What this is
- This research investigates the link between long-term exposure to PM2.5 and () in adults and the elderly in Guangdong, China.
- It analyzes data from 6628 participants to assess how PM2.5 affects components like fasting blood glucose and triglyceride levels.
- Findings indicate that increased PM2.5 exposure correlates with higher risks of and its components, emphasizing the need for environmental improvements.
Essence
- Long-term exposure to PM2.5 is associated with an increased risk of and its components, including elevated fasting blood glucose and triglycerides.
Key takeaways
- A 10-μg/m³ increase in PM2.5 exposure raises the odds of developing by 17%. This suggests that air quality improvements could reduce prevalence.
- Participants exposed to higher PM2.5 levels showed a 36% increase in the risk of hypertriglyceridemia. This highlights the significant impact of air pollution on lipid metabolism.
- The study found that rural residents and those over 45 years are particularly vulnerable to the adverse effects of PM2.5 exposure, indicating a need for targeted public health interventions.
Caveats
- The cross-sectional design limits the ability to establish causality between PM2.5 exposure and . Longitudinal studies are needed for confirmation.
- Data on secondary diseases related to were incomplete, which may affect the accuracy of the results.
- Interactions between PM2.5 and other indoor pollutants were not fully explored, potentially confounding the findings.
Definitions
- Metabolic syndrome (MetS): A cluster of metabolic disorders including abdominal obesity, hypertension, dyslipidemia, and hyperglycemia.
AI simplified
Introduction
Metabolic syndrome (MetS) is a cluster of metabolic disorders including abdominal obesity, hypertension, hypertriglyceridemia, low high-density lipoprotein cholesterol (HDL-c) and hyperglycemia [1]. MetS has been recognized as an urgent public health concern because it affects 20–30% of the global population, of which the standardized prevalence of MetS is around 24.2% in China [2, 3]. Previous studies showed that MetS was associated with an increased risk of cardiovascular diseases (CVDs), diabetes mellitus, cancers and other chronic non-communicable diseases [4, 5]. Evidence suggests that MetS-related adverse health outcomes may be enhanced not only by genetic factors, physical inactivity and unhealthy diet [6–9], but also by environmental pollutant exposure [10, 11], including air pollution.
Accumulating studies have added to the evidence that the inhalation of particulate matte ≤ 2.5 µm (PM2.5) might lead to pulmonary oxidative stress, systemic inflammation, vascular dysfunction and atherosclerosis [12–16]. Previous studies suggested that PM2.5 was the major risk factor for adverse health outcomes including hypertension [12], obesity [13], elevated fasting blood glucose (FBG) [14, 15], waist circumference [16] and dyslipidemia [17], which were crucial components in the diagnosis of MetS. However, the effects of PM2.5 on blood pressure [18, 19], fasting blood glucose [20, 21] and obesity [22–24] still remained inconsistent. Furthermore, the evidence concerning the associations of air pollution and MetS is still scarce. To our knowledge, only a few studies have reported the detrimental effects of long-term exposure to ambient air pollution on MetS [17, 25–28], which were mainly conducted in the developed countries such as Korea, North America or Saudi Arabia [17, 25, 26]. Only two epidemiological studies evaluated the associations between PM2.5 and the prevalence of MetS in the developing countries such as China [27, 28] among adolescents and children [27], and adults and elderly [28]. In addition, the effects of PM2.5 on specific components on MetS in Chinese population was limited based on the prior evidence.
As one of the most developed provinces in southern China, there has been considerable lifestyle and dietary changes during these decades in Guangdong, resulting in the increase of MetS and stroke, coronary heart disease, and cancers [29]. Meanwhile, air pollution has become one of the most severe environmental problem in Guangdong [30]. In the CAPES study, despite a relatively low concentrations of PM, there was a higher risk of the total, cardiovascular and respiratory mortality attributed to PM in Guangzhou (the capital city of Guangdong province), compared with the heavy industry cities in northeastern China, where PM pollution was more severe [31]. The relatively higher concentration of the toxic component including polybrominated diphenyl ethers (PBDEs) found in PM2.5 in southern China [32, 33] might help provide the evidence for the stronger association between PM and mortality.
