Effects of ambient particulate exposure on blood lipid levels in hypertension inpatients

Background With modernization development, multiple studies of atmospheric particulate matter exposure conducted in China have confirmed adverse cardiovascular health effects. However, there are few studies on the effect of particulate matter on blood lipid levels in patients with cardiovascular disease, especially in southern China. The purpose of this study was to investigate the association between short- and long-term exposure to ambient particulate matter and the levels of blood lipid markers in hypertension inpatients in Ganzhou, China. Methods Data on admission lipid index testing for hypertension inpatients which were divided into those with and without arteriosclerosis disease were extracted from the hospital's big data center from January 1, 2016 to December 31, 2020, and air pollution and meteorology data were acquired from the China urban air quality real time release platform from January 1, 2015 to December 31, 2020 and climatic data center from January 1, 2016 to December 31, 2020, with data integrated according to patient admission dates. A semi-parametric generalized additive model (GAM) was established to calculate the association between ambient particulate matter and blood lipid markers in hypertension inpatients with different exposure time in 1 year. Results Long-term exposure to particulate matter was associated with increased Lp(a) in three kinds of people, and with increased TC and decreased HDL-C in total hypertension and hypertension with arteriosclerosis. But particulate matter was associated with increased HDL-C for hypertension inpatients without arteriosclerosis, at the time of exposure in the present study. It is speculated that hypertension inpatients without arteriosclerosis has better statement than hypertension inpatients with arteriosclerosis on human lipid metabolism. Conclusion Long-term exposure to ambient particulate matter is associated with adverse lipid profile changes in hypertension inpatients, especially those with arteriosclerosis. Ambient particulate matter may increase the risk of arteriosclerotic events in hypertensive patients.

Background: With modernization development, multiple studies of atmospheric particulate matter exposure conducted in China have confirmed adverse cardiovascular health e ects. However, there are few studies on the e ect of particulate matter on blood lipid levels in patients with cardiovascular disease, especially in southern China. The purpose of this study was to investigate the association between short-and long-term exposure to ambient particulate matter and the levels of blood lipid markers in hypertension inpatients in Ganzhou, China.
Methods: Data on admission lipid index testing for hypertension inpatients which were divided into those with and without arteriosclerosis disease were extracted from the hospital's big data center from January , to December , , and air pollution and meteorology data were acquired from the China urban air quality real time release platform from January , to December , and climatic data center from January , to December , , with data integrated according to patient admission dates. A semi-parametric generalized additive model (GAM) was established to calculate the association between ambient particulate matter and blood lipid markers in hypertension inpatients with di erent exposure time in year.
Results: Long-term exposure to particulate matter was associated with increased Lp(a) in three kinds of people, and with increased TC and decreased HDL-C in total hypertension and hypertension with arteriosclerosis. But particulate matter was associated with increased HDL-C for hypertension inpatients without arteriosclerosis, at the time of exposure in the present study. It is speculated that hypertension inpatients without arteriosclerosis has better statement than hypertension inpatients with arteriosclerosis on human lipid metabolism.
Conclusion: Long-term exposure to ambient particulate matter is associated with adverse lipid profile changes in hypertension inpatients, especially those with arteriosclerosis. Ambient particulate matter may increase the risk of arteriosclerotic events in hypertensive patients. KEYWORDS ambient particulate matter, blood lipid markers, hypertension, arteriosclerosis, generalized additive model (GAM) . Introduction Air pollution has become an environmental risk factor that seriously affects health, bringing a huge economic burden of disease and health loss to society (1,2). According to a Comment by the World Health Organization (WHO), the number of premature deaths caused by air pollution has exceeded 7 million each year (3). Among air pollutants, ambient Most studies have found that ambient particulate matter exposure can increase the incidence of cardiovascular disease events (5, 6), among which hypertension as a risk factor for most cardiovascular diseases has been confirmed by most studies to be related to the health hazard effects of ambient particulate matter exposure (7,8). Blood lipids are the general term for neutral fats (triglycerides) and lipids (phospholipids, glycolipids, sterols, steroids) in plasma (9). Adverse alterations in lipid levels in humans can cause deposition of lipids in the vessel wall, which in turn triggers vascular stiffening they are also recognized as a risk factor for cardioand cerebrovascular disease (10)(11)(12), and play an important role in chronic diseases such as hypertension, myocardial infarction, and ischemic stroke (13)(14)(15)(16). It has been found that long-term exposure to ambient particulate matter can change the lipid level in the population (17, 18), but there is no consistency about the lipid effects of air pollution described by different studies (19,20).
Ganzhou City is located in the southern of China. It has a typical humid subtropical monsoon climate. The precipitation is concentrated in spring and summer, the climate is mild, the heat and rainfall are abundant, the duration of cold and hot air flow is short, and the frost-free period is long. In recent years, Ganzhou's rare earth mining industry, tourism, furniture construction and other industries have developed rapidly, driving the steady growth of the city's economy. With the influx of population, the continuous increase in traffic flow, and the urbanization, the urban vitality of Ganzhou has been greatly improved, but it has also increased pollution emissions, posing challenges to air governance. As the economy develops, there are no studies on associations between air pollution exposure and blood lipids in Ganzhou. Therefore, this study collected ambient particulate matter exposure concentrations in Ganzhou City and blood lipid detection data of hypertensive inpatients in a tertiary hospital, and explored the impacts of ambient particulate matter pollution on blood lipid levels in hypertensive populations.

