Effects of Volume Replacement for Urinary Losses from Mannitol Diuresis on Brain Water in Normal Rats

Background/Objective: It is frequently recommended that urine output following perioperative mannitol administration be replaced 1:1 with an isotonic crystalloid solution. It is possible that this strategy could increase brain water by reducing the serum osmolality achieved with prior mannitol administration. Therefore, brain water content of rats treated with mannitol alone or mannitol plus normal saline (NS) was studied over a range of urinary replacement ratios. Methods: Male Wister rats received mannitol 3.2 gm/100 gm infused over 45 min followed by hourly determinations of urine output (UO). Control animals received no additional therapy, whereas animals undergoing intervention received hourly replacement of their urinary losses with 0.9% NS in decreasing NS:UO ratios (1:1, 1:2, 1:3). Three hours after completion of the mannitol infusion, a final tally of UO was made. At that time in all animals, blood was obtained for determination of hemoglobin and electrolyte concentrations and plasma osmolality. Following that, the animals were sacrificed to determine brain water content. Additional groups underwent the same protocol but for 5 h with 1:1 urinary replacement, or received a volume of NS equal to that of the mannitol administered to all other control and intervention animals. Results: 1:1 replacement of urinary loss with NS following mannitol administration was associated with brain water content indistinguishable from control animals receiving only a volume of NS equal to that of the mannitol administered to all other groups. Regression analysis demonstrated a decrease in the final brain water content of 0.67% (CI95 0.43–0.92, p < 0.001) per replacement level as NS:UO replacement ratios were decreased from 1:1 to 1:2 and, finally to 1:3. At the final NS:UO replacement ratio of 1:3, brain water content was indistinguishable from the control group receiving mannitol without NS replacement (p = 0.48) For 1:1 replacement following mannitol, brain water did not differ between experiments of 3 or 5 h duration (p = 0.52). Conclusions: In rats, NS replacement of UO 1:1 following mannitol administration leads to brain water content no different than if NS had been given in place of mannitol. Only when the NS:UO replacement ratio was 1:3, brain water was similar to that of control animals receiving mannitol alone. The recommendation to replace UO 1:1 with an equal volume of isotonic crystalloid following perioperative mannitol administration must recognize how this strategy could elevate brain water content compared to less vigorous replacement of UO.