An extensive body of research has shown that vulnerable populations are consistently at greater risk of contracting SARS-CoV-2 and experiencing more severe symptoms (1). Ensuring equal access to COVID-19 vaccination remains a key public health challenge.

Numerous studies from different parts of the world have demonstrated that non-use of COVID-19 vaccination is more prevalent among populations with lower educational attainment, lower incomes, and among racialized minorities (2–10, 77).

The reasons behind non-use of vaccination have been the subject of extensive analysis. One set of factors concerns access to healthcare. Physical access to vaccination sites has been identified as a barrier to preventive practices (9, 11, 12). Financial constraints have also played a role, particularly in countries where vaccines were not provided free of charge (13). Other studies have emphasized that distrust toward healthcare professionals can also limit vaccine uptake (14).

Beyond these classic socio-economic determinants of healthcare non-use (15, 16), other distributed—namely, the unprecedented speed of its rollout, occurring barely a year after the pandemic began. In a context where governments and scientists assumed leading roles in managing the health crisis, it is crucial to analyze vaccination behavior in relation to levels of trust in governmental institutions (17, 18). In France, as in many other countries, the government strongly relied on scientific expertise to justify its pandemic response. Therefore, studying the impact of trust in government on vaccination behavior also entails accounting for trust in scientists (19).

Since disadvantaged social groups are often more distrustful of government institutions (20), the strong governmental involvement in vaccination programs—and the resulting politicization of these efforts—may have exacerbated social inequalities in vaccine uptake. Lack of trust in both government and scientific authorities has thus emerged as a significant barrier to COVID-19 vaccination (13, 21–24).

Another defining feature of the COVID-19 crisis has been the widespread circulation of opinions on social media, which fueled public controversies and the proliferation of conspiracy beliefs (25, 26). Several studies have shown that such conspiracy beliefs are associated with lower levels of vaccine uptake (27, 28).

Taking into account these various social determinants of non-uptake of vaccination calls for the examination of a dimension that remains relatively underexplored in the COVID-19 literature—namely, the impact of experiences of discrimination on vaccination behavior. Numerous studies have shown that populations subjected to racial discrimination are more likely to report general vaccine hesitancy (29–34). However, research examining the relationship between broader experiences of discrimination and vaccination behavior remains much more limited. In France, as in many other countries, those most reluctant to receive the vaccine were, as previously underlined, often drawn from the most socioeconomically disadvantaged populations, those with lower levels of educational attainment, and racialized minority groups (19, 35). These same social groups are also those most frequently exposed to various forms of discrimination (36, 57). However, what about individuals who have faced forms of discrimination other than those associated with race or socioeconomic status?

The World Health Organization defines structural discrimination as “the rules, norms, routines, patterns of attitudes and behaviors in institutions and other societal structures that represent obstacles to groups or individuals in achieving the same rights and opportunities as those available to most of the population,” and identifies it as a major determinant of health equity (16).

A growing body of research has established strong links between discrimination and health outcomes (37–40). Discrimination results from both individual behaviors and institutional practices, which in either case create barriers to healthcare access and are associated with lower quality of care. Structural discrimination also has broad and cumulative negative effects on key social determinants of health, including increased poverty and reduced access to education, employment, healthcare services, and safe and adequate living environments (41).

The few studies that have examined the impact of discrimination on vaccination practices have mostly limited their analyses to racialized minority groups. These studies attribute greater vaccine hesitancy among these populations to factors such as language barriers, fear of deportation, and limited access to vaccination (12), to experiences of everyday racism (42), or to heightened distrust of the healthcare system, partly resulting from encounters with healthcare professionals perceived as discriminatory (43, 44). Only one study—conducted by Wulkotte in Germany—has shown that experiences of discrimination are associated with lower vaccination uptake not only among migrant populations but also among non-migrants (45).

Experiences of discrimination that may influence vaccine non-use are not limited to belonging to racialized groups and can also be rooted in place of residence, particularly for individuals living in medical deserts (areas with severely limited access to healthcare services) (46). This spatial dimension of discrimination may intersect with racial criteria, as illustrated by public policies that have allowed environmental degradation in Overseas Territories, often at the expense of residents' health. One striking example is the widespread use of chlordecone, a carcinogenic pesticide banned in the United States as early as 1973 and in mainland France by 1990, but which continued to be used in the French Caribbean until 1993 to boost banana crop yields. The denunciation of the health and environmental consequences of chlordecone exposure has since become a symbol of protest against the discriminatory treatment of overseas populations in public policy decisions made from metropolitan France (47).