. . Blood lipid markers in hypertensive hospitalized patients
The hospitalization records of cardiovascular medicine from January 1, 2016 to December 31, 2020 were extracted from the big data center of a tertiary hospital (two campuses) in Ganzhou City, including the patient's admission date, main diagnosis, age, gender, and biochemical tests. According to the 10th revised International Classification of Diseases (ICD-10), whose code is Hypertension (ICD-10: I10-I13), we screened the information of patients with "hypertension" as their discharge diagnosis and matched them with data of blood lipid markers collected during hospitalization. Included lipid markers included triglycerides (TG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), and lipoprotein [Lp(a)]. Since LP(a) is skewed distribution data, we logarithmically transform LP(a). The hypertensive inpatients were divided into "hypertension with arteriosclerosis population" and "hypertension without arteriosclerosis population" according to whether other discharge diagnoses included arteriosclerosis. The sum of the two populations was called "total hypertension population." . . Air pollution and meteorology data of Ganzhou City

. . Statistical analysis
SPSS was used to describe the mean ± standard deviation (Mean ± SD) of blood lipid indexes in three groups of "hypertension with arteriosclerosis, " "hypertension without arteriosclerosis" and "total hypertension." Age, gender and hospitalization date were used as covariates for covariance analysis to compare the differences in the mean concentrations of blood lipid indexes among the three groups. If differences were statistically significant, Bonferroni's method was used for multiple comparisons (21). The central and discrete trends of air pollution and meteorological data were described, and the differences in air pollution concentrations between adjacent years were compared using the Kruskal-Wallis test (22). Spearman correlation analysis was used to evaluate the associations between air pollutants and meteorological data (23).
R4.0.4 was used to establish a semi-parametric generalized additive model (GAM) (24,25) to adjust the temperature, atmospheric relative humidity, age and gender of patients and balance the time trend to calculate the effect of changes in the unit concentration of air pollution on blood lipids in the study population. On this basis, choose the lag time period in which the ambient particulate matter has the strongest correlation with the blood lipid concentration of the population in the lag of 1 year to establish a dual-pollutant model, and observe the impacts of ambient particulate matter exposure on the blood lipid concentration of the three groups of people after adjusting for the concentration of gaseous pollutants. The partial regression coefficient and 95% confidence interval of the effect of ambient particulate matter on the concentration of blood lipid markers were obtained, and converted into the change value of blood lipid concentration by exp[ (β×10)-1] × 100. That is, the corresponding change in blood lipid concentration for every 10 µg/m 3 increase in PM 2.5 or PM 10 concentration, and its 95% confidence interval.

