Mental illness remains a significant public health challenge globally, contributing to considerable morbidity, mortality, and economic burden [1]-[3]. In the United States, mental health disorders are among the leading causes of disability-adjusted life years (DALYs), with millions requiring hospitalization each year due to acute psychiatric episodes or for long-term psychiatric care [4][5]. Florida, in particular, ranks among the top five states in terms of the prevalence of mental illness [6][7]. The Floridian aging population structure, rapid population growth, diverse ethnic communities, vulnerability to natural disasters, and county-level variations in healthcare access create unique mental health challenges [8]. Mental disorders such as depression, anxiety, bipolar disorder, and schizophrenia are not only highly prevalent but also among the top contributors to disability and health service utilization, including psychiatric hospitalizations [5][9].

Hospitalization rates for mental disorders vary considerably across geographic regions and are strongly influenced by socioeconomic inequalities, demographic differences, and environmental exposures [10]. These contextual disparities affect access to mental health services, quality of care, and recovery outcomes. For instance, populations in lower-income or rural areas may face barriers to accessing timely and appropriate treatment, which can lead to more frequent or prolonged hospitalizations [6][11]. Additionally, environmental conditions, such as exposure to pollution, limited green space, or urban stressors, may not be randomly distributed across space, potentially acting as social determinants of mental health [12][13].

Despite the recognition of these challenges, limited research has explored the geographic variation in hospitalization rates for mental disorders at granular or county levels. Most existing studies focus on individual-level predictors or broad national trends, often overlooking small-area variation that can reveal important patterns of service need and inequity. Understanding the spatial distribution of mental health-related hospitalizations is essential for identifying regional disparities, informing mental health policy, and optimizing resource allocation. Such insights are particularly critical in a state like Florida, characterized by a heterogeneous population and wide variability in healthcare infrastructure and access.

This study addresses a notable gap by investigating the spatial patterns of mental health-related hospitalizations across counties in Florida. The main research questions guiding this study are:

1) What are the spatial patterns and hotspots of hospitalization rates for mental disorders across Florida counties? 2) How do socioeconomic, demographic, and healthcare access factors influence these spatial patterns? 3) What are the implications of these spatial disparities for mental health policy and targeted resource allocation? Answering these questions has practical significance. Identifying geographic areas with disproportionate hospitalization burdens can help policymakers and health planners allocate mental health resources more equitably and efficiently, ensuring that services reach communities with the greatest need.

While prior studies have documented the role of socio-environmental and demographic factors in shaping mental health outcomes, these investigations often neglect spatial dependencies or clustering effects. Socioeconomic determinants such as poverty, unemployment, and income inequality have consistently been linked to increased risk of mental illness [11]. Similarly, access to care, including the density of mental health professionals, availability of psychiatric beds, and insurance coverage, remains a key barrier in many underserved areas [14]. Environmental stressors, including urban density, pollution, and lack of access to restorative green spaces, have also been found to exacerbate psychological distress [12][15].

Regarding the role of socio-environmental and demographic factors, spatial epidemiological techniques have been successfully used in other public health contexts—such as cancer mapping [16], cardiovascular outcomes [17], and infectious disease surveillance [18], to detect geographic clusters and inform localized interventions. However, their application to mental health outcomes, particularly hospitalization rates, remains underdeveloped. Recent advances in Bayesian spatial modeling offer promising tools for identifying contextual and spatial effects that influence health outcomes at small-area levels [19].

This study leverages these advanced spatial methodologies to assess the geographic distribution of hospitalization rates for mental disorders in Florida. Specifically, we apply Bayesian spatial models to examine associations between hospitalization rates and county-level socioeconomic, demographic, and environmental variables. By integrating spatial statistics with public health data, we aim to uncover hidden geographic patterns that may otherwise go undetected in traditional analyses, thereby contributing novel insights to the field of mental health services research.

