This study was a descriptive, cross-sectional, and quantitative approach to examine KAB related to the PHD adhering to the Checklist for Reporting of Survey Studies (Multimedia Appendix 1). It constitutes the behavioral component of the BIOQUALIM project [22], a transdisciplinary One Health research program assessing the role of the PHD in the primary and tertiary prevention of NCDs through 4 complementary studies (clinical, chemical, behavioral, and psychosocial).

This study was conducted in France at the Hospices Civils de Lyon (HCL) between April 15, 2024, and June 15, 2024. HCL is the second-largest public university hospital network in France, encompassing 13 hospitals and approximately 24,000 employees across the Lyon metropolitan area [23]. It covers a wide range of clinical specialties, research units, and training departments. The workforce includes physicians, nurses, midwives, pharmacists, technicians, allied health staff, and administrative personnel, constituting a highly diverse and representative health care environment.

The participants included (1) were aged ≥18 years, (2) spoke fluent French, and (3) were health care professionals at HCL.

This study was approved by the ethics committee of HCL (approval IRB00014232 2024 06 13_12, issued on June 19, 2024) and complied with the Declaration of Helsinki and the French data protection authority MR-004 norms. Written informed consent was obtained from all participants.

Figure 1 shows the flowchart of the study. Of the 8000 persons assessed for eligibility, 1229 (15.36%) were included, and 1104 (13.8%) fully completed the questionnaire. This sample size, well above the calculated minimum, ensured high statistical precision (–2.7% to +2.7% margin of error), whereas the voluntary recruitment process may limit full population representativeness.

Table 1 shows the sociodemographic and geographic characteristics of the participants. Most participants were women (927/1104, 83.97%) aged between 30 and 59 years (882/1104, 79.89%), lived in small households (990/1104, 89.67%), and were in intermediate professions (720/1104, 65.22%). Omnivorous (702/1104, 63.59%) and flexitarian (331/1104, 29.98%) diets were predominant. Half of the participants (598/1104, 54.17%) lived in urbanized areas (quintile 4-5), and 69.02% (762/1104) resided in wealthy metropolitan areas. A high proportion (876/1104, 79.3%) belonged to municipalities in the top 2 income quintiles. Weekly food expenses varied, with 46.01% (508/1104) spending less than €100 (US $116.56) and 49.55% (547/1104) spending more.

To better understand variations in KAB toward the PHD, a cluster analysis was conducted using standardized KAB scores. The internal consistency of the solution was supported by a total within-cluster sum of squares of 103,797. This clustering allocation (Table 1) and analysis (Figure 2 and Table 2) revealed 3 distinct profiles in health care professionals’ engagement with the PHD.

Cluster 1, named “knowledge-attitude gap” (481/1104, 43.57%), exhibited a relatively high knowledge score (32.8) but moderate scores in both attitude (13.8) and behavior (14.6), resulting in a total KAB score of 61.2. Cluster 2, named “high intent, low action” (318/1104, 28.8%), had the highest knowledge (33.1) and attitude (23.3) scores but the lowest behavior score (10.2), yielding a total KAB score of 76.6. Cluster 3, named “behavior-driven alignment” (305/1104, 27.63%), had moderate knowledge (21.4) and attitude (11.7) scores and the highest behavior score (15.7), resulting in a total KAB score of 48.8.

Cluster 1 (“knowledge-attitude gap”; 481/1104, 43.57% of the participants) was predominantly composed of middle-aged women in intermediate professions with a largely omnivorous diet and living in wealthy metropolitan areas (Table 1). Cluster 2 (“high intent, low action”; 318/1104, 28.8% of the participants) included more older participants and had the highest proportion of flexitarians. This group showed more diversity in occupational status and place of residence, with more frequent location in low- and middle-income municipalities. Cluster 3 (“behavior-driven alignment”; 305/1104, 27.63% of the participants) was the youngest group, also predominantly female, with mixed diets and a more frequent location in rural areas. Weekly food expenses were similar across clusters, mostly ranging from €51 to €150 (US $59.44-$174.83).

A multinomial logistic regression evaluated associations between individual characteristics and cluster membership considering the “behavior-driven alignment” cluster (cluster 3) as the reference category (Figure 3).

The comparison of cluster 1, “knowledge-attitude gap,” with cluster 3, “behavior-driven alignment” (Figure 3A), demonstrated that cluster 1 was significantly associated with being female (OR 7.08, 95% CI 2.98-16.8; P<.001) and of older age, especially ≥60 years (OR 7.78, 95% CI 2.72-22.20; P<.001) and 50 to 59 years (OR 2.86, 95% CI 1.41-5.79; P=.004). Omnivorous individuals were less likely than flexitarians to belong to cluster 1 (OR 0.43, 95% CI 0.26-0.72; P=.001). Living in periurban areas (OR 3.66, 95% CI 1.32-10.19; P=.01) or precarious population districts (OR 4.99, 95% CI 1.10-22.73; P=.04) also increased the probability of alignment with this cluster. Other variables showed no significant associations.

