ABSTRACT
Background Few studies have investigated the association between air pollution and arterial stiffness in Chinese population, and the findings are inconsistent. The problem of multicollinearity exists when modeling multiple air pollutants simultaneously. Objective To investigate potential association between air quality index (AQI) and population brachial-ankle pulse wave velocity (baPWV) in Beijing. Methods This study retrieved medical examination data of 2971 participants from the Beijing Health Management Cohort, who were under 60 years old and not yet retired, from January 1, 2015 to December 31, 2019. The most recent medical examination data available were utilized for this analysis. AQI data from 35 air pollution monitoring sites in Beijing and meteorological data (including atmospheric pressure, air temperature, wind speed, and relative humidity) from 16 meteorological monitoring stations from January 1, 2014 to December 31, 2019 were collected. An average AQI exposure level for 365 d before the date of physical examination for each participant was computed using inverse distance weighting. Multiple linear regression analysis was employed to investigate the relationship between AQI and baPWV in Beijing, after adjusting for confounding variables including age, gender, body mass index, mean arterial pressure, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, triglycerides, fasting blood glucose, atmospheric pressure, temperature, wind speed, relative humidity, medication history of diabetes, medication history of hypertension, cardiovascular disease, education, smoking status, drinking status, and physical activity intensity. Subgroup analysis was performed by age, sex, presence of diabetes, and presence of hypertension. Results AQI demonstrated an overall decreasing trend during the study period and was lower in the northern regions and higher in the southern regions of Beijing. After adjusting the confounding variables, each 10 unit increase in AQI was associated with 6.18 (95%CI: 1.25, 11.10) cm·s−1 increase in baPWV in all participants, 8.05 (95%CI: 2.32, 13.79) cm·s−1 increase in the participants <50 years, 15.82 (95%CI: 8.33, 23.31) cm·s−1 increase in the female group, 10.10 (95%CI: 4.66, 15.55) cm·s−1 increase in the participants without diabetes, and 9.41 (95%CI: 4.21, 14.62) cm·s−1 increase in the participants without hypertension. However, there was no statistically significant association observed between AQI and baPWV in the age group ≥50 years, the male group, the diabetic group, and the hypertensive group (P>0.05). Conclusion An increase in long-term AQI levels is associated with an elevation in the degree of arterial stiffness. Individuals under 50 years old, females, without hypertension or diabetes are susceptible populations to arterial stiffness when being exposed to air pollution. Improving air quality may contribute to prevent arterial stiffness.
ABSTRACT
Background Evidence about the association between air pollution and carotid intima-media thickness (CIMT) is inconsistent, and limited studies have explored the relationship between gaseous pollutants and CIMT. Additionally, personal activity patterns and infiltrated ambient pollution are not comprehensively considered to estimate individual exposure to air pollutants. Objective To investigate the relationship between long-term time-weighted individual exposure to ambient pollutants [fine particulate matter (PM2.5), inhalable particulate matter (PM10), nitrogen dioxide (NO2), sulfur dioxide (SO2), ozone (O3), and carbon monoxide (CO)] and the progression of CIMT. Methods This study was performed among 554 participants in the Beijing Health Management Cohort who were free of atherosclerotic lesions on carotid artery at baseline. Daily concentrations of pollutants were predicted at both residential and work addresses based on land-use regression model. With additional consideration of personal indoor and outdoor activity patterns at both addresses and exposure to ambient pollutants from traffic transportation, individual time-weighted concentration was calculated. Indoor exposure was estimated by infiltrated ambient pollutants (based on infiltration factors and land-use regression model). Personal activity patterns included type, time, location, and frequency. Exposure to ambient pollutants from different traffic transportations was estimated by the average outdoor pollutant concentrations at both residential and work addresses combined within filtration factors and time spent on commuting. Multiple linear regression was conducted to assess the association of time-weighted individual pollutant exposure and the central position of CIMT progression. Quantile regression was applied to explore the relationship between time-weighted individual pollutant exposure and the progression of CIMT on different percentiles. Results The median value of CIMT progression was 369.49 μm·year−1. PM2.5, PM10, SO2, and O3 were associated with CIMT progression in the multiple linear regression model. The largest effect sizes of PM2.5, PM10, and SO2 were obtained for one-year exposure (regression coefficient: 66.910, 64.077, and 191.070, respectively), and two-year exposure for O3 (regression coefficient: 62.197). The results of quantile regression demonstrated different effect sizes for pollutants among different percentiles on CIMT progression. Significant associations between CIMT progression and PM2.5 from P30 to P50, CO from P10 to P40, and PM10 from P30 to P60 were observed. Two-year and three-year exposures to NO2 (P10, P20 and P40) were also associated with CIMT progression. The association between SO2 and the progression of CIMT was proved on all percentiles, and larger effect sizes of one-year and two-year exposures to SO2 (except P90) were demonstrated with increasing percentiles. The upward trend for the coefficients was clearly presented from P50 to P80. Specifically, the coefficient of two-year exposure to SO2 ranged from 136.583 (P50) to 277.330 (P80). No statistically significant association was observed between O3 and CIMT progression on any percentile (P>0.05), and the results were inconsistent with those of the multiple linear regression. Conclusion Individual time-weighted exposures to PM2.5, PM10, SO2, NO2, and CO have the potential to promote the progression of CIMT, and the adverse effect of ambient pollution on atherosclerotic lesion is identified.
