Open Access
Open Peer Review

This article has Open Peer Review reports available.

How does Open Peer Review work?

Associations of air pollution exposure with blood pressure and heart rate variability are modified by oxidative stress genes: A repeated-measures panel among elderly urban residents

Environmental Health201615:47

https://doi.org/10.1186/s12940-016-0130-3

Received: 4 November 2015

Accepted: 18 March 2016

Published: 25 March 2016

Abstract

Background

Oxidative stress has been suggested as a major cause of elevated blood pressure (BP) and reduced heart rate variability (HRV) due to air pollution. We hypothesized that the associations of air pollution exposure with BP and HRV are modified by oxidative stress gene polymorphisms.

Methods

Between 2008 and 2010, we conducted up to 5 surveys of 547 elderly participants, measured their BP and HRV, and genotyped 47 single nucleotide polymorphisms (SNPs) in 18 oxidative stress genes. Linear mixed models were constructed to evaluate the associations of particulate matter ≤10 μm, nitrogen dioxide, and sulfur dioxide with BP and HRV, as well as the modifications of these associations by the genotyped SNPs.

Results

Single-SNP analyses revealed interactions between air pollution and 15 SNPs (for BP) and 33 SNPs (for HRV) (all, P for interaction < 0.05). When we generated genetic risk scores for BP and HRV, using the SNPs with interactions in the single-SNP models, we found that associations of air pollution exposure with BP and HRV were modified by the genetic risk scores (P for interaction < 0.05).

Conclusions

These results strongly suggest that the associations of air pollution with BP and HRV are mediated by oxidative stress pathways.

Keywords

Air pollution Heart rate variability Blood pressure Oxidative stress Genetic factors

Background

Air pollution exposure is associated with increased cardiovascular morbidity and mortality [1], which are one of the most prominent health outcomes related to air pollution [2]. Previous studies have suggested that changes in blood pressure [3] and heart rate variability [4] could be induced by air pollution, and these changes may be responsible for the reported increase in cardiovascular morbidity and mortality due to air pollution [5]. Oxidative stress has been indicated as a major cause of the associations of air pollution exposure with increased blood pressure and decreased heart rate variability [6, 7]. Therefore, it is biologically plausible that the associations of air pollution with blood pressure and heart rate variability are modified by a genetic predisposition to oxidative stress [8].

Despite the importance of gene-environment interactions for identifying and assessing potential mechanistic pathways for health conditions [9, 10], few studies have evaluated the air pollution-gene interaction with regard to blood pressure or heart rate variability, and their results have been inconsistent [5, 1118]. Furthermore, these studies have been conducted mostly in a few populations of middle-aged or elderly Caucasians, which limits the external generalizability and does not exclude the possibility of random error or population stratification. In addition, previous studies have focused on individual single nucleotide polymorphisms (SNPs) or genes, lowering the ability to comprehensively assess the oxidative stress pathways with regard to blood pressure and heart rate variability, because single SNPs typically have small effects and may affect specific phenotypes cumulatively by working together with other SNPs.

Therefore, in the present study, we hypothesized that oxidative stress-related genetic polymorphisms could modify the associations of air pollution exposure with blood pressure and heart rate variability. We evaluated the hypothesis using a repeated-measures panel of elderly urban residents in Seoul, Korea, and used genetic risk scores instead of individual SNPs or genes to summarize the cumulative effects of genetic polymorphisms in the oxidative stress pathway as a potential mechanisms behind the associations of air pollution exposure with blood pressure and heart rate variability [19].

Methods

Study design and population

The Korean Elderly Environmental Panel (KEEP) study was a community-based repeated-measures study that aims to evaluate the relationship between environmental risk factors and adverse health outcomes in an elderly population. We recruited 560 non-institutionalized elderly individuals who regularly visited a community welfare center that is located in Seongbuk-Gu district, Seoul, Republic of Korea. The inclusion criteria were an age of ≥60 years and the ability to communicate and follow the instructions of the survey staff. Information regarding the participants’ sociodemographic characteristics, lifestyle, and medical history was obtained using a structured questionnaire, which was administered by trained interviewers at the baseline survey. Between August 2008 and August 2010, we conducted up to 5 surveys for each participant with measurement of blood pressure, heart rate variability, and anthropometric parameters.

Among the 560 participants, we subsequently excluded 7 individuals for having no blood pressure measurement and 6 individuals for having no heart rate variability information. Therefore, we included data from 547 participants in the final analyses. All participants provided their written informed consent, and the study protocol was approved by the ethical review board at Seoul National University Hospital (C-704-040-205).

Air pollutant concentrations and meteorological factors

We estimated the individuals’ exposures to air pollutants, such as particulate matter ≤10 μm (PM10), nitrogen dioxide (NO2), and sulfur dioxide (SO2), using 24-h monitoring data that were obtained from the Korea National Institute of Environmental Research, Incheon, Republic of Korea. We used the daily mean values for PM10, NO2, and SO2 from the monitoring center that was nearest to each participant’s residence. The mean distance between the monitoring centers and the participants’ residences was <1 km. Detailed information regarding the measurement methods has been described elsewhere [20]. Daily mean temperatures and dew point temperatures were measured at the monitoring center that was nearest to each participant’s residence, and these data were obtained from the Korea Meteorological Administration, Seoul, Republic of Korea. Next, we calculated the apparent temperature using the following formula [2124]: Apparent temperature = –2.653 + [0.994 × daily mean temperature (°C)] + (0.0153 × [daily dew point temperature (°C)]2).

