Zoology.ubc.ca

RESEARCH ARTICLE
Dehydration Effects on the Risk of Severe
Decompression Sickness in a Swine Model
FAHLMAN A, DROMSKY DM. Dehydration effects on the risk of
randomly may well contain physically or biologically severe decompression sickness in a swine model. Aviat Space Envi-
definable variables that have not yet been identified as ron Med 2006; 77:102– 6.
important factors, and, therefore, have not been exper- Background: Several physiological factors have been suspected of
affecting the risk of decompression sickness (DCS), but few have been imentally controlled. Diving textbooks are filled with thoroughly studied during controlled conditions. Dehydration is a po- potential factors that may alter DCS risk, but many of tential factor that could increase the risk of DCS. It has been suggested these have not been tested under controlled conditions.
that hydration may enhance inert gas removal or increase surface ten- Consequently, the challenge is to identify those that sion of the blood. Hypothesis: Dehydration increases DCS risk. Meth-
ods: Littermate pairs of male Yorkshire swine (n ϭ 57, mean Ϯ 1 SD
significantly alter DCS risk under a controlled experi- 20.6 Ϯ 1.7 kg) were randomized into two groups. The hydrated group mental setting. Clearly defining risk factors, their ef- received no medication and was allowed ad lib access to water during fect(s), and magnitude would be of great benefit for a a simulated saturation dive. The dehydrated group received intravenous wide variety of people such as aviators, astronauts, 2 mg ⅐ kgϪ1 Lasix (a diuretic medication) without access to water commercial and military divers, caisson workers, sport throughout the dive. Animals were then compressed on air to 110 ft ofseawater (fsw, 4.33 ATA) for 22 h and brought directly to the surface at divers, and hyperbaric chamber personnel.
a rate of 30 fsw ⅐ minϪ1 (0.91 ATA ⅐ minϪ1).
Fluid balance is potentially one of these risk factors non-fatal central nervous system (CNS) or cardiopulmonary DCS were for DCS that has received little consideration. Many recorded. Results: In the hydrated group University of British Columbia Library
diving textbooks report the perception that DCS risk pulmonary DCS ϭ 9, CNS DCS ϭ 2, Death ϭ 4. In the depends on hydration state. However, on closer scru- (n ϭ 26): DCS ϭ 19, cardiopulmonary DCS ϭ 19, CNS DCS ϭ 6,Death ϭ 9. Dehydration significantly increased theWed, 15 Mar 2006 07:48:21 tiny, the evidence to support these notions is often DCS and death. Specifically, it increased the risk of cardiopulmonary anecdotal, contradictory, or at best suggestive that more DCS, and showed a trend toward increased CNS DCS. In addition, investigation is in order (5). One animal study in rats dehydrated subjects manifested cardiopulmonary DCS sooner and showed that the DCS incidence in dehydrated animals showed a trend toward more rapid death (p Ͻ 0.1). Conclusion: Hydra-
was 71% while in hydrated rats it was 55% (19). These tion status at the time of decompression significantly influences theincidence and time to onset of DCS in this model.
differences were not statistically significant, but the au- Keywords: diving, swine, saturation diving, diuretics, hydration.
thor concluded that a trend was apparent, suggestingincreased susceptibility in dehydrated animals. It isphysiologically plausible that dehydration could alter IT HAS BEEN KNOWN for more than a century that inert gas removal by reducing blood flow to poorly the risk of decompression sickness (DCS) is a func- perfused tissues, or that it may decrease surface tension tion of the volumes of gases dissolved in tissues and the and thereby facilitate bubble formation (5). However, rate and magnitude of decompression from a higher hydration in humans did not lead to improved inert gas pressure (1). For the diver, gas supersaturation occurs while breathing inert gas in hyperbaria and then reduc- If it can be demonstrated that fluid balance signifi- ing the pressure. The same phenomenon has also been cantly affects DCS risk, this may offer a relatively easy reported in people rapidly exposed to hypobaria, as in means of reducing susceptibility without resorting to high altitude flight or extravehicular activities in space prolonged decompression or subsequent recompres- (6). Safety has been significantly improved by empiri- sion therapy. Therefore, in this trial, furosemide (Lasix; cally tested tables that limit the pressure exposure and Aventis, Bridgewater, NJ) was used to induce experi- decompression rate. However, some decompression mental dehydration and assess its effect on DCS mani- events are associated with the development of DCS festations after direct ascent from saturation conditions.
symptoms, even though the tables have been followedwell within the limits (12).
