Spatial spillover effect of environmental factors on the tuberculosis occurrence among the elderly: a surveillance analysis for nearly a dozen years in eastern China

Tuberculosis (TB) is one of the deadliest infectious diseases, causing substantial concern globally. The single agent of this disease, Mycobacterium Tuberculosis (MTB), is transmitted by the respiratory tract causing lesions in nearly all tissues and organs [1]. In 2022, an estimated 10.6 million people worldwide fell ill with TB, with an incidence of 133 cases per 100,000 people and 1.3 million deaths [1]. Although continuous efforts had been implemented in high burden countries, China, as one of 30 high burden countries, still accounts for 7.1% of global cases [1]. Among the affected population, the elderly group had contributed to a substantial part [2]. Along with aging, influencing factors such as the decreased immune function, cognitive deficiency, and insufficient social and family care might also lead to increased morbidity and PTB transmission risk in this special group [3, 4]. Thus, increasing attentions was directed on lowering the reactivation of latent pulmonary TB (PTB) infection and preventing new infections among the elderly.

Increasing evidence demonstrated that the dominated drivers of PTB development not only included personal factors such as gender, malnutrition, Acquired Immune Deficiency Syndrome (AIDS), diabetics, smoking and alcohol consumption, but were also associated with environmental factors and spatial locations [5‐7]. For environmental factors, available research showed that atmospheric pollutants such as PM_2.5 and SO₂ were associated with notification rate of PTB, partly even with a lag effect [8, 9]. It was possible that atmospheric pollutants may affect the susceptibility of TB in individual level by the possible mechanism of inducing damage in the tracheobronchial mucosa, and triggering systemic immune responses through inhibition of the synthesis and secretion of inflammatory mediators [8]. Besides, meteorological factor such as humidity can influence the transmission of MTB in the environment, thereby altering the risk of infection in the population [6]. However, the spillover effect caused by environmental factors is denoted that these factors not only affect the local epidemic but also impact surrounding regions, which was explored for TB onset in limited literature. In addition, the pattern of communicable diseases generally showed a diversity in spatial distribution. Thus, spatiotemporal analysis has been widely used in epidemiological research to identify temporal and spatial clusters of infectious diseases. However, among the elderly, the spillover effect of environmental factors and the spatial characteristics of PTB onset remain unclear [10, 11].

Our study aimed to analyze and determine the spatial and temporal distribution characteristics and risk clustering of the elderly PTB in Zhejiang Province, eastern China, as well as to identify environmental factors that have direct and spillover effects on notification rate of PTB. These findings may provide important empirical evidence for health policy formulation and public health resource allocation.

With the challenges of an aging population, the burden of PTB in China has persisted as a public health concern [3]. This study described the general characteristics and spatiotemporal distribution, and explored the influence of environmental factors on the occurrence of PTB among the elderly. These findings may inform future targeted interventions for PTB prevention and control for this high-risk population as well as promoting the allocation of health resources.

In this study, the notification rate of PTB among the elderly population was decreased, which was consistent with the general population in the province [15]. Reasons for this decrease are likely multifactorial but may be mostly due to the implementation of DOTS and the National TB Control Program, which include strengthening government dominance, refining public health systems, improving case notification, and increasing the financial budget for PTB in Zhejiang province, China. In addition, this trend may be partly attributed to improvements at the provincial-level in pathogenic diagnosis using new technology, such as use of GeneXpert MTB/RIF, since 2017. This method allows detection of the MTB complex in less than two hours, promoting early identification and treatment of PTB cases and reducing the transmission of PTB among elderly people [20]. This also explains the successive increase in the proportion of pathogen positivity since 2017 [21]. Furthermore, the coronavirus disease pandemic led to an excessive reduction in PTB detection and case notification at early stage, which may have also caused an underestimation of the notification rate in recent years [22].

Notified PTB cases among the elderly fluctuated, with the highest number in summer and autumn and lowest in February. In winter, factors such as vitamin D deficiency due to reduced sunlight, and indoor air pollution and crowd gathering due to coldness and Chinese New Year may increase the risk of MTB infection and transmission, respectively. Thus, given the potential incubation time between infection and PTB onset and the potential delayed time, the high infection rate in the former may cause a high activated risk among the elderly in the summer and autumn [23]. Therefore, elderly people should be aware of the risk of PTB infection in winter, and public health sectors should strengthen health education to reduce clustering, strengthen nutrition, and shorten the potential delay from symptom occurrence to the behavior of health-seeking.