ABSTRACT
To compare the performance of generalized additive model (GAM) and long short-term memory recurrent neural network (LSTM-RNN) on the prediction of daily admissions of respiratory diseases with comorbid diabetes. Daily data on air pollutants, meteorological factors and hospital admissions for respiratory diseases from Jan 1st, 2014 to Dec 31st, 2019 in Beijing were collected. LSTM-RNN was used to predict the daily admissions of respiratory diseases with comorbid diabetes, and the results were compared with those of GAM. The evaluation indexes were calculated by five-fold cross validation. Compared with the GAM, the prediction errors of LSTM-RNN were significantly lower [root mean squared error (RMSE): 21.21±3.30 vs. 46.13±7.60, <0.01; mean absolute error (MAE): 14.64±1.99 vs. 36.08±6.20, <0.01], and the value was significantly higher (0.79±0.06 vs. 0.57±0.12, <0.01). In gender stratification, RMSE, MAE and values of LSTM-RNN were better than those of GAM in predicting female admission (all <0.05), but there were no significant difference in predicting male admission between two models (all >0.05). In seasonal stratification, RMSE and MAE of LSTM-RNN were lower than those of GAM in predicting warm season admission (all <0.05), but there was no significant difference in value (>0.05). There were no significant difference in RMSE, MAE and between the two models in predicting cold season admission (all >0.05). In the stratification of functional areas, the RMSE, MAE and values of LSTM-RNN were better than those of GAM in predicting core area admission (all <0.05). has lower prediction errors and better fitting than the GAM, which can provide scientific basis for precise allocation of medical resources in polluted weather in advance.
Subject(s)
Female , Humans , Male , Beijing/epidemiology , Diabetes Mellitus/epidemiology , Hospitalization , Memory, Short-Term , Neural Networks, ComputerABSTRACT
Objective To develop a new risk model for predicting type 2 diabetes (T2D) in a rural Chinese population in north China.Methods A village-based cohort study was performed.Data from Handan Eye Study conducted from 2006-2013 comprising 4 132 participants aged 30 years old (1 793 male and 2 339 female) with complete diabetes data at baseline and follow-up were analyzed.The blood biomarkers of T2D incident risk were screened and a new risk model was derived by using unconditional stepwise logistic regression after adjustment of age,body mass index (BMI),waist circumference,and family history of diabetes in random two-thirds of the sample cohort (selected randomly).In addition,a simple point system for T2D risk was built according to the procedures as described in Framingham Study,and the new risk score was subsequently validated in the final one-third of the sample cohort.Results The new risk score included age (8 points),BMI (6 points),waist circumference (8 points),family history of diabetes (9 points),fasting plasma glucose (23 points),and triglycerides (4 points).The score ranged from 0 to 58.The AUC was 0.802 (0.780-0.822) in the validation sample.At the optimal cutoff value of 27,the sensitivity and specificity were 70.27％ (58.50％-80.30％) and 80.83％ (78.60％-82.90％) respectively.Conclusions A new risk model for predicting T2D have been developed in a rural Chinese population in north China,and the risk score can be used in rural basic health care settings after validation.
ABSTRACT
Objective To investigate the risks of self-rated health in the ≥55-year elderly in Beijing and the occurrence of stroke.Methods The subjects (n=2 101;aged ≥55) from Beijing longitudinal study of aging (BLSA) were collected by Xuanwu Hospital,Capital Medical University from January 1992 to December 2016.One hundred and twenty-one subjects with stroke at baseline and 92 with incomplete information were excluded,and finally,1 888 elderly patients without cerebrovascular disease at baseline were included in the analysis.Based on the actual situation,the self-rated health was to identify an item that matched their current state from good,general to poor.The deadline for the survey was December 31,2012.The competitive risk model was used to assess the health self-rated status and the risk of stroke.Non-stroke deaths,including cancer and car accidents were treated as competitive events.Results Of the 1 888 subjects enrolled,946 (50.1%) self-rated health were good,616 (32.6%) were general,and 326 (17.3%) were poor;438 (23.2%) had stroke,751 (37.8%) had non-stroke death,and 699 (37.0%) were right censored data.Using the competing risk model and adjusting the age,sex,living area,marital status,education level,smoking,alcohol consumption,physical exercise,hypertension,diabetes mellitus,coronary heart disease,and body mass index,the occurrence of stroke in patients with poor self-rated health was 1.44 times (95%CI 1.11-1.87,P<0.01) as good as those who were good.Conclusion In the self-rated health of the elderly ≥55 years old in Beijing,the people with poor self-rated health increased the occurrence of stroke after considering the competitive risks.
ABSTRACT
This article introduces definitions of three special tests, namely, non-inferiority test (to verify that the efficacy of the experimental drug is clinically not inferior to that of the positive control drug), equivalence test (to verify that the efficacy of the experimental drug is equivalent to that of the control drug) and superiority test (to verify that the efficacy of the experimental drug is superior to that of the control drug), and methods of sample size estimation under the three different conditions. By specific examples, the article introduces formulas of sample size estimation for the three special tests, and their SAS realization in detail.