The present study demonstrated that plasma leptin levels are positively related to vascular endothelial function in overweight T2D patients, but not in lean patients. Of importance, plasma leptin level was a significant contributing factor to FMD, independent of BMI, blood pressure and other traditional cardiovascular risk factors, in overweight T2D patients. Our findings suggest that leptin exerts its positive and vasodilator effect on endothelial function in overweight diabetic patients with an elevated risk for cardiovascular diseases.
A number of clinical studies have investigated the association between plasma leptin levels and vascular endothelial function [16, 29–31]. Plasma leptin levels were reported to be negatively associated with EDV measured by forearm plethysmography in the elderly subjects , and with FMD of the brachial artery in patients with nonalcoholic fatty disease  and polycystic ovarian syndrome . However, the association of plasma leptin with vascular endothelial function was not independent of the other variables including BMI in these studies. Plasma leptin-to adiponectin ratio was also shown to be negatively associated with FMD in healthy elderly subjects . In the other studies, plasma leptin levels were not associated with FMD of the brachial artery [32–37]. FMD was evaluated after weight reduction by low-calorie diet for obese subjects in several studies [38–41]. Increased FMD after weight loss was negatively , positively , or not [40, 41] associated with change in plasma leptin levels. These previous cross-sectional and interventional studies demonstrate inconsistent results on the relationship between leptin and endothelial function, possibly because the study subjects or the method used to estimate endothelial function differs. The findings of our study contrast with those of these previous studies in that the association of leptin levels with FMD was independent of confounding cardiovascular risk factors such as age, BMI, SBP, and lipids in overweight T2D patients. In addition, moderately elevated plasma leptin levels unexpectedly exhibited a positive relation with FMD in overweight patients with T2D in this study. These patients are also generally at a high risk for atherosclerosis and cardiovascular disease.
Functional leptin receptors are expressed in vascular endothelium . Several experimental studies showed that acutely administered leptin induces endothelium-dependent vasorelaxation by stimulating the release of endothelial NO or the endothelium-derived hyperpolarizing factor (EDHF) in rats [8, 9, 42], or by neuronal NO synthase in mice . Endothelium-independent vasodilation by leptin was also identified in the saphenous vein and internal mammary artery ex vivo in humans with coronary artery disease . A direct vasodilator effect of acute leptin infusion has also been investigated in several human studies. Nakagawa et al.  reported forearm vasodilatation by intra-arterial infusion of leptin, and coronary artery vasodilation by intra-coronary leptin infusion in a separate study . The leptin-induced vasodilatation was independent of NO and possible involvement of other vasoactive agents such as the EDHF or prostacyclin was suggested in those human studies [10, 11]. Brook et al.  demonstrated that brachial FMD increased 2 hours following subcutaneous injection of recombinant human leptin without altering blood pressure in non-obese adults. These in vitro and in vivo data indicate a direct vasodilator effect of leptin on the endothelium through endothelial NO or other factors, and this could support our finding, at least in part, that an independent and positive association exists between plasma leptin levels and FMD in overweight subjects.
On the other hand, a number of studies have shown that leptin regulates the sympathetic nervous system, endothelin-1 production, and renin-angiotensin system [4, 46], all of which contribute to vasoconstriction and may counteract the depressor effect of leptin on vascular function. Moreover, a large number of studies have indicated that leptin regulates immune function and cytokine secretion, upregulates C-reactive protein production, and increases oxidative stress in endothelial cells, all of which promote the pathophysiology of atherogenesis including endothelial dysfunction [1, 4, 47, 48]. Indeed, obesity or long-term hyperleptinemia was shown to reduce NO bioavailability of the aortic endothelium mice  and rats , and to attenuate NO-dependent vasodilation of the coronary artery in dogs . Schinzari et al.  demonstrated in human study that leptin infusion enhanced EDV in lean subjects, but not in patients with obesity-related MetS. High leptin concentrations were also reported to be associated with impaired EDV measured by forearm plethysmography  and inversely with adenosine-stimulated coronary blood flow  in human subjects, but those associations were not independent of BMI and/or insulin levels. These experimental and clinical studies indicate that the NO-dependent vasodilatory effects of leptin become impaired, and by this mechanism, leptin may contribute to endothelial dysfunction and the progression of atherosclerosis in patients with obesity and/or MetS. These studies imply that the selective leptin resistance seen in obesity may not be limited to appetite and body weight control, but may involve the hemodynamic actions of leptin, thus leading to the pro-atherogenic effects of leptin on vascularture . However, to date, no previous study has demonstrated an independent association of leptin with brachial artery FMD assessed by ultrasound, in human subjects with obesity and/or MetS.
