidity, Santé publique France collaborated with a scientist from the National University of Singapore on a new approach to studying the health impact of the main multimorbidity combinations (dyads, triads or tetrads) by gender, age, and socioeconomic and geographic factors.
The results published this month in the journal PLOS One  also shed light on the aggregation process at the origin of multimorbidity and make it possible to identify the interactions between the most deleterious diseases that should be considered a priority. These elements will help to refine surveillance indicators and, in turn, even prevention actions.
3 questions to Joël Coste, Santé publique France
Evaluation of the multimorbidity burden must go well beyond the simple count of chronic diseases, as it is most often done in epidemiological literature.
It must take into account the different impacts of morbid associations on health status, the synergy of their effects and their aggregation mechanisms. Some morbid associations, because of their high impact on health, their multiplying synergistic effect and common risk factors, deserve greater attention.
The initial work (1) that led to the formalisation of the burden qualification approach described in this article was based on data from two population-based surveys (Health and Social Protection Survey ESPS 2010-14 and Handicap-Household Health Survey HSM 2008) concerning more than 60 pathologies.
The consequences of these pathologies have been studied in terms of activity limitation, perceived health and mortality, the three main criteria usually taken into account to estimate the burden of diseases. But measuring and characterizing the burden of multimorbidity was a new approach.
The second step, as presented in our recently published article, was to characterise the epidemiology of multimorbidity in the French adult population and to explore its diversity by gender, age, and several socioeconomic and territorial indicators.
The approach began by identifying pathologies that have significant implications in terms of activity limitations, perceived health status or mortality. Forty-eight chronic pathologies independently affecting one of these indicators were selected. We then characterised each group (dyads, triads, etc.) of associated pathologies in terms of frequency and the mechanisms explaining their association (shared risk factors, causal relation, etc.). The most frequent multimorbid associations include cardiometabolic, osteoarticular and mental pathologies, which maintain causal relationships between each other or can be explained by common risk factors (obesity, insufficient activity, etc.). Finally, we evaluated the impact as well as the cumulative and synergistic effects of these associations to identify those with the most deleterious effects on health status. They include complicated diseases (such as organ failure), but also pathologies affecting the sense organs and the locomotor system (for activity limitations), as well as mental illness (for perceived health). The synergistic effects of associated pathologies vary from addition (the effects simply add up between the pathologies when associated in dyads or triads: cardio-metabolic affections, low back pain, osteoporosis, trauma sequelae, depression and anxiety); to multiplication (the effects are multiplied between the pathologies when associated: obesity, COPD, migraine, certain osteo-articular pathologies).
The prevalence of multimorbidity in France has been estimated at 30%, based on associations of pathologies that have an impact on health status (perceived health, activity limitations or mortality) and taking as a criterion the presence of two chronic diseases in the same individual during the year. This prevalence is 39% if we take as a criterion the presence of two chronic diseases throughout life. However, the prevalence of multimorbidity is higher (+23% to +31%) and occurs earlier (5 to 15 years) in women than in men; it is higher in older subjects, but not negligible in those aged 35 to 44 years (11% and 20% in men and women of this age group, respectively); and it is higher in subjects with the lowest levels of education and income (+84% to +104%), as well as among manual workers (+96%).
A particularly clear inverse dose-response relationship was observed with education level, suggesting causality of the relationship. However, the influence of geographic and territorial indicators was very small and negligible once socioeconomic factors were taken into account. Additional analyses identified multimorbid associations particularly marked by social inequalities, which included low back pain, osteoarthritis, COPD, and anxiety.
Santé Publique France has made multimorbidity a priority theme in the surveillance and prevention of chronic diseases.
The next step will be to assess the evolving trends of multimorbidity in connection with health inequity based on the 2022 data from the Drees Autonomy survey and the SNDS survey (national health data system), which will allow us to explore certain morbid associations further.
Similarly, the relevance of multimorbidity as a predictor of negative health events will be tested by the risk assessment of COVID-19/severe COVID-19 associated with multimorbidity and its components: to this end, a systematic review with European partnerships is currently in progress within the agency.
The first results already allow us to underline that multimorbidty monitoring and prevention should begin as early as mid-life and probably even sooner in disadvantaged groups. Particular attention should be paid to the main elementary groupings of two, three or four pathologies, which concern 75% of multimorbid subjects in France, especially the hypertension-lumbar pain associations (ranked 1st in France) and those concerning obesity-hypertension (ranked 2nd), which generally share the same determinants. The series of associations between common chronic pain pathologies (migraine, low back pain, arthrosis of peripheral joints) are also widely related to the same determinants (obesity and mental illness) and carry the same risks regarding (over)consumption of analgesics. The same is true for the main associations affected by socioeconomic health inequity (involving low back pain, COPD, and anxiety), for which educational determinants appear causal.
These results support the prevention and health promotion strategy currently under development at Santé publique France, which recommends action before mid-life (40-55 years old) in order to promote behaviours and environments that are health-positive and that reduce the burden of chronic diseases and multimorbidity, thus supporting healthy aging.
 Coste J, Valderas JM, Carcaillon-Bentata L. The epidemiology of multimorbidity in France: variations by gender, age and socioeconomic factors, and implications for surveillance and prevention. PLoS One. 2022 Apr 6;17(4):e0265842
(1) Coste J, Valderas JM, Carcaillon-Bentata L (2021). Estimating and characterizing the burden of multimorbidity in the community: A comprehensive multistep analysis of two large nationwide representative surveys in France. PLOS Medicine 18(4): e1003584.