Skip to main content
NCBI home page
BMJ Public Health logoLink to BMJ Public Health
. 2026 May 21;4(2):e004258. doi: 10.1136/bmjph-2025-004258

Contextual impact of socio-economic inequalities on health: a case study of COVID-19 mortality among populations with and without diabetes in France

Marie Guion 1,2,✉, Sarah Goria 3, Sandrine Fosse-Edorh 1,0, Emmanuel Cosson 2,4,0
PMCID: PMC13201992  PMID: 42205606

Abstract

Aims

The aim was to describe the impact of socio-economic inequalities (SEI) on COVID-19 mortality in people aged 45 years and over living with or without diabetes during the two epidemic waves of 2020 (wave 1 (W1): March–May; wave 2 (W2): September–December) in European France.

Methods

People living with pharmacologically treated diabetes were identified using a validated algorithm from the French National Health Data System. COVID-19 mortality in 2020 was obtained from the French Medical Causes of Death database. SEI were measured using the French Deprivation Index at municipality level (FDep; with Q1 corresponding to people living in the least deprived). Age-standardised COVID-19 mortality was calculated by FDep quintiles for each month of the year and by sex. The relative risk of COVID-19 mortality associated with FDep quintiles (with Q1 as the reference) according to diabetes status and epidemic wave was estimated by sex using an adjusted log-linear Poisson model.

Results

During W1, when only collective measures were available (lockdown), no social gradient was observed, and it was even the least deprived women with diabetes (Q1) who had a higher risk of COVID-19 mortality. Among men with diabetes, only the most deprived (Q5) had a slightly higher risk of COVID-19 mortality than the least deprived (Q1). The same results were overall found in people without diabetes.

During W2, after the implementation of the first individual preventive measures (mask), COVID-19-related mortality followed a positive SEI gradient among men and women with and without diabetes, with higher rates in the most deprived groups (Q5). Moreover, the influence of SEI was greater in women without diabetes in the most deprived groups (Q3, Q4, Q5) than in women with diabetes.

Conclusion

Our study highlights the importance of integrating the principles of equity, which give consideration to the most vulnerable, into the development of individual prevention measures.

Keywords: COVID-19, Epidemiology, Diabetes Mellitus

WHAT IS ALREADY KNOWN ON THIS TOPIC

  • The COVID-19 pandemic has severely affected the population, including those with diabetes, leading to preventive measures whose impact on the population remains unknown.

WHAT THIS STUDY ADDS

  • During the COVID-19 pandemic, the effect of socio-economic inequalities on COVID-19 mortality differed between the two epidemic waves, during which different preventive intervention strategies were implemented (collective versus individual).

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

  • These results highlight the importance of incorporating principles of equity, which take into account the most vulnerable, into the development of individual prevention measures.

Introduction

Inequalities in mortality are present throughout the world, including in countries where access to healthcare is facilitated, such as France. In this country, inequalities in mortality could even be exacerbated compared with other European countries.1 Yet for certain populations at risk, such as people living with diabetes, the socio-economic inequalities are even greater.2 The prevention messages aimed at the general population are more beneficial to the least deprived,3 who are more likely to have a higher level of health literacy.4 The social inequalities in health could therefore be aggravated when insufficiently targeted preventive measures are deployed, as is described in the theory of fundamental causes.3 5

Taking the recent example of the COVID-19 pandemic,6 we hypothesised that the effects of socio-economic inequalities on COVID-19 mortality may have worsened in France between the two epidemic waves of 2020. During the first wave (W1), general preventive measures were deployed (total lockdown), whereas during the second wave (W2), more individual preventive measures were introduced (mask wearing, frequent handwashing). Indeed, the theory of fundamental causes5 states that socio-economic status is a fundamental cause of inequalities in mortality, in the sense that it ‘embodies an array of resources, such as money, knowledge, prestige, power and beneficial social connections’, which can be used ‘to avoid disease risks or to minimise the consequences of disease once it occurs’ so that ‘as opportunities for avoiding disease expand so health inequalities continue to exist’.3 This analysis is therefore essential to better anticipate potential future health crises,6 7 particularly among the most vulnerable populations.

As such, the aim of this study was to compare the influence of social inequalities on COVID-19 mortality during the two epidemic waves of 2020 in European France. More specifically, analyses were performed in the population living with diabetes, a major risk factor for COVID-19 mortality8,10 and in whom social inequalities have already been identified,11 and compared with people without diabetes.

Methods

Database

The French National Health Data System (Système national des données de santé (SNDS)), a medical-administrative database, collects individual and anonymised healthcare data on the beneficiaries of the country’s various health insurance schemes, covering virtually its entire population.12 The SNDS encompasses several databases, including the Inter-Scheme Consumption Data (Données de consommation inter-régimes (DCIR)) and the Medical Causes of Death.

The DCIR contains information on the reimbursement of outpatient care (ie, medical procedures, biological tests, medical devices, drugs) and sociodemographic information on the beneficiaries, such as age, sex, date of death and municipality of residence.

The French Medical Causes of Death database, managed by the Center for Epidemiology of Medical Causes of Death (Centre d’épidémiologie sur les causes médicales de Décès), contains an exhaustive list of all people who have died in France,13 as well as their underlying and associated causes of death and other sociodemographic variables, such as municipality of residence, year of birth, date of death and sex. The underlying causes of death are defined by WHO codes as “the disease or injury which initiated the train of morbid events leading directly to death, or the circumstances of the accident or violence which produced the fatal injury”,14 in accordance with the WHO International Classification of Diseases, 10th Revision (ICD-10) rules. The other causes mentioned on the certificate are the associated causes.