From the Naval Medical Research Center, Silver Spring, MD.
It has been generally accepted that susceptibility to This manuscript was received for review in September 2005. It was DCS is caused by both physiological and psychological accepted for publication in November 2005.
factors. Despite this, it appears that a component of Address reprint requests to: Diana Temple, Naval Medical Re- DCS risk remains approachable only by probabilistic search Center, Diving Medicine, 503 Robert Grant Ave., Silver Spring,MD 20910-7500; TempleD@nmrc.navy.mil.
mathematical treatment as a random event (23). How- Reprint & Copyright by Aerospace Medical Association, Alexan- ever, physiological phenomena that appear to occur Aviation, Space, and Environmental Medicine • Vol. 77, No. 2 • February 2006 DEHYDRATION & DCS IN SWINE—FAHLMAN & DROMSKY rate was closely followed until a depth of about 33 fsw(2 ATA). Due to piping restrictions, the remainder of The experiments reported herein were conducted ac- the decompression required 1.5–2 min.
cording to the guidelines on laboratory animal use (18).
On reaching surface the animals were fitted with Before commencing, the Institutional Animal Care and individual monitors that measured heart rate and he- Use Committee reviewed and approved all aspects of moglobin saturation (VetOx 4404, Heska, Ft. Collins, this protocol. The institutional animal care facility is CO) and then transferred to individual 36“ ϫ 23“ ϫ 22“ clear Plexiglas observation pens without access to food Subjects: Neutered male Yorkshire swine littermates or water. Onset of severe DCS (neurological or cardio- from a closed breeding colony (n ϭ 57, mean Ϯ 1 SD pulmonary dysfunction) was recorded to the nearest 20.6 Ϯ 1.7 kg) were examined by a veterinarian on minute by dedicated observers. Disease and symptom arrival and housed in individual runs where water was onset times are referenced to the time the animals freely available. Their daily feedings consisted of 2% by reached surface. Neurological/central nervous system bodyweight of laboratory animal feed (Harlan Teklad, (CNS) DCS was defined as motor weakness (dimin- Madison, WI). Animals remained in the care facility for ished limb strength, repeated motor incoordination, or a minimum of 72 h before experiments.
inability to stand after being righted by the investiga- Predive preparation: Animals were transported to the tor), paralysis (complete limb dysfunction, areflexia, laboratory in plastic transport kennels (22“ ϫ 32“ ϫ 22“; hypotonia), or cranial nerve dysfunction (10). An ani- Vari-Kennel, R.C. Steele, Brockport, NY), placed in a mal was diagnosed with cardiopulmonary (CP) DCS if Panepinto sling (Charles River, Wilmington, MA) and it sustained the following parameters for 1 min or more: anesthetized by intramuscular injection of ketamine (20 mg ⅐ kgϪ1) and xylazine (1 mg ⅐ kgϪ1). Using sterile technique, 5 cm of a customized Tygon catheter (model H) Ͼ 150 bpm, and arterial O2 saturation #RPC-040, Braintree Scientific, Braintree, MA) was in- (SpO2) Ͻ 80%. This condition was usually accompanied serted into the left external jugular vein and external- by respiratory distress, as evidenced by open-mouthed, ized in the posterior midline at the T1 level. After labored breathing, central cyanosis, inversion of the closing the incision, the catheter was connected to an normal inspiratory/expiratory ratio, and production of injection port. Animals then received 500 mg chloram- frothy white sputum (10). All subjects with signs of phenicol to reduce infection risk and 2 ml heparinized severe DCS were removed to a Panepinto sling and saline (2 U ⅐ mlϪ1) to maintain catheter given 2.5 mg diazepam i.v. as necessary to alleviate their distress. Skin DCS and behavioral features (e.g., down using waterproof surgical tape. After a complete limb lifting) indicative of milder DCS were noted but not classified as positive cases for this study. After the 1-h observation period the subjects were weighed.
