Different physiological mechanisms underlie an adverse cardiovascular disease risk profile in men and women

Abstract

Cardiovascular diseases (CVD) affect about one third of the population and are the leading cause of mortality. The prevalence of CVD is closely linked to the prevalence of obesity because obesity is commonly associated with metabolic abnormalities that are important risk factors for CVD, including insulin resistance, pre-diabetes, and type 2 diabetes, atherosclerotic dyslipidemia, endothelial dysfunction and hypertension. Women have a more beneficial traditional CVD risk profile (lower fasting plasma glucose, less atherogenic lipid profile) and a lower absolute risk for CVD than men. However, the relative risk for CVD associated with hyperglycemia and dyslipidemia is several-fold higher in women than in men. The reasons for the sex differences in CVD risk associated with metabolic abnormalities are unclear but could be related to differences in the mechanisms that cause hyperglycemia and dyslipidemia in men and women, which could influence the pathogenic processes involved in CVD. In this article we review the influence of a person’s sex on key aspects of metabolism involved in the cardiometabolic disease process, including insulin action on endogenous glucose production, tissue glucose disposal, and adipose tissue lipolysis, insulin secretion and insulin plasma clearance, postprandial glucose, fatty acid, and triglyceride kinetics, hepatic lipid metabolism, and myocardial substrate use. We conclude that there are marked differences in many aspects of metabolism in men and women that are not all attributable to differences in the sex hormone milieu. The mechanisms responsible for these differences and the clinical implications of these observations are unclear and require further investigation.

Introduction

Cardiovascular diseases (CVD), including atherosclerosis, hypertension, myocardial infarction, stroke, and heart failure, affect ~10% of young and middle-aged (<65years) and ~30% of older (≥65years) adults and are the leading causes of mortality, accounting for >25% of all deaths ^(1–3). More than 80% of CVD-related deaths are due to ischemic heart disease and stroke ⁽³⁾. The prevalence of CVD is closely linked to the prevalence of obesity because obesity is commonly associated with metabolic abnormalities that are important risk factors for CVD, including insulin resistance, pre-diabetes, and type 2 diabetes (T2D), atherosclerotic dyslipidemia, endothelial dysfunction and hypertension ^(4–13). Insulin is a key regulator of glucose and lipid metabolism and also regulates sympathetic nerve activity and endothelial function; accordingly, resistance to the effects of insulin is a key pathogenic mechanism involved in CVD ⁽¹⁴⁾. In addition, increases in plasma glucose and triglycerides (TG) per se are directly involved in causing the cellular pathogenic changes associated with hypertension and atherosclerosis ^{(15, 16)}.

Premenopausal women have a more beneficial traditional CVD risk profile (lower fasting plasma glucose ^(17–22) and less atherogenic lipid profile, characterized by lower plasma TG and apolipoprotein B [apo-B] containing particles, higher HDL-cholesterol [HDL-C], and more large and fewer small HDL particles ^{(19, 23)}) and a lower absolute risk for CVD than men. The observed sex differences in the metabolic CVD risk profile are attributed to the sex hormone milieu, particularly the protective effect of estradiol, but it is becoming clear that chronological age per se has a major and possibly greater influence on cardiometabolic function in women than menopause ^(24–28). In addition, the relative risk for CVD associated with hyperglycemia and dyslipidemia is several-fold higher in women than in men ^{(5, 7, 10, 12, 29–33)}. A meta-analysis of 64 studies, including a total of 858,507 people, found the relative risk for CVD associated with T2D is ~45% greater in women than in men ⁽²⁹⁾; another, smaller meta-analysis, which focused on young and middle-aged (<60 years old) adults only, found the relative risk for CVD associated with T2D is approximately three times as high in women than men ⁽³⁰⁾. Moreover, women with atherosclerotic dyslipidemia (hypertriglyceridemia and/or low HDL-C) have a two to four times greater risk for CVD than women with normal plasma lipids whereas atherosclerotic dyslipidemia increases the risk for CVD by only 25%−50% in men ^{(5, 7)}. The reasons for the sex differences in absolute CVD risk and CVD risk associated with increased glucose and TG concentrations are unclear but could be related to differences in the mechanisms that cause hyperglycemia and dyslipidemia in men and women, which could influence the pathogenic processes involved in CVD. For example, the CVD risk associated with increased plasma TG concentration is not simply determined by the total amount of TG but dependent on the number of circulating TG-containing lipoprotein particles at any given TG concentration (i.e., lots of small TG-poor vs few large, TG-rich particles), and the T2D risk associated with impaired glucose tolerance depends on the shape of the plasma glucose profile after glucose ingestion, which is likely determined by variations in insulin secretion, plasma clearance, and target tissue action ^(34–40). In this article, we will review and highlight important differences in insulin kinetics and action, basal and postprandial glucose and lipid metabolism, and myocardial substrate use between men and women.

