In hemodialysis patients, brachial-ankle pulse wave velocity (baPWV) levels are affected by particulate matter with an aerodynamic diameter of 10 μm or less (PM10). We conducted this study to determine whether there is an association between short- and long-term PM10 exposure and baPWV in apparently healthy adults aged 40 years and older.
A total of 1,628 subjects who underwent health examinations between 2006 and 2009 were included in the study. On the basis of the day of medical screening, the 1–3-day and 365-day moving averages of PM10 concentrations were used to evaluate the association between short- and long-term exposure to PM10 and high baPWV (≥the third quartile of baPWV, 1,534 cm/s) using logistic regression models. Additional subgroup analyses were conducted according to age, sex, obesity (body mass index ≥25.0 kg/m2), and comorbidities such as metabolic syndrome.
No statistically significant associations were identified between short-term and long-term exposure to PM10 and baPWV in any of the subjects and subgroups. A 10-μg/m3 increase in the 2-day moving average of PM10 exposure was marginally associated with high baPWV in non-obese subjects (odds ratio, 1.059; P=0.058). This association in non-obese subjects was significantly different from that in obese subjects (P=0.038).
This study did not show statistically significant associations between short-term and long-term exposure to PM10 and baPWV in apparently healthy subjects. With short-term exposure to PM10, non-obese subjects showed a marginally unfavorable association with baPWV. Further studies are necessary to validate and elucidate the mechanism underlying the effect of PM10 on baPWV.
According to the World Health Organization burden of disease from ambient air pollution reports, air pollution, especially that caused by particulate matter (PM), is a risk factor associated with acute lower respiratory infections, chronic obstructive pulmonary disease, lung cancer, ischemic heart disease (IHD), and stroke [
Aortic PWV is an independent predictor of cardiovascular morbidity and mortality in patients with end-stage renal disease and essential hypertension as well as in the general population [
Therefore, this study aimed to determine whether the correlation between long- and short-term exposure to PM10 and baPWV exists in apparently healthy cancer-free adults aged 40 years and older and to identify susceptible subgroups.
We initially investigated cancer-free patients who underwent medical screening examinations at Seoul National University Health Promotion Center between January 1, 2006, and December 31, 2009. Since the reference values of baPWV are influenced by age, participants aged 40 years and older were included in this study. Medical and demographic data were collected from chart reviews and online databases at our hospital. Patients with missing variables, such as blood pressure, fasting blood glucose, cholesterol levels (including high-density lipoprotein [HDL] cholesterol and triglycerides), waist circumference, body mass index (BMI), smoking status, drinking status, regular exercise, residence location, baPWV measurements, and antihypertensive and antidiabetic medication status were excluded from the study (
This study was conducted in accordance with the tenets of the Declaration of Helsinki and the need to obtain patient consent was waived by the Seoul National University Hospital Institutional Review Board (IRB approval no., E-1803-108-932).
Air pollution levels were recorded using a network of 265 monitoring stations near the patients’ living areas in South Korea. Data on air quality status obtained by the Korean Ministry of Environment were used [
Hypertension and diabetes status, including the medication details, were ascertained using self-administered questionnaires. Patients were classified as nonsmokers, ex-smokers, and current smokers using self-administered structured questionnaires, which were rechecked by trained nurses. Heavy alcohol consumption was defined as alcohol consumption of more than 30 g/d (210 g/wk) [
Metabolic syndrome was defined according to the National Cholesterol Education Program Adult Treatment Panel III (NCEP-ATP III) and American Heart Association/National Heart Lung and Blood Institute (AHA/NHLBI) [
BaPWV was measured using a Vascular Profiler 1000 (VP-1000; Omron Healthcare, Kyoto, Japan), with blood pressure cuffs placed around the arms and legs in the supine position, and was calculated using the equation: baPWV=(D1-D2)/t, where D1 is the distance between the heart and ankle, D2 is the distance between the heart and brachium, and t is the transit time between brachial arterial waves and tibial arterial waves. The baPWV was measured by a trained technician.
Previously, a baPWV cutoff value of 1,547 cm/s, which is close to the highest quartile value of the outcome variable of baPWV (=1,534 cm/s) in this study, has been suggested as an independent predictor of asymptomatic coronary artery disease evaluated by coronary computed tomography angiography. Considering previous studies, we used a cutoff value of 1,534 cm/s for our data analysis.
