The COVID-19 pandemic has not only highlighted the importance of infectious diseases in global health, but has also exacerbated existing social inequalities (1–3). Studies from multiple countries have reported higher COVID-19 risk among the socially deprived individuals (4–7). Across Europe, low-income groups and racialized minorities have been found particularly affected during the pandemic (8). In the United States, higher socioeconomic status (SES) was linked with earlier SARS-CoV-2 infection, while later on, higher COVID-19 incidence was found in lower SES group (9). In the United Kingdom, Black and South Asian populations were at greater risk of infection, due to overcrowded housing, and front-line occupational exposure (10). Similarly, in France, marginalized populations, including racialized minorities and those living in the most deprived areas experienced disproportionately high infection rates, alongside healthcare professionals (11–13). These inequalities were likely due to occupational exposure under lockdown measures, and overcrowded living conditions (1, 11). However, the social pattern of infection has changed over place and time, and the underlying mechanisms are yet not fully understood.

Explaining these complex disparities necessitates a research framework that goes beyond single-axis analyses and considers the intersecting influence of diverse social mechanisms. The concept of intersectionality offers such framework to understand how multiple social relations (e.g., race, ethnicity, gender, class) intersect at the micro level of individual experience and reflect broader systems of privilege and oppression at the macro, social-structural level (e.g., racism, sexism, classism) (14, 15). Scholars have advanced contemporary understandings of intersectionality by explaining how multiple forms of oppression interlock, creating patterns different from singular forms of discrimination or their mere addition. They emphasized that deep-rooted systems of power and oppression perpetuate domination, privilege, and persistent inequality (15–17).

This research framework has been widely applied in qualitative studies to examine social inequalities; however, its integration into quantitative health research is more recent (18). Likewise, adopting an intersectional approach is essential for understanding and addressing inequalities in the context of the COVID-19 pandemic (19). Although a few studies attempted to do so for other COVID-19 outcomes such as vaccine uptake (20), mortality (21), and mental health (22, 23), similar analyses for the risk of infection remains, to our knowledge, largely underexplored.

As a mini review this paper followed a traditional or narrative literature review approach (19, 24–26), which allows flexibility in exploring and synthesizing existing research on a rather large topic. We provide an overview of the existing quantitative evidence by conducting a search of peer-reviewed empirical studies examining social inequalities in COVID-19 infection in high income countries, with a particular focus on France. To identify relevant articles we searched Google Scholar and PubMed using different combinations of the keywords divided into four groups: (1) COVID-19 related terms (“COVID-19,” “SARS-CoV-2 infection,” “Seroprevalence of SARS-CoV-2”); (2) intersectionality related (“intersectionality,” “intersectional”); (3) social dimensions (“social inequalities,” “socioeconomic inequalities,” “gender,” “race,” “ethnicity,” “class,” “social class,” occupational class”); and (4) geographic scope (“high income countries,” “France”). At least one term from each keyword group was consistently included in the search strategy, for example: “intersectionality AND gender AND race AND class AND COVID-19 AND France.” We also identified articles using a snowball search method by reviewing the reference lists of studies retrieved from the initial keyword search. We included both original research and review articles published in English and French, between 2020 to September 2024 that were relevant for our research questions. Firstly, we summarize key findings about inequalities across different social groups, then we report on common methodological approaches, and finally, we identify research gaps and needs regarding the adoption of an intersectional lens.

The existing literature provided valuable insights into social inequalities in the risk of SARS-CoV-2 infection. However, it was predominantly focused on unidimensional analyses examining social determinants such as gender, race/ethnicity, income or education in isolation. These studies employed a variety of statistical methods, ranging from descriptive analyses to advanced modeling techniques, to explore the effects of different types of power relations on infection risk. Many studies begin with descriptive analyses to summarize infection rates across different socio-demographic groups. For example, studies often compared infection rates/seroprevalence by sex, income levels, occupations, or geographic regions, providing a baseline understanding of disparities (29, 39, 40). Further, it was noted that in a majority of studies, multivariate analyses such as linear, logistic and multinomial regression models were commonly used to examine the association between social determinants and the risk of infection, net of other factors (13, 28, 32, 40, 58, 59, 63–65). These models estimated the likelihood of infection based on explanatory variables, both individual and contextual, such as age, sex/gender, income, education, housing conditions, migration status, comorbidities, occupational exposure, region/state, etc.

