Use of serum osmolality to identify heart disease stage in dogs and relationship to mathematical chloride correction

Abstract Background Heart failure‐associated hypochloremia can be depletional from diuretics or dilutional from water retention. Serum osmolality reflects water balance but has not been evaluated in dogs with heart disease. Hypothesis To determine if serum osmolality is related to heart disease stage and amount of mathematical correction of serum chloride (Cl−) concentrations in healthy dogs and dogs with myxomatous mitral valve degeneration (MMVD). Animals Seventy‐seven dogs (20 healthy, 25 Stage B MMVD, 32 Stage C/D MMVD). Methods Serum Cl− concentrations were mathematically corrected. Osmolality was calculated (calOsm) and directly measured by freezing point depression (dmOsm) and compared by Bland‐Altman analysis. Biochemical variables and osmolality were compared among healthy, Stage B, and Stage C/D dogs. Correlations were explored between osmolality and biochemical variables. Median and range are presented. P < .05 was considered significant. Results The calOsm was different among groups (P = .003), with Stage B (310 mOsm/kg; 306, 316) and C/D dogs (312 mOsm/kg; 308, 319) having higher calOsm than healthy dogs (305 mOsm/kg; 302, 308). Osmolality methods were moderately correlated (P < .0001, rs = .46) but with proportional bias and poor agreement. The amount of Cl− correction was negatively correlated with calOsm (P < .0001, rs = −.78) and dmOsm (P = .004, rs = −.33). Serum bicarbonate concentration was negatively correlated with Cl− (P < .0001, rs = −.67). Conclusions and Clinical Importance Dogs with Stage B and Stage C/D heart disease had higher calOsm than healthy dogs. Osmolality was inversely related to the amount of Cl− correction, which supports its use in assessing relative body water content. Poor agreement between calOsm and dmOsm prevents methodological interchange.