. . Characteristics of lipid markers in study subjects
Among the hypertensive hospitalized patients, 2,269 cases of TC and TG were matched respectively, among which 685 cases had arteriosclerosis, 2,268 cases of HDL-C had 685 cases of arteriosclerosis, 2,266 cases of LDL-C had 685 cases of arteriosclerosis, 2,241 cases of Lp(a) of which 674 had arteriosclerosis. As shown in Table 1, the mean values of the five lipid markers in the total hypertensive population were: TG 1.74 mmol/l, TC 4.45 mmol/l, HDL-C 1.1 mmol/l, LDL-C 2.69 mmol/l, and LP (a) 1.81 µmol/L. There were no significant differences in lipid concentrations among the three groups by covariance analysis.  As shown in Table 3, positive correlations were observed among all air pollutants except O 3 , for which ambient particulate matter had stronger correlations with NO 2 and SO 2 . There was a negative correlation between both air temperature and humidity and with ambient particulate matter, but positive correlations between air temperature and O 3 or SO 2. as well as humidity and CO were observed.

. . Blood lipids e ects on hypertension inpatients with arteriosclerosis
As shown in Figure 3, it was observed in the hypertensive population with arteriosclerosis that PM 2.5 and PM 10   . . Blood lipids e ects on hypertension inpatients without arteriosclerosis TC and TG in hypertension inpatients without arteriosclerosis were not associated with ambient particulate matter exposure during the study period. As show in Figure 4, HDL-C and LDL-C were both significantly positively correlated with ambient particulate matter at lag0-6, and HDL-C showed a positive correlation with ambient particulate matter again at lag0-89 to lag0-209. During the study period, PM 2.5 was positively correlated with Lp(a) concentration in hypertensive population without arteriosclerosis only at lag 0-359. Lp

. . Blood lipids e ects after adjusting for gaseous pollutants
The positive correlation between ambient particulate matter and TC in the population with total hypertension and hypertension with arteriosclerosis remained unchanged after adjusting for gaseous pollutants, but it disappeared significantly after adjusting for NO 2 concentration. The positive correlation between particulate matter and Lp(a) of the three groups of people remained unchanged after adjusting for the effects of gaseous pollutants, but the significant effects of particulate matter disappeared after adjusting for SO 2 , O 3 and NO 2 . In the total hypertensive population, ambient particulate matter was negatively correlated with HDL-C, and the correlation changed positively after CO adjustment. The correlation between PM 10 and HDL-C changed from positive to negative after adjusting for NO 2 and SO 2 in the hypertensive population without arteriosclerosis. The positive correlation between ambient particulate matter and LDL-C levels in hypertensive patients without arteriosclerosis was reversed after adjusting for CO and NO 2 , but there was not significant. The detailed results are shown in Figure 6. These results  revealed that long-term exposure to ambient particulate matter is associated with adverse lipid profile changes in hypertension inpatients, especially those with arteriosclerosis