Discriminatory practices stemming from healthcare professionals' own judgments can also play a significant role in the quality of medical care provided. For instance, in clinical situations where symptoms and severity are identical—such as chest pain—men tend to be taken more seriously than women, and white patients more than those belonging to racialized minorities (48). Similarly, emergency physicians encountering Roma patients—who combine “ethnoracial otherness” with socioeconomic, residential, and administrative precarity—often attribute their more frequent use of hospital services to “cultural difference,” despite their reasons for visiting emergency departments being comparable to those of the general population (49). These discriminatory processes are frequently based on healthcare providers' perception of otherness, leading to mechanisms of delegitimization and differentiation (50). Such dynamics have also been observed in the management of hypertensive disorders during pregnancy among women of sub-Saharan African origin (51). Discrimination related to sexual orientation or gender identity can also have a significant impact on healthcare access (52, 53).

To better understand the relationship between discrimination and COVID-19 vaccination, we propose two central hypotheses. Structural discrimination occurs systematically across society and is embedded in dominant cultural norms, legal systems, and economic structures—not solely within the healthcare sector. Our first hypothesis is that the relationship between experiences of discrimination and non-vaccination may vary depending on the type of discrimination experienced (54). Such experiences may occur in interactions with healthcare professionals, but also in areas such as access to employment, housing, or public services, core domains where discriminatory practices are well documented (57). The key issue here is to examine how these experiences influence vaccination practices, with the understanding that discriminatory treatment does not have the same consequences for attitudes and practices depending on whether or not it occurs in relation to institutional representatives. Discrimination experienced within institutional settings may have a particularly strong impact on trust and compliance with public health recommendations.

Our second hypothesis posits that frequent exposure to discrimination reduces the likelihood of vaccine uptake. Experiencing discrimination undermines trust and confidence in a society's social, economic, political, and cultural systems (16). Discrimination acts as a social reminder that signals to affected individuals that they do not occupy the same place in society as others. As Pollak (55) showed in his study on gay men during the AIDS epidemic, the repeated reminder of belonging to a stigmatized minority can foster a sense of social exclusion that is incompatible with engagement in preventive health behaviors. We thus hypothesize the existence of a cumulative effect of discrimination on vaccine non-use.

Based on a large-scale quantitative survey conducted among the general population, this article aims to explore the relationship between experiences of discrimination and non-use of COVID-19 vaccination. Unlike previous studies that focus on specific subpopulations, our analysis considers a wide range of discriminatory contexts—including interactions with healthcare professionals, employment, housing, and public services—as well as the frequency of such experiences.

Although COVID-19 vaccine uptake was high among the population in France, a significant minority remained reluctant to take up this free and widely accessible intervention. Our results indicate that experiences of discrimination are associated with specific and cumulative differences in vaccination practices: discrimination encountered during interactions with healthcare professionals, and to a lesser extent with public services, was more strongly associated with non-vaccination than discrimination related to employment or housing. Importantly, these associations were observed across all groups, whether racialized or not, as shown by the analysis of factors associated with non-vaccination that accounts for the interaction between discrimination and racialized minority status, even though experiences of discrimination alone do not fully explain the lower vaccination rates among racialized groups.

The social profile of unvaccinated individuals in autumn 2022 was similar to that observed during the previous wave of the EpiCov survey conducted in summer 2021 (59). This stability suggests that vaccine non-uptake does not primarily reflect delays in access or temporary barriers, but rather represents a sustained pattern of non-vaccination from the vaccine's widespread availability in June 2021 through the administration of the questionnaire in autumn 2022. The results indicate that this non-use is partly explained by lived experiences of discrimination occurring across various social spheres (employment, health, public services) and linked to diverse grounds (sex, age, weight, etc.).

Whereas the existing studies on the relationship between discrimination and vaccine non-uptake have limited their analyses to racialized individuals disadvantaged by the healthcare system (12, 42, 60–63), our approach considers multiple types of discrimination across different social domains. Our findings reveal that discrimination experienced within healthcare settings, during interactions with public services, and in the workplace, each increases the likelihood of being unvaccinated, indicating that discrimination experiences may play a role in shaping vaccine acceptance. More specifically, discrimination in housing did not appear to affect vaccination practices, whereas the association was notably strong for discrimination in healthcare and, to a lesser extent, in public service contexts. In other words, discrimination originating from social or state institutions exerted the greatest impact on vaccine non-uptake.

One might hypothesize that these results primarily reflect a heightened distrust of public authorities among the least vaccinated social groups, both in general and specifically regarding their management of the COVID-19 pandemic. However, analyses controlling for trust in government showed that the associations with experiences of discrimination remained consistent, with significant odds ratios (see Appendix 5). The strong influence of discrimination experiences in healthcare—and to a lesser extent in public services—may also relate to the unique status of the COVID-19 vaccination campaign, which was both recommended by healthcare professionals and heavily promoted by public authorities.

