- Email: [email protected]
Premature cardiovascular disease in chronic renal failure
REVIEW
Review
Premature cardiovascular disease in chronic renal failure Colin Baigent, Kate Burbury, David Wheeler There is a remarkable lack of reliable information about the determinants of risk of cardiovascular disease (CVD) among patients with chronic renal failure. Indeed, such patients have often been deliberately excluded from randomised trials of treatments of CVD, perhaps because of concerns about drug safety. But the absolute risk of CVD among them may be large, so the potential absolute benefits of treatments may also be large, and may well exceed any increased hazards. Hence, as well as further investigation of the underlying mechanisms of cardiac disease, it would be helpful to have some large-scale randomised trials in a wide range of renal patients of interventions (such as cholesterol-lowering drugs, antihypertensives, aspirin, B-vitamins, and antioxidant vitamins) that are of proven or suspected benefit in other settings. If safe and effective treatments can be identified, and started early in the natural history of renal failure, the exceptionally high risk of CVD experienced by these patients could be decreased before and after end-stage renal failure has occurred. Populations with certain chronic metabolic disorders (eg, diabetes and chronic renal failure) are at substantially increased risk of cardiovascular disease (CVD). Compared with the general population, few epidemiological studies of the determinants of CVD have been done in these groups. This situation is unfortunate, because the absolute risks of myocardial infarction, stroke, and congestive heart failure among some such individuals may be high, so the absolute benefits of effective treatments could also be large. At least part of the reason for this disparity is that “sick” populations are difficult to study. In healthy populations, measurement of risk factors some years before the onset of CVD should minimise any effects of pre-existing disease itself on risk-factor levels, thereby avoiding important distortions of any true associations. This is not the case in sick populations. For example, among individuals aged 85 years or older, various chronic illnesses that induce a compensatory decrease in cholesterol synthesis are also associated with an increased risk of death, producing artefactual negative associations between cholesterol and mortality.1 This phenomenon, known as confounding by disease (or reverse causality), may limit the extent to which standard observational studies can identify the true determinants of CVD in such populations. People with chronic renal failure are one such sick population at substantially increased risk of CVD. Despite the effectiveness of dialysis in the prevention of the immediately fatal consequences of end-stage renal failure, mortality remains high among dialysis patients, and many of these deaths involve cardiac disease.2 The relative hazard is largest among the young (figure 1): cardiac mortality in the USA for dialysis patients younger than 45 years is more than 100 times greater than in the general population.3 Although only small numbers of individuals progress to end-stage renal failure and require dialysis, Lancet 2000; 356: 147–52 Clinical Trial Service Unit & Epidemiological Studies Unit, Radcliffe Infirmary, Oxford, OX2 6HE, UK (C Baigent BMBCh, K Burbury MBBS); and Department of Nephrology, University Hospital (Birmingham) NHS Trust, Queen Elizabeth Medical Centre, Birmingham, UK (D Wheeler MD) Correspondence to: Dr Colin Baigent
THE LANCET • Vol 356 • July 8, 2000
less severe forms of renal impairment are much more common and are associated with substantial morbidity. In the USA, for example, although only about 0·1% of the population currently requires dialysis,2 more than ten times this number 2·5% have renal impairment sufficient to produce a serum creatinine concentration in excess of about 150 mol/L,4 and observational studies have shown that such individuals are at increased risk of CVD.5 This review considers some of the possible causes of this premature cardiovascular disease among patients with renal impairment and discusses ways in which the risk might eventually be reduced.
CVD among dialysis patients In the general population, cardiac disease is usually caused by atheromatous lesions in the coronary arteries. However, among patients commencing dialysis, the main cardiac abnormality is left ventricular hypertrophy, which is found in about 75% of them.6 This disorder is often accompanied by other circulatory abnormalities, including ventricular dilatation,6 arterial (and especially aortic) stiffening,7 and coronary atherosclerosis with prominent calcification.8 Clinical manifestations of congestive heart failure are already present in about a third of new dialysis patients, angina in about a quarter, and a history of myocardial infarction in about 10%.2 Once patients start dialysis, the occurrence of new atheromatous coronary-artery disease is difficult to determine clinically, mainly because symptoms of myocardial ischaemia can occur in the absence of angiographically significant coronary atheroma;9 thus only about a quarter of cardiac deaths on US dialysis programmes are attributed to myocardial infarction, and the remainder are categorised as sudden or arrhythmic (or some other type).2 Routine cause-of-death data in such patients may, however, be subject to serious misclassification errors,10 so some of the “sudden or arrhythmic” deaths could be due to undocumented myocardial infarction. Since left ventricular hypertrophy and atheromatous coronary-artery disease must have developed over a period of at least a few years, establishment of when cardiovascular disease first becomes significant during the
147
For personal use only. Not to be reproduced without permission of The Lancet.
REVIEW
Figure 1: Cardiovascular mortality (death due to arrhythmias, cardiomyopathy, cardiac arrest, myocardial infarction, atherosclerotic heart disease, and pulmonary oedema) in dialysis patients From reference 3, with permission.
natural history of progressive renal failure would be useful. Observational studies have indicated (but not proved) that this may well occur at a very early stage. Among middle-aged men in one community study, a moderate increase in serum creatinine concentration (⭓130 mol/L) was associated with an age-adjusted relative risk of 1·5 for coronary-artery disease and 3·0 for stroke.5 Such an association might arise because renal dysfunction causes an increased risk of cardiovascular disease, or because cardiovascular disease (such as renalartery stenosis or heart failure) causes renal dysfunction, or because some other factor (eg, hypertension or diabetes) causes both renal dysfunction and cardiovascular disease. Moreover, combinations of these explanations could apply to a greater or lesser degree as renal disease progresses. Some cardiovascular risk factors that are altered in association with renal impairment, including raised blood pressure and dyslipidaemia, are established causes of CVD in the general population, whereas others (eg, microalbuminuria, anaemia, or raised concentrations of lipoprotein (a), homocysteine, or various acute-phase reactants) are of less certain relevance. The latter group of risk factors may be direct causes of CVD in chronic renal failure, markers of pre-existing CVD, or markers of other factors that themselves increase the risk of CVD, but they are worth studying irrespective of their precise role. Prospective observational studies designed to assess the effects of particular risk factors among dialysis patients may be distorted by the effects of reverse causality, but in principle this difficulty should be kept to a minimum by restricting attention to younger individuals with only minor renal impairment and little comorbid disease. Alternatively, at least for the study of reversible risk factors (eg, blood cholesterol), randomised trials of drugs that modify them (eg, 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors) should avoid such confounding altogether. Unfortunately, however, neither type of epidemiological study has yet been done among predialysis patients. A complementary approach to these questions might be to assess the particular risk-factor disturbances observed
148
Figure 2: 10-year death rates from coronary heart disease (CHD) among individuals aged 35–57 Individuals with diabetes whose cholesterol was low (<4·7 mmol/L) had higher CHD rates than non-diabetic individuals with high cholesterol concentrations (>6·3 mmol/L); personal communication, J Neaton, MRFIT study.
in early renal failure, and then to derive the relative risks for CVD that would be expected to be associated with these abnormalities from observational studies of the general population. This approach has been useful in other high risk populations, such as those with diabetes. In patients with diabetes, despite a higher absolute risk, the relation between cholesterol and coronary-artery disease death is similar to that in the general population (figure 2). Moreover, randomised trials of cholesterollowering therapy have indicated similar proportional (and larger absolute) reductions in risk among patients with coronary-artery disease who also had diabetes. This review therefore seeks first to quantify the change in each risk factor that might typically be observed in association with early renal impairment, and then to use the currently available epidemiological evidence to assess the possible relative risks of coronary-artery disease and of congestive heart failure that might eventually be observed if these abnormalities were to persist long-term. Standard methods of correcting for the effects of underestimation due to fluctuations in the levels of risk factors within individuals over time have been employed throughout.11
Effects of early renal impairment on CVD risk Hypertension Experimental and clinical studies have shown that renal damage can cause hypertension through plasma volume expansion, sodium retention, overactivity of both the sympathetic nervous system and the renin-angiotensinaldosterone axis, and accumulation of circulating endogenous vasoactive substances.12 Without effective treatment, blood pressure gradually increases as glomerular filtration rate declines,13 while hypertension causes further renal damage and establishes a vicious circle. Early renal failure might typically result in a 10–20 mm Hg increase in diastolic blood pressure until, and unless, renal impairment is identified and treated. Left ventricular hypertrophy also seems to develop progressively in association with deteriorating renal function, beginning at a very early stage. In a Canadian study of progressive renal failure, left ventricular
THE LANCET • Vol 356 • July 8, 2000
For personal use only. Not to be reproduced without permission of The Lancet.
