Environmental Health and Preventive Medicine最新文献

Background: Hypertension is a serious chronic disease that can significantly lead to various cardiovascular diseases, affecting vital organs such as the heart, brain, and kidneys. Our goal is to predict the risk of new onset hypertension using machine learning algorithms and identify the characteristics of patients with new onset hypertension.

Methods: We analyzed data from the 2011 China Health and Nutrition Survey cohort of individuals who were not hypertensive at baseline and had follow-up results available for prediction by 2015. We tested and evaluated the performance of four traditional machine learning algorithms commonly used in epidemiological studies: Logistic Regression, Support Vector Machine, XGBoost, LightGBM, and two deep learning algorithms: TabNet and AMFormer model. We modeled using 16 and 29 features, respectively. SHAP values were applied to select key features associated with new onset hypertension.

Results: A total of 4,982 participants were included in the analysis, of whom 1,017 developed hypertension during the 4-year follow-up. Among the 16-feature models, Logistic Regression had the highest AUC of 0.784(0.775∼0.806). In the 29-feature prediction models, AMFormer performed the best with an AUC of 0.802(0.795∼0.820), and also scored the highest in MCC (0.417, 95%CI: 0.400∼0.434) and F1 (0.503, 95%CI: 0.484∼0.505) metrics, demonstrating superior overall performance compared to the other models. Additionally, key features selected based on the AMFormer, such as age, province, waist circumference, urban or rural location, education level, employment status, weight, WHR, and BMI, played significant roles.

Conclusion: We used the AMFormer model for the first time in predicting new onset hypertension and achieved the best results among the six algorithms tested. Key features associated with new onset hypertension can be determined through this algorithm. The practice of machine learning algorithms can further enhance the predictive efficacy of diseases and identify risk factors for diseases.

Background: Evidence on the combined effects of air pollutants and greenspace exposure on pulmonary tuberculosis (PTB) treatment is limited, particularly in developing countries with high levels of air pollution.

Objective: We aimed to examine the individual and combined effects of long-term exposure to air pollutants on PTB treatment outcomes while also investigating the potential modifying effect of greenspace.

Methods: This population-based study included 82,784 PTB cases notified in Zhejiang Province, China, from 2015 to 2019. The 24-month average concentrations of particulate matter with an aerodynamic diameter ≤2.5 µm (PM_2.5), ozone (O₃), nitrogen dioxide (NO₂), and sulfur dioxide (SO₂) before PTB diagnosis were estimated using a dataset derived from satellite-based machine learning models and monitoring stations. Greenspace exposure was assessed using the annual China Land Cover Dataset. We conducted analyses using time-varying Cox proportional hazards models and cumulative risk indices.

Results: In individual effect models, each 10 µg/m³ increase in PM_2.5, NO₂, O₃, and SO₂ concentrations was associated with hazard ratios for PTB treatment success of 0.95 (95% confidence interval (CI): 0.93-0.97), 0.92 (95% CI: 0.91-0.94), 0.98 (95% CI: 0.97-0.99), and 1.52 (95% CI: 1.49-1.56), respectively. In combined effect models, long-term exposure to the combination of air pollutants was negatively associated with PTB treatment success, with a joint hazard ratio (JHR) of 0.79 (95% CI: 0.63-0.96). Among the pollutants examined, O₃ contributed the most to the increased risks, followed by PM_2.5 and NO₂. Additionally, areas with moderate levels of greenspace showed a reduced risk (JHR = 0.81, 95% CI: 0.62-0.98) compared with the estimate from the third quantile model (JHR = 0.68, 95% CI: 0.52-0.83).

Conclusions: Combined air pollutants significantly impede successful PTB treatment outcomes, with O₃ and PM_2.5 accounting for nearly 75% of this detrimental effect. Moderate levels of greenspace can mitigate the adverse effects associated with combined air pollutants, leading to improved treatment success for patients with PTB.

Background: The mechanisms distinguishing metabolically healthy from unhealthy phenotypes within the same BMI categories remain unclear. This study aimed to investigate the associations between regional fat distribution and metabolically unhealthy phenotypes in Chinese adults across different BMI categories.

Methods: This cross-sectional study involving 11833 Chinese adults aged 20 years and older. Covariance analysis, adjusted for age, compared the percentage of regional fat (trunk, leg, or arm fat divided by whole-body fat) between metabolically healthy and unhealthy participants. Trends in regional fat percentage with the number of metabolic abnormalities were assessed by the Jonckheere-Terpstra test. Odds ratios (ORs) and their 95% confidence intervals (CIs) were estimated by logistic regression models. All analyses were performed separately by sex.