The comparison of cluster 2, “high intent, low action,” with cluster 3, “behavior-driven alignment” (Figure 3B), demonstrated that the cluster 2 profile was more likely associated with women (OR 5.58, 95% CI 1.45-21.52; P=.01) and individuals aged 50 to 59 years (OR 5.58, 95% CI 1.45-21.53; P=.01). A marginal trend of alignment with this cluster was noted for those in periurban areas (OR 6.82, 95% CI 0.99-47.10; P=.05). No other factors were significantly associated.

The results of the mediation analysis (Table 3) revealed that knowledge significantly predicted attitudes (b=0.2893 (0.2230-0.3556); SE 0.0338; t₁₁₀₂=8.56; P<.001). In turn, attitudes significantly predicted behaviors (b=0.1798, 95% CI 0.1472-0.2124; SE 0.0166; t₁₁₀₂=10.83; P<.001). However, the direct effect of knowledge on behaviors controlling for attitudes was not statistically significant (b=0.0072, 95% CI −0.0306 to 0.0449; SE 0.0192; t₁₁₀₂=0.37; P=.71).

The indirect effect of knowledge on behaviors via attitudes was significant (b=0.0520), with a 95% CI (0.0379-0.0682) that did not include 0. This indicates a full mediation effect: knowledge influences behaviors only indirectly through its effect on attitudes.

In a French context and among HCL health care professionals, this study addressed global health by identifying key behavioral typologies regarding the PHD. The PHD represents a transformative framework for addressing interlinked crises of human health and environmental sustainability. However, our findings highlight persistent gaps in its adoption even among health-competent populations. Cluster analysis of KAB identified three distinct profiles: (1) “knowledge-attitude gap”; (2) “high intent, low action”; and (3) “behavior-driven alignment.” These profiles underline both opportunities and barriers shaping dietary transitions within health care systems.

The largest cluster, “knowledge-attitude gap” (481/1104, 43.57% of the participants) demonstrated high knowledge about the PHD but failed to adopt corresponding dietary behaviors. This knowledge-behavior disconnect has been noted across public health domains, particularly among health care workers [31,32], who have dual responsibilities: modeling health-promoting behaviors and guiding patients toward sustainable lifestyles. Similar gaps have been identified in tobacco cessation [33] and preventive oncology [34], suggesting that professional knowledge does not ensure personal action. Our analysis is consistent with the KAB framework, indicating that, within this sample, knowledge appeared to influence behaviors indirectly through attitudes. Thus, affective and motivational factors may be as critical as informational ones. Environmental factors, institutional norms, and perceived behavioral control likely moderate this pathway, highlighting the importance of structural interventions alongside education [19,35]. Our mediation analysis suggests that targeted interventions should emphasize strengthening attitudes and motivations rather than simply providing information. Approaches such as motivational interviewing, normative messaging, and role-modeling within health care teams might significantly reduce the knowledge-attitude gap [36,37].

The “high intent, low action” cluster (318/1104, 28.8% of the participants) demonstrated high knowledge and positive attitudes but low PHD-aligned behaviors. This profile illustrates a different form of behavioral inertia: intention without implementation. Contributing factors may include workplace culture, lack of time, workplace food options, or insufficient social support. Previous studies show that, without standardized and accessible sustainable consumption options, even health-conscious individuals struggle to act on their intentions [38]. This profile was more prevalent among older and flexitarian participants, suggesting generational or cultural barriers to adopt the PHD. These individuals may benefit from interventions promoting action planning and institutional support, such as providing plant-based meals in hospitals or developing peer-led sustainability programs.

Conversely, the “behavior-driven alignment” cluster (305/1104, 27.63% of the participants), characterized by the highest behavioral alignment, had only moderate knowledge and attitude scores. This challenges the assumption that knowledge is a prerequisite for sustainable dietary practice. It underscores the role of social and contextual determinants on food choices. Behavioral science research suggests that daily routines, cultural norms, and environmental prompts can foster health-promoting behaviors even without strong cognitive drivers [39,40]. This cluster, predominantly young and periurban, may reflect evolving dietary behaviors driven by cultural trends, peer influence, or institutional exposure rather than formal health education [41,42]. Exploiting these archetypes could inform peer-led interventions within health systems.

Collectively, these profiles reveal that no group simultaneously achieved high knowledge and high behavioral adherence to the PHD. This overall pattern underscores the persistence of a structural knowledge-behavior gap across all segments. The absence of a cluster combining both high knowledge and high adoption of PHD practices reflects the well-established knowledge-behavior gap in health psychology and sustainability research. Numerous studies have shown that knowledge or intention alone seldom predicts actual behavior as the translation into action depends on psychological, social, and contextual moderators [43,44]. Self-efficacy and perceived behavioral control play pivotal roles in bridging this gap, influencing whether individuals feel capable of acting upon what they know [45]. Despite the proliferation of evidence-based guidelines, barriers to adherence persist (eg, guideline complexity, environmental context, and resource constraints)—as documented in recent implementation research [46,47]. Frameworks highlight multistage processes in which motivation, environmental triggers, and sustained self-regulation determine whether intentions lead to consistent behavioral adoption [48]. Similar intention-action discrepancies have been observed in sustainable food consumption, where positive attitudes coexist with low behavioral uptake due to contextual and social constraints [49]. Altogether, these findings suggest that promoting sustainable dietary practices among health care staff requires multilevel interventions that combine education with environmental and organizational enablers.