REVIEW
hypertrophy (determined echocardiographically) was found in about a third of patients with mild renal insufficiency (defined as calculated creatinine clearance of 50–75 mL/min) and in about half of those with severe renal insufficiency (clearance <25 mL/min).14 During the follow-up of this cohort, high blood pressure was associated with a higher rate of progression of left ventricular hypertrophy. A negative association between blood pressure and mortality has been observed in dialysis populations,15 but since blood pressure may fall as a consequence of structural heart disease, such associations could be a consequence—at least in part—of reverse causality.16 It may be more reliable to consider prospective observational studies in the general population. In such studies, the relation between risk of coronary-artery disease (plotted on a doubling, or logarithmic, scale) and blood pressure is roughly linear, with a prolonged 10 mm Hg higher diastolic blood pressure associated, between the ages of about 45 and 65 years, with a 1·6-fold higher risk of coronary-artery disease,17 and with a 1·8fold higher risk of congestive heart failure.17,18 On the basis of these studies, therefore, the prolonged 10–20 mm Hg higher diastolic blood pressure observed in untreated (or undertreated) people with early renal disease might eventually be expected to produce a 1·6–2·5-fold higher risk of coronary-artery disease, and a 1·8-fold to 3·0-fold higher risk of congestive heart failure (table); these relative risks may be substantially higher among young individuals.19 Dyslipidaemia LDL cholesterol—Dyslipidaemia in renal failure is characterised by an accumulation of partly metabolised triglyceride-rich particles (predominantly VLDL and IDL remnants) mainly due to abnormal lipase function, causing hypertriglyceridaemia and low HDL-cholesterol concentrations.20 Insulin resistance is also common among patients with moderate renal impairment.21 Once patients reach end-stage renal failure, cholesterol concentrations are typically similar to those in the general population (or lower),20 but this pattern often conceals a highly abnormal lipid subfraction profile with a predominance of atherogenic small, dense LDL-particles.22 Cross-sectional studies have suggested that lipid and apolipoprotein abnormalities are already present among individuals with early impairment and increase in severity as renal dysfunction progresses.23 As was the case for blood pressure, inverse associations observed among dialysis patients between blood cholesterol and all-cause24 or cardiovascular25 mortality may have been distorted by reverse causality. Prospective observational studies in the general population have shown that the relation between risk of coronary-artery disease and blood cholesterol is roughly log-linear, and that within the range studied there is no threshold below which a lower cholesterol concentration is not associated with a lower risk of coronary-artery disease. In middle age, each 1 mmol/L increase in LDL-cholesterol is associated with around a two-fold higher risk of coronary-artery disease,26 with similar associations observed in both low-risk and highrisk populations.27 Early renal impairment with microalbuminuria might typically be associated with an increase of about 0·5 mmol/L in LDL-cholesterol,28 and increases of at least 1 mmol/L (and often much more) occur in association with “nephrotic-range” proteinuria (ie, >3 g per 24 hours).29 In middle-aged patients, such
THE LANCET • Vol 356 • July 8, 2000
differences in LDL-cholesterol, if sustained, would be expected to produce around a 1·4-fold higher risk of coronary-artery disease among microalbuminuric patients, and a two-fold higher risk among those with heavy proteinuria—and, as for blood pressure, these relative risks may be substantially larger among younger patients.30 Moreover, the existence of other disturbances in cholesterol metabolism (eg, a high proportion of smalldense LDL particles) may imply still larger excess risks. HDL-cholesterol—Lower concentrations of HDLcholesterol are also associated with an increased risk of coronary-artery disease, independently of LDLcholesterol concentrations, and a meta-analysis of observational studies suggested that a prolonged 0·1 mmol/L lower concentration of HDL-cholesterol would be associated with around a 1·2-fold higher risk of coronary-artery disease.31 Among patients with early renal failure, HDL-cholesterol concentrations might typically be around 0·1–0·2 mmol/L lower than in a healthy control population,23 which might then be expected to produce a 1·2–1·4-fold higher risk of coronary-artery disease, independently of any changes in LDL-cholesterol concentrations (table). Lipoprotein (a) Early renal impairment is associated with an increase of about 0·2–0·4 mol/L in mean lipoprotein (a) concentrations, and cross-sectional data suggest that lipoprotein (a) concentrations are negatively correlated with glomerular filtration rate.32 Substantially larger increases (eg, of 0·8–1·2 mol/L)) may occur in conjunction with nephrotic-range proteinuria.33 Whereas genetic causes of raised lipoprotein (a) concentrations produce a selective increase in low-molecular-weight isoforms, it is mainly the high-molecular-weight isoforms that increase during renal failure, suggesting that different biochemical mechanisms are involved.33 Lipoprotein (a) concentrations fall after successful transplantation,34 and since this is unlikely to result from changes in medication, renal dysfunction per se appears to be at least partly responsible for the raised lipoprotein (a) concentrations among individuals with renal impairment. Prospective studies in the general population have indicated that raised lipoprotein (a) concentrations are associated with an increased risk of coronary-artery disease (and of other Typical change in risk factor
Approximate relative risk*
Coronary-artery disease Diastolic blood pressure† LDL-cholesterol HDL-cholesterol Lipoprotein (a) Fibrinogen Homocysteine
↑10–20 mm Hg ↑0·5–1·0 mmol/L‡ ↓0·2–0·4 mmol/L ↑0·2–0·4 mol/L‡ ↑1 g/L‡ ↑5 mol/L
1·6–2·5 1·4–2·0 1·2–1·4 <1·5 1·8 <1·5
Congestive heart failure Diastolic blood pressure† Haemoglobin
↑10–20 mm Hg ↓10–30 g/L
1·8–3·0 1·3–2·2
*Relative risks estimated for middle-aged populations, from meta-analyses, and corrected for regression dilution bias. Relative risks may be substantially larger among younger individuals. †In prospective observational studies, the risk associated with a 5 mm Hg lower diastolic blood pressure was similar to that of about 10 mm Hg lower systolic blood pressure.17 ‡Individuals with substantial proteinuria may develop differences that are substantially larger (though in the same direction) than the means that are shown.
Typical differences in risk factors for coronary-artery disease and congestive heart failure between patients with early renal impairment and a normal middle-aged population
149
For personal use only. Not to be reproduced without permission of The Lancet.
REVIEW
forms of vascular disease). A meta-analysis of prospective studies has shown that, compared with individuals in the bottom third of the baseline lipoprotein (a) distribution, individuals in the top third had a 1·7-fold higher risk of coronary-artery disease.35 In the general population, lipoprotein (a) concentrations have a highly skewed distribution, with a mean of about 0·8 mol/L and SD of 0·6 mol/L, which would indicate that even if the association is entirely causal (which it may well not be), a persistent increase of 0·2–0·4 mol/L would be associated with a relative risk of less than 1·5 (table). However, the range of values represented by the difference between the bottom and top thirds is highly dependent on the particular assay used and the particular group under study,33 so it is difficult to predict the size of any causal effects precisely. Microalbuminuria Microalbuminuria may result from a specific renal abnormality (eg, altered haemodynamics or glomerular basement-membrane selectivity) or it may reflect a more generalised increase in vascular permeability secondary to vascular endothelial dysfunction.36,37 The prevalence of microalbuminuria (ie, urinary albumin excretion of 20–200 g/min) is raised among individuals with type I or type II diabetes37 or hypertension,28 and is a useful marker of the likelihood of progressive renal impairment because it can be detected in the absence of a definite increase in serum creatinine. A meta-analysis of prospective observational studies among type II diabetic patients suggested that the presence of microalbuminuria at baseline is associated with about a two-fold increased risk of cardiovascular events (although since only one of the six studies reporting cardiovascular events corrected for confounding variables, this may well be an exaggeration).38 Microalbuminuria is also associated with an increased risk of coronary-artery disease among nondiabetic individuals,39 and with an increased risk of left ventricular hypertrophy and myocardial ischaemia among patients with essential hypertension.28 However, since microalbuminuria is strongly associated with known risk factors for coronary-artery disease (raised blood LDLcholesterol and raised blood pressure), whether microalbuminuria is itself an independent cause of cardiovascular disease remains uncertain. Homocysteine Even small reductions in glomerular-filtration rate are associated with an increase in plasma total homocysteine40 and, among individuals with established renal impairment, total homocysteine is strongly negatively correlated with serum creatinine.41 Individuals with early renal impairment typically have increases of around 5 mol/L in plasma total homocysteine.40,41 As has been observed for lipoprotein (a), total homocysteine concentrations fall in individual patients after renal transplantation,42 suggesting that renal mechanisms are at least partly responsible for the increase in total homocysteine among individuals with renal impairment. Prospective observational studies in the general population have shown that a 3–5 mol/L higher blood total homocysteine corresponds to about a 1·3-fold greater risk of coronary-artery disease, but increased total homocysteine concentrations are of uncertain causal relevance.43 Even if this association is causal, however, a 5 mol/L higher concentration of total homocysteine in
150
the early stages of renal failure would be expected to confer no more than about a 1·5-fold increased risk of coronary-artery disease (table). Acute-phase reactants Acute or chronic inflammation produces systemic changes, including altered concentrations of various blood components that are collectively referred to as acutephase reactants. These include fibrinogen, albumin, and C-reactive protein.44 There is little evidence that hypoalbuminaemia or raised C-reactive protein concentrations are direct causes of CVD. However, hyperfibrinogenaemia may plausibly contribute to an increased risk of CVD, since fibrinogen modulates coagulation and increases blood viscosity, and prospective studies in the general population have consistently shown positive associations with the risk of CVD.45 Increases in mean plasma fibrinogen concentrations of about 1·0 g/L have been reported among patients with early renal failure, with substantially larger increases (over 2·0 g/L) among those with heavy proteinuria.46 A meta-analysis of prospective studies in the general population suggested that an increase in fibrinogen of about 1 g/L corresponded to a 1·8-fold increased risk of coronary-artery disease,45 so if the association between fibrinogen and coronary-artery disease is mainly one of cause and effect, then the difference in fibrinogen associated with early renal failure could produce about a 1·8-fold increase in the risk of coronary-artery disease (table). Uncertainties about the causal relevance of fibrinogen remain, however, since genetic variations associated with lifelong differences in fibrinogen have not been consistently associated with differences in the risk of coronary-artery disease.47 Anaemia Erythropoietin is produced in the kidney, and anaemia resulting from a deficiency of erythropoietin is present in the majority of patients with end-stage renal failure; smaller reductions in haemoglobin occur in mild-tomoderate renal impairment. For example, among patients without left ventricular hypertrophy in a Canadian cohort of predialysis patients, mean baseline haemoglobin concentrations among those with creatinine clearances of more than 50, 25–50, and less than 25 mL/min were 145, 130, and 120 g/L respectively, and during the subsequent follow-up of this group of patients with early renal disease, each 5 g/L lower haemoglobin was associated with about a 1·3-fold increased incidence of left ventricular growth.14 Similarly, in patients with more severe renal impairment who were on dialysis, each 10 g/L decrease in mean haemoglobin was associated with about a 1·3-fold greater incidence of congestive heart failure.48 Such studies cannot exclude the possibility, however, that other unmeasured toxins that accumulate in association with renal dysfunction and which correlate with reductions in haemoglobin, were themselves a cause of left ventricular growth or cardiac dysfunction. But, if this moderate association is one of cause and effect and begins in early renal failure, then the risk of congestive heart failure would be predicted to increase by about 1·3–2·2-fold in association with decreases of 10–30 g/dL in haemoglobin (table). In principle, the independent effects of lower haemoglobin concentration on CVD in renal failure could be measured by randomised trials that compare the use of erythropoietin to attain different target haemoglobin
THE LANCET • Vol 356 • July 8, 2000
For personal use only. Not to be reproduced without permission of The Lancet.
