*3.1. Vascular Parameters*

Brachial FMD and carotid IMT were normally distributed. Brachial FMD was significantly lower (7.5 ± 18, 95% CI 7.3–7.7%, *p* < 0.001), but carotid IMT was significantly greater (0.68 ± 0.13, 95% CI 0.67–0.69 mm, *p* < 0.0001) in Northern Chinese, compared with their Southern counterparts (8.1 ± 3.0, 95% CI 7.8–8.5% and 0.57 ± 0.13, 95% CI 0.56–0.58 mm, respectively) (Table 2). The GTN responses of the two groups were similar.

**Table 2.** Vascular Parameters in Northern–Southern China Locations.


Compared with Southern China † *p* < 0.0001; †† *p* = 0.01. FMD: Flow-mediated Dilation. GTN: Glyceryltrinitrate Dilation. IMT: Intima-media Thickness.

#### *3.2. Determinants of Risk Factors for Impaired Brachial FMD*

On multivariate regression analyses, in Southern Chinese, brachial FMD was inversely related to PM2.5 (beta = −0.274, *p* = 0.001), age (beta= −0.238, *p* < 0.005), but not to gender, smoking status, BMI, MS, homocysteine, LDL-C or MTHFR. (Model R<sup>2</sup> = 0.202, F = 4.026, *p* < 0.0001) (Table 3). In Northern Chinese, brachial FMD was related to gender (beta = −0.329, *p* = 0.009), but not to other traditional risk factors (Model R<sup>2</sup> = 0.211, F = 3.802, *p* < 0.0001). In the overall 1323 Chinese cohort, lower brachial FMD was related to older age, male gender and northern location, but not to PM2.5 (Model R2 = 0.190, F = 7.802, *p* < 0.0001).



\* Model R2 = 0.202; F-value = 4.026; *p* < 0.0001. \*\* Model R2 = 0.211; F-value = 3.802; *p* < 0.0001. \*\*\* Model R2 = 0.190; F-value = 7.802; *p* < 0.0001. BMI: Body Mass Index. FMD: Flow-mediated Dilation. LDL-C: Low Density Lipoprotein Cholesterol. MTHFR: Methylenetetrahydrofolate Reductase Gene Polymorphisms. PM2.5: Particulate Matters < 2.5 um in Diameter.

> On multivariate regression analyses, carotid IMT in Southern Chinese was significantly related to PM2.5 (beta = 0.334, *p* < 0.0001), independent of age (beta = 0.393, *p* < 0.0001) and gender (beta = 0.146, *p* = 0.043) (Model R<sup>2</sup> = 0.451, F = 13.3, *p* < 0.0001) (Table 4). In Northern Chinese, carotid IMT was related to age (beta = 0.385, *p* < 0.0001), smoking status (beta = 0.157, *p* = 0.01), MS (beta = 0.110, *p* = 0.039), homocysteine (beta = 0.137, *p* = 0.014) and LDL-C (beta = 0.145, *p* = 0.0003), but not to PM2.5 (beta = 0.033, *p* = 0.471). For the overall cohort, carotid IMT was related to PM2.5 (beta = 0.368, *p* < 0.0001), independent

of other atherosclerotic risk factors including age, male gender, BMI, MS, HC, LDL-C and northern location (beta = 0.206, *p* = 0.002) (Model R2 = 0.362, F = 27.1, *p* < 0.0001). No PM2.5 and age (beta = 0.305, *p* = 0.125), or PM2.5 and location (beta = 0.093, *p* = 0.056) interactions were identified.


**Table 4.** Determinants of Risk Factors for Carotid IMT.

\* Model R<sup>2</sup> = 0.451; F-value = 13.3; *p* < 0.0001. \*\* Model R<sup>2</sup> = 0.335; F-value = 7.67; *p* < 0.0001. \*\*\* Model R<sup>2</sup> = 0.362; F-value = 27.1; *p* < 0.0001. BMI: Body Mass Index. IMT: Intima-media thickness. LDL-C: Low Density Lipoprotein Cholesterol. MS: Metabolic Syndrome. MTHFR: Methylenetetrahydrofolate Reductase Gene Polymorphisms. PM2.5: Particulate Matters < 2.5 um in Diameter.