Genetic polymorphisms

We used the QIAamp DNA Blood Mini Kit (Qiagen, Valencia, CA, USA) to extract genomic DNA from samples of the participants’ peripheral blood lymphocyte. We analyzed 18 genes (AhR, ANKK1, CAT, COMT, CYP1A1, CYP1B1, CYP2B6, EPHX1, GSTP1, HSPA1L, MPO, MTHFR, NAT2, NOS3, NQO1, PON1, PTGS2, and SOD2) that are related to oxidative stress, and genotyped 47 SNPs from these genes. The SNPs were selected based on a priori knowledge of being associated with oxidative stress status, a previous study from the National Center for Biotechnology Information, as well as minor allele frequencies of ≥5 % for the Japanese and Chines populations in the HapMap (http://hapmap.ncbi.nlm.nih.gov/) to consider public health implications. Polymorphisms in ANKK1, CAT, CYP1B1, EPHX1, HSPA1L, MPO, NOS3, PON1, PTGS2, and SOD2 were identified using a TaqMan fluorogenic 5’ nuclease assay (ABI, Foster City, CA, USA); polymorphisms in AhR, COMT, CYP1A1, CYP2B6, MTHFR, NAT2, and NQO1 were identified using the Sequenom Mass ARRAY platform [25]; and polymorphisms in GSTP1 were identified using a multiplex polymerase chain reaction method [26]. Among the 47 genotyped SNPs, we excluded one SNP (rs2965753) from the present analyses because it was not within the Hardy-Weinberg equilibrium (P = 9.3126 × 10-7 using the chi square test).

Blood pressure and heart rate variability

Blood pressure and heart rate variability were measured between 10:00 AM and 12:00 PM. After ≥10 min of rest, a trained medical technologist measured the participant’s blood pressure using an autonomic sphygmomanometer (HEM-780; Omron, Kyoto, Japan). Once the first measurement was finished, the participant was asked to sit and rest for another 10 min, and their blood pressure was subsequently re-measured. We averaged the two blood pressure measurements and used the mean value as a main outcome variable. We also calculated the mean arterial pressure by adding one-third of the systolic blood pressure and two-thirds of the diastolic blood pressure.

To measure heart rate variability, participants were asked to attach 3 limb leads to both wrists and their left ankle, and to relax in the seated position for ≥5 min. Heart rate variability was automatically analyzed via electrocardiography using a heart rate variability analyzing device (SA-3000P; Medicore, Seoul, Republic of Korea). We analyzed the standard deviation of normal-to-normal intervals (SDNN) and the root mean square of successive differences (RMSSD) for time domain measures, and low frequency power (0.04–0.15 Hz, LF) and high frequency power (0.15–0.40 Hz, HF) for frequency domain measures.

Statistical analysis

We log-transformed the variables that exhibited log-normal distributions (e.g., SDNN, RMSSD, LF, and HF) and used these values for our analyses. In the KEEP study, exposures and outcomes were repeatedly measured for each participant up to 5 times, and we constructed the long format data set, which stacks information obtained from each survey in the row. To consider intra-individual correlation due to repeated-measures data structure, linear mixed models were constructed. We evaluated the associations of exposure to air pollutants (PM10, NO2, and SO2) with blood pressure (systolic blood pressure, diastolic blood pressure, and mean arterial pressure) and heart rate variability (SDNN, RMSSD, LF, and HF) using these models. We also applied daily lag structures up to 3 days, and reported the results from the models with the best fit (which were determined using the Akaike information criterion). After we analyzed the minor allele frequency and Hardy-Weinberg equilibrium for each SNP, we added interaction terms for each air pollutant and the SNP to the main models, which contained lower order terms and covariates to assess the interaction. All SNPs were modeled as an additive model, which is known to provide good performance, even in cases where the true genetic model is not known [27, 28]. The general form of the linear mixed models used in this analysis is presented in the Additional file 1.

We calculated the genetic risk scores for blood pressure and for heart rate variability by summing the number of risk alleles for the SNPs that exhibited interactions (P value for interaction < 0.05) with any air pollutant with regard to blood pressure or heart rate variability. To satisfy the assumption that each SNP in the genetic risk score was independently associated with risk, we estimated the linkage disequilibrium between the SNPs within the same gene by calculating |D’| values and selected one SNP within each linkage disequilibrium block. Furthermore, we categorized the genetic risk scores for blood pressure and heart rate variability into tertiles. We then assessed the association between air pollution exposure and the outcome of interest within each tertile using linear mixed models, after we had performed nonparametric analyses using the generalized additive mixed models (see Additional file 1). Heterogeneity in the associations according to the genetic risk scores was assessed by adding and testing the product term of the tertile score and each air pollutant.

All models were adjusted for covariates that were selected a priori: age (years), sex, smoking status (current smoker, ex-smoker, non-smoker, did not answer), alcohol drinking (current drinker, non-drinker, did not answer), body mass index (kg/m2), hypertension medication (no, yes), and apparent temperature. All covariates except for body mass index and apparent temperature were included in the models as time-independent. Although demographic and lifestyle factors may not influence air pollution exposure, we included them to block any potential backdoor path [29]. All analyses were performed using SAS software (version 9.4; SAS Institute Inc., Cary, NC, USA), R software (version 3.1.0; Comprehensive R Archive Network: http://cran.r-project.org), and Haploview software (version 4.2; http://www.broadinstitute.org/haploview). All P-values were two-sided.

