Impact of sex and diet-induced weight loss on vascular insulin sensitivity in type 2 diabetes

Abstract

Vascular insulin resistance, a major characteristic of obesity and type 2 diabetes (T2D), manifests with blunting of insulin-induced vasodilation. Although there is evidence that females are more whole body insulin sensitive than males in the healthy state, whether sex differences exist in vascular insulin sensitivity is unclear. Also uncertain is whether weight loss can reestablish vascular insulin sensitivity in T2D. The purpose of this investigation was to 1) establish if sex differences in vasodilatory responses to insulin exist in absence of disease, 2) determine whether female sex affords protection against the development of vascular insulin resistance with long-term overnutrition and obesity, and 3) examine if diet-induced weight loss can restore vascular insulin sensitivity in men and women with T2D. First, we show in healthy mice and humans that sex does not influence insulin-induced femoral artery dilation and insulin-stimulated leg blood flow, respectively. Second, we provide evidence that female mice are protected against impairments in insulin-induced dilation caused by overnutrition-induced obesity. Third, we show that men and women exhibit comparable levels of vascular insulin resistance when T2D develops but that diet-induced weight loss is effective at improving insulin-stimulated leg blood flow, particularly in women. Finally, we provide indirect evidence that these beneficial effects of weight loss may be mediated by a reduction in endothelin-1. In aggregate, the present data indicate that female sex confers protection against obesity-induced vascular insulin resistance and provide supportive evidence that, in women with T2D, vascular insulin resistance can be remediated with diet-induced weight loss.

INTRODUCTION

Insulin actions on the vascular endothelium lead to Akt and nitric oxide synthase (eNOS) activation, nitric oxide production, and vasodilation, facilitating the delivery of insulin and glucose to target tissues such as skeletal muscle (1–5). Concurrently, insulin stimulates the formation of the vasoconstrictor peptide endothelin-1 (ET-1) in endothelial cells, which can limit nitric oxide-mediated vasodilation during hyperinsulinemia (6–11).

Impaired endothelial insulin signaling, and consequent blunting of insulin-induced dilation and blood flow, is a distinguished feature of obesity and type 2 diabetes (T2D) (8–13). In fact, increasing evidence reveals that vascular insulin resistance is an early event in the genesis of obesity-induced systemic insulin resistance that precedes insulin resistance in other metabolically active tissues (9, 14–16). Notably, disruption of insulin signaling in endothelial cells contributes to whole body insulin resistance, impaired glucose homeostasis, and vascular disease development (17–19).

In the healthy state, whole body insulin sensitivity is greater in females than males, and this sex difference has been attributed to many factors, including differences in sex hormones, fat distribution, estrogen signaling in adipose tissue and liver, and skeletal muscle β-oxidation capacity (20–26), among others. However, it is possible that an inherent sexual dimorphism in vascular insulin sensitivity also exists and further contributes to sex differences in glucose control. Moreover, although recent data in humans indicate that young healthy female individuals are relatively protected against impairments in vascular insulin sensitivity induced by the short-term adoption of an obesogenic lifestyle (27), the extent to which female sex affords protection against the development of vascular insulin resistance with long-term overnutrition and obesity requires further investigation. Finally, once vascular insulin resistance is fully manifested, as it occurs in adults with obesity that progresses into T2D, it remains unknown whether diet-induced weight loss can attenuate vascular insulin resistance in both men and women.

To begin to address these questions, we first assessed sex differences in insulin-induced dilation in isolated femoral arteries from healthy mice, as well as sex differences in leg blood flow responses to systemic insulin infusion in a large cohort of healthy humans. After demonstration that, in the healthy state, sex differences in vascular insulin sensitivity are not observed, we next examined if female mice exposed to an obesogenic diet are protected against vascular insulin resistance compared with male counterparts. Finally, following the observation that the impairment in vasodilator responses to insulin in adults with obesity that advances into T2D is not influenced by sex, we tested in a subset of men and women if weight loss induced by a low-energy, low-glycemic diet enhances insulin-stimulated leg blood flow, as well as other indices of vascular function. Follow-up mechanistic experiments were performed in cultured endothelial cells and in isolated arteries from a female swine model of T2D.

METHODS

Ethics and Approvals

All human study procedures conformed to the Declaration of Helsinki and were approved by the University of Missouri Institutional Review Board (Protocol 2012869, 2008181, 2012106, 2013126, and 2020286). The clinical trial involving the dietary intervention was registered at ClinicalTrials.gov (NCT03648996). Written informed consent was obtained from all subjects before any procedures. All animal study procedures received prior approval by the University of Missouri Animal Care and Use Committee and were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. To aid the reader, hereafter men/women and males/females are used when referring to human and animal studies, respectively.

Human Studies

Healthy men (n = 33) and women (n = 33), 18 to 50 yr of age and nonobese [body mass index (BMI) of <30 kg/m²], from whom we had assessed insulin-stimulated leg blood flow, were included in a retrospective analysis designed to examine the effects of sex on vascular insulin sensitivity. A separate cohort of men (n = 19) and women (n = 18) with a diagnosis of T2D, 35 to 75 yr of age and overweight or obese (BMI = 25–50 kg/m²) and from whom we also had measures of insulin-stimulated leg blood flow were used to examine the effects of sex in the setting of T2D. All participants were free of a history of recent cardiovascular events (<12 mo), uncontrolled hypertension (≥180 mmHg systolic or 100 mmHg diastolic), autoimmune diseases, renal or hepatic diseases, active cancer, use of immunosuppressant therapy, excessive alcohol consumption (>14 drinks/wk for men; >7 drinks/wk for women), current tobacco use, and pregnancy (confirmed by negative pregnancy test on the morning of the study visit). Subjects were recruited from the local community and University of Missouri Hospital and Clinics. It should be noted that data from participants included in this sex-specific retrospective analysis were included in previous publications examining other questions (10, 27, 28).

