Geospatial Analysis of Tuberculosis Incidence in Relation to SocioEconomic and Environmental Indicators in Romania (2015-2021)

Jenna Zabroski¹Daniel Peptenatu^2,3,4Beatrice Mahler⁵Amr Soliman⁶Andreea Karina Gruia^3,7*Alexandra Grecu^3,7*Ioana Munteanu^8,9Aurel Bǎloi^10,3,7

• ¹Cancer Epidemiology Education in Special Populations (CEESP), NewYork, United States
• ²Faculty of Interdisciplinary Studies, University of Bucharest, Bucharest, Romania
• ³Interdisciplinary Center for Advanced Studies (CISA-ICUB), University of Bucharest, Bucharest, Romania
• ⁴Faculty of Geography, University of Bucharest, Bucharest, Romania
• ⁵Universitatea de Medicina si Farmacie Carol Davila din Bucuresti, Bucharest, Romania
• ⁶Department of Community Health and Social Medicine, School of Medicine, New York, United States
• ⁷Faculty of Administration and Business, University of Bucharest, Bucharest, Romania
• ⁸Universitatea Titu Maiorescu Facultatea de Medicina, Bucharest, Romania
• ⁹Institutul de Pneumoftiziologie Marius Nasta, Bucharest, Romania
• ¹⁰Graphit Innovation Factory, Drobeta Turnu Severin, Romania

Tuberculosis (TB) is a global public health burden. Romania experiences some of the highest rates of TB compared to other European countries. TB incidence is spatially clustered throughout Romania, and the reasoning behind this is likely linked to various environmental and socio-economic factors. This study aimed to investigate the spatial distribution of TB incidence per 1,000 population between 2015 and 2021 and examined its correlation with particulate matter (PM2.5) levels, living conditions, and level of educational indicators. TB incidence data from 2015 to 2021 aggregated at the level of 3,181 administrative territorial units (UATs) in Romania were analyzed. Pearson correlation coefficients examined linear associations among total TB incidence and exploratory indicators. A Principal Component Analysis (PCA) was conducted due to the presence of multicollinearity in the Pearson correlation matrix, which grouped key housing indicators into one hybrid living condition index (PCA1). Bivariate Moran's I analyses assessed localized spatial clustering between TB incidence and two exploratory indicators: PM2.5 levels and PCA1. A spatial lag regression model accounted for spatial dependence between TB incidence and three explanatory predictors: PM2.5 levels, the percentage of the population that is illiterate, and PCA1. All significance tests were conducted at a threshold of p<0.05. TB incidence is spatially clustered in Romania. PM2.5 levels (r = 0.261, p<0.0001) and PCA1 (r = 0.338, p<0.0001) are positively and significantly associated with TB incidence. Illiteracy showed no significant association with TB incidence. The spatial lag model confirmed spatial autocorrelation (Rho = 0.436, p<0.0001) and explains 28.4% of the variation in TB incidence across Romania. This is the first geospatial study in Romania that explores the link between TB incidence and exploratory indicators of air pollution, living conditions, and education level using epidemiological data obtained in the national tuberculosis surveillance and control program. Greater exposure to air pollution and worsened living conditions are correlated with higher TB incidence rates. Policymakers should highlight the need for geographically targeted interventions to improve TB control and screening in Romania.