We performed several sensitivity analyses. First, we evaluated the interaction of each air pollutant with each SNP using the 2-df joint test, which simultaneously tests the main effect of the SNP and its interaction with the environmental factor [30]. To perform the 2-df joint test, we performed a likelihood ratio test using a full model that included terms for the SNP, air pollutant, their interaction, and any covariates, as well as a nested model that excluded the SNP and interaction terms, using the SAS macro MIXED_FIT and R package lme4. Second, we determined the genetic risk scores using different set of SNPs, which were selected from each linkage disequilibrium block, and assessed whether the results were robust. Third, we weighted the follow-up observations using the inverse probability of having a follow-up response, in order to reduce the potential selection bias that is caused by non-random loss to follow-up [31]. Logistic regression was performed to estimate the probability of a follow-up, and the covariates included age, sex, body mass index, years of schooling, blood pressure, season, and outdoor temperature at the prior visit. We gave a weight of 1 to the first observation and the inverse probability of follow-up to each follow-up observation [32].

Results

Table 1 shows the participants’ baseline characteristics. All participants were ≥60 years old, and the mean age was 70.7 years. Among the 547 individuals, 404 (73.9 %) were women, 467 (85.4 %) were non-smokers, and 412 (75.3 %) were non-drinkers. The average body mass index was 24.7 kg/m2.
Table 1

Characteristics of the Participants at Enrollment in the Korean Elderly Environmental Panel Study (2008–2010)

 

All (n = 547)

Men (n = 143)

Women (n = 404)

Age (year)

70.7 (5.2)

71.4 (4.4)

70.4 (5.5)

Height (cm)

154.7 (7.7)

164.3 (5.3)

151.3 (5.0)

Weight (kg)

59.3 (9.0)

65.7 (9.8)

57.0 (7.4)

Smoking status

Current smoker

30 (5.5)

29 (20.3)

1 (0.3)

Ex-smoker

36 (6.6)

32 (22.4)

4 (1.0)

Non-smoker

467 (85.4)

79 (55.2)

388 (96.0)

Did not answer

14 (2.6)

3 (2.1)

11 (2.7)

Alcohol consumption

Current drinker

118 (21.6)

77 (53.9)

41 (10.2)

Nondrinker

412 (75.3)

61 (42.7)

351 (86.9)

Did not answer

17 (3.1)

5 (3.5)

12 (3.0)

Blood pressure (mmHg)

Systolic blood pressure

131.8 (16.7)

130.8 (17.0)

132.1 (16.6)

Diastolic blood pressure

74.5 (9.9)

73.9 (10.2)

74.7 (9.8)

Mean arterial pressure

93.6 (11.5)

92.9 (11.9)

93.8 (11.4)

Heart rate variability

SDNN (ms)

26.9 (1.7)

25.4 (1.7)

27.4 (1.6)

RMSSD (ms)

20.5 (2.0)

18.4 (2.1)

21.3 (1.9)

LF (ms2)

90.83 (3.5)

79.0 (3.5)

95.4 (3.5)

HF (ms2)

66.9 (3.8)

50.6 (4.1)

73.8 (3.6)

HF high frequency power for frequency domain, LF low frequency power for frequency domain, RMSSD root mean square of successive differences for time domain, SDNN standard deviations of normal-to-normal intervals for time domain

Categorical data are shown as n (%) and continuous data as mean (SD), except for heart rate variability, which is presented as geometric mean (geometric SD)

Table 2 presents the air pollutant levels and meteorological factors on the days when the health examinations were conducted. The mean concentrations of PM10, NO2, and SO2 were 42.6 μg/m3, 36.5 ppb, and 4.0 ppb, respectively. The mean temperature and dew point were 16.9 °C and 6.2 °C, respectively. The air pollutant levels and meteorological factors on the health examination days were similar over the 3 previous lag days (data not shown).
Table 2

Air Pollutant Levels and Meteorological Factors in Seongbuk-Gu (Seoul, Republic of Korea) on the Survey Days, Korean Elderly Environmental Panel Study, 2008–2010

 

Mean (SD)

Median

Range

IQR

PM10 (μg/m3)

42.6 (24.7)

37.1

7.6, 151.3

25.4

NO2 (ppb)

36.5 (12.6)

34.9

9.8, 81.0

16.8

SO2 (ppb)

4.0 (2.1)

3.5

1.0, 13.9

2.4

Mean temperature (°C)

16.9 (9.0)

18.5

-7.2, 29.2

15.0

Dew point (°C)

6.2 (10.8)

7.7

-25.6, 21.9

17.2

IQR interquartile range, NO 2 nitrogen dioxide, PM 10 particulate matter ≤10 μm, SD standard deviation, SO 2 sulfur dioxide