REVIEW
concentrations. Partial correction of anaemia among dialysis patents has been shown to reduce left ventricular mass,49 but the effects of increasing haemoglobin on cardiovascular events among predialysis patients are unknown. In the dialysis population, however, a large randomised trial compared normal haematocrit (target 42%) to low haematocrit (target 30%) among 1233 patients with a history of cardiac disease, with follow-up scheduled to continue for 3 years.50 Unfortunately, that study was stopped prematurely because of a slight, and non-significant, excess of cardiovascular deaths in the normal haematocrit group (125 of 618 normal vs 112 of 615 low). However, since left ventricular hypertrophy is already established once patients require dialysis, erythropoietin might be somewhat more effective if used at an earlier stage in the natural history of renal failure, and this question should perhaps be revisited by further randomised trials.
Discussion Early renal failure is associated with changes in known and suspected cardiovascular risk factors, some of which seem to be a direct result of renal impairment. Higher blood presure, higher LDL-cholesterol concentrations, and lower HDL-cholesterol concentrations are all likely to be of independent causal relevance to coronary-artery disease. Extrapolation from observational studies among middle-aged individuals in the general population suggests that the early exposure to abnormalities of these three factors seen in early renal disease might well produce large excess relative (and absolute) risks of coronary-artery disease. For example, if a middle-aged patient with early renal impairment and microalbuminuria was exposed to a prolonged 10 mm Hg higher diastolic blood pressure, a 0·5 mmol/L higher LDL-cholesterol concentration, and a 0·1 mmol/L lower HDL-cholesterol concentration, the risk of coronary-artery disease is predicted to increase by around three-fold (1·6⫻1·4⫻1·2; table). Likewise, a prolonged 20 mm Hg higher diastolic blood pressure, 1 mmol/L higher LDL-cholesterol concentration, and a 0·2 mmol/L lower HDL-cholesterol concentration in a patient with more severe hypertension and nephrotic-range proteinuria might be associated with a seven-fold increased risk (2·5⫻2·0⫻1·4; table). Similarly, if the causal relevance of anaemia to the development of congestive heart failure is truly independent of other factors, then typical exposures in early renal failure would be predicted to produce increases of around two-fold to seven-fold (1·8⫻1·3 to 3·0⫻2·2) in the risk of congestive heart failure (table). The impact of these risk factors on the risk of coronary-artery disease and of congestive heart failure may well be substantially larger among younger patients.19,30 Hence, when taken together with type-1 diabetes, which is a common cause of both cardiac disease and renal failure in young dialysis patients, these features of renal impairment may provide at least a partial explanation for the 100-fold increase in CVD risk observed among them. Large-scale prospective observational studies among younger individuals with early renal impairment, with remeasurement of exposures as renal impairment progresses and long-term follow-up of clinical outcomes, would help to quantify the independent contribution from each of the known risk factors, and from emerging risk factors of less certain causal relevance. However, whether or not the known risk factors are
THE LANCET • Vol 356 • July 8, 2000
increased, and whether or not any such increases can account for most of the increased incidence of CVD in some particular group of renal patients, reduction of exposure to such risk factors may still produce substantial reductions in the subsequent incidence of CVD events. For example, patients with established coronary-artery disease are at increased risk of myocardial infarction even when they have “average” LDL-cholesterol concentrations, and randomised trials have shown that lowering LDL-cholesterol concentrations to below average levels reduced this risk.27 Hence, in patients with renal failure, reduction of LDL-cholesterol concentrations may well lower the risk of cardiovascular disease among those with average (or even below average) LDLcholesterol concentrations.51 Similarly, reducing blood pressure among patients with “average” blood pressure in early renal failure may well reduce the risk of coronaryartery disease and of congestive heart failure. So, even if independent “uraemic” risk factors for cardiovascular disease remain to be discovered, reduction of exposure to known risk factors for coronary-artery disease (such as cholesterol and blood pressure) might well substantially reduce the damaging cardiovascular effects of renal disease. Consider an example in which two known and two unknown “uraemic” risk factors are each assumed independently to double the risk of CVD, and suppose that treatments exist that can eventually reverse the full effects of known risk factors. The 16-fold increased risk of CVD in a patient who has been exposed to both of the known and both of the unknown “uraemic” risk factors could be reduced by three-quarters (to a four-fold increased risk) just by treating the known risk factors. Thus it may not be necessary first to identify the risk factors peculiar to a particular disease before being able to reduce some of the excess risks resulting from them. This point may be particularly relevant among dialysis patients and renal transplant recipients, in whom poorly understood factors related to uraemia, the dialysis procedure, or to immunosuppressive regimens may further promote the development of cardiovascular disease. Colin Baigent is a career-track scientist with the UK Medical Research Council. David Wheeler received support for research into CVD in renal failure from the British Heart Foundation. Helpful comments on earlier drafts were provided by Jane Armitage, Rory Collins, John Danesh, Rob Foley, Richard Peto, and Cathie Sudlow. Onima Chowdhury provided helpful editorial assistance.
References 1
2
3
4
5
6
7
8
Corti M-C, Guralnik JM, Salive ME, et al. Clarifying the direct relation between total cholesterol levels and death from coronary heart disease in older persons. Ann Intern Med 1997; 126: 753–60. US Renal Data System. 1999 Annual Data Report. Bethesda: National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, April 1999. Foley RN, Parfrey PS, Sarnak MJ. Clinical epidemiology of cardiovascular disease in chronic renal disease. Am J Kidney Dis 1998; 32 (suppl 3): S112–19. Jones CA, McQuillan GM, Kusek JW, et al. Serum creatinine levels in the US population: Third National Health and Nutrition Examination Survey. Am J Kidney Dis 1998; 32: 992–99. Wannamethee SG, Shaper AG, Perry IJ. Serum creatinine concentration and risk of cardiovascular disease. Stroke 1997; 28: 557–63. Foley RN, Parfrey PS, Harnett JD, et al. Clinical and echocardiographic disease in patients starting end-stage renal disease therapy. Kidney Int 1995; 47: 186–92. London GM, Marchais SJ, Guerin AP, Metivier F, Pannier B. Cardiac hypertrophy and arterial alterations in end-stage renal disease; hemodynamic factors. Kidney Int 1993; 43 (suppl 41): S42–49. Schwarz U, Buzello M, Ritz E, et al. Morphology of coronary
151
For personal use only. Not to be reproduced without permission of The Lancet.
