OBJECTIVE: This study was designed to test the hypothesis that the addition of chloride to a lactate infusion would reduce the frequency of panic attacks. METHOD: The subjects included 14 healthy volunteers and 20 patients meeting the DSM-IV criteria for panic disorder. All subjects received an infusion of lactate dissolved in 0.9% sodium chloride and an infusion of lactate dissolved in 5% dextrose in water on separate days in a random-order, double-blind procedure. Blood pressure, heart rate, and panic symptoms were measured at 3-minute intervals during the infusions. The occurrence of panic attacks was ascertained through the subjects' reports of losing control, panicking, or "going crazy" and the presence of at least four Research Diagnostic Criteria symptoms of a panic attack. RESULTS: Fifteen (75%) of the patients with panic disorder reported a panic attack during one of the infusions or both; no healthy volunteers had a panic attack. The patients with panic disorder were significantly more likely to have a panic attack during the lactate/sodium chloride infusion than during the infusion of lactate/5% dextrose in water. The number of panic attack symptoms reported at 3-minute intervals did not differ between the two types of infusion. CONCLUSIONS: The coadministration of glucose resulted in a reduced sensitivity to the panicogenic effects of lactate. The hypothesis that adding chloride to the infusion would reduce the frequency of lactate-induced panic attacks was not supported.Abstract Teaser