Considering the current MetS epidemic, the more toxic effect of PM2.5 in south China, the inconsistent effects of PM2.5 on specific components of MetS, and the limited information of the association between PM2.5 and MetS, we explored the effects of ambient PM2.5 pollution on MetS and its components [blood pressure, triglyceride (TG), high-density lipoprotein-cholecsterol (HDL-c), fasting blood glucose (FBG) and waist circumference] in Guangdong, China. To address the knowledge gap, our findings would provide important public health implications which aimed to reduce the detrimental impact of ambient air pollution of PM2.5 on CVDs and MetS in China.
Materials and methods
Study design and participants
This study was conducted using a multistage, probability-based sampling strategy, based on the Chronic Disease and Risk Factors Surveillance in 2015 in Guangdong province, China. 14 surveillance points were randomly selected. Between October 2015 and February 2016, adults aged 18 years who were living in the current residence for at least 6 months were recruited. All participants were interviewed face-to-face by using a structured questionnaire, which has been described previously [34, 35]. In addition, participants underwent anthropometric measurements (blood pressure, fasting glucose, blood pressure, waist circumstance, height and weight) and blood sample collection by the well-trained public health practitioners from the local health stations or community health service centers. The study protocol was approved by the ethics review committee of the National Center for Chronic and Non-Communicable Disease Control and Prevention, China Center for Disease Control and Prevention. All participants were provided written informed consent. Inclusion and exclusion criteria of participants have been reported previously [36].
MetS definition
| Conditions | Recommended threshold | |
|---|---|---|
| For Men | For women | |
| Elevated TG levels | ≥ 1.7 mmol/l (150 mg/dl) | ≥ 1.7 mmol/l (150 mg/dl) |
| Decreased HDL-c levels | < 1.0 mmol/l (40 mg/dl) for males | < 1.3 mmol/l (50 mg/dl) |
| Elevated blood pressure | Elevated blood pressure | Elevated blood pressure |
| Elevated FBG levels | FBG ≥ 5.6 mmol/l (100 mg/dl) | FBG ≥ 5.6 mmol/l (100 mg/dl) |
| Central obesity | waist circumference ≥ 90 cm | waist circumference ≥ 80 cm |
Assessment of long-term exposure to air pollution
We used the spatiotemporal land-use regression (LUR) model to estimate the two-year average exposure of ambient air pollutants including PM2.5, particulate matter < 10 µm (PM10), sulfur dioxide (SO2), nitrogen dioxide (NO2) and ozone (O3) at individual levels. The details of the data and prediction process has been published previously [38], which were as follows:The spatiotemporal LUR model was built with the following predictors: population density, road length, land-use data (farmland, blue space, living land, and green space), and ambient visibility. Two smooth temporal basis functions were analyzed to estimate the secular trend of air pollution. The R2 was 88.86% with the root mean square error (RMSE) of 5.65%, based on the findings of the tenfold cross-validation.Residence address was extracted from the questionnaire and included into the model to forecast the weekly average air pollution between April 2013 and December 2016.The two-year averaged air pollutant concentrations before the investigation date were estimated for each individual.
Covariates
The following covariates were incorporated to examine the potential confounding and mediating effects: age, sex (man and woman), race (Han and minority), region (urban and rural), occupation (physical work and non-physical work), education level (none, primary school education, middle school education, university education or higher), marital status (none, primary school education, middle school education and university education or higher), household income (< 30, 30–50, 50–100, 100–200 and ≥ 200 × 1000 RMB), weight change in the past year (an increase of > 2.5 kg, unchanged < 2.5 kg, a decease of > 2.5 kg and unclear), alcohol consumption, exercise, family history of diabetes (no and yes), exercise (no and yes), alcohol consumption (no and yes), passive smoking (no and yes), cigarette smoking (non-smoker and smoker), biomass fuel use (no and yes), body-mass index (BMI) (under weight, normal and overweight/obese), grain consumption, vegetable and fruit consumption and red meat consumption. The definition of the covariates is summerized in E-Table 1 [34, 35, 39, 40].
