This cross-sectional study was conducted in France from October 2023 to February 2024. This study was performed in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) [27] checklist (Table S1 in Multimedia Appendix 1).

An email invitation was sent out to 348,715 civil servants with a detailed information letter about the study. Persons who agreed to participate and who met the inclusion criteria were included. For people who agreed to participate, a questionnaire was used to check the inclusion and exclusion criteria. Participants completed a questionnaire about their sociodemographic characteristics (eg, age, sex, education level, professional category, and frequency of dental visits), for which the answer to each question was not compulsory in order to respect people’s privacy. Then, they completed 3 questionnaires (in French): the OH knowledge questionnaire [30], the Oral Health Literacy Instrument (OHLI) [30] and the Short Test of Functional Health Literacy in Adults (s-TOFHLA) [33].

The main outcome was to determine the level of OHL of civil servants in France. The secondary outcomes were to determine the level of OH knowledge; to determine the level of general HL; to identify the determinants of an adequate level of OHL, OH knowledge, and HL; and to assess the correlation between the level of OHL and the level of HL.

The education level was categorized as level 1, or low (<baccalaureate); level 2, or moderate (from baccalaureate to baccalaureate plus 2 years); and level 3, or high (>baccalaureate plus 2 years). Data were analyzed using R (version 3.6.0; R Foundation for Statistical Computing). Sociodemographic numerical and categorical variables were calculated and reported as number (percentage) and mean (SD), respectively. The scores for OH knowledge, OHLI-F, and s-TOFHLA were calculated and reported as mean (SD) and 95% CIs. The influence of the variables (eg, age, gender professional category, and professional situation) on the OHLI scores (OHLI-F, OH knowledge, and s-TOFHLA) was tested using a classic ANOVA. This analysis was supplemented by logistic regressions where the variables to be explained were binarized scores using a threshold equal to 75 (1 if the score is greater than or equal to 75, 0 otherwise). The R packages gtsummary and ggstats were used in these analyses.

The comparison between the OHLI-F reading comprehension and numeracy scores was performed using the Welch 2-sample t test and a paired t test (both 2-tailed).

Concerning the correlation matrix between scores, the Pearson correlation parameter was computed and tested for nullity via a usual t test. The correlation was assumed low or null for scores between 0.00 and 0.25, low for scores between 0.26 and 0.49, moderate for scores between 0.50 and 0.69, high for scores between 0.70 and 0.89, and very high for scores between 0.90 and 1.00 [34].

The study was conducted in accordance with the Declaration of Helsinki and approved by the Research Ethics Committee of the University Hospital Centre of Rouen (E2023-4; October 15, 2023). The platform Claroline connect from the University Claude Bernard Lyon 1 used for the online questionnaire was in accordance with the Regulation EU 2016/679 of the European Parliament and the Council of 27 April 2016 on the protection of individuals with regard to the processing of personal data and on the free movement of such data.

Figure 1 represents the flowchart of the study. A total of 348,715 persons were assessed for eligibility, among whom 343,236 were excluded (n=1639 did not meet the inclusion criteria and n=341,597 refused to participate). Finally, 5479 persons were included and 1917 completed the 3 questionnaires.

Table 1 presents the sociodemographic characteristics and behavioral attitudes to OH management of the participants. Participants had a mean age of 45.6 (SD 14.7) years. Among participants, 54.2% were women (n=1038), 70.9% were in a relationship (n=1359), and 52.8% were retired (n=1012). For education, 47.4% of participants had a level of 3 (n=909). Most of the participants reported visiting a dentist annually (n=1336, 69.7%), within the last 6 months (n=888, 46.3%), and for a simple check-up (n=1508, 78.7%).

Table 2 presents descriptive statistics for the OHLI-F, OH knowledge questionnaire, and s-TOFHLA.

