*3.3. Blood Pressure at Rest and During the Cold Pressor Test*

Table 3 shows the BP at rest and during the CPT before and after the supplementations. There were no significant between-group differences in the resting BP at baseline and no time-by-group interactions. There were significant time-by-group interactions for changes (Δ) in brachial SBP (*p* = 0.02), brachial MAP (*p* = 0.04), aortic SBP (*p* = 0.03), and aortic MAP (*p* = 0.04) responses to the CPT. CIT+GSH supplementation reduced ΔSBP compared to the placebo and CIT (*p* < 0.05 for both), and reduced ΔMAP compared to the placebo (*p* < 0.05) but not to CIT (Figure 4). No significant time-by-group interactions were observed for ΔDBP, Δ augmentation index normalized to a heart rate of 75, and Δ reflection time (Tr). Attenuation of aortic ΔMAP was significantly related to the improvement in FMD% (*r* = −0.33, *p* < 0.05).

#### **Table 3.** Brachial and aortic blood pressures at rest and during CPT.


Values are the mean ± SD. Abbreviations: AIx, augmentation index normalized to the heart rate of 75; CIT, citrulline; CIT+GSH, CIT+glutathione; DBP, diastolic blood pressure; MAP, mean arterial pressure; SBP, systolic blood pressure; Tr, reflection time. *p*-values of ANOVA time-by-group interaction.

#### *3.4. Serum Biomarkers*

Table 4 shows serum biomarkers before and after the interventions. Significant timeby-group interactions were observed for ARG (*p* = 0.005), ORN (*p* = 0.04), and ARG/ADMA ratio (*p* = 0.007). ARG and ARG/ADMA ratios were significantly increased by CIT supplementation compared with placebo (*p* = 0.01 for both) and CIT+GSH (*p* = 0.008 and *p* = 0.04, respectively) (Figure 5A,D). The ARG/ADMA ratio significantly increased after CIT+GSH (*p* = 0.04). ORN was increased after CIT compared to CIT+GSH (*p* = 0.05) but not to the placebo (*p* = 0.12). Arginase I levels decreased after CIT (*p* = 0.01) and tended (*p* = 0.07) to

decrease after CIT+GSH, but there was no significant time-by-group interaction. Glucose, insulin, homeostatic model assessment for insulin resistance, glutathione peroxidase, superoxide dismutase, oxidized LDL, and malondialdehyde did not significantly change after 4 weeks in any group.

**Figure 4.** Changes (Δ) in (**A**) the brachial systolic blood pressure SBP (SBP), (**B**) brachial mean arterial pressure (MAP), (**C**) aortic SBP, and (**D**) aortic MAP to the cold pressor test before and after supplementations. Abbreviations: CIT, citrulline; CIT+GSH, CIT+glutathione. \* *p* < 0.05 before vs. after, † *p* < 0.05 vs. placebo and CIT, ‡ *p* < 0.05 vs. placebo.



Data are the mean ± SD. *p*-values of ANOVA time-by-group interaction. Abbreviations: ADMA, asymmetric dimethylarginine; HOMA-IR, homeostatic model assessment for insulin resistance; GPx, glutathione peroxidase; SOD, superoxide dismutase; Ox-LDL, oxidized LDL; MDA, malondialdehyde. \* *p* < 0.05 vs. before; ‡ *p* < 0.01 vs. before; † *p* < 0.05 vs. placebo and CIT+GSH; # *p* < 0.05 vs. CIT+GSH.

**Figure 5.** Changes (Δ) in serum levels of arginine (**A**), ornithine (**B**), and the arginine/ADMA ratio (**C**) from 0–4 weeks in the three groups. Values are the mean ± SE. Abbreviations: ARG, arginine; ADMA, asymmetric dimethylarginine; CIT, citrulline; CIT+GSH, CIT+glutathione. \* *p* < 0.05 vs. placebo and CIT+GSH; † *p* < 0.05 vs. CIT+GSH.

#### **4. Discussion**

This study examined the effects of 4 weeks of CIT and CIT+GSH supplementation on vascular function and BP responsiveness in healthy postmenopausal women. We found that CIT+GSH supplementation improved endothelial function through an increase in the ARG/ADMA ratio. Although CIT supplementation increased serum ARG levels and the ARG/ADMA ratio, it did not statistically improve FMD. In addition, CIT+GSH supplementation attenuated BP responsiveness to CPT. These data show that CIT+GSH supplementation has protective cardiovascular effects at rest and during sympathetic stimulation (CPT) in healthy postmenopausal women.

