Plasma Asymmetric Dimethylarginine Concentrations Are Elevated in Obese Insulin-Resistant Women and Fall with Weight Loss

Context: Plasma asymmetric dimethylarginine (ADMA) concentrations are higher in apparently healthy, insulin-resistant (IR) individuals and decrease in response to thiazolidenedione treatment.

Objective: The objective of the study was to determine whether ADMA concentrations would also fall when insulin sensitivity is enhanced with weight loss in obese individuals.

Design/Setting/Patients/Intervention: Twenty obese women classified as IR or insulin sensitive (IS) on the basis of their steady-state plasma glucose (SSPG) concentration during the insulin suppression test underwent 12 wk of dietary weight loss.

Outcome Measures: Plasma glucose, insulin, and ADMA were measured at baseline and after weight loss; change in insulin resistance was quantified by repeating the SSPG after the dietary intervention.

Results: Although weight loss was similar in the two groups, significant improvements in SSPG, glucose, and insulin concentrations were confined to the IR group. Baseline plasma ADMA concentrations (mean ± sd) were higher in IR subjects (1.69 ± 0.44 vs. 1.18 ± 0.45 μmol/liter, P = 0.02) and decreased to 1.20 ± 0.22 μmol/liter (P < 0.001) with weight loss. In contrast, ADMA levels did not change with a similar extent of weight loss in the IS group.

Conclusion: Plasma ADMA levels are higher in obese, IR women than in equally obese, IS women and decrease in response to weight loss when associated with enhancement of insulin sensitivity.

PLASMA CONCENTRATIONS OF asymmetric dimethylarginine (ADMA), an endogenous inhibitor of nitric oxide synthase (1), are elevated in several clinical syndromes associated with insulin resistance and increased cardiovascular disease (CVD) risk (2–9). We have recently shown that plasma ADMA concentrations were also elevated in apparently healthy, insulin-resistant (IR) individuals and that treatment of these subjects with the thiazolidenedione (TZD) compound, rosiglitazone, resulted in both enhanced insulin sensitivity and a fall in plasma ADMA concentration (10). Although these findings were consistent with the notion that insulin resistance and compensatory hyperinsulinemia were responsible for the elevated plasma ADMA concentrations, the actions of TZD compounds are not limited to improving insulin sensitivity (11), and it is possible that the rosiglitazone-induced decrease in plasma ADMA concentrations could have resulted from direct effect of the drug, unrelated to its ability to enhance insulin sensitivity.

The present study, initiated to extend the observations discussed above, addressed the following two questions: 1) would plasma ADMA concentrations be elevated in IR individuals, compared with insulin-sensitive (IS) persons when both groups were obese; and 2) would elevated plasma ADMA concentrations in IR individuals also fall when insulin sensitivity was enhanced by nonpharmacological means? To answer these questions we took advantage of the fact that insulin sensitivity improves when obese, IR individuals lose weight but does not change with weight loss in IS, obese individuals (12, 13). Therefore, we compared plasma ADMA concentrations in equally obese women, divided into IR and IS subgroups, and repeated the measurements of both insulin sensitivity and ADMA concentrations after weight loss. The results demonstrated that plasma ADMA concentrations were higher at baseline in the IR, obese individuals and that weight loss in this group was associated with both an improvement in insulin sensitivity and a decrease in plasma ADMA concentrations. In contrast, neither variable changed when IS, obese persons lost weight. These findings provide additional evidence that plasma ADMA concentrations are increased in IR individuals and demonstrate that they will decline when insulin sensitivity is improved by nonpharmacological means.

Subjects and Methods

The study population consisted of volunteers from the San Francisco Bay Area who responded to newspaper advertisements seeking healthy, overweight women interested in weight loss. All potential subjects gave written, informed consent to participate in weight loss studies approved by the Stanford Human Subjects Committee. Subjects were required to have a body mass index between 29 and 35 kg/m², to be in apparently good health, to be nonpregnant and nonlactating, to have stable weight and no history of eating disorder or endocrine cause of obesity, and to have fasting plasma glucose less than 126 mg/dl. No subjects were taking blood pressure or lipid-lowering medications.