OHLI-F scores ranged from 15.8 to 97.4, with a mean of 80.9 (SD 7.9). An adequate OHLI-F score was obtained by 84% of participants (n=1610). For the reading comprehension part, the participants obtained a mean score of 43.4 (SD 3.6), while for the numeracy part, the mean score was 37.5 (SD 6.0). The OH knowledge score ranged from 82.4 to 94.1, with a mean 87.6 (SD 7.9); the score was adequate for 90.6% of participants (n=1736). Consequently, 9.5% (n=181) and 16% (n=307) of participants had inadequate or marginal OH knowledge and OHLI-F scores, respectively. The scores for s-TOFHLA were higher than for OHLI-F, and s-TOFHLA scores ranged from 97.2 to 100.0, with a mean of 98.2 (SD 2.8). An adequate s-TOFHLA score was obtained by 99.9% of participants (n=1915).

The OHLI-F reading comprehension score was significantly higher than the OHLI-F numeracy score (P<.001 with a Welch 2-sample t test and paired t test).

The variance analysis of OHLI-F is presented in Table 3. Among the sociological variables analyzed, age, education level, and professional category significantly impacted OHLI-F, OH knowledge, and s-TOFHLA levels. The score for OH knowledge was also significantly impacted by gender.

The levels of OHL, OH knowledge, and general HL according to age, education level, and professional category are presented in Figure 2. Scores increased between ages 19 and 49, then decreased after age 49. They increased with the level of education and with the professional category.

Table 4 presents the matrix of correlations between OHLI-F, OHLI knowledge, and s-TOFHLA. OHLI-F was highly correlated with OH knowledge, but OHLI-F and OH knowledge had a low correlation with s-TOFHLA.

The WHO has emphasized the need to combat the burden of oral diseases and to include oral diseases in initiatives to prevent NCDs [1]. Improving OH and reducing OH inequalities is linked to the level of OHL [35]. It is therefore essential to know the level of OHL in order to implement appropriate preventive measures [36,37]. In order to reach a large number of people, these preventive actions must be deployed in the workplace. In France, civil servants represent several socioprofessional levels with different levels of education; they therefore represent an excellent study model with results that are transferable to the general population as a whole.

To our knowledge, this study is one of the largest studies in terms of number of participants analyzing OHL and general HL levels [31,38]. In addition, it is the first to focus on civil servants in France. Thus, this study represents an antecedent to the implemention of efficient preventive actions in the workplace. The population studied had a number of strengths that allow the results to be applied to the general population. First, this population included people of all ages (18 years to more than 80 years), marital status, and levels of education. Second, the public service population covers the majority of socioprofessional categories, with a wide variety of public occupations (eg, clerks, customs officers, librarians, police officers, state architects, magistrates, port officers, and prison staff). Third, this population is composed of different level of socioprofessional categories; the 3 categories (A, B, and C) cover jobs with different responsibilities, qualifications, remuneration, and recruitment conditions [39,40]. In this study, 51.7% (n=990) of participants belonged to category A, 35.5% (n=681) to category B, and 12.8% (n=246) to category C, which is in line with the distribution of civil servants in France who belong to category A (56%), B (24%), and C (21%) [24]. Fourth, this population has access to workplace prevention initiatives even after retirement.

Our results indicated that civil servants had an OHLI-F mean score of 80.9 (SD 7.9) and an OH knowledge mean score of 87.6 (SD 10.5). The majority had an adequate level of OHL (n=1610, 84%), OH knowledge (n=1736, 90.6%), and HL (n=1915, 99.9%). Consequently, 9.5% (n=181) and 16% (n=387) of participants had inadequate or marginal OH knowledge and OHLI-F scores, but none had an inadequate level of general HL. A low correlation between HL and OHL was observed. French civil servants had a higher level of HL than OHL and had a significantly higher level of OHLI-F reading comprehension (mean 43.4, SD 3.6) than OHLI-F numeracy (37.5, SD 6.0).

Age, education level, and professional category were the 3 determinants of the level of OHL, OH knowledge, and HL. Levels increased with age up to 49 years, with education level, and with professional category (C < B < A). Conversely, the professional situation (active or retired) did not represent a determinant of the level of OHL, OH knowledge, and HL. Working people obtained nonsignificantly higher OHLI scores than retirees, while no difference was observed between the 2 groups for the OH knowledge score.