Subjects were categorized into two groups: those less than the third quartile and those greater than or equal to the third quartile of baPWV (1,534 cm/s, 75th percentile). P-values were calculated using the t-test for continuous variables and chi-square test for categorical variables in the baPWV subgroups (< or ≥1,534 cm/s). Univariate and multivariate logistic regression analyses were performed using a baPWV cutoff value of 1,534 cm/s as an outcome variable and PM10 as exposure variables (short- and long-term), including age, sex, the use of antihypertensive medication and anti-diabetic medication, systolic blood pressure, daily mean temperature, year of examination, and residence area of participants as covariates. An effect modification analysis was conducted to investigate vulnerable subgroups by including an interaction term of PM10 exposure and a subgroup in addition to the covariates in the main models. The odds ratios (ORs) and 95% confidence intervals were estimated. Statistical analysis was performed using Stata ver. 16.0 (Stata Corp., College Station, TX, USA).
A total of 1,628 patients were eligible for the study, as shown in
In the study population, age, hypertensive medication, diabetes mellitus medication status, blood pressure (both systolic and diastolic), and glucose levels were statistically different between the high (≥1,534) and low (<1,534) baPWV groups. More specifically, patients with high baPWV were mostly older, were more likely to take hypertensive medication, had higher blood pressure, used more diabetes medication, and had higher glucose levels. There were no statistically significant differences between PM10 levels at Lag 0 to Lag 3 days and in moving averages (
When we assessed the short-term PM10 exposure through the 24-hour moving average of Lag 0 to Lag 3 days prior to medical examination, we found no statistically significant associations in the total study population and subgroups. When we investigated the subgroups according to age, sex, BMI, anti-hypertensive and anti-diabetes medication status, and metabolic syndrome with short-term exposure to PM10 from Lag 0 to Lag 3, participants with a lower BMI (<25.0 kg/m2) presented a slightly greater risk of high PWV compared to their counterparts with a moving average of Lag 0–1. In particular, a 10-μg increase in the 24-hour moving average of PM10 exposure of Lag 0–1 was marginally associated with a 6% elevated risk of high baPWV in non-obese subjects. Long-term exposure assessed through a 365-day moving average of PM10 (PM10 Lag 0–365) was not associated with an increased risk of high baPWV among all subjects and subgroups.
Previous studies on the association between PM10 and baPWV showed inconsistent results due to differences in study populations and small sample sizes. In a systematic review, based on the published literature until January 2017, Zanoli et al. [
In 2018, in a well-characterized large community-based cohort of the Framingham Heart study [
Several mechanisms underlying the association between PM and PWV have been postulated. Mild and transitory inflammatory stimuli have been proposed as possible links between arterial stiffening [
The possible mechanism underlying the differential effect of short-term exposure to PM10 on high PWV in non-obese subjects may be as follows. Previously, Tang et al. [
This study used various laboratory data and clinical measurements to determine specific vulnerable subgroups and explain the potential links underlying the relationship between PM and inflammation, fitness, level of adiposity, and PWV. However, our study had some limitations. Although we analyzed the association between PM10 and baPWV in short- and long-term exposures, we did not analyze PM2.5, because PM2.5 data were only available for the city of Seoul between 2006 and 2009, whereas PM10 was available for the entire nation. The study population we used underwent private health screening tests in a single university hospital; thus, this group may not be representative of the general population. Furthermore, we used both short- and long-term PM exposure assessments with fixed monitoring data, which may cause non-differential misclassification at the individual level of exposure and underestimate the association in long-term exposure-related adverse health effects. Finally, because we did not have information regarding the timing of the activities of each patient, we could not ascertain the exact location and timing of exposure to PM10, which could lead to a bias if the PM10 levels between the patient’s workplace and residence differ. Nevertheless, because exposure misclassification could lead to a null hypothesis, we can presume that our results could underestimate the effects of PM10 compared to those that could be obtained by measuring PM10 exposure levels with more precision. In other words, because we did not install 24-hour air pollution monitoring devices for each patient, we could not consider the air pollution levels when commuting to and from work; thus, a more sophisticated modeling approach is required in future studies.
In conclusion, we showed the probable adverse effect of short-term exposure to PM10 on high baPWV in non-obese subjects compared to their obese counterparts. Further large and prospective studies are necessary to validate and elucidate the mechanism underlying of the influence of PM on baPWV, especially in healthy, non-obese subjects.
No potential conflict of interest relevant to this article was reported.
Supplementary materials can be found via
Differential effect of short-term exposure of PM10 on brachial-ankle pulse wave velocity, according to age, gender, BMI, hypertension, diabetes status, and MetS, all with PM10 moving averages.
Inclusion and exclusion criteria of the eligible study population. SNUH, Seoul National University Hospital; PWV, pulse wave velocity; HPDP, Health Promotion Disease Prevention Center.