Statistical analysis
We analyzed the characteristics between the groups with MetS and without MetS, by demonstrating the mean and standard deviation for continuous variables and frequencies for categorical variables. The t-test was performed to analyze the distribution of continuous variables, and when indicated, appropriate transformation was applied. A contingency table and Chi-squared test was performed for analyzing the frequencies of categorical variables. The normality and equality of variance was assessed by using the Shapiro–Wilk’s test and Bartlett’s test, respectively. The odds ratios (ORs) and 95% confidence intervals (95%CIs) were calculated for determining the association between ambient air pollutant exposure to PM2.5 and the presence of MetS and its components by using the generalized linear mixed model, based on the three stepwise models to confirm the validity of findings. Family was treated as random effect by calculating the intraclass correlation coefficient (ICC). We compared the Akaike's information criterion value of these three models to avoid over-fitting. The magnitude of collinearity was assessed based on the variance inflation factor (VIF). The VIF of 5 or greater indicated collinearity among the variables. Variables with the evidence of a significant collinearity were excluded from the model. The Spearman’s rank correlation test was used to determine the relationship between pollutants. Strong, moderate, and weak correlations were defined as the coefficients (rs) greater than 0.60, 0.30 to 0.60, and less than 0.30, respectively. Since strong and moderate correlation was identified between PM2.5 and other pollutant models, we only applied the single pollutant model (PM2.5) to avoid covariance. We further stratified the study participants by the region, sex, age, cigarette smoking, alcohol consumption, exercise, BMI, grain consumption, vegetable and fruit consumption and red meat consumption, to study the significant associations between PM2.5 and MetS, high TG and FBG in each stratum. We also included the interaction terms in the generalized linear mixed effect models to test the interactions between PM2.5 and MetS, high TG and FBG in each subgroup. All statistical analyses were performed with R software (version 4.0.2). The threshold of statistical significance for P value was set to be 0.05.
Results
Associations of long-term PMexposure with MetS, high TG and high FBG in different stratum (*represents thevalue for interaction with significance).) Metabolic syndrome,) High triglyceride,) High fasting blood glucose 2.5 p A B C
| Characteristics | (= 6628)Totaln | Metabolic Syndrome | p | |
|---|---|---|---|---|
| (= 1691)Eventn | (= 4937)Non-Eventn | |||