More particularly, the determinants of an adequate level of OHLI were age, education level, and occupational category. For OH knowledge, persons in a relationship had a significantly higher score, and persons with a low professional category (B or C) had a less-adequate score.

Regarding the level of OHL and OH knowledge, civil servants had a higher level than observed in the French population, obtaining an OHLI mean score of 73.5 (SD 13.6) and a mean OH knowledge score of 79.8 (SD 15.3) [30]. This difference could be due to the fact that the sample size was smaller in the study by Clément et al [30]. Compared with other countries, OHLI scores are higher in Canada (87.2, SD 10.2) [28] and the United States (81.6, SD 13.6) [41] but lower in Russia (77.2, SD 14.5) [42], Malaysia (75.1, SD 15.6) [29], and Chile (37.1, SD 9.0) [43]. The low score observed in Chile may be due to the fact that the Spanish OHLI questionnaire used was not well adapted, because there are phonological, grammatical, and lexical variations in Spanish in Hispanic America [43]. The OH knowledge score observed in the French civil servants was higher than in the Russian population (63.8, SD 16.9), American population (60.8, SD 3.1), Canadian population (57.0, SD 26.1), and Chilean population (10.0, SD 3.6).

Regarding the reading comprehension and numeracy scores, in all countries that measured OHLI with this scale, the scores were similar [28,30,41-43], except for France [30]. Indeed, in France, for the population as a whole, the reading comprehension score (41.4, SD 6.4) was higher than the numeracy score (34.2, SD 8.5) [30], as observed in the population of civil servants in this study. Previous studies have shown that low numeracy reduces adherence to medication, skews perceptions of the benefits and risks of screening, hinders access to treatment, undermines risk communication (restricting prevention initiatives for the most vulnerable), and seems to have a negative effect on medical outcomes [44]. Low numeracy is also associated with higher sensitivity to external factors that do not affect objective numerical data but are linked to biases in judgment and decision-making (eg, the influence of framing and ratio bias effects) [45]. Thus, in France, health promotion actions should consider this specifically and should target the improvement of numeracy. For this, prevention initiatives targeting people with low numeracy should be implemented with the aim of improving their understanding of digital information. These actions would benefit all patients, even those with high numeracy [46].

Regarding the determinants of the level of OHL and HL, the education level was also underlined in the studies of Clément et al [30], Ramlay et al [29], McCarlie et al [41], and Blizniuk et al [42]. In the study of Sabbahi et al [28], the education level did not have a significant impact. To our knowledge, for the OHLI questionnaire, the link between professional category or professional situation and OHL or HL have not been studied. However, this study demonstrated that these professional factors are major determinants of OHL and HL. Ehmann et al [47] demonstrated a bidirectional relation between work and HL. Work is a determinant of HL and the benefits of adequate HL have an impact on all life activities, including work. In addition, age appears to be a determinant of OHL, as previously described [48,49]. Although there is considerable individual variability, age-related decline in literacy is linked to a reduction in comprehension and fluency of reasoning as a result of physiological aging.

Regarding the correlation between OHLI and s-TOFHLA, a weak correlation was observed, in contrast to previous studies that demonstrated strong correlations between the Malay version of the OHLI and s-TOFHLA [29] and between the OHLI and TOFHLA [28]. This could be explained by the fact that s-TOFHLA shows errors in identifying people with low HL in specific dimensions of HL, such as numeracy, which was the case in our sample. The correlation between OHLI and s-TOFHLA therefore needs to be qualified, and efforts are needed to develop more-encompassing and practical strategies for identifying those with low HL for use in research and clinical practice [50]. Therefore, in order to assess the correlation between OHL and HL, it might be interesting to use a questionnaire that assesses HL on the basis of reading comprehension and numeracy, such as TOFHLA [51].

