点击显示 收起
1 From the Department of Medicine, Division of Nephrology, San Francisco General Hospital; the University of California San Francisco; and the Department of Medicine, Division of Endocrinology and Metabolism, Stanford University School of Medicine, Stanford, CA.
2 Supported by research grants HL-08506 and RR-00070 from the National Institutes of Health. 3 Address reprint requests to FS Facchini, Box 1341 UCSF, San Francisco, CA 94110. E-mail: fste2000{at}yahoo.com.
ABSTRACT
Background: It is not known whether total circulating lipid hydroperoxides are increased in insulin-resistant individuals and whether this correlates with depletion of liposoluble antioxidant vitamins that are consumed during lipid peroxidation.
Objective: The goal of this study was to define the relation between resistance to insulin-mediated glucose disposal and plasma concentrations of lipid hydroperoxides and liposoluble antioxidant vitamins in healthy volunteers.
Design: Insulin-mediated glucose disposal was determined in 36 healthy, nondiabetic volunteers by measuring their steady-state plasma insulin (SSPI) and glucose (SSPG) concentrations in response to a 180-min constant infusion of octreotide, insulin, and glucose. In addition, fasting plasma concentrations of lipid hydroperoxides and liposoluble antioxidant vitamins were determined by using the FOX 2 assay and liquid chromatography.
Results: Statistically significant direct relations were observed between SSPG and mean arterial blood pressure (r = 0.44, P = 0.008) and plasma lipid hydroperoxide concentrations (r = 0.42, P = 0.01), whereas significant inverse correlations were found between SSPG and -carotene (r = -0.58, P = 0.0002), ß-carotene (r = -0.49, P = 0.004), lutein (r = -0.35, P = 0.04), -tocopherol (r = -0.36, P = 0.04), and -tocopherol (r = -0.45, P = 0.007).
Conclusions: Variations in insulin-mediated glucose disposal in healthy individuals are significantly related to plasma concentrations of lipid hydroperoxides and liposoluble antioxidant vitamins. These findings suggest that total plasma lipid peroxidation is increased in insulin-resistant individuals at an early, preclinical stage, ie, well before the development of glucose intolerance and type 2 diabetes.
Key Words: Insulin resistance • carotenoids • tocopherols • lipid hydroperoxides • oxidative stress • lipid peroxidation • glucose disposal
INTRODUCTION
We recently published the results of a study of nondiabetic individuals showing that the more insulin resistant an individual, the more oxidized his or her circulating LDL particles (1). Evidence also exists that circulating lipid peroxidation products are increased in type 2 diabetes (2). Because insulin resistance precedes the development of type 2 diabetes, the current study was initiated to broaden these earlier finding and to evaluate whether greater total plasma lipid peroxidation might occur at a preclinical stage, ie, in individuals who are at greater risk of type 2 diabetes but whose glucose tolerance is still normal. To achieve this goal we studied healthy individuals with various degrees of insulin resistance but normal glucose tolerance and quantified total circulating plasma lipid hydroperoxides. In addition, we measured concentrations of the main lipid-phase antioxidant defenses, ie, carotenoids and tocopherols.
SUBJECTS AND METHODS
Thirty-six healthy volunteers with a mean (±SEM) age of 47 ± 2 y and body mass index (BMI; in kg/m2) of 25 ± 1 (range: 19–32) were studied at the Stanford Hospital and San Francisco General Hospital General Clinical Research Centers. The study was approved by the hospitals' institutional review board and experimental procedures were in accordance with institutional guidelines. All subjects gave written, informed consent to participate.
Subjects were considered healthy on the basis of a physical examination, measurement of routine blood chemistry indexes, cell blood count, urinalysis, and a resting 12-lead electrocardiogram. Blood pressure was measured in the early morning with an automatic oscillometric device (Dinamap 8360; Critikon, Tampa, FL). Mean arterial blood pressure was calculated as one-third of pulse pressure plus diastolic pressure. All subjects were normotensive (blood pressure <145/85 mm Hg) and glucose tolerant (3). None of the subjects smoked.
After subjects had fasted overnight, blood was drawn and plasma was separated from whole blood in heparin-treated tubes by centrifugation (at 2800 x g for 15 min at 4°C) and frozen immediately at -80°C until analyzed. Plasma lipid hydroperoxides were measured by oxidation of ferrous iron in the presence of xylenol orange (4). In this method, also called the FOX 2 assay, hydroperoxides oxidize ferrous iron to ferric iron, which forms a complex with xylenol orange to yield a chromophore that can be detected spectrophotometrically at 560 nm. No lipid extraction step is necessary for lipoprotein analysis because the 90% methanol in a 25-mmol H2SO4/L solution in which the assay is performed denatures the lipoprotein sufficiently for access of ferrous iron to lipid hydroperoxides. The FOX 2 assay measures mainly total plasma lipid hydroperoxides (ie, the sum of cholesterol, triacylglycerol, and phospholipid hydroperoxides contained in all lipoprotein fractions) because detection of aqueous hydroperoxides requires more sensitive methods (4). With use of the FOX 2 assay, there are no extraction-related losses of lipid hydroperoxides (4), which allows more precise quantitation of authentic lipid hydroperoxides, provided that standard conditions are observed (5).