Descriptive statistics for the outcome variable (county-level hospitalizations due to mental disorders) and their associated attributes are presented in Table 1. In 2022, the average number of hospitalizations for mental disorders across Florida counties was 3069, with considerable variation. The lowest count was recorded in Glades County (35 hospitalizations), while Miami-Dade County had the highest at 28,718. The median age across counties was 44.26 years, and the median household income was $61,786. The mean obesity prevalence was 68.14%, and the average poverty rate was 14.92%. Across the 67 counties, substantial variability was observed in key sociodemographic and environmental factors, including the

Table 1. Descriptive statistics of county-level variables.

percentage of adults with at least a bachelor’s degree, household income, proximity to healthy food stores, poverty levels, and the demographic composition in terms of White and veteran populations. These differences were also reflected in the Standardized Hospitalization Rate (SHR) values.

The spatial distribution of SHR values is depicted in Figure 1, which reveals non-random spatial variation in mental health-related hospitalizations. Higher-than-expected hospitalization rates were clustered in the northern, central, and eastern regions of Florida, as indicated by red and orange shading. However, SHR may be an unstable metric, particularly in counties with small populations, since it assumes homogeneity across counties and does not account for local variability. Therefore, to generate more reliable estimates and assess spatial dependency, we fitted a Bayesian spatial model (BSM), specified in Equation (1).

Figure 1.Spatial distribution of standardized hospitalization rates of mental disorders in Florida (2022).

Table 2 summarizes the results of the Bayesian model. Posterior estimates and 95% credible intervals (CIs) were used to evaluate the association between hospitalization rates and county-level covariates. Looking at the 95% credible interval (CI) of parameters of BSM, we observe that counties with higher percentage of people with health insurance had lower hospitalization rates ( β ^ : −0.0826, CI: (−0.1636, −0.0017), which does not include zero) after adjusting for other variables. The results also show that there is a positive and strong relationship between the percentage of poverty and hospitalization rate due to mental disorders ( β ^ : 0.2679, CI: (0.0608, 0.4768)) after adjusting for other county-level covariates. Similarly, there is a strong and direct association between the hospitalization rate and percentage of residents who are obese ( β ^ : 0.1417, CI: (0.0371, 0.2465). But this association is moderated by poverty (obesity × poverty, β ^ : −0.0044, CI: (−0.0073, −0.0015)) or white (obesity × white, β ^ : −0.0014, CI: (−0.0025, −0.0003)). These interaction terms suggest that in counties with higher poverty rates or a greater percentage of White residents, the positive association between obesity and hospitalization weakens or even reverses. No statistically significant associations were observed between hospitalization rates and other variables, such as unemployment rate, income level, education level, segregation index, or veteran population percentage.

Table 2.Estimated posterior mean (PM) and standard deviation (SD) of population parameters of Model (1) along with a lower limit ( L C L ) and upper limit ( U C I ) of 95% equal-tail credible interval (CI).

Using the fitted BSM, we computed posterior hospitalization rates (PHR) for all counties and mapped these in Figure 2. The map indicates that counties in the northern and central regions of Florida had the highest hospitalization burdens, defined by PHR values exceeding 1.

To identify hospitalization hotspots, we calculated the exceedance probability, defined as the posterior probability that a county's hospitalization rate exceeds a given threshold (e.g., PHR > 1). Figure 3 displays exceedance probabilities for all 67 counties. High exceedance probabilities (close to 1) indicate counties where hospitalization burdens are significantly above the expected level, while values near 0 suggest below-threshold rates. Looking at Figure 3, one sees that the identified hotspots (red color) for hospitalization burden due to mental disorders are

Figure 2. Spatial distribution of posterior hospitalization rates of mental disorders in Florida (2022).

Figure 3.Exceedance probability that a county’s hospitalization rate for mental disorders is greater than expected.

scattered across Florida. In the northwest, the hotspot counties in 2022 were Escambia, Gadsden, Jefferson and Sant Rosa. Likewise, in the northeast, the counties were Alachua, Bradford, Clay, Duval and Putnam. Hotspot counties in the central east region were Indian River, Martin, Orange, Seminole and Volusia. In the central west part, Hernando, Pasco and Pinellas counties were identified as having high hospitalization rates, and the southeast Broward, Miami-Dade and Monroe happened to have high burden of hospitalization rates for mental disorders. These spatial patterns underscore the geographic inequities in the burden of mental health hospitalization across the state. Identifying such hotspots is vital for targeted interventions, enabling policymakers and public health practitioners to prioritize high-burden areas for resource allocation, prevention strategies, and expanded access to mental health care services.

