Importance Evaluation Based on Random Forest Algorithms: Insights into the Relationship between Negative Air Ions Variability and Environmental Factors in Urban Green Spaces

A machine learning model to predict the risk of depression in US adults with obstructive sleep apnea hypopnea syndrome: a cross-sectional study

Objective

Depression is very common and harmful in patients with obstructive sleep apnea hypopnea syndrome (OSAHS). It is necessary to screen OSAHS patients for depression early. However, there are no validated tools to assess the likelihood of depression in patients with OSAHS. This study used data from the National Health and Nutrition Examination Survey (NHANES) database and machine learning (ML) methods to construct a risk prediction model for depression, aiming to predict the probability of depression in the OSAHS population. Relevant features were analyzed and a nomogram was drawn to visually predict and easily estimate the risk of depression according to the best performing model.

Study design

This is a cross-sectional study.

Methods

Data from three cycles (2005-2006, 2007-2008, and 2015-2016) were selected from the NHANES database, and 16 influencing factors were screened and included. Three prediction models were established by the logistic regression algorithm, least absolute shrinkage and selection operator (LASSO) algorithm, and random forest algorithm, respectively. The receiver operating characteristic (ROC) area under the curve (AUC), specificity, sensitivity, and decision curve analysis (DCA) were used to assess evaluate and compare the different ML models.

Results

The logistic regression model had lower sensitivity than the lasso model, while the specificity and AUC area were higher than the random forest and lasso models. Moreover, when the threshold probability range was 0.19-0.25 and 0.45-0.82, the net benefit of the logistic regression model was the largest. The logistic regression model clarified the factors contributing to depression, including gender, general health condition, body mass index (BMI), smoking, OSAHS severity, age, education level, ratio of family income to poverty (PIR), and asthma.

Conclusion

This study developed three machine learning (ML) models (logistic regression model, lasso model, and random forest model) using the NHANES database to predict depression and identify influencing factors among OSAHS patients. Among them, the logistic regression model was superior to the lasso and random forest models in overall prediction performance. By drawing the nomogram and applying it to the sleep testing center or sleep clinic, sleep technicians and medical staff can quickly and easily identify whether OSAHS patients have depression to carry out the necessary referral and psychological treatment.

Identifying the dominant impact factors and their contributions to heatwave events over mainland China

The heatwave frequency and intensity have significantly changed as the climate warms and human activities increase, which poses a potential risk to human society. However, the impact factors that determine the change of heatwave events remain unclear. Here, we estimated the heatwave events based on data from 2474 in-suit gauges during 1960-2018 at daily scale in China. Besides, we explored possible drivers and their contributions to the change of heatwave based on correlation analysis, multiple linear regression (MLR), and random forest (RF) in different subregions of China. The results show that the temporal changes of all heatwave metrics exhibit significant differences between the period 1960-1984 and the period 1985-2019. Spatially, the heatwave frequency and duration significant increase in the southern China (S), eastern arid region (EA), northeastern China (NE), Qinghai-Tibet region (QT) and western arid and semi-arid region (WAS). The occurrence of the first heatwave event in a year tends to be earlier in S, NE, EA, WAS, and QT than before. Based on the regression modelling and RF, human activities play an important role in heatwave intensity in all subregions of China. For heatwave frequency, urbanization generate a dominant influence in NE, EA, and QT, with relative contributions (RC) ranging from 32.8 % to 38.9 %. Long-term climate change exerts the dominant influence in C, N, and S. Moreover, the first day of the yearly heatwave event (HWT) in NE is significantly influenced by climate change, with RC of 33.9 % for temperature variation (TEM). Our findings could provide critical information for understanding the causes of heatwave across different regions of China in the context of rapid urbanization and climate change.

Sexual hormones in a coastal river adjacent to the Bohai Sea: Characteristic pollutants and dominantly influencing factors