FMD decreases progressively throughout the stages of the menopausal transition in healthy women [4]. Low FMD is a valuable biomarker for predicting future cardiovascular events in apparently healthy adults [39]. In this study, we showed that 4 weeks of CIT+GSH supplementation increased brachial FMD by 2.9% in healthy postmenopausal women. Consistent with our findings, a non-placebo controlled study observed a beneficial effect of CIT supplementation for 4 weeks on FMD in middle-aged patients with vasospastic angina and low FMD [40]. Similarly, 2 weeks of ARG supplementation increased FMD by 3.1% in healthy older men with age-related endothelial dysfunction [21]. A previous meta-analysis suggested that ARG supplementation improves FMD in individuals with baseline FMD values < 7.0% [41]. These previous findings suggest that CIT and ARG improve FMD in individuals with low FMD. Our data suggest that improvements in FMD following CIT+GSH mainly occurred in women with a low baseline FMD. Importantly, we found that the increase in FMD after 4 weeks of CIT+GSH was 0.91 SD greater vs. placebo and 0.67 SD greater vs. CIT alone. Given that for each 1 SD increase in FMD there is an associated 50% lower risk of cardiovascular events [42], our findings are clinically important.

Recent data showed that FMD values lower than 5.4% are indicative of impaired endothelial function in apparently healthy adults [43]. Based on this FMD reference value, 85% of women in the CIT+GSH group had low FMD, and thus, had impaired endothelial function at baseline. Evidence indicates that cardiometabolic risk factors, including BP and FBG, are determinants of low FMD [43]. Independent of traditional cardiovascular risk factors, low FMD predicts CVD risk [44]. In our study, an elevated baseline FBG was observed in the CIT+GSH group. Based on FBG values in the CIT+GSH group, four and seven participants had prediabetes or an increased risk of incident prediabetes, respectively [45]. It is known that prediabetes can negatively impact the age-related decline in FMD [46,47]. We demonstrate that CIT+GSH supplementation increased the mean FMD above 5.4% in a group of postmenopausal women with elevated FBG. Considering that a 4.3% increase in FMD is associated with 50% CVD risk reduction [42], the 2.8%

improvement in FMD following 4 weeks of CIT+GSH may reduce the risk of cardiovascular events in apparently healthy postmenopausal women.

Age associated arterial stiffening affects the aorta to a greater extent than peripheral muscular arteries [48]. In older women, 10 years of aging have a greater impact on aortic PWV (+2.4 m/s) than brachial PWV (+0.19 m/s), indicating that peripheral arteries have less stiffening than the aorta [48]. A widely used measure of arterial stiffness is brachialankle PWV (baPWV), a segment that includes central (aortic PWV) and peripheral (brachial and leg PWV) arteries [49]. Since peripheral arteries have relatively more smooth muscle cells than collagen fibers [48], brachial PWV and leg PWV are more responsive to NOmediated vasodilation than the aorta [50,51]. Oral CIT supplementation (5.6 g daily) for 1 week significantly reduced baPWV by ~0.14 m/s in middle-aged men with high baseline baPWV [27]. A further study clarified that the beneficial effect of CIT (6 g daily) on baPWV was due to a reduction of leg PWV (~0.40 m/s) with no effect on aortic PWV in obese postmenopausal women [29]. In the present study, aortic PWV was not reduced by CIT and CIT+GSH supplementations. This ineffectiveness may be attributed to the normal baseline values of our participants, which are considered to be ~7.6 m/s for adults aged ≥50 years [38]. Importantly, CIT+GSH decreased brachial PWV by ~0.40 m/s in postmenopausal women with increased risk of prediabetes. This finding is in agreement with previous studies [27,29] and confirms that peripheral arteries are more responsive than the aorta to dietary supplementation with NO precursors. The decrease in brachial PWV observed in the current study may be attributed to enhanced endothelial-mediated vasodilation [51].

Impaired FMD and increased sympathetic activity are associated with the increased risk of incident hypertension in healthy postmenopausal women [13,14]. Our participants had a normal or elevated SBP at rest. Our observation that the resting BP was unaffected by CIT and CIT+GSH supplementations is consistent with previous studies in normotensive adults [30,32] and middle-aged adults with elevated SBP and FBG [27]. Thus, we used the CPT as a systemic sympathetic vasoconstrictor stimulus to evaluate the efficacy of dietary supplements for attenuating BP reactivity to CPT [18,19]. We observed reductions of brachial and aortic SBP and MAP responses to CPT after CIT+GSH but not CIT supplementation. Previous evidence of blunted BP responsiveness to CPT after 2 weeks of CIT (6 g/day) was observed in healthy young [31] and older adults [32]. However, we did not measure FMD in those previous studies. In mice, CIT supplementation attenuated cold hypersensitivity by improving endothelial function [52]. In the current study, the attenuation of aortic MAP reactivity to CPT was related to improvements in FMD. This finding suggests that improved FMD with CIT+GSH supplementation may reduce the risk of cardiovascular events related to the augmented aortic BP load during conditions with increased sympathetic stimulation [16,19,53].

