Dominic N. Reeds, MD, Geriatrics and Nutritional Science and the Center for Human Nutrition, Washington University School of Medicine, Box 8031, 660 South Euclid, St Louis, MO 63110 E-mail: firstname.lastname@example.org
Highly active antiretroviral therapy (HAART) has markedly prolonged life expectancy in patients infected with the human immunodeficiency virus (HIV). Use of HAART has been associated with development of a metabolic syndrome that may increase the risk of developing ischemic heart disease. The purpose of this review is to discuss the incidence and clinical manifestations of the metabolic syndrome in patients with HIV infection and contributing factors, as well as whether cardiovascular risk is disproportionately elevated in patients with HIV-associated metabolic syndrome.
Highly active antiretroviral therapy (HAART) has transformed infection with the human immunodeficiency virus (HIV) from a rapidly progressive and uniformly fatal disease into a chronic manageable condition. Life expectancy following the diagnosis of HIV infection is now more than 35 years.1 Enthusiasm for the use of HAART has been tempered by the emergence of a cluster of metabolic disorders that have many features of the metabolic syndrome (MS): increased waist circumference, impaired fasting glucose, hypertriglyceridemia, low plasma high-density lipoprotein (HDL) concentration, and hypertension.2 Almost 20% of patients with HIV infection who are receiving treatment meet criteria from the Third Report of the Adult Treatment Panel for MS, and about 50% have 2 of the 5 diagnostic criteria.3
In patients without HIV infection, MS is associated with a 2-fold increase in risk of ischemic heart disease4; however, it is unclear whether this is true of the HIV-associated MS (HIV-MS). At first glance, HIV-MS appears to be no different from MS in HIV-seronegative persons. Surrogate markers, however, suggest that HIV-MS may be more proatherogenic than MS because of differences in the manifestations of the diagnostic criteria. This review will discuss differences in the phenotype of MS between patients with and without HIV infection. The findings of trials assessing the risk of ischemic heart disease in patients with HIV infection will also be discussed.
Adipose Tissue Distribution and Function
Central adiposity and elevated waist:hip ratio are common in patients with HIV infection.5 In MS, both the subcutaneous and visceral adipose tissue depots are enlarged. In contrast, individuals with HIV-MS lose subcutaneous adipose tissue and preserve or gain trunk fat mass.6 As a result, for any given amount of trunk fat, patients with HIV-MS will have a larger proportion of visceral fat. Visceral adipose tissue is more resistant to the antilipolytic effects of insulin,7 resulting in increased free fatty acid (FFA) release and elevated portal venous fatty acid concentration, which may promote both hepatic steatosis and insulin resistance. Increased portal vein FFA concentration directly stimulates very low-density lipoprotein-triglyceride (VLDL-TG) secretion by the liver,8 raising plasma triglyceride concentration and lowering plasma HDL levels. Spillover of fatty acids released from visceral fat depots may accumulate in skeletal muscle, resulting in intramyocellular fat deposition and peripheral insulin resistance.9 Visceral fat accumulation is also associated with greater circulating levels of proinflammatory cytokines including C-reactive protein. Adipocyte dysfunction and reduced production of insulin-sensitizing adipokines, such as adiponectin, may also promote insulin-resistant glucose metabolism in HIV-MS.10 In combination, these data suggest that for any given waist:hip ratio, patients with HIV-MS may have dysfunctional adipocytes, more visceral fat, and a higher FFA concentration, which in combination exacerbate dyslipidemia and glucose intolerance and increase circulating levels of proinflammatory, proatherogenic cytokines (Figure).