The DCIR and French Medical Causes of Death databases were probabilistically matched on several variables (years, months and days of death; sex; years, months and days of birth; department and municipality of residence), in order to link the information in the two databases. The match rate between the two databases, which varies according to the year studied, was 87.2% for 2020.15

Study population

The population studied includes all people living in European France who have had at least one healthcare reimbursement per year, known as consumers. It represents around 65.5 million people. For our analyses, we retained only those individuals for whom we had complete data for our variables of interest (ie, age, sex, French Deprivation Index (FDep), municipal population density).

Concerning the underlying causes of death, we selected death caused by COVID-19 with the ICD-10 codes U07.2 and U07.1.

Diabetes indicators

People with diabetes were identified using a validated algorithm based on reimbursements for antidiabetic drugs (class A10 of the Anatomical Therapeutic Chemical Classification) on at least three separate dates in a calendar year, or on two dates if at least one large package of an antidiabetic drug had been dispensed.16

The COVID-19 mortality rates for people with diabetes, both annual and by wave, were calculated as the ratio between the number of people with diabetes who died from COVID-19 during the year 2020 or during W1 or W2 and the number of person-years (PY) with diabetes alive on 1 January 2020.

The monthly COVID-19 mortality rates were also calculated in relation to the people who were alive on the first day of each month in order to take account of the mortality over time during this period.

Socio-economic indicators

Socio-economic inequalities were measured using an ecological indicator of deprivation, called the FDep V.2015.17 Social disadvantage measured by FDep is defined as an accumulation of material and social disadvantages, measured at the municipal level in European France (which has 36 000 municipalities). It is made up of four indicators (unemployment rate, percentage of blue-collar workers in the active population, percentage of high school graduates and median household income) from routinely measured spatial data. FDep population-weighted quintiles were considered. Thus, the first quintile (Q1) includes people living in the least deprived areas and the last quintile (Q5) includes those living in the most deprived areas. FDep quintiles were assigned to each person based on their municipality of residence.

Statistical analyses

All analyses were stratified by sex. We included people over 45 years of age because it is from this age that our event of interest, COVID-19 mortality, is highest (very few cases before 45 years) (online supplemental table 1). Crude rates of mortality from COVID-19 over the whole period and for W1 and W2 were estimated for each 5-year age group from 45 to 89 years and then for those aged 90 years and over. Age-standardised COVID-19 mortality rates by FDep quintile for the whole period and by month were estimated using the 2013 European standard population.18 The excess mortality of people with diabetes compared with people without diabetes was calculated using the standardised mortality ratio (SMR) and 95% CIs. The SMR was defined as the ratio between the observed number of deaths of people with diabetes and the expected number of deaths of people with diabetes obtained by using the mortality rate of the population without diabetes.

To explore the association between COVID-19 mortality, diabetes status (people with or without diabetes), time period (W1: March–May 2020, W2: September–December 2020 and W3: all other months in 2020, results not shown) and FDep of the municipality of residence (five quintiles), we defined a log-linear Poisson model in a hierarchical Bayesian framework. The model took into account age, with a second-order random walk, population density of the municipality of residence (1: large urban centre; 2: intermediate urban centre, urban belts and small towns, 3: rural towns, rural with dispersed housing and rural with very dispersed housing19), time period, diabetes status and FDep quintiles (with Q1 as the reference) and the interaction between these last three terms. The offset term included in the model was the population living at the beginning of each month by age, FDep quintiles, region and population density. Spatially structured and unstructured random effects at the regional level were included to account for overdispersion and were modelled using the Besag-York-Mollié model.20 Adjacency was used to define neighbours: all regions sharing a border with the region of interest were defined as neighbours. The integrated nested Laplace approximation (INLA) approach was used to compute the posterior marginals of all the parameters of interest.21

Analyses were performed using the SAS Enterprise Guide (V.4.3, SAS, Cary, USA) for the descriptive results and the R software environment using the INLA and CARBayes packages.

Results are presented as relative risks (RR) of COVID-19 mortality associated with FDep quintiles by diabetes status and period, and their 95% credibility intervals.

Results

Study population

On 1 January 2020, there were 3275 million people living with pharmacologically treated diabetes in European France and 62 218 million people living without diabetes. In 2020, a total of 56 190 people in our study population died from COVID-19.

Of the 29 million people aged 45 years and over alive on 1 January 2020 (3019 million people with pharmacologically treated diabetes), 55 894 died from COVID-19. Of these, 12 396 (60% men, mean age of death from COVID-19 for men 78.5 years and for women 82.4 years) were receiving pharmacological treatment for diabetes and 43 498 (49% men, mean age of death from COVID-19 for men 81.5 years and for women 86.5 years) were without diabetes. Figure 1 shows the flow chart for the study.

Figure 1. Flow chart. FDep, French Deprivation Index.

Figure 1

Open in a new tab

During W1 (March–May 2020), 5366 people with diabetes (60.5% men) and 19 245 people without diabetes (49.4% men) died from COVID-19. During W2 (September–December 2020), 6779 people with diabetes (59.9% men) and 23 262 people without diabetes (48.7% men) died from COVID-19.

COVID-19 mortality rates

Over the whole of 2020, the crude COVID-19 mortality rates for men and women aged 45 years and over with diabetes were 441 and 372 per 100 000 PY, respectively. During W1, these rates were 769 and 637 per 100 000 PY, respectively and during W2 they were 721 and 613 per 100 000 PY, respectively.

For men and women without diabetes aged 45 years and over, these rates were 178 and 153 per 100 000 PY in 2020, 319 and 268 per 100 000 PY in W1 and 285 and 247 per 100 000 PY in W2, respectively.

