Journal List > J Korean Med Sci > v.34(Suppl 1) > 1139091

TOOLS
Jung, Na, Lee, and Jang: Elderly Mortality and Exposure to Fine Particulate Matter and Ozone
Original Article
Preventive & Social Medicine

Journal of Korean Medical Science 2019; 34(48): e311.

Published online: 15 November 2019

DOI: https://doi.org/10.3346/jkms.2019.34.e311

Elderly Mortality and Exposure to Fine Particulate Matter and Ozone

En-Joo Jung1, Wonwoong Na2, Kyung-Eun Lee3, Jae-Yeon Jang1

1Department of Preventive Medicine, Ajou University School of Medicine, Suwon, Korea.

2SK Hynix SHE R&D Center, Icheon, Korea.

3Occupational Safety and Health Research Institute, Korea Occupational Safety and Health Agency, Ulsan, Korea.

Address for Correspondence: En-Joo Jung, MD, MPH. Department of Preventive Medicine, Ajou University School of Medicine, 164 World cup-ro, Yeongtong-gu, Suwon 16499, Republic of Korea. iyness@naver.com

Received 11 June 2019       Accepted 15 October 2019

© 2019 The Korean Academy of Medical Sciences.

The Korean Academy of Medical Sciences

(open-access):

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (https://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Background

The effects on particulate matter (PM) and ozone on health are being reported by a number of studies. The effects of these air pollutants are likely to be stronger in the elderly population, but studies in this regard are scarce. The purpose of this study was to study the effects of PM ≤ 2.5 μm and ozone on chronic health effects of the elderly population.

Methods

In order to analyze the health status of the elderly population, National Statistical Office Mortality records were used. In this study, we calculated the number of deaths in Seoul of people who were 60 years or older between 2002 and 2012. The current study analyzed each disorder separately and the lag effect. PM and ozone were analyzed using the single exposure model, as well as the adjusted multi exposure model.

Results

In the single exposure analysis with PM2.5 as the exposure variable, with the increase of 10 μg/m3, the number of deaths increased by 1.0039 fold, and vascular disease 1.0053 fold. In the multi exposure model adjusting for ozone, the number of deaths increased by 1.0037 fold, and vascular disease 1.0049 fold. In the single exposure analysis with ozone as the exposure variable, with the increase of 10 ppb, the number of deaths increased by 1.0038 fold, and in the multi exposure model adjusting for PM2.5, the number of deaths increased by 1.0027 fold. These results differed depending on the period or season. There was a 5-day lag effect between PM2.5 and deaths in the multi exposure model, and 1.0028 fold when adjusted for ozone. There was a 1-day lag effect in single exposure models with ozone as the main variable, and 1.0027 fold increase in deaths.

Conclusion

In our study, an increase in the number of deaths in the elderly population in accordance with the increase in the PM2.5 and ozone was found. The association found in our study could also influence socioeconomic burden. Future studies need to be performed in regards to younger population, as well as other air pollutants.

Graphical Abstract

jkms-34-e311-abf001

Keywords: Aged, Fine Particulate Matter, Ozone, Mortality, Disease Subgroups

INTRODUCTION

In the past, a number of studies have considered particulate matter (PM) and ozone in terms of morbidity and mortality.1234567 There have also been studies focusing on the effects of air pollutants on the elderly population, in which hospital visits (e.g., inpatient or outpatient) and mortality due to cardiopulmonary disease or asthma were researched.891011121314 Furthermore, similar studies have been conducted in Korea, in which the associations between air pollutants with depression and insulin resistance,1516 as well as acute stroke death were found.17 Previous studies considering the relationship between air pollution and elderly health have used time series analysis,18192021222324 often using PM and ozone as exposure variables, and mortality as the outcome variable. In 2003, Fischer et al.25 used the mortality record in the Netherlands and found associations between PM10, pneumonia and mortality as well as in regards to ozone, cardiovascular problems, pneumonia and mortality in people 75 years or older. Also, a study published by Goldberg et al.26 in 2013 found that people, aged 65 years or older, with a history of heart failure or atrial fibrillation had and increased mean percent change of mortality due to PM2.5, and those with a history of coronary artery disease had an increased mean percent change of mortality due to ozone. Furthermore, Di et al.27 reported in 2017 that an increase of 10 μg/m3 in PM2.5 and 10 ppb in warm-season ozone were statistically significantly associated with a relative increase in all-cause daily mortality rate. Despite the importance of prior studies, they have been limited to specific illnesses such as cardiovascular or respiratory diseases, and all-cause mortality. Additionally, a study conducted in Korea assessed the association between air pollutants and mortality. Kim et al.,28 reported in 2017 that PM10 level was significantly associated with all-cause mortality and cardiovascular disease depending on the season. Another study further postulated that the effects of air pollutants (e.g., SO2, Co), on mortality in the elderly population differed in high temperature.29 Furthermore, a study found an increased all-cause mortality incidence when highly dense PM10 level persisted.30 However, as far as we know, studies considering the effects of PM2.5 and ozone on elderly mortality using single and two pollutants model are scarce. Our study aimed to study the effects of PM2.5 and ozone on the health of community-dwelling elders in Seoul area using mortality data of the Korea National Statistical Office. Furthermore, additional analysis was conducted to distinguish between illnesses as a cause of mortality. Different models taking into account various lag periods were postulated to assess possible delayed effects.