REVIEW
9
10
11
12
13
14
15
16
17
18
19
20 21
22
23
24
25
26
27
28
29
atherosclerotic lesions in patients with end-stage renal failure. Nephrol Dial Transplant 2000; 15: 218–23. Rostand SG, Kirk KA, Rutsky EA. Dialysis-associated ischaemic heart disease: insights from coronary angiography. Kidney Int 1984; 25: 653–59. Perneger TV, Klag MJ, Whelton PK. Cause of death in patients with end-stage renal disease: death certificates vs registry reports. Am J Public Health 1993; 83: 1735–38. Clarke R, Shipley M, Lewington S, et al. Underestimation of risk associations due to regression dilution in long-term follow-up of prospective studies. Am J Epidemiol 1999; 150: 341–53. Edmunds ME, Russell GI. Hypertension in renal failure. In: Swales JD, ed. Textbook of hypertension. Oxford: Blackwell Scientific Publications, 1994: 798–810. Buckalew VM Jr, Berg RL, Wang SR, Porush JG, Rauch S, Schulman G, and the Modification of Diet in Renal Disease Study Group. Prevalence of hypertension in 1795 subjects with chronic renal disease: the Modification of Diet in Renal Disease Study baseline cohort. Am J Kidney Dis 1996; 28: 811–21. Levin A, Thompson CR, Ethier J, et al. Left ventricular mass index increase in early renal disease: impact of decline in hemoglobin. Am J Kidney Dis 1999; 34: 125–34. Zager PG, Nikolic J, Brown RH, et al. “U” curve association of blood pressure and mortality in dialysis patients. Kidney Int 1998; 54: 561–69. Foley RN, Parfrey PS, Harnett JD, Kent GM, Murray DC, Barre PE. Impact of hypertension on cardiomyopathy, morbidity and mortality in end-stage renal disease. Kidney Int 1996; 49: 1379–85. MacMahon S. Blood pressure and the risks of cardiovascular disease. In: Swales JD, ed. Textbook of hypertension. Oxford: Blackwell Scientific Publications, 1994: 46–57. Levy D, Larson MG, Vasan RS, Kannel W, Ho KKL. The progression from hypertension to congestive heart failure. JAMA 1996; 275: 1557–62. Prospective Studies Collaboration. Cholesterol, diastolic blood pressure, and stroke: 13 000 strokes in 450 000 people in 45 prospective cohorts. Lancet 1995; 346: 1647–53. Oda H, Keane WF. Lipid abnormalities in end stage renal disease. Nephrol Dial Transplant 1998; 13 (Suppl 1): 45–49. Fliser D, Pacini G, Engelleiter R, et al. Insulin resistance and hyperinsulinemia are already present in patients with incipient renal disease. Kidney Int 1998; 53: 1343–47. Rajman I, Harper L, McPake D, Kendall MJ, Wheeler DC. Lowdensity lipoprotein subfraction profiles in chronic renal failure. Nephrol Dial Transplant 1998; 13: 2281–87. Grützmacher P, März W, Peschke B, Gross W, Schoeppe W. Lipoproteins and apolipoproteins during the progression of chronic renal disease. Nephron 1988; 50: 103–11. Lowrie EG, Lew NL. Death risk in hemodialysis patients: the predictive value of commonly measured variables and an evaluation of death rate differences between facilities. Am J Kidney Dis 1990; 15: 458–82. Degoulet P, Legrain M, Réach I, et al. Mortality risk factors in patients treated by chronic hemodialysis: Report of the Diaphane Collaborative Study. Nephron 1982; 31: 103–10. Law MR, Wald NJ, Wu T, Hackshaw A, Bailey A. Systematic underestimation of association between serum cholesterol concentration and ischaemic heart disease in observational studies: data from the BUPA study. BMJ 1994; 308: 363–66. Baigent C, Armitage J. Cholesterol reduction among patients at increased risk of coronary heart disease. Proc R Coll Physicians Edinb 1999; 29 (Suppl 5): 10–15. Bianchi S, Bigazzi R, Campese VM. Microalbuminuria in essential hypertension: significance, pathophysiology, and therapeutic implications. Am J Kidney Dis 1999; 34: 973–95. Warwick GL, Packard CJ. Lipoprotein metabolism in the nephrotic syndrome. Nephrol Dial Transplant 1993; 8: 385–96.
152
30 Law MR, Wald NJ, Thompson SG. By how much and how quickly does reduction in serum cholesterol concentration lower risk of ischaemic heart disease? BMJ 1994; 308: 367–73. 31 Gordon DJ, Probstfield JL, Garrison RJ, et al. High-density lipoprotein cholesterol and cardiovascular disease: four prospective American studies. Circulation 1989; 79: 8–15. 32 Sechi LA, Zingaro L, De Carli S, Sechi G, Catena C, Falleti E. Increased serum lipoprotein (a) levels in patients with early renal failure. Ann Intern Med 1998; 129: 457–61. 33 Kronenberg F, Utermann G, Dieplinger H. Lipoprotein[a] in renal disease. Am J Kidney Dis 1996; 27: 1–25. 34 Kronenberg F, König P, Lhotta K, et al. Apolipoprotein[a] phenotype-associated decrease in lipoprotein[a] plasma concentrations after renal transplantation. Arterioscler Thromb 1994; 14: 1399–404. 35 Danesh J, Collins R, Peto R. Lipoprotein (a) and coronary heart disease: meta-analysis of prospective studies. Circulation 2000 (in press). 36 Jensen JS, Borch-Johnsen K, Jensen G, Feldt-Rasmussen B. Microalbuminuria reflects a generalised transvascular albumin leakiness in clinically healthy subjects. Clin Sci 1995; 88: 629–33. 37 Gilbert RE, Cooper ME, McNally PG, O’Brien RC, Taft J, Jerums G. Microalbuminuria; prognostic and therapeutic implications in diabetes mellitus. Diabet Med 1994; 11: 636–45. 38 Dinneen SF, Gerstein HC. The association of microalbuminuria and mortality in non-insulin-dependent diabetes mellitus: a systematic overview of the literature. Arch Intern Med 1997; 157: 1413–18. 39 Yudkin JS, Forrest RD, Jackson CA. Microalbuminuria as predictor of vascular disease in non-diabetic subjects. Lancet 1988; ii: 530–33. 40 Arnadottir M, Hultberg B, Nilsson-Ehle P, Thysell H. The effect of reduced glomerular filtration rate on plasma total homocysteine concentration. Scand J Clin Lab Invest 1996; 56: 41–46. 41 Chauveau P, Chadefaux B, Coudé M, et al. Hyperhomocysteinaemia, a risk factor for atherosclerosis in chronic uremic patients. Kidney Int 1993; 43 (suppl 41): S72–77. 42 van Guldener C, Janssen MJFM, Stehouwer CDA, et al. The effects of renal transplantation on hyperhomocysteinaemia in dialysis patients, and the estimation of renal homocysteine extraction in patients with normal renal function. Neth J Med 1998; 52: 58–64. 43 Danesh J, Lewington S. Plasma homocysteine and coronary heart disease: systematic review of published epidemiological studies. J Cardiovasc Risk 1998; 5: 229–32. 44 Gabay C, Kushner I. Acute-phase proteins and other systemic responses to inflammation. N Engl J Med 1999; 340: 448–54. 45 Danesh J, Collins R, Appleby P, Peto R. Association of fibrinogen, C-reactive protein, albumin, or leukocyte count with coronary heart disease. JAMA 1998; 279: 1477–82. 46 Irish A. Cardiovascular disease, fibrinogen and the acute phase response: associations with lipids and blood pressure in patients with chronic renal disease. Atherosclerosis 1998; 137: 133–39. 47 Scarabin P-Y, Bara L, Ricard S, et al. Genetic variation at the -fibrinogen locus in relation to plasma fibrinogen concentrations and risk of myocardial infarction: the ECTIM study. Arterioscler Thromb 1993; 13: 886–91. 48 Foley RN, Parfrey PS, Harnett JD, Kent GM, Murray DC, Barre PE. The impact of anaemia on cardiomyopathy, morbidity and mortality in end-stage renal disease. Am J Kidney Dis 1996; 28: 53–61. 49 Pascual J, Teruel JL, Moya JL, Liaño F, Jiménez-Mena M, Ortuño J. Regression of left ventricular hypertrophy after partial correction of anaemia with erythropoietin in patients on hemodialysis: a prospective study. Clin Nephrol 1991; 35: 280–87. 50 Besarab A, Bolton WK, Browne JK, et al. The effects of normal as compared with low hematocrit values in patients with cardiac disease who are receiving hemodialysis and epoietin. N Engl J Med 1998; 339: 584–90. 51 Baigent C, Wheeler D. Should we reduce blood cholesterol to prevent cardiovascular disease among patients with chronic renal failure? Nephrol Dial Transplant 2000 (in press).
THE LANCET • Vol 356 • July 8, 2000
For personal use only. Not to be reproduced without permission of The Lancet.