Figure 2 shows the crude mortality rates for COVID-19 in the different age groups for men and women from age 45 years according to their diabetes status, as well as the ratios for people with versus without diabetes for the entire 2020 period (figure 2A,D) and for W1 (figure 2B,E) and W2 (figure 2C,F). COVID-19 mortality increased with age in both sexes, and was higher in men than in women. Mortality was also higher in people with diabetes than without, regardless of age group, with a higher ratio of diabetes versus non-diabetes in the younger age groups. This ratio was higher during W1 than W2 in women (figure 2E,F).

Figure 2. COVID-19 mortality rate according to diabetes status by period and age group, European France 2022. COVID-19 mortality rates by age group, according to diabetes status (with or without diabetes) and mortality ratios (diabetes versus no diabetes) in men for the whole of 2020 (A), from March to May 2020 (B) and from September to December 2020 (C), and in women for the whole of 2020 (D), from March to May 2020 (E) and from September to December 2020 (F), people living in European France aged 45 years and over. PY, person-years.

Figure 2

Open in a new tab

Standardised COVID-19 mortality rates and excess mortality by socio-economic level for the two epidemic waves

Overall, for the whole of 2020, the age-standardised mortality rates for COVID-19 in men aged 45 years and over with or without diabetes were 339 and 200 per 100 000 PY, respectively. In women, they were 215 and 110 per 100 000 PY, respectively. The SMR adjusted for age by sex of people with diabetes related to people without diabetes was approximately 1.7 in men (1.69 (95% CI 1.66 to 1.73)) and women (1.75 (95% CI 1.70 to 1.80)).

Figure 3 shows the age-standardised COVID-19 mortality rates by socio-economic level in men and women according to their diabetes status for each month in 2020. Regardless of sex, people aged 45 years and over with diabetes (figure 3A,C) had higher COVID-19 mortality rates across all FDep quintiles than those without diabetes (figure 3B,D) during W1 and W2. Men had higher mortality rates than women, regardless of diabetes status, across all FDep quintiles. During W1, regardless of sex and diabetes status, people living in Q1 had slightly higher mortality rates than people living in the other quintiles, with the exception of the month of May for men with diabetes. In contrast, during W2, it was generally people living in Q5 who had higher mortality rates.

Figure 3. COVID-19 mortality rate by diabetes status for each month of 2020, European France. Age-standardised COVID-19 mortality rates in men with diabetes (A) and without diabetes (B) and in women with diabetes (C) and without diabetes (D) aged 45 years and over for each month of 2020 in each FDep quintile in European France. The red zone depicts the first wave, the blue zone the second wave and the grey zone the non-wave months. FDep, French Deprivation Index (with Q1 corresponding to the least deprived municipalities); PY, person-years.

Figure 3

Open in a new tab

Influence of social deprivation on COVID-19 mortality according to diabetes status during the two epidemic waves

Figure 4 shows the influence of socio-economic inequalities on COVID-19 mortality according to sex and diabetes status during the two epidemic waves.

Figure 4. Adjusted relative risks (aRR) of COVID-19 mortality by period, diabetes status and socio-economic status, European France. aRR (95% credibility intervals) of COVID-19 mortality associated with the quintiles of the FDep (Q1: the reference) according to diabetes status (with or without diabetes) in men during wave 1 (March–May 2020) (A) and wave 2 (September–December 2020) (B) and in women in wave 1 (C) and wave 2 (D) aged 45 years and over in European France in 2020 (log-linear Poisson model). diab, people with diabetes; FDep, French Deprivation Index (with Q1 corresponding to the least deprived municipalities); no-diab, people without diabetes.

Figure 4

Open in a new tab

During W1, in men, the RR of mortality did not differ from Q2, Q3 and Q4 versus Q1, and increased significantly in Q5 regardless of diabetes status (figure 4A). In women, the RR of mortality was significantly lower in Q2, Q3 and Q4 versus Q1 regardless of diabetes status, and also significantly lower in Q5 for those without diabetes (figure 4C). In contrast, during W2, regardless of sex and diabetes status, the RR of mortality versus Q1 differed significantly and increased according to a positive gradient from Q2 to Q5 (figure 4B and D), except for women without diabetes, where no gradient was present (figure 4D). Finally, in W1 and W2, the RR of mortality for all deprivation quintiles versus Q1 was similar in men and women with or without diabetes, except for women in W2 (figure 4D), where the RR of mortality for Q3, Q4 and Q5 versus Q1 was higher in women without diabetes than in women with diabetes.

Discussion

Principal results

This study, based on virtually all deaths from COVID-19 in people aged 45 years and over in European France in 2020, shows that COVID-19 mortality was much more severe in people with diabetes, in men and in people over the age of 75 years. In addition, we show that the association between socio-economic inequalities and COVID-19 mortality differed during W1 and W2. During W1, these inequalities were weakly associated with COVID-19 mortality in men and disadvantaged women living in the least deprived municipalities, regardless of diabetes status. During W2, a socio-economic mortality gradient emerged, to the disadvantage of the most deprived, regardless of sex or diabetes status.