Procedure: The morning after catheterization, the an- Close observation continued until 4 h post-surface, at imals were brought to the laboratory, weighed, and which time they were euthanized by cardioplegia with placed in a Panepinto sling. The animals were then bolus i.v. injection of 40 ml of 4-Molar potassium chlo- randomized to either ad lib access to water during the ride solution. Previous experiments using this model hyperbaric exposure (hydrated group), or i.v. infusion have shown that all cases of DCS presented within this of Lasix (2 mg ⅐ kgϪ1) delivered over 10 min follow by 4-h period (8). All animals that developed severe DCS no access to water during the hyperbaric exposure (de- or expired from their disease were immediately sent for hydrated group). All animals were placed in a modified necropsy as previously detailed to clinically verify the transport kennel that allowed direct visualization diagnosis based on observed symptoms (10).
Analysis: A priori calculations using the Chi-square Subject pairs were placed in a manually controlled test indicated that sample sizes of 26 subjects per group hyperbaric chamber (656 ft3 internal volume, WSF In- would detect a 50% change in incidence with p ϭ 0.05 dustries, Buffalo, NY) and compressed to a 110 fsw (4.33 and 95% power. Independent variables included age, ATA) using air. Compression progressed in phases, pre-dive weight, weight change during the dive, and beginning with 5 fsw ⅐ minϪ1 (0.15 ATA ⅐ minϪ1) to a treatment group. Differences in independent variables depth of 33 fsw (2 ATA). If the animal showed no between groups were determined by t-tests or Mann- distress or other evidence of middle ear barotraumas, Whitney in the case of unequal variances. A survival the compression rate was increased to 10 fsw ⅐ minϪ1 analysis, using a log-rank test, was used to compare the (0.30 ATA ⅐ minϪ1). The compression rate was further time to symptom onset for each of the two groups. The increased to 20 fsw ⅐ minϪ1 (0.45 ATA ⅐ minϪ1) beyond influence of specific independent variables on outcome 99 fsw (4 ATA) if the animals tolerated the descent well.
was determined using logistic regression and likelihood Animal comfort was the limiting factor in all descent ratio testing in the manner described by Hosmer and rates. Temperature was maintained between 26.7– Lemeshow (14). The logistic regression analysis was 29.4°C, humidity between 50 –75%, and CO performed incorporating four independent variables; tion Ͻ 0.3%. The animals were constantly monitored via pre-weight, weight loss, age, and group (hydrated or closed-circuit television cameras through observation dehydrated). Initially, a univariate analysis on each in- ports. After 22 h, the animals were returned to surface dependent variable was performed; only those vari- (1 ATA) at a nominal rate of 30 fsw ⅐ minϪ1 (0.91 ATA ables with a p-value Ͼ 0.20 (Wald test) were included in ⅐ minϪ1) with no decompression stops. In practice that a multivariate analysis. Exclusion of a variable from the Aviation, Space, and Environmental Medicine • Vol. 77, No. 2 • February 2006 DEHYDRATION & DCS IN SWINE—FAHLMAN & DROMSKY TABLE I. SUMMARY OF OUTCOMES BY GROUP.
CNS ϭ central nervous system; CP ϭ cardiopulmonary DCS; Both ϭ animals manifested both CNS and CP DCS; Death ϭ animal expired fromtheir disease and mean time of severe DCS for all animals (All DCS), for those with CNS or CP DCS, and the mean time of death after surfacing.