Insulin-mediated glucose disposal was quantified by a modification (14) of the insulin suppression test as originally described and validated (15, 16). Briefly, subjects were infused for 180 min with octreotide (0.27 μg/m²·min), insulin (32 mU/m²·min), and glucose (267 mg/m²·min). Blood was drawn at 10-min intervals from 150 to 180 min of the infusion to measure the plasma glucose and insulin concentrations, and the average of these four values was considered to represent the steady-state plasma insulin (SSPI) and glucose (SSPG) concentrations for each individual. Because SSPI concentrations are similar in all subjects, the SSPG concentration provides a direct measure of the relative ability of insulin to mediate disposal of an infused glucose load: the higher the SSPG concentration, the more IR the individual.

Based on the SSPG concentrations during the insulin suppression test, subjects who qualified as IR or IS (but not intermediate) were eligible to proceed with the dietary weight loss intervention. Whereas insulin-mediated glucose disposal varies continuously in healthy nondiabetic individuals, precluding an objective definition of an individual as being either IS or IR (17), we previously demonstrated that CVD developed to a significantly greater degree in the third of the healthy subject population that was the most IR (18, 19). Thus, for the purposes of this study, we defined individuals with SSPG concentrations in the top tertile of the distribution observed in 490 healthy volunteers (including normal weight) (17) as IR and those in the lowest tertile as IS. Plasma was obtained after an overnight fast for measurement of fasting plasma l-arginine and ADMA concentrations as previously described (3, 10).

After baseline measurement (3, 10) of plasma ADMA concentrations in the IR and IS groups, we measured plasma glucose and insulin responses to standardized test meals comprising (as percent calories) 42% carbohydrate, 43% fat, and 15% protein. The meals were given at 0800 h (20% of daily caloric requirement) and 1200 h (40% of daily caloric requirement), and blood was drawn for measurement of plasma glucose and insulin concentrations before the meal at 0800 h and then at hourly intervals for 8 consecutive hours. Daylong glucose and insulin responses were calculated, using the trapezoidal method, as the area under the curve of the nine time points, with equal weighting of each time point. These values, along with individual time points were used for statistical analyses as described below.

The dietary intervention consisted of a 12-wk period of weight loss via hypocaloric diet. Subjects were individually instructed on a calorie-restricted diet, using the Harris-Benedict equation (20) for resting energy requirements, multiplied by an activity factor, less 1000 kcal/d, to produce a weight loss of approximately 0.9 kg/wk. For all subjects, macronutrient composition of the diets was 42% carbohydrate, 43% fat, and 15% protein. Subjects met with research nutritionists at weekly intervals during the dietary period to enhance compliance and evaluate progress of weight loss. At these visits subjects were also reminded to maintain their usual level of physical activity (to minimize possible confounding by changes in exercise level), and were questioned as to any changes. After completion of the weight-loss phase, subjects were assigned a weight-maintenance diet for 2 wk before repeating the baseline measurements described above. Only those volunteers who completed the study and lost at least 5% of initial body weight are included in the present analysis.

The results are expressed as mean ± sd. Student’s unpaired t test (two tailed) was used to evaluate the statistical significance of baseline differences between the IR and IS groups. Due to small numbers, women who were premenopausal and/or took estrogen replacement were combined for comparison with postmenopausal women (without estrogen replacement), defined as cessation of menses for at least 1 yr. Fisher’s exact test was used to compare the menopausal hormone status. Student’s paired t test was used to evaluate the change in ADMA, weight, and SSPG within each group. ANOVA was used to compare the change in ADMA, with group (IR vs. IS) as the main effect and percent weight change and baseline ADMA as covariates. Daylong insulin and glucose values were compared at baseline (between groups) and before and after weight loss (within groups) via repeated-measures ANOVA. Pearson’s correlation coefficients were calculated to assess the potential association between change in SSPG, daylong insulin response (area under the curve), weight, and change in ADMA concentrations. P < 0.05 was considered statistically significant. All analyses were done with Systat (version 10.2; Richmond, CA).