Comparisons of results

Risk factors for COVID-19 mortality

The vast majority of studies have shown that people with diabetes had a greater risk of dying from COVID-19 than those without diabetes. Similar to our study, some studies have shown a twofold increased risk of COVID-19 mortality in people with diabetes.22 However, this increased risk tended to reduce with advancing age, probably because the populations without diabetes were also accumulating an increasing number of other comorbidities.9

It has also been widely shown that increasing age and male sex were risk factors for COVID-19 mortality in the populations with or without diabetes.10 11 23

Socio-economic inequalities and COVID-19 mortality in people with diabetes

Few studies have examined the association between socio-economic inequalities and COVID-19 mortality in the population with diabetes. A study in England showed that COVID-19 mortality was significantly higher in people with diabetes in the most deprived quintiles than in the least deprived quintiles.11 However, this study only covered the start of the COVID-19 pandemic (up to 11 May 2020) and did not address the evolution in these socio-economic inequalities according to epidemic period. The difference observed with this study during the first wave of the epidemic can be explained by the fact that the behaviours of the populations, preventive measures and epidemic waves differed from country to country.

Socio-economic inequalities and epidemic periods: what is happening around the world?

Studies in the general population are however more numerous. Studies in France have shown that the association between COVID-19 incidence and socio-economic status varied according to the period of the epidemic.24 25 Indeed, the authors sought to compare the risk of COVID-19 infection before and during lockdown according to socio-economic status, and showed a higher risk of infection before lockdown among the least deprived compared with the most deprived; however, this risk fell more sharply during lockdown among the least deprived.25

Moreover, if we look specifically at temporal changes in the influence of socio-economic inequalities on COVID-19 mortality, a general population study in the USA showed that COVID-19 mortality rates varied over time. Initially, the highest rates were in the least disadvantaged areas, but these rates gradually shifted towards the most disadvantaged areas, a result consistent with that observed in our study. However, the same study suggests that after a certain period, a new reversal of the trend took place.26 Another USA population-based study27 showed that COVID-19 incidence and mortality were higher among the least deprived at the start of the epidemic (January–May 2020), but that this association reversed over time, to the detriment of the most deprived. The same dynamic has been seen in our study, which also has the advantage of showing national and stratified data on diabetes, which is one of the main risk factors for COVID-19 mortality.10 11 23

Other general population studies have investigated the association between COVID-19 mortality and socio-economic inequalities, but only during the first wave of COVID-19 (up to May 2020). Some studies showed higher mortality in the least deprived populations,28 29 other studies showed no difference according to socio-economic status,30 and the vast majority of studies showed higher mortality in the most deprived populations.31 32 Given that COVID-19 was an emerging disease, knowledge of this disease and the measures put in place by governments developed gradually and in different ways depending on the countries, with one example being the USA, Brazil and Senegal whose governments have not embraced social distancing measures,33,35 making it more difficult to compare studies carried out in different countries.36

Explanations of our results

Differing influence of socio-economic status on COVID-19 mortality between wave 1 and wave 2

It has been hypothesised that the development of an emerging disease must be placed in a temporal context,37 because it evolves in the population through different phases according to the knowledge and understanding of its risk factors, and of its prevention and treatment,36 which vary according to socio-economic status.38

In France in 2020, collective preventive measures were deployed during W1, with a near-total lockdown of the population (except for key roles). Before lockdown was introduced, exposure to the virus may have been higher among the most deprived, potentially due to factors such as living in densely populated areas and overcrowded housing39 and having jobs that bring them into contact with the public (cashiers, bus drivers, etc).24 However, this may have been earlier among the least deprived, because they have a greater social network and ability to travel and their jobs as healthcare workers bring them into direct contact with people who are sick.40 41 This greater upstream exposure of the least deprived may be one possible explanation for the absence of a socio-economic gradient during the first wave of the epidemic.25 Complementarily, this absence of a social gradient could also be explained by the initial spread of the virus, which, in the absence of preventive measures, is relatively random and affects the population with a similar probability.

Indeed, during W2 in France, greater awareness of the pandemic was associated with more individual preventive measures deployed: mask wearing in public places, social distancing, use of hand sanitiser gel, etc. However, it has been shown that new health interventions are mainly adopted by the least deprived people when first deployed.442,44 Several reasons are often suggested to explain the difference in the adoption of new behaviours according to socio-economic status, such as the understanding of messages, their acceptability and their application,42 45 which are higher among the least deprived people who have better health literacy.46 In contrast, one study showed that people with a high socio-economic status engaged less in social distancing but were greater adopters of hygiene behaviours, although this conduct reduced as the number of hospitalisations increased. The authors suggested that adapting and habituation to risk should also be taken into account in studies of emerging diseases.47 Lastly, during the second wave of the epidemic, the massive deployment of telework only benefited the least deprived, as the most deprived were more likely to have jobs requiring them to work outside the home.24 Thus, the appearance of a socio-economic gradient in COVID-19 mortality during the second wave may be related to better understanding and application of the prevention messages and lower exposure by the least deprived, compared with the most deprived.

Differing influence of socio-economic status on COVID-19 mortality in men and women

In addition, our results show different dynamics in the effects of socio-economic inequalities on COVID-19 mortality according to sex during the first wave of the epidemic. While the least deprived women had a greater risk of COVID-19 mortality than the other quintiles, the most deprived men had the highest risk. Potentially, the least deprived women may have been more exposed to the virus before the first lockdown, as they are the most likely to take on responsibilities in terms of childcare (extracurricular activities, childcare during teleworking) and household chores (shopping).48

We show that the influence of socio-economic inequalities on COVID-19 mortality was similar in people with and without diabetes, except in women during the second period: the influence of socio-economic inequalities was greater in women without diabetes than in women with diabetes. We can hypothesise that women with diabetes had been more sensitive to the prevention messages because they knew that their disease was a risk factor for COVID-19 complications. We also know that women are more attentive to their health and to preventive behaviours than men,47 49 which may help explain why this difference is not found in men.