Canthaxanthin, - and ß-carotene, cryptoxanthin, lutein, lycopene, retinol, tocopherols, and zeaxanthin were measured by liquid chromatography (6). Plasma cholesterol, triacylglycerol, and lipoprotein fractions were measured with use of routine techniques (1).
After fasting blood specimens were collected, an insulin suppression test was performed to quantify insulin-mediated glucose disposal (7). Subjects were infused for 180 min with octreotide (5 µg/min), glucose (13.3 mmol•min/m2), and insulin (180 nmol•min/m2). The solution was infused continuously into an indwelling polytetrafluoroethylene catheter placed in a superficial antecubital vein. Venous blood samples were obtained from a similar catheter inserted in the contralateral antecubital vein and kept patent by a slow infusion of 0.9% NaCl. Blood was collected every 60 min during the first 2 h and every 10 min during the last half hour for measurement of plasma glucose (8) and insulin (9) concentrations. The mean values of the 4 measurements made during the last half hour were used as the steady-state plasma glucose (SSPG) and insulin (SSPI) concentrations. Octreotide inhibits endogenous insulin secretion, and infused glucose and insulin inhibit endogenous glucose production. Under these circumstances, the higher the SSPG, the more insulin resistant the individual.
Results are expressed as means ± SEs. Values not normally distributed (eg, SSPG, triacylglycerol, lipid hydroperoxides, and - and ß-carotene) were log transformed; for simplicity, log is omitted in the text and tables. Linear and multiple regression models were used to identify potential associations between the study variables and to adjust for covariates. We used a P for trend analysis to compare subjects in the most insulin-sensitive tertile with those in the most insulin-resistant one.
RESULTS
Mean values of clinical characteristics and plasma concentrations of lipid hydroperoxides, carotenoids, retinol, and tocopherols and regression coefficients between SSPG and the other experimental variables are shown in Table 1. SSPG was not significantly related to age, sex, or BMI. However, significant direct relations were found between SSPG and mean arterial blood pressure, plasma triacylglycerol, and lipid hydroperoxides, whereas significant inverse relations were found between SSPG and -carotene, ß-carotene, lutein, -tocopherol, and -tocopherol. No significant relations were found between SSPG and retinol, lycopene, the other carotenoids, or -tocopherol. Lipid hydroperoxides were significantly related to triacylglycerol (r = 0.31, P < 0.05) but not to cholesterol, mean arterial blood pressure, or BMI. In the relation between SSPG and lipid hydroperoxides, adjustment for age, sex, BMI, and triacylglycerol by multiple regression was performed and only SSPG retained predictive power (R2 = 0.69, P < 0.01; Table 2).
View this table:
TABLE 1.. Regression analysis between insulin resistance, clinical characteristics, and fasting plasma concentrations of lipid hydroperoxides, carotenoids, retinol, and tocopherols in healthy volunteers1
View this table:
TABLE 2.. Multiple regression analysis between lipid hydroperoxides as the dependent variable and age, sex, BMI, triacylglycerols, and steady-state plasma glucose (SSPG) as the independent variables
We also normalized lipophilic antioxidant concentrations to plasma triacylglycerol and total cholesterol concentrations. After this analysis, only the inverse relation between insulin resistance and -tocopherol (-tocopherol/mmol triacylglycerol + total cholesterol) was strengthened (r = -0.43, P< 0.01), indicating that dilution of the same number of -tocopherol molecules in a greater lipid mass is better correlated with impaired insulin sensitivity.
To further evaluate the relation between SSPG and the other variables of interest, comparisons were made between the lowest and highest tertiles in terms of SSPG. These results are shown in Table 3, and the findings are consistent with those shown in Table 1. Specifically, the most insulin-resistant tertile had significantly higher plasma lipid hydroperoxide concentrations and significantly lower concentrations of - and ß-carotene, lutein, -tocopherol (normalized for triacylglycerol and total cholesterol content), and -tocopherol.
View this table:
TABLE 3.. Effect of differences in insulin resistance on experimental variables1
The relation between plasma concentrations of lipid hydroperoxides and liposoluble vitamins is shown in Table 4. Concentrations of several carotenoids and tocopherols were significantly related to plasma lipid hydroperoxides.