An interquartile-range increase in the air pollutants (PM10, NO2, and SO2) was positively associated with systolic blood pressure (PM10: β = 0.93, 95 % confidence interval [CI]: 0.23, 1.63; NO2: β = 0.96, 95 % CI: 0.18, 1.74; SO2: β = 1.60, 95 % CI: 0.79, 2.42), diastolic blood pressure (PM10: β = 0.62, 95 % CI: 0.20, 1.03; NO2: β = 0.77, 95 % CI: 0.31, 1.24; SO2: β = 0.74, 95 % CI: 0.26, 1.23), and mean arterial pressure (PM10: β = 0.71, 95 % CI: 0.22, 1.19; NO2: β = 0.82, 95 % CI: 0.28, 1.36; SO2: β = 1.02, 95 % CI: 0.46, 1.58). However, air pollution exposure was not associated with heart rate variability (Table 3). In multiple pollutant models which include all air pollutants and covariates, SO2 levels were still associated with systolic blood pressure (β = 1.43, 95 % CI: 0.60, 2.26), diastolic blood pressure (β = 0.59, 95 % CI: 0.09, 1.09), and mean arterial pressure (β = 0.86, 95 % CI: 0.29, 1.44), while other associations were not found (data not shown).
Table 3

Associations of PM10, NO2, and SO2 with Blood Pressure and Heart Rate Variability per Interquartile-Range Increase in Air Pollutant Concentration Using Linear Mixed Models, Korean Elderly Environmental Panel Study, 2008–2010

 

PM10

NO2

SO2

 

No. of Obs.

Estimate

95 % CI

No. of Obs.

Estimate

95 % CI

No. of Obs.

Estimate

95 % CI

SBP

1719

0.93

0.23, 1.63

1719

0.96

0.18, 1.74

1719

1.60

0.79, 2.42

DBP

1719

0.62

0.20, 1.03

1719

0.77

0.31, 1.24

1719

0.74

0.26, 1.23

MAP

1719

0.71

0.22, 1.19

1719

0.82

0.28, 1.36

1719

1.02

0.46, 1.58

SDNN

1719

-0.01

-0.03, 0.02

1719

-0.02

-0.05, 0.01

1719

0.01

-0.02, 0.03

RMSSD

1718

-0.01

-0.04, 0.02

1718

-0.02

-0.05, 0.02

1718

0.01

-0.02, 0.05

LF

1719

0.004

-0.06, 0.06

1719

-0.02

-0.10, 0.05

1719

0.04

-0.02, 0.11

HF

1719

-0.03

-0.09, 0.03

1719

-0.05

-0.13, 0.02

1719

-0.004

-0.07, 0.06

CI confidence interval, DBP diastolic blood pressure, HF high frequency power for frequency domain, LF low frequency power for frequency domain, MAP mean arterial pressure, No. of Obs. number of observations used in the analysis, NO 2 nitrogen dioxide, PM 10 particulate matter ≤10 μm, RMSSD root mean square of successive differences for time domain, SBP systolic blood pressure, SDNN standard deviations of normal-to-normal intervals for time domain, SO 2 sulfur dioxide

We constructed separate models to assess the interaction of 46 SNPs in 18 genes with air pollution exposure. Among these models, 15 SNPs exhibited interactions (P value for interaction < 0.05) with the air pollutants with regard to blood pressure. We selected 11 SNPs (rs4646421, rs1801133, rs2917670, rs2066853, rs4680, rs854560, rs5277, rs769218, rs1799983, rs7830, and rs2234922) from these 15 SNPs, because we observed linkage disequilibrium between rs4646422 and rs4646421 (|D’| = 0.98), rs1437135 and rs2917670 (|D’| = 0.93), rs769218 and rs769217 (|D’| = 1), and rs1799983 and rs2853796 (|D’| = 0.94). The genetic risk score for blood pressure was calculated by summing the number of risk alleles for the 11 SNPs. The genetic risk score for heart rate variability was calculated using 17 SNPs (rs1695, rs4646422, rs2855658, rs10012, rs1801133, rs1800566, rs3745274, rs1799931, rs854560, rs662, rs2758331, rs5277, rs769217, rs2227956, rs2853796, rs7830, and rs1051740) that were selected using the same procedure. In the present study, we did not use weighted genetic risk scores due to methodological difficulties in considering and summarizing different interaction effects among SNPs in gene-environment interaction study.

We conducted nonparametric analyses using generalized additive mixed models to assess the association of air pollution exposure with blood pressure and heart rate variability within each genetic risk score tertile (low: 6–11 risk alleles, moderate: 12–13, high: 14–19 for blood pressure; low: 8–12, moderate: 13–14, high: 15–21 for heart rate variability). The directions of the associations were different between the low risk group and the high risk group, especially for heart rate variability (Figs. 1 and 2, see Additional file 1: Figures S1–4). The associations of air pollution exposure with blood pressure and heart rate variability were modified by the genetic risk scores for the corresponding outcomes (see Additional file 1: Table S1 and Additional file 1: Table S2). However, the genetic risk scores themselves were not associated with blood pressure or heart rate variability (data not shown).
Fig. 1

Associations between ambient PM10 concentration and blood pressure in generalized additive mixed models, stratified according to tertiles of genetic risk score for blood pressure, Korean Elderly Environmental Panel Study, 2008–2010. Models were adjusted for age, sex, smoking status, alcohol drinking, body mass index, hypertension medication, and apparent temperature. The plots show the associations of ambient PM10 concentration with systolic blood pressure (a), diastolic blood pressure (b), and mean arterial blood pressure (c) in the low genetic risk score group; the associations with systolic blood pressure (d), diastolic blood pressure (e), and mean arterial blood pressure (f) in the moderate genetic risk score group; and the associations with systolic blood pressure (g), diastolic blood pressure (h), and mean arterial blood pressure (i) in the high genetic risk score group. PM10, particulate matter ≤10 μm. Solid lines, spline curves; shaded area, 95 % confidence interval