METHODS

Study sample and data collection

The current study utilized the mortality data from Korean National Statistical Office to analyze the mortality data of those 60 years or older living in Seoul area between 2002 to 2012. Daily mortality incidents during the study period were collected and used as the outcome variable measure. Because more than 95% of the death certificates in Seoul are completed by medical doctors, these reports are assumed to be credible. There were 311,526 overall mortality incidents and 28,320 yearly mortality incidents during the study period. The overall as well as disease-specific mortality incidents are shown in Table 1.
Table 1

Summary statistics for air pollutants, meteorological factors and daily overall/disease-specific deaths with subgroups of disease in Seoul, 2002–2012

jkms-34-e311-i001
Variables Mean ± SD Min P25 P50 P75 Max
Daily deaths
Total 78 ± 11 42 70 77 85 124
Vascular diseasea 16 ± 5 4 13 16 19 37
Chronic respiratory disease 3 ± 1 1 1 2 3 10
Lung cancer 5 ± 2 1 3 5 7 17
Hypertension 2 ± 1 1 1 1 2 6
Diabetes 4 ± 2 1 3 4 5 15
Non-trauma 73 ± 11 41 65 72 80 115
Air pollutants
PM2.5, µg/m3 29 ± 18 3 17 25 36 366
Ozone, ppb 17 ± 1 2 10 16 23 63
Meteorological factors
Temperature, °C 12.7 ± 10.3 −14.5 4.0 14.3 21.9 31.8
Relative humidity, % 61.0 ± 15.1 19.9 49.6 61.1 71.9 96.5
Barometric pressure, hPa 1,005.7 ± 7.7 981.0 999.9 1,005.9 1,011.9 1,026.8
SD = standard deviation, Max = maximum, Min = minimum, P = percentile; PM = particulate matter.
aCardiovascular disease and coronary heart disease.
The number of daily overall as well as disease-specific mortality incidents were analyzed. In terms of the disease-specific mortality, illnesses such as vascular disease, chronic respiratory disease, lung cancer, diabetes and hypertension were classified according to the Charlson Comorbidity Index (CCI) based on the International Classification of Diseases 10th revision (ICD-10). CCI is widely used as a way to control for comorbidities. It was included in our analysis, because the study sample, which was the elderly population, tends to have higher number of comorbidities compared with other age groups. Moreover, the Quan algorithm used in this study showed high predictability in terms of mortality incidence of hospital in-patients.31 The usefulness of the Quan algorithm was further demonstrated in a study of Korean patients suffering from myocardial infarction. In this study, Charlson comorbidity ICD-10 conversion algorithm, observation period, as well as the predictive power according to the range of data and diagnosis of the disease were most accurate when the Quan algorithm was used.32 In regards to vascular disease, myocardial infarction (I21.x, I22.x, and 125.2), congestive heart failure (I09.9, I11.0, I13.0, I13.2, I25.5, I42.0, I42.5-I42.9, I43.x, I50.x, and P29.0), peripheral vascular disease (I70.x, I71.x, I73.1, I73.8, I73.9, 177.1, 179.0, I179.2, K55.1, K55.8, K55.9, Z95.8, and Z95.9), and cerebrovascular disease (G45.x, G46.x, H34.0, and I60.x-I69.x) were included. Also, (127.8, 127.9, J40.x-J47.x, J60.x-J67.x, J68.4, J70.1, and J70.3) for chronic respiratory disease, (C30.x–C39.x) for lung cancer, and in the case of diabetes, both diabetes without other complications and with other complications (E10.x–E14.x) were included.33 Lastly, although hypertension is not included in CCI, because it is a highly prevalent disease in the Korean elderly population,34 it was included in the study with the classification of I10.x.

Exposure variables

Both PM2.5 and ozone level were measured hourly. The amount of PM was measured using ‘Beta-ray absorption method’, which measures the amount of β -ray absorbed by a certain substance, considering the characteristic of the β-ray being absorbed more in bigger substances. To measure the ozone level, an ultraviolet ray with the wavelengths of 253.65 nm, which is most easily absorbed by the ozone, was shot at the ozone. The amount of the β-ray, which was absorbed by the ozone and decreased, was measured to calculate the ozone level. The amount of PM2.5 from 2002 to 2012 was acquired through the Climate Environment Division of Seoul, and the ozone level as well as other weather factors were acquired through the Ministry of Environment. The daily average of PM2.5 and ozone measurements were analyzed with meteorological factors (e.g., daily average of temperature, atmospheric pressure, and humidity) to account for other weather variation.

Statistical analysis

The number of disease-specific daily mortality incidents and 2 air pollutants (e.g., PM2.5 and ozone) were recorded according to date, and analyzed by the time series method (Table 2). More detailed information regarding the statistical model is described in a previous study.35 In this study, a single exposure model (i.e., air pollutants were treated as single air pollutant) and a multiple exposure model (i.e., each air pollutant was adjusted for the other air pollutant) were used. The purpose of using these models was to ascertain whether or not the association between an air pollutant and mortality remained significant even after adjusting for the other air pollutant. For example, in a single exposure model in terms of the PM only included the effect of the PM2.5, but the multiple exposure model was able to account for the effects of both the PM2.5 and ozone. Air pollutants such as sulfur dioxide or nitrogen dioxide were not included in the current study, because according a previous study,35 adjusting for these factors decreased the statistical power in regards to PM2.5, and a strong correlation between the main exposure variable (e.g., PM2.5) and sulfur dioxide or nitrogen dioxide.
Table 2

Estimated relative risk (95% intervals) in daily disease-specific mortality per 10 μg/m3 increase in PM2.5 and 10 ppb increase in ozone concentration in Seoul

jkms-34-e311-i002
Models Total Vascular disease Chronic pulmonary disease Lung cancer Hypertension Diabetes
PM2.5
Model 1 1.0062 (1.0053–1.0071) 1.0081 (1.0036–1.0125) 1.0086 (0.9991–1.0182)a 1.0027 (0.9945–1.0110) 0.9926 (0.9787–1.0065) 1.0075 (0.9988–1.0162)a
Model 2 1.0039 (1.0029–1.0048) 1.0053 (1.0005–1.0101) 1.0047 (0.9945–1.0150) 1.0003 (0.9914–1.0092) 0.9970 (0.9824–1.0117) 1.0073 (0.9980–1.0166)
Model 3 1.0037 (1.0027–1.0046) 1.0049 (1.0001–1.0098) 1.0022 (0.9917–1.0128) 1.0014 (0.9923–1.0105) 0.9997 (0.9848–1.0147) 1.0058 (0.9962–1.0153)
Ozone
Model 1 1.0029 (1.0006–1.0051) 0.9979 (0.9860–1.0098) 1.0156 (0.9897–1.0415) 0.9965 (0.9769–1.0161) 0.9980 (0.9677–1.0283) 1.0304 (1.0078–1.0529)
Model 2 1.0038 (1.0013–1.0063) 1.0041 (0.9908–1.0175) 1.0156 (0.9866–1.0447) 0.9930 (0.9708–1.0152) 0.9772 (0.9434–1.0110) 1.0210 (0.9956–1.0464)
Model 4 1.0027 (1.0002–1.0052) 1.0018 (0.9884–1.0153) 1.0143 (0.9850–1.0435) 0.9921 (0.9696–1.0145) 0.9790 (0.9447–1.0133) 1.0189 (0.9934–1.0444)
Model 1 was a univariate model. Model 2 was generated using over-dispersed generalized linear models and polynomial distributed lag model for cumulative exposures over the same day and lag days, adjusted for calendar day (natural smooth function with 4 × 12 df), day of the week, temperature (lag 0, natural smooth function, 3 df), pressure (lag = 0), and humidity (lag 0). Model 3 was generated using multivariate model and ozone (lag 0, natural smooth function, 3 df). Model 4 was generated using multivariate model and PM2.5 (lag 0, natural smooth function, 3 df).
PM = particulate matter, df = degrees of freedom.
Bolds are P value less than 0.05.
aP value was less than 0.1.
To analyze the different seasonal effects of air pollution on mortality, study periods were divided into warm periods and the rest into cold periods. Based on average sunshine hours, we considered May to August as the warm period and the rest as the cold period. The association between air pollutants and daily overall mortality differed depending on the season. Period-specific mortality rate was described with disease-specific mortality rate (Fig. 1 and Table 3).
Fig. 1

