-
The purpose of this study was to evaluate the impact of PM2.5 on lung function indicators within the same group under different pollution conditions, so we performed the survey three times in December 2018, March 2019, and April 2019. December 2018 showed low-pollution weather even though it was during heating season; March 2019 showed high-pollution weather also during the heating season; April 2019 showed high-pollution weather despite being during the non-heating season. Each survey was conducted for five days and included three parts: part 1 was exposure monitoring, including the hourly concentration of PM2.5 in classrooms, dormitories, the canteen, and playgrounds for five consecutive days during the survey; part 2 consisted of questionnaire surveys, including a parent questionnaire and a schoolchild questionnaire; part 3 was a physical examination on the last day of each survey, including height, weight, and pulmonary function. The study protocol was approved by the ethics review committee of the Chinese Center for Disease Control and Prevention (
No. 201803) and the Biomedical Ethics Committee of Peking University (No. IRB00001052-18089). Parents and schoolchildren were initially informed about the study, and all participants and their parents signed a written informed consent form. -
A total of 64 schoolchildren attend grades 4th−5th of the selected boarding school in a District of Beijing, China. The inclusion criteria were as follows: age between 9 and 12 years, willingness to participate in the study and commitment to complete the survey, and boarding at the school during the study. The exclusion criteria were as follows: respiratory infections, acute or chronic lung diseases before the survey; and leaving school during the study. A total of 51 children participated in our research, including 19 boys and 32 girls.
-
During the survey period, we placed PM2.5 measurement devices (Agris, WP6930S) in classrooms, dormitories, the canteen, and playgrounds for five consecutive days. All of the instruments were uniformly calibrated. The monitoring process was as follows: According to the schedule of students’ classes, the data was read every 40–60 min on average, and three calibrated direct-reading instruments were used in the same place each time, scattered within the same room far away from the doors and windows or at the center of playground, 1.5 m above the ground. Each instrument was reads three times, and the average value of nine reading across the three instruments was calculated and considered the average exposure concentration of the location during that period.
During the survey period, schoolchildren kept an activity log every day, where they recorded their individual daily indoor and outdoor activities and travel times.
According to the indoor and outdoor activity time of schoolchildren and indoor and outdoor PM2.5 concentrations, we estimated the individual daily exposure after time-weighted averaging with the following formula:
$${E_{\text{i}}} = \sum {{C_{\text{j}}}{t_{{\text{ij}}}}} /\sum {{t_{{\text{ij}}}}} ,$$ where Ei refers to the average individual exposure concentration of subject i to PM2.5 in micro-environment j, Cj refers to the concentration of PM2.5 in micro-environment j, and tij refers to the time subject i was in micro-environment j[28].
We calculated the 24-h individual daily exposure concentration from lag 0 d to lag 4 d and the average sliding individual daily exposure concentration from lag 0−1 d to lag 0−4 d. Lag 0 d referred to the day of physical examination; lag 1 d refers to 1 d before the day of physical examination; lag 2 d refers to 2 d before the day of physical examination; lag 3 d refers to 3 d before the day of physical examination; lag 4 d refers to 4 d before the day of physical examination; lag 0−1 d refers to the day of physical examination to 1 d before the day of physical examination; lag 0−2 d refers to the day of physical examination to 2 d before the day of physical examination; lag 0−3 d refers to the day of physical examination to 3 d before the day of physical examination; lag 0−4 d refers to the day of physical examination to 4 d before the day of physical examination. The daily exposure concentration on the day of physical examination day (lag 0 d) was calculated by taking the 15-haverage from 0:00 to 15:00 (midnight to 3:00 p.m.). We chose 15:00 because the physical examinations were ended at that time.