Fig. 2

Associations between ambient PM10 concentration and heart rate variability in generalized additive mixed models, stratified according to tertiles of genetic risk score for heart rate variability, Korean Elderly Environmental Panel Study, 2008–2010. Models were adjusted for age, sex, smoking status, alcohol drinking, body mass index, hypertension medication, and apparent temperature. The plots show the associations of ambient PM10 concentration with SDNN (a), RMSSD (b), LF (c), and HF (d) in the low genetic risk score group; the associations with SDNN (e), RMSSD (f), LF (g), and HF (h) in the moderate genetic risk score group; and the associations with SDNN (i), RMSSD (j), LF (k), and HF (l) in the high genetic risk score group. HF, high frequency power for frequency domain; LF, low frequency power for frequency domain; PM10, particulate matter ≤10 μm; RMSSD, root mean square of successive differences for time domain; SDNN, standard deviations of normal-to-normal intervals for time domain. Solid lines, spline curves; shaded area, 95 % confidence interval

Several sensitivity analyses revealed robust results. First, we observed very similar results for the interaction testing of each SNP using the 2-df joint test. Second, we did not observe any appreciable change when we selected different sets of SNPs from each linkage disequilibrium pair and constructed alternative genetic risk scores. Third, inverse probability weighting for the follow-up observations caused no appreciable changes (data not shown).

Discussion

In the present study, air pollution exposure was positively associated with blood pressure among non-institutionalized elderly participants. Although we did not observe an association between air pollution exposure and heart rate variability among the whole study population, air pollution exposure was inversely associated with heart rate variability in the high genetic risk group, after we stratified the participants according to genetic risk score which was calculated based on the SNPs in oxidative stress genes. The associations of air pollution exposure with blood pressure and heart rate variability were modified by the genetic risk scores, suggesting that the associations are mediated, at least partly, by oxidative stress pathways.

A limited number of studies have investigated the interactions between air pollution exposure and genetic polymorphisms with regard to blood pressure or heart rate variability. Moreover, these studies were conducted in a few study populations that consisted of middle-aged or elderly Caucasians. The association between black carbon exposure and blood pressure was reported to be modified by polymorphisms in microRNA processing genes [18], but not by polymorphisms in oxidative stress genes [5, 14]. Polymorphisms in several genes including oxidative stress genes have been demonstrated to modify the associations between particulate matter ≤2.5 μm or traffic-related PM10 and heart rate variability [1113, 1517]. In the present study, which was conducted among an Asian population which is different from Caucasian populations in the distribution of oxidative stress gene polymorphisms and inflammatory marker levels related to cardiovascular disease risk [33, 34], we investigated the interactions between exposure to air pollutants (PM10, NO2, and SO2) and 46 SNPs in 18 oxidative stress genes on blood pressure and heart rate variability. Our results suggest that there is a common oxidative stress-related mechanism that affects the associations of air pollution exposure with blood pressure and heart rate variability.

Among 46 SNPs in 18 oxidative stress-related genes, 15 SNPs in 11 genes showed interactions with air pollutants with regard to blood pressure, and 17 SNPs in 17 genes showed interactions with regard to heart rate variability. The present results suggest that the observed interactions between air pollutants and genetic predisposition to oxidative stress on blood pressure and heart rate variability may be attributable to cumulative effects of various risk alleles and not to specific SNPs. Because we genotyped SNPs exhibiting ≥5 % minor allele frequency for the Japanese and Chinese populations in the HapMap, our results can be considered to support common disease-common variant hypothesis, which is characterized by the condition of high allele frequencies with low relative risk [35]. Despite the methodological difficulties in considering and summarizing different interaction effects among SNPs in the gene-environment interaction analysis and assessing potential gene-gene interactions, the current study has originality in investigating gene-environment interactions rather than main effects of genetic polymorphisms.

Previous studies have reported heterogeneous results regarding the association between air pollution exposure and heart rate variability, both in terms of its magnitude and its direction [36]. For example, some studies reported that air pollution exposure was associated with a decline in heart rate variability [37, 38], whereas other studies have reported a null association or an association with an increase in heart rate variability [39, 40]. These inconsistencies might be attributable to differences in the study populations’ genetic characteristics, and especially in genes that are related to oxidative stress. In the present study, we observed opposite directions for the associations among participants with high and low genetic risk scores calculated using their oxidative stress gene polymorphisms. These opposing directions might explain why we did not observe an association between air pollution exposure and heart rate variability among the whole study population.

The mechanism for the positive association between air pollution exposure and heart rate variability in the low genetic risk group is not clear. Previous studies have suggested that low-dose oxidative stress that is induced by environmental pollutants could stimulate an beneficial adaptive effect, by increasing the synthesis of antioxidants and promoting mitochondrial function [4143]. Although the level of exposure to air pollutants is the same, people with reduced genetic predisposition to oxidative stress might actually benefit from low-dose oxidative stress challenges via air pollution. However, data regarding the protective influence of low-level air pollution exposure are limited [4446]. Moreover, the role of genetic polymorphisms in this process has not been investigated. Further studies are needed to fully understand the gene-environment interactions that we observed.