Mean PM2.5, ozone and daily deaths with regression line.

PM = particulate matter.
jkms-34-e311-g001
Table 3

Estimated relative risk (95% intervals) in daily disease-, period-specific mortality per 10 μg/m3 increase in PM2.5 and 10 ppb increase in ozone concentration in Seoul

jkms-34-e311-i003
Models Total Vascular disease Chronic pulmonary disease Lung cancer Hypertension Diabetes
PM2.5 (cold period)
Model 1 1.0063 (1.0053–1.0073) 1.0082 (1.0032–1.0132) 1.0057 (0.9948–1.0165) 1.0037 (0.9943–1.0131) 0.9871 (0.9718–1.0025)a 1.0066 (0.9966–1.0165)
Model 2 1.0047 (1.0036–1.0058) 1.0063 (1.0008–1.0118) 1.0012 (0.9893–1.0131) 1.0023 (0.9920–1.0126) 0.9958 (0.9794–1.0122) 1.0084 (0.9976–1.0191)
Model 3 1.0046 (1.0035–1.0057) 1.0061 (1.0005–1.0116) 0.9988 (0.9866–1.0110) 1.0027 (0.9923–1.0131) 0.9977 (0.9811–1.0142) 1.0077 (0.9967–1.0186)
PM2.5 (warm period)
Model 1 1.0077 (1.0056–1.0097) 1.0060 (0.9957–1.0163) 1.0213 (0.9996–1.0430)a 1.0020 (0.9841–1.0198) 1.0191 (0.9850–1.0531) 1.0183 (0.9986–1.0379)a
Model 2 1.0025 (1.0004–1.0045) 1.0001 (0.9893–1.0110) 1.0132 (0.9901–1.0362) 0.9961 (0.9772–1.0150) 1.0109 (0.9755–1.0464) 1.0065 (0.9860–1.0269)
Model 3 1.0016 (0.9993–1.0038) 0.9997 (0.9881–1.0113) 1.0106 (0.9861–1.0350) 0.9948 (0.9744–1.0153) 1.0263 (0.9865–1.0661) 0.9986 (0.9768–1.0204)
Ozone (cold period)
Model 1 0.9953 (0.9920–0.9987) 0.9927 (0.9752–1.0103) 0.9890 (0.9497–1.0284) 0.9962 (0.9661–1.0263) 1.0164 (0.9715–1.0612) 1.0137 (0.9802–1.0472)
Model 2 1.0020 (0.9982–1.0058) 1.0071 (0.9869–1.0273) 0.9998 (0.9544–1.0451) 0.9962 (0.9616–1.0307) 0.9858 (0.9347–1.0370) 1.0027 (0.9639–1.0416)
Model 4 1.0039 (1.0000–1.0078) 1.0092 (0.9887–1.0296) 0.9993 (0.9530–1.0456) 0.9927 (0.9576–1.0278) 0.9945 (0.9425–1.0465) 1.0087 (0.9691–1.0483)
Ozone (warm period)
Model 1 1.0118 (1.0086–1.0151) 1.0027 (0.9852–1.0202) 1.0316 (0.9950–1.0684)a 1.0057 (0.9777–1.0337) 0.9811 (0.9346–1.0276) 1.0596 (1.0274–1.0917)
Model 2 1.0046 (1.0010–1.0083) 1.0018 (0.9815–1.0221) 1.0178 (0.9750–1.0604) 1.0001 (0.9672–1.0329) 0.9703 (0.9171–1.0236) 1.0422 (1.0049–1.0795)
Model 4 1.0023 (0.9983–1.0063) 0.9956 (0.9739–1.0173) 1.0181 (0.9722–1.0640) 1.0059 (0.9699–1.0418) 0.9478 (0.8869–1.0086)a 1.0378 (0.9977–1.0780)a
Model 1 was a univariate model. Model 2 was generated using over-dispersed generalized linear models and polynomial distributed lag model for cumulative exposures over the same day and lag days, adjusted for calendar day (natural smooth function with 4 × 12 df), day of the week, temperature (lag 0, natural smooth function, 3 df), pressure (lag = 0), and humidity (lag 0). Model 3 was generated using multivariate model and ozone (lag 0, natural smooth function, 3 df). Model 4 was generated using multivariate model and PM2.5 (lag 0, natural smooth function, 3 df).
PM = particulate matter, df = degrees of freedom.
Bolds are P value less than 0.05.
aP value was less than 0.1.
Because the overall mortality rate was close to a Poisson distribution, and the explanatory variables of mortality were mostly nonlinear, the Generalized Linear Model (GLM) was used. In this model, the main exposure variable was the daily average of each air pollutant level. GLM combines the time series analysis, Poisson distribution and natural spline to calculate the short-term relationship between PM2.5, ozone and disease-specific mortality. The Partial autocorrelation function was used to induce the optimal degrees of freedom (df) in the natural spline of the calendar time, which was 4 per year according to Peng and Dominici.36 Therefore, the season and long-term trend was controlled for by the natural 3rd smooth spline with the calendar time (df = 4 × 12). Considering previous studies,3738 temperature and other pollutants (df = 3) at the time of mortality incident were integrated in to the natural 3rd smooth spline model. Day of the week (DOW), which was a categorical variable, was used as a dummy variable to control for all DOW related variables. The final model is shown below.
log[E(Yt-i)] = α + β1PM2.5t-i or β1O3t-i + ns (Temp0, df = 3) + Humi + Press + ns (PM2.50 or O30, df = 3) + ns (Time, df = 4 × 12) + β2DOW
E(Yt-i) is the expected number of deaths; PM2.5t-i or O3t-i are the average amount of PM2.5 and ozone level in Seoul with i being the lag day. For instance, when lag = 2, the average amount of PM2.5 or the Ozone level is included in the analysis; ns is the natural 3rd spline; Temp0 is the daily average temperature (°C), Humi is the daily relative humidity (%), and Press is the daily relative atmospheric pressure (hPa); PM2.50 and O30 are the densities of PM2.5 and ozone; Time indicated the long-term trend as well as seasonal effects with calendar days.
Considering the lag model, in which the resulting effect is delayed compared to the exposure, the model consists a delay (lag). For example, lag1 indicates an exposure a day prior to the resulting incident. Different lag effects were observed in prior studies. In terms of ischemic stroke, a latency between 0 and 1 was apparent (i.e., the day of as well as a day prior to the exposure),39 for chronic obstructive pulmonary disease, a latency of 0–3 (i.e., the day of, as well as one day, two days and three days prior to the exposure),40 and in terms of cardiovascular mortality and morbidity, a latency of 1–5 days was observed.41 With a careful consideration of prior results, the current study analyzed a latency up to 5 days making a single lag model (lag0–lag5). Disease-specific mortality rate was described according to the day of the exposure (Table 4).
Table 4