Results: In non-obese individuals, metabolically unhealthy participants exhibited higher percent trunk fat and lower percent leg fat compared to healthy participants. Additionally, percent trunk fat increased and percent leg fat decreased with the number of metabolic abnormalities. After adjustment for demographic and lifestyle factors, as well as BMI, higher percent trunk fat was associated with increased odds of being metabolically unhealthy [highest vs. lowest quartile: ORs (95%CI) of 1.64 (1.35, 2.00) for men and 2.00 (1.63, 2.46) for women]. Conversely, compared with the lowest quartile, the ORs (95%CI) of metabolically unhealthy phenotype in the highest quartile for percent arm and leg fat were 0.64 (0.53, 0.78) and 0.60 (0.49, 0.74) for men, and 0.72 (0.56, 0.93) and 0.46 (0.36, 0.59) for women, respectively. Significant interactions between BMI and percentage of trunk and leg fat were observed in both sexes, with stronger associations found in individuals with normal weight and overweight.

Conclusions: Trunk fat is associated with a higher risk of metabolically unhealthy phenotype, while leg and arm fat are protective factors. Regional fat distribution assessments are crucial for identifying metabolically unhealthy phenotypes, particularly in non-obese individuals.

Although salt-related behaviors may influence urinary salt excretion in early childhood, this relationship remains unclear. This study aimed to examine salt-related behaviors using data from a salt check sheet and urinary salt excretion parameters using spot urine samples from 4-year-old children. This cross-sectional study included all 4-year-old children who underwent health checkups in Ohnan Town, Shimane Prefecture. The study sample consisted of 109 children (49 boys). Measures from spot urine samples included estimated salt excretion (g/day) and the sodium-potassium (Na/K) ratio. Salt-related behaviors were assessed using a salt check sheet that was completed by the parents or guardians. The associations between salt-related behaviors and urinary salt excretion parameters were analyzed using a generalized linear model. The median (M) and interquartile range (IQR) for urinary measures in 4-year-old children were as follows: estimated salt excretion (M = 4.4, IQR: 3.3-6.2) and Na/K ratio (M = 2.3, IQR: 1.4-3.3). The low frequency of consumption of high-salt foods ("such as pickles, pickled plums, etc." and "noodles such as udon and ramen") was associated with low salt excretion and low Na/K ratio. However, in the case of "consumption of udon, ramen, or other soups", the Na/K ratio was higher for "About half a bowl" and "Some" than for "An entire bowl." Additionally, for "eating out or having convenience-store-bought bento (lunch plate) for lunch", the Na/K ratio was higher for "No" than for "Almost every day." In conclusion, the frequency of high-sodium food intake was associated with both urinary sodium excretion and the Na/K ratio in 4-year-old children. Longitudinal investigations using the 24-hour urine collection method are needed to confirm these salt-related behaviors.

Background: Chronic kidney disease (CKD) poses a major global health challenge, often foreshadowing poor patient outcomes. The C-reactive protein to albumin ratio (CAR) serves as a pivotal biomarker, demonstrating a strong correlation with adverse outcomes in cardiovascular disease (CVD). This study sought to examine the correlation between CAR and the risk of all-cause and cardiovascular mortality in patients with CKD stages 3-5.

Methods: This study utilized data of CKD patients from the National Health and Nutrition Examination Survey (NHANES) from 1999 to 2010, with follow-up to December 31, 2019. The optimal CAR cutoff value was identified utilizing the method of maximally selected rank statistics. Multivariable Cox proportional hazards regression model, restricted cubic splines (RCS) model, and subgroup analysis were employed to assess the association between CAR and mortality among CKD patients.

Results: During a median (with interquartile range) follow-up period of 115 (112,117) months among 2,841 CKD individuals, 1,893 deaths were observed, including 692 deaths due to CVD events. Based on the RCS analysis, a non-linear correlation was observed between CAR and mortality. Using 0.3 as the optimal CAR cutoff value, the cohort was divided into high and low groups. In the fully adjusted model, CKD patients with high CAR values exhibited an elevated risk of all-cause mortality (hazard ratio [HR] 1.53, 95% confidence interval [CI] 1.28-1.83, P < 0.001) and cardiovascular mortality (HR 1.48, 95% CI 1.08-2.02, P = 0.014). Compared to the population aged >65 years (HR 1.32, 95% CI 0.99-1.76, P = 0.064), the risk of cardiovascular mortality was significantly higher in those aged ≤65 years (HR 2.19, 95% CI 1.18-4.09, P = 0.014) with elevated CAR levels.