-
The parents questionnaire and the schoolchildren respiratory health questionnaire were based on the Respiratory History and Symptom Questionnaire (ATS-DLD-78-C) recommended by the American Thoracic Association[29]. The parents questionnaire was filled up by the child’s guardian and included the following: parents’ contact information, family ethnicity, child’s birth date, home address, child’s long-term disease status with respect to such conditions as bronchitis, asthma, history of respiratory allergies, and history of allergies to food, dust, detergents, and chemicals. The schoolchildren questionnaire was filled up by the children themselves, mainly to obtain information about recent respiratory symptoms, such as cough, sputum, sore throat, medication, and exercise. Medication referred to any use of anti-cold medicine in the past week. Exercise was divided into three levels: high, medium, and low, defined as having exercised more than 2 h per day, between 1 and 2 h, and less than 1 h. Before the start of the study, the parents questionnaire and the schoolchildren questionnaire were designed to exclude schoolchildren with acute and chronic respiratory diseases. After the end of the study (that is, the fifth day, Friday), the schoolchildren questionnaire was conducted again to collect medication and exercise information during the survey period.
-
Pulmonary function was performed using a portable pulmonary function testing device (SpirometerHI-801, Omron), the indexes included VC, FVC, FEV1, PEF, FEF25%, FEF50%, FEF75%, and FEF25%–75%. The detailed method was as follows. (1) Pulmonary function test parameters were selected. (2) For the FVC test, the subject covered his/her mouth, breathed calmly, inhaled as much as possible, then exhaled as much as possible, and returned to normal breathing after the last expiration. (3) For the FEV1 test, the subject covered his/her mouth, breathed calmly, inhaled as much as possible, and then expelled the air as quickly as possible. These pulmonary function tests need to be repeated at least three times. Comparing the three measurements, the difference between the maximum and minimum values of FVC must be less than 0.15 L (ΔFVC < 0.15 L), the difference between the maximum and minimum values of FEV1 must be less than 0.15 L (ΔFEV1 < 0.15 L), and the difference between the maximum and minimum values of PEF must be less than 0.67 L/s (ΔPEF < 0.67 L/s). The measurement result with the largest total value of FVC + FEV1 was selected as the final test result.
-
Basic information of schoolchildren, individual PM2.5 exposure data, and pulmonary function data were analyzed using basic statistical techniques. Continuous variables were expressed as mean (Mean, M) ± standard deviation (SD) or median and interquartile range (IQR) and were analyzed using the Analysis of Variance (ANOVA) or Wilcox symbol quality test, respectively. Categorical variables were expressed in terms of frequency and percentage. The Pearson chi-square test and Fisher’s exact test were used for statistical comparison among groups. Using PM2.5, gender, age, body mass index (BMI), allergy, medication, and exercise as fixed variables, survey times and ID as random variables, and pulmonary function indicators as dependent variables, a linear mixed-effects model was built to examine the influence on pulmonary function after different delays. The basic model was as follows:
$${Y_{{\text{ij}}}} = \sum {{\beta _{{\text{ij}}}}{X_{{\text{ij}}}}} + {\gamma _{\text{i}}}{Z_{\text{i}}} + {\gamma _{\text{j}}}{W_{\text{j}}} + u,$$ where Y refers to pulmonary function indicators VC, FVC, FEV1, FEV1/FVC, PEF, FEF25%, FEF50%, FEF75%, and FEF25%-75%, X refers to fixed variables PM2.5, gender, age, BMI, allergy, medication, and exercise, Z refers to the random variable ‘ID’, W refers to the random variable ‘survey time,’ β refers to the coefficient of the fixed variable, γ refers to the coefficient of the random variable, i refers to the i-th research object, j refers to the j-th survey, and u refers to the error.
Sensitivity analysis was conducted to evaluate the stability of the model. The effects of PM2.5 on all of the pulmonary function indicators were calculated with units of PM2.5 increase of 10 μg/m3.
All of the data were statistically analyzed using R 3.5.1 and the lme4 packages.
-
A total of 51 schoolchildren participated in the present study (Table 1); 37.25% were boys and 62.75% were girls, and they ranged in ages from 10.13−10.74 years old. BMIs were 18.12 −18.56 kg/m2, there were 8 schoolchildren who had histories of allergies to food, dust, detergents, or chemicals, 5, 4, and 6 schoolchildren had taken anti-cold medicine during the three survey periods, and more than 40% of schoolchildren exercised more than 2 h/d. The rate of exercise in the third survey was significantly different from in the other two surveys, P < 0.05.