Estimated relative risk (95% interval) in daily mortality per 10 μg/m3 increase in PM2.5 and 10 ppb increase in ozone concentrations in Seoul, according to lag days

jkms-34-e311-i004
Models Lag 0 Lag 1 Lag 2 Lag 3 Lag 4 Lag 5
PM2.5
Model 1 1.0062 (1.0053–1.0071) 1.0060 (1.0051–1.0069) 1.0058 (1.0049–1.0067) 1.0056 (1.0047–1.0065) 1.0054 (1.0045–1.0063) 1.0052 (1.0043–1.0062)
Model 2 1.0036 (1.0026–1.0046) 1.0034 (1.0025–1.0044) 1.0033 (1.0023–1.0042) 1.0031 (1.0022–1.0041) 1.0030 (1.0020–1.0039) 1.0029 (1.0019–1.0038)
Model 3 1.0035 (1.0025–1.0045) 1.0033 (1.0024–1.0043) 1.0032 (1.0022–1.0041) 1.0030 (1.0020–1.0040) 1.0029 (1.0019–1.0038) 1.0028 (1.0018–1.0037)
Ozone
Model 1 1.0029 (1.0006–1.0051) 1.0018 (0.9996–1.0041) 1.0008 (0.9986–1.0031) 0.9998 (0.9976–1.0021) 0.9988 (0.9966–1.0011) 0.9979 (0.9956–1.0001)a
Model 2 1.0035 (1.0010–1.0061) 1.0027 (1.0003–1.0052) 1.0020 (0.9995–1.0044) 1.0012 (0.99988–1.0037) 1.0006 (0.9982–1.0030) 0.9999 (0.9976–1.0023)
Model 4 1.0025 (1.0000–1.0051) 1.0017 (0.9992–1.0042) 1.0010 (0.9985–1.0034) 1.0003 (0.9979–1.0027) 0.9996 (0.9972–1.0020) 0.9990 (0.9966–1.0014)
Model 1 was a univariate model. Model 2 was generated using over-dispersed generalized linear models and polynomial distributed lag model for cumulative exposures over the same day and lag days, adjusted for calendar day (natural smooth function with 4 × 12 df), day of the week, temperature (lag 0, natural smooth function, 3 df), pressure (lag = 0), and humidity (lag 0). Model 3 was generated using multivariate model and ozone (lag 0, natural smooth function, 3 df). Model 4 was generated using multivariate model and PM2.5 (lag 0, natural smooth function, 3 df).
PM = particulate matter, df = degrees of freedom.
Bolds are P value less than 0.05.
aP value was less than 0.1.
The results of the statistical analysis in the current study were based on the relative risk of daily mortality due to 10 µg/m3 increase in the PM density and 10 ppb increase in the ozone density. A P value below 0.05 was considered statistically significant, and all analyses were conducted using R statistical software, version 3.3.3 (R Foundation for Statistical Computing, Vienna, Austria).

Ethics statement

Because the data in this study were anonymized and impersonalized, we did not obtain individual informed consent from the subjects. The study was approved by Ajou University Hospital for Institutional Review Board approval (approval No. AJIRB-SBR-EXP-19-327).