The p-values represent differences in DCS outcome (Yates Chi-square) or mean values among groups (log-rank test for time of onset).
multivariate analysis was based on the log-likelihood risk of severe DCS and death as compared with the ratio test. Statistical significance was set at the p Ͻ 0.05 hydrated animals (p Ͻ 0.01). Specifically, dehydration level and p-values 0.05 Ͻ p Ͻ 0.1 were considered a increased the risk of CP DCS (p Ͻ 0.01), and showed a trend toward increased CNS DCS (p ϭ 0.069) in thismodel. In no case was pre-weight an important covari- ate, suggesting that the significant differences in pre- There were no significant differences in age between weight did not affect the DCS outcome.
groups (mean Ϯ SE; hydrated: 71 Ϯ 1 d; dehydrated:70 Ϯ 1 d). Statistically significant pre-dive weight dif- DISCUSSION
ferences existed between the two groups (p Ͻ 0.05, Some studies have tried to find a correlation between ANOVA, mean Ϯ SE; hydrated: 20.1 Ϯ 0.3 kg; dehy- DCS risk and variables such as body temperature, drated: 21.0 Ϯ 0.3 kg), but was only a significant pre- bodyweight, exercise, gender, adiposity, age, serum dictor of CNS DCS. Dehydrated animals lost signifi- cholesterol, sensitivity to complement activation, Dopp- cantly more weight during the dive than the hydrated ler bubble grades, patent foramen ovale, and hydration group (p Ͻ 0.01, ANOVA, mean Ϯ SE; hydrated: 0.8 Ϯ status, but most of these have had contradictory results 0.1 kg; dehydrated: 1.7 Ϯ 0.1 kg). The weight loss, (see the references in 13). The only physiological vari- able that has been undisputedly correlated with DCS risk is bodyweight in rats (17). Fluid status, on the other hand, is considered an important factor that can alter The animals had DCS manifestations, case IP : 137.82.96.26 DCS risk, but only a few controlled studies have been tions, and histopathology similar to Wed, 15 Mar 2006 07:48:21 conducted to verify this connection. A study in rats described (9). Overall, 28/57 animals sustained severe concluded that there appears to be a connection be- DCS, and 13/57 succumbed to the disease. Hydrated tween increased DCS risk and dehydration (19). Previ- animals had significantly less DCS (32.3%) compared ous work demonstrated that swine receiving i.v. injec- with dehydrated animals (73.0%, p Ͻ 0.01) and there tions of normal saline before a 22-h hyperbaric was a trend for a lower death rate (hydrated 12.9%, exposure to 110 fsw had a lower incidence of severe dehydrated 34.6%, p Ͻ 0.1, Yates Chi-square). The over- DCS as compared with a historical control group (11).
all DCS onset time and CP DCS onset time was signif- However, comparing historical data complicates evalu- icantly shorter in the dehydrated group than the hy- ation as minor differences in the experimental design drated group (p Ͻ 0.01, log-rank test) and there was a may significantly alter outcome. Therefore, the purpose trend toward a more rapid time to onset for CNS DCS of the current investigation was to compare the DCS and death. Table I summarizes results by group.
incidence in hydrated and dehydrated animals in a Neurological DCS was observed in 8/57 animals in this study and appeared Ͻ 1 h after surfacing, devel- Swine have well recognized anatomical and physio- oping rapidly over the course of a few minutes. It logic similarities to humans, and volumes have been manifested as progressive weakness of one or more written about their use as biomedical research models limbs, most commonly involving the hind limbs. Most (21). In recent years they have been successfully used to animals with evidence of CNS DCS began to show study a variety of diving-related conditions (2,3,20) and recovery within the 4-h observation period. CP DCS methods to improve decompression safety (8,16). The occurred in 28/57 animals, presenting as progressive pathological findings of livid skin DCS, multiple punc- tachypnea and tachycardia, with respiratory rates often tate spinal and cerebral hemorrhages, and profuse pul- exceeding 100 breaths ⅐ minϪ1 (Ͼ 300% of baseline) and monary congestion after no-stop decompression are sustained heart rates near 200 bpm (200% of baseline), consistent with previous observations in other animal combined with declining hemoglobin saturation often models of DCS, as well as human studies (2,4,22).
accompanied by production of frothy white sputum. If Outcome criteria in this study were necessarily se- the animal did not recover, it manifested central cyano- vere, in part because the more subtle manifestations of sis, increasing respiratory distress, eventually declined the disease often cannot reliably be detected in an ani- mal model. More invasive and, therefore, more sensi- The logistic regression analysis suggested that dehy- tive testing methods were precluded by the need to dration with Lasix significantly increased the overall observe the untreated natural history of the disease.