Study perspectives

Our results raise the question of the genesis of socio-economic inequalities during epidemics (in our study, that of COVID-19 in European France in 2020). Our study shows the dynamic between socio-economic inequalities and COVID-19 mortality at the start of the pandemic, but did not measure this dynamic later. For example, although vaccination is provided in high-income countries free of charge or at low cost (available free of charge in 2021 in France), we hypothesise that vaccination programmes may have generated additional pandemic-related inequalities.31 50 51 It is therefore vital to pursue research in this area in order to better understand and better prevent the negative impacts of the COVID-19 pandemic on these populations, particularly over the long term (loss of employment, cessation of treatment, mental health, violence, decompensation, etc).6

Contributions to science and public health (guidelines)

Our findings will help to improve knowledge of the COVID-19 epidemic and other pandemics. In particular, they suggest that the impact of public policies is far from negligible when it comes to engendering social inequalities in health. There are at least four factors that explain these social inequalities in health during epidemics: exposure, transmission, vulnerability and access to healthcare.31 52 We would add that differential exposure to prevention messages must also be taken into account. The challenge for public policy is to adapt these action levers to the different social groups so that they are effective for everyone. This could be achieved through co-designing policies, by involving in the decision-making healthcare professionals and patients as closely as possible to their life context, or by developing interventions that respect the concept of ‘proportionate universalism’.53 54 In urgent situations, such as during the COVID-19 health crisis, anticipating these principles when implementing preventive measures as part of the preparation for any health crisis would help to limit the occurrence or aggravation of social inequalities in health.

Benefits of the study

Our study has many strengths. The use of the SNDS databases made it possible to include virtually the entire population of European France.12 The use of a validated algorithm to identify people with pharmacologically treated diabetes16 enabled us to quantify the mortality of pharmacologically treated people with diabetes in a quasi-exhaustive way. As the study focused on the year 2020, the results were not influenced by possible variants of the COVID-19 virus (with different severity), nor by potential differences in vaccination (available in 2021). Finally, by incorporating density into the model, we were able to take into account the more likely spread of the virus in densely populated municipalities.

Limitations of the study

However, our study has certain limitations. The SNDS databases do not contain information on certain risk factors strongly associated with diabetes, such as ethnic origin,55 family history of diabetes or body mass index.56 57 Yet it has been shown that some ethnic origins were more exposed to COVID-19 than others,7 and that this exposure varied according to the epidemic period studied.58 We did not have access to certain other important information that could constitute potential confounding factors, such as household size, type of housing, use of transportation or occupation, which have been associated with COVID-19 mortality risk.40 Given that some of these factors evolved during the pandemic (notably transportation use and remote work), their omission may have influenced the estimates made between the two epidemic waves. Also, the algorithm used in the SNDS to identify people with diabetes has certain limitations. It does not distinguish between the different types of diabetes.59 It also does not identify people with undiagnosed diabetes,57 60 which probably underestimates the excess COVID-19 mortality of people with diabetes. Also, because the linkage between our two data sources was not 100%, we are also probably underestimating the number of people who died from COVID-19 with or without diabetes, but probably without differential bias according to diabetes status and epidemic wave. In addition, the fact that around 13% of deceased individuals were not matched could introduce selection bias if, for instance, unmatchable records are systematically associated with specific socio-economic or demographic profiles. However, there is no reason to expect differential bias between the two waves of the epidemic, since the linkage is performed on an annual basis. Also, the care of some elderly people in institutions is not included in the SNDS, even though this population has been heavily affected by COVID-19 mortality. Finally, use of the FDep based on municipality of residence is not a simple approximation of individual characteristics, and an approximation of the individual’s living environment, inducing an ecological bias. Consequently, the use of the FDep as a socio-economic indicator may lead to an underestimation of the actual social inequality gaps observed in our study.

Conclusion

Our study suggests that the preventive measures implemented during the COVID-19 pandemic may have contributed to the emergence or reinforcement of social inequalities in health. We show that taking temporality into account is essential for understanding the dynamics between emerging pathologies and social inequalities in health in epidemiological studies. Taking COVID-19 as an example, during the first wave of COVID-19, when only collective interventions were deployed, there were few social inequalities in COVID-19 mortality. During the second wave, preventive measures were incorporated on a more individual basis, and a habitual gradient of inequality was observed. Our study highlights the importance of incorporating principles of equity, which take into account the most vulnerable using the principle of proportionate universalism, into the development of individual prevention measures. In particular, this would have involved increased distribution of free masks and hand sanitiser, targeted support for people with chronic conditions such as diabetes who work in essential jobs that do not allow for remote working and the maintenance or even strengthening of medical monitoring during periods of restriction. By considering social and medical vulnerabilities simultaneously, such measures might have contributed to reduce inequalities in exposure to the virus, limited the risk of severe forms of the disease and reduced mortality inequalities. However, prevention policies should also address underlying social determinants, such as income insecurity, housing conditions and limited access to health information, with the aim of reducing the vulnerability of disadvantaged populations and, consequently, inequalities in mortality. These findings are crucial when preparing responses to new emerging diseases, as well as developing strategies for preventing and controlling non-communicable diseases such as diabetes.

Supplementary material

online supplemental file 1
bmjph-4-2-s001.docx (26.9KB, docx)
DOI: 10.1136/bmjph-2025-004258

Acknowledgements

We would like to thank the Fondation de France and Santé publique France for funding the doctoral thesis of MG. We would also like to thank Thomas Rohban for guiding us in drafting these results. Finally, we would like to thank Santé publique France’s English translation service.

The funder had no role in the study design, data collection and analysis, decision to publish or preparation of the manuscript.

