Microalbuminuria in essential hypertension: redefining the threshold

Over the last few years, assessment of the small amounts of urinary albumin excretion, the so-called microalbuminuria, has become an integrated marker of cardiovascular and/or renal risk in both diabetic 1 and non-diabetic elderly populations 2. In essential hypertension, measurement of microalbuminuria has increased despite the scarce evidence supporting any prognostic value 3 and the absence of evidence bolstering the contention that a reduction in microalbuminuria is followed by a decrease in risk. Beginning with the studies of microalbuminuria in essential hypertension, a clustering of risk factors has been reported in microalbuminuric subjects 4. Moreover, increased rates of early end-organ damage, defined as increased left ventricular mass 5 and/or carotid wall thickness 6, have been described. The development of both microalbuminuria and left ventricular hypertrophy (LVH) run in parallel due to the impact of not only haemodynamic, but also non-haemodynamic factors 7. Studies using an echo-based diagnosis of LVH revealed a high prevalence of echo-LVH among the microalbuminuric subjects 8. In this issue of the journal, the largest ever study regarding the association between microalbuminuria and ECG-based diagnosis of LVH is published 9. The authors found that microalbuminuria was related to the magnitude of ECG-LVH, independent of older age, office blood pressure, diabetes, serum creatinine or smoking, although the presence of diabetes, heavy smokers and several ethnic groups (African American, Hispanic and Asian) increased the rates of microalbuminuria. The percentage of subjects with microalbuminuria was surprisingly small (23%), taking account of the fact that the patients recruited into the LIFE study represent an older population with ECG-LVH. This figure may be influenced by previous antihypertensive treatments, since assessment of urinary albumin excretion (UAE) was performed 1–2 weeks after the withdrawal of prior antihypertensive therapy. Nevertheless, if we accept that LVH is a barometer of a high blood pressure load and a powerful surrogate for cardiovascular morbidity and mortality, then a high percentage of patients with ECG-LVH would be expected to have microalbuminuria. The fact that only 23% of patients have microalbuminuria in the LIFE study implies that the threshold UAE rate for defining ‘microalbuminuria’ in people with hypertension is too high. There is ample evidence supporting this viewpoint. In the majority of the published studies involving people with hypertension, microalbuminuria was defined as a UAE rate > 30 mg/24 h or > 20 μg/min in an overnight urine collection. A simpler, more practical alternative, as used in the LIFE study 9, is the measurement of the albumin : creatinine ratio in a single morning first void urine specimen (> 3.5 mg albumin/mmol creatinine was the threshold for the diagnosis of microalbuminuria). These reference values derive from studies performed in diabetic subjects, in whom the determinants of UAE differ, at least in part, from those acting in hypertension alone. The main purpose of documenting microalbuminuria in people with hypertension is that it provides a bedside marker of enhanced cardiovascular risk. Evidence from cross-sectional and prospective studies supports the adoption of a lower UAE rate for the detection of enhanced risk in individuals with hypertension. In the PREVEND study, a high–normal UAE (10–20 mg/l) was associated with hypertension 10. Moreover, continuous relationship between left ventricular mass and UAE rate from the range of normoalbuminuria was observed in newly diagnosed patients with essential hypertension 11. Furthermore, a baseline level of UAE in the normal range has been a predictor for the occurrence of microalbuminuria 12,13. In a 3-year follow-up study in initially untreated normoalbuminuric hypertensive subjects, each increment of 1 mg/24 h for baseline UAE increased the risk of microalbuminuria during antihypertensive treatment by 6% 13. A small number of prospective studies demonstrating the prognostic value of UAE in terms of morbidity and mortality have been published 14–18. Damsgaard et al. 14 conducted a follow-up study in an elderly population and demonstrated that subjects with UAE rate > 7.5 μg/min had a higher mortality rate compared to those with lower values. The median albumin excretion rate in the 31 individuals who died was 15 μg/min. In another study, from a population-based cohort of 2085 consecutive subjects 15, 79 developed ischaemic