RESULTS

Between 2002 and 2012, PM2.5 level had a mean value of 29 µg/m3 (standard deviation [SD], 18 µg/m3; MAX, 366 µg/m3), and ozone level had a mean value of 17 ppb (SD, 1 ppb; MAX, 63 ppb). In the same period the average for temperature was 12.7°C, humidity 61% and air pressure was 1,005.7 hPa. There were an average of 78 (SD, 11) overall daily mortality, 16 (SD, 5) mortality due to vascular disease, 3 (SD, 1) due to chronic respiratory disease, 5 (SD, 2) due to lung cancer, 2 (SD, 1) due to hypertension, and 4 (SD, 2) due to diabetes (Table 1).
The association between air pollutants and daily overall mortality differed depending on the season. Considering the warm-period and the cold-period, increased mortality risk in association with increased PM2.5 was apparent in the cold-period (R2 = 3.04e–05; P = 0.230), but not in the warm-period (R2 = 0.0014; P = 0.001). Contrastingly, the association between ozone and mortality was apparent in the warm-period (R2 = 0.0242; P = 5.5e–55), but not in the cold-period (R2 = 0.000228; P = 0.041) (Fig. 1).
In regards to daily mortality, as the level of particulate level increased by 10 µg/m3, the overall mortality rate in the multiple exposure model, adjusting for ozone level, increased by 1.0037 (95% confidence interval [CI], 1.0027–1.0046; P < 0.05), and mortality due to vascular disease by 1.0049 (95% CI, 1.0001–1.0098; P < 0.05). In the multiple exposure model, adjusting for PM2.5, ozone was associated with 1.0027 (95% CI, 1.0002–1.0052; P < 0.05) fold higher overall mortality rate and, 1.0304 (95% CI, 1.0078–1.0529; P < 0.05) fold higher risk of mortality due to diabetes in the single exposure model (Table 2). Considering the warm and cold period separately, in the multiple exposure model, with ozone as the main exposure factor, increase in the overall mortality rate 1.0046 (95% CI, 1.0010–1.0083; P < 0.05) and mortality due to diabetes 1.0422 (95% CI, 1.0049–1.0795; P < 0.05) were significant in the warm-period. Moreover, in the multiple exposure model, adjusting for PM2.5, the overall mortality rate increased by 1.0039 (95% CI, 1.0000–1.0078; P < 0.05) due to ozone in the cold-period. In association with PM2.5, overall mortality increased by 1.0046 fold and mortality due to vascular disease by 1.0061 fold in the multiple exposure model adjusting for ozone during the cold-period. Also, during the warm-period, the overall mortality increased by 1.0025 fold (95% CI, 1.0004–1.0045; P < 0.05) (Table 3).
The relative risk with the lag effect of mortality due to increase in ozone level is shown in Table 4 and Fig. 2. The relative risk of mortality due to increase in PM2.5 level in the single exposure model was 1.0062 (95% CI, 1.0053–1.0071; P < 0.05) on the day of death, and 1.0060 (95% CI, 1.0051–1.0069; P < 0.05) one day after death. The results of the lag effect are as follows: lag 0 (day of death): 1.0035, 95% CI, 1.0025–1.0145; lag 1: 1.0033, 95% CI, 1.0024–1.0043; lag 2: 1.0032, 95% CI, 1.0022–1.0041; lag 3: 1.0030, 95% CI, 1.0020–1.0040; lag 4: 1.0029, 95% CI, 1.0019–1.0038; lag 5: 1.0028, 95% CI, 1.0018–1.0037; all P < 0.05). In the model with ozone as the exposure factor, adjusting for meteorological factors, the lag effect remained significant until one day after death, and was smaller compared to the model considering PM as the main exposure factor: lag 0 (day of death): 1.0035, 95% CI, 1.0010–1.0061; lag 1: 1.0027, 95% CI, 1.0003–1.0052; all P < 0.05). Lastly, in the multiple exposure model of ozone including PM, the result was only significant in lag 0 (day of death): 1.0025, 95% CI, 1.0000–1.0051; P < 0.05).
Fig. 2

Estimated relative risk (95% CI) in daily mortality per 10 μg/m3 increase in PM2.5 and 10 ppb increase in ozone concentrations in Seoul, according to lag days (0–5). Estimates were generated using over-dispersed generalized linear models and polynomial distributed lag model for cumulative exposures over the same day and lag days, adjusted for calendar day (natural smooth function with 4 × 12 df), day of the week, temperature (lag 0, natural smooth function, 3 df), pressure (lag = 0), humidity (lag 0) and month.

CI = confidence interval, PM = particulate matter, df = degrees of freedom.
jkms-34-e311-g002