Review
Premature cardiovascular disease in chronic renal failure Colin Baigent, Kate Burbury, David Wheeler There is a remarkable lack of reliable information about the determinants of risk of cardiovascular disease (CVD) among patients with chronic renal failure. Indeed, such patients have often been deliberately excluded from randomised trials of treatments of CVD, perhaps because of concerns about drug safety. But the absolute risk of CVD among them may be large, so the potential absolute benefits of treatments may also be large, and may well exceed any increased hazards. Hence, as well as further investigation of the underlying mechanisms of cardiac disease, it would be helpful to have some large-scale randomised trials in a wide range of renal patients of interventions (such as cholesterol-lowering drugs, antihypertensives, aspirin, B-vitamins, and antioxidant vitamins) that are of proven or suspected benefit in other settings. If safe and effective treatments can be identified, and started early in the natural history of renal failure, the exceptionally high risk of CVD experienced by these patients could be decreased before and after end-stage renal failure has occurred. Populations with certain chronic metabolic disorders (eg, diabetes and chronic renal failure) are at substantially increased risk of cardiovascular disease (CVD). Compared with the general population, few epidemiological studies of the determinants of CVD have been done in these groups. This situation is unfortunate, because the absolute risks of myocardial infarction, stroke, and congestive heart failure among some such individuals may be high, so the absolute benefits of effective treatments could also be large. At least part of the reason for this disparity is that “sick” populations are difficult to study. In healthy populations, measurement of risk factors some years before the onset of CVD should minimise any effects of pre-existing disease itself on risk-factor levels, thereby avoiding important distortions of any true associations. This is not the case in sick populations. For example, among individuals aged 85 years or older, various chronic illnesses that induce a compensatory decrease in cholesterol synthesis are also associated with an increased risk of death, producing artefactual negative associations between cholesterol and mortality.1 This phenomenon, known as confounding by disease (or reverse causality), may limit the extent to which standard observational studies can identify the true determinants of CVD in such populations. People with chronic renal failure are one such sick population at substantially increased risk of CVD. Despite the effectiveness of dialysis in the prevention of the immediately fatal consequences of end-stage renal failure, mortality remains high among dialysis patients, and many of these deaths involve cardiac disease.2 The relative hazard is largest among the young (figure 1): cardiac mortality in the USA for dialysis patients younger than 45 years is more than 100 times greater than in the general population.3 Although only small numbers of individuals progress to end-stage renal failure and require dialysis, Lancet 2000; 356: 147–52 Clinical Trial Service Unit & Epidemiological Studies Unit, Radcliffe Infirmary, Oxford, OX2 6HE, UK (C Baigent BMBCh, K Burbury MBBS); and Department of Nephrology, University Hospital (Birmingham) NHS Trust, Queen Elizabeth Medical Centre, Birmingham, UK (D Wheeler MD) Correspondence to: Dr Colin Baigent
THE LANCET • Vol 356 • July 8, 2000
less severe forms of renal impairment are much more common and are associated with substantial morbidity. In the USA, for example, although only about 0·1% of the population currently requires dialysis,2 more than ten times this number 2·5% have renal impairment sufficient to produce a serum creatinine concentration in excess of about 150 mol/L,4 and observational studies have shown that such individuals are at increased risk of CVD.5 This review considers some of the possible causes of this premature cardiovascular disease among patients with renal impairment and discusses ways in which the risk might eventually be reduced.
CVD among dialysis patients In the general population, cardiac disease is usually caused by atheromatous lesions in the coronary arteries. However, among patients commencing dialysis, the main cardiac abnormality is left ventricular hypertrophy, which is found in about 75% of them.6 This disorder is often accompanied by other circulatory abnormalities, including ventricular dilatation,6 arterial (and especially aortic) stiffening,7 and coronary atherosclerosis with prominent calcification.8 Clinical manifestations of congestive heart failure are already present in about a third of new dialysis patients, angina in about a quarter, and a history of myocardial infarction in about 10%.2 Once patients start dialysis, the occurrence of new atheromatous coronary-artery disease is difficult to determine clinically, mainly because symptoms of myocardial ischaemia can occur in the absence of angiographically significant coronary atheroma;9 thus only about a quarter of cardiac deaths on US dialysis programmes are attributed to myocardial infarction, and the remainder are categorised as sudden or arrhythmic (or some other type).2 Routine cause-of-death data in such patients may, however, be subject to serious misclassification errors,10 so some of the “sudden or arrhythmic” deaths could be due to undocumented myocardial infarction. Since left ventricular hypertrophy and atheromatous coronary-artery disease must have developed over a period of at least a few years, establishment of when cardiovascular disease first becomes significant during the
147
For personal use only. Not to be reproduced without permission of The Lancet.
REVIEW
Figure 1: Cardiovascular mortality (death due to arrhythmias, cardiomyopathy, cardiac arrest, myocardial infarction, atherosclerotic heart disease, and pulmonary oedema) in dialysis patients From reference 3, with permission.
natural history of progressive renal failure would be useful. Observational studies have indicated (but not proved) that this may well occur at a very early stage. Among middle-aged men in one community study, a moderate increase in serum creatinine concentration (⭓130 mol/L) was associated with an age-adjusted relative risk of 1·5 for coronary-artery disease and 3·0 for stroke.5 Such an association might arise because renal dysfunction causes an increased risk of cardiovascular disease, or because cardiovascular disease (such as renalartery stenosis or heart failure) causes renal dysfunction, or because some other factor (eg, hypertension or diabetes) causes both renal dysfunction and cardiovascular disease. Moreover, combinations of these explanations could apply to a greater or lesser degree as renal disease progresses. Some cardiovascular risk factors that are altered in association with renal impairment, including raised blood pressure and dyslipidaemia, are established causes of CVD in the general population, whereas others (eg, microalbuminuria, anaemia, or raised concentrations of lipoprotein (a), homocysteine, or various acute-phase reactants) are of less certain relevance. The latter group of risk factors may be direct causes of CVD in chronic renal failure, markers of pre-existing CVD, or markers of other factors that themselves increase the risk of CVD, but they are worth studying irrespective of their precise role. Prospective observational studies designed to assess the effects of particular risk factors among dialysis patients may be distorted by the effects of reverse causality, but in principle this difficulty should be kept to a minimum by restricting attention to younger individuals with only minor renal impairment and little comorbid disease. Alternatively, at least for the study of reversible risk factors (eg, blood cholesterol), randomised trials of drugs that modify them (eg, 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors) should avoid such confounding altogether. Unfortunately, however, neither type of epidemiological study has yet been done among predialysis patients. A complementary approach to these questions might be to assess the particular risk-factor disturbances observed
148
Figure 2: 10-year death rates from coronary heart disease (CHD) among individuals aged 35–57 Individuals with diabetes whose cholesterol was low (<4·7 mmol/L) had higher CHD rates than non-diabetic individuals with high cholesterol concentrations (>6·3 mmol/L); personal communication, J Neaton, MRFIT study.
in early renal failure, and then to derive the relative risks for CVD that would be expected to be associated with these abnormalities from observational studies of the general population. This approach has been useful in other high risk populations, such as those with diabetes. In patients with diabetes, despite a higher absolute risk, the relation between cholesterol and coronary-artery disease death is similar to that in the general population (figure 2). Moreover, randomised trials of cholesterollowering therapy have indicated similar proportional (and larger absolute) reductions in risk among patients with coronary-artery disease who also had diabetes. This review therefore seeks first to quantify the change in each risk factor that might typically be observed in association with early renal impairment, and then to use the currently available epidemiological evidence to assess the possible relative risks of coronary-artery disease and of congestive heart failure that might eventually be observed if these abnormalities were to persist long-term. Standard methods of correcting for the effects of underestimation due to fluctuations in the levels of risk factors within individuals over time have been employed throughout.11
Effects of early renal impairment on CVD risk Hypertension Experimental and clinical studies have shown that renal damage can cause hypertension through plasma volume expansion, sodium retention, overactivity of both the sympathetic nervous system and the renin-angiotensinaldosterone axis, and accumulation of circulating endogenous vasoactive substances.12 Without effective treatment, blood pressure gradually increases as glomerular filtration rate declines,13 while hypertension causes further renal damage and establishes a vicious circle. Early renal failure might typically result in a 10–20 mm Hg increase in diastolic blood pressure until, and unless, renal impairment is identified and treated. Left ventricular hypertrophy also seems to develop progressively in association with deteriorating renal function, beginning at a very early stage. In a Canadian study of progressive renal failure, left ventricular
THE LANCET • Vol 356 • July 8, 2000
For personal use only. Not to be reproduced without permission of The Lancet.