A subset of men (n = 7) and women (n = 9) with T2D underwent a 6-mo, low-energy dietary intervention to examine the effects of weight loss on vascular insulin sensitivity. Subjects with a habitual diet already containing low amounts of sugars (<5% of total energy intake) and a history of changes in body weight (≥10%) within the past 6 mo were excluded from this trial. The characteristics of the subjects enrolled in the dietary intervention are provided in Table 1. Of the nine women, one was premenopausal, and the other eight were postmenopausal. None of them was on hormone-replacement therapy.

Table 1.

Subject characteristics, anthropometrics, and blood profile parameters before and after the 6-mo dietary intervention

	Before	After
Age, yr	58.8 ± 3.1	59.4 ± 3.1*
Sex (females/males)	(9/7)
Height, cm	168.8 ± 2.9
Weight, kg	106.8 ± 5.8	99.7 ± 5.3*
BMI, kg/m2	36.8 ± 1.5	34.4 ± 1.5*
Body fat, %	44.1 ± 1.5	42.6 ± 1.7*
Lean mass, kg	57.1 ± 3.3	54.7 ± 3.0*
Blood glucose, mg/dL	120.5 ± 6.4	112.9 ± 11.5
Plasma insulin, μIU/mL	11.6 ± 1.6	7.3 ± 4.9*
White blood cell count, ×10/L	7.11 ± 0.32	6.37 ± 0.34*
Plasma AST, IU/L	21.19 ± 1.82	17.73 ± 1.60*
Plasma ALT, IU/L	27.31 ± 2.77	19.47 ± 1.99*
Plasma uric acid, mg/dL	5.61 ± 0.34	5.93 ± 0.30
Plasma IFN-γ, pg/mL	0.55 ± 0.17	0.90 ± 0.33
Plasma IL-10, pg/mL	4.45 ± 0.92	3.22 ± 0.49
Plasma IL-6, pg/mL	1.73 ± 0.30	1.56 ± 0.34
Plasma nitrite, nM	25.54 ± 3.77	26.31 ± 3.0
Systolic blood pressure, mmHg	137 ± 5	131 ± 6
Diastolic blood pressure, mmHg	78 ± 2	76 ± 3
Heart rate, beats/min	71 ± 6	64 ± 3
Medications (n)
Biguanide	12	12
Dipeptidyl peptidase-IV inhibitor	0	0
Glucagon-like peptide-1 agonist	4	5
Sodium-glucose cotransporter 2 inhibitor	4	4
Sulfonylurea	8	8
Insulin	7	7
Thiazolidinedione	1	1
Angiotensin-converting-enzyme inhibitor/	5	5
Angiotensin II receptor antagonist	7	7
β Blocker	3	3
Calcium channel blocker	3	3
Thiazide	6	6
Fibric acid agent	2	2
HMG-CoA reductase inhibitor	16	16
Omega-3-acid ethyl ester	2	2

Data are presented as means ± SE, n = 16. Data were analyzed using paired t test. Nonnormally distributed data (fasting blood glucose, AST, ALT, IFN-γ, IL-10, IL-6, nitrite, diastolic blood pressure, and heart rate) were analyzed using the Wilcoxon signed-rank test. *P < 0.05. ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; HMG-CoA, 3-hydroxy-3-methyl-glutaryl-coenzyme A.

Low-energy, low-glycemic dietary intervention.

The goal of the dietary intervention was to induce weight loss by reducing intake of dietary sugars to <5% of energy and total daily energy intake by 500 kcal. For the first 2 wk (controlled dietary phase), participants were provided all meals by the study team, which were prepared at the University of Missouri Nutritional Center for Health. Subjects were instructed to not eat any other foods. Within week 2, subjects individually attended a grocery store tour with the study dietitian to identify high-fructose and high-sugar foods to avoid. After 2 wk, subjects began to prepare their own meals while following the dietary regimen provided by the research dietitian (self-provided dietary phase). Participants were in contact with the dietitian and study team to have their questions answered and received social support to maintain dietary adherence. Subjects were allowed to consume nonsugar-containing beverages, including water, low- or no-energy-containing sodas, and regular coffee or tea without creamer, which could be sweetened with noncaloric sweeteners. Consumption of sugar-sweetened beverages and fruit juices that are high in fructose and/or added sugar was discouraged during the dietary intervention. Habitual alcohol consumption could be continued if drinks containing fructose and/or added sugars (e.g., margaritas, fruit-containing drinks) were avoided. Dietary counseling occurred weekly for the first 6 wk and again at weeks 8, 12, 16, and 20. Subjects were also asked to self-report body weight at weeks 3, 4, 5, 6, 8, 16, and 20 for monitoring purposes. Dietary records were analyzed using Nutrition Data System for Research software (University of Minnesota, Minneapolis, MN).

Before and after the dietary intervention, participants reported to the Clinical Research Center after an overnight fast and after refraining from caffeine for at least 12 h and vigorous physical activity for at least 24 h. At these visits, subjects underwent measurements of body weight, height, and body composition via dual-energy X-ray absorptiometry, as well as assessments of brachial artery flow-mediated dilation (an index of endothelium-dependent vasodilation) and carotid-to-femoral pulse-wave velocity (an index of arterial stiffness), all as previously described (10, 13, 27, 29). Subsequently, insulin-stimulated leg blood flow and skeletal muscle perfusion were assessed as described below.

Assessment of insulin-stimulated leg blood flow and skeletal muscle perfusion.

