Abstract
PURPOSE: The precise effects of non-steroidal anti-inflammatory drugs on the neuroendocrine hydro-electrolytic regulation are not precisely understood. The aim of this pilot study was to evaluate, in healthy subjects, the neuroendocrine response of the antidiuretic system to intravenous diclofenac infusion. METHODS: For this single-blinded, cross-over study, we recruited 12 healthy subjects (50% women). Test sessions were divided into three observation times (pre-test; test; 48 h post-test), which were repeated equally on two different occasions, with the administration of diclofenac (75 mg in saline solution 0.9% 100 cc) on 1 day, or placebo (saline solution 0.9% 100 cc) on another day. The night before the test the subjects were asked to collect a salivary cortisol and cortisone sample, which was repeated on the night of the procedure session. Serial urine and blood samples were collected on the test day (for osmolality, electrolytes, ACTH, cortisol, copeptin, MR-proADM, MR-proANP; the last three represent more stable and analytically reliable molecules than their respective active peptides). Moreover, the subjects were evaluated with the bioimpedance vector analysis (BIVA) before and after the test. Forty-eight hours after the end of the procedure urine sodium, urine potassium, urine osmolality, serum sodium and copeptin were revaluated together with BIVA. RESULTS: No significant changes in circulating hormone levels were observed; anyway, 48 h after diclofenac, BIVA showed a significant water retention (p < 0.00001), especially in extracellular fluid (ECF) (16.47 ± 1.65 vs 15.67 ± 1.84, p < 0.001). Salivary cortisol and cortisone tended to increase only the night after placebo administration (p = 0.054 cortisol; p = 0.021 cortisone). CONCLUSION: Diclofenac resulted in an increased ECF at 48 h, but this phenomenon seems to be associated with a greater renal sensibility to the action of vasopressin rather than with an increase in its secretion. Moreover, a partial inhibitory effect on cortisol secretion can be hypothesized.