ADMA competes with ARG for binding to eNOS, thereby a decrease in ARG/ADMA leads to reduced NO production [54]. Evidence supports that the ARG/ADMA ratio is a better biomarker of endothelial function than circulating ARG and ADMA levels alone [7,55]. Our participants had normal ARG/ADMA values at baseline [56]. However, the CIT+GSH group tended to have a lower ARG/ADMA ratio, which has been associated with hyperglycemia [57]. In the present study, CIT and CIT+GSH supplementations increased the ARG/ADMA ratio, suggesting increased ARG availability for NO production [54]. This finding is consistent with Ochiai et al. [27] who reported an increase in plasma ARG/ADMA due to the isolated increase in ARG levels. CIT supplementation increases circulating ARG via de novo synthesis in the kidneys [54]. Serum ARG introduction to endothelial cells via the cationic amino acid y+ transporter 1 depends on both ARG and ADMA levels, since they compete for the transporter [58]. Therefore, a greater ARG availability displaces ADMA from eNOS binding, thereby improving NO synthesis [7,27,40,54,59]. A recent study reported that 2 g of CIT for 4 weeks increased NOx levels in type 2 diabetes patients as a result of arginase inhibition [23]. These findings suggest that a low dose of CIT may improve endothelial function in individuals with hyperglycemia. In the study by Ochiai et al., the increase in ARG/ADMA was evident after CIT supplementation in men with FBG in the prediabetes category [27]. Therefore, augmented ARG/ADMA ratio may explain the improvement in endothelial function following CIT and ARG supplementations in adults with hyperglycemia [27] and low baseline FMD [21,40,60].

Despite greater increases in the serum ARG levels and ARG/ADMA ratio, FMD did not improve after CIT supplementation alone. This discrepancy in the findings may be explained by the healthy status of the participants. In agreement with our findings, ARG supplementation for 4 weeks increased serum ARG levels by almost double but failed to improve FMD in healthy postmenopausal women [61]. Similarly, 6 g of CIT supplementation efficiently increased circulating ARG levels and ARG/ADMA without affecting FMD in healthy adults [24]. In the present study, the increase in ARG levels positively increased the ARG/ADMA ratio, indicating improved ARG availability for NO production. Similarly, previous studies failed to show improvement in endothelial function assessed as increased NOx after oral ARG and CIT supplementations in healthy adults [25,62]. Thus, increased ARG availability via CIT or ARG supplementation may not improve endothelial function in individuals with normal FMD [24,61]. Of note, CIT is concurrently produced with NO by eNOS, and recycled to de novo ARG [63]. It is possible that CIT supplementation stimulated the production of NO and CIT, but the effect on FMD was not evident in the absence of endothelial dysfunction. Nonetheless, the increase in the ARG/ADMA ratio by CIT may have vascular benefits, since an increase in this ratio by 1 SD may decrease CVD risk by 20% [64].

Oxidative stress markers were not improved by both supplementations. In healthy adults, 200–1000 mg of GSH daily for 4 weeks did not improve microvascular endothelial function and malondialdehyde levels [65,66]. Recently, whole blood GSH was elevated after 1 month with 250 or 1000 mg of GSH supplementation [67]. Despite no improvement in the oxidative stress markers, CIT+GSH increased FMD by improving eNOS function via increased ARG availability [33]. The ability of CIT+GSH to increase NO production was demonstrated in human umbilical vein endothelial cells after exposure for 24 h to the combination but not to CIT and GSH alone [34]. Thus, GSH provides an augmenting effect to the conversion of ARG to NO [34].

There are some limitations in the present study. The sample size was relatively small and included healthy postmenopausal women. FBG was not considered for randomization, which resulted to be higher in the CIT+GSH group. Future studies might investigate the effects of CIT and CIT+GSH in individuals with prediabetes. A GSH dose greater than 200 mg daily for longer than 4 weeks of supplementation may reduce markers of oxidative stress, as previously shown [67]. We measured serum levels of superoxide dismutase and glutathione peroxidase, two intracellular enzymes, rather than intracellular enzymatic activity. A statistical limitation could be a "regression to the mean" phenomenon when examining changes in FMD from baseline. However, there were no significant decreases in FMD in the placebo group after the 4-week intervention, which supports our conclusion of CIT+GSH being a viable route to improve endothelial function in healthy postmenopausal women.