Intravenous catheters were inserted into the antecubital veins of both arms for blood draws and infusion. Approximately 20 min after placement of catheters, and while the subject was resting in the supine position, assessments of superficial femoral artery blood flow followed by quadriceps muscle (vastus lateralis) microvascular perfusion were obtained via Doppler ultrasound and contrast-enhanced ultrasound (iE33; Philips Medical Systems), respectively, as detailed previously (10, 27). Skeletal muscle microvascular perfusion measures are only reported in individuals with T2D who underwent the dietary intervention. Insulin (Humulin R U-100) was prepared via dilution in 250 mL of 0.9% saline along with 5 mL of blood taken from the subject to a final concentration of 500 mU/mL, as previously described (10, 27, 28, 30, 31). Insulin infusion started with two priming infusion rates over 10 min (160 to 80 mU/m² body surface area/min), followed by a steady infusion rate of 40 mU/m² body surface area/min for 60 min. During insulin infusion, blood glucose was determined every ∼5 min at bedside (YSI 2300 STAT PLUS glucose analyzer) and maintained at baseline levels by infusing a dextrose solution at a variable rate. Plasma was also obtained and stored at −80°C for analysis of insulin, which was assessed using a commercially available kit (ALPCO Cat. No. 80-INSHU-E10.1, Salem, NH). Femoral artery blood flow and quadriceps muscle microvascular perfusion were reassessed at the end of the 60-min insulin infusion, and data are presented as percent of change from preinsulin infusion values. Hemodynamic variables including heart rate and blood pressure were monitored throughout, as previously described (10, 27, 28, 30, 31).

Assessment of biochemical parameters.

Fasted plasma and whole blood samples were immediately frozen and sent to the University of Missouri Diabetes Diagnostic Laboratory for insulin and hemoglobin A1c (HbA1c) analysis, respectively. Plasma samples were also sent to the University of Minnesota Advanced Research and Diagnostic Laboratory for analysis of glucose, uric acid, liver enzymes, and lipid profile. The homeostasis model assessment of insulin resistance (HOMA-IR) was calculated as follows: [fasting glucose (mg/dL) × fasting insulin (μU/mL)]/405. Concentrations of ET-1 in plasma were assessed using an ELISA kit (No. EIAET1, Invitrogen, Thermo Fisher Scientific) following the manufacturer’s instructions. Plasma inflammatory cytokines [IL-6, IL-10, IL-12, IFN-γ, monocyte chemoattractant protein-1 (MCP-1), and TNF-α] concentrations were measured using the MILLIPLEX Human Cytokine/Chemokine Magnetic Bead Panel–Immunology Multiplex Assay (HCYTOMAG-60K) in a Bioplex 200 (Bio-Rad), also according to the manufacturer’s instructions. Nitrite, a by-product of nitric oxide, was assessed in plasma samples using the gold-standard method of ozone-based reductive chemiluminescence (CLD88, Eco Physics, Ann Arbor, MI) according to the manufacturer’s recommendations and as previously described (27, 32).

Animal and In Vitro Studies

Experiments in isolated femoral arteries from mice.

Male and female wild-type (C57BL/6) mice fed normal chow, 21–33 wk of age, and from which we had collected measures of insulin-induced dilation in isolated femoral arteries were included in a retrospective analysis designed to examine the effects of sex in the absence of disease. Two weeks before being euthanized, a subset of these male and female mice underwent glucose tolerance testing (GTT), as previously described (33). To assess whether female sex offers protection against the development of vascular insulin resistance with long-term overnutrition and obesity, 5-wk-old male and female wild-type (C57BL/6) mice were randomly assigned to either remain on standard chow diet or fed a obesogenic diet (5.06 kcal/g of food consisting of 60.1% kcal from fat, 14.9% kcal from protein, and 25.0% kcal from carbohydrate; F1850, Bio Serv) for 28 wk. Mice were kept on a 12-h light:dark cycle with ad libitum access to water and food. Two weeks before euthanasia, mice underwent insulin tolerance testing, as described (34). At termination, the femoral artery was excised for assessment of insulin and sodium nitroprusside-induced dilation, as described previously (34). Briefly, femoral arteries were isolated and cannulated onto glass micropipettes, pressurized at 70 mmHg without flow, and warmed to 37°C in commercial pressure myography chambers (Living Systems Instrumentation, Burlington, VT). Arteries were preconstricted with phenylephrine (10 µM), and vasodilatory responses to insulin (10⁻⁹ to 10⁻⁵ M) and sodium nitroprusside (10⁻⁸ to 10⁻⁴ M) were determined. Arteries that failed to respond to the preconstrictor were considered unviable and were excluded from analysis. It should be noted that some phenotypic data from these wild-type mice were used in previous publications addressing unrelated questions (33, 34).

Experiments in isolated skeletal muscle resistance arteries from swine.

To examine the role of ET-1 in suppressing insulin-induced vasodilation in T2D, experiments were performed in isolated skeletal muscle resistance arteries from healthy control pigs and our recently established swine model of cardiometabolic disease that develops T2D and heart failure. The five female control pigs (final body weight = 46.2 ± 0.8 kg; fasting plasma glucose = 92.1 ± 6.8 mg/dL; total cholesterol = 76.1 ± 5.0 mg/dL; triglycerides = 31.5 ± 5.8 mg/dL) and four female T2D pigs (final body weight = 77.4 ± 3.8 kg; fasting plasma glucose = 115.3 ± 6.1 mg/dL; total cholesterol = 976.5 ± 132.9 mg/dL; triglycerides = 129.6 ± 30.1 mg/dL) used for these experiments were described in detail by Olver et al. (35). After euthanasia, two resistance arteries were isolated from the medial portion of the long head of the triceps, and insulin-induced dilation was assessed using pressure myography, as previously described (16). Arterial responsiveness to 80 mM KCl was used to ensure viability. Next, one artery was treated with the ET-1 receptor A inhibitor BQ-123 (1 µM; B150, Sigma) in the bath for 30 min, whereas the other artery was kept untreated, before assessing vascular reactivity to increasing insulin concentrations (10⁻⁹ to 10⁻⁵ M). Vasodilatory responses to insulin were calculated as percent maximal dilation, which was achieved under Ca²⁺-free conditions after completion of the protocol. It should be noted that in our original article describing this swine model, vascular responses to insulin in untreated arteries were presented in the supplemental material (35).