| Age (year), mean (SD) | 50.12 (14.73) | 54.09 (12.83) | 48.76 (15.09) | < 0.001* |
| Sex, n (%) | < 0.001* | |||
| Man | 2955 (44.6) | 677 (40.0) | 2278 (44.6) | |
| Women | 3673 (55.4) | 1014 (60.0) | 2659 (55.4) | |
| Race, n (%) | 0.57 | |||
| Han | 6562 (99.0) | 1672 (98.9) | 4890 (99.0) | |
| Minority | 66 (1.0) | 19 (1.1) | 47 (1.0) | |
| Region, n (%) | 0.092 | |||
| Urban | 3613 (54.5) | 892 (52.7) | 2721 (55.1) | |
| Rural | 3015 (45.5) | 799 (47.3) | 2216 (44.9) | |
| Occupation, n (%) | 0.273 | |||
| Physical work | 5070 (76.5) | 1310 (77.5) | 3760 (76.2) | |
| Non-physical work | 1558 (23.5) | 381 (22.5) | 1177 (23.8) | |
| Educational level, n (%) | < 0.001* | |||
| None | 836 (12.6) | 303 (17.9) | 533 (10.8) | |
| Primary school education | 2246 (33.9) | 625 (37.0) | 1621 (32.8) | |
| Middle school education | 2905 (43.8) | 656 (38.8) | 2249 (45.6) | |
| University education or higher | 641 (9.7) | 107 (6.3) | 534 (10.8) | |
| Marriage status, n (%) | < 0.001* | |||
| Unmarried | 357 (5.4) | 46 (2.7) | 311 (6.3) | |
| Married | 5968 (90.0) | 1544 (91.3) | 4424 (89.6) | |
| Widowed or divorced | 303 (4.6) | 101 (6.0) | 202 (4.1) | |
| Household income (× 1000 RMB) | 0.038* | |||
| < 30 | 1029 (15.5) | 261 (15.4) | 768 (15.6) | |
| 30 ≤ Household income < 50 | 1171 (17.7) | 321 (19.0) | 850 (17.2) | |
| 50 ≤ Household income < 100 | 1218 (18.4) | 280 (16.6) | 938 (19.0) | |
| 100 ≤ Household income < 200 | 514 (7.8) | 122 (7.2) | 392 (7.9) | |
| ≥ 200 | 171 (2.6) | 34 (2.0) | 137 (2.8) | |
| Refuse to answer or don't know | 2525 (38.1) | 673 (39.8) | 1852 (37.5) | |
| Behaviors factors | ||||
| Cigarette smoking | 0.003* | |||
| Nonsmoker | 4428 (66.8) | 1180 (69.8) | 3248 (65.8) | |
| Smoker | 2200 (33.2) | 511 (30.2) | 1689 (34.2) | |
| Alcohol consumption, n (%) | 0.027* | |||
| No | 3929 (59.3) | 1041 (61.6) | 2888 (58.5) | |
| Yes | 2699 (40.7) | 650 (38.4) | 2049 (41.5) | |
| Exercise, n (%) | 0.019* | |||
| No | 5479 (82.7) | 1430 (86.6) | 4049 (82.0) | |
| Yes | 1149 (17.3) | 261 (15.4) | 888 (18.0) | |
| Family history of diabetes mellitus, n (%) | 0.524 | |||
| No | 6222 (93.9) | 1582 (93.6) | 4640 (94.0) | |
| Yes | 406 (6.1) | 109 (6.4) | 297 (6.0) | |
| Weight change in the past 12 months, n (%) | 0.221 | |||
| Increase in > 2.5 kg | 609 (9.2) | 164 (9.7) | 445 (9.0) | |
| Unchanged (< 2.5 kg) | 4743 (71.6) | 1217 (72.0) | 3526 (71.4) | |
| Decease in > 2.5 kg | 596 (9.0) | 132 (7.8) | 464 (9.4) | |
| Unclear | 680 (10.3) | 178 (10.5) | 502 (10.2) | |
| Household air pollution exposure | ||||
| Passive smoking, n (%) | 0.598 | |||
| No | 1531 (32.0) | 393 (31.2) | 1138 (32.3) | |
| Yes | 3250 (68.0) | 868 (68.8) | 2382 (67.7) | |
| Biomass fuel | 0.323 | |||
| No | 5136 (77.5) | 1325 (78.4) | 3811 (77.2) | |