Some studies also included stratified and interaction analyses that provided additional insights into the variability of risks across different subgroups (28, 29, 61, 62, 65, 66). Furthermore, more advanced methods such as mediation analyses, Poisson regression model, Cox proportion hazard model, negative binomial and machine learning models has also been applied in a few studies (12, 58, 66–70). However, none of the existing studies incorporated an intersectional framework, often overlooking the compounded and intersecting effects of multiple social positions, limiting an in-depth understanding of the mechanisms underlying inequalities in infection risk.

To the best of our knowledge, only a few studies from the United States, United Kingdom and Europe have incorporated intersectionality for examining inequalities in the context of the pandemic focusing on outcomes such as mortality, mental health or vaccination (20–23, 71, 72). However, no such evidence was found from France. Using an intersectional approach in the analyses, some of these studies highlighted how socially constructed categories shaped by structural power relations, intersect to produce intricate layers of advantage and disadvantage. For example, the study by Morales (71) conducted in the US, demonstrated how gender intersected with individuals’ socioeconomic status (measured by education, household income and employment status) to influence COVID-19 vaccine hesitancy. They further found that poverty and employment influenced vaccine hesitancy among women but not men (71). However, one of the key limitations of the stratification is that it does not fully capture the complex interactions between multiple social dimensions as they are experienced within different subgroups, which can influence study outcomes (73). In a study from Sweden (20), researchers examined sociodemographic disparities in COVID-19 vaccination uptake using national register data. Logistic regression models were applied, along with an intersectional approach that included multiple cross-classified subgroups. The intersectional variable was constructed by combining categories of age, sex, income, country of birth, and occupational status, resulting in 72 strata. The findings showed that non-vaccinated individuals were more likely to be younger, male, have lower income, be unemployed, or born outside Sweden. Vaccine coverage ranged from 32% to 96% across intersectional strata/subgroups. These findings highlighted that focusing on single sociodemographic factors and group averages, without accounting for differences/heterogeneity within groups, may obscure important variations likely explained by complex social mechanisms (20).

Further, a study examined how intersecting demographic and socioeconomic factors shaped health-related quality of life (HRQoL) and mental health in children and adolescents in Geneva, Switzerland during the pandemic (22). The study applied Multilevel analysis of individual heterogeneity and discriminatory accuracy (MAIHDA) model by nesting individuals within 48 intersectional social strata based on sex, age, immigrant background, parental education, and financial hardship in Bayesian multilevel logistic models. Mental health was assessed using items from the Pediatric Global Health (PGH-7) and the Strengths and Difficulties Questionnaire (SDQ), while health-related quality of life (HRQoL) was measured using the short-form version of the Pediatric Quality of Life Inventory (PedsQL). The outcome variables of the study were poor HRQoL, poor parent-reported mental health and mental health difficulties. The range of the predicted probability for poor HRQoL varied between 3.4% (for 6–11 years old Swiss girls, with highly educated parents and no financial problems) to 34.6% (for 12–17 years old non-Swiss girls with highly educated parents and financial problems). The analyses highlighted those strata involving adolescents and financial problems consistently showed worse HRQoL. Whereas, in contrast, predicted frequency of mental health difficulties (based on the SDQ score) range varied less across groups (4.4–6.5%). Therefore, the intersectional analysis revealed diverse and outcome-specific social patterns in HRQoL and mental health and suggest that post-pandemic efforts to reduce HRQoL inequities should focus on adolescents from financially disadvantaged families, while mental health interventions should broadly support all children and adolescents (22).