heart disease. When adjusted for other risk factors, the relative risk of ischaemic heart disease associated with microalbuminuria, defined as an albumin/creatinine ratio of only > 0.65 mg/mmol, was 2.3. Thus, this doubling of risk occurred at a UAE level of only one-fifth the commonly used threshold for the diagnosis of microalbuminuria. Similarly, in a population-based cohort aged 50–75 years, who were followed prospectively for 5 years, microalbuminuria, defined as albumin/creatinine ratio > 2.0 mg/mmol, was associated with a four-fold increase in cardiovascular mortality and an approximately two-fold increase in all-cause mortality 16. Furthermore, in a cohort of postmenopausal women living in Utrecht, the cardiovascular age-adjusted mortality rate for hypertensive women who were in the highest quintile of UAE was 4.3-fold greater compared to that observed in women without detectable UAE 17. The highest quintile corresponded to an albumin/creatinine ratio > 2.41 mg/mmol. Finally, in the cohort of subjects included in the HOPE study 18, the relative risk of the primary end-point in the fourth quartile was 1.97 (albumin/creatinine ratio > 1.62 mg/mmol) compared to the lowest quartile of albumin/creatinine ratio. For every 0.4 mg/mmol increase in albumin/creatinine ratio level, the adjusted hazard of major cardiovascular events increases by 5.9%. In all of these studies, the UAE threshold required to define increased risk in people with hypertension was far below the threshold UAE for the current definition of microalbuminuria. This probably reflects the fact that mircoalbuminuria was originally defined on the basis of increased risk for developing diabetic nephropathy. It is now clear that its significance extends beyond nephropathy and it is a marker for generalized vascular dysfunction in the diabetic and non-diabetic population. If, we accept that the UAE rate associated with increased cardiovascular risk in people with hypertension should be redefined, how do we go about it? There are several possible approaches to defining reference values for ‘normal’ and ‘abnormal’ UAE rates in people with hypertension. Longitudinal population-based studies of sufficient size and ethnic diversity are required to define the relationship between UAE rates and outcome. The LIFE study reported in this issue will provide data on the association between baseline UAE rate and outcome over 5 years in a high-risk, older population. Ongoing and future studies should continue to include baseline estimations of UAE rates to develop a comprehensive database on the relationship between UAE rate and outcome in a wide range of populations at various levels of risk. It will also be important to define whether therapy-induced reductions in UAE rate are associated with a corresponding and proportionate reduction in risk. In this regard, the monitoring of UAE rate might ultimately become useful as a surrogate for efficiency of blood pressure control and cardiovascular risk reduction. UAE decreases in proportion to blood pressure reduction in people with hypertension 19. Inhibition of the renin–angiotensin system either with angiotensin converting enzyme inhibition or angiotensin II receptor antagonism (AIIRA) is particularly effective at reducing UAE rates, with the response being greater than that seen with other forms of antihypertensive therapy 20. What is unclear is whether such differential effects of different classes of drug therapy on UAE rate will impact on outcome. In this regard, the LIFE study has compared the AIIRA losartan with the β-blocker atenolol. If there are the anticipated differences in the antiproteinuric actions of these two therapies, despite similar blood pressure control, it will be of great interest to see whether this relates to an impact on outcome. Even if there is not, this does not mean that the detection of UAE rate is of no importance. We already know that, despite antihypertensive treatment and whatever the class of drug used, a progressive increase of UAE is occasionally observed in some patients 13. This appears to represent suboptimal systolic blood pressure control and/or a progressive rise in baseline glucose values 14. As the data accumulate from the ongoing trials and the normal range for UAE in hypertension is better defined, the measurement of UAE rates may serve a dual purpose. Microalbuminuria is a powerful way of identifying those at risk for multiple cardiovascular risk factor intervention. Moreover, a failure to regress microalbuminuria may indicate the inadequacy of that intervention. Time will tell.