| Yes | 1492 (22.5) | 366 (21.6) | 1126 (22.8) | |
| Grain consumption(g/daily), medium (IQR) | 400.00 (376.90) | 394.39 (371.73) | 400.00 (380.00) | 0.212 |
| Vegetable and Fruit consumption (g/daily), medium (IQR) | 308.00 (298.35) | 308.00 (293.33) | 308.33 (299.00) | 0.891 |
| Red Meat consumption (g/daily), medium (IQR) | 71.43 (95.96) | 53.57 (96.29) | 80.00 (115.86) | 0.019* |
| Ambient air pollution exposure (μg/m), mean (SD)3 | ||||
| PM2.5 | 37.2 (4.8) | 37.4 (4.5) | 37.1 (4.8) | 0.04* |
| PM10 | 55.4 (5.0) | 55.2 (4.8) | 55.5 (5.1) | 0.041* |
| SO2 | 16.1 (3.8) | 16.2 (3.9) | 16.1 (3.8) | 0.227 |
| NO2 | 26.0 (12.6) | 25.3 (11.4) | 26.2 (12.9) | 0.007* |
| O3 | 56.2 (6.4) | 56.7 (5.9) | 56.1 (6.5) | 0.005* |
| Anthropometry | ||||
| BMI (kg/m2), mean (SD) | 23.04 (3.36) | 25.65 (3.21) | 22.15 (2.92) | < 0.001* |
| BMI category, n (%) | < 0.001* | |||
| Under weight | 504 (56.4) | 15 (0.9) | 489 (9.9) | |
| Normal | 3741 (7.6) | 504 (29.8) | 3237 (65.6) | |
| Overweight/ Obese | 2383 (36.0) | 1172 (69.3) | 1211 (24.5) | |
| MetS, n (%) | 1691 (25.5) | 1691 (100.0) | - | |
| Central obesity, n (%) | 2038 (30.7) | 1236 (73.1) | 802 (16.2) | < 0.001* |
| High TG, n (%) | 1379 (20.8) | 998 (59.0) | 381 (7.7) | < 0.001* |
| Low HDL-c, n (%) | 2759 (41.6) | 1325 (78.4) | 1434 (29.0) | < 0.001* |
| Hypertension, n (%) | 3339 (50.4) | 1430 (84.6) | 1909 (38.7) | < 0.001* |
| High FBG, n (%) | 1606 (24.2) | 920 (54.4) | 686 (13.9) | < 0.001* |
| Summary statistics | Spearman correlation coefficients | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Mean | Median | Minimum | Maximum | IQR | PM2.5 | PM10 | SO2 | NO2 | O3 | |
| PM(μg/m)2.53 | 37.17 | 38.3 | 27.99 | 46.96 | 8.84 | 1 | 0.71* | 0.52* | 0.6* | -0.49* |
| PM(μg/m)103 | 55.43 | 55.09 | 42.17 | 67.33 | 7.58 | 1 | 0.63* | 0.75* | -0.51* | |
| SO(μg/m)23 | 16.12 | 15.92 | 9.31 | 22.28 | 5.44 | 1 | 0.37* | -0.35* | ||
| NO(μg/m)23 | 25.98 | 23.07 | 7.94 | 62.68 | 18.07 | 1 | -0.68* | |||
| O(μg/m)33 | 56.23 | 56.96 | 40.54 | 68.83 | 7.38 | 1 | ||||
| Variables | MetS | Central obesity | High TG | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| AIC | OR (95%CI) | P | AIC | OR (95%CI) | P | AIC | OR (95%CI) | P | |||
| Model 1 | 7522.5 | 1.14(1.01, 1.29) | 0.039* | 8181.9 | 1.04 (0.93, 1.16) | 0.516 | 6757.9 | 1.36 (1.34, 1.38) | < 0.001* | ||
| Model 2 | 5807.8 | 1.17(1.15, 1.19) | < 0.001* | 4075.2 | 1.02 (0.85, 1.23) | 0.806 | 6190.5 | 1.40 (1.21, 1.62) | < 0.001* | ||
| Model 3 | 5807.8 | 1.17(1.01, 1.35) | 0.042* | 4060.1 | 0.98 (0.82, 1.18) | 0.813 | 6183 | 1.36 (1.18, 1.58) | < 0.001* | ||
| Variables | Low HDL-c | Hypertension | High FBG | ||||||||
| AIC | OR (95%CI) | P | AIC | OR (95%CI) | P | AIC | OR (95%CI) | P | |||