Acute effect of PM10 on brachial-ankle pulse wave velocity, according to the subgroups of age (A), gender (B), body mass index (BMI) (C), hypertension (D), diabetes mellitus (DM) status (E), and metabolic syndrome (MetS) (F). The PM10 moving averages are shown on the x-axis. *Asterisks indicate statistically significant differences between subgroups with 95% confidence intervals (P-value for interaction <0.05). Adjustments were made for age, sex, BMI, district (region), systolic blood pressure, average daily temperature, hypertensive medication, DM medication, and year of examination. PM10, particulate matter with an aerodynamic diameter of 10 μm or less; OR, odds ratio.
Baseline characteristics according to baPWV cutoff of 1,534 (cm/s)
Characteristic | Total | PWV ≥1,534 (cm/s) | PWV <1,534 (cm/s) | P-value | |
---|---|---|---|---|---|
No. of patients | 1,628 | 408 | 1,220 | ||
Age (y) | 56.0±8.4 | 62.1±8.3 | 54.0±7.5 | <0.001 | |
Male | 862 (53.0) | 229 (56.1) | 633 (51.9) | 0.13 | |
Body mass index (kg/m2) | 24.4±3.1 | 24.7±3.0 | 24.3±3.1 | 0.01 | |
Diabetic medication | 146 (9.0) | 75 (18.4) | 71 (5.8) | <0.001 | |
Hypertensive medication | 438 (26.9) | 181 (44.4) | 257 (21.1) | <0.001 | |
Heavy alcohol drinking | 139 (8.5) | 29 (7.1) | 110 (9.0) | 0.23 | |
Smoking status | 0.15 | ||||
Never smoker | 932 (57.2) | 232 (56.9) | 700 (57.4) | ||
Ex-smoker | 436 (26.8) | 121 (29.7) | 315 (25.8) | ||
Current smoking | 260 (16.0) | 55 (13.5) | 205 (16.8) | ||
Systolic BP (mm Hg) | 130.0±16.1 | 141±14.4 | 125.7±14.7 | <0.001 | |
Diastolic BP (mm Hg) | 78.2±10.6 | 83.7±10.6 | 76.3±9.9 | <0.001 | |
Cholesterol | 203.9±35.6 | 205.6±34.6 | 203.4±35.9 | 0.27 | |
High-density lipoprotein (mg/dL) | 54.6±13.4 | 53.2±13.2 | 55.0±13.5 | 0.02 | |
Glucose (mg/dL) | 95.6±20.5 | 103.4±26.5 | 93.0±17.3 | <0.001 | |
Metropolitan living | 1,007 (61.9) | 270 (66.2) | 737 (60.4) | 0.04 | |
Capital living | 866 (53.2) | 231 (56.6) | 635 (52.1) | 0.11 | |
Average daily temperature (°C) | 13.9±9.5 | 13.4±9.8 | 14.1±9.4 | 0.18 | |
Lag days of PM10 (μg/m3) | |||||
Lag 0 | 57.3±32.2 | 58.2±33.2 | 57.0±31.9 | 0.52 | |
Lag 0–1 | 55.9±28.9 | 56.8±29.9 | 55.7±28.5 | 0.49 | |
Lag 0–2 | 55.2±26.7 | 55.5±26.4 | 55.1±26.8 | 0.79 | |
Lag 0–3 | 54.6±24.9 | 54.4±23.8 | 54.6±25.3 | 0.88 | |
Lag 0–365 | 57.7±6.5 | 57.4±6.5 | 57.8±6.5 | 0.24 |
Values are presented as mean±standard deviation for continuous variables and or number (%) for categorical variables.
baPWV, brachial-ankle pulse wave velocity; BP, blood pressure; PM10, particulate matter with an aerodynamic diameter of 10 μm or less; PM10 Lag 0–365, 1-year average PM10 level; PM10 Lag 0, 24-hour average PM10 on the day of health examination; PM10 Lag 0–n, average PM10 levels on the concurrent day and n previous days.