. Discussion
In recent years, Ganzhou has stepped up its air pollution control actions. Since 2017, the concentration of ambient particulate matter has decreased significantly year by year. Ganzhou City belongs to southern China and has not established a unified heating system. However, in the cold winter, residents will generally have individual combustion or energy-consuming heating behaviors, resulting in an increase in the emission of ambient particulate matter and a negative correlation between the concentration of particulate matter and meteorological factors. In addition, the low temperatures in winter may cause poor diffusion of pollutants in the air, leading to widespread peaks of pollutants in the cold season (26). Similar to most southern cities, the main sources of ambient particulate matter emissions in Ganzhou are construction dust and vehicle exhaust from urbanization. From 2015 to 2020, the average daily concentration of PM 10 in Ganzhou was 59.89 µg/m 3 , which was higher than the first-level standard of the national ambient air quality standard (GB3095-2012), but lower than the second-level standard (the firstlevel standard was 40 µg/m 3 , and the second-level standard was 70 µg/m 3 ). The average daily concentration of PM 2.5 was 38.17 µg/m 3 , which was higher than the NAQS secondary standard (15 µg/m 3 for the primary standard and 35 µg/m 3 for the secondary standard).   There are these lipid transfer proteins which are HDL-C, LDL-C, and very low-density lipoprotein cholesterol. High density lipoprotein (HDL) can uptake low-density lipoprotein, cholesterol, triglycerides, sedimented from the intimal lining of the vessel wall for excretion to the liver, which helps to resist the stiffening of blood vessels caused by hyperlipidemia (27) and maintain cardiovascular health (28)(29)(30). Low density lipoprotein (LDL) is a lipoprotein particle that carries cholesterol into tissue cells and carries cholesterol accumulation across the arterial wall to cause arteriosclerosis (31). Lipoprotein [Lp (a)] is synthesized by the liver and is a specialized cholesterol rich macromolecular lipoprotein that promotes atherosclerosis (32).
This study found differences in the effects of ambient particulate matter on TC, HDL-C, LDL-C, and Lp (a) content among hypertension with or without arteriosclerosis. Combined with the analysis of biochemical associations among lipid markers to identify differences in the associations of each lipid marker with ambient particulate matter among different populations, we raise several conjectures.
Ambient particulate matter exposure exhibited health hazard effects of increasing TC levels and decreasing HDL-C levels for both the hypertensive with arteriosclerosis population and the total hypertensive population. However, in the hypertensive without arteriosclerosis population, there are no association with TC levels was present but increasing HDL-C levels emerged with ambient particulate matter exposure. There was a similar trend for Lp (a) levels to increase with higher ambient particulate matter concentrations in the three groups. A positive association with short-term particulate matter exposure was observed only in people with hypertension without arteriosclerosis. It is speculated that ambient particulate matter exposure mainly cause increased cholesterol and lipoprotein synthesis and decreased high-density lipoprotein synthesis in hypertensive patients. But in patients without arteriosclerosis, the regulatory mechanisms of lipid homeostasis are normal, so that lipid transfer proteins such as HDL-C and LDL-C were temporarily increases to maintain stable total cholesterol levels. When cholesterol synthesis continues increasing with long-term exposure, beneficial HDL-C is more producted to metabolize excess cholesterol. Meanwhile LDL-C is inhibited temporarily to protect blood vessels. The phenomenon of compensatory increases in HDL-C and LDL-C with ambient particulate matter exposure has thus emerged in hypertensive population without vascular stiffening.
Long-term exposure to air pollution was found to be associated with increased TC and LP (a) levels in the elderly, obese adolescents and cardiovascular disease (CVD) patients in other studies on associations between air pollution and lipid concentrations (33)(34)(35), and additional reports showed that lower HDL-C levels were associated with particulate matter exposure (19,34,36). Two studies of the association between air pollution and blood lipids in adults in China both showed that particulate matter had a deleterious effect on blood lipid markers, and this effect was more pronounced in people who were overweight or obese (37). These studies are similar to the results of ambient particulate matter and lipid-related changes that we observed in hypertensive patients with arteriosclerosis and total hypertension. Particulate matter exposure may have a stronger effect on raising TC and lowering HDL-C in individuals already at high risk, a study of older U.S. men suggests (38). This corresponds to our findings and conjectures.
The cardiovascular harm of ambient particulate matter exposure has been widely confirmed, and it has been proposed that oxidative stress (39,40), inflammatory response (41,42), and DNA methylation (43,44) are the main damage mechanisms. Dyslipidemia is the most sensitive metabolic risk factor to air pollution exposure (45). Ambient particulate matter can cause systemic inflammation and oxidative stress, inducing adverse lipid metabolism and oxidation (46). The inflammatory response further mediates a lipid compensatory response against the invasion of hazardous substances from ambient particulate matter and aids tissue repair. When the repair of injury is saturated by lipid compensatory responses, cascading repetitive stimulation of the inflammatory response enhances atherosclerotic lesion formation (47). Additionally ambient particulate matter can also lead to specific gene methylation related to lipid metabolism by reducing the activity of DNA methyltransferases (48). The above studies exhibited different mechanistic pathways by which ambient particulate matter altered blood lipids, which could directly provoke adverse lipid metabolism and also provoke lipid compensatory responses in the pre hazard period, corresponding to the differences in relevant lipid changes among different populations in this study and the speculation raised previously.