Regulation of plasma glucose and TG concentrations

Plasma glucose concentration is maintained by a balance between hepatic, and to a lesser extent, renal glucose production, meal glucose appearance in plasma, and tissue glucose uptake. Insulin is a major regulator of endogenous glucose production and tissue glucose uptake (see ^(41–43) for excellent and detailed reviews). Insulin suppresses endogenous glucose production, both by acting directly on hepatocytes, and indirectly by inhibiting glucagon production and adipose tissue lipolysis ⁽⁴³⁾. Endogenous glucose production is very sensitive to the inhibitory effect of insulin and small increases in plasma insulin above basal values are sufficient to completely suppress it ^(44–46). Insulin stimulates tissue (predominantly muscle) glucose uptake in a dose dependent manner and the maximal stimulatory effect of insulin on glucose disposal far exceeds the normal postprandial rise in plasma insulin ⁽⁴⁴⁾. Insulin is also a potent inhibitor of adipose tissue lipolysis and fatty acid release into plasma, and small increases in plasma insulin above basal values are sufficient to completely suppress it ^{(45, 47, 48)}. Insulin also regulates hepatic TG synthesis and secretion, both directly and indirectly by regulating adipose tissue lipolysis. Insulin stimulates hepatic de novo lipogenesis, inhibits VLDL-particle (apoB-100) and TG secretion, and regulates the availability of adipose-derived fatty acids for hepatic TG synthesis ^{(41, 49)}. In healthy people, insulin secretion, plasma insulin clearance, and insulin sensitivity are tightly coordinated and both insulin secretion and insulin clearance often change simultaneously in opposite directions to compensate for changes in insulin sensitivity; relative insulin insufficiency due to an imbalance among insulin secretion, plasma clearance, and sensitivity causes an increase in plasma glucose, fatty acid, and TG concentrations, and ultimately pre-diabetes, T2D, and atherosclerotic dyslipidemia ^(50–52).

Basal plasma glucose concentration and flux in men and women

Plasma glucose concentration after an overnight fast is generally slightly (~10%) lower in women than in men ^{(17–22, 53)}, but it is unclear whether this is due to less glucose production or more efficient plasma clearance in women than in men. The results from studies that evaluated basal endogenous glucose production are equivocal. In most studies basal endogenous glucose production, expressed per kg body weight or per kg fat-free mass, was not different in men and women, irrespective of adiposity status and age ^(54–62). However, in some studies, basal endogenous glucose production, expressed per kg body weight or fat-free mass was less ^{(18, 63)} and in others it was greater ^{(64, 65)} in women compared wtih age-matched men. The reasons for the differences in results among studies are unclear but are likely related to differences in the prevailing plasma insulin concentration (because insulin is a potent inhibitor of endogenous glucose production ⁽⁴⁵⁾) and the duration of fasting, which affects hepatic glucose production differently in men and women ^{(55, 66)}.

Basal plasma free fatty acid (FFA) concentration and flux in men and women

Plasma FFA concentration after an overnight fast is generally greater in women than men ^{(53, 67–69)}. The difference in FFA concentration is largely due to the greater fat mass relative to fat-free mass, not differences in adipose tissue lipolytic activity and/or plasma clearance rate (reviewed below), in women than men. We measured FFA appearance rate (Ra) in plasma, an index of adipose tissue lipolytic activity ⁽⁷⁰⁾, in lean, overweight, and obese (including severely obese) men and women and found basal FFA Ra in plasma, is directly related to fat mass, and the relationship between fat mass and FFA appearance in plasma is not different in men and women ⁽⁶⁸⁾. However, FFA Ra in relationship to fat-free mass, or unit of plasma volume, or resting energy expenditure is approximately 50% greater in women than in men ^{(68, 71, 72)} because women have more fat mass than men for any given amount of fat-free mass ⁽⁷³⁾, and fat-free mass is the primary determinant of resting energy expenditure ^{(74, 75)}.