Characteristic sexual hormones (SHs) and the factors that dominantly influence their occurrence in coastal ecosystems are less understood. This study verified the relationships between SHs and environmental factors and further inferred the possible controlling mechanisms of SH distribution. A characteristic pollutant of SHs was first proposed by determining the contamination level and ecological risks of SHs (seven species) in a coastal river adjacent to the Bohai Sea. The results showed that the 17β-oestradiol (17β-E2), estriol (E3), and 17α-ethynylestradiol (EE2) had high mean concentrations of 11.20 (±1.31), 10.17 (±4.91), and 16.71 (±0.88) ng L^-1, respectively, in the river water. The concentration of estrone (E1) was positively related to microbial substances of DOMs (p < 0.05). The humification index (HIX) had a negative relationship with E3 (p < 0.05). In water, the distribution of total SHs was regulated by the HIX and fluorescence index (FI), which might be related to photodegradation reactions. The 17α-oestradiol (17α-E2) and EE2 were related to humified organic matter, while E3 and androstenedione (ADD) were influenced by sewage input. The 17β-E2, E1, and 17α-E2 may be derived from animal sources, while E3, ADD, EE2, and progesterone were from human activities. Oestrogens, including E1, 17α-E2, 17β-E2, and EE2, displayed higher ecological risks than androgens and progesterone, with medium to high risk in most sites. The 17β-E2 was regarded as a characteristic pollutant of SHs throughout the river system, which displayed the highest risk. This paper may provide a reference for SH risk management and control.

Temporal dynamics of negative air ion concentration and its relationship with environmental factors: Results from long-term on-site monitoring

Negative air ions (NAIs) play an important role in evaluating forest health effects and promoting human physical and mental health. In this paper, long-term on-site monitoring of NAI concentration, air temperature, and relative humidity was conducted in real time over 24 h, from July 2019 to March 2021, to explore the temporal dynamic patterns of NAIs. We found that the daily dynamics of NAI concentration showed a bimodal curve. The peak concentrations usually occurred in the early morning (5:00-7:00) and afternoon (15:00-17:00), and the lowest concentrations usually occurred at noon (11:00-13:00). At the monthly scale, NAI concentrations were relatively high in February and August and low in May and December, and at the seasonal scale, NAI concentration was significantly higher in summer than in other seasons. Autumn had the second highest NAI concentration. There was no significant difference in NAI concentration between winter and spring. A comprehensive analysis shows that the AQI was the most key factor affecting NAI concentrations compared to temperature and relative humidity, especially the two indicators of particulate matter and ozone, and that NAI concentration had a negative correlation with these indicators and was significantly higher under favorable air quality conditions than under polluted air conditions. NAI concentrations and air temperature showed marked piecewise characteristics, with NAIs increasing linearly with rising temperature only if the Ta was separated into three ranges of -5 °C-10 °C, 10 °C-30 °C, and 30 °C-40 °C. With rising relative humidity, NAI concentration increased in accordance with a quadratic function. Our research provides new insights into the NAI temporal dynamics patterns and its driving factors, and will aid in scheduling outdoor recreation and forest health activities for urban people.

Perfluorinated compounds in a river basin from QingHai-Tibet Plateau: Occurrence, sources and key factors

The occurrence of perfluorinated compounds (PFCs) in different environmental media in the QingHai-Tibet Plateau has been limitedly investigated. In this study, the water, sediments, soils and agricultural product samples were collected in the Huangshui River basin, and contents of the PFCs and values of water parameters were determined. This study investigated dominantly regulating factors of the distribution of PFCs in the water emphatically, explored the sources and assessed potential risks of the PFCs integrally. The results showed that perfluorohexanesulfonic acid, perfluorooctanesulfonic acid, perfluorobutanoic acid (PFBA) and perfluorooctanoic acid presented high maximum concentrations of 3207.42, 3015.96, 1941.89 and 826.4 ng L-1 in the water, respectively. There were 12 PFCs detected in crops, with the maximum concentration of 5206.86 ng g-1 for PFBA. The significantly positive correlation (p < 0.05) was observed between the concentrations of PFBA in crops and that in adjacent rivers, indicating that the irrigation most likely contributed to the accumulation of PFBA in the studied crops. The occurrence of the PFCs in the water during the dry season was dominantly regulated by fluorescent dissolved organic matters via the hydrophobic interaction, while it was primarily regulated by the total nitrogen and electrical conductivity via electrostatic interaction during the wet season. The PFCs in the water were mainly from the wastewater discharged from wastewater treatment plants and carpet factories, while the resuspension of the PFCs in sediments was also an important contribution especially in wet season. The PFCs in the river has posed sustained risk to the public health, especially children.