This study examined the spatial distribution of hospitalization rates due to mental disorders across counties in Florida and explored their associations with key socioeconomic, demographic, and environmental covariates. Our findings reveal significant geographic variation in mental health-related hospitalizations, with elevated rates concentrated in the northern and central regions of the state. These spatial disparities suggest that mental health burdens are not evenly distributed and may reflect underlying social and environmental inequities.

Consistent with previous research emphasizing the role of contextual determinants in mental health outcomes [5][11], our study found that counties with higher poverty rates exhibited significantly greater hospitalization rates for mental disorders. This supports the notion that economic deprivation contributes to poor mental health through pathways such as chronic stress, inadequate housing, food insecurity, and limited access to preventive care [1][10]. In contrast, higher levels of health insurance coverage were associated with reduced hospitalization rates, underscoring the protective role of access to healthcare services. These findings highlight the urgent need for interventions that expand coverage and reduce structural barriers to care in underserved areas.

A notable contribution of this study is the identification of obesity as a significant positive predictor of hospitalization rates for mental disorders. This aligns with literature indicating a bidirectional relationship between obesity and mental illness, potentially mediated by stigma, inflammation, and comorbidities such as diabetes or cardiovascular disease [9]. However, we found that the strength and direction of this association varied by county-level poverty and racial composition, suggesting that obesity's impact on mental health is context-dependent. In counties with higher poverty or a larger White population, the positive association between obesity and hospitalization was attenuated or reversed. These moderating effects point to complex interdependencies between physical health, socioeconomic status, and social environments that warrant further investigation. Environmental and socioeconomic factors have been found to significantly impact mental health in various mechanisms [30]-[32]. Though not clearly known, some mechanisms may explain moderation of race and poverty for the association of obesity and higher rates of hospitalization for mental disorder.

Our application of Bayesian spatial modeling provides a methodological advantage by accounting for spatial dependencies and heterogeneity in population structure, issues that may bias estimates based on traditional standardized rates. The identification of high-burden counties through posterior hospitalization rates and exceedance probabilities offers a more robust understanding of localized mental health challenges. Hotspot counties, particularly in the northwest (e.g., Escambia, Gadsden, Jefferson), northeast (e.g., Duval, Putnam), and central regions (e.g., Orange, Hernando), should be prioritized for targeted mental health interventions and capacity-building efforts.

These findings have practical implications for public health planning. By identifying geographic areas with disproportionate hospitalization burdens, state and local health agencies can tailor interventions that improve service accessibility, invest in community-based preventive care. Moreover, spatial epidemiology can complement ongoing mental health surveillance systems by uncovering patterns that are not visible through aggregate state-level analyses.

Several limitations should be acknowledged. First, the ecological nature of the study precludes causal inferences at the individual level. Second, while we included several important county-level covariates, unmeasured confounders such as mental health stigma, health workforce shortages, or behavioral health service quality may also influence hospitalization rates. Third, the cross-sectional design captures spatial patterns at a single time point and may not reflect temporal trends. Future studies employing multilevel modeling of small-area level and individual level variables would enhance understanding of evolving mental health disparities. Evaluation of the association between mental disorder and hospitalization at a finer spatial unit would be useful in future study. In addition, potential confounders, for instance, a county-level density of mental health providers, are worth including in a future study as well.

In conclusion, this study provides new evidence on the spatial distribution and contextual determinants of mental health-related hospitalization rates across Florida counties. Using Bayesian spatial modeling, we identified significant geographic disparities, with elevated hospitalization burdens in the northern and central regions of the state. Higher poverty rates, lower insurance coverage, and obesity prevalence were found to be key drivers of hospitalization, with important moderating effects from racial composition and socioeconomic conditions. These findings underscore the importance of spatially targeted mental health strategies that account for both place-based and population-based factors.

Investments in public health infrastructure, particularly in hotspot counties, are critical to reducing the burden of mental disorders and promoting mental health equity. Our study highlights the value of integrating spatial epidemiological tools into mental health surveillance and planning efforts, ultimately contributing to more effective and equitable mental health systems.

The datasets were obtained from publicly archived located at: https://www.flhealthcharts.gov/ChartsReports/rdPage.aspx?rdReport=NonVitalInd.Data.