| Model 1 | 8989 | 1.00 (0.90, 1.12) | 0.944 | 9171 | 1.04 (0.93, 1.16) | 0.506 | 7302.1 | 1.17 (1.02, 1.35) | 0.023* | ||
| Model 2 | 8358.3 | 0.98 (0.87, 1.11) | 0.784 | 7680.7 | 1.02 (0.90, 1.16) | 0.721 | 6871.9 | 1.15 (1.01, 1.33) | 0.047* | ||
| Model 3 | 8318.3 | 0.99 (0.88, 1.12) | 0.867 | 7658.5 | 1.03 (0.91,1.16) | 0.682 | 6823.2 | 1.18 (1.02, 1.36) | 0.026* | ||
| Variable | MetS | Pinter | High TG | Pinter | High FBG | Pinter |
|---|---|---|---|---|---|---|
| OR (95%CI) | OR (95%CI) | OR (95%CI) | ||||
| Region | 0.054 | 0.004* | < 0.001* | |||
| Urban (= 3613)n | 1.03 (0.85, 1.24) | 1.15 (0.96, 1.36) | 0.79 (0.66, 0.95)* | |||
| Rural (= 3015)n | 1.38 (1.11, 1.70)* | 1.71 (1.37, 2.13)* | 1.87 (1.55, 2.25)* | |||
| Sex | 0.275 | 0.848 | 0.517 | |||
| Men (= 2955)n | 1.24 (1.01, 1.53)* | 1.34 (1.11, 1.62)* | 1.23 (1.02, 1.47)* | |||
| Women (= 3673)n | 1.09 (0.90, 1.31) | 1.36 (1.12, 1.66)* | 1.13 (0.95, 1.35) | |||
| Age | 0.412 | 0.083 | 0.049* | |||
| < 45 years(= 2316)n | 1.07 (0.81, 1.39) | 1.21 (0.96, 1.54) | 0.89 (0.68, 1.16) | |||
| ≥ 45 years(= 4312)n | 1.19 (1.01, 1.39)* | 1.43 (1.22, 1.69)* | 1.26 (1.09, 1.45)* | |||
| Cigarette smoking | 0.793 | 0.637 | 0.081 | |||
| Nonsmoker(= 4428)n | 1.16 (0.98, 1.38) | 1.35 (1.15, 1.59)* | 1.14 (0.98, 1.34) | |||
| Smoker(= 2200)n | 1.17 (0.91, 1.50) | 1.42 (1.20, 1.68)* | 1.33 (1.07, 1.65)* | |||
| Alcohol consumption | 0.156 | 0.195 | 0.261 | |||
| Non-drinker (= 3929)n | 1.20 (0.96, 1.50) | 1.32 (1.06, 1.64)* | 1.35 (1.10, 1.66)* | |||
| Drinker (= 2699)n | 1.14 (0.95, 1.36) | 1.37 (1.16, 1.61)* | 1.10 (0.93, 1.29) | |||
| Exercise | 0.269 | 0.015* | 0.04* | |||
| No (= 5479)n | 1.17 (1.01, 1.35)* | 1.37 (1.18, 1.58)* | 1.28 (1.11, 1.46)* | |||
| Yes (= 1149)n | 1.14 (0.82, 1.60)* | 0.97 (0.72, 1.31)* | 0.95 (0.68, 1.33) | |||
| BMI | 0.795 | 0.681 | 0.349 | |||
| Underweight(= 504)n | 1.02 (0.65, 1.58) | 1.25 (0.48, 3.27) | 1.72 (0.98, 3.02) | |||
| Normal (= 3741)n | 1.16 (0.93, 1.43) | 1.47 (1.20, 1.79)* | 1.21 (1.02, 1.44)* | |||
| Over weight/Obese (= 2383)n | 1.14 (0.96, 1.36) | 1.28 (1.06, 1.53)* | 1.12 (0.93, 1.36) | |||
| Grain consumption | 0.713 | 0.897 | < 0.001* | |||
| < 400 g/daily(= 3232)n | 1.32 (1.08, 1.62)* | 1.38 (1.14, 1.68)* | 1.59 (1.32, 1.91)* | |||
| ≥ 400 g/daily (= 3396)n | 1.09 (0.90, 1.34) | 1.33 (1.11, 1.61)* | 0.96 (0.80, 1.15) | |||
| Vegetable and Fruit consumption | 0.979 | 0.869 | 0.419 | |||
| < 400 g/daily(= 3858)n | 1.25 (1.03, 1.51)* | 1.41 (1.18, 1.69)* | 1.43 (1.20, 1.70)* | |||
| ≥ 400 g/daily (= 2770)n | 1.14 (0.92, 1.41) | 1.36 (1.10, 1.66)* | 0.91 (0.75, 1.11) | |||
| Red Meat consumption | 0.312 | 0.332 | 0.86 | |||
| < 100 g/daily (= 3970)n | 1.16 (0.97, 1.39) | 1.29 (1.09, 1.53)* | 1.33 (1.13, 1.57)* | |||