Experiments in cultured endothelial cells.

Human umbilical vein endothelial cells (No. CC-2519, Lonza, Morristown, NJ) were cultured in VascuLife EnGS (Lifeline Cell Technology, Frederick, MD) medium with 2% fetal bovine serum (FBS) under standard conditions (37°C, 5% CO₂). Experiments were performed when cells reached ∼90% confluency. Plasma samples from subjects with T2D (n = 16; 7 men and 9 women) collected before versus after the dietary intervention were pooled within the timepoint and used for cellular treatments. Specifically, endothelial cells were treated for 24 h with pooled plasma (30% of final volume of medium with 0.5% FBS) from each timepoint. In a separate experiment, endothelial cells were treated for 24 h with a mixture of proinflammatory cytokines (10 ng/mL each) found to be reduced in circulation after the dietary intervention, namely, IL-12, TNF-α, and MCP-1. After treatments, cell culture supernatants were collected and stored at −80°C for subsequent analysis. Quantification of ET-1 in supernatants from all experiments was assessed using an ELISA kit (No. EIAET1, Invitrogen, Thermo Fisher Scientific) following the manufacturer’s instructions. For the plasma-endothelial cell experiment, the initial concentration of ET-1 in media (with 30% plasma) was also measured and subtracted from the final concentration in the supernatant to calculate net production of ET-1 by endothelial cells over the 24-h period. Nitrite concentration in the supernatants of cytokine-endothelial cell experiments was assessed using the same chemiluminescence method stated earlier. Samples were not available for determination of nitrite production in plasma-endothelial cell experiments. Specific protein content was assessed in endothelial cell lysates prepared in radioimmunoprecipitation assay buffer (RIPA) buffer (Invitrogen, Cat. No. 1861278, Waltham, MA) and phosphatase inhibitors (Invitrogen, Cat. No. 1862495). Proteins within samples were separated in Criterion Tris-Glycine eXtended-PAGE precast gels (Bio-Rad) and transferred onto polyvinylidene difluoride membranes. Specific proteins were probed using the following antibodies: anti-pAkt^Ser473 (1:1,000, No. 4060, Cell Signaling), anti-Akt (1:1,000, No. 5298, Santa Cruz), and anti-eNOS (1:1,000, No. 32027, Cell Signaling). Individual protein band intensities were quantified via densitometry via Bio-Rad ChemiDoc XRS+ System (Bio-Rad). Stain-free technology (Bio-Rad) was used to determine protein loading and for normalization purposes.

Statistical Analysis

Data are presented as means ± SE. Shapiro–Wilk’s test was performed for assessment of data distribution. Sex and/or treatment-related differences in outcomes were determined using two-tailed paired and unpaired t test, and two-way repeated-measures ANOVA, as appropriate. Bonferroni post hoc was performed when significant interactions were found. When data were not normally distributed, nonparametric tests, Mann–Whitney U (Wilcoxon rank-sum test) and Wilcoxon signed-rank tests, were performed. The results were considered significant when P < 0.05. Statistical analyses were performed using GraphPad, Prism (version 9.0).

Our dietary intervention was powered to detect a mean difference in leg blood flow response to insulin between pre- and postvisits of 50%. Based on previous work, this represents approximately two-thirds of the distance in mean response between individuals with T2D and age-matched healthy controls (10). Assuming a correlation of r = 0.6 between serial measurements on the same subject, a sample size of 12 subjects would be sufficient to provide 80% power to detect a significant effect of the intervention. To account for possible dropouts and/or data loss in some variables, we enrolled 16 subjects.

RESULTS

Sex Does Not Influence Vascular Insulin Sensitivity in Healthy Mice or Humans

As displayed in Fig. 1A, insulin-induced dilation in isolated femoral arteries was not different between male and female chow-fed mice. Vasodilatory responses to sodium nitroprusside were also not affected by sex. As expected, female mice exhibited enhanced glucose tolerance compared with males, suggestive of increased whole body insulin sensitivity.

Figure 1.

Sex does not influence vascular reactivity to insulin in healthy mice or humans. A: insulin- and sodium nitroprusside-induced dilation in isolated femoral arteries from C57BL6 male (n = 19) and female (n = 17) mice fed chow, and glucose tolerance testing (GTT) from a subset of these animals (males, n = 11; females, n = 10). Arteries were preconstricted with phenylephrine (Phe). B: plasma insulin and blood glucose concentrations following 60 min of systemic insulin infusion with coinfusion of dextrose to prevent hypoglycemia in healthy men (n = 30–33) and women (n = 30–32). Leg blood flow and vascular conductance responses to 60 min of insulin infusion are presented as percent changes from preinsulin infusion (men, n = 33; women, n = 33). Data are presented as means ± SE. Data were analyzed using two-way repeated-measures ANOVA or unpaired t test, as appropriate. Nonnormally distributed data (leg blood flow and vascular conductance) were analyzed using the Mann–Whitney U (Wilcoxon rank-sum test) test. *P < 0.05. AUC, area under the curve.

Plasma insulin and blood glucose concentrations during systemic infusion of insulin in healthy men and women are presented in Fig. 1B and demonstrate successful achievement of hyperinsulinemia while preventing hypoglycemia. As shown, leg blood flow responses to insulin infusion, as well as changes in leg vascular conductance, were not different between men and women. Mean arterial pressure was maintained during insulin infusion in both men (0 min = 88 ± 2 mmHg, 60 min = 89 ± 1 mmHg, P = 0.30) and women (0 min = 84 ± 1 mmHg, 60 min = 84 ± 1 mmHg, P = 0.97). In both men and women, heart rate increased during insulin infusion (men: 0 min = 60 ± 2 beats/min, 60 min = 63 ± 2 beats/min, P < 0.05; women: 0 min = 65 ± 2 beats/min, 60 min = 69 ± 2 beats/min, P < 0.05).