Insulin action on glucose metabolism in men and women

Potential sex differences in insulin action on glucose metabolism have been evaluated by using the homeostasis model assessment of insulin resistance (HOMA-IR) ^{(e.g.,76, 77)}, the oral glucose tolerance test (OGTT) ^{(e.g.,18, 78, 79)}, the intravenous glucose tolerance test (IVGTT) ^{(e.g.,21, 22)} and the gold-standard hyperinsulinemic-euglycemic clamp technique, with or without simultaneous glucose tracer infusion ^{(e.g.,18, 47, 54, 57, 58, 80–83)}. We focus on the results from studies that used the hyperinsulinemic-euglycemic clamp procedure in conjunction with glucose tracers (stable isotope- or radiolabeled) to distinguish the effects of insulin on glucose production and glucose disposal (not those that only report the “M-value”, i.e., the glucose infusion rate during the clamp) and those that used the arterio-venous balance technique or dynamic PET imaging to provide a direct measure of tissue glucose uptake rates. The HOMA-IR and the IVGTT-derived insulin sensitivity indices do not provide direct information about the effect of insulin on organ-specific glucose kinetics and the OGTT provides a standard 75 g dose of glucose to subjects regardless of body size, which makes the interpretation of the results difficult because women are generally smaller than men ^{(79, 84, 85)}.

Insulin action on endogenous glucose production.

Endogenous glucose production is very sensitive to changes in plasma insulin and even small increases in plasma insulin concentration above values observed after an overnight fast can almost completely inhibit it ^{(45, 86)}. A study that used a relatively low-dose insulin infusion rate that sub-maximally suppressed endogenous glucose production found endogenous glucose production was more sensitive to the inhibitory effect of insulin in women than men (greater relative suppression in women) ⁽⁵⁷⁾. Several other studies evaluated the effects of higher (near maximally suppressive) doses of insulin on endogenous glucose production and found near maximally suppressed endogenous glucose production rates were not different in men and women ^{(18, 54, 58)}.

Insulin action on glucose disposal.

Comparing whole body glucose disposal rates in men and women is difficult because of differences in body size and body composition in men and women. In healthy lean men, skeletal muscle accounts for the majority (>75%) of whole body insulin-stimulated glucose disposal ^{(86, 87)}. However, both muscle and adipose tissue are highly sensitive to insulin ^(88–90) and insulin-stimulated tissue glucose uptake rates in various adipose tissue depots range from 25% to >50% the rates measured in muscle ⁽⁶³⁾. Accordingly, the contribution of adipose tissue to total (whole body) glucose disposal depends on a person’s adiposity. Whole body insulin-stimulated glucose disposal rate expressed per kg fat-free or lean body mass and adjusted for plasma insulin concentration was often not different in men or age-matched (young or older) women ^{(47, 57, 81)} but glucose uptake rate per leg lean mass (AV-balance technique) or uptake into muscle (assessed by using dynamic PET imaging) was greater in lean women than lean age-matched men ^{(54, 82, 83)}.

FFA-induced insulin resistance of glucose metabolism in men and women

Plasma FFA are important negative regulators of insulin action in liver and muscle. An experimentally-induced (intravenous lipid and heparin infusion) increase in plasma FFA concentration before and during a hyperinsulinemic-euglycemic clamp impairs insulin action in liver and muscle in a dose-dependent manner ^(91–94). The adverse effect of FFA on insulin action lasts for almost 4 h after cessation of lipid infusion ⁽⁹⁵⁾. The observed greater insulin sensitivity of both endogenous glucose production ⁽⁵⁷⁾ and muscle glucose disposal ^{(54, 82, 83)} in women compared with men is therefore intriguing considering basal FFA release from adipose tissue in relationship to fat-free mass is markedly greater in women than in men ^{(68, 71, 72)}. Several studies therefore tested the susceptibility of men and women to FFA-induced insulin resistance. In some studies, women were less susceptible to FFA-induced insulin resistance of glucose disposal ^{(54, 58)}, whereas others reported no sex difference in FFA-mediated insulin resistance ^{(94, 96)}; however, this could have been due to statistical power because a trend for a lesser impairment in women than men (46% vs 60% impairment) was observed ⁽⁹⁶⁾. Only one study evaluated the effect of increased plasma FFA concentration on insulin-mediated suppression of endogenous glucose production and found FFA impaired it similarly in men and women ⁽⁵⁸⁾.