. /fpubh. .  exposure were more significant than those of PM 2.5 in the dualpollutant model. Therefore, we believe that compared with gaseous pollutants, ambient particulate matter exposure is more strongly associated with increased TC, Lp(a) and decreased HDL-C, and the harmful effect of PM 10 is greater than that of PM 2.5 . A longitudinal study in Shijiazhuang, China also suggests that ambient particulate . /fpubh. . matter may have a greater impact on lipid health than gaseous pollutants (49). In a hypertensive population with arteriosclerosis, the positive associations between ambient particulate matter exposure and HDL-C and LDL-C were partially reversed after adjustment for gaseous pollutants. Based on the results of the dual-pollutant model, we believe that ambient particulate matter exposure is not the main pollutant affecting the elevated HDL-C and LDL-C concentrations in this population. Combined with the previous conjecture, we suggest that the phenomenon of higher HDL-C associated with particulate matter exposure in people with hypertension without vascular stiffness, on the one hand, may be due to compensatory increases in HDL-C caused by the body's own protective mechanisms against lipid homeostasis. On the other hand, gaseous pollutants, which were positively correlated with each other and with particulate matter, exhibited stronger associations with HDL-C elevation. Based on the above results, we speculate that ambient particulate matter exposure can cause adverse changes in blood lipid levels in hypertensive patients and increase the risk of arteriosclerosis events. This study has several strengths. First, we applied the detailed blood lipid detection data of inpatients collected by the hospital's big data center, and the real-time monitoring data of air pollution concentration collected by the environmental protection department to ensure the accuracy of the data source. We selected hypertensive patients as research subjects and grouped them according to whether they were accompanied by arteriosclerosis, so as to study the correlation between ambient particulate matter exposure and the risk of blood lipids in hypertensive patients with different disease states. Second, the data analysis applied a semiparametric generalized additive model. After adjusting for the effects of gender, age, weather, and time, the correlation between ambient particulate matter pollution and blood lipids in different lag time periods in a year was more accurately explored, and the relationship between each blood lipid index and air pollution exposure was calculated. Third, this study is the first to explore the relationship between air pollution and human blood lipid concentrations in southern Jiangxi, China, and its findings can .
provide reference for other regions with similar development levels and geographic latitudes. This study also has some limitations. First, in addition to environmental factors, blood lipid levels are affected by a variety of factors, including genetics, behavior, and medication (50,51). This study was not able to collect this information, so the results ignore the role of some valuable confounding factors and may be biased. Secondly, the air pollution concentration is the average of the data of the five monitoring points in Ganzhou City, and it is impossible to accurately understand the actual exposure concentration of air pollution of each research object. This exposed misclassification is likely to reduce the significance of the association (52). In addition, our study only analyzed the lipid effects associated with particulate matter exposure during a one-year period, and did not investigate the effect of a longer lag. Significant associations of LP (a) with ambient PM generally appeared at lag 359 days. Looking at the trends of LP (a) changes associated with exposure at different lag times, more significant associations with LP (a) increases may occur at longer exposure times.

. Conclusions
Ambient particulate matter exposure was associated with higher TC, LP (a) and lower HDL-C in hypertensive patients, and PM 10 exposure was more strongly associated with changes in lipid markers than PM 2.5 . There are association of long-term exposure to ambient particulate matter with the risk of arteriosclerosis in hypertensive patients, and such impacts have stronger harmful effects in patients with arteriosclerosis.

Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Ethics statement
The studies involving human participants were reviewed and approved by the Ethics Committee of Gannan Medical University. The patients/participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.

Author contributions
YG: conceptualization, writing-reviewing and editing, and funding acquisition. CL and LH: methodology, software, data curation, formal analysis, writing-original draft, review, and editing. XZho: software, formal analysis, investigation, and data curation. KH and MG: sensitivity analysis. XZha: conceptualization, data acquisition, supervision, and funding acquisition. All authors contributed to the article and approved the submitted version.