Comment on: Ekinci et al. Dietary Salt Intake and Mortality in Patients With Type 2 Diabetes. Diabetes Care 2011;34:703-709

Ekinci et al. (1) report that lower 24-h urinary sodium excretion is paradoxically associatedwith an increased risk of all-cause and cardiovascular mortality in type 2 diabetes. They conclude, “Such data call into question universal recommendations that all adults should endeavor to reduce their salt intake.” The authors are to be commended for basing their prospective observational study on repeated 24-h urine excretions. However, we believe that their conclusions need to be treated with caution for several reasons. First, as the authors also state, this is an observational study, prone to unmeasured and residual confounding that may have played a dominant role in eliciting these paradoxical results. Second, and possibly a consequence of the first, the studied population may not be a representative population. In particular, Ekinci et al. showed that a lower systolic blood pressure was also associated with a significantly increased risk of allcause and cardiovascular mortality. Does this mean that we should question the value of blood pressure lowering as well? Moreover, subjects with a lower 24-h urinary sodium excretion were older, more often female, and less often obese. It is therefore likely that these subjects were smaller, had lower energy requirements, and accordingly lower overall food intake and proportionally lower sodium intake. It would be important therefore, to verify whether adjustment for body surface area or 24-h creatinine excretion, as a proxy for physical fitness, would affect the results. Third, a lack of a full range of sodium intake (in particular to the low end) may have hampered a correct conclusion. The lower tertile had,150 mmol Na/day cutoff, where the authors do not report the average of this group. This may still have been higher than the recommended intake of 5–6 g of salt daily. This would render it difficult to accept excess activation of the renin-angiotensin-aldosterone system (RAAS) or sympathetic nervous system as a mechanism for the increased mortality rates, as these neurohormonal pathways are typically activated when salt consumption is below 3 g/day (sodium excretion;50mmol/day). Indeed in a recent study in subjects with type 1 diabetes, both a very low (,50 mmol/day) and high urinary sodium excretion (.200 mmol/day) were associated with increased mortality, while subjects with the guideline recommended salt intake of 5–6 g were at lowest risk (2). These contrasting results are physiologically difficult to explain and illustrate again the complexity of the observational studies. No interaction was detected with the modality of antihypertensive therapy. This is surprising as moderate salt restriction consistently potentiates the effect of RAAS blockade on blood pressure, albuminuria, and hard renal end points (3). In patients with renal disease, salt restriction even overcomes resistance to RAAS blockade, likely by additional inhibition of tissue angiotensin-converting enzyme (4). Although observational studies can be useful to generate interesting hypotheses, they never provide the true answer. As the authors mention themselves, randomized controlled trials are necessary to truly assess the impact of salt reduction on mortality. We submit that the currently available data suggest that such trials may demonstrate that dietary salt reduction reduces the risk of mortality (5). HIDDO J. LAMBERS HEERSPINK, PHARMD PHD ARJAN KWAKERNAAK, MD PHD DICK DE ZEEUW, MD PHD GERJAN NAVIS, MD PHD