In the field of global public health, diabetes has become a serious challenge that cannot be ignored. It is like an invisible “health killer” quietly affecting the lives of hundreds of millions of people. Statistical data shows that more than 500 million people worldwide suffer from diabetes, and this number is expected to continue to rise, possibly reaching 1.3 billion in the future. In Brazil, the prevalence of diabetes has also been increasing, rising from 6.2% in 2013 to 7.7% in 2019, putting a heavy strain on the local healthcare system and social economy.
Although it is widely known that green spaces are beneficial to health, such as relieving stress and increasing opportunities for exercise, the relationship between neighborhood green spaces and diabetes in Brazil has lacked in – depth research. People in different regions have different access to green spaces. Does this affect the incidence and treatment of diabetes? To answer these questions, researchers from several institutions in Brazil conducted a study titled “Neighborhood green space and diabetes mellitus in Brazil”, and the relevant results were published in the journal iScience.
To explore the association between neighborhood green spaces and diabetes, the researchers used a variety of professional techniques. They obtained data on diabetic inpatients from January 1, 2008, to December 31, 2018, from the public dataset of the Brazilian Ministry of Health, covering patients’ basic information and detailed hospitalization records. When measuring the level of green spaces, the Normalized Difference Vegetation Index (NDVI) was used. This data was processed and provided by the Brazilian Agricultural Research Corporation (EMBRAPA) based on the products of the Moderate Resolution Imaging Spectroradiometer (MODIS) on the Terra and Aqua satellites. Through specific calculations and screenings, the average NDVI under different buffer sizes was obtained. At the same time, the researchers included air pollution variables (PM2.5 and O3), climatic factors (temperature, relative humidity, and precipitation), and socioeconomic status (average household income and years of education) as covariates in the analysis, and used a quasi – Poisson regression model to evaluate the relationship between green spaces and diabetes – related hospitalizations.
The study results showed that during the study period, a total of 1,713,760 people were hospitalized due to diabetes, with 1,017,104 female patients and 698,524 male patients. Among different age groups, the number of inpatients in the age group over 65 years old was the largest (610,969), while the number in the age group of 15 – 45 years old was the smallest (416,095). The average NDVI was 0.35 (SD = 0.16), and the average NDVI gradually increased as the buffer size increased. The average concentrations of PM2.5 and O3 were 17.66μg/m3 (SD = 23.69) and 21.40 ppb (SD = 7.58), respectively. The average temperature was 22.71℃ (SD = 4.06), and the average relative humidity was 77.40% (SD = 12.93).
Regarding the relationship between neighborhood green spaces and diabetes – related hospitalizations, significant differences were observed among different regions. In the northern region of Brazil, green spaces had a significant protective effect on diabetes – related hospitalizations among people aged 15 – 45, showing a significant negative correlation under all buffer sizes. For example, in the 0 m buffer zone, for every increase of one interquartile range (IQR) in NDVI, the relative risk (RR) of hospitalization for the overall population (all ages and genders) was 0.981 (95% confidence interval [CI]: 0.980; 0.982). In the central – western region, significant negative correlations were also observed in the 0, 500, and 1,000 m buffer zones. However, in the northeastern region, except for females aged 15 – 45 in the 0 m buffer zone, significant positive correlations were found in all buffer zones, indicating that an increase in green spaces may be associated with a higher rate of diabetes – related hospitalizations. The southeastern region also mainly showed positive correlations. In the southern region, positive correlations were found in males of all age groups and individuals aged 46 – 65, while the results for other gender/age groups were not significant.
Further stratified analysis by NDVI tertiles revealed that in the high and middle tertiles, most subgroups in the southeastern and southern regions showed significant negative correlations, indicating that high NDVI helps reduce the risk of diabetes – related hospitalizations; the negative correlation was more obvious in the central – western region at the middle tertile. At the low tertile, the pattern changed. Except for some subgroups in the northern region that still showed positive correlations, many subgroups in other regions showed significant negative correlations, suggesting that lower levels of green spaces are also associated with reduced diabetes – related hospitalizations in these areas. In addition, the study also confirmed a non – linear relationship between green spaces and the risk of diabetes – related hospitalizations, presenting a complex U – shaped or multi – stage pattern. The results of sensitivity analysis showed that after adjusting for various factors and using different models, the study results remained robust.
This study is of great significance as it clearly points out that the relationship between green spaces and diabetes – related hospitalizations is significantly affected by regional differences. In some regions, green spaces have a protective effect on diabetes – related hospitalizations; in other regions, the results are the opposite. These differences may be related to regional socioeconomic conditions, accessibility of medical resources, functions and accessibility of green spaces, as well as some unmeasured confounding factors. This study provides important references for urban planners and policymakers, helping them develop targeted strategies to optimize the distribution of green spaces, improve their accessibility, promote health equity, reduce the burden of diabetes – related hospitalizations, promote sustainable urban development, and safeguard public health.