Footnotes

Funding: This research was carried out as part of a doctoral thesis funded by the Fondation de France (00107045).

Provenance and peer review: Not commissioned; externally peer reviewed.

Patient consent for publication: Not applicable.

Ethics approval: Not applicable.

Data availability free text: The aggregated data are accessible to researchers on reasonable request to the corresponding author. Requests for data require the approval of Santé Publique France.

Patient and public involvement: Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

Data availability statement

Data are available on reasonable request.

References

  • 1.Mackenbach JP, Stirbu I, Roskam A-JR, et al. Socioeconomic inequalities in health in 22 European countries. N Engl J Med. 2008;358:2468–81. doi: 10.1056/NEJMsa0707519. [DOI] [PubMed] [Google Scholar]
  • 2.Guion M, Mandereau-Bruno L, Goria S, et al. Eleven-year trends in socioeconomic inequalities in the prevalence and incidence of pharmacologically treated type 2 diabetes in France, 2010-2020. Diabetes Metab. 2024;50:101509. doi: 10.1016/j.diabet.2024.101509. [DOI] [PubMed] [Google Scholar]
  • 3.Mackenbach JP, Looman CWN, Artnik B, et al. “Fundamental causes” of inequalities in mortality: an empirical test of the theory in 20 European populations. Sociol Health Illn. 2017;39:1117–33. doi: 10.1111/1467-9566.12562. [DOI] [PubMed] [Google Scholar]
  • 4.Tajdar D, Lühmann D, Fertmann R, et al. Low health literacy is associated with higher risk of type 2 diabetes: a cross-sectional study in Germany. BMC Public Health. 2021;21:510. doi: 10.1186/s12889-021-10508-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Link BG, Phelan J. Social conditions as fundamental causes of disease. J Health Soc Behav. 1995:80–94. https://core.ac.uk/reader/77145022?utm_source=linkout Available. [PubMed] [Google Scholar]
  • 6.Bambra C, Riordan R, Ford J, et al. The COVID-19 pandemic and health inequalities. J Epidemiol Community Health. 2020;74:964–8. doi: 10.1136/jech-2020-214401. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Khalatbari-Soltani S, Cumming RC, Delpierre C, et al. Importance of collecting data on socioeconomic determinants from the early stage of the COVID-19 outbreak onwards. J Epidemiol Community Health. 2020;74:620–3. doi: 10.1136/jech-2020-214297. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Hartmann-Boyce J, Rees K, Perring JC, et al. Risks of and From SARS-CoV-2 Infection and COVID-19 in People With Diabetes: A Systematic Review of Reviews. Diabetes Care . 2021;44:2790–811. doi: 10.2337/dc21-0930. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Barron E, Bakhai C, Kar P, et al. Associations of type 1 and type 2 diabetes with COVID-19-related mortality in England: a whole-population study. Lancet Diabetes Endocrinol. 2020;8:813–22. doi: 10.1016/S2213-8587(20)30272-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Semenzato L, Botton J, Drouin J, et al. Chronic diseases, health conditions and risk of COVID-19-related hospitalization and in-hospital mortality during the first wave of the epidemic in France: a cohort study of 66 million people. Lancet Reg Health Eur . 2021;8:100158. doi: 10.1016/j.lanepe.2021.100158. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Holman N, Knighton P, Kar P, et al. Risk factors for COVID-19-related mortality in people with type 1 and type 2 diabetes in England: a population-based cohort study. Lancet Diabetes Endocrinol. 2020;8:823–33. doi: 10.1016/S2213-8587(20)30271-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Tuppin P, Rudant J, Constantinou P, et al. Value of a national administrative database to guide public decisions: From the système national d’information interrégimes de l’Assurance Maladie (SNIIRAM) to the système national des données de santé (SNDS) in France. Rev Epidemiol Sante Publique. 2017;65 Suppl 4:S149–67. doi: 10.1016/j.respe.2017.05.004. [DOI] [PubMed] [Google Scholar]
  • 13.Rey G. Death certificate data in France: Production process and main types of analyses. Rev Med Interne. 2016;37:685–93. doi: 10.1016/j.revmed.2016.01.011. [DOI] [PubMed] [Google Scholar]
  • 14.WHO Cause of death. [26-Jun-2024]. https://www.who.int/standards/classifications/classification-of-diseases/cause-of-death Available. Accessed.
  • 15.Roquefeuil L, Rey G, Bounebache K, et al. Documentation à l’attention des utilisateurs concernant l’intégration des causes médicales de décès dans le système national des données de santé - version 4 2020. 2024. https://documentation-snds.health-data-hub.fr/snds/formation_snds/documents_cnam/guide_cepidc/Avant_propos.html#avant-propos Available.
  • 16.Fuentes S, Cosson E, Mandereau-Bruno L, et al. Identifying diabetes cases in health administrative databases: a validation study based on a large French cohort. Int J Public Health. 2019;64:441–50. doi: 10.1007/s00038-018-1186-3. [DOI] [PubMed] [Google Scholar]
  • 17.Rey G, Jougla E, Fouillet A, et al. Ecological association between a deprivation index and mortality in France over the period 1997 - 2001: variations with spatial scale, degree of urbanicity, age, gender and cause of death. BMC Public Health. 2009;9:33. doi: 10.1186/1471-2458-9-33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Pace M, Lanzieri G, Glickman M, et al. Revision of the European Standard Population. Publications Office of the European Union; 2013. Available. [DOI] [Google Scholar]
  • 19.INSEE. https://www.insee.fr n.d. Available.