Items First survey (n = 51) Second survey (n = 51) Third survey (n = 51) Age (mean ± SD) 10.13 ± 0.63 10.74 ± 0.84 10.74 ± 0.84 Gender (n, %) Boys 19 (37.25) 19 (37.25) 19 (37.25) Girls 332 (62.75) 332 (62.75) 332 (62.75) BMI (kg/m2, mean ± SD) 18.12 ± 3.68 18.44 ± 3.32 18.56 ± 3.45 Allergy (n, %) Yes 8 (15.69) 8 (15.69) 8 (15.69) No 43 (84.31) 43 (84.31) 43 (84.31) Taking anti-cold medicine (n, %) Yes 5 (9.80) 4 (7.80) 6 (11.76) No 46 (90.20) 47 (92.16) 45 (88.24) Average daily exercise (h/d), (n, %) > 2 22 (43.14) 24 (47.06) 25 (49.02)* 1−2 17 (33.33) 16 (31.37) 10 (19.61)* < 1 112 (23.53) 11 (21.57) 16 (31.37)* Note. n, sample size; SD, standard deviation. *Means significant difference compared with other groups, P < 0.05. Table 1. Basic characteristics of participants
-
Figure 1 shows the average daily indoor and outdoor PM2.5 concentrations and individual exposure concentrations during the survey. As shown, during the first survey (December 3–7, 2018), the outdoor PM2.5 concentrations < 40 μg/m3, indicating good air quality. On some days of the second (March 4–8, 2019) and third (April 21–25, 2019) surveys, PM2.5 concentrations reached > 100 μg/m3, which were classified as slight pollution. A certain difference between indoor and outdoor concentrations was observed in the same time period. The individual PM2.5 exposure concentrations (calculated based on time activity patterns and micro-environmental PM2.5 concentrations) were between indoor and outdoor PM2.5 concentrations.
Table 2 shows the average individual exposure levels on different days, including daily average concentration on the day of physical examination, daily average concentration, and average sliding concentration 0 to 4 d before the physical examination. The highest exposure concentration was measured at lag 4 d of the second survey, and the lowest individual exposure concentration was measured on the day of the first pulmonary function test. The median of average individual daily PM2.5 concentrations at physical examination day was 6.00 μg/m3, the median concentration of lag 1 d to lag 4 d was 17.41 μg/m3, 30.75 μg/m3, 41.73 μg/m3, and 57.75 μg/m3, respectively, and on lag 0–1 d to lag 0–4 d it was 15.26 μg/m3, 18.7 μg/m3, 24.46 μg/m3, and 40.02 μg/m3, respectively.
Exposure
dayFirst survey (n = 51),
M (P25, P75)Second survey (n = 51),
M(P25, P75)Third survey (n = 51),
M (P25, P75)Total median,
M (P25, P75)Lag 4 d 17.15 (16.80, 17.15) 102.27 (100.39, 102.27) 57.75 (57.75, 57.75) 57.75 (17.15, 100.39) Lag 3 d 13.29 (13.29, 13.29) 41.73 (41.66, 41.73) 87.63 (87.63, 89.05) 41.73 (13.29, 87.63) Lag 2 d 30.75 (30.75, 30.75) 9.18 (9.18, 9.18) 65.95 (65.95, 67.42) 30.75 (9.18, 65.95) Lag 1 d 2.32 (2.32, 2.32) 17.41 (17.41, 17.41) 24.51 (23.95, 24.51) 17.41 (2.32, 23.95) Lag 0 d 1.94 (1.94, 1.94) 29.50 (29.50, 29.50) 6.00 (4.00, 6.00) 6.00 (1.94, 29.50) Lag 0–1 d 2.14 (1.94, 2.32) 23.46 (17.41, 29.50) 15.26 (6.00, 24.51) 15.26 (2.14, 23.46) Lag 0–2 d 11.67 (5.67, 26.97) 18.70 (9.18, 26.97) 32.16 (23.95, 65.95) 18.7 (11.67, 31.79) Lag 0–3 d 12.08 (6.78, 30.76) 24.46 (17.41, 41.66) 46.03 (23.95, 87.63) 24.46 (12.08, 46.03) Lag 0–4 d 13.10 (6.78, 26.97) 40.02 (26.97, 100.39) 48.37 (24.51, 57.75) 40.02 (13.10, 48.37) Note. n, sample size; M, median; P25, lower quartile; P75, upper quartile. Table 2. Individual PM2.5 exposure concentrations (μg/m3)
-
Table 3 shows the mean and standard deviation of the three surveys of pulmonary function indicators. The first survey pulmonary function indicators, that is, VC, FVC, FEV1 and FEV1/FVC were significantly different from the other two surveys (P < 0.01). No significant difference was observed for other indicators.