Effects of Different Site Conditions on the Concentration of Negative Air Ions in Mountain Forest Based on an Orthogonal Experimental Study

The negative air ions (NAI) in a forest play an important and positive role in promoting the health of people using the forest for recreation. The purpose of this study was to explore the environmental characteristics that can effectively represent high concentrations of NAI in mountain forests to help the recreational users to seek out sites with high NAI concentrations for personal health reasons. In order to achieve this goal, we selected the mountain forest of Taibai Mountain National Forest Park, Shaanxi Province, China, as the research object and adopted an orthogonal experimental design with three factors and three levels to study the effects of terrain, altitude, and forest canopy density on the forest NAI concentrations. The results show that obvious peak–valley fluctuation occurs during 6:31 a.m. to 18:30 p.m., with the highest concentration of NAI at 8:00 a.m. (Average: 163 ions/cm3) and the lowest at 16:00 p.m. (Average: 626 ions/cm3). The altitude (p < 0.01) and canopy density (p < 0.05) were found to significantly affect NAI concentrations. The combination of site conditions in the mountain forest observed to have the highest NAI concentrations was valley topography, low altitude, and high canopy density. In addition, the highest NAI concentration was between 14:00 p.m. and 16:00 p.m., under this combination, which was thus identified as the most suitable time for health-promotion activities in mountain forests. The results provide insights into the NAI concentration characteristics and variations, along with identifying important environmental factors for the selection of health-promotion activities in mountain forests.

Comprehensive Evaluation of Healthcare Benefits of Different Forest Types: A Case Study in Shimen National Forest Park, China

Studies have shown that contact with nature plays a crucial role in the amelioration of human health. Forest therapy has recently received widespread attention as a novel and subsidiary treatment approach for stress recovery and health promotion. However, there is a lack of ample research on the comprehensive evaluation of the forest healthcare benefits. Moreover, it is not entirely clear what kind of forest types and seasons are suitable for forest therapy activities and how healthcare forests should be constructed and managed. From September 2019 to January 2020 and May to August 2020., five forest types of Phyllostachys edulis forest, subtropical evergreen broad-leaved forest, Liquidambar formosana forest, Cunninghamia lanceolata forest, coniferous and broad-leaved mixed forest and a forestless control group in Shimen National Forest Park, Guangzhou City, Guangdong Province, China were selected. Variations in the character of negative air ion concentration, air oxygen content, human comfort index and phytoncide relative content were analyzed. Principal component analysis and systematic clustering were used to construct forest comprehensive healthcare index and evaluation grade in order to assess the healthcare benefits of different forest types. In terms of negative air ion concentration, the subtropical evergreen broad-leaved forest was far ahead of the other forest types throughout the year, while the forestless control group was the worst. All stands reached the annual maximum in summer, followed by spring, autumn and winter. From the perspective of air oxygen content, summer > spring > autumn > winter, among them, all forest stands clearly exceeded the normal atmospheric oxygen content (20.9%) in the first three quarters. Moreover, the air oxygen content of coniferous and broad-leaved mixed forest was the highest in five forest types; the forestless control group was the lowest. Judging from the human comfort index, in the whole year, all forest types, including the forestless group, were at the comfortable level and above. However, the five forest types still differed greatly in diverse seasons, among which Phyllostachys edulis forest and subtropical evergreen broad-leaved forest were superior to Liquidambar formosana forest, Cunninghamia lanceolata forest, coniferous and broad-leaved mixed forest in spring and summer, while it was in reverse in autumn and winter. In view of the phytoncide relative content, the subtropical evergreen broad-leaved forest was the highest, followed by the Cunninghamia lanceolata forest. The relative content of phytoncide was released more in summer, second, by spring, autumn and winter. Furthermore, establishing forest comprehensive healthcare index (FCHI = 0.1NAICi + 0.35AOCi + 0.27HCIi + 0.28PRCi), according to the FCHI value, it was divided into five rating levels. Overall, the comprehensive healthcare index of the five forest stands distinctly outperformed the forestless control group in all seasons. In addition, the five forest types were at level I in spring and summer. From the comprehensive data of the whole year, the comprehensive healthcare index of the coniferous and broad-leaved mixed forest was the best, followed by the subtropical evergreen broad-leaved forest. The forest environment has a favorable influence on the human body and mind, so it is suggested that citizens go to the forest environment regularly for healthcare and physical and mental washing. In terms of the forest healthcare benefits, the best seasons for forest therapy in Shimen National Forest Park are spring and summer; autumn is suitable as well. When planning and constructing the forest therapy bases in Shimen National Forest Park in the future, coniferous and broad-leaved mixed forests should be allocated more in the stand transformation to promote forest healthcare benefits. Protecting and developing the landscape resources of the subtropical evergreen broad-leaved forests should be paid close attention, as well as making rational use of their health activity space.