| ≥ 100 g/daily (= 2658)n | 1.19 (0.96, 1.49) | 1.38 (1.13, 1.70)* | 1.02 (0.83, 1.24) |
Discussion
Understanding the impacts of long-term exposure to ambient PM2.5 on MetS is crucial, because 25.5% of the population had MetS in the studied regions of southern China. This study was conducted to elucidate the key research question regarding whether exposure to ambient PM2.5 would increase the risk of having MetS and confer a detrimental impact on its specific components in Guangdong province. Information regarding the associations between PM2.5 and the prevalence of MetS with its specific components in China remains scarce. Reassuringly, we found that long-term exposure to ambient PM2.5 pollution was significantly associated with an increased risk of MetS. In addition, long-term exposure to PM2.5 increased the risk of high TG and high FBG. Furthermore, the participants living in rural area, aged greater than 45 years, having less exercises and < 400 g/daily grain intake were more susceptible to the adverse effects of ambient PM2.5 exposure.
Although previous studies and the current study were conducted in different geographical areas, with differences in the population characteristics, pollutant concentrations or sources, exposure duration and exposure measurement, it is worth mentioning that positive associations of long-term ambient PM2.5 pollution exposure with MetS remained consistent and that the magnitudes of the effect estimates observed in these studies were comparable. The normative aging study in New York [17] and a cross-sectional study in China [27] found that 10 μg/m3 increase in ambient PM2.5 was associated with a 10% to 31% higher risk of MetS among children, adolescents and elderly population. A nationwide population-based cohort study in Korea showed that each 10 µg/m3 increase in one-year averaged concentration of PM2.5 was associated with a 7% higher risk of MetS in adults [25]. Likewise, the Chinese health study found that each 10 μg/m3 increase in the long-term exposure to PM2.5 was associated with 5% higher risk of MetS in 15,477 adults from 33 communities in northeast China [28]. We have detected the largest magnitude of effect estimates of the association between PM2.5 and MetS in adults. Compared with other heavy industry cities in northeast China, higher risk of total, cardiovascular and respiratory mortality was found in Guangzhou, where the concentration of PM was relatively low [31]. The relatively high concentration of the toxic components (e.g. PBDEs) in PM2.5 detected in southern China [32, 33] might help explain the paradoxically larger effect estimates of the association between PM and total/cardiovascular/respiratory disease mortality and MetS, in the scenario of the lower concentration of PM in Guangdong.