Pulmonary function indexes First survey (mean ± SD) Second survey (mean ± SD) Third survey (mean ± SD) VC (L) 2.06 ± 0.38* 2.44 ± 0.58 2.34 ± 0.35 FVC (L) 1.95 ± 0.39* 2.17 ± 0.43 2.22 ± 0.38 FEV1 (L) 1.80 ± 0.34* 1.86 ± 0.38 1.94 ± 0.37 FEV1/FVC (%) 0.93 ± 0.08* 0.85 ± 0.10 0.87 ± 0.09 PEF (L/s) 3.36 ± 0.91 3.39 ± 1.02 3.48 ± 0.81 FEF75% (L/s) 1.49 ± 0.44 1.28 ± 0.52 1.46 ± 0.52 FEF50% (L/s) 2.59 ± 0.71 2.44 ± 0.72 2.53 ± 0.73 FEF25% (L/s) 3.12 ± 1.03 3.18 ± 0.96 3.21 ± 0.84 FEF25%–75% (L/s) 2.25 ± 0.75 2.15 ± 0.68 2.32 ± 0.66 Note. SD, standard deviation; *P < 0.01, between this group and the other two groups. VC, vital capacity; FVC, forced vital capacity; FEV1, forced expiratory volume in one second; FEV1/FVC, expiratory volume in one second/forced vital capacity; PEF, peak expiratory flow; FEF25%, 25% forced expiratory flow; FEF50%, 50% forced expiratory flow; FEF75%, 75% forced expiratory flow; FEF25%–75%, 25%−75% forced expiratory flow. Table 3. Three measurements of pulmonary function
-
With lag 4 d exposure data, we excluded three independent variables (medication, allergy, and exercise) one by one for model sensitivity analysis (Table 4). We found that after excluding three independent variables separately, all of the pulmonary function indicators did not change significantly, indicating that the results have a high degree of confidence.
Variables VC (mL) FVC (mL) FEV1 (mL) FEV1/FVC (%) PEF (mL/s) FEF25% (mL/s) FEF50% (mL/s) FEF75% (mL/s) FEF25%−75% (mL/s) Basic model −37.29
(−54.51, −20.08)−32.55
(−50.26, −15.26)3.32
(−14.98, 21.61)−8.55
(−13.06, −4.04)−8.12
(−47.35, 31.10)−22.04
(−33.81, −10.28)−23.85
(−52.68, 4.99)−19.99
(−49.23, 9.24)−18.38
(−37.96, 1.20)− Medicine −37.97
(−55.26, −20.68)−32.69
(−50.43, −15.67)4.13
(−14.25, 22.51)−8.47
(−12.99, −3.96)−4.07
(−43.81, 35.66)−21.86
(−33.93, −9.79)−22.15
(−51.27, 6.97)−18.44
(−47.78, 10.91)−18.66
(−38.31, 0.93)− Allergy −37.79
(−55.10, −20.48 )−32.71
(−50.45, −15.13)3.98
(−14.57, 22.54)−8.03
(−12.56, −3.49)−6.21
(−45.55, 33.13)−21.94
(−33.78, −10.10)−22.87
(−51.79, 6.05)−18.30
(−47.56, 10.97)−18.67
(−38.46, 0.56)− Exercise −39.07
(−55.83, −22.31)−32.14
(−50.13, 15.69)4.90
(−13.75, 23.54)−8.30
(−12.68, −3.92)−2.08
(−40.33, 36.17)−22.10
(−33.99, −10.21)−21.70
(−49.68, 6.28)−17.38
(−45.70, 10.94)−18.56
(−38.11, 1.13)Note. The basic model independent variables were PM2.5, age, gender, BMI, exercise, allergies, and medication, and the dependent variables were indicators of pulmonary function. Allergies referred to history of allergies to food, dust, detergents, and chemicals. Each child’s long-term disease status such as bronchitis, asthma, history of respiratory allergies was excluded at the begin of the study. Medication referred to the situation of taking anti-cold medicine during the survey period. BMI was divided into three levels, < 18.5, 18.5−23.5, and > 23.5. All of the indicators were calculated based on each 10 μg/m3 increase in PM2.5. VC, vital capacity; FVC, forced vital capacity; FEV1, forced expiratory volume in one second; FEV1/FVC, expiratory volume in one second/forced vital capacity; PEF, expiratory flow; FEF25%, 25% forced expiratory flow; FEF50%, 50% forced expiratory flow; FEF75%, 75% forced expiratory flow; FEF25%–75% , 25%−75% forced expiratory flow. Table 4. Model sensitivity analysis (using lag 4 d data)