REVIEW
hypertrophy (determined echocardiographically) was found in about a third of patients with mild renal insufficiency (defined as calculated creatinine clearance of 50–75 mL/min) and in about half of those with severe renal insufficiency (clearance <25 mL/min).14 During the follow-up of this cohort, high blood pressure was associated with a higher rate of progression of left ventricular hypertrophy. A negative association between blood pressure and mortality has been observed in dialysis populations,15 but since blood pressure may fall as a consequence of structural heart disease, such associations could be a consequence—at least in part—of reverse causality.16 It may be more reliable to consider prospective observational studies in the general population. In such studies, the relation between risk of coronary-artery disease (plotted on a doubling, or logarithmic, scale) and blood pressure is roughly linear, with a prolonged 10 mm Hg higher diastolic blood pressure associated, between the ages of about 45 and 65 years, with a 1·6-fold higher risk of coronary-artery disease,17 and with a 1·8fold higher risk of congestive heart failure.17,18 On the basis of these studies, therefore, the prolonged 10–20 mm Hg higher diastolic blood pressure observed in untreated (or undertreated) people with early renal disease might eventually be expected to produce a 1·6–2·5-fold higher risk of coronary-artery disease, and a 1·8-fold to 3·0-fold higher risk of congestive heart failure (table); these relative risks may be substantially higher among young individuals.19 Dyslipidaemia LDL cholesterol—Dyslipidaemia in renal failure is characterised by an accumulation of partly metabolised triglyceride-rich particles (predominantly VLDL and IDL remnants) mainly due to abnormal lipase function, causing hypertriglyceridaemia and low HDL-cholesterol concentrations.20 Insulin resistance is also common among patients with moderate renal impairment.21 Once patients reach end-stage renal failure, cholesterol concentrations are typically similar to those in the general population (or lower),20 but this pattern often conceals a highly abnormal lipid subfraction profile with a predominance of atherogenic small, dense LDL-particles.22 Cross-sectional studies have suggested that lipid and apolipoprotein abnormalities are already present among individuals with early impairment and increase in severity as renal dysfunction progresses.23 As was the case for blood pressure, inverse associations observed among dialysis patients between blood cholesterol and all-cause24 or cardiovascular25 mortality may have been distorted by reverse causality. Prospective observational studies in the general population have shown that the relation between risk of coronary-artery disease and blood cholesterol is roughly log-linear, and that within the range studied there is no threshold below which a lower cholesterol concentration is not associated with a lower risk of coronary-artery disease. In middle age, each 1 mmol/L increase in LDL-cholesterol is associated with around a two-fold higher risk of coronary-artery disease,26 with similar associations observed in both low-risk and highrisk populations.27 Early renal impairment with microalbuminuria might typically be associated with an increase of about 0·5 mmol/L in LDL-cholesterol,28 and increases of at least 1 mmol/L (and often much more) occur in association with “nephrotic-range” proteinuria (ie, >3 g per 24 hours).29 In middle-aged patients, such
THE LANCET • Vol 356 • July 8, 2000
differences in LDL-cholesterol, if sustained, would be expected to produce around a 1·4-fold higher risk of coronary-artery disease among microalbuminuric patients, and a two-fold higher risk among those with heavy proteinuria—and, as for blood pressure, these relative risks may be substantially larger among younger patients.30 Moreover, the existence of other disturbances in cholesterol metabolism (eg, a high proportion of smalldense LDL particles) may imply still larger excess risks. HDL-cholesterol—Lower concentrations of HDLcholesterol are also associated with an increased risk of coronary-artery disease, independently of LDLcholesterol concentrations, and a meta-analysis of observational studies suggested that a prolonged 0·1 mmol/L lower concentration of HDL-cholesterol would be associated with around a 1·2-fold higher risk of coronary-artery disease.31 Among patients with early renal failure, HDL-cholesterol concentrations might typically be around 0·1–0·2 mmol/L lower than in a healthy control population,23 which might then be expected to produce a 1·2–1·4-fold higher risk of coronary-artery disease, independently of any changes in LDL-cholesterol concentrations (table). Lipoprotein (a) Early renal impairment is associated with an increase of about 0·2–0·4 mol/L in mean lipoprotein (a) concentrations, and cross-sectional data suggest that lipoprotein (a) concentrations are negatively correlated with glomerular filtration rate.32 Substantially larger increases (eg, of 0·8–1·2 mol/L)) may occur in conjunction with nephrotic-range proteinuria.33 Whereas genetic causes of raised lipoprotein (a) concentrations produce a selective increase in low-molecular-weight isoforms, it is mainly the high-molecular-weight isoforms that increase during renal failure, suggesting that different biochemical mechanisms are involved.33 Lipoprotein (a) concentrations fall after successful transplantation,34 and since this is unlikely to result from changes in medication, renal dysfunction per se appears to be at least partly responsible for the raised lipoprotein (a) concentrations among individuals with renal impairment. Prospective studies in the general population have indicated that raised lipoprotein (a) concentrations are associated with an increased risk of coronary-artery disease (and of other Typical change in risk factor
Approximate relative risk*
Coronary-artery disease Diastolic blood pressure† LDL-cholesterol HDL-cholesterol Lipoprotein (a) Fibrinogen Homocysteine
↑10–20 mm Hg ↑0·5–1·0 mmol/L‡ ↓0·2–0·4 mmol/L ↑0·2–0·4 mol/L‡ ↑1 g/L‡ ↑5 mol/L
1·6–2·5 1·4–2·0 1·2–1·4 <1·5 1·8 <1·5
Congestive heart failure Diastolic blood pressure† Haemoglobin
↑10–20 mm Hg ↓10–30 g/L
1·8–3·0 1·3–2·2
*Relative risks estimated for middle-aged populations, from meta-analyses, and corrected for regression dilution bias. Relative risks may be substantially larger among younger individuals. †In prospective observational studies, the risk associated with a 5 mm Hg lower diastolic blood pressure was similar to that of about 10 mm Hg lower systolic blood pressure.17 ‡Individuals with substantial proteinuria may develop differences that are substantially larger (though in the same direction) than the means that are shown.
Typical differences in risk factors for coronary-artery disease and congestive heart failure between patients with early renal impairment and a normal middle-aged population
149
For personal use only. Not to be reproduced without permission of The Lancet.
REVIEW
forms of vascular disease). A meta-analysis of prospective studies has shown that, compared with individuals in the bottom third of the baseline lipoprotein (a) distribution, individuals in the top third had a 1·7-fold higher risk of coronary-artery disease.35 In the general population, lipoprotein (a) concentrations have a highly skewed distribution, with a mean of about 0·8 mol/L and SD of 0·6 mol/L, which would indicate that even if the association is entirely causal (which it may well not be), a persistent increase of 0·2–0·4 mol/L would be associated with a relative risk of less than 1·5 (table). However, the range of values represented by the difference between the bottom and top thirds is highly dependent on the particular assay used and the particular group under study,33 so it is difficult to predict the size of any causal effects precisely. Microalbuminuria Microalbuminuria may result from a specific renal abnormality (eg, altered haemodynamics or glomerular basement-membrane selectivity) or it may reflect a more generalised increase in vascular permeability secondary to vascular endothelial dysfunction.36,37 The prevalence of microalbuminuria (ie, urinary albumin excretion of 20–200 g/min) is raised among individuals with type I or type II diabetes37 or hypertension,28 and is a useful marker of the likelihood of progressive renal impairment because it can be detected in the absence of a definite increase in serum creatinine. A meta-analysis of prospective observational studies among type II diabetic patients suggested that the presence of microalbuminuria at baseline is associated with about a two-fold increased risk of cardiovascular events (although since only one of the six studies reporting cardiovascular events corrected for confounding variables, this may well be an exaggeration).38 Microalbuminuria is also associated with an increased risk of coronary-artery disease among nondiabetic individuals,39 and with an increased risk of left ventricular hypertrophy and myocardial ischaemia among patients with essential hypertension.28 However, since microalbuminuria is strongly associated with known risk factors for coronary-artery disease (raised blood LDLcholesterol and raised blood pressure), whether microalbuminuria is itself an independent cause of cardiovascular disease remains uncertain. Homocysteine Even small reductions in glomerular-filtration rate are associated with an increase in plasma total homocysteine40 and, among individuals with established renal impairment, total homocysteine is strongly negatively correlated with serum creatinine.41 Individuals with early renal impairment typically have increases of around 5 mol/L in plasma total homocysteine.40,41 As has been observed for lipoprotein (a), total homocysteine concentrations fall in individual patients after renal transplantation,42 suggesting that renal mechanisms are at least partly responsible for the increase in total homocysteine among individuals with renal impairment. Prospective observational studies in the general population have shown that a 3–5 mol/L higher blood total homocysteine corresponds to about a 1·3-fold greater risk of coronary-artery disease, but increased total homocysteine concentrations are of uncertain causal relevance.43 Even if this association is causal, however, a 5 mol/L higher concentration of total homocysteine in
150
the early stages of renal failure would be expected to confer no more than about a 1·5-fold increased risk of coronary-artery disease (table). Acute-phase reactants Acute or chronic inflammation produces systemic changes, including altered concentrations of various blood components that are collectively referred to as acutephase reactants. These include fibrinogen, albumin, and C-reactive protein.44 There is little evidence that hypoalbuminaemia or raised C-reactive protein concentrations are direct causes of CVD. However, hyperfibrinogenaemia may plausibly contribute to an increased risk of CVD, since fibrinogen modulates coagulation and increases blood viscosity, and prospective studies in the general population have consistently shown positive associations with the risk of CVD.45 Increases in mean plasma fibrinogen concentrations of about 1·0 g/L have been reported among patients with early renal failure, with substantially larger increases (over 2·0 g/L) among those with heavy proteinuria.46 A meta-analysis of prospective studies in the general population suggested that an increase in fibrinogen of about 1 g/L corresponded to a 1·8-fold increased risk of coronary-artery disease,45 so if the association between fibrinogen and coronary-artery disease is mainly one of cause and effect, then the difference in fibrinogen associated with early renal failure could produce about a 1·8-fold increase in the risk of coronary-artery disease (table). Uncertainties about the causal relevance of fibrinogen remain, however, since genetic variations associated with lifelong differences in fibrinogen have not been consistently associated with differences in the risk of coronary-artery disease.47 Anaemia Erythropoietin is produced in the kidney, and anaemia resulting from a deficiency of erythropoietin is present in the majority of patients with end-stage renal failure; smaller reductions in haemoglobin occur in mild-tomoderate renal impairment. For example, among patients without left ventricular hypertrophy in a Canadian cohort of predialysis patients, mean baseline haemoglobin concentrations among those with creatinine clearances of more than 50, 25–50, and less than 25 mL/min were 145, 130, and 120 g/L respectively, and during the subsequent follow-up of this group of patients with early renal disease, each 5 g/L lower haemoglobin was associated with about a 1·3-fold increased incidence of left ventricular growth.14 Similarly, in patients with more severe renal impairment who were on dialysis, each 10 g/L decrease in mean haemoglobin was associated with about a 1·3-fold greater incidence of congestive heart failure.48 Such studies cannot exclude the possibility, however, that other unmeasured toxins that accumulate in association with renal dysfunction and which correlate with reductions in haemoglobin, were themselves a cause of left ventricular growth or cardiac dysfunction. But, if this moderate association is one of cause and effect and begins in early renal failure, then the risk of congestive heart failure would be predicted to increase by about 1·3–2·2-fold in association with decreases of 10–30 g/dL in haemoglobin (table). In principle, the independent effects of lower haemoglobin concentration on CVD in renal failure could be measured by randomised trials that compare the use of erythropoietin to attain different target haemoglobin
THE LANCET • Vol 356 • July 8, 2000
For personal use only. Not to be reproduced without permission of The Lancet.