The spatial autocorrelation analysis indicated that the epidemics were geographically dependent. The hot- and cold-spot regions identified in the local LISA maps were similar to the most likely risk cluster and the two secondary protection clusters identified in the spatial-temporal scan statistics. Studies has shown that older adults commonly have an estimated 90% PTB due to reactivation of LTBI acquired earlier in life rather than due to recent transmission [24]. Combined with the limited economic level in the western region of Zhejiang Province, we speculated that related influencing factors such as malnutrition should be explored, and possible interventions such as vitamin A and C supplementation and standardized preventive treatment should be implemented among the latent infection population [25]. In addition, due to mass X-ray screening of people aged over 60 or 65 years, especially among the elderly with no symptoms, active cases were identified earlily and given standardized treatment regimes. This also would help decrease local PTB epidemics in the general population.

In the Spatial Durbin Model, our study provided strong evidence that meteorological factors were vital factors affecting the occurrence of PTB among the elderly. There was a significant positive association between annual mean precipitation and PTB occurrence, which was consistent with the results of Qin T et al. [26]. The spillover effect of precipitation was approximately twice as significant as its direct effect, possibly due to Zhejiang Province’s coastal location, where frequent air currents aid in the formation and dispersal of droplets and suspended particles related to TB [26]. These particles can spread in all directions with the airflow, significantly impacting neighboring areas. Additionally, TB can be transmitted through various mediums, including surface water and groundwater formed by precipitation [27]. It could cause a longer distance of dissemination in space, leading to the expansion of spatial spillover effects. Furthermore, areas with higher tuberculosis notification rate often exhibited spatial clustering, further intensifying the impact of precipitation-induced spatial spillover effects [28]. Moreover, the results indicated that the high annual mean temperature could reduce the tuberculosis occurrence in local region and had a significantly negative spillover effect. It was suggested that high temperatures may stimulate the immune system’s response, leading to increased inflammation and production of immune effector molecules, thereby enhancing the ability to clear TB. The increase in local temperature reduces outdoor gatherings and activities among the elderly [29], potentially inhibiting transmission in the external environment. Hence, the increased local temperature exerts an indirect influence on surrounding areas through the spatial spillover effects. Also, our study found different effects between air pollutants and the risk of PTB occurrence among elderly. SO₂ has no direct effect on the notification rate of PTB. Previous study found no significant correlation between SO₂ and the risk of TB when long-term exposure or exposure to abnormally high concentrations of pollutants was ignored [30, 31], which is consistent with our findings. However, SO₂ has a notable negative spatial spillover effect on the health of the elderly. One potential explanation is that low-level SO₂ exhibits antimicrobial properties by reacting with enzymes and proteins within the cell membranes of microorganisms, thereby disrupting their structure and function, leading to the inhibition of microbial growth and reproduction [32]. Therefore, considering the distance-decay-effect of SO₂, short-term exposure to low-level SO₂ exhibited an protective effect on the elderly in the surrounding region during the diffusion process [33]. Exposure to PM_2.5 may increase the risk of PTB among the elderly in the local population which is consistent with previous studies [34]. Interestingly, there was no spatial spillover effect between PM_2.5 and the tuberculosis occurrence in our study. This may be attributed to the “Low-low” distribution of air pollutants in Zhejiang province, as well as the special geographical location and meteorological conditions near the sea, which help dissipate PM_2.5 [35]. Additionally, greenspace can contribute to the dispersal of PM_2.5 through deposition and filtration [36], thereby reducing the negative impact of PM_2.5 on the health of the elderly in the surrounding counties.

Despite its strengths, this study had some limitations. First, like other surveillance data, some PTB cases among the elderly may not be notified owing to a delay in seeking health care or not visiting medical institutions. Underestimation of the PTB notification rate in this specific population was unavoidable. Second, in 2019 and 2021, the administrative regions of several counties in Zhejiang Province had changed, and we integrated adjacent regions as a whole, which might ignore the spatial-temporal correlation within the integrated region. Third, the environmental factors obtained in this study were from the value of annual mean between 2010 and 2020, which may have affected the further analysis in 2021 and ignored the lag effect, leading to potential bias.

Targeted interventions among the elderly in Western Zhejiang may be more efficient than broad, province-wide interventions. Decreasing environmental pollution levels, such as PM_2.5, and enhancing the diversity of greenspace would be beneficial in controlling PTB occurrence while the low annual mean temperature and high annual mean precipitation in local and neighboring areas tend to have higher PTB onset among the elderly.