The associations of the increases in interquartile range for each air pollutant (PM10, NO2, and SO2) with outcomes of interest were comparable with respect to directions and magnitudes (Table 3). When we constructed multiple pollutant models that include all air pollutants and covariates, SO2 levels remained to be associated with higher blood pressure (data not shown). However, this result should be interpreted cautiously because precise assessment of independent contributions of each air pollutant is difficult due to high correlation among individual air pollutants (data not shown).

The gene-environment interaction may also explain why only a small portion of heritability can be explained by common risk variants in genome-wide association studies [9, 47]. In the present study, genetic risk scores were not associated with blood pressure or heart rate variability, although interactions with air pollution exposure were observed. Similarly, air pollution exposure was not associated with heart rate variability when we analyzed the whole study population. However, when we stratified the study population according to genetic risk, air pollution exposure was inversely associated with heart rate variability in the high genetic risk group. Our results suggest that genetic or environmental factors cannot be accurately evaluated if only the main effects are considered without considering potential interactions.

The present study has several limitations. First, we used monitoring data as a proxy for individual air pollution exposures, which introduces the possibility of misclassification. However, estimating the individual exposure from the monitoring data based on the participant’s residence appears to be reasonable, given the fact that most of the participants were retired or unemployed due to their age. Second, the present study was conducted among elderly adults and our findings may not generalize to younger populations. Third, although we evaluated a relatively large number of oxidative stress-related genetic polymorphisms, some studies have reported that non-oxidative stress-related genes might also interact with air pollution exposure with regard to blood pressure or heart rate variability [1, 16, 18].

Conclusions

We found that the associations of air pollution exposure with blood pressure and heart rate variability were modified by genetic risk scores calculated using polymorphisms in oxidative stress genes, which suggest that an oxidative stress-related mechanism may contribute to the associations. Because the effects of air pollution exposure may differ by genetic predispositions, future air quality guidelines should take into account genetically susceptible populations and standards should be set to the lower levels considering the potential adverse health effects in these vulnerable subgroups.

Abbreviations

BP: 

blood pressure

HF: 

high frequency power for frequency domain

HRV: 

heart rate variability

LF: 

low frequency power for frequency domain

PM10

particulate matter ≤10 μm

RMSSD: 

root mean square of successive differences for time domain

SDNN: 

standard deviations of normal-to-normal intervals for time domain

SNP: 

single-nucleotide polymorphism

Declarations

Acknowledgments

This work was supported by the R&D Program for Society of the National Research Foundation (grant number: 2014M3C8A5030619) and the Basic Science Research Program (2013R1A1A3A04004612) through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning, Republic of Korea.

Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Authors’ Affiliations

(1)
Department of Preventive Medicine, Seoul National University College of Medicine
(2)
Department of Bioscience and Bioengineering, Sejong University
(3)
Seoul Metropolitan Institute of Public Health and Environment
(4)
Institute of Environmental Medicine, Medical Research Center
(5)
Environmental Health Center, Seoul National University College of Medicine