HIV infection itself is associated with low plasma HDL and elevated plasma triglyceride concentration.11 Hypertriglyceridemia is strikingly common in HIV-MS, approaching 50% of all patients.12 Dyslipidemia in patients infected with HIV is indicative of insulin resistance for glucose and lipid metabolism and of hepatic steatosis.13 The pathogenesis of hypertriglyceridemia in HIV-MS appears to arise predominantly from increased hepatic secretion of VLDL-TG rather than reduced clearance.14 Hepatic FFA availability is the major regulator of VLDL-TG secretion,15 and these FFA derive from 3 main sources: (a) plasma FFA from systemic (ie, subcutaneous fat) or nonsystemic pools (ie, visceral fat), (b) intrahepatic triglyceride stores (ie, hepatic steatosis), and (c) de novo lipogenesis. There is evidence that all 3 factors contribute to dyslipidemia in patients with HIV infection. De novo lipogenesis,16 resting lipolytic rate,17 and hepatic triglyceride stores are increased in HIV-associated hypertriglyceridemia.13 In combination, these factors result in dramatic elevations in VLDL-TG secretion in patients infected with HIV. Suppression of lipolysis, using niacin or the niacin analog acipimox, reduces plasma triglyceride levels by 20% to 30% in persons with HIV-MS and MS, suggesting that increased FFA availability contributes to dyslipidemia in both patient populations.18,19 Triglyceride clearance may also be impaired in HIV-MS because of reduced lipoprotein lipase activity20; however, this appears to play a minor role.
Impaired glucose tolerance and hyperinsulinemia are strongly associated with ischemic heart disease in the general population. Fasting blood glucose concentration is determined primarily by insulin-mediated suppression of liver glucose output. Monitoring of fasting blood glucose concentration is sensitive for predicting insulin-resistant glucose metabolism in HIV-negative populations but may be insensitive in patients positive for HIV-MS. Indeed, one study found that in normoglycemic patients with HIV-associated fat redistribution, oral glucose tolerance testing revealed impaired glucose tolerance in 35% and type 2 diabetes in 7%, compared with 5% and 0.5%, respectively, in matched participants in the Framingham Offspring Study.21 What is the reason for this discrepancy? Antiretroviral therapy, in particular the use of selected protease inhibitors (PIs), reduces insulin-stimulated glucose uptake22 and impairs pancreatic β-cell function.23 This could explain the clinical observation of fasting normoglycemia but postprandial hyperglycemia in persons with HIV-MS.
It also appears that the nature of the defects in muscle metabolism in HIV-MS is fundamentally different from that in MS because a number of differences exist between the 2 disease states. In contrast to patients with type 2 diabetes, persons with HIV-MS have elevated whole-body proteolytic rates that fail to suppress with insulin infusion,24 indicative of a shared signaling defect in amino acid and glucose metabolism. Further, during exercise, fatty acid uptake and oxidation by skeletal muscle is impaired in patients with HIV-MS,25 possibly from mitochondrial toxicity associated with antiretroviral drugs. Reduced oxidation of intramyocellular lipids may result in accumulation of toxic lipid species, contributing to insulin-resistant glucose and amino acid metabolism. Reducing plasma FFA concentration with acipimox dramatically improves insulin action in skeletal muscle but does not affect hepatic glucose output in patients with HIV-MS.26 Taken together, these data suggest that elevated plasma FFA, accumulation of intramyocellular lipids, and impaired myocellular fatty acid oxidation play a more pathogenic role in insulin-resistant glucose metabolism in HIV-MS than in MS.
Ischemic heart disease is the leading cause of death in patients with MS. The process of atherosclerosis begins early in life and as a result, processes that affect vascular endothelium early in life may have a profound effect on the risk of developing ischemic heart disease as patients enter middle age. Patients with HIV-MS experience a period of untreated viremia followed by exposure to potentially toxic antiretroviral drugs. As a result, the effects of HIV replication and antiretroviral therapy on vascular endothelium and their interaction with the manifestations of MS are critical determinants for the risk of developing ischemic heart disease. Endothelial function is widely accepted as a preclinical marker for elevated cardiovascular risk in patients with-out HIV infection. HIV infection may directly impair endothelial function27; however, some studies have found that antiretroviral therapy and dyslipidemia may worsen endothelial function.28 Long-term HAART, especially PI-based therapy, may ultimately contribute to atherosclerotic events. McComsey and colleagues29 found that HIV-infected children who are receiving HAART have greater carotid intima-media thickness (CIMT) than uninfected controls. Curiously, duration of HAART, rather than traditional risk factors for cardiovascular disease (CVD), correlated most closely with CIMT in children. Early studies in adults found use of PI to be associated with elevated CIMT,30 although recent studies refute this finding.31 Together, these studies suggest that in patients infected with HIV, the virus and possibly selected antiretroviral medications may further increase CVD risk through direct effects on endothelial function.