DISCUSSION

In this study, we analyzed the short-term effects of PM2.5 on overall mortality as well as mortality due to vascular disease in the elderly. In terms of ozone, the overall mortality and mortality due to diabetes were effected. Furthermore, we observed a lag effect of PM2.5 and ozone between the day of and a few days prior to death on overall mortality. The increased vulnerability of the elderly against air pollutants was shown in a previous study, in which the mortality rate due to air pollutants was 2-3 fold higher for those 85 or older compared to those 65 and younger.18 Furthermore, in terms of the vulnerability to PM in the elderly could be explained through three mechanisms: 1) a reduced physical reserve due to the body's response to counter-act the harmful effects of PM; 2) increased inflammation due to PM; and 3) weakened health due to sarcopenia.42
The association between mortality, PM2.5 and ozone in the elderly population was found in previous studies. Di et al.27 found an increase in the mortality rate in accordance with an increase in the PM2.5 density, but in terms of ozone density, the association was only significant during the warm period. Also, Fann et al.9 found that a yearly increase of PM2.5 and an average increase in ozone density in 8 hours during the summer season were associated with 1.1 million, and 36,000 life years lost. Our results were somewhat consistent with previous findings and indicated that PM2.5 was associated with increased overall mortality in all situations, but ozone was only significantly associated with overall mortality rate during the warm period. The cause of these results is supported by previous studies proposing that photochemical formation is weaker during the cold period, and pulverized ozone density consumes the reaction between ozone and nitrogen oxides. Therefore, ozone density is negatively correlated with nitrogen oxide density, and nitrogen dioxide is positively correlated with disease related health status, causing a difference in mortality rates due to ozone density between cold and warm periods.434445464748495051
In this study, increase in the PM2.5 significantly increased the mortality rate due to vascular disease, but the effect of ozone was not ascertained. Although a number of studies have found that PM10 and PM2.5 increases the risk of cerebrovascular disease, the cause of it has not been clearly identified. However, it is possible that the PM10 or PM2.5 triggers oxidative stress of vascular endothelial cells, reduces angiogenesis, and increases vascular inflammation, which then can lead to cerebrovascular disease. Consequently, inadequate microvascular function caused by the PM2.5 could be the underlying mechanism between the association between PM2.5 and cerebrovascular disease.525354 In a study by Halonen et al.,41 hospitalization due to PM2.5 of the elderly, suffering from either asthma or COPD, increased significantly, but the association between cardiovascular health and PM2.5 was not very strongly supported. In contrast, Ren et al.55 found that 10 ppb increase in ozone density was associated with 6.5% increase in mortality due to stroke. However, because the participants included in the study were people 35 years or older, a further study only considering elderly population could show different results.
Unlike previous studies showing associations between lung cancer,5657 hypertension and air pollution,2658 our study did not find any significant association between these factors. This difference could be accounted for by the possibility that the number of mortality incidents due to lung cancer was relatively small, limiting the statistical power. In regards to hypertension, it is noteworthy to consider that the diagnosis of a disease for the elderly, suffering from multiple diseases, is often made for convenience depending on the medical prescription, subsequently not correctly reflecting the correct incident rate.59
The association between ozone density and mortality due to diabetes during the warm period was significant, but was much weaker in regards to PM2.5. In a previous study looking at the association between air pollutants and the markers of insulin resistance in Korean population, PM10 and ozone were both associated with an increase in insulin resistance of the elderly population. Additionally, certain genotype was shown to increase the effects of air pollutants on insulin resistance.15 In a study conducted in Italy, increase in PM10, PM2.5 and ozone density was associated with diabetes incidents, and in case on ozone density, the association remained significant even after controlling for other sources of pollutants.60 In our study, the small number of mortality due to diabetes could have been the cause that limited its association with PM2.5, but the fact that the effect of ozone density remained significant in the exposure model could suggest that ozone acts as an independent risk factor in terms of mortality due to diabetes.
PM2.5 and ozone on the day of as well as a few days prior to death was associated with increased mortality in our results. This association indicates that the two air pollutants increase mortality incidents and that there is a lag effect in the relationship. The lag effect included in our study controls for seasonal overdispersion. Also, a multiple exposure model, including confounding variables often associated with air pollution and mortality, as well as a single exposure model, only including the main exposure variables, were used to estimate the lag effect. In the past, epidemiological studies have reported contrasting results. A study published by Costa et al.61 in 2017 reported that PM10 on the day of and a day prior to death was associated with increased non-accidental mortality incidents as well as mortality due to cerebrovascular disease, and chronic lung disease. Furthermore, an accumulation of PM10 during 10 days prior to death significantly increased the mortality incidents due to cerebrovascular disease and chronic lung disease. Also, while a study from Japan published in 2013 found that a maximum 8 hours of 10 ppb increase in ozone was associated with 0.69% increase in mortality incidence,62 studies conducted in Canada and Finland did not find any lag effects in terms of the effect of air pollutant on the association between heart failure and cardiopulmonary disease with mortality.6364 Accordingly, each air pollutant influenced mortality incidents differently in our study. Unlike the PM2.5, which showed 5 days of lag effect, the effects of ozone density were much smaller, only showing significant effects on the day of and a day prior to the death. Because we used daily averages to determine the mortality of ozone in this paper, the results might be underestimated compared to those that result in 1 hour or 8 hours maximum. However, the 5 days lag effect was apparent even in the multiple exposure model, controlling for ozone density, the amount of PM2.5 exposure on the day of as well as several days before death needs to be considered.
There are several limitations of this study. First, because the study only included those 60 years and older, if age is a confounding factor in the relationship between air pollutants and medical use, it is difficult to generalize the results to the younger generation. Second, important factors such as influenza epidemic, pollen, and index of holidays were not accounted for, it is possible that confounders were not controlled completely.
Despite these limitations, this study does have various strengths. First, the current study used the population representative data of the Korean National Statistical Office to derive a significant association between PM2.5, ozone and mortality. Second, because the results included overall mortality as well as mortality due to highly prevalent illnesses in the elderly population, the association of interest was found in terms of more various conditions.
The results of the current study provide meaningful findings. Although both PM2.5 and ozone were significantly associated with overall mortality, the lag effect was longer in terms of the PM2.5. Also, because the results differed according to warm and cold periods, it could be assumed that the period season changes the effects of the PM2.5 on mortality. In future studies, the research should be expanded to wider age cohorts, and how different air pollutants influence medical use should be considered in more various types of air pollutants.

Notes

Disclosure The authors have no potential conflicts of interest to disclose.

Author Contributions

  • Conceptualization: Jung EJ, Jang JY.

  • Data curation: Jung EJ, Lee KE, Na W.

  • Methodology: Jung EJ.

  • Validation: Jung EJ, Lee KE.

  • Writing - original draft: Jung EJ.

  • Writing - review & editing: Jung EJ.