Results

SSPG concentrations were determined in 84 volunteers; 65 classified as being either IR or IS by the criteria defined in Subjects and Methods. Forty-four of these individuals completed the weight-loss study, and the results of 20 volunteers (10 IR and 10 IS) who lost at least 5% of initial body weight were included in the current analysis. The baseline characteristics of the two experimental groups are shown in Table 1, and it is clear that the only significant difference between them was their SSPG concentrations. None of the subjects engaged in strenuous physical activity (e.g. marathon training), and they were requested to maintain the current level of activity throughout the study duration.

Figure 1 indicates that both the IR and IS groups lost a statistically significant amount of weight (8.6 ± 3.9 kg, 10.3%, and 7.3 ± 2.8 kg, 8.7%, respectively), and the amount of weight lost did not differ significantly between groups. Figure 1 also demonstrates that SSPG concentration decreased in association with weight loss by 42% in the IR group (P < 0.001), whereas the values were similar before and after weight loss in IS individuals. Although insulin sensitivity was clearly enhanced with weight loss in the IR subjects, it should be noted that their SSPG concentrations after weight loss were still higher than those of IS individuals.

Daylong plasma glucose and insulin concentrations are shown in Fig. 2. The baseline daylong glucose response was similar in the two groups (P = 0.32), and weight loss did not lead to a significant decline in glucose concentrations in either the IR (P = 0.24) or IS (P = 0.69) group.

Figure 2 also illustrates the impact of weight loss on the daylong insulin response of the two groups. It is obvious that baseline plasma insulin concentrations were higher in the IR subjects (repeated measures ANOVA, P = 0.005) and that they decreased significantly when these individuals lost weight (repeated measures ANOVA, P = 0.047). In contrast, there was no significant change in the daylong insulin response after weight loss in the IS group. Similar to the changes in SSPG concentration, daylong plasma insulin levels in the IR subjects declined to a value approximately halfway between where they were at baseline and the values in the IS subjects.

Plasma ADMA concentrations before and after weight loss in the two experimental groups are shown in Fig. 3. These results demonstrate that plasma ADMA concentrations were significantly higher (P = 0.02) in the IR women (1.69 ± 0.44 μmol/liter) than the equally obese, IS women (1.18 ± 0.45 μmol/liter). Furthermore, ADMA concentrations declined significantly in association with weight loss (P < 0.001) in the IR group, whereas they did not change in the IS group. Even after adjustment for percent weight change and baseline ADMA concentrations, as compared with the IS group, the IR group experienced a significantly greater decline in ADMA concentrations (P = 0.028). As a result, plasma ADMA levels after weight loss were identical in the IR (1.20 ± 0.22 μmol/liter) and IS (1.20 ± 0.60 μmol/liter) groups.

In addition to having higher plasma ADMA concentrations, the plasma l-arginine to ADMA concentration ratio was significantly lower in the IR subjects at baseline (48.1 ± 13 vs. 74.4 ± 31.8, P < 0.05). Unlike ADMA concentrations, weight loss did not lead to statistically significant changes in l-arginine concentrations. Consequently, the plasma l-arginine to ADMA concentration ratio increased in the IR subjects but not the IS subjects, resulting in the difference between the two groups being no longer statistically significant (63.0 ± 13.5 vs. 79.8 ± 44.9).

The relationships between the change in ADMA and the change in SSPG, daylong plasma insulin concentration, and body weight in the entire study population are shown in Fig. 4. It is apparent that there was a significant association between the decline in insulin resistance (SSPG) and the fall in ADMA concentration (r = 0.70, P = 0.001) and a significant association of lesser magnitude between the decrease in daylong plasma insulin concentrations and the decline in ADMA concentration (r = 0.48, P = 0.04). On the other hand, there was no correlation between the amount of weight lost and the fall in plasma ADMA concentration.

Discussion

The results of the current study demonstrate that obese, apparently healthy women who are IR have higher plasma ADMA concentrations than do equally obese, apparently healthy women who are IS. When considered in conjunction with our earlier findings (10), we have now demonstrated that plasma ADMA concentrations are higher in IR, compared with IS, nonobese men and women, obese women, and patients with essential hypertension. The fact that plasma ADMA concentrations are also elevated in patients with type 2 diabetes, CVD, and renal failure (2–5, 7), clinical syndromes known to be associated with insulin resistance (6, 8, 9), lends further support for the existence of a relationship between insulin resistance and increases in plasma ADMA concentrations.