REVIEW
concentrations. Partial correction of anaemia among dialysis patents has been shown to reduce left ventricular mass,49 but the effects of increasing haemoglobin on cardiovascular events among predialysis patients are unknown. In the dialysis population, however, a large randomised trial compared normal haematocrit (target 42%) to low haematocrit (target 30%) among 1233 patients with a history of cardiac disease, with follow-up scheduled to continue for 3 years.50 Unfortunately, that study was stopped prematurely because of a slight, and non-significant, excess of cardiovascular deaths in the normal haematocrit group (125 of 618 normal vs 112 of 615 low). However, since left ventricular hypertrophy is already established once patients require dialysis, erythropoietin might be somewhat more effective if used at an earlier stage in the natural history of renal failure, and this question should perhaps be revisited by further randomised trials.
Discussion Early renal failure is associated with changes in known and suspected cardiovascular risk factors, some of which seem to be a direct result of renal impairment. Higher blood presure, higher LDL-cholesterol concentrations, and lower HDL-cholesterol concentrations are all likely to be of independent causal relevance to coronary-artery disease. Extrapolation from observational studies among middle-aged individuals in the general population suggests that the early exposure to abnormalities of these three factors seen in early renal disease might well produce large excess relative (and absolute) risks of coronary-artery disease. For example, if a middle-aged patient with early renal impairment and microalbuminuria was exposed to a prolonged 10 mm Hg higher diastolic blood pressure, a 0·5 mmol/L higher LDL-cholesterol concentration, and a 0·1 mmol/L lower HDL-cholesterol concentration, the risk of coronary-artery disease is predicted to increase by around three-fold (1·6⫻1·4⫻1·2; table). Likewise, a prolonged 20 mm Hg higher diastolic blood pressure, 1 mmol/L higher LDL-cholesterol concentration, and a 0·2 mmol/L lower HDL-cholesterol concentration in a patient with more severe hypertension and nephrotic-range proteinuria might be associated with a seven-fold increased risk (2·5⫻2·0⫻1·4; table). Similarly, if the causal relevance of anaemia to the development of congestive heart failure is truly independent of other factors, then typical exposures in early renal failure would be predicted to produce increases of around two-fold to seven-fold (1·8⫻1·3 to 3·0⫻2·2) in the risk of congestive heart failure (table). The impact of these risk factors on the risk of coronary-artery disease and of congestive heart failure may well be substantially larger among younger patients.19,30 Hence, when taken together with type-1 diabetes, which is a common cause of both cardiac disease and renal failure in young dialysis patients, these features of renal impairment may provide at least a partial explanation for the 100-fold increase in CVD risk observed among them. Large-scale prospective observational studies among younger individuals with early renal impairment, with remeasurement of exposures as renal impairment progresses and long-term follow-up of clinical outcomes, would help to quantify the independent contribution from each of the known risk factors, and from emerging risk factors of less certain causal relevance. However, whether or not the known risk factors are
THE LANCET • Vol 356 • July 8, 2000
increased, and whether or not any such increases can account for most of the increased incidence of CVD in some particular group of renal patients, reduction of exposure to such risk factors may still produce substantial reductions in the subsequent incidence of CVD events. For example, patients with established coronary-artery disease are at increased risk of myocardial infarction even when they have “average” LDL-cholesterol concentrations, and randomised trials have shown that lowering LDL-cholesterol concentrations to below average levels reduced this risk.27 Hence, in patients with renal failure, reduction of LDL-cholesterol concentrations may well lower the risk of cardiovascular disease among those with average (or even below average) LDLcholesterol concentrations.51 Similarly, reducing blood pressure among patients with “average” blood pressure in early renal failure may well reduce the risk of coronaryartery disease and of congestive heart failure. So, even if independent “uraemic” risk factors for cardiovascular disease remain to be discovered, reduction of exposure to known risk factors for coronary-artery disease (such as cholesterol and blood pressure) might well substantially reduce the damaging cardiovascular effects of renal disease. Consider an example in which two known and two unknown “uraemic” risk factors are each assumed independently to double the risk of CVD, and suppose that treatments exist that can eventually reverse the full effects of known risk factors. The 16-fold increased risk of CVD in a patient who has been exposed to both of the known and both of the unknown “uraemic” risk factors could be reduced by three-quarters (to a four-fold increased risk) just by treating the known risk factors. Thus it may not be necessary first to identify the risk factors peculiar to a particular disease before being able to reduce some of the excess risks resulting from them. This point may be particularly relevant among dialysis patients and renal transplant recipients, in whom poorly understood factors related to uraemia, the dialysis procedure, or to immunosuppressive regimens may further promote the development of cardiovascular disease. Colin Baigent is a career-track scientist with the UK Medical Research Council. David Wheeler received support for research into CVD in renal failure from the British Heart Foundation. Helpful comments on earlier drafts were provided by Jane Armitage, Rory Collins, John Danesh, Rob Foley, Richard Peto, and Cathie Sudlow. Onima Chowdhury provided helpful editorial assistance.
References 1
2
3
4
5
6
7
8
Corti M-C, Guralnik JM, Salive ME, et al. Clarifying the direct relation between total cholesterol levels and death from coronary heart disease in older persons. Ann Intern Med 1997; 126: 753–60. US Renal Data System. 1999 Annual Data Report. Bethesda: National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, April 1999. Foley RN, Parfrey PS, Sarnak MJ. Clinical epidemiology of cardiovascular disease in chronic renal disease. Am J Kidney Dis 1998; 32 (suppl 3): S112–19. Jones CA, McQuillan GM, Kusek JW, et al. Serum creatinine levels in the US population: Third National Health and Nutrition Examination Survey. Am J Kidney Dis 1998; 32: 992–99. Wannamethee SG, Shaper AG, Perry IJ. Serum creatinine concentration and risk of cardiovascular disease. Stroke 1997; 28: 557–63. Foley RN, Parfrey PS, Harnett JD, et al. Clinical and echocardiographic disease in patients starting end-stage renal disease therapy. Kidney Int 1995; 47: 186–92. London GM, Marchais SJ, Guerin AP, Metivier F, Pannier B. Cardiac hypertrophy and arterial alterations in end-stage renal disease; hemodynamic factors. Kidney Int 1993; 43 (suppl 41): S42–49. Schwarz U, Buzello M, Ritz E, et al. Morphology of coronary
151
For personal use only. Not to be reproduced without permission of The Lancet.
