In a recent article published in the American Journal of Kidney Diseases, Nigwekar et al examined data from an incident dialysis patient cohort (ArMORR) and concluded that hyponatremia was related in cross-sectional analyses with markers of mineral metabolism, and in a longitudinal analysis with one year mortality. The authors’ work builds on previous literature relating hyponatremia to death in a prevalent dialysis cohort, and additionally generates hypotheses about a link through mineral metabolism pathways.
The study draws patients from a well-established cohort, with unique features including ascertainment of biochemical laboratory data at the first dialysis treatment, and comorbid conditions added prospectively. As such, the investigators were able to study a truly incident patient population, in comparison to other dialysis cohorts that enrolled prevalent patients at 90 days. This incident population allowed the investigators to examine the role of hyponatremia and its association with mineral metabolism parameters, independent of dialysis treatment practices.
The findings from this study showing a positive association between hyponatremia and mortality (HR 1.42, CI 1.19-1.69) were robust to adjustment for several potential confounders. In addition, there were cross sectional associations relating hyponatremia to a two-fold higher risk of hypercalcemia (HR 2.03, CI 1.14-3.59), and a greater than 30% higher risk of elevated alkaline phosphatase levels and hypoparathyroidism. These findings should be considered hypothesis generating, as they were not observed in propensity matched analyses, and as such, may be due to residual confounding.
The link of hyponatremia with death, however, was consistent in multivariate and propensity based analyses, and confirms findings from previous studies in the general and prevalent dialysis populations. In these studies, hyponatremia may be a surrogate for advanced age, frailty, and worsening congestive heart failure/liver disease. It may also be a causal abnormality that can be modified with dialysis prescription or medications. The study by Nigwekar et al. further raises the possibility for causation, and highlights the need for prospective studies to truly test this hypothesis.
Navdeep Tangri, MD, PhD, FRCPC
Assistant Professor, Division of Nephrology, University of Manitoba