References

1. Ensor KB, Raun LH, Persse D. A case-crossover analysis of out-of-hospital cardiac arrest and air pollution. Circulation. 2013; 127(11):1192–1199.
2. Hao Y, Balluz L, Strosnider H, Wen XJ, Li C, Qualters JR. Ozone, fine particulate matter, and chronic lower respiratory disease mortality in the United States. Am J Respir Crit Care Med. 2015; 192(3):337–341.
3. Linn WS, Szlachcic Y, Gong H Jr, Kinney PL, Berhane KT. Air pollution and daily hospital admissions in metropolitan Los Angeles. Environ Health Perspect. 2000; 108(5):427–434.
4. Shields KN, Cavallari JM, Hunt MJ, Lazo M, Molina M, Molina L, et al. Traffic-related air pollution exposures and changes in heart rate variability in Mexico City: a panel study. Environ Health. 2013; 12:7.
5. Tallon LA, Manjourides J, Pun VC, Mittleman MA, Kioumourtzoglou MA, Coull B, et al. Erectile dysfunction and exposure to ambient air pollution in a nationally representative cohort of older men. Environ Health. 2017; 16(1):12.
6. To T, Zhu J, Larsen K, Simatovic J, Feldman L, Ryckman K, et al. Progression from asthma to chronic obstructive pulmonary disease. Is air pollution a risk factor? Am J Respir Crit Care Med. 2016; 194(4):429–438.
7. Yamazaki S, Shima M, Yoda Y, Oka K, Kurosaka F, Shimizu S, et al. Association between PM2.5 and primary care visits due to asthma attack in Japan: relation to Beijing's air pollution episode in January 2013. Environ Health Prev Med. 2014; 19(2):172–176.
8. Téllez-Rojo MM, Romieu I, Ruiz-Velasco S, Lezana MA, Hernández-Avila MM. Daily respiratory mortality and PM10 pollution in Mexico City: importance of considering place of death. Eur Respir J. 2000; 16(3):391–396.
9. Fann N, Lamson AD, Anenberg SC, Wesson K, Risley D, Hubbell BJ. Estimating the national public health burden associated with exposure to ambient PM2.5 and ozone. Risk Anal. 2012; 32(1):81–95.
10. Koken PJ, Piver WT, Ye F, Elixhauser A, Olsen LM, Portier CJ. Temperature, air pollution, and hospitalization for cardiovascular diseases among elderly people in Denver. Environ Health Perspect. 2003; 111(10):1312–1317.
11. Morgan G, Corbett S, Wlodarczyk J. Air pollution and hospital admissions in Sydney, Australia, 1990 to 1994. Am J Public Health. 1998; 88(12):1761–1766.
12. Orellano P, Quaranta N, Reynoso J, Balbi B, Vasquez J. Effect of outdoor air pollution on asthma exacerbations in children and adults: systematic review and multilevel meta-analysis. PLoS One. 2017; 12(3):e0174050.
13. Schwartz J, Litonjua A, Suh H, Verrier M, Zanobetti A, Syring M, et al. Traffic related pollution and heart rate variability in a panel of elderly subjects. Thorax. 2005; 60(6):455–461.
14. Wheida A, Nasser A, El Nazer M, Borbon A, Abo El Ata GA, Abdel Wahab M, et al. Tackling the mortality from long-term exposure to outdoor air pollution in megacities: lessons from the Greater Cairo case study. Environ Res. 2018; 160:223–231.
15. Kim JH, Hong YC. GSTM1, GSTT1, and GSTP1 polymorphisms and associations between air pollutants and markers of insulin resistance in elderly Koreans. Environ Health Perspect. 2012; 120(10):1378–1384.
16. Lim YH, Kim H, Kim JH, Bae S, Park HY, Hong YC. Air pollution and symptoms of depression in elderly adults. Environ Health Perspect. 2012; 120(7):1023–1028.
17. Hong YC, Lee JT, Kim H, Ha EH, Schwartz J, Christiani DC. Effects of air pollutants on acute stroke mortality. Environ Health Perspect. 2002; 110(2):187–191.
18. Cakmak S, Dales RE, Vidal CB. Air pollution and mortality in Chile: susceptibility among the elderly. Environ Health Perspect. 2007; 115(4):524–527.
19. Carlsen HK, Forsberg B, Meister K, Gíslason T, Oudin A. Ozone is associated with cardiopulmonary and stroke emergency hospital visits in Reykjavík, Iceland 2003–2009. Environ Health. 2013; 12:28.
20. Goldberg MS, Burnett RT, Brook J, Bailar JC 3rd, Valois MF, Vincent R. Associations between daily cause-specific mortality and concentrations of ground-level ozone in Montreal, Quebec. Am J Epidemiol. 2001; 154(9):817–826.
21. Kan H, London SJ, Chen G, Zhang Y, Song G, Zhao N, et al. Season, sex, age, and education as modifiers of the effects of outdoor air pollution on daily mortality in Shanghai, China: the Public Health and Air Pollution in Asia (PAPA) study. Environ Health Perspect. 2008; 116(9):1183–1188.
22. Martins LC, Pereira LA, Lin CA, Santos UP, Prioli G, Luiz Odo C, et al. The effects of air pollution on cardiovascular diseases: lag structures. Rev Saude Publica. 2006; 40(4):677–683.
23. Szyszkowicz M. Ambient air pollution and daily emergency department visits for ischemic stroke in Edmonton, Canada. Int J Occup Med Environ Health. 2008; 21(4):295–300.
24. Wong CM, Atkinson RW, Anderson HR, Hedley AJ, Ma S, Chau PY, et al. A tale of two cities: effects of air pollution on hospital admissions in Hong Kong and London compared. Environ Health Perspect. 2002; 110(1):67–77.
25. Fischer P, Hoek G, Brunekreef B, Verhoeff A, van Wijnen J. Air pollution and mortality in The Netherlands: are the elderly more at risk? Eur Respir J Suppl. 2003; 40:34s–38s.
26. Goldberg MS, Burnett RT, Stieb DM, Brophy JM, Daskalopoulou SS, Valois MF, et al. Associations between ambient air pollution and daily mortality among elderly persons in Montreal, Quebec. Sci Total Environ. 2013; 463-464:931–942.
27. Di Q, Dai L, Wang Y, Zanobetti A, Choirat C, Schwartz JD, et al. Association of short-term exposure to air pollution with mortality in older adults. JAMA. 2017; 318(24):2446–2456.
28. Kim SE, Honda Y, Hashizume M, Kan H, Lim YH, Lee H, et al. Seasonal analysis of the short-term effects of air pollution on daily mortality in Northeast Asia. Sci Total Environ. 2017; 576:850–857.
29. Park AK, Hong YC, Kim H. Effect of changes in season and temperature on mortality associated with air pollution in Seoul, Korea. J Epidemiol Community Health. 2011; 65(4):368–375.
30. Kim SE, Bell ML, Hashizume M, Honda Y, Kan H, Kim H. Associations between mortality and prolonged exposure to elevated particulate matter concentrations in East Asia. Environ Int. 2018; 110:88–94.
31. Sundararajan V, Quan H, Halfon P, Fushimi K, Luthi JC, Burnand B, et al. Cross-national comparative performance of three versions of the ICD-10 Charlson index. Med Care. 2007; 45(12):1210–1215.
32. Kim KH. Comparative study on three algorithms of the ICD-10 Charlson comorbidity index with myocardial infarction patients. J Prev Med Public Health. 2010; 43(1):42–49.
