Biochemical characteristics associated with spironolactone use in hospitalized patients with hypotonic hyponatraemia

Background Spironolactone has been associated with hypotonic hyponatraemia. However, its accompanying biochemical pattern – especially with concomitant loop diuretics – remains underdescribed. We characterized biochemical and epidemiological profiles of hospitalized patients with established hypotonic hyponatraemia according to spironolactone exposure.Methods Single-centre retrospective cohort of non-thiazide patients admitted with hypotonic hyponatraemia. Biochemical parameters were compared between spironolactone users (n=107) and non-users (n=427). The spironolactone-alone subgroup (n=27) was compared between (i) the non-diuretic group (n=374) and (ii) the spironolactone+furosemide group (n=80). We performed correlations between (a) admission serum sodium (s-Na) and in-hospital mortality; (b) spironolactone dose and biochemical parameters, also adjusted for furosemide dose; (c) serum potassium (s-K) and relevant parameters. Multivariable regression analyses were performed for s-K, s-Na, fractional excretion (FE)-K, FE-Cl, FE-H2O and serum urea (s-urea) as dependent variables to assess associations with spironolactone while accounting for potential confounders.Results After eGFR adjustment, spironolactone users showed higher s-K (p<0.001), s-urea (p<0.001) and s-uric acid (s-UA; p=0.01). The spironolactone-alone group showed higher s-K (p=0.003), and s-urea (p=0.001) compared to the non-diuretic group; lower dose (p=0.04), s-urea (p=0.04), serum creatinine (s-creatinine) (p<0.001), s-UA (p=0.02) and chronic heart failure prevalence (p<0.001); higher eGFR (p=0.002), and hypertension prevalence (p=0.048), compared to the spironolactone+furosemide group. Multivariable regression analyses identified independent associations of spironolactone with higher s-K (b=0.33, p=0.001) and with higher log-transformed s-urea (b=0.22, p<0.001), corresponding to a 24.5% increase in s-urea.Conclusions Spironolactone was independently associated with higher s-K and with higher s-urea levels. This pattern was not reflected by reduced urinary potassium excretion. Importantly, no clear evidence that concomitant furosemide materially modified this biochemical pattern was observed in the present analyses. Awareness of this pattern may support the phenotypic characterization of hyponatraemic patients receiving spironolactone. However, prospective studies are warranted to evaluate the diagnostic utility of this biochemical pattern in hyponatraemic patients receiving spironolactone.