Hyponatremia is the electrolyte abnormality most often encountered in patients admitted for heart failure (HF).1–3 Its prevalence ranges between 18 and 30%,4–8 although it is lower in stable HF outpatients (10-17%).9–12 It also varies with the severity and duration of HF, the criteria used to define it, medication prescribed and drug doses, particularly of diuretics. Most studies define hyponatremia as Na <135 mEq/l, although some use quartiles, and a linear relation has been observed between risk of short- and long-term total mortality and in-hospital cardiovascular morbidity and mortality and sodium <140 mEq/l. Other studies have demonstrated a U-curve for values between 140 and 145 mEq/l.4–12 The value of natremia as a risk indicator is thus evident even for values considered normal; for every 3 mEq/l fall in blood sodium the risk of in-hospital mortality rises by 19.5% in HF with systolic dysfunction and by 8.6% in HF with preserved systolic function, and mortality 60-90 days after discharge rises by 10-18%.4,5 Pathophysiologically, hyponatremia is multifactorial, usually due to hypervolemia related to hyperactivity of the sympathetic and rein-angiontensin-aldosterone systems and their effects on the kidneys. This hyperactivity is in turn secondary to reduced cardiac output and consequent reduced stimulation of aortic and carotid baroreceptors, which also induces an excessive increase in non-osmotic secretion of vasopressin and hence fluid retention.1–3 However, diuretic therapy can transform this to normovolemia or even hypovolemia.1–3 Recent reports suggest that chronic inflammation mediated by interleukin-6 may play a part in this increased vasopressin secretion.13

The relationship between blood pressure (BP) and HF, which may appear a simple one due to the known importance of high BP in the etiopathogenesis of HF, becomes more complex in light of the results of various studies that report worse prognosis in HF patients with lower office systolic BP, particularly below 120 mmHg, on admission, at discharge or even as outpatients, especially if they have reduced ejection fraction (EF).14–17 A meta-analysis of 10 studies including over 8000 patients with stable chronic HF showed a reduction in mortality of 13% for every 10 mmHg higher systolic BP on admission.17 However, some authors describe a U-shaped curve for long-term mortality with values between 120 and 139 mmHg, with slight differences between reduced and preserved EF.15

Most studies on the prognostic value of blood sodium levels in HF have shown that hyponatremia is associated with significantly lower office BP, which is also independently associated with higher risk.4–12 The few studies on the prognostic value of ambulatory BP monitoring (ABPM) in HF show that non-dipper profiles (inverted dipper or riser) are more common in these patients, possibly related to increased sympathetic activity, and that these profiles are significantly associated with worse prognosis than for a dipper profile,18–20 as has been shown in hypertensive patients and in population-based studies.21,22 However, to the best of my knowledge no studies have set out to relate natremia levels with BP values recorded by ABPM or with possible changes in circadian patterns.

The study by Arévalo-Lorido et al.23 published in this issue of the Journal sets out to analyze precisely this relationship in 175 patients with stable HF, mostly of hypertensive etiology, assessed by ABPM. The authors found a significant correlation between serum sodium levels and systolic BP by ABPM in multivariate analysis, as well as a low prevalence of dipper pattern (20.5%), as has been reported both in patients hospitalized for acute HF and in those with stable chronic HF.18–20 Furthermore, although they did not specifically examine the subgroup with Na <135 mEq/l, the usual definition of hyponatremia, they found that in the lowest quartile (133-139 mEq/l), both mean 24-hour and mean daytime systolic BP levels by ABPM were consistently lower than in the other quartiles. It should be noted that there were no significant differences in drugs prescribed in the different quartiles, including diuretics. There was also a lower prevalence of dipper pattern and a higher proportion of risers in the lowest quartile. Although, as pointed out above, non-dipper profiles are common in HF with either reduced or preserved EF, the riser pattern has recently been particularly strongly associated with HF with preserved EF and with increased risk.19,20 In the present study, most patients in all quartiles of serum sodium levels had preserved EF, which may explain the high proportion of risers (38%), but this contrasts with the lower proportion seen in hypertension (5-19%).

The authors point out certain limitations of their study, including the fact that they were unable to form a true hyponatremic group, i.e. with sodium level <135 mEq/l (although several studies have shown increased risk with levels below 140 mEq/l5,9,10), and that they could not analyze HF with reduced and preserved EF separately, due to the preponderance of the latter, which could have further clarified some of the inter-relationships that have recently been reported, especially the increased risk seen in HF patients with riser profile and with preserved (but not reduced) EF.20 Nevertheless, we consider their work to be original and of considerable interest, as it confirms the previously described association between low sodium levels and lower office BP in stable HF patients, only this time with 24-hour ABPM, with its added diagnostic and prognostic value. Furthermore, they raise the possibility that the changes in circadian profile found mostly in the lowest serum sodium quartile, particularly the lower prevalence of night-time dipping and the significantly higher proportion of riser profile, may improve risk stratification, higher risk being associated with lower sodium levels, as described in studies of both hospitalized4–8 and outpatient9–12 HF populations. The study may also contribute toward a better understanding of the pathophysiology involved, with specific serum sodium levels being used as markers or risk factors for overall mortality and cardiovascular morbidity and mortality in HF, and possibly as therapeutic targets.