**3. Results**

The patient characteristics and biochemical parameters of the recruited subjects are summarized in Table 1. Controls and patients were matched for age and gender, and there were no significant di fferences between the distribution of education levels or median values (25% and 75% percentiles) of BMI, triglycerides, or total cholesterol between the groups (Table 1). Increased values of hs-CRP were observed among patients with periodontitis, CAD, and periodontitis + CAD in comparison with healthy subjects (*p* < 0.001). Patients with CAD and periodontitis + CAD had a similar proportion of previous CVD events (atrial fibrillation, angina pectoris, stroke, heart failure) and took more CVD drugs (antihypertensive, statins, low-dose aspirin, beta-blockers). Patients with CAD and periodontitis + CAD presented lower serum retinol, α-carotene, β-carotene, γ-tocopherol, β-cryptoxanthin, lutein, zeaxanthin, and lycopene levels compared to periodontitis and healthy controls (Table 1).

Table 2 shows dental variables in patients with periodontitis, CAD, periodontitis + CAD, and controls. Patients with periodontitis and with periodontitis + CAD presented a lower median number of teeth and higher median values of periodontal parameters (CAL, PD, BOP, PI) compared with CAD and control subjects (*p* < 0.001). Moreover, the median values of periodontal parameters were significantly higher in patients in the periodontitis and periodontitis + CAD groups compared to patients with CAD and healthy controls (*p* < 0.001, Kruskal–Wallis test) (Table 2).



the Mann–Whitney test. CAD, ischemic heart disease; CVD, cardiovascular disease; hs-CRP, C-reactive protein.


**Table 2.** Clinical dental variables of recruited subjects.

Data are expressed as median (25th and 75th percentile). \*\* *p* < 0.001 significant differences vs. control subjects calculated by the Mann–Whitney test. §§ *p* < 0.001 significant differences vs. periodontitis patients calculated by the Mann–Whitney test. ## *p* < 0.001 significant differences vs. CAD patients calculated by the Mann–Whitney test. CAL, clinical attachment level; PD, probing pocket depth; BOP, bleeding on probing; PI, plaque index.

### *vitamin C Evaluation*

Median (25th and 75th percentile) serum and salivary vitamin C levels are presented in Figure 2. The median concentrations of serum and salivary vitamin C were lower in the CAD (*p* < 0.01) and in the periodontitis + CAD (*p* < 0.001) groups compared to controls. Serum and salivary vitamin C concentrations were also significantly decreased in patients of the periodontitis + CAD group in comparison with periodontitis patients (*p* < 0.01; Figure 2). Overall, the *p*-value for trend analysis performed (Jonckheere–Terpstra test) indicated that serum vitamin C progressively decreased in patients with periodontitis, CAD, and periodontitis + CAD (*p* < 0.001; Figure 2).

**Figure 2.** Median values (25% and 75% percentiles) of serum and salivary vitamin C levels in each group of subjects. \* *p* < 0.05, \*\* *p* < 0.01, and \*\*\* *p* < 0.001 significant differences vs. control subjects (derived by Kruskal–Wallis test). §§ *p* < 0.01 significant differences vs. periodontitis patients. *p* < 0.001 (obtained by Jonckheere–Terpstra test).

There was no statistically significant correlation between salivary and serum vitamin C levels (rs = 0.157, *p* = 0.087; Figure 3).

**Figure 3.** Correlation analysis of serum and salivary vitamin C levels in all enrolled subjects.

Moreover, across all subjects, serum/salivary vitamin C concentrations correlated negatively (rs = –0.378, *p* < 0.001)/(rs = −0.427, *p* < 0.001) with hs-CRP levels (Figure 4).

**Figure 4.** Correlation analysis of serum and salivary vitamin C levels with CRP values in all enrolled subjects.

The adjusted multivariate linear regression analysis, aimed at assessing the possible association of periodontitis and CAD on serum and salivary vitamin C levels, showed that hs-CRP (*p* < 0.001) was the only statistically significant predictor variable for serum vitamin C; hs-CRP (*p* < 0.001) and triglycerides (*p* = 0.028) were the statistically significant predictor variables for salivary vitamin C (Table 3).


**Table 3.** Uni- and multivariate linear regression model for serum and salivary vitamin C levels in all enrolled subjects.

Age was included as a continuous variable. For periodontitis and CAD, controls served as reference. For gender, males served as reference. For education, primary school served as a reference.