Female Sex Confers Protection against Obesity-Induced Vascular Insulin Resistance in Mice

As shown in Fig. 2A, male and female mice fed an obesogenic diet for 7 mo had increased body weight and adiposity compared with chow-fed mice. Notably, as displayed in Fig. 2B, diet-induced obesity impaired femoral artery insulin-induced dilation in males but not females. Vasodilatory responses to sodium nitroprusside were not significantly affected by diet in either males or females. As demonstrated by an insulin tolerance test (Fig. 2C), both male and female mice fed the obesogenic diet exhibited evidence of impaired whole body insulin sensitivity relative to chow-fed mice.

Figure 2.

Female sex confers protection against obesity-induced vascular insulin resistance in mice. A: body weight and percent body fat in C57BL6 male (n = 12 or 13/group) and female (n = 15–17/group) mice fed chow or an obesogenic diet for 28 wk. B: insulin and sodium nitroprusside (SNP)-induced dilation in isolated femoral arteries from male (n = 8/group) and female (n = 9 or 10/group) mice fed chow or an obesogenic diet. Arteries were preconstricted with phenylephrine (Phe). C: insulin tolerance testing (ITT) in male (n = 11/group) and female (n = 6–11/group) mice fed chow or an obesogenic diet. Data are presented as means ± SE. Data were analyzed using two-way repeated-measures ANOVA or unpaired t test, as appropriate. *P < 0.05. AUC, area under the curve.

Men and Women with Obesity and T2D Exhibit Similar Leg Blood Flow Responses to Insulin Infusion

Plasma insulin and blood glucose concentrations during systemic infusion of insulin in men and women with T2D are presented in Fig. 3A and demonstrate successful achievement of hyperinsulinemia while preventing hypoglycemia. As shown, leg blood flow responses to insulin infusion, as well as changes in leg vascular conductance, were not different between men and women with T2D. Such hemodynamic responses to insulin in men and women with T2D were blunted (P < 0.05) when compared with the responses in healthy subjects. Mean arterial pressure was maintained during insulin infusion in men (0 min = 100 ± 3 mmHg, 60 min = 98 ± 3 mmHg, P = 0.20) and women (0 min = 99 ± 2 mmHg, 60 min = 97 ± 2, P = 0.10). Heart rate was also unchanged during insulin infusion in both men and women (men: 0 min = 75 ± 8 beats/min, 60 min = 75 ± 6 beats/min, P = 0.41; women: 0 min = 70 ± 3 beats/min, 60 min = 71 ± 4 beats/min, P = 0.65).

Figure 3.

Men and women with type 2 diabetes (T2D) exhibit comparable levels of vascular insulin resistance, but diet-induced weight loss is effective at correcting it, particularly in women. A: plasma insulin and blood glucose concentrations after 60 min of systemic insulin infusion with coinfusion of dextrose to prevent hypoglycemia in men (n = 18 or 19) and women (n = 16–18) with T2D. Leg blood flow and vascular conductance responses to 60 min of insulin infusion are presented as percent changes from preinsulin infusion (men, n = 19; women, n = 18). B: self-reported dietary macronutrient composition before the dietary intervention (i.e., pre) and macronutrient composition prescribed for the 2-wk controlled dietary phase in the subset of men (n = 7) and women (n = 9) with T2D who underwent the 6-mo low-energy, low-glycemic dietary intervention. Body weight (percent from starting body weight), HOMA-IR, and HbA1c before and after the 6-mo dietary intervention (men, n = 7; women, n = 9). C: leg blood flow, vascular conductance, and skeletal muscle (vastus lateralis) perfusion responses to 60 min of insulin infusion before and after the 6-mo dietary intervention are presented as percent changes from preinsulin infusion (men, n = 5–7; women, n = 8 or 9). Representative images captured at end diastole during contrast-enhanced ultrasound of the vastus lateralis at minutes 0 and 60 of insulin infusion in a woman with T2D are also displayed. Leg blood flow data for five subjects are not available due to low image resolution for accurate diameter or velocity measurements (confirmed by an independent investigator) or inability to secure intravenous access, thus preventing infusion to be performed. Skeletal muscle perfusion data for five subjects are not available due to low image resolution for perfusion analysis, contraindication to the use of microbubbles, or inability to secure intravenous access. Blood pressure measurements at 60 min of insulin infusion were missing for one subject with T2D. Because mean arterial pressure is needed for calculation of leg vascular conductance, regression analysis was used to predict that missing mean arterial pressure value. Also, for skeletal muscle perfusion, two subjects were deemed outliers based on the robust regression and outlier removal (ROUT) method (Q = 5%) and removed from analysis. D: brachial artery flow-mediated dilation (FMD) and carotid-to-femoral pulse-wave velocity (cfPWV) before and after the 6-mo dietary intervention (men, n = 5–7; women, n = 8). For three subjects, the images were not of sufficient resolution for accurate diameter measurements (confirmed by an independent investigator), and thus, these videos were not processed for FMD analysis. For one subject, the carotid pulse-wave signal was of insufficient quality for detection of the foot of pressure waveform; thus, cfPWV data could not be generated. Data are presented as means ± SE. Data were analyzed using two-way repeated-measures ANOVA or paired and unpaired t test, as appropriate. Nonnormally distributed data (leg blood flow, vascular conductance, and HOMA-IR) were analyzed using the Mann–Whitney U (Wilcoxon rank-sum) and Wilcoxon signed-rank test. *P < 0.05. In C, although data from men and women for leg blood flow, leg vascular conductance, and skeletal muscle (Skm) perfusion are presented in separate subpanels, the dietary effect remained significant for leg blood flow and leg vascular conductance when men and women were combined. In D, the dietary effect on FMD was only significant when men and women were combined. HbA1c, hemoglobin A1c, HOMA-IR, homeostasis model assessment of insulin resistance.