REVIEW
9
10
11
12
13
14
15
16
17
18
19
20 21
22
23
24
25
26
27
28
29
atherosclerotic lesions in patients with end-stage renal failure. Nephrol Dial Transplant 2000; 15: 218–23. Rostand SG, Kirk KA, Rutsky EA. Dialysis-associated ischaemic heart disease: insights from coronary angiography. Kidney Int 1984; 25: 653–59. Perneger TV, Klag MJ, Whelton PK. Cause of death in patients with end-stage renal disease: death certificates vs registry reports. Am J Public Health 1993; 83: 1735–38. Clarke R, Shipley M, Lewington S, et al. Underestimation of risk associations due to regression dilution in long-term follow-up of prospective studies. Am J Epidemiol 1999; 150: 341–53. Edmunds ME, Russell GI. Hypertension in renal failure. In: Swales JD, ed. Textbook of hypertension. Oxford: Blackwell Scientific Publications, 1994: 798–810. Buckalew VM Jr, Berg RL, Wang SR, Porush JG, Rauch S, Schulman G, and the Modification of Diet in Renal Disease Study Group. Prevalence of hypertension in 1795 subjects with chronic renal disease: the Modification of Diet in Renal Disease Study baseline cohort. Am J Kidney Dis 1996; 28: 811–21. Levin A, Thompson CR, Ethier J, et al. Left ventricular mass index increase in early renal disease: impact of decline in hemoglobin. Am J Kidney Dis 1999; 34: 125–34. Zager PG, Nikolic J, Brown RH, et al. “U” curve association of blood pressure and mortality in dialysis patients. Kidney Int 1998; 54: 561–69. Foley RN, Parfrey PS, Harnett JD, Kent GM, Murray DC, Barre PE. Impact of hypertension on cardiomyopathy, morbidity and mortality in end-stage renal disease. Kidney Int 1996; 49: 1379–85. MacMahon S. Blood pressure and the risks of cardiovascular disease. In: Swales JD, ed. Textbook of hypertension. Oxford: Blackwell Scientific Publications, 1994: 46–57. Levy D, Larson MG, Vasan RS, Kannel W, Ho KKL. The progression from hypertension to congestive heart failure. JAMA 1996; 275: 1557–62. Prospective Studies Collaboration. Cholesterol, diastolic blood pressure, and stroke: 13 000 strokes in 450 000 people in 45 prospective cohorts. Lancet 1995; 346: 1647–53. Oda H, Keane WF. Lipid abnormalities in end stage renal disease. Nephrol Dial Transplant 1998; 13 (Suppl 1): 45–49. Fliser D, Pacini G, Engelleiter R, et al. Insulin resistance and hyperinsulinemia are already present in patients with incipient renal disease. Kidney Int 1998; 53: 1343–47. Rajman I, Harper L, McPake D, Kendall MJ, Wheeler DC. Lowdensity lipoprotein subfraction profiles in chronic renal failure. Nephrol Dial Transplant 1998; 13: 2281–87. Grützmacher P, März W, Peschke B, Gross W, Schoeppe W. Lipoproteins and apolipoproteins during the progression of chronic renal disease. Nephron 1988; 50: 103–11. Lowrie EG, Lew NL. Death risk in hemodialysis patients: the predictive value of commonly measured variables and an evaluation of death rate differences between facilities. Am J Kidney Dis 1990; 15: 458–82. Degoulet P, Legrain M, Réach I, et al. Mortality risk factors in patients treated by chronic hemodialysis: Report of the Diaphane Collaborative Study. Nephron 1982; 31: 103–10. Law MR, Wald NJ, Wu T, Hackshaw A, Bailey A. Systematic underestimation of association between serum cholesterol concentration and ischaemic heart disease in observational studies: data from the BUPA study. BMJ 1994; 308: 363–66. Baigent C, Armitage J. Cholesterol reduction among patients at increased risk of coronary heart disease. Proc R Coll Physicians Edinb 1999; 29 (Suppl 5): 10–15. Bianchi S, Bigazzi R, Campese VM. Microalbuminuria in essential hypertension: significance, pathophysiology, and therapeutic implications. Am J Kidney Dis 1999; 34: 973–95. Warwick GL, Packard CJ. Lipoprotein metabolism in the nephrotic syndrome. Nephrol Dial Transplant 1993; 8: 385–96.
152
30 Law MR, Wald NJ, Thompson SG. By how much and how quickly does reduction in serum cholesterol concentration lower risk of ischaemic heart disease? BMJ 1994; 308: 367–73. 31 Gordon DJ, Probstfield JL, Garrison RJ, et al. High-density lipoprotein cholesterol and cardiovascular disease: four prospective American studies. Circulation 1989; 79: 8–15. 32 Sechi LA, Zingaro L, De Carli S, Sechi G, Catena C, Falleti E. Increased serum lipoprotein (a) levels in patients with early renal failure. Ann Intern Med 1998; 129: 457–61. 33 Kronenberg F, Utermann G, Dieplinger H. Lipoprotein[a] in renal disease. Am J Kidney Dis 1996; 27: 1–25. 34 Kronenberg F, König P, Lhotta K, et al. Apolipoprotein[a] phenotype-associated decrease in lipoprotein[a] plasma concentrations after renal transplantation. Arterioscler Thromb 1994; 14: 1399–404. 35 Danesh J, Collins R, Peto R. Lipoprotein (a) and coronary heart disease: meta-analysis of prospective studies. Circulation 2000 (in press). 36 Jensen JS, Borch-Johnsen K, Jensen G, Feldt-Rasmussen B. Microalbuminuria reflects a generalised transvascular albumin leakiness in clinically healthy subjects. Clin Sci 1995; 88: 629–33. 37 Gilbert RE, Cooper ME, McNally PG, O’Brien RC, Taft J, Jerums G. Microalbuminuria; prognostic and therapeutic implications in diabetes mellitus. Diabet Med 1994; 11: 636–45. 38 Dinneen SF, Gerstein HC. The association of microalbuminuria and mortality in non-insulin-dependent diabetes mellitus: a systematic overview of the literature. Arch Intern Med 1997; 157: 1413–18. 39 Yudkin JS, Forrest RD, Jackson CA. Microalbuminuria as predictor of vascular disease in non-diabetic subjects. Lancet 1988; ii: 530–33. 40 Arnadottir M, Hultberg B, Nilsson-Ehle P, Thysell H. The effect of reduced glomerular filtration rate on plasma total homocysteine concentration. Scand J Clin Lab Invest 1996; 56: 41–46. 41 Chauveau P, Chadefaux B, Coudé M, et al. Hyperhomocysteinaemia, a risk factor for atherosclerosis in chronic uremic patients. Kidney Int 1993; 43 (suppl 41): S72–77. 42 van Guldener C, Janssen MJFM, Stehouwer CDA, et al. The effects of renal transplantation on hyperhomocysteinaemia in dialysis patients, and the estimation of renal homocysteine extraction in patients with normal renal function. Neth J Med 1998; 52: 58–64. 43 Danesh J, Lewington S. Plasma homocysteine and coronary heart disease: systematic review of published epidemiological studies. J Cardiovasc Risk 1998; 5: 229–32. 44 Gabay C, Kushner I. Acute-phase proteins and other systemic responses to inflammation. N Engl J Med 1999; 340: 448–54. 45 Danesh J, Collins R, Appleby P, Peto R. Association of fibrinogen, C-reactive protein, albumin, or leukocyte count with coronary heart disease. JAMA 1998; 279: 1477–82. 46 Irish A. Cardiovascular disease, fibrinogen and the acute phase response: associations with lipids and blood pressure in patients with chronic renal disease. Atherosclerosis 1998; 137: 133–39. 47 Scarabin P-Y, Bara L, Ricard S, et al. Genetic variation at the -fibrinogen locus in relation to plasma fibrinogen concentrations and risk of myocardial infarction: the ECTIM study. Arterioscler Thromb 1993; 13: 886–91. 48 Foley RN, Parfrey PS, Harnett JD, Kent GM, Murray DC, Barre PE. The impact of anaemia on cardiomyopathy, morbidity and mortality in end-stage renal disease. Am J Kidney Dis 1996; 28: 53–61. 49 Pascual J, Teruel JL, Moya JL, Liaño F, Jiménez-Mena M, Ortuño J. Regression of left ventricular hypertrophy after partial correction of anaemia with erythropoietin in patients on hemodialysis: a prospective study. Clin Nephrol 1991; 35: 280–87. 50 Besarab A, Bolton WK, Browne JK, et al. The effects of normal as compared with low hematocrit values in patients with cardiac disease who are receiving hemodialysis and epoietin. N Engl J Med 1998; 339: 584–90. 51 Baigent C, Wheeler D. Should we reduce blood cholesterol to prevent cardiovascular disease among patients with chronic renal failure? Nephrol Dial Transplant 2000 (in press).
THE LANCET • Vol 356 • July 8, 2000
For personal use only. Not to be reproduced without permission of The Lancet.