  • 20.Besag J, York J, Mollié A. Bayesian image restoration, with two applications in spatial statistics. Ann Inst Stat Math. 1991;43:1–20. doi: 10.1007/BF00116466. [DOI] [Google Scholar]
  • 21.Rue H, Martino S, Chopin N. Approximate Bayesian Inference for Latent Gaussian models by using Integrated Nested Laplace Approximations. Journal of the Royal Statistical Society Series B. 2009;71:319–92. doi: 10.1111/j.1467-9868.2008.00700.x. [DOI] [Google Scholar]
  • 22.Kumar A, Arora A, Sharma P, et al. Is diabetes mellitus associated with mortality and severity of COVID-19? A meta-analysis. Diabetes Metab Syndr. 2020;14:535–45. doi: 10.1016/j.dsx.2020.04.044. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Cariou B, Hadjadj S, Wargny M, et al. Phenotypic characteristics and prognosis of inpatients with COVID-19 and diabetes: the CORONADO study. Diabetologia. 2020;63:1500–15. doi: 10.1007/s00125-020-05180-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Bajos N, Counil E, Franck J-E, et al. Social inequalities and dynamics of the early COVID-19 epidemic: a prospective cohort study in France. BMJ Open. 2021;11:e052888. doi: 10.1136/bmjopen-2021-052888. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Bajos N, Jusot F, Pailhé A, et al. When lockdown policies amplify social inequalities in COVID-19 infections: evidence from a cross-sectional population-based survey in France. BMC Public Health. 2021;21:705. doi: 10.1186/s12889-021-10521-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Neelon B, Mutiso F, Mueller NT, et al. Spatial and temporal trends in social vulnerability and COVID-19 incidence and death rates in the United States. PLoS One. 2021;16:e0248702. doi: 10.1371/journal.pone.0248702. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Clouston SAP, Natale G, Link BG. Socioeconomic inequalities in the spread of coronavirus-19 in the United States: an examination of the emergence of social inequalities. Soc Sci Med. 2021;268:113554. doi: 10.1016/j.socscimed.2020.113554. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Zaldo-Aubanell Q, Campillo I López F, Bach A, et al. Community Risk Factors in the COVID-19 Incidence and Mortality in Catalonia (Spain). A Population-Based Study. Int J Environ Res Public Health. 2021;18:3768. doi: 10.3390/ijerph18073768. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.García CN. Socioeconomic, demographic and healthcare determinants of the COVID-19 pandemic: an ecological study of Spain. BMC Public Health. 2021;21:606. doi: 10.1186/s12889-021-10658-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Fortunato F, Lillini R, Martinelli D, et al. Association of socio-economic deprivation with COVID-19 incidence and fatality during the first wave of the pandemic in Italy: lessons learned from a local register-based study. Int J Health Geogr. 2023;22:10. doi: 10.1186/s12942-023-00332-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.McGowan VJ, Bambra C. COVID-19 mortality and deprivation: pandemic, syndemic, and endemic health inequalities. Lancet Public Health. 2022;7:e966–75. doi: 10.1016/S2468-2667(22)00223-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Williamson EJ, Walker AJ, Bhaskaran K, et al. Factors associated with COVID-19-related death using OpenSAFELY. Nature New Biol. 2020;584:430–6. doi: 10.1038/s41586-020-2521-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Fridman I, Lucas N, Henke D, et al. Association Between Public Knowledge About COVID-19, Trust in Information Sources, and Adherence to Social Distancing: Cross-Sectional Survey. JMIR Public Health Surveill. 2020;6:e22060. doi: 10.2196/22060. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Algan Y, Cohen D, Davoine E, et al. Trust in scientists in times of pandemic: Panel evidence from 12 countries. Proc Natl Acad Sci U S A. 2021;118:e2108576118. doi: 10.1073/pnas.2108576118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Diallo AI, Faye A, Tine JAD, et al. Factors associated with the acceptability of government measures to address COVID-19 in Senegal. Rev Epidemiol Sante Publique. 2022;70:109–16. doi: 10.1016/j.respe.2022.03.123. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Gadeyne S, Rodriguez-Loureiro L, Surkyn J, et al. Are we really all in this together? The social patterning of mortality during the first wave of the COVID-19 pandemic in Belgium. Int J Equity Health. 2021;20:258. doi: 10.1186/s12939-021-01594-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Clouston SAP, Rubin MS, Phelan JC, et al. A Social History of Disease: Contextualizing the Rise and Fall of Social Inequalities in Cause-Specific Mortality. Demography. 2016;53:1631–56. doi: 10.1007/s13524-016-0495-5. [DOI] [PubMed] [Google Scholar]
  • 38.Quinn SC, Kumar S. Health inequalities and infectious disease epidemics: a challenge for global health security. Biosecur Bioterror. 2014;12:263–73. doi: 10.1089/bsp.2014.0032. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Gibson M, Petticrew M, Bambra C, et al. Housing and health inequalities: a synthesis of systematic reviews of interventions aimed at different pathways linking housing and health. Health Place. 2011;17:175–84. doi: 10.1016/j.healthplace.2010.09.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Dubost C, Pollak C, Rey S. État des lieux et perspectives. Paris: Direction de la recherche, des études, de l’évaluation et des statistiques; 2020. Les inégalités sociales face à l’épidémie de covid-19; p. 40.https://drees.solidarites-sante.gouv.fr/sites/default/files/2020-10/DD62.pdf Available. [Google Scholar]