Regarding the complexity of metabolic alterations that constitute MetS, many studies have investigated the association between long- and short-term exposure of PM2.5 and its specific components [15, 17, 21, 42–45]. Several population-based studies have reported harmful effects of ambient PM2.5 on FBG, yet the results were inconsistent. Though Alderete et al. did not identify a statistically significant association between long-term exposure to PM2.5 and FBG in Los Angeles Latino children [21], several other studies investigating the harmful effects of PM2.5 on FBG has supported our findings in different population [15, 42, 43]. The Normative Aging Study found that exposure to high levels of PM2.5 within 28 days was associated with an increased level of FBG [43]. A cross-sectional study revealed a positive association between exposure to PM2.5 and increased FBG among primary school children in China [15]. Few studies have investigated the relationship between PM2.5 and high TG. We are awared of only three studies which were conducted in specific populations or yielded different results from this study. Similar to the results from 587 elderly individuals in the US [17] and 73,117 subjects with known CVDs and risk factors in southern Israel [44], we have identified the adverse impact of PM2.5 on TG. However, none of the significant association was found in the population-based cross-sectional study conducted in northeast China [45]. Similar to the results of Wallwork RS et al. [17], we did not reveal a significant association between PM2.5 and abdominal obesity, low HDL-c and hypertension, which are the essential components of MetS that are often presented as the underlying and/or preceding other components [46] and cardiovascular events [47, 48]. PM2.5 might activate the metabolic mechanisms such as inflammation, which might increase the risk of developing elevated FBG and hypertriglyceridemia without substantially increasing the risk of abdominal obesity, low HDL-c or hypertension.
As seen in other air pollutant studies, the health effects shown in our study were relatively small. However, regarding the broad extent of the exposed population and the continuous nature of exposure, health implications of ambient PM2.5 exposures should be considered at the population level rather than at the individual level [49, 50]. Metabolic risk factors have long been hypothesized as the mediators between air pollutants and CVDs [45, 51, 52]. A previous study showed that participants with an existing metabolic risk factor had a higher risk of CVDs than those without [45]. The results of high TG and high FBG attributed to PM2.5 based on our analyses may help provide the evidence to support these hypotheses. In addition, MetS, high FBG and TG can be translated into adverse health outcomes of CVDs and diabetes mellitus [4, 5]. Participants with type 2 diabetes and hypertriglyceridemia may be more susceptible to the cardiovascular effects of PM2.5 than those without cardiometabolic risk factors. Small differences in the glucose/TG control within the normal range could be translated into the clinically meaningful variation in CVDs and diabetes mellitus risk [53]. These metabolic associations may represent the intermediate factors that help explain the detrimental effect of increased exposure to PM2.5 on CVDs and diabetes mellitus morbidity and mortality. Nevertheless, our findings were not unexpected because air pollution exposure and metabolic risk factors have been closely associated with the heightened inflammatory responses, which is implicated in the development of CVD [52]. Thus, participants with high TG and high FBG might be more susceptible to the detrimental effects of PM2.5, which could help interpret a higher CVD prevalence.
There were limitations regarding the study design and data interpretation. The causality between ambient PM2.5 exposure and MetS and its components cannot be confirmed owning to the cross-sectional study design. Second, data on the secondary MetS diseases were also not fully collected. Although we have excluded participants with CVDs, other diseases including hyperlipidemia and renal hypertension were not available, which might have influenced on the results. Third, the information on multiple food intake was limited regarding the importance of such variable on the etiology of MetS. Furthermore, there could be interactions between PM2.5 and multiple indoor air pollutants (e.g., mold, household fuels, allergens, tobacco smoke, cooking, furniture, paints, cleaning agents) [54], which cannot be readily disentangled.
However, our findings remain robust. We conducted the LUR model to determine PM2.5 exposure at a specific address to safeguard the accuracy of the exposure assessment. Additionally, our association analyses were based on multiple models, with the results not being materially altered. Because the long-term health risk of TG and FBG may be important predictors for future risks of CVDs and diabetes mellitus, efforts should be endeavored to minimize the concentration and exposure to PM2.5 pollution.
Conclusion
In conclusion, this study adds to the comprehensive evidence of the association between long-term exposure to PM2.5 and MetS. Dyslipidemia especially high triglyceride and FBG impairment is strongly associated with PM2.5 levels. However, further prospective studies are needed to confirm our findings.
Supplementary Information
Additional file 1.