In the late 1990s, HIV-related mortality declined by 60% in the United States. Unfortunately, between 1999 and 2004, deaths from non-HIV-related causes rose by ≈33%, of which ≈25% were due to CVD.32 Initial case series suggested that HAART use, in particular PI-based HAART, may be associated with increased cardiovascular risk. Retrospective studies have provided conflicting results, reporting either increased,33 decreased,34 or unchanged35 rates of ischemic heart disease end points in patients receiving HAART. In 2003, initial reports from a prospective cohort of more than 23,000 patients receiving HAART found the risk of myocardial infarction (MI) was increased by 26% over 4 to 6 years of HAART use. Further, the risk of MI was closely related to the duration of therapy.36 Recently, after a median follow-up period of 4.5 years, patients receiving PI-based therapy within this cohort had a lower, but still elevated, relative risk of MI even after adjustment for CVD risk factors including lipid levels.37 Recent data also show that HIV-infected patients receiving HAART have an approximately 2-fold increased risk of hospital admission for ischemic heart disease, but this risk did not increase with duration of HAART.38 These findings suggest that HAART directly increases the risk of ischemic heart disease and that this risk may rise with longer duration of therapy. Thus, it would be expected that conservative therapy that limited drug exposure could reduce these risks. The Strategies for Management of Antiretroviral Therapy (SMART) Study Group39 evaluated the effect of intermittent therapy for 4 years in 5400 patients infected with HIV, hypothesizing that intermittent therapy would confer antiviral benefits but limit metabolic and cardiovascular toxicity. Participants received either continuous HAART (viral suppression group) or intermittent, CD4+ cell count-guided HAART (drug conservation group). Curiously, the hazard ratio for fatal or nonfatal CVD was greater in the drug conservation group, suggesting that (1) drug toxicity is unrelated to CVD in HIV infection, (2) intermittent therapy itself contributes to CVD, or (3) inadequately treated HIV infection increases CVD risk. In summary, HAART in and of itself appears to confer a slightly increased risk of CVD; however, to date, no prospective studies have evaluated the interaction of HIV-MS with CVD risk.
Why the Lack of Evidence?
At present, there are no compelling data demonstrating an increased CVD risk in patients with HIV-MS, probably because of the inherent limitations of studies performed to date. Atherosclerosis takes many years to develop following acquisition of CVD risk factors, and HAART has only been available for about 10 years. Hence, it is likely that epidemiologic differences in CVD rates will emerge but are not yet apparent because of the limited follow-up period. In addition, obtaining appropriate groups for comparison with HIV-infected populations remains challenging. Reference databases may not be well matched for HIV-infected populations because of differences in socioeconomic status, lifestyle, and rates of tobacco or recreational drug use.
While prospective data are currently lacking, surrogate markers are highly suggestive of increased CVD risk in HIV-MS. It is likely that CVD risk will worsen as a result of the emerging epidemic of obesity in HIV-infected persons.40 Weight loss and increased physical activity improve the metabolic and CVD risk profiles in the general population with MS. In HIV-MS, data now support the use of conservative measures such as resistance and aerobic exercise training41 and lifestyle modification including use of a low-fat diet42 to improve CVD risk. Conceptually, it seems logical to recommend weight loss to patients with HIV-MS. However, there is currently insufficient evidence to support this recommendation. Further safety and efficacy studies of weight loss in patients with HIV-MS need to be conducted because, at least in some settings, obesity may be protective against progression of HIV infection.43
The author was supported in part by NIH K23 -RR019508.