Impaired Insulin-Stimulated Leg Blood Flow in Individuals with T2D Can Be Corrected with Diet-Induced Weight Loss, an Effect Driven by Women

Subject characteristics are summarized in Table 1. Figure 3B depicts energy content and macronutrient composition of the diets before and at the end of the 2-wk controlled dietary phase. This low-energy, low-glycemic diet was effective at reducing body weight, HOMA-IR, HbA1c (Fig. 3B), and other metabolic outcomes including markers of liver function and chronic inflammation (Table 1). Notably, these improvements in body composition and metabolic outcomes caused by the intervention were accompanied by augmented vascular insulin sensitivity, as evidenced by the vascular responses to insulin infusion depicted in Fig. 3C. Specifically, Fig. 3C illustrates that leg blood flow and leg vascular conductance responses to insulin were improved after the 6-mo dietary intervention. When data were split by sex, the effect was significant in women (P < 0.05) but not in men (Fig. 3C). Similarly, skeletal muscle perfusion in response to insulin infusion was also improved with the dietary intervention but only in women (Fig. 3C). Following the dietary intervention, insulin infusion also caused an increase in heart rate (0 min = 65 ± 3 beats/min, 60 min = 67 ± 3 beats/min, P < 0.05), a response not observed at the baseline visit. In addition, improved indices of vascular insulin sensitivity with diet-induced weight loss were also accompanied with an increase in flow-mediated dilation but not by a reduction in pulse wave velocity (Fig. 3D). Brachial artery characteristics during the brachial artery flow-mediated dilation assessments are summarized in Table 2.

Table 2.

Brachial artery characteristics during the flow-mediated dilation assessments before and after the 6-mo dietary intervention

	Before	After
Baseline diameter, mm	3.86 ± 0.20	3.78 ± 0.1
Baseline mean shear rate, s−1	168 ± 14	180 ± 20
Peak diameter, mm	4.08 ± 0.19	4.02 ± 0.1
Time-to-peak dilation, s	103 ± 14	91 ± 12
Shear rate AUC, a.u.	39,601 ± 6,332	44,696 ± 6,920

Data are presented as means ± SE; n = 13. Data were analyzed using paired t test. AUC, area under the curve.

Possible Role of Reduced ET-1 in Mediating the Beneficial Effects of Diet-Induced Weight Loss on the Vasodilatory Actions of Insulin

As illustrated in Fig. 4, the 6-mo dietary intervention caused a reduction in plasma levels of the proinflammatory cytokines IL-12, TNF-α, and MCP-1, as well as a reduction in plasma ET-1. No changes were observed in plasma IFN-γ, IL-10, or IL-6 (Table 1). Also, no changes were observed in plasma nitrite, a by-product of nitric oxide (Table 1). As a follow-up experiment, cultured endothelial cells were treated with pooled plasma collected before and after the dietary intervention (Fig. 5A). As shown, endothelial cells exposed to plasma collected after the dietary intervention produced less ET-1 than cells exposed to plasma collected before the dietary intervention. Endothelial cells treated with postdiet plasma also exhibited greater Akt phosphorylation, relative to cells treated with prediet plasma, although no changes in eNOS protein content were noted (Fig. 5A). We also report that treatment of endothelial cells with a mixture of cytokines (IL-12, TNF-α, and MCP-1) shown to be reduced in plasma with the dietary intervention increased ET-1 production (Fig. 5B). Cytokine treatment of endothelial cells also led to reduced Akt phosphorylation and reduced eNOS expression; however, concentration of nitrite in the cell culture supernatant was not affected by cytokine exposure (Fig. 5B). Finally, in Fig. 5C, we show that blunted insulin-induced dilation in isolated skeletal muscle arteries from female T2D pigs was rescued when arteries were treated with the ET-1 receptor A antagonist BQ-123, thus providing further evidence that excessive ET-1 signaling is a potential key driver of impaired insulin-induced dilation in T2D.

Figure 4.

Diet-induced weight loss reduces circulating inflammatory cytokines and endothelin-1 (ET-1) in type 2 diabetes. Fasting plasma inflammatory cytokines (IL-12, TNF-α, and MCP-1) and ET-1 before and after the 6-mo dietary intervention (men, n = 7; women, n = 9). Data are presented as means ± SE. Data were analyzed using paired t test. Nonnormally distributed data (IL-12, TNF-α, and ET-1) were analyzed using the Wilcoxon signed-rank test. *P < 0.05. MCP-1, monocyte chemoattractant protein-1.

Figure 5.

Possible role of reduced endothelin-1 (ET-1) in mediating the beneficial effects of diet-induced weight loss on the vasodilatory actions of insulin. A: ET-1 concentration in the supernatant of endothelial cells exposed to plasma (30% of final volume of medium) collected before (i.e., pre-) vs. after (i.e., post-) the dietary intervention. Initial concentration of ET-1 in media (with 30% plasma) was subtracted from final concentration in the supernatant to calculate net production of ET-1 by endothelial cells over the 24-h period (n = 10/condition). In cell lysates, p-Akt/Akt and eNOS were determined via Western blotting. Representative images are included to the right. Plasma concentrations of cytokines and ET-1 were as follows: prediet plasma IL-12 = 13.59 pg/mL, MCP-1 = 98.77 pg/mL, TNF-α = 12.49 pg/mL, ET-1 = 4.55 pg/mL; postdiet plasma IL-12 = 9.96 pg/mL, MCP-1 = 71.37 pg/mL, TNF-α = 10.47 pg/mL, ET-1 = 2.31 pg/mL. Cartoon illustrating the experimental design is included on the left. B: ET-1 and nitrite concentration in the supernatant of endothelial cells exposed to a cytokine cocktail (IL-12, TNF-α, and MCP-1; each at 10 ng/mL) for 24 h (n = 11/condition). In cell lysates, p-Akt/Akt and eNOS were determined via Western blotting. Representative images are included to the right. C: insulin-induced dilation in isolated skeletal muscle (Skm) arteries from female healthy or T2D pigs. Arteries were treated with vs. without the ET-1 receptor A inhibitor, BQ-123 (1 µM, 30 min) before (and during) the insulin dose-response curve (n = 4/group and condition). Arteries were preconstricted with U-46619. Data are presented as means ± SE. Data were analyzed using two-way repeated-measures ANOVA or unpaired t test, as appropriate. Nonnormally distributed data (ET-1 from cytokine cocktail experiments) were analyzed using the Mann–Whitney U (Wilcoxon rank-sum). In A and B, we also probed for p-eNOS/eNOS in cell lysates, and no significant treatment effects were found (data not shown). *P < 0.05. AUC, area under the curve; Cc, cytokine cocktail; EC, endothelial cell; MCP-1, monocyte chemoattractant protein-1; NOS, nitric oxide synthase; T2D, type 2 diabetes; V, vehicle; WB, Western blotting.

