*3.2. Determinants of Stroke Recurrence*

Among the investigated factors, the multiple analysis results (Table 3) showed that HTN (odds ratio, OR, 1.93; 95% confidence interval, 95%CI, 1.23, 3.04; *p* = 0.004) and CCI (OR, 1.55; 95%CI; 1.08, 2.23; *p* = 0.017) were associated with a higher likelihood of SR. Whereas PAB was associated with a lower likelihood of SR (OR, 0.87; 95%CI, 0.80, 0.95; *p* = 0.003). Moreover, hemorrhage stroke had a lower likelihood of recurrence compared to infarction stroke (OR, 0.55; 95%CI, 0.34, 0.90; *p* = 0.019).

#### **Table 3.** Determinants of stroke recurrence (*n* = 951).



**Table 3.** *Cont.*

Abbreviation: OR, odds ratio; CI, confidence interval; HL, health literacy; CCI, Charlson comorbidity index; PAB, pro-oxidant–antioxidant balance. \* Results of bivariate logistic regression analysis. \*\* Results of multivariate logistic regression analysis adjusted for age, education, ability to pay for medication, stroke classification, depressive symptoms, and CCI.

#### *3.3. Modification Impact of PAB on the Association between HTN and Stroke Recurrence*

As shown in Table 4, compared to stroke patients without HTN and with the lowest PAB scores, those with HTN had an increased likelihood of SR (OR, 2.06; 95%CI, 1.28, 3.32; *p* = 0.003). However, the likelihood of SR decreased in stroke patients with HTN and with every one-point increment of PAB (OR, 0.83; 95%CI, 0.66, 0.94; *p* = 0.022). Moreover, the results of interaction were visualized in Figure 1. Simple slope analysis showed that the negative impact of HTN on SR was lowered by higher PAB values from −1SD (OR, 2.43; 95%CI, 1.29. 4.57; *p* = 0.005) to the mean (OR, 1.82; 95%CI, 1.16, 2.88; *p* = 0.008) and +1SD (OR, 1.36; 95%CI, 1.13, 2.58; *p* = 0.003).

**Table 4.** Interaction of hypertension and pro-oxidant–antioxidant balance on stroke recurrence (*n* = 951).


Abbreviation: OR, odds ratio; CI, confidence interval; HTN, hypertension; PAB, pro-oxidant–antioxidant balance. \* Results of bivariate logistic regression analysis. \*\* Results of multivariate logistic regression analysis, adjusted for age, education, ability to pay for medication, stroke classification, depressive symptoms, and comorbid conditions.

**Figure 1.** Simple slope plot for the interaction between hypertension and pro-oxidant–antioxidant balance on stroke recurrence (*n* = 951). HTN, hypertension; PAB, pro-oxidant–antioxidant balance; SD, standard deviation.

#### **4. Discussion**

The current findings emphasize the positive impact of PAB on lowering SR, especially regarding the moderating role of PAB in mitigating the negative impact of HNT on SR.

Although OS is involved in the pathological mechanism of stroke, the study of OS in stroke patients is challenging to undertake because of the complexly interrelated processes. Further, OS is caused by numerous exogenous factors, and the combined (either addictive or synergistic) influences of those factors on the OS process should be considered. Therefore, a global estimation of the PAB score was established based on various dietary, lifestyle, and medication components [25]. In the literature, a low PAB score (reflecting excessive OS) was associated with poor health outcomes (such as chronic kidney diseases and allcause mortality) [25], but little is known about PAB score and SR. Previous studies directly estimated the plasma PAB concentration and found heterogeneity associations between PAB and SR. Although the redox unbalances were greater in patients within one month of stroke onset than in their controls [52], the serum PAB concentration at the acute stroke phase was not associated with SR within five years [53]. To our knowledge, the current study was the first to find that a higher health behavior-based PAB score reflecting antioxidant properties could reduce the likelihood of SR.

The relationship between OS and inflammation is a vicious circle that links to HTN [54]. In stroke patients, persistent inflammatory responses are activated in those with HTN, which leads to an additional chain of OS procedures, and the HTN-OS correlation contributes to increasing SR [55]. That is why eliminating the free radical agents in the OS processes via "scavengers" is able to decrease the adverse consequences of stroke [55]. Previous studies showed that although the association between the antioxidative and anti-inflammatory capacities of a single dietary component with the hypotensive impact was not clear, adding such a component may become a new method for controlling the balance in OS–inflammation crosstalk [56]. Our findings supported the rationale that a greater PAB score could minimize the deleterious influence of HTN on SR probability. In the current study, the PAB score was estimated based on diet quality, fruit and vegetable intake, and physical activity in regard to anti-oxidative capacity. Therefore, the interplay of those components should be taken into consideration for the balanced regulation of OS and inflammation, as well as for preventing SR.

In recent decades, dietary nutrients have been used a novel intervention for building cardiovascular and neurological health due to their antioxidant properties [57]. However, the mechanism regarding the relationship between diet nutrients and neuro-cardiological diseases is not yet well defined, that metabolic control (including lipid, ROS, and energy metabolisms), immunological regulation (including astrocyte and microglial activation), and epigenetic modification (including non-coding RNA, DNA methylation, and histone modification) were assumed to be involved. Regarding dietary nutrients, the intervention programs focused on two forms of diet, including dietary supplements (such as vitamins and plant extracts) and dietary restriction (such as the DASH diet with decreasing cholesterol, total fat, and saturated fat and increasing fibers and minerals). Moreover, fruits and vegetables are rich sources of vitamins and plant extracts containing polyphenolic components, which are potent antioxidants and are reported to be beneficial to cardiovascular health [58]. However, the excessive intake of antioxidative dietary nutrients (e.g., iron and vitamin A) could lead to the side effects, such as cell death, because of ROS overproduction, causing OS. Additionally, there are anti-nutrient compounds contained in fruits and vegetables (such as tannins, oxalates, and lectins), which may threaten health [59]. Therefore, the approach of PAB needs to consider both the individual pro-oxidant and antioxidant components and their mutual effects.

The present study is the first to assess the potential role of PAB score in altering the relationship between HTN and SR. However, several limitations should be noted. First, the associations were recognized, but the causal inference could not be established in a cross-sectional study. Second, as the study participants were recruited in stable conditions of stroke, the findings could apply only to mild and moderate stroke patients but not to those with severe stroke conditions. However, the sample size was satisfactory for representativeness, with a power of evidence of up to 90%. Third, the questionnaires were responded to via self-reporting, and the "standard serving" size of food was not defined, which may lead to underestimation.