This study had several limitations. First, the s-TOFHLA was chosen to assess literacy levels because it is validated in French, easy to use, and short [33], but this questionnaire only includes a reading comprehension section and not a numeracy section. The TOFHLA [51], the long version of the s-TOFHLA, could have been chosen as it includes both sections, like the OHLI, and was used as a model for the development of the OHLI. However, it is not available in French and takes 22 minutes to complete. Thus, completing the OHLI, OH knowledge, and HL questionnaires would have led to participant exhaustion and thus bias in the assessment. Moreover, the other HL questionnaires available in French (the Functional, Communicative and Critical Health Literacy scale [52], the Health Literacy Questionnaire [53], the European Health Literacy Survey Questionnaire [54], and the Single-Item Literacy Screener [55]) are not based on reading comprehension and numeracy.

Second, the OHLI-F was used to assess the OHL level because it is the only questionnaire validated in French. Other OHL questionnaires containing reading comprehension and numeracy sections could have been used, such as the Comprehensive Oral Health Literacy Scale [56] or the Test of Functional Health Literacy in Dentistry [57].

Third, even if our results seem generalizable to the entire public service, this study only included state agents. This profession remains the most important among the 5.67 million French civil servants, divided into 3 parts (state, territorial, and hospital), including 2.52 million for the state public service, that is, 1.46 million people after excluding the national education department [24].

Fourth, as the aim of this study was to determine the level of literacy required to implement OH promotion programs in the workplace, the composition of our population (n=1190, 59% were aged over 60 years and n=1012, 52.8% were retired) could represent a bias. However, they are a target population because aging affects their literacy, OH, and general health levels. In France, a particularity is that retirees can access workplace programs through their health insurance.

Finally, it is possible that our results are related to selection bias, as some of the nonrespondents may have refused to answer the HL questionnaires to hide their low level of HL, thus underestimating the prevalence of low HL [58].

The purpose of this research was to investigate the level of HL and, more specifically, OHL among civil servants in France in order to contribute to the formulation of prevention strategies. These strategies will target increasing the knowledge of OHL in this professional category of the state civil service, which includes more than 5 million active and retired civil servants [24,59]. These prevention strategies should be oriented toward a community-based learning model focusing on the most vulnerable risk groups. For this, the workplace appears to be an opportunity [20,21,60]. Even if most civil servants had an adequate level of OHL and global HL, 2 high-risk groups have been identified for the implementation of a public health literacy program. First, the category C agents have been determined to be the most vulnerable socioprofessional category, yet these employees constitute the vast majority of territorial civil servants (75.6%), half of hospital civil servants (48%), and only 20% of state civil servants. Second, retired people are a particularly vulnerable group, with low levels of HL. The average retirement age for civil servants (excluding military personnel) is 62 years. This group represents 2.5 million people in the state civil service, and is composed only of people older than 62 years [59].

As the Report of the French National Health Commission advocates [61], our results suggest the implementation of an ambitious action program, with a particular focus on category C civil servants and retired civil servants. Our main recommendations focus on three areas: (1) create a favorable environment for the development of HL, which requires advocacy with both national and local decision-makers; (2) mobilize relevant resources and mechanisms to develop HL among vulnerable civil servants; and (3) develop, evaluate, investigate, and disseminate best practices specific to the at-risk groups, which to date require further clarification and evaluation. Preventive measures should be targeted primarily at category C civil servants and retired civil servants. For example, activities such as “teach-back,” which has a real role to play in literacy, could help to improve OHL [62]. In addition, tools could be proposed such as (1) modern technologies to improve the learning process, (2) face-to-face or online meetings and courses for empowerment and health promotion, (3) freely accessible electronic books or journals for information, and (4) specific applications targeting HL teaching [63].

The civil service represents 19.8% of total jobs in France, to which it is possible to add the 5.5 million retirees who are part of the same administrative system. Targeting this population to raise awareness and promote OHL, which has a considerable impact on health outcomes, is therefore an important public health issue. The workplace is at the heart of the health promotion strategy, with a focus on vulnerable groups such as category C professionals and older or retired civil servants. Future investigations will need to detect and investigate appropriate preventive communication procedures adapted to the literacy level of professional categories of workers and focused on numeracy skills. This should make it possible to achieve greater disease equity and combat the burden of NCDs.