Conclusions: A notable correlation exists between the elevation of CAR and increased all-cause and cardiovascular mortality, suggesting its potential as an independent indicator for evaluating the prognosis of patients with CKD stages 3-5.

Background: Previous studies have demonstrated that short-term exposure to ambient particulate matter elevates the risk of ischemic stroke in major urban areas of various countries. However, there is a notable gap in research focusing on remote areas inhabited by ethnic minorities and the cumulative effects of air pollutants. Our study conducted in the area aims to explore the potential association between ischemic stroke and air pollutants and contribute to improving health outcomes among the community.

Methods: This retrospective observational study was conducted at the Xing'an League People's Hospital in Inner Mongolia. The medical records of 4,288 patients admitted for IS between November 1, 2019, and October 31, 2020, were reviewed. Data on demographics (age and sex), air pollutants (PM₁₀, PM_2.5, NO₂, NO, CO, and O₃), and meteorological factors (daily average temperature, daily average wind speed, and daily average atmosphere pressure) were collected and analyzed. The statistical analysis included descriptive statistics, Poisson distribution analysis to evaluate the adverse effects of atmospheric pollutants on daily hospitalizations, and subgroup analysis to determine whether gender and age could modify the impact on hospitalizations.

Results: A substantial correlation was revealed in single-day lags model. The peak delayed effects of PM₁₀, PM_2.5, SO₂, and NO₂ were observed at lag8 (PM₁₀ (OR = 1.016, 95%CI 1.002, 1.030), PM_2.5 (OR = 1.027, 95%CI 1.007, 1.048), SO₂ (OR = 1.153, 95%CI 1.040, 279) and NO₂ (OR = 1.054, 95%CI 1.005, 1.105)) while males exhibited a consistent trend from lag0 to lag8 (PM₁₀ (OR = 1.035, 95%CI 1.018, 1.053), PM_2.5 (OR = 1.056, 95%CI 1.030, 1.082), SO₂ (OR = 1.220, 95%CI 1.072, 1.389), NO₂ (OR = 1.126, 95%CI 1.061, 1.120), CO (OR = 10.059, 95%CI 1.697, 59.638) and O₃ (OR = 0.972, 95%CI 0.946, 0.999)). When gender and age were considered, a positive impact was also observed after three days cumulative effect in males.

Conclusions: There is a significant cumulative effect of exposure to air pollution on IS hospital admissions, especially the males and patients under the age of 65. Our results also suggested that a notable association between CO and NO₂ in two-pollutant models.

Background: High temperatures are known to be associated with an increased risk of self-harm, but the influence of demographic changes and country-level indicators on the burden of heat-related self-harm remains unclear. This study examined the key factors driving changes in self-harm mortality linked to high temperatures and explored their impact at the country level.

Methods: This is an ecological study that analyzes data from the 2021 Global Burden of Disease (GBD) study, the World Bank, and the Climate Research Unit (CRU) were analyzed. Decomposition analyses were used to identify key factors driving changes in high temperature-related self-harm mortality between 1990 and 2021. A panel data model assessed the impact of national indicators on heat-related self-harm mortality.

Results: In 2021, 14,885 deaths globally were attributed to heat-related self-harm, a 41.94% increase from 1990, with low-middle SDI regions accounting for 47.84% of these deaths. While the global death rate from heat-related self-harm declined slightly over this period, South Asia and low-middle SDI regions contributed most to the decline. However, population aging exacerbated mortality rates. Demographic and meteorological factors were also linked to heat-related self-harm.

Conclusion: The global decline in heat-related self-harm mortality is largely driven by reductions in females, low-middle SDI regions, and South Asia. However, population aging and growth in these regions have added to the mortality burden, slowing the overall decline. Factors such as population density are also associated with heat-related self-harm. Targeted measures are needed to mitigate heat-induced self-harm more effectively in future.

Background: The objective of this study was to investigate the potential link between myopia in adolescents and exposure to per- and polyfluoroalkyl substances (PFASs).

Methods: This investigation included 1971 subjects with accessible PFAS level data, myopia status, and associated variables from four cycles of the National Health and Nutritional Examination Survey (NHANES). The investigation focused on specific PFAS compounds found in the serum, including perfluorohexane sulfonate (PFHxS), perfluorononanoic acid (PFNA), perfluorooctanoic acid (PFOA), and perfluorooctane sulfonic acid (PFOS), chosen for their frequent detection. Owing to the skewed nature of the PFAS level data, the PFAS levels were log-transformed (Ln-PFAS) prior to analysis. Logistic regression, restricted cubic spline modeling, subgroup analysis, and sensitivity analysis were used to examine the associations between exposure to PFASs and the onset of myopia.