  • 41.Elliott J, Bodinier B, Whitaker M, et al. COVID-19 mortality in the UK Biobank cohort: revisiting and evaluating risk factors. Eur J Epidemiol. 2021;36:299–309. doi: 10.1007/s10654-021-00722-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.OECD . Health for Everyone?: Social Inequalities in Health and Health Systems. OECD Publishing; 2019. Available. [DOI] [Google Scholar]
  • 43.Riou J, Panczak R, Althaus CL, et al. Socioeconomic position and the COVID-19 care cascade from testing to mortality in Switzerland: a population-based analysis. Lancet Public Health. 2021;6:e683–91. doi: 10.1016/S2468-2667(21)00160-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Victora CG, Vaughan JP, Barros FC, et al. Explaining trends in inequities: evidence from Brazilian child health studies. Lancet. 2000;356:1093–8. doi: 10.1016/S0140-6736(00)02741-0. [DOI] [PubMed] [Google Scholar]
  • 45.Metzger C, Yaï J, Nassany O, et al. Factors associated with compliance with barrier gestures and social distancing measures in France. J Epidemiol Popul Health. 2024;72:202194. doi: 10.1016/j.jeph.2024.202194. [DOI] [PubMed] [Google Scholar]
  • 46.Svendsen MT, Bak CK, Sørensen K, et al. Associations of health literacy with socioeconomic position, health risk behavior, and health status: a large national population-based survey among Danish adults. BMC Public Health. 2020;20:565. doi: 10.1186/s12889-020-08498-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Martin-Lapoirie D, McColl K, Gallopel-Morvan K, et al. Health protective behaviours during the COVID-19 pandemic: Risk adaptation or habituation? Soc Sci Med. 2024;342:116531. doi: 10.1016/j.socscimed.2023.116531. [DOI] [PubMed] [Google Scholar]
  • 48.Denton M, Prus S, Walters V. Gender differences in health: a Canadian study of the psychosocial, structural and behavioural determinants of health. Soc Sci Med. 2004;58:2585–600. doi: 10.1016/j.socscimed.2003.09.008. [DOI] [PubMed] [Google Scholar]
  • 49.Montaut A. Études et Résultats; 2010. Santé et recours aux soins des femmes et des hommes: premiers résultats de l’enquête handicap-santé 2008.https://drees.solidarites-sante.gouv.fr/publications/etudes-et-resultats/sante-et-recours-aux-soins-des-femmes-et-des-hommes-premier Available. [Google Scholar]
  • 50.Nagata JM, Hernández-Ramos I, Kurup AS, et al. Social determinants of health and seasonal influenza vaccination in adults ≥65 years: a systematic review of qualitative and quantitative data. BMC Public Health. 2013;13:388. doi: 10.1186/1471-2458-13-388. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Bambra C. Pandemic inequalities: emerging infectious diseases and health equity. Int J Equity Health. 2022;21:6. doi: 10.1186/s12939-021-01611-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Blumenshine P, Reingold A, Egerter S, et al. Pandemic influenza planning in the United States from a health disparities perspective. Emerg Infect Dis . 2008;14:709–15. doi: 10.3201/eid1405.071301. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Affeltranger B, Potvin L, Ferron C, et al. Proportionate universalism: Towards “real equality” of prevention in France. Sante Publique. 2018;30:13–24. doi: 10.3917/spub.184.0013. [DOI] [PubMed] [Google Scholar]
  • 54.Marmot M, Allen J, Goldblatt P, et al. London: Institute of Health Equity; 2010. Fair society, healthy lives: the marmot review: strategic review of health inequalities in england post-2010.http://www.parliament.uk/documents/fair-society-healthy-lives-full-report.pdf Available. [Google Scholar]
  • 55.Wemrell M, Bennet L, Merlo J. Understanding the complexity of socioeconomic disparities in type 2 diabetes risk: a study of 4.3 million people in Sweden. BMJ Open Diabetes Res Care. 2019;7:e000749. doi: 10.1136/bmjdrc-2019-000749. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Ouedraogo E, Allard L, Bihan H, et al. The association of metabolic syndrome and COVID-19 deterioration. Nutr Metab Cardiovasc Dis. 2021;31:3236–42. doi: 10.1016/j.numecd.2021.08.036. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Lailler G, Fuentes S, Kab S, et al. Prevalence and risk factors associated with prediabetes and undiagnosed diabetes in France: The national CONSTANCES cohort. Diabetes Epidemiology and Management . 2023;10:100121. doi: 10.1016/j.deman.2022.100121. [DOI] [Google Scholar]
  • 58.Nafilyan V, Islam N, Mathur R, et al. Ethnic differences in COVID-19 mortality during the first two waves of the Coronavirus Pandemic: a nationwide cohort study of 29 million adults in England. Eur J Epidemiol. 2021;36:605–17. doi: 10.1007/s10654-021-00765-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Fuentes S, Hrzic R, Haneef R, et al. Identifying type 1/type 2 diabetes in medico-administrative database to improve health surveillance, medical research and prevention in diabetes: Algorithm development and application. Diabetes Epidemiology and Management . 2023;10:100137. doi: 10.1016/j.deman.2023.100137. [DOI] [Google Scholar]
  • 60.Lailler G, Piffaretti C, Fuentes S, et al. Prevalence of prediabetes and undiagnosed type 2 diabetes in France: Results from the national survey ESTEBAN, 2014-2016. Diabetes Res Clin Pract. 2020;165:108252. doi: 10.1016/j.diabres.2020.108252. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

online supplemental file 1
bmjph-4-2-s001.docx (26.9KB, docx)
DOI: 10.1136/bmjph-2025-004258

Data Availability Statement

Data are available on reasonable request.

Articles from BMJ Public Health are provided here courtesy of BMJ Publishing Group

RESOURCES