Insulin action on adipose tissue lipolysis in men and women

Adipose tissue is very sensitive to the antilipolytic effect of insulin ⁽⁴⁸⁾, so even small differences in plasma insulin concentration can have marked effects on FFA appearance in plasma. The results from studies that evaluated the effect of sex on insulin-mediated suppression of FFA release into the circulation are inconsistent and difficult to interpret because different doses of insulin were used and plasma insulin concentrations were either not reported or markedly (~30%) different in men and women ^{(47, 97, 98)}. However, one of these studies evaluated the dose response relationship between plasma insulin concentration and FFA rate of appearance in plasma in lean and overweight and obese men and women and found the half-maximum effective (EC₅₀) insulin concentration was not different in men and women but greater in obese than non-obese subjects ⁽⁴⁷⁾, suggesting no sex differences in insulin sensitivity of adipose tissue lipolysis but obesity-associated insulin-resistance in both men and women.

Insulin secretion in men and women

A large cohort study that included 380 healthy young subjects found plasma C-peptide concentration (an index of insulin secretion) after an overnight fast was greater in women than men ⁽²¹⁾. Potential sex differences in glucose-stimulated insulin secretion have been evaluated by using both intravenous and oral glucose tolerance tests. During the IVGTT, a body weight-adjusted dose of glucose is provided, whereas the same standard dose of glucose (75 g) is given to everyone during the OGTT, which makes the interpretation of the results from OGTTs difficult, because women are generally smaller than men ^{(79, 84, 85)}. The acute C-peptide response to an intravenous glucose challenge was not different in men and age-matched women ⁽²¹⁾, but the acute insulin response was greater in women than men ^{(21, 22)}, suggesting similar glucose-induced insulin secretion but impaired insulin clearance in women compared with men (reviewed in more detail below). The interpretation of the results from studies that evaluated insulin secretion after mixed meal ingestion ^{(64, 80)} is complicated because different meals were used in different studies and meal energy and carbohydrate contents were not always adjusted for differences in body weight and energy expenditure in men and women. One study provided a body weight adjusted meal (10 kcal/kg and 1.2 g dextrose/kg) to both young and older men and women ⁽⁶⁴⁾ and found the early rise in plasma C-peptide was not different in women and men, but women had slightly higher C-peptide concentrations during the later postprandial period (~60 min after starting the meal).

Insulin clearance in men and women

The effect of sex on plasma insulin clearance is unclear because of conflicting results from different studies. A study that used the hyperinsulinemic-euglycemic pancreatic clamp technique in conjunction with arterial and hepatic vein blood sampling in young and older adults found whole body insulin clearance was greater in women than in men, and this was due to greater non-splanchnic insulin clearance in women whereas hepatic/splanchnic insulin clearance was lower in women than men ⁽⁹⁹⁾. Another study reported impaired steady-state insulin clearance during a hyperinsulinemic-euglycemic clamp in women compared with men, but insulin clearance was calculated as the insulin infusion rate divided by plasma insulin concentration ⁽⁸⁰⁾, which ignores residual endogenous insulin secretion during the clamp ⁽¹⁰⁰⁾. Studies that used a mathematical modelling approach to estimate whole body and regional plasma insulin clearance after mixed meal ingestion found postprandial non-splanchnic insulin clearance was greater in young Caucasian women than men and splanchnic insulin clearance was significantly less or tended to be less in women than men ⁽⁶⁴⁾; however, in young Asian and older Caucasian subjects, plasma insulin clearance rates were not different in women and men ^{(64, 80)}. Data obtained during an IVGTT, suggests impaired insulin clearance in women compared with men because the acute C-peptide response, which provides a measure of insulin secretion, was not different in men and women but insulin concentration was greater in women than men ⁽²¹⁾.

