Salt-The Forgotten Renal Risk Factor: Sodium Intake Affects Urinary Albumin Excretion Especially in Overweight Subjects. J Intern Med 256: 324-330, 2004.

Abstract

Decades ago, following the work of Ambard and Beaujard (1), dietary sodium restriction became a mainstay of the treatment of patients with renal disease on both sides of the Atlantic, although admittedly based on very limited evidence (2,3). Subsequently, the adverse effect of salt was ascribed to hypervolemia and hypertension. After diuretics and effective antihypertensive agents had become available, dietary restriction of salt had fallen out of fashion in the perhaps premature belief that any adverse effect of salt can be prevented by augmenting natriuresis using diuretics, although at least in experimental studies this was by no means the case (4). With respect to BP and cardiovascular risk, salt has recently seen a renaissance. In the Dietary Approaches to Stop Hypertension (DASH) Study, lowering of dietary salt from usual to low intake with simultaneous further modifications of the diet decreased systolic BP by 7.1 mmHg overall and by 11.5 mmHg in hypertensive individuals (5). A meta-analysis concluded that the BP difference between intake of 12 g and 3 g of salt, i.e., sodium chloride, was 5.6/3.2 mmHg in hypertensive and 3.5/1.8 mmHg in normotensive subjects (6). In a prospective Finnish study, a 100 mmol/d difference in sodium intake was associated with a relative risk of coronary heart disease of 1.5 and of all cause mortality of 1.2, independent of BP and classic cardiovascular risk factors (7). In contrast to impressive experimental studies, for instance in intact rats and in antithymocyte serum nephritis as well as postangiotensin infusion (8–10), very little clinical information has been presented. It is therefore of considerable interest that in a cross-sectional cohort study of 7850 subjects aged 28 to 75 yr in the city of Groningen (The Netherlands), a significant positive relationship was demonstrated between dietary sodium intake and urinary albumin excretion. The association was independent of gender, age, BP, body mass index (BMI), waist-to-hip ratio, serum cholesterol, or glucose concentration. The difference between the lowest quintile of sodium excretion (mean, 70.5 mmol/d) and the highest (mean, 220 mmol/d) was 7.5 mg albumin/d (5.3 to 13.3) versus 11.1 mg albumin/d (7.3 to 21.7). It must be admitted that other indicators of food intake were also positively associated with albuminuria, but the effect of sodium intake on urinary albumin excretion was independent of such other food constituents. Of interest with respect to potential pathogenetic mechanisms is the observation that a significant interaction was noted between urinary sodium and BMI. At any given sodium intake, subjects with a higher BMI had higher urinary albumin excretion than subjects with a lower BMI. In a large cohort, this observation amplifies and extends previous observations of the Montpellier group (11). It also confirms with more reliable assessment of sodium intake the previous observations in the National Health and Nutrition Examination Survey (NHANES I) study (12). The study has the merit of putting squarely back onto the map a hotly debated issue that has caused considerable controversy in the past, i.e., sodium chloride intake in hypertensive and particularly in renal patients. It does not answer, however, which mechanism(s) account(s) for the link. One can only speculate on the basis of animal experiments. High salt has been shown to increase left ventricular mass, AT1 receptor expression, and aldosterone synthase activity in Wistar Kyoto rats despite unchanged BP and suppression of the renin system (13). In addition, salt was also shown to induce myocardial as well as renal fibrosis (14). In the aortic endothelium (15) and in glomeruli (16), high salt increases the activities of the p38 mitogen-activated protein (MAP) kinase. In the kidney, high salt intake also increases oxidative stress (8). Of particular interest is the activation of TGF-β induced by high salt (17,18). An interesting interrelation exists between nitric oxide (NO) and TGF-β: NO inhibits TGF-β (19). The high salt sensitivity of BP in patients with renal failure may be explained by the finding that the BP response to high salt is dependent on NO production (20). In humans, the increase in renal plasma flow in response to salt loading was reduced by inhibition of NO production (21), of considerable interest because the endogenous NO synthase inhibitor L-NMMA is increased very early on in renal failure (22). In summary, there is no lack of potential pathomechanisms in the kidney (18) to speculate about the interpretation of the above findings; what is lacking is strong human evidence. In a small, retrospective series, a relation between progression and salt intake was noted (23). Certainly this issue deserves more attention. In this respect, the paper by Verhave et al. (24) will hopefully serve to wake up the renal community: There may be a chance out there to improve renal outcomes (25).