View this table:
TABLE 4.. Regression analysis between concentrations of lipid hydroperoxides and carotenoids, retinol, and tocopherols in 36 healthy volunteers
DISCUSSION
Although the results of this study are straightforward, their interpretation is far from simple. Before attempting the more complicated task of creating a coherent formulation that ties our results together, it seems useful to first place our findings in the context of published data. To begin with, we showed in healthy, nondiabetic volunteers that the more insulin resistant the subjects were, the higher their total plasma concentration of lipid hydroperoxides. Therefore, observations made earlier of increased lipid hydroperoxides in persons with type 2 diabetes (2), a state in which virtually 100% of patients have some degree of insulin resistance, can now be extended to otherwise healthy persons with impaired insulin action. The demonstration of a general relation between insulin resistance and total plasma lipid hydroperoxides at an early stage, ie, well before glucose intolerance ensues, renders it unlikely that these changes are secondary to loss of glycemic control. In addition, the fact that SSPG remained the only predictor of lipid hydroperoxides after adjustment for triacylglycerol concentrations indicates that higher peroxidation of polyunsaturated fatty acids from triacylglycerol is probably not a major factor in explaining the observed differences.
The second finding of this study was that the more insulin resistant the healthy volunteers were, the lower their plasma concentrations of carotenoids and tocopherols. This result was seen both when the study population was considered as a whole and when the more insulin-sensitive individuals were compared directly with the more insulin-resistant ones. Although we believe this to be the first demonstration of this relation, it is consistent with other published data. For example, Ford et al (10) reported that fasting insulin concentrations are inversely correlated with concentrations of all serum carotenoids. Because the plasma insulin concentration is highly correlated with direct measures of insulin resistance (11), Ford et al's results and ours seem comparable. In addition, we previously published evidence that insulin-mediated glucose disposal is inversely related to self-reported dietary intakes of vitamins A and E (12).
When assimilating these findings, it is tempting to speculate that the lower the intake of liposoluble antioxidant vitamins, the greater the degree of insulin resistance and the higher the fasting plasma insulin concentration. However, the fact that degree of insulin resistance is related to both the plasma concentrations of carotenoids and tocopherols and the plasma concentrations of lipid hydroperoxides does not necessarily mean that these 2 variables are related to each other. On the other hand, it is difficult to avoid consideration of this possibility. The results of our attempt to address this issue are shown in Table 4, in which we report that plasma concentrations of several carotenoids and tocopherols were significantly correlated with lipid hydroperoxides.
One limitation of the current study is that we did not measure plasma ascorbic acid concentrations. Vitamin C is an essential water-soluble antioxidant able to effectively inhibit in vitro lipid peroxidation initiated by water-soluble prooxidants, such as aqueous extracts of cigarette smoke (13, 14). Although it is intuitive that different prooxidants may be more or less important in different individuals, there is the possibility that in nonsmokers (such as our healthy volunteers), oxidation of lipoproteins may be initiated by other species. These other species may not necessarily be water soluble (15) or may result from autooxidation in the presence of dioxygen, which, being very liposoluble itself, tends to become compartmentalized in micellar structures, such as lipoproteins and biomembranes. In this setting, lipophilic antioxidants, such as tocopherols and carotenoids, seem more relevant in inhibiting self-propagation of the autoxidation reaction. Therefore, although the role of ascorbate as a physiologic antioxidant cannot be negated, particularly in smokers, focus of the current study was on liposoluble antioxidant vitamins that are normally embedded in the lipoprotein structure and that inhibit lipoprotein oxidation no matter how started.
Given these observations in combination with our previous finding of an inverse correlation between insulin resistance and reported dietary intake of liposoluble antioxidant vitamins, we propose the following hypothesis. Decreased dietary intake of liposoluble antioxidant vitamins contributes to a decrease in plasma concentration of carotenoids and tocopherols, which in turn impairs the ability of insulin to stimulate glucose disposal by muscle. Insulin resistance results in an increase in lipid peroxidation, a process accentuated by the associated decrease in plasma antioxidant content. Alternatively, lipid peroxidation, secondary to a decrease in antioxidant vitamins, might impair insulin action.
Even if no dietary intake data were collected in the current study population, measurements made previously in a similar population (12) and data reported by others (10, 16–18) suggest that lipophilic antioxidant vitamins (in particular tocopherols) may modulate insulin action. Furthermore, low vitamin E concentrations, which occur more commonly in insulin-resistant subjects, were shown to predict diabetes (16), whereas increased consumption of raw vegetables (which supply most dietary tocopherols and carotenoids) was shown to have the opposite effect (19). Although, as suggested by recent reports (20, 21), carotenoids and tocopherols may not be the direct mediators of such effects, it seems plausible that some lipophilic antioxidant micronutrient, perhaps yet to be characterized, modulates insulin sensitivity, at least to some extent, and therefore changes risk of type 2 diabetes.
In conclusion, we showed that plasma concentrations of lipid hydroperoxides were higher in insulin-resistant volunteers who were otherwise healthy than in insulin-sensitive ones, presumably indicating enhanced lipid peroxidation, reduced clearance, or both. Such aberration was associated with significantly lower plasma concentrations of the lipid-phase antioxidants carotenoids and tocopherols. Although the associative nature of our findings does not allow us to establish cause and effect, an integrated analysis of the available literature strongly suggests that tocopherols and perhaps carotenoids should be added to the growing list of environmental factors known to modulate insulin effects.
REFERENCES