References

  1. Koulova A, Frishman WH. Air pollution exposure as a risk factor for cardiovascular disease morbidity and mortality. Cardiol Rev. 2014;22:30–6.View ArticleGoogle Scholar
  2. von Klot S, Gryparis A, Tonne C, Yanosky J, Coull BA, Goldberg RJ, Lessard D, Melly SJ, Suh HH, Schwartz J. Elemental carbon exposure at residence and survival after acute myocardial infarction. Epidemiol Camb Mass. 2009;20:547–54.View ArticleGoogle Scholar
  3. Kubesch N, De Nazelle A, Guerra S, Westerdahl D, Martinez D, Bouso L, Carrasco-Turigas G, Hoffmann B, Nieuwenhuijsen MJ. Arterial blood pressure responses to short-term exposure to low and high traffic-related air pollution with and without moderate physical activity. Eur J Prev Cardiol. 2015;22:548–57.View ArticleGoogle Scholar
  4. Sarnat JA, Golan R, Greenwald R, Raysoni AU, Kewada P, Winquist A, Sarnat SE, Dana Flanders W, Mirabelli MC, Zora JE, Bergin MH, Yip F. Exposure to traffic pollution, acute inflammation and autonomic response in a panel of car commuters. Environ Res. 2014;133:66–76.View ArticleGoogle Scholar
  5. Mordukhovich I, Wilker E, Suh H, Wright R, Sparrow D, Vokonas PS, Schwartz J. Black carbon exposure, oxidative stress genes, and blood pressure in a repeated-measures study. Environ Health Perspect. 2009;117:1767–72.Google Scholar
  6. Miller MR. The role of oxidative stress in the cardiovascular actions of particulate air pollution. Biochem Soc Trans. 2014;42:1006–11.View ArticleGoogle Scholar
  7. Mills NL, Donaldson K, Hadoke PW, Boon NA, MacNee W, Cassee FR, Sandström T, Blomberg A, Newby DE. Adverse cardiovascular effects of air pollution. Nat Clin Pract Cardiovasc Med. 2009;6:36–44.View ArticleGoogle Scholar
  8. Ren C, Park SK, Vokonas PS, Sparrow D, Wilker E, Baccarelli A, Suh HH, Tucker KL, Wright RO, Schwartz J. Air pollution and homocysteine: more evidence that oxidative stress-related genes modify effects of particulate air pollution. Epidemiol Camb Mass. 2010;21:198–206.View ArticleGoogle Scholar
  9. Ober C, Vercelli D. Gene-environment interactions in human disease: nuisance or opportunity? Trends Genet TIG. 2011;27:107–15.View ArticleGoogle Scholar
  10. Uher R. Gene-environment interactions in common mental disorders: an update and strategy for a genome-wide search. Soc Psychiatry Psychiatr Epidemiol. 2014;49:3–14.View ArticleGoogle Scholar
  11. Adam M, Imboden M, Boes E, Schaffner E, Künzli N, Phuleria HC, Kronenberg F, Gaspoz J-M, Carballo D, Probst-Hensch N. Modifying effect of a common polymorphism in the interleukin-6 promoter on the relationship between long-term exposure to traffic-related particulate matter and heart rate variability. PLoS One. 2014;9:e104978.View ArticleGoogle Scholar
  12. Baccarelli A, Cassano PA, Litonjua A, Park SK, Suh H, Sparrow D, Vokonas P, Schwartz J. Cardiac autonomic dysfunction: effects from particulate air pollution and protection by dietary methyl nutrients and metabolic polymorphisms. Circulation. 2008;117:1802–9.View ArticleGoogle Scholar
  13. Chahine T, Baccarelli A, Litonjua A, Wright RO, Suh H, Gold DR, Sparrow D, Vokonas P, Schwartz J. Particulate air pollution, oxidative stress genes, and heart rate variability in an elderly cohort. Environ Health Perspect. 2007;115:1617–22.View ArticleGoogle Scholar
  14. Levinsson A, Olin A-C, Modig L, Dahgam S, Björck L, Rosengren A, Nyberg F. Interaction effects of long-term air pollution exposure and variants in the GSTP1, GSTT1 and GSTCD genes on risk of acute myocardial infarction and hypertension: a case-control study. PLoS One. 2014;9:e99043.View ArticleGoogle Scholar
  15. Park SK, O’Neill MS, Wright RO, Hu H, Vokonas PS, Sparrow D, Suh H, Schwartz J. HFE genotype, particulate air pollution, and heart rate variability: a gene-environment interaction. Circulation. 2006;114:2798–805.View ArticleGoogle Scholar
  16. Ren C, Baccarelli A, Wilker E, Suh H, Sparrow D, Vokonas P, Wright R, Schwartz J. Lipid and endothelium-related genes, ambient particulate matter, and heart rate variability--the VA Normative Aging Study. J Epidemiol Community Health. 2010;64:49–56.View ArticleGoogle Scholar
  17. Schwartz J, Park SK, O’Neill MS, Vokonas PS, Sparrow D, Weiss S, Kelsey K. Glutathione-S-transferase M1, obesity, statins, and autonomic effects of particles: gene-by-drug-by-environment interaction. Am J Respir Crit Care Med. 2005;172:1529–33.View ArticleGoogle Scholar
  18. Wilker EH, Baccarelli A, Suh H, Vokonas P, Wright RO, Schwartz J. Black carbon exposures, blood pressure, and interactions with single nucleotide polymorphisms in MicroRNA processing genes. Environ Health Perspect. 2010;118:943–8.View ArticleGoogle Scholar
  19. Zanobetti A, Baccarelli A, Schwartz J. Gene-air pollution interaction and cardiovascular disease: a review. Prog Cardiovasc Dis. 2011;53:344–52.View ArticleGoogle Scholar
  20. Kim JH, Hong Y-C. GSTM1, GSTT1, and GSTP1 polymorphisms and associations between air pollutants and markers of insulin resistance in elderly Koreans. Environ Health Perspect. 2012;120:1378–84.View ArticleGoogle Scholar
  21. Alessandrini E, Zauli Sajani S, Scotto F, Miglio R, Marchesi S, Lauriola P. Emergency ambulance dispatches and apparent temperature: a time series analysis in Emilia-Romagna, Italy. Environ Res. 2011;111:1192–200.View ArticleGoogle Scholar
  22. Almeida SP, Casimiro E, Calheiros J. Effects of apparent temperature on daily mortality in Lisbon and Oporto, Portugal. Environ Health Glob Access Sci Source. 2010;9:12.Google Scholar