The results of this study expand our understanding of the complex relationship between obesity, insulin resistance, endothelial dysfunction, and CVD. At the simplest level, they provide further evidence that the prevalence of CVD risk factors varies significantly in obese individuals as a function of their degree of insulin sensitivity. Thus, we have previously shown that glucose intolerance, hyperinsulinemia, dyslipidemia, and elevated levels of C-reactive protein occur primarily in those obese individuals who are also IR (13, 14). The current study extends these earlier findings and provides evidence that elevated ADMA concentrations should be added to the list of CVD risk factors that occur primarily in the subset of obese individuals that are also IR.

In the current study, we showed that weight loss with concurrent enhancement of insulin sensitivity leads to a substantial and statistically significant decline in ADMA concentration among IR individuals. In essence, the ADMA concentrations in these individuals were normalized to those of their IS counterparts, even though their insulin sensitivity did not improve to the same degree. This finding is important for two reasons. First, it suggests a mechanism by which weight loss may lead to enhanced endothelial function and lower CVD risk. Second, the current findings extend those of our prior study, in which we found that enhancement of insulin sensitivity with TZD treatment (10) is capable of decreasing plasma ADMA concentrations. Thus, the current results, in which we used a different means of enhancing insulin action weight loss, provide support for the hypothesis that reductions in ADMA in the prior study were a function of enhanced insulin action rather than a direct drug effect. Our results are strengthened by the use, in the current study, of a control group of equally obese individuals who were not IR at baseline and who did not experience a change in insulin action despite weight loss similar in magnitude to that seen in the IR group. The fact that in the IS group, ADMA concentrations were low to begin with and did not change with weight loss suggests that insulin resistance, rather than obesity per se, is a potential mediator of ADMA concentrations. Thus, the results of our two studies strongly suggest that insulin resistance is associated with elevations in ADMA concentrations and that enhanced insulin sensitivity, irrespective of the specific intervention, has the ability to lower ADMA concentrations.

Although recognition of the importance of elevated plasma ADMA concentrations as indicating increased CVD risk is relatively recent, the evidence that this is the case is accumulating rapidly (7, 21–23). Furthermore, the biological function of ADMA, competitive inhibition of nitric oxide synthase bioactivity, provides a logical explanation by which ADMA potentiates CVD risk. In addition to its action as a vasodilator, nitric oxide inhibits platelet aggregation, leukocyte adhesion, proliferation of vascular smooth muscle, and endothelial generation of superoxide anion, all of which are key processes in atherogenesis (24). Although it is tempting to speculate that endothelial function would improve when plasma ADMA concentrations fall in overweight, IR individuals, this functional variable was not quantified in our study. However, results of several previous weight-loss studies (25–28), using a variety of different experimental protocols, have concluded that endothelial vasodilatory function improves in association with weight loss. Thus, the observed improvement in ADMA concentrations and resultant l-arginine to ADMA ratios described in this study when IR volunteers lost weight provides a unifying explanation for this beneficial effect.

In conclusion, the current results, taken together with our previous findings, indicate that plasma ADMA concentrations are elevated in IR, hyperinsulinemic individuals and that they decrease significantly in association with enhanced insulin sensitivity. These observations emphasize the potential clinical benefit of interventions that improve insulin sensitivity and decrease plasma ADMA concentrations. Indeed, elevations in ADMA may represent a reversible abnormality that contributes to the increased CVD risk seen in IR individuals.

Acknowledgments

This work was supported by National Institutes of Health Grants RR2HLL406, RR 000070, RR16071-01, and 1RO1HL73084-01.

T.M., M.S., C.L., F.A., J.B., G.M.R., and P.S.T. have nothing to declare.

Abbreviations:

• ADMA,

  Asymmetric dimethylarginine;

• CVD,

  cardiovascular disease;

• IR,

  insulin-resistant;

• IS,

  insulin-sensitive;

• SSPG,

  steady-state plasma glucose;

• SSPI,

  steady-state plasma insulin;

• TZD,

  thiazolidenedione.