To our knowledge, this is the first study to explore the association of leptin with vascular endothelial function in patients with T2D. We found an independent and positive association between plasma leptin levels and FMD in overweight (BMI ≥ 25 kg/m2), but not lean subjects, even after adjustment for other confounders including age, BMI, SBP, HbA1c, and lipid levels. The positive association between plasma leptin levels and FMD in patients with diabetes is contrary to previous studies that show an association between hyperleptinemia and impaired endothelial function in patients with obesity and/or MetS. There are several possible explanations for this discrepancy. First, the degree of obesity in our Japanese patients was much less than that of the studies performed in European countries, which demonstrated a resistance to leptin-induced vasoreactivity in human subjects with obesity or MetS [15, 49]. BMI and plasma leptin levels were 33.6 kg/m2 and 10.3 ng/mL, respectively, in obese subjects from the study by Sundell and colleagues , and 38 kg/m2 and 21.2 ng/mL, respectively, in subjects with MetS in the study by Schinzari and colleagues . These BMI and leptin levels are much higher than those of our overweight subjects (BMI, 28.4 kg/m2; leptin levels, 6.0 ng/mL). Moreover, the plasma leptin levels of our overweight subjects were similar to those of the healthy  and control subjects , at 4.3 ng/mL and 8.7 ng/mL, respectively, in two studies from Europe. Therefore, the vasodilator effect of leptin could have still been activated in our overweight subjects because their leptin levels were not very hyperleptinemic. In addition, plasma leptin did not exhibit a significant association with FMD in our lean subjects. This could be due to their low leptin levels (2.5 ng/mL), at which leptin has not been found to exert significant vasodilation in Japanese subjects . Second, forearm plethysmography reflects endothelial function of the resistance artery mediated mainly by the EDHF, whereas FMD reflects the conduit artery by NO . An animal study showed that even if the endothelial NO synthase-derived NO production is impaired or absent, leptin can induce neuronal NO synthase in the endothelium to maintain endothelium dependent-vasorelaxation in a mouse model of obesity with hyperleptineima or angiotensin-II-induced vascular dysfunction . Therefore, in this study, NO-mediated vasodilation assessed with FMD of the brachial artery was observed in overweight diabetic subjects with mildly elevated plasma leptin levels. Third, overweight subjects in this study were significantly younger than the lean subjects. Endothelial dysfunction assessed by FMD is recognized as an early marker of vascular damage, contributing to the initiation and progression of atherosclerosis . Although intima-media thickness of the carotid artery did not significantly differ between groups (lean, 1.06 ± 0.60 mm; overweight, 0.99 ± 0.45 mm; p = 0.456), it could be speculated that the overweight subjects in our study had less advanced atherosclerosis, and were thus able to respond to the vasodilator effect of moderately elevated plasma leptin levels at the time of FMD measurement.
Our results further demonstrated that plasma leptin levels were also independently and positively associated with NMD in both lean and overweight subjects and those associations were found only after adjusting for other confounders including age, obesity, BP, and lipids. The vasodilator response to exogenous NO reflects vascular smooth muscle function and is reported to be impaired independently of endothelial dysfunction in subjects at risk for atherosclerosis . Apart from the endothelium, leptin was also shown to directly target vascular smooth muscle cells via NO-dependent  and NO/endothelium independent  manner. Thus, the correlation between leptin and NMD in both lean and overweight subjects may reflect the smooth muscle-dependent vasodilator effect of leptin, which can be observed even in lean T2D patients with low plasma leptin levels.
There were a few limitations of our study. First, this was a cross-sectional study; therefore, a causal relationship between plasma leptin and FMD cannot be clarified. Second, the patients with T2D in this study were receiving various anti-atherogenic drug interventions such as anti-hypertensive agents, statins and insulin therapy that can exert considerable effects on FMD of the brachial artery and related atherosclerotic risk factors. To minimize the effect of such treatments, we adjusted for patient treatment status in our multivariate analyses. Third, our overweight subjects were significantly younger than the lean subjects. A positive relationship between BMI and FMD was also found in the univariate analysis. Since all potential confounding risk factors could not be adjusted for with the consecutive inclusion of our subjects, factors including age and BMI were adjusted for and the independent association of leptin was confirmed in the multivariate analyses. Fourth, no healthy controls were used to compare our findings, and we could not confirm that FMD was impaired in our study population of T2D patients. Last, this study included a very low number of morbidly obese patients with a BMI ≥ 30 kg/m2 (n = 15, 8.8%); thus, our results are only applicable to normal or overweight T2D patients. Leptin could contribute differently to FMD in morbidly obese patients with more severe leptin resistance and hyperleptinemia than overweight subjects [15, 49, 52].
Further studies with a larger population that includes T2D patients with a wide range of BMI are required to validate these findings. Furthermore, prospective and interventional studies assessing changes in both plasma leptin levels and FMD are warranted to clarify whether plasma leptin levels are predictive of vascular endothelial function in patients with obesity and T2D.