Aviation, Space, and Environmental Medicine • Vol. 77, No. 2 • February 2006 DEHYDRATION & DCS IN SWINE—FAHLMAN & DROMSKY TABLE II. LOGISTIC REGRESSION RESULTS FOR ANIMAL AGE, PRE-DIVE WEIGHT, WEIGHT LOSS DURING THE DIVE, AND TREATMENT (DF ϭ 1 FOR ALL) FOR ANIMALS WITH ACCESS TO WATER (HYDRATED) OR WITHOUT ACCESS TO WATER THROUGHOUT THE HYPERBARIC EXPOSURE AND GIVEN 2 MG ⅐ KGϪ1 LASIX (DEHYDRATED).
Parameter estimates (Ϯ SE), log likelihood (LL), and p-value for the log likelihood ratio test compared to intercept only (NULL) models.
This was prompted by the very real possibility that a dehydration may increase the risk of DCS and would impede the rescue effort. Thus, our results are sugges- tive of a simple and effective means to reduce DCS risk swered was not how many subjects could benefit from during a DISSUB rescue. In addition, the substantial recompression; theoretically they all would.
effect of dehydration in the current study warrants further consideration for military, commercial, and rec- chamber to prevent CNS morbidity or life-threatening reational divers as well as astronauts and aviators as an important factor that affects DCS risk.
Bodyweight is frequently used as an easily measured, In conclusion, after direct ascent from saturation con- highly sensitive indicator of overall fluid loss. This is ditions, dehydrated animals manifest severe CP DCS used both in a clinical situation (e.g., congestive heart sooner and more often than their hydrated counter- rate failure patients) and as a standard technique in parts. They also show a clear trend toward a higher rate sport events (e.g., triathlon races, wrestling and tennis of CNS DCS and more rapid death in this model.
tournaments, and long-distance cycling events). In thisstudy, without access to food or water, the weight lossis attributable to water losses from both normal dehy- dration, compression diuresis, and from the Lasix.
The authors are indebted to Chief Petty Officers Tony Ruopoli and Rob Hale; Petty Officers Harold Boyles, William Dow, and Thomas Regression analysis revealed that weight loss met or Robertson; Mrs. Catherine Jones; and Mr. Melvin Routh for their approached statistical significance. Addition of weight excellent technical assistance during the experiments. Thanks are also loss did not improve the log likelihood measurement due to the staff of the Laboratory Animal Medicine and Science when compared with the dehydrated group alone (Ta-
Department and Technical Services Department at NMRC. We are ble II, severe DCS and CP DCS). The reason for this is
also grateful to Ms. Diana Temple for her help in preparation of themanuscript.
that the two variables are highly correlated, as is also Naval Sea Systems Command work unit 63713N M000099.01B-1610 evident by the changing parameter estimates when supported this trial. The opinions and assertions contained herein are both are included. Thus, in the final analysis, dehydra- the private ones of the author and are not to be construed as official tion is the most significant predictor of severe and CP or reflecting the views of the Department of the Navy or the Navalservice at large. U.S. Government employees did this work as part of DCS and death in this study. In addition, hydrated their official duties; therefore, it may not be copyrighted and may be animals had a longer time to symptom onset of severe DCS and a trend toward a prolonged time to death.
This has important implications in the event of a dis-abled submarine (DISSUB) rescue operation, as hydrat- ing survivors prior to decompression is a simple means 1. Boycott AE, Damant GCC, Haldane JS. The prevention of decom- of reducing DCS incidence and severity. For example, pression-air illness. J Hyg 1908; 8:342– 443.
2. Broome JR, Dick EJ Jr. Neurological decompression illness in severe dehydration and hypothermia are common in swine. Aviat Space Environ Med 1996; 67:207–13.