Logistic regression analysis with baPWV cutoff 1,534 cm/s as the dependent variable (N=1,628)
Variable | Unadjusted OR (95% CI) | P-value | Adjusted OR (95% CI) |
P-value |
---|---|---|---|---|
Age | 1.13 (1.12–1.15) | <0.001 | 1.13 (1.11–1.15) | <0.001 |
Male | 1.19 (0.95–1.49) | 0.14 | 0.99 (0.78–1.26) | 0.95 |
BMI | 1.05 (1.01–1.09) | 0.01 | 1.08 (0.95–1.23) | 0.41 |
Systolic BP | 1.07 (1.06–1.08) | <0.001 | 1.08 (1.07–1.09) | <0.001 |
Diastolic BP | 1.07 (1.06–1.09) | <0.001 | 1.10 (1.08–1.12) | <0.001 |
Triglycerides | 1.002 (1.001–1.003) | 0.002 | 1.002 (1.001–1.004) | <0.001 |
Cholesterol | 1.002 (0.999–1.005) | 0.27 | 1.005 (1.001–1.009) | 0.01 |
High-density lipoprotein | 0.99 (0.98–1.00) | 0.02 | 0.99 (0.98–1.00) | 0.19 |
Glucose | 1.02 (1.02–1.03) | <0.001 | 1.02 (1.01–1.02) | <0.001 |
Lag days of PM10 | ||||
Lag 0 | 1.001 (0.998–1.005) | 0.52 | 1.002 (0.998–1.006) | 0.36 |
Lag 0–1 | 1.001 (0.997–1.005) | 0.50 | 1.002 (0.998–1.007) | 0.36 |
Lag 0–2 | 1.001 (0.996–1.005) | 0.79 | 1.001 (0.996–1.006) | 0.76 |
Lag 0–3 | 1.000 (0.995–1.004) | 0.88 | 1.000 (0.994–1.005) | 0.88 |
Lag 0–365 | 0.990 (0.973–1.007) | 0.24 | 0.996 (0.969–1.023) | 0.75 |
baPWV, brachial-ankle pulse wave velocity; OR, odds ratio; CI, confidence interval; BMI, body mass index; BP, blood pressure; PM10, particulate matter with an aerodynamic diameter of 10 μm or less; PM10 Lag 0–365, 1-year average of PM10 level; PM10 Lag 0, 24-hour average of PM10 on the day of health examination; PM10 Lag 0–n, average PM10 levels on the concurrent day and n previous days.
ORs are expressed as per 1 μg/m3 increase in average residential PM10 concentration. Logistic regression was done for predictors and baPWV cutoff 1,534 cm/s which are categorical variables. In adjusted models, adjustment was done for age, sex, BMI, district (region), systolic BP, average daily temperature, hypertensive medication, diabetes medication, and year of examination.
The association of PM10 Lag 0–365 (long-term exposure to PM10) with baPWV in subgroups according to clinical risk factors, demographics, and laboratory findings
Variable | Over PWV cutoff/subgroup population | PWV cutoff 1,534 cm/s |
|
---|---|---|---|
OR (95% CI) | P-value | ||
Age ≥65 y | 160/278 | 0.961 (0.902–1.023) | 0.21 |
Age <65 y | 248/1,350 | 1.004 (0.972–1.037) | 0.82 |
Male | 229/862 | 1.001 (0.962–1.042) | 0.95 |
Female | 179/766 | 0.988 (0.943–1.035) | 0.61 |
BMI ≥25.0 kg/m2 | 181/639 | 0.991 (0.948–1.036) | 0.69 |
BMI <25.0 kg/m2 | 227/989 | 1.001 (0.961–1.042) | 0.97 |
With hypertension | 181/438 | 0.985 (0.935–1.038) | 0.55 |
No hypertension | 227/1,190 | 0.991 (0.954–1.028) | 0.62 |
With diabetes | 75/146 | 1.076 (0.957–1.211) | 0.22 |
No diabetes | 333/1,482 | 0.987 (0.957–1.018) | 0.41 |
Never smoker | 232/932 | 1.014 (0.974–1.056) | 0.50 |
Ex-smoker | 121/436 | 0.986 (0.934–1.041) | 0.62 |
Current smoker | 55/260 | 0.955 (0.873–1.045) | 0.32 |
Regular exercise | 195/720 | 1.009 (0.963–1.057) | 0.72 |
Not regular | 213/908 | 0.982 (0.944–1.021) | 0.33 |
Metabolic syndrome | 182/454 | 0.971 (0.926–1.019) | 0.24 |
No metabolic | 226/1,174 | 1.006 (0.967–1.046) | 0.78 |
Adjustment was done for age, sex, BMI, district (region), systolic blood pressure, average daily temperature, hypertensive medication, diabetes medication, and year of examination.
PM10, particulate matter with an aerodynamic diameter of 10 μm or less; PM10 Lag 0–365, 1-year average of PM10 level; baPWV, brachial-ankle pulse wave velocity; PWV, pulse wave velocity; OR, odds ratio; CI, confidence interval; BMI, body mass index.