Postprandial glucose kinetics in men and women

Postprandial glucose kinetics in young and older men and women were evaluated by using a triple tracer mixed meal metabolic testing protocol ⁽⁶⁴⁾. The meal provided 10 kcal/kg and contained 1.2 g dextrose/kg. Endogenous glucose production was rapidly and nearly completely suppressed during the first 60 min after meal ingestion and then returned to basal values in both men and women (both young and old). However, meal glucose appearance in plasma was faster in women than in men (both young and old). Differences in glucose absorption in men and women have also been observed during an OGTT ⁽¹⁸⁾, but the results cannot be directly compared with the meal test or among men and women because both men and women received 75 g of glucose during the OGTT, so women received much more glucose relative to their body weight and metabolic rate than men.

Postprandial fatty acid kinetics in men and women

Postprandial endogenous and meal fatty acid appearance in plasma in men and pre-menopausal women has been evaluated by using a dual tracer (oral and intravenous) mixed meal testing protocol ⁽¹⁰¹⁾. Meal ingestion suppressed FFA rate of appearance in plasma rapidly and nearly completely for almost 4 h in both men and women whereas meal-derived fatty acid appearance in plasma tended to be greater in men than women ⁽¹⁰¹⁾. Postprandial lipemia and the organ distribution and metabolic fate of FFA entering the systemic circulation from adipose tissue lipolysis and meals are markedly different in men and pre-menopausal women. After an overnight fast, a smaller proportion of plasma FFA flux is oxidized to CO₂ in women than men ⁽¹⁰²⁾, even though women convert plasma FFA more rapidly to readily oxidized ketones ⁽¹⁰³⁾. The greater non-oxidative disposal of FFA in women appears to be targeted to adipose tissue because a greater proportion of both plasma FFA and meal-derived fatty acids are stored in subcutaneous adipose tissue in women than in men (~25% vs ≤10%, respectively) whereas uptake into liver, muscle, and visceral fat after mixed or high fat meal ingestion is not different in men and women ^(104–108). The postprandial increase in plasma TG after consuming a mixed or high fat meal is less in women than men, even though the same amount of meal fat is oxidized in women and men ^(108–110) and less meal fat is cleared by splanchnic tissues in women than in men ⁽¹¹¹⁾. The difference in postprandial lipemia between women and men was observed regardless of whether or not the meal was adjusted for individual subject’s energy needs and therefore smaller relative to body weight in men than women. These results suggest markedly impaired peripheral TG clearance after meal intake in men compared with women. In addition, it was found that adding carbohydrates to an oral lipid load decreased postprandial lipemia in women but not in men ⁽¹¹²⁾. The differences in postprandial lipid metabolism in men and pre-menopausal women are at least in part due to differences in the sex hormone milieu ^(113–115). However, an independent effect of chronological age on postprandial lipemia has also been observed and it was as pronounced, if not more pronounced than that of menopause ⁽²⁵⁾. Moreover, subcutaneous adipose tissue fatty acid storage is even greater in postmenopausal than premenopausal women ⁽¹¹⁶⁾, suggesting the observed sexual dimorphism in adipose tissue fatty acid storage is not due to differences in female sex steroids.