A similar kind of study from the United Kingdom aimed to examine mental health inequalities among young adults during the pandemic (23). The study assessed mental health using measures of anxiety, depression, loneliness, and life satisfaction and formed intersectional strata using categories tied to social power, including age, sex, race/ethnicity, sexual orientation, and socioeconomic position. Using the MAIHDA method (23), it explored patterns of mental health measures across multiple intersecting positions and examined whether these intersections revealed effects beyond those attributable to any single position considered separately. The analysis revealed significant mental health inequalities across different intersectional strata. Much of these inequalities were driven by the additive effects of the variables used to define the intersections, with some of the largest gaps associated with sexual orientation, followed by cohort/generation, and birth sex with sexual minorities, females, and younger people (in their teens/20s) showing worse levels. Further, intersectional effects were reported mostly in intersections defined by the combinations of marginalized and privileged social positions (e.g., lower than expected life satisfaction was observed in South Asian men in their 30s from a sexual minority and disadvantaged childhood social class, followed by Black heterosexual men in their 30s from an advantaged childhood social class) (23). These sub-groups would likely have remained invisible in a fully categorical approach. Compared to the classical approaches, the results of the study make significant contribution to the literature by highlighting those inequalities are not confined to groups with exclusively advantaged or disadvantaged positions and may vary depending on the context. These intersectional effects reflect the influence of marginalization and/or privilege shaped by the interlocking systems of oppression in the studied outcomes (74) and would have been made difficult to unveil otherwise.

Intersectional analyses challenge the notion that inequalities are confined solely to groups with uniformly advantaged or disadvantaged positions. Instead, they emphasize that such disparities are contingent on specific social contexts and interactions. Moreover, such analyses provide a detailed understanding of intersections that might remain undetected while using classical approaches, and provide estimates of the heterogeneity within those intersections. This highlights the importance of Hankivsky’s theorization of intersectionality that the significance of specific combinations of social (power) relations might be better revealed inductively during the research process rather than predetermined (75). The examples (20, 22, 23) pointed above and their strong conceptual foundations highlight the crucial importance of including an intersectional lens while analyzing social inequalities in the risk of SARS-CoV-2 infection. This is likely to provide a critical framework for structuring research questions and hypotheses, emphasizing the need to move beyond assumptions of homogeneity across intersecting social categories, and help to solve enigmas of inconsistent results across studies conducted in apparently similar contexts, such as those we reported about gender considered in isolation.

In spite of its promises, there are some challenges attached to the implementation the framework in empirical studies. Researchers already working within this paradigm have highlighted the lack of clear guidelines for quantitative approaches to effectively capture the complexity of intersecting social positions (14, 76). However, recent review papers have made significant progress in addressing this gap and laid the foundation by discussing theoretical and analytical approaches that are well-suited for studying intersectionality in quantitative health research (18, 73, 77, 78). For instance, researchers emphasized the importance of including intersectional group measures beyond sex/gender and race, such as sexual orientation, disability, religion, nativity, and immigration status, in order to capture a broader spectrum of social inequalities. Further, it is suggested to explicitly define the intersectional positions analyzed, their relation to social power, and to ensure alignment between theoretical approaches, methods, and interpretations. The inclusion of McCall’s intersectionality frameworks- anticategorical (deconstructing traditional categories), intracategorical (focusing on underexplored intersections), and intercategorical (using existing categories to study inequality) was recommended, with an emphasis on their explicit application and clear justification for their use (79). Moreover, in this context, Public Health Agency of Canada has also provided a checklist to strengthen the integration of intersectionality theory in quantitative health inequality analyses (78).

Another challenge, specific to the risk of SARS-CoV-2 infection and also a potential limitation of this review, is the likely complex dynamics over place and time, particularly in relation to the introduction of vaccination. The timing of vaccine rollout varied across countries and likely played an important role in shaping inequalities. While we aimed to include papers covering the pre-vaccination phase, it is possible that some studies may have combined data from pre and post-vaccination periods. This may complicate the interpretation of infection patterns and the comparability of findings across studies. Future research should account for these temporal shifts to better capture the structural drivers of infection risk.

In the French context, the lack of intersectional research on SARS-CoV-2 infection risk is particularly compelling, given the evidence of significant social inequalities during the pandemic. As discussed above, studies have documented higher infection risks among low-income people, and racialized minorities but no significant gender difference, yet these analyses were unidimensional. Incorporating an intersectional lens by accounting for multiple social dimensions would provide a more holistic understanding of these inequalities and their dynamics. In line with previous work (19, 80, 81), supporting the adoption of an intersectional perspective in analyzing the effects of social inequalities related to the COVID-19 crisis, such approaches are also crucial for informing policy responses to effectively tackle future pandemics. To address this gap, future research could use data from the French EpiCoV cohort which offers strong potential to analyze and uncover new insights into social inequalities and serve as a valuable foundation for studies in other high-income countries.