-
With the control of PM2.5, gender, age, BMI, allergy, medication, and exercise, we used the linear mixed-effects model to analyze the lag effects of individual PM2.5 exposure on pulmonary function. As shown in Figure 2, PM2.5 had no significant immediate effect on pulmonary function on the day of physical examination, but PM2.5 had a lagged effect on pulmonary function on lag days. Acute PM2.5 exposure can cause a decrease in VC, FVC, FEV1, FEV1/FVC, FEF25%, FEF50%, and FEF25%–75% on lag days; the largest declines were observed in FEF25%–75%, FEV1/FVC, FEF75%, and FEV1 on lag 0–1 d, 80.44 mL/s, 35.85%, 78.58 mL/s, and 61.34 mL, respectively, while the largest decreases were observed in FEF25% (83.68 mL/s) on lag 1 d, VC (32.34 mL) on lag 4 d, and FVC (37.76 mL) on lag 0–4 d.
-
The subgroup analysis revealed that PM2.5 has inconsistent effects on pulmonary function between boys and girls (Figures 3–4), and the decline in most indicators was greater for boys than for girls. Each 10 μg/m3 increase in PM2.5 on the day of physical examination caused a decrease in FEV1 and FEV1/FVC in boys, that is, 62.26 mL and 25.85%, respectively. Among average sliding concentrations, each 10 μg/m3 increase in PM2.5 on lag 0–1 d caused a decrease in FEF25%–75% and FVC in boys of 99.92 mL/s and 82.11 mL, respectively; on lag 0–2 d and lag 0–3 d it caused a decrease in FEV1 in boys of 106.58 mL and 65.10 mL, respectively; and on lag 0–4 d it caused a decrease in FEF25%–75% and FVC in boys of 66.92 mL/s and 16.50 mL, respectively. No significant changes were found in girls at the same number of days. Among average daily concentrations, each 10 μg/m3 increase in PM2.5 on lag 1 day caused decreases in both boys and girls, in such indexes as FEV1, FVC, FEV1/FVC, FEF25%, and FEF25%–75%, and the decreases were more pronounced in boys than in girls. The same results were obtained for FEF25%–75% on lag 3 d, for FEV1, FEV1/FVC, FEF25%, and FEF25%–75% on lag 4 d. The changes in rest indexes were not statistically significant (Figures 3–4).
Acute Effects of Individual Exposure to Fine Particulate Matter on Pulmonary Function in Schoolchildren
doi: 10.3967/bes2020.086
- Received Date: 2020-04-08
- Accepted Date: 2020-06-12
-
Key words:
- PM2.5 /
- FEV1 /
- VC /
- PEF /
- FEF25%–75%
Abstract:
Citation: | YANG Xiao Yan, WEN Bo, HAN Feng, WANG Chong, ZHANG Shao Ping, WANG Jun, XU Dong Qun, WANG Qin. Acute Effects of Individual Exposure to Fine Particulate Matter on Pulmonary Function in Schoolchildren[J]. Biomedical and Environmental Sciences, 2020, 33(9): 647-659. doi: 10.3967/bes2020.086 |