Basal hepatic lipid metabolism in men and women

In a series of studies, we evaluated VLDL-TG and VLDL-apoB-100 kinetics by using stable isotope labeled tracer techniques in conjunction with compartmental modeling analysis in lean and obese men and women. The results from these studies revealed that a person’s sex affects the kinetics of both the particle (apo-B100) per se and the TG moiety of particles, often independently suggesting differences in the lipid load of particles. The differences between men and women are not only due to differences in the sex hormone milieu and are dependent on subjects’ adiposity status. We found: i) lean young women produce fewer but TG-richer VLDL particles than men ^{(69, 72, 117)}, ii) ovarian hormone deficiency after menopause increases VLDL-TG but not VLDL-apoB-100 (VLDL particle) secretion rate ⁽¹¹⁸⁾, iii) testosterone treatment has no effect on VLDL-TG and VLDL-apoB-100 kinetics, but estradiol given to postmenopausal women with obesity stimulates VLDL-TG plasma clearance ^{(119, 120)}, iv) increased VLDL-TG concentrations in obese compared with lean men results from over-secretion of VLDL-TG whereas increased VLDL-TG concentrations in obese compared with lean women results in part from VLDL-TG over-secretion but mostly from impaired VLDL-TG removal from plasma ^{(69, 117)}, and v) obese women, but not obese men, are resistant to the inhibitory effects of combined hyperglycemia-hyperinsulinemia on hepatic VLDL-TG secretion whereas no differences in the hyperglycemia-hyperinsulinemia induced suppression of VLDL-TG secretion was observed in lean men and women ⁽¹²¹⁾. A higher VLDL-TG secretion rate in women with abdominal obesity compared with lean women was also observed by others ⁽¹²²⁾ and was mostly due to an increase in the secretion of large and to a lesser extent small VLDL (50% and 12% increase, respectively). VLDL-TG plasma clearance rate in that study was also ~15%−20% less in obese compared with lean women, but the difference did not reach statistical significance ⁽¹²²⁾. In addition, it was found that menopause increased hepatic TG secretion specifically in the small VLDL fraction, and decreased or tended to decrease the secretion of both small and large VLDL particles ⁽¹²²⁾. The observed differences in VLDL-TG secretion rate between men and women are most likely due to differences in the incorporation of systemic plasma fatty acids into VLDL-TG ⁽⁷²⁾, rather than differences in hepatic de novo lipogenesis (DNL) ⁽¹¹⁰⁾.

Effect of fructose ingestion on hepatic de novo lipogenesis in men and women

Fructose stimulates hepatic DNL and high fructose consumption is associated with hepatic steatosis and hypertriglyceridemia ^(123–125). Consumption of a high- compared with a low-fructose drink (containing 100 g sugar with either 60% or 20% fructose) significantly increased postprandial hepatic DNL in women (peak DNL ~20% vs ~7%) but not in men (~7% after both meals) ⁽¹²⁶⁾. This suggests women are more susceptible to fructose-induced hepatic steatosis and non-alcoholic fatty liver disease. It is worth noting that the stimulatory effect of fructose on DNL is most likely a secondary phenomenon because very little fructose is directly converted to fatty acids and fructose-to-fatty acid conversion was only observed in men but not in women ⁽⁶⁵⁾. This is consistent with recent findings that suggest fructose metabolism occurs predominantly in the small intestine, where it is converted to glucose, lactate, and glycerol ⁽¹²⁷⁾.

Myocardial substrate utilization in men and women

A series of elegant studies that used dynamic PET imaging have demonstrated marked differences in myocardial substrate use in men and women. Myocardial oxygen consumption is greater in healthy lean women than healthy lean men and women’s hearts use less glucose and fewer dietary fatty acids as a source of energy than men ^{(106, 128, 129)}. Obesity reduces myocardial glucose uptake and oxidation in men, but not in women ⁽¹²⁹⁾. Insulin-stimulated myocardial glucose uptake rate, on the other hand, is not different in healthy young men and women ⁽⁸³⁾. These findings could have important clinical implications, because myocardial perfusion and fuel use are directly linked with cardiac function ^{(130, 131)}.

Summary and Conclusion

There are marked differences in many aspects of glucose and lipid metabolism in men and women. Women compared with men: i) are more sensitive to the inhibitory effect of insulin on glucose production and the stimulatory effect of insulin on muscle glucose disposal, ii) have greater adipose tissue FFA release relative to fat-free mass and resting energy and are less susceptible to the adverse effect of FFA on insulin action in muscle, iii) have altered meal glucose absorption kinetics, possibly due to different gastric emptying rates ⁽¹³²⁾, iv) have greater basal and postprandial non-oxidative fatty acid disposal and fatty acid storage in adipose tissue and reduced postprandial lipemia, and v) are more susceptible to fructose-induced DNL. Moreover, hepatic and plasma lipid metabolism is markedly affected by sex and the observed metabolic differences between men and women depend on subjects’ adiposity and age. On the other hand, no major differences between men and women have been observed for the antilipolytic effect of insulin and acute glucose-induced insulin secretion. The effect of sex on plasma insulin clearance is unclear because of conflicting results from different studies. We conclude that sex needs to be considered when interpreting data reported in the literature and planning new studies. Carefully designed studies are needed to determine the mechanisms responsible for the observed sexual dimorphism in metabolism and to disentangle the effects of chronological and biological (pre/post menopause) age on metabolism in women.