Results: PFOA levels were significantly associated with myopia risk (OR: 1.33; 95% CI: 1.05-1.69; P = 0.019). More specifically, with respect to the first quartile, the second quartile (OR_Q2: 1.69; 95% CI: 1.16-2.46; P = 0.007), third quartile (OR_Q3: 1.45; 95% CI: 1.03-2.03; P = 0.035), and highest quartile (OR_Q4: 1.58; 95% CI: 1.12-2.21; P = 0.010) of participants presented with increased myopia risk. Mediation analysis revealed that PFOA and myopia risk were partially mediated by serum albumin (ALB), with a mediation percentage of 22.48% (P = 0.008). A nonlinear inverted U-shaped relationship was identified between the level of PFOA and myopia risk (P for nonlinearity = 0.005).

Conclusion: Our findings suggest a potential link between exposure to PFOA and the likelihood of myopia development in young individuals and a mediating effect of serum ALB on this relationship. Notably, PFOA was identified as a key PFAS significantly contributing to the observed link between PFAS exposure and myopia risk. The potential threat of PFOA to myopia should be examined further.

Background: Depression among older adults is an important public health issue, and air and noise pollution have been found to contribute to exacerbation of depressive symptoms. This study examined the association of exposure to air and noise pollutants with clinically-newly-diagnosed depressive disorder. The mediating role of individual pro-inflammatory markers was explored.

Methods: We linked National Health Insurance claim data with 2998 healthy community-dwellers aged 55 and above who participated in the Healthy Aging Longitudinal Study between 2009 and 2013. Newly diagnosed depressive disorder was identified using diagnostic codes from the medical claim data. Pollutants were estimated using nationwide land use regression, including PM_2.5 and PM₁₀, carbon monoxide, ozone, nitrogen dioxide, sulfur dioxide, and road traffic noise. Cox proportional hazard models were employed to examine the association between pollutants and newly developed depressive disorders. The mediating effect of serum pro-inflammatory biomarkers on the relationship was examined.

Results: Among the 2998 participants, 209 had newly diagnosed depressive disorders. In adjusted Cox proportional hazard models, one interquartile range increase in PM_2.5 (8.53 µg/m³) was associated with a 17.5% increased hazard of developing depressive disorders. Other air pollutants and road traffic noise were not linearly associated with depressive disorder incidence. Levels of serum tumor necrosis factor receptor 1 mediated the relationship between PM_2.5 and survival time to newly onset depressive disorder.

Conclusion: PM_2.5 is related to an increased risk of newly developed depressive disorder among middle-aged and older adults, and the association is partially mediated by the pro-inflammatory marker TNF-R1.

Background: Chronic constipation is a long-term problem that decreases children's quality of life. Information and communication technology devices have developed rapidly in recent decades and have had various impacts on children. This prospective cohort study examined the association between television/digital versatile disc (TV/DVD) screen exposure time at age 1 and the risk of chronic constipation at age 3.

Methods: Data from 63,697 infants in the Japan Environment and Children's Study (JECS) were analyzed. We divided participants into five groups according to TV/DVD exposure time per day: no exposure (0 h), short exposure (<1 h), middle exposure (1.0-<2.0 h), long exposure (2.0-<4.0 h), and very long exposure (≥4 h). Logistic regression analysis was performed to assess the association between TV/DVD exposure time and the risk of constipation. For logistic regression analysis, odds ratios (ORs) were adjusted for sex, parents' education, household income, nursery school, feeding contents, and obesity. The interaction between the sexes was also examined.

Results: The prevalence of constipation for males, females, and all participants at age 3 was 9.3, 11.0, and 10.1%, respectively. The TV/DVD screen time distribution per day at age 1 was 10.6% for none, 34.1% for short, 29.9% for middle, 19.2% for long, and 6.2% for the very long exposure group. After adjusting for confounding factors, a dose-response pattern was identified between TV/DVD exposure time and constipation in all participants (p for trend < 0.001). The adjusted ORs increased progressively in the short (OR 1.15, 95% confidence interval [CI] 1.04-1.27), middle (OR 1.22, 95% CI 1.11-1.35), long (OR 1.37, 95% CI 1.24-1.52), and very long exposure groups (OR 1.53, 95% CI 1.35-1.74). This association was not significantly different between the sexes (p for interaction = 0.36).

Conclusions: Longer TV/DVD exposure time at age 1 was associated with the risk of chronic constipation at age 3. Excessive screen exposure may need to be avoided from infancy to decrease the risk of chronic constipation in later years.