  23. Ballester J, Robine J-M, Herrmann FR, Rodó X. Long-term projections and acclimatization scenarios of temperature-related mortality in Europe. Nat Commun. 2011;2:358.View ArticleGoogle Scholar
  24. Wichmann J, Andersen Z, Ketzel M, Ellermann T, Loft S. Apparent temperature and cause-specific emergency hospital admissions in Greater Copenhagen, Denmark. PLoS One. 2011;6:e22904.View ArticleGoogle Scholar
  25. Gabriel S, Ziaugra L, Tabbaa D. SNP genotyping using the Sequenom MassARRAY iPLEX platform. Curr Protoc Hum Genet Editor Board Jonathan Haines Al 2009, Chapter 2:Unit 2.12.Google Scholar
  26. Kim J-H, Park S-G, Lee K-H, Choi J-H, Ha E-H, Myung S-K, Hong Y-C. GSTM1 and GSTP1 polymorphisms as potential factors for modifying the effect of smoking on inflammatory response. J Korean Med Sci. 2006;21:1021–7.View ArticleGoogle Scholar
  27. Balding DJ. A tutorial on statistical methods for population association studies. Nat Rev Genet. 2006;7:781–91.View ArticleGoogle Scholar
  28. Qi L, Ma J, Qi Q, Hartiala J, Allayee H, Campos H. Genetic risk score and risk of myocardial infarction in Hispanics. Circulation. 2011;123:374–80.View ArticleGoogle Scholar
  29. Greenland S, Pearl J, Robins JM. Causal diagrams for epidemiologic research. Epidemiol Camb Mass. 1999;10:37–48.View ArticleGoogle Scholar
  30. Kraft P, Yen Y-C, Stram DO, Morrison J, Gauderman WJ. Exploiting gene-environment interaction to detect genetic associations. Hum Hered. 2007;63:111–9.View ArticleGoogle Scholar
  31. Robins JM, Rotnitzky A, Zhao LP. Analysis of semiparametric regression models for repeated outcomes in the presence of missing data. J Am Stat Assoc. 1995;90:106–21.View ArticleGoogle Scholar
  32. McCracken J, Baccarelli A, Hoxha M, Dioni L, Melly S, Coull B, Suh H, Vokonas P, Schwartz J. Annual ambient black carbon associated with shorter telomeres in elderly men: Veterans Affairs Normative Aging Study. Environ Health Perspect. 2010;118:1564–70.View ArticleGoogle Scholar
  33. Nelson HH, Wiencke JK, Christiani DC, Cheng TJ, Zuo ZF, Schwartz BS, Lee BK, Spitz MR, Wang M, Xu X. Ethnic differences in the prevalence of the homozygous deleted genotype of glutathione S-transferase theta. Carcinogenesis. 1995;16:1243–5.View ArticleGoogle Scholar
  34. Miller MA, Cappuccio FP. Ethnicity and inflammatory pathways - implications for vascular disease, vascular risk and therapeutic intervention. Curr Med Chem. 2007;14:1409–25.View ArticleGoogle Scholar
  35. Peng B, Kimmel M. Simulations provide support for the common disease-common variant hypothesis. Genetics. 2007;175:763–76.View ArticleGoogle Scholar
  36. Pieters N, Plusquin M, Cox B, Kicinski M, Vangronsveld J, Nawrot TS. An epidemiological appraisal of the association between heart rate variability and particulate air pollution: a meta-analysis. Heart Br Card Soc. 2012;98:1127–35.Google Scholar
  37. He F, Shaffer ML, Li X, Rodriguez-Colon S, Wolbrette DL, Williams R, Cascio WE, Liao D. Individual-level PM2.5 exposure and the time course of impaired heart rate variability: the APACR Study. J Expo Sci Environ Epidemiol. 2011;21:65–73.View ArticleGoogle Scholar
  38. Lee M-S, Eum K-D, Fang SC, Rodrigues EG, Modest GA, Christiani DC. Oxidative stress and systemic inflammation as modifiers of cardiac autonomic responses to particulate air pollution. Int J Cardiol. 2014;176:166–70.View ArticleGoogle Scholar
  39. Bartell SM, Longhurst J, Tjoa T, Sioutas C, Delfino RJ. Particulate air pollution, ambulatory heart rate variability, and cardiac arrhythmia in retirement community residents with coronary artery disease. Environ Health Perspect. 2013;121:1135–41.Google Scholar
  40. Shields KN, Cavallari JM, Hunt MJO, Lazo M, Molina M, Molina L, Holguin F. Traffic-related air pollution exposures and changes in heart rate variability in Mexico City: a panel study. Environ Health Glob Access Sci Source. 2013;12:7.Google Scholar
  41. Calabrese V, Cornelius C, Mancuso C, Lentile R, Stella AMG, Butterfield DA. Redox homeostasis and cellular stress response in aging and neurodegeneration. Methods Mol Biol Clifton NJ. 2010;610:285–308.View ArticleGoogle Scholar
  42. Lee D-H, Jacobs DR. Hormesis and public health: can glutathione depletion and mitochondrial dysfunction due to very low-dose chronic exposure to persistent organic pollutants be mitigated? J Epidemiol Community Health. 2015;69:294–300.View ArticleGoogle Scholar
  43. Luna-López A, González-Puertos VY, López-Diazguerrero NE, Königsberg M. New considerations on hormetic response against oxidative stress. J Cell Commun Signal. 2014;8:323–31.View ArticleGoogle Scholar
  44. Daniels MJ, Dominici F, Samet JM, Zeger SL. Estimating particulate matter-mortality dose-response curves and threshold levels: an analysis of daily time-series for the 20 largest US cities. Am J Epidemiol. 2000;152:397–406.View ArticleGoogle Scholar
  45. Franklin M, Zeka A, Schwartz J. Association between PM2.5 and all-cause and specific-cause mortality in 27 US communities. J Expo Sci Environ Epidemiol. 2007;17:279–87.View ArticleGoogle Scholar
  46. Joseph PM. Can fine particulate matter explain the paradoxical ozone associations? Environ Int. 2008;34:1185–91.View ArticleGoogle Scholar
  47. Thomas D. Gene--environment-wide association studies: emerging approaches. Nat Rev Genet. 2010;11:259–72.View ArticleGoogle Scholar

Copyright

© Kim et al. 2016

Advertisement