Beyond knowledge and attitudes, professional identity and perceived institutional norms significantly influence behavior. When sustainability is not embedded in the clinical culture, a disconnect can arise between personal values and visible action. Positioning the PHD as a personal choice and a professional standard could reinforce adoption and enable culture change [50]. Interventions integrating social role-modeling, normative messaging, and organizational alignment with PHD principles may be more effective than education alone.

The coexistence of these 3 profiles within a single health system reveals that health care professionals, despite having received scientific training and facing NCDs, do not uniformly model sustainable dietary behaviors. Lifestyle behaviors among health care professionals reflect inconsistencies found in the general population [51]. Given their direct involvement in managing NCDs and their role as trusted communicators, health care professionals’ dietary practices influence public health far beyond personal well-being [52]. By adopting and advocating for sustainable diets aligned with frameworks such as the PHD, which integrates human and environmental health goals, they can enhance clinical credibility and support institutional shifts toward healthier food policies [53,54].

Both the WHO [18] and the EAT-Lancet Commission [1] emphasize the PHD potential to reduce premature mortality, alleviate environmental degradation, and contribute to the United Nations Sustainable Development Goals [55]. Therefore, aligning health care professional behavior with these recommendations is a systemic opportunity. A PHD-adherent workforce is not only healthier but also more credible in promoting planetary health.

Health care institutions must recognize that individual knowledge is not sufficient to drive change. Behavioral interventions need embedding within broader institutional commitments to sustainability, including food supply policies, cafeteria reforms, cooking staff education, and food environment restructuring. Differentiating between structural barriers (eg, food access) and psychological barriers (eg, habit and motivation) should enable targeted actions on both fronts. Systems combining policy shifts with environmental engagements such as menu redesign, sustainable meal defaults, or workplace programs are more likely to succeed [56]. By integrating PHD principles into daily operations and wellness programs, health systems can act as exemplars and catalysts for the planetary health agenda, contributing directly to climate resilience, disease prevention, and institutional leadership [57].

This study has several limitations. First, it was conducted in a single hospital network and among health care professionals only using a convenience sampling strategy. These features may limit external validity and prevent direct extrapolation to the general population or other institutional contexts. In addition, analyses were descriptive by design, intended to identify behavioral profiles rather than develop predictive models. Second, self-reported data may have introduced social desirability bias, particularly among health care professionals who may overreport sustainable behaviors. Third, the cross-sectional design limits the ability to infer causality or examine the progression from knowledge to behavior over time, highlighting the need for longitudinal studies. Fourth, the sample was disproportionately female, reflecting HCL staffing demographics (75% women) but potentially influencing the results and limiting representativeness. Fifth, the study did not distinguish between clinical staff directly involved in patient care and administrative personnel. In addition, variables such as seniority or cultural or ethnic background were not collected due to institutional and regulatory restrictions, which may have constrained the exploration of factors influencing KAB. Sixth, although structural and institutional influences were conceptually discussed, they were not directly measured, limiting empirical assessment of their moderating effects. Finally, while the PHD offers a globally relevant framework, its applicability varies across cultural and socioeconomic contexts, requiring local adaptation. Future research should explore institutional culture, perceived norms, and structural facilitators to better understand their specific roles in shaping dietary adoption.

In French hospital settings (like the HCL), the 3 KAB profiles can guide targeted action: improve motivation and perceived control for the knowledge-attitude gap; remove practical barriers and adjust the food environment for high intent, low action; and maintain momentum for behavior-driven alignment. Indeed, given its scale and structure, HCL provides a relevant pilot environment to test and refine such approaches before broader implementation across other health care systems. As one of the largest public hospital networks in France [23], it offers a realistic institutional platform to evaluate feasibility and inform national deployment strategies.

Short-term institutional priorities are to align cafeteria offerings with the PHD (default options, placement, and pricing), deliver brief training with clear normative cues, and track change over time using the current KAB scoring framework (light-touch monitoring).

Medium-term priorities are to replicate the approach in other French health care contexts and link results to routine indicators (diet adherence, purchasing, and waste) to evaluate health and sustainability impacts. If routine standardized measurement is needed, a separate program could undertake formal psychometric validation; this is not required to implement the aforementioned actions.

Our study reveals heterogeneity in how health care professionals relate to the PHD. While knowledge remains important, attitudes and contextual factors strongly shape behavior. The identification of a behaviorally aligned cluster with only moderate knowledge underlines the power of cultural and environmental levers. To advance planetary health, health care must move beyond education toward institutional alignment. Within hospital systems, health care professionals may play an important role in promoting alignment with planetary health principles, potentially contributing to both personal and institutional sustainability goals.

Practical steps such as policy reform, sustainable food guidelines, staff training, and integrating sustainability into curricula are essential to embed these values within health care systems and support meaningful, long-term dietary transformation.