DISSUB victims (15). It has long been suggested that 3. Buttolph TB, Dick EJ Jr, Toner CB, et al. Cutaneous lesions in Aviation, Space, and Environmental Medicine • Vol. 77, No. 2 • February 2006 DEHYDRATION & DCS IN SWINE—FAHLMAN & DROMSKY swine after decompression: histopathology and ultrastructure.
13. Fahlman A, Tikuisis P, Himm JF, et al. On the likelihood of Undersea Hyperb Med 1998; 25:115–21.
decompression sickness during H2 biochemical decompression 4. Catron PW, Thomas LB, McDermott JJ, et al. Failure of heparin, in pigs. J Appl Physiol 2001; 91:2720 –9.
superoxide dismutase, and catalase to protect against decom- 14. Hosmer DW, Lemeshow S. Applied logistic regression. New pression sickness. Undersea Biomed Res 1987; 14:319 –30.
5. Conkin J. A literature survey: fluid balance in animals and man 15. House CM, House JR, Oakley EHN. Findings from a simulated and its influence on decompression sickness. Houston, TX: disabled submarine survival trial. Undersea Hyper Med 2000; Technology Incorporated; 1983:1–27.
6. Conkin J, Kumar KV, Powell MR, et al. A probabilistic model of 16. Kayar SR, Fahlman A, Lin WC, Whitman WB. Increasing ac- hypobaric decompression sickness based on 66 chamber tests.
tivity of H2-metabolizing microbes lowers decompression Aviat Space Environ Med 1996; 67:176 – 83.
sickness risk in pigs during H2 dives. J Appl Physiol 2001; 7. Conkin J, Waligora JM, Horrigan DJJ. Effect of hydration on nitrogen washout in human subjects. Houston, TX: NASA, 17. Lillo RS, Parker EC, Porter WR. Decompression comparison of helium and hydrogen in rats. J Appl Physiol 1997; 82:892–901.
8. Dromsky DM, Spiess BD, Fahlman A. Treatment of decompres- 18. NRC. Guide for the care and use of laboratory animals. Washing- sion sickness in swine with intravenous perfluorocarbon emul- ton, DC: National Academy Press; 1996.
sion. Aviat Space Environ Med 2004; 75:301–5.
19. Philp RB. Decompression sickness in experimental animals. In: 9. Dromsky DM, Toner CB, Fahlman A, Weathersby PK. Prophylac- Lambertsen CJ, ed. Underwater physiology. Proceedings of the tic treatment of severe decompression sickness with methyl- third symposium on underwater physiology; 23-25 March prednisolone. Undersea Biomed Res 1999; 26:15.
1966; Washington, DC. New York, NY: Academic Press; 1967: 10. Dromsky D, Toner CB, Survanshi S, et al. The natural history of severe decompression sickness after rapid ascent from air sat- 20. Sharpe RP, Broome JR. Spinal cord lipid levels in a porcine model uration in a porcine model. J Appl Physiol 2000; 89:791– 8.
of spinal cord decompression sickness. Undersea Hyperb Med 11. Dromsky DM, Weathersby PK, Fahlman A. Prophylactic high dose methylprednisolone fails to treat severe decompression 21. Swindle M. Swine as models in biomedical research. Ames, IA: sickness in swine. Aviat Space Environ Med 2003; 74:21– 8.
12. Elliott DH, Kindwall EP. Manifestations of the decompression 22. Van Rensselaer H. The pathology of the caisson disease. Med Rec disorders. In: Bennett PB, Elliott PB, eds. The physiology and medicine of diving. San Pedro, CA: Best Publishing Co.; 1982: 23. Weathersby PK, Homer LD, Flynn ET. On the likelihood of de- compression sickness. J Appl Physiol 1984; 57:815–25.
Aviation, Space, and Environmental Medicine • Vol. 77, No. 2 • February 2006

Source: http://www.zoology.ubc.ca/~fahlman/pub-link/Diving/lasix/Fahlman-ASEM-77-102-106-2006.pdf