DISCUSSION

The main findings of the present investigation are several fold. First, we show that nonobese male and female mice exhibit a similar degree of femoral artery insulin-induced dilation. This is consistent with the observation that no sex differences were found in leg blood flow responses to hyperinsulinemia in a large cohort of healthy participants. Second, we provide evidence that female mice are protected against overnutrition and obesity-induced impairments in vasodilatory responses to insulin. Third, we show that women who are obese and present with T2D do manifest vascular insulin resistance to a similar extent as men, yet diet-induced weight loss is effective at enhancing insulin-stimulated leg blood flow (an effect driven by women). Finally, we provide circumstantial evidence that these beneficial effects of diet-induced weight loss may be mediated by reduced ET-1 production, likely secondary to decreases in inflammatory cytokines in circulation. Collectively, data from this investigation indicate that female sex confers protection against obesity-induced vascular insulin resistance and that, in adult obese women with T2D, vascular insulin resistance can be ameliorated with diet-induced weight loss.

Insulin-stimulated blood flow plays an important role in the delivery of glucose and insulin into target tissues such as skeletal muscle, the primary site for glucose disposal (1–5). Here we provide evidence in healthy mice and humans that sex does not influence vascular reactivity to insulin. Accordingly, these data do not support the hypothesis that enhanced whole body insulin sensitivity in healthy females can be, in part, attributed to greater vasodilatory responses to insulin. It is likely that factors such as differences in sex hormones, fat distribution, estrogen signaling in adipose tissue and liver, and skeletal muscle β-oxidation capacity (20–26), among others, may be the major underlying mechanisms driving sex differences in glucose metabolism in healthy individuals.

With development of obesity, insulin resistance in the vasculature can precede the impairments in insulin sensitivity in other tissues involved in glucose metabolism (15, 16). However, the influence of sex in the development and progression of vascular insulin resistance remains largely unknown. In this regard, we recently reported that a short-term (i.e., 10 days) intervention that caused a positive energy balance, through reduced ambulatory activity and increased consumption of sugar-sweetened carbonated beverages, impaired vascular insulin sensitivity in young healthy men but not women (27). On the basis of these findings, our next logical step was to determine the extent to which female sex affords protection against the development of vascular insulin resistance when obesity persists. To address this question, we subjected male and female mice to an obesogenic diet for 7 mo. In line with our previous findings in humans, we show that diet-induced obesity led to impairments in insulin-induced dilation in male but not female mice, further underscoring the notion that a sexual dimorphism exists in the development of vascular insulin resistance triggered by obesity. Undoubtedly, research is warranted to elucidate the mechanisms by which female sex affords such vascular protection. Along these lines, our data suggest that the vascular smooth muscle of female mice fed the obesogenic diet tends to become more responsive to nitric oxide. This could be a compensatory mechanism responsible for preservation for vasodilator actions of insulin in females.

An important observation is that, in humans, once metabolic dysfunction progresses into T2D, the magnitude of impairment in leg blood flow responses to insulin infusion is comparable between men and women, suggesting that the protection afforded by female sex is not sustained. Because vascular insulin resistance in obesity and T2D contributes to impaired glucose disposal and to the pathogenesis of vascular dysfunction and atherosclerotic disease (12, 17–19, 36, 37), efforts are devoted to identifying therapeutic strategies to restore vascular insulin sensitivity. To this end, herein we examined the effects of a 6-mo low-energy, low-glycemic dietary intervention on vascular insulin sensitivity in a small cohort of overweight or obese men and women with T2D. We show this dietary intervention was effective at reducing body weight by 6.34%, HOMA-IR by 1.9 units, and HbA1c by 0.9%, as well as successful at improving markers of liver function and chronic inflammation. Changes in body composition and metabolic outcomes in individuals with T2D were accompanied by improved vasodilatory responses to insulin infusion. Of note, after the dietary intervention, systemic insulin infusion caused an increase in heart rate, a response also observed in young healthy individuals. An increase in heart rate during hyperinsulinemia is likely required to prevent a fall in blood pressure in the face of peripheral vasodilation (28, 30, 31, 38). Accordingly, the increased heart rate during insulin infusion after the dietary intervention may be secondary to improved peripheral vasodilator function. In this regard, improvements in vascular insulin sensitivity with diet-induced weight loss were also paralleled with enhancement of other standard measures of endothelial function, such as flow-mediated dilation. However, no changes in aortic stiffness, as assessed by carotid-to-femoral pulse-wave velocity, were observed. There is precedence in the literature to suggest that lifestyle interventions may not be effective at producing aortic destiffening effects after substantial structural changes in the artery wall and severe stiffness have developed (39).

