1. Materials and Methods: Animals: -Yorkshire pigs (Sus scrofa), castrated males, n=34, body mass range 17-22 kg Groups: 1) Untreated controls 2) Surgical controls * : injected with 60 mL saline into caecum and large intestine 3) Treated animals * : injected with varying volumes (12-83 mL) and activities (200- 2200  mol CH 4 /min) of Methanobrevibacter smithii into the caecum and large intestine * Surgery performed under anesthesia, abdomen opened to allow access to caecum and large intestine. Animals studied immediately after recovery. Dive simulation: -Chamber pressurized to 24 bar (21.6-22.9 bar PH 2, 0.3-0.5 bar PO 2 ) for 3 hr. -Decompression rate 0.9 bar/min to 11 bar; animals observed for 1 hr for DCS - Euthanized in chamber on confirmation of DCS or at end of hour. Measurements: -Chamber gases analyzed by gas chromatography for H 2, O 2, He, N 2, and CH 4 -CH 4 output rate (  mol CH 4 /min) from the chamber was used as an indicator of the CH 4 production rate from pigs. -Severe symptoms of DCS included: walking difficulties, fore and/or hind limb paralysis, falling, convulsions. Statistical analysis: -Variables: body mass, PH 2, %H 2, pre- and post decompression CH 4 output rate, activity injected, and ambient temperature, were tested for the univariate case for significance against outcome. - Only variables with a P<0.25 (Wald’s test) were considered and tested for the multivariate case. - Backward and forward steps were used to test for the multivariate case. Results: 1) Animals injected with methanogens had a 43 % incidence of DCS which was significantly lower than the 80% DCS incidence of the untreated and surgical control animals. 2) There was a positive correlation between increasing methanogenic activity injected into the pig versus CH 4 output from the chamber. 3) Body mass, pre- and post-decompression CH 4 output rate, and activity injected were the only univariate variables with a P<0.25 (Wald’s test) and further considered for the multivariate case. 4) Activity injected was the only significant variable, and there was a negative correlation between activity injected versus DCS incidence (P<0.02, likelihood ratio test). Abstract: BACKGROUND: Biochemical decompression, a novel approach for decreasing decompression sickness (DCS) risk by increasing the tissue washout rate of the diluent gas, was tested in an animal model during simulated H 2 dives. Logistic regression was used to show that the DCS risk was negatively correlated with the biochemical activity used to decrease the tissue gas burden. METHODS: To increase the washout rate of H 2, we used a H 2 -metabolizing microbe (Methanobrevibacter smithii) that converts H 2 to CH 4. Pigs (Sus scrofa, 19.4  1.3 kg) were either untreated controls (n=10), saline-injected controls (n=10), or injected with M. smithii into the large intestine (n=14; varying total activity 200-2200 mmol CH 4 /min). The pigs were placed in a dry hyperbaric chamber and compressed to 24 bar (20.6-22.9 bar H 2, 0.3-0.5 bar O 2 ). Chamber gases (O 2, N 2, He, H 2, and CH 4 ) were monitored using gas chromatography throughout the dive. After 3 h, the pigs were decompressed to 11 bar at 0.9 bar/min, and observed for severe symptoms of DCS for 1 h. RESULTS: Pigs with M. smithii had a 46% lower incidence of DCS compared to controls (6/14 vs. 16/20). Using logistic regression techniques, we showed that the DCS risk rate decreased with increasing total activity of microbes injected (P = 0.015, Log-Likelihood ratio = 5.965, 1 df). CONCLUSION: Reducing the tissue concentration of the inert gas significantly reduced the risk of DCS in a pig model. (NMRDC #61153N MR04101.00D-1103). Decompression sickness incidence for untreated control (UC), surgical control (SC), and treated (T) pigs at 24 bar. ‡Incidence in T (6/14) was significantly lower than in UC and SC animals (P<0.05, 1-tailed  2 - test). The DCS incidence in the UC group (9/10) was not significantly different from that of the SC group (7/10, P > 0.29, Fisher’s Exact test). Error bars represent 95% binomial confidence limits. Total activity injected was positively correlated (P<0.01) with average CH 4 output rate from chamber during the last hour at 24 bar. Introduction: A novel method for safer decompressions from hydrogen dives involves the use of H 2 -metabolizing microbes (Kayar et al., Am. J. Physiol. 275: R677-682, 1998). These microbes convert some of the H 2 dissolved in the diver’s tissues into methane (4 H 2 + CO 2 => CH 4 + 2 H 2 O). The conversion decreases the incidence of DCS by reducing the body burden of excess H 2 released upon decompression (Fahlman et al., FASEB J. 13:A408, 1999). In this study, we further hypothesized that increasing H 2 wash-out rate by increasing the methanogenic activity supplied to the animal will decrease the DCS incidence in a dose-dependent manner. DECOMPRESSION SICKNESS RISK CORRELATED WITH ACTIVITY OF H 2 -METABOLIZING MICROBES INJECTED IN PIGS PRIOR TO DIVES IN H 2. A Fahlman, SR Kayar, WJ Becker, WC Lin, and WB Whitman. Naval Medical Research Center, Bethesda, MD 20889-5607 and Dept. of Microbiology, Univ. of Georgia, Athens, GA 30602.. CONCLUSION: -Injection of methanogenic microbes reduces the DCS incidence in pigs during simulated H 2 dives. - The DCS incidence decreased with increasing methanogenic activity injected, supporting a dose -response effect. ‡ Points of Contact: NMRI-Principal Investigator Dr. Susan Kayar, E-mail: kayars@nmripo.nmri.nnmc.navy.mil Research Assistant Andreas Fahlman, E-mail: fahlmana@nmripo.nmri.nnmc.navy.mil