This finding aligns with previous research showing that trust in healthcare professionals plays a key role in vaccine hesitancy (8, 14, 64). Individuals reporting discrimination tend to have lower trust in these actors (65). However, the concept of trust in the healthcare system is complex and difficult to interpret (66), as perceptions vary widely depending on geographical location (with significant territorial inequalities), institutional status (public hospital vs. private clinic), and professional status (private practitioner vs. hospital-based clinician).

The results also highlighted the cumulative influence of discrimination on vaccine non-uptake, which increased with the frequency of discriminatory experiences—except in cases involving health professionals, where even a single encounter appears sufficient to generate distrust toward the healthcare system. In the literature, this association has primarily been demonstrated for individuals belonging to racialized minorities: racial discrimination is presented as a key factor explaining the lack of trust in health authorities among these minorities, and this mistrust is itself linked to lower vaccination uptake (30, 42, 67). Our findings confirm this pattern: racialized minorities were particularly affected by discrimination from institutional actors and were also among the groups with the lowest vaccination rates. While immigrants from racialized minority groups exhibited lower vaccination rates compared to the majority population, vaccine non-uptake was even more pronounced among their descendants.

Interestingly, although women reported higher rates of discrimination than men, their COVID-19 vaccine uptake was higher. A more detailed analysis reveals that this exception is not truly one. Among women reporting no discrimination in the 5 years preceding the survey, 5.7% remained unvaccinated, whereas this proportion rose to 9% for those discriminated against in the workplace, 12.3% for those facing discrimination in contact with public services, and 13.9% in healthcare settings. Thus, even among women, a gradient was observed according to the sphere in which discrimination occurred, in relation to vaccination status. Even though women face high discrimination rates, their strong pro-vaccine behavior tied to their gender identity (likely linked to social norms of care-giving or higher engagement with preventive health) outweighs the negative impact of their discrimination experiences, resulting in a net positive uptake rate.

Multivariate analysis showed that the association between experiences of discrimination and vaccine non-uptake was not limited to racialized minorities but was observed across all social groups, consistent with findings from Wulkotte's study (45). Hence, adopting a broad spectrum of discrimination experiences makes it possible to show that their consequences are not limited to racialized populations. However, since these minorities are more frequently exposed to discrimination overall, they are consequently more likely to be affected by the link between discrimination and vaccine non-uptake.

As with any general population survey, the EpiCov study has limitations that should be acknowledged. Homeless individuals were not included in the sampling frame, as they are not affiliated with the tax registry that served as the sampling base for this survey. Yet, these individuals face major barriers to accessing healthcare in general and vaccination in particular (11, 68). Consequently, the rate of vaccine non-uptake is likely underestimated, and the social determinants of vaccine non-uptake within these groups may differ from those identified in this study.

Moreover, to avoid recall bias regarding distant episodes of discrimination, we limited the analysis to events occurring within the past 5 years without further chronological specification. However, we cannot exclude that other significant events, such as the deterioration of a close relative's health due to SARS-CoV-2 infection, may have occurred since the discrimination episode and also influenced vaccination decisions.

It is also important to note that our analyses rely on self-reported experiences of discrimination. Some may argue that questionnaire responses do not necessarily reflect the objective reality of lived experiences. This methodological challenge is well-known among researchers in this field (69). We adopted formulations used in specific French surveys on discrimination, which have been shown to produce results consistent with records of discrimination by institutional bodies tasked with combating discrimination, such as the Defender of Rights (58).

Finally, we did not account for the sources of information individuals used regarding vaccination. Misinformation campaigns have been widespread in France and other countries, especially on social media (70). It can be assumed that socially disadvantaged individuals were most exposed to such misinformation (10). Since vaccine uptake is lowest in these groups, including this factor might have enhanced our understanding of the drivers of vaccine non-uptake.

Considering a broad spectrum of discrimination types and their association with vaccine non-uptake has revealed the structuring role of experiences perceived as injustices, particularly in encounters with institutions meant to guarantee and embody equal treatment. Our findings confirm that inequalities in healthcare access—even when services are free and widely available, as was the case with COVID-19 vaccination—cannot be attributed solely to beliefs or cultural differences. Instead, they may be associated with discriminatory practices that have shaped the worldview of those affected.

This result nuances studies that interpret vaccine hesitancy solely as a consequence of misinformation (71) or susceptibility to conspiracy theories (72). The significant role of discrimination, particularly in healthcare settings, underscores the importance of past experiences in shaping relationships with the body and medicine. Health inequalities are not only a striking manifestation of the embodiment of social structures (73) but also the product of unequal and discriminatory experiences in care and treatment.

COVID-19 vaccination has prevented approximately 160,000 deaths and 1.5 million hospitalizations in France (74), and nearly twenty million deaths worldwide (75). These facts should be considered when designing future vaccination campaigns (76). Moreover, integrating lived experiences with state and healthcare institutions may help explain observed social disparities in the uptake of preventive measures such as cervical or colorectal cancer screening and other widely accessible health goods that remain unevenly utilized.