Notably, on closer examination of the data, it became apparent that the improvement in vascular insulin sensitivity induced by the dietary regimen was particularly driven by women. Although the number of participants was small, it is tempting to speculate that loss of lean body mass with the dietary intervention, which occurred primarily in men, may have contributed to their attenuated improvement in insulin-stimulated leg blood flow. Because the goal of the dietary intervention was to reduce total daily energy intake by 500 kcal, it is also possible that women were in a larger relative energy deficit, and this may be why they exhibited a more homogenous response than men. We expect that findings from this small study will stimulate the conduct of a larger scale randomized clinical trial to confirm that diet-induced weight loss restores vascular insulin sensitivity more profoundly in obese women with T2D compared with men. These future studies should consider subjecting the participants to the same relative energy deficit. Finally, a logical extension of this work could also involve the superimposition of resistance exercise as an adjuvant therapeutic approach to prevent muscle wasting with diet-induced weight loss.

To begin to shed some light into possible mechanisms underlying the improvement in insulin-stimulated leg blood flow with diet-induced weight loss, we assessed circulating levels of the vasoconstrictor peptide ET-1. We found that plasma ET-1 concentrations were reduced after the dietary intervention. Because ET-1 is primarily produced in endothelial cells, next we considered the possibility that changes in the circulating milieu caused by the dietary intervention could have influenced the vascular endothelium and contribute to its reduction in ET-1 production. In support of this, we show that endothelial cells exposed to plasma collected after the dietary intervention produced less ET-1 than cells exposed to plasma collected before the dietary intervention. Although it remains unknown what specific factors in plasma are responsible for the suppressed production of ET-1 following the dietary intervention, it is possible that reduced presence of proinflammatory cytokines plays a role. In this regard, treatment of endothelial cells with a mixture of cytokines (IL-12, TNF-α, and MCP-1), found to be reduced with the dietary intervention, also promoted ET-1 production. To better understand the balance between ET-1 and the nitric oxide system in these in vitro experimental models, we probed for Akt and eNOS in cell lysates. Consistent with the notion that Akt signaling is a determinant of eNOS expression (40), we found that cytokine exposure-induced reduction of Akt phosphorylation was accompanied by a reduction in eNOS protein content. However, cytokine exposure did not reduce nitrite levels in the cell culture supernatant. Also, although treatment of endothelial cells with postdiet plasma increased Akt phosphorylation, these effects did not translate to changes in eNOS. Accordingly, in general, the effects on ET-1 production appeared to be a more consistent finding across experimental models than modulation of eNOS expression and nitrite levels, also not affected by the dietary intervention.

Collectively, we found that ET-1 levels are reduced with diet-induced weight loss, likely owing to changes in the circulating milieu (e.g., reduced proinflammatory cytokines) that drive endothelial cells to produce less ET-1. The reduction of ET-1 in plasma following weight loss could also be secondary to changes in other circulating factors beyond cytokines that result from improved overall metabolic function, including improved liver function and lipid profile. Regardless, reduced endothelial production of ET-1 may be a principal mechanism by which the dietary intervention led to improvements in insulin-stimulated leg blood flow in T2D. Indeed, we show that impaired insulin-induced dilation in isolated skeletal muscle arteries from female T2D pigs can be restored when arteries are treated with the ET-1 receptor A antagonist BQ-123, underscoring the important role of ET-1 in suppressing insulin-induced dilation in T2D.

To our knowledge, this report contains the largest data set in humans involving insulin infusion studies coupled with ultrasound-based measures of blood flow, and these data are strengthened by complementary experiments in cultured endothelial cells and isolated arteries from small and large animals. In aggregate, we show in healthy mice and humans that sex does not influence vascular reactivity to insulin, suggesting that enhanced whole body insulin sensitivity in females is likely not attributed to greater vasodilatory responses to insulin. Furthermore, we provide evidence that female sex confers protection against obesity-induced vascular insulin resistance in mice, and that in obese individuals who present with T2D, impairments in insulin-stimulated leg blood flow can be attenuated with diet-induced weight loss, particularly in women.

DATA AVAILABILITY

Data will be made available upon reasonable request.

GRANTS

This work was supported, in part, by the National Institutes of Health Grants R21 DK116081 and R01 HL142770 (to C.M.-A.), R01 HL137769 and K01 HL125503 (to J.P.), R01 HL151384 (to L.A.M-L. and J.P.), R00 HL130339 and R01 HL153523 (to J.K.L.), and R01 HL112998 (to C.A.E.); American Heart Association postdoctoral fellowship 16POST27760052 (to T.D.O.); an Investigator-Sponsored Research Grant from Lantheus Medical Imaging, Inc (to C.M.-A); and funds from the Margaret W. Mangel Faculty Research Catalyst Fund (to J.K.L. and J.P.).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

C.M.-A., R.N.S., and J.P. conceived and designed research; T.D.O., C.A.E., E.J.P., J.K.L., L.A.M.-L., J.P., R.N.S., J.A.S., L.K.P., K.B., F.I.R.-P., N.J.M., L.F.-S., and N.S. performed experiments; C.M.-A., R.N.S., J.A.S., L.K.P., K.B., F.I.R.-P., N.J.M., L.F.-S., N.S., T.D.O., C.A.E., E.J.P., J.K.L., L.A.M.-L., and J.P. analyzed data; C.M.-A., R.N.S., J.A.S., L.K.P., K.B., F.I.R.-P., N.J.M., L.F.-S., N.S., T.D.O., C.A.E., E.J.P., J.K.L., L.A.M.-L., and J.P. interpreted results of experiments; F.I.R.-P. and L.F.-S. prepared figures; C.M.-A., R.N.S., E.J.P., and J.P. drafted manuscript; C.M.-A., R.N.S., J.A.S., L.K.P., K.B., F.I.R.-P., N.J.M., L.F.-S., N.S., T.D.O., C.A.E., E.J.P., J.K.L., L.A.M.-L., and J.P. edited and revised manuscript; C.M.-A., R.N.S., J.A.S., L.K.P., K.B., F.I.R.-P., N.J.M., L.F.-S., N.S., T.D.O., C.A.E., E.J.P., J.K.L., L.A.M.-L., and J.P. approved final version of manuscript.