33. Quan H, Sundararajan V, Halfon P, Fong A, Burnand B, Luthi JC, et al. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 2005; 43(11):1130–1139.
34. Kim NR, Kim HC. Prevalence and trends of isolated systolic hypertension among Korean adults: the Korea National Health and Nutrition Examination Survey, 1998–2012. Korean Circ J. 2015; 45(6):492–499.
35. Jung EJ, Na W, Lee KE, Jang JY. A study on the influence on medical care for the elderly by exposure to fine particulate matter and ozone. J Environ Health Sci. 2019; 45(1):30–41.
36. Peng RD, Dominici F. Statistical Methods for Environmental Epidemiology with R: a Case Study in Air Pollution and Health. New York, NY: Springer;2008.
37. Xu M, Guo Y, Zhang Y, Westerdahl D, Mo Y, Liang F, et al. Spatiotemporal analysis of particulate air pollution and ischemic heart disease mortality in Beijing, China. Environ Health. 2014; 13:109.
38. Lim YH, Park AK, Kim H. Modifiers of diurnal temperature range and mortality association in six Korean cities. Int J Biometeorol. 2012; 56(1):33–42.
39. Ca M, Aj C, Ecol H, Access R. Pollution à l’ozone et survenue d’accidents vasculaires cérébraux à Dijon entre 1994 et 2004. Environ Risque Santé. 2007; 6:421–422.
40. Fung KY, Khan S, Krewski D, Chen Y. Association between air pollution and multiple respiratory hospitalizations among the elderly in Vancouver, Canada. Inhal Toxicol. 2006; 18(13):1005–1011.
41. Halonen JI, Lanki T, Tiittanen P, Niemi JV, Loh M, Pekkanen J. Ozone and cause-specific cardiorespiratory morbidity and mortality. J Epidemiol Community Health. 2010; 64(9):814–820.
42. Eckel SP, Louis TA, Chaves PH, Fried LP, Margolis AH. Modification of the association between ambient air pollution and lung function by frailty status among older adults in the Cardiovascular Health Study. Am J Epidemiol. 2012; 176(3):214–223.
43. Cai J, Chen R, Wang W, Xu X, Ha S, Kan H. Does ambient CO have protective effect for COPD patient? Environ Res. 2015; 136:21–26.
44. Cai J, Zhao A, Zhao J, Chen R, Wang W, Ha S, et al. Acute effects of air pollution on asthma hospitalization in Shanghai, China. Environ Pollut. 2014; 191:139–144.
45. Cao J, Li W, Tan J, Song W, Xu X, Jiang C, et al. Association of ambient air pollution with hospital outpatient and emergency room visits in Shanghai, China. Sci Total Environ. 2009; 407(21):5531–5536.
46. Chen R, Chu C, Tan J, Cao J, Song W, Xu X, et al. Ambient air pollution and hospital admission in Shanghai, China. J Hazard Mater. 2010; 181(1-3):234–240.
47. Hua J, Yin Y, Peng L, Du L, Geng F, Zhu L. Acute effects of black carbon and PM2.5 on children asthma admissions: a time-series study in a Chinese city. Sci Total Environ. 2014; 481:433–438.
48. Huang W, Tan J, Kan H, Zhao N, Song W, Song G, et al. Visibility, air quality and daily mortality in Shanghai, China. Sci Total Environ. 2009; 407(10):3295–3300.
49. Kan H, Chen B. A case-crossover analysis of air pollution and daily mortality in Shanghai. J Occup Health. 2003; 45(2):119–124.
50. Kan H, London SJ, Chen G, Zhang Y, Song G, Zhao N, et al. Differentiating the effects of fine and coarse particles on daily mortality in Shanghai, China. Environ Int. 2007; 33(3):376–384.
51. Zhao W, Cheng J, Li D, Duan Y, Wei H, Ji R, et al. Urban ambient air quality investigation and health risk assessment during haze and non–haze periods in Shanghai, China. Atmos Pollut Res. 2013; 4(3):275–281.
52. Brook RD, Bard RL, Morishita M, Dvonch JT, Wang L, Yang HY, et al. Hemodynamic, autonomic, and vascular effects of exposure to coarse particulate matter air pollution from a rural location. Environ Health Perspect. 2014; 122(6):624–630.
53. Chen S, Wu X, Hu J, Dai G, Rong A, Guo G. PM2.5 exposure decreases viability, migration and angiogenesis in human umbilical vein endothelial cells and human microvascular endothelial cells. Mol Med Rep. 2017; 16(3):2425–2430.
54. Zhang X, Staimer N, Tjoa T, Gillen DL, Schauer JJ, Shafer MM, et al. Associations between microvascular function and short-term exposure to traffic-related air pollution and particulate matter oxidative potential. Environ Health. 2016; 15(1):81.
55. Ren C, Melly S, Schwartz J. Modifiers of short-term effects of ozone on mortality in eastern Massachusetts--a case-crossover analysis at individual level. Environ Health. 2010; 9(1):3.
56. Moolgavkar SH. Air pollution and daily mortality in three U.S. counties. Environ Health Perspect. 2000; 108(8):777–784.
57. Kloog I, Ridgway B, Koutrakis P, Coull BA, Schwartz JD. Long- and short-term exposure to PM2.5 and mortality: using novel exposure models. Epidemiology. 2013; 24(4):555–561.
58. Lin H, Guo Y, Zheng Y, Di Q, Liu T, Xiao J, et al. Long-term effects of ambient PM2.5 on hypertension and blood pressure and attributable risk among older Chinese Adults. Hypertension. 2017; 69(5):806–812.
59. Moon OR, Kim CY, Kim MK. Individual variations in the code of the international classification of disease for similar outpatient conditions among general practitioners. Korean J Health Policy Adm. 1992; 2(1):66–79.
60. Orioli R, Cremona G, Ciancarella L, Solimini AG. Association between PM10, PM2.5, NO2, O3 and self-reported diabetes in Italy: a cross-sectional, ecological study. PLoS One. 2018; 13(1):e0191112.
61. Costa AF, Hoek G, Brunekreef B, Ponce de Leon AC. Air pollution and deaths among elderly residents of São Paulo, Brazil: an analysis of mortality displacement. Environ Health Perspect. 2017; 125(3):349–354.
62. Ng CF, Ueda K, Nitta H, Takeuchi A. Seasonal variation in the acute effects of ozone on premature mortality among elderly Japanese. Environ Monit Assess. 2013; 185(10):8767–8776.
63. Halonen JI, Lanki T, Yli-Tuomi T, Tiittanen P, Kulmala M, Pekkanen J. Particulate air pollution and acute cardiorespiratory hospital admissions and mortality among the elderly. Epidemiology. 2009; 20(1):143–153.
64. Vanasse A, Talbot D, Chebana F, Bélanger D, Blais C, Gamache P, et al. Effects of climate and fine particulate matter on hospitalizations and deaths for heart failure in elderly: a population-based cohort study. Environ Int. 2017; 106:257–266.
TOOLS
ORCID iDs

En-Joo Jung
https://orcid.org/0000-0002-0757-6032

Wonwoong Na
https://orcid.org/0000-0002-3128-0565

Kyung-Eun Lee
https://orcid.org/0000-0001-5112-7747

Jae-Yeon Jang
https://orcid.org/0000-0002-3913-8758

Similar articles