Alcohol use as a risk factor for DCS

Brief summary: 1. there are no data showing an association between alcohol use and DCS risk; 2. everybody thinks there are; 3. there are theoretical reasons to suspect a link, but these remain unstudied and unproven. Read on for more detail.


It is often suggested that the consumption of alcohol before or after diving predisposes the diver to the development of decompression sickness (DCS). As Brylske (1994) states, "The bottom line is that while the whole issue of alcohol's contribution to DCS is speculative, this speculation is based on sound physiological principles" (pp. 24, 26). There are a number of postulated mechanisms whereby alcohol use could increase DCS risk.

1. Alcohol lowers surface tension, favoring nucleation and bubble stability and growth (Brylske, 1994; Eckenhoff, 1989). However, the effect of alcohol in vivo is unknown (Eckenhoff, 1989).

2. The vasodilation (dilation of blood vessels) caused by alcohol may lead to increased blood flow and thereby to increased nitrogen absorption (Brylske, 1994; Eckenhoff, 1989; PADI, 1984). However, Zhang et al. (1989) note that vasodilation could also improve nitrogen washout from tissues.

3. Alcohol leads to diuresis and dehydration (Brylske, 1994; Dovenbarger, 1988; Smith, 1993), which may decrease the fluid volume of blood, altering the absorption and elimination of nitrogen (Brylske, 1994; see also Aharon-Peretz et al., 1993). However, the dehydrating effects of alcohol are minimal for moderate drinking, and the body returns to homeostasis.

Current policy

Alcohol use is described as a risk factor for DCS in the dive training manuals of all the major diving certification agencies as well as numerous books and articles on diving and DCS. DAN cautions against drinking before or after diving, citing alcohol's effect on judgment (Smith, 1993) and the dehydrating effects of alcohol, which may inhibit nitrogen off-gassing (Dovenbarger, 1988). The training manual of the Professional Association of Dive Instructors (PADI) instructs divers to avoid alcohol before, between, and two hours after diving, stating emphatically that alcohol "causes dilation (enlarging) of the blood capillaries, increasing circulation. Drinking alcohol prior to a dive will cause increased nitrogen absorption during a dive. Drinking alcohol after a dive will release nitrogen much too quickly, possibly contributing to bubble formation" (PADI, 1984; p. 185). Books on deep diving similarly caution against drinking, often giving no particular reason (Gilliam & Von Maier, 1992; Lippmann, 1992). (Gilliam has changed his thinking on this point somewhat. In a later article [Gilliam, 1994], he criticized some of the conventional wisdom of diving, including the rule "you can't dive if you have a beer with lunch." Gilliam noted that alcohol has the same dehydrating effect as caffeinated beverages, and that the effect of moderate consumption of these beverages is of little consequence. He concluded that "a few sacred cows continue to moo long after their milk ran dry" [p. 78].)


In contrast to the certainty with which many have associated drinking with DCS, there are no data demonstrating such an increase in risk. In essence, statements about alcohol-related risk are based on conventional wisdom rather than research. As Brylske (1994) writes, "Although no relationship between alcohol and DCS has ever been proven in a controlled scientific study, common sense tells us that there might be a substantially increased risk to anyone who consumes alcohol before diving" (p. 24).

Statistics available from diving accidents reported to DAN in 1991 (Dovenbarger, 1994) show that 38% of dive accident victims drank the night before the dive that resulted in DCS, 15% drank after the dive, and 1-2% drank pre-dive or between dives. There were also nine diving-related fatalities involving drugs and/or alcohol in 1991. Dovenbarger (1994) states that these accident statistics "seem to lead to a fairly straightforward correlation" between alcohol consumption and DCS (pp. 23, 26). This statement is incorrect, because DAN's statistics are insufficient to establish any such correlation. (A "correlation" between any two variables can only be established by showing that as one variable changes, the other variable changes also. You cannot establish a correlation when one of the variables--in this case, diving accidents--does not vary.) A conclusion about alcohol-related risk and DCS cannot be reached without information on the existence of alcohol in non-DCS dives, and this information is not available in DAN's statistics. That is, to show that alcohol is related to the incidence of DCS, one would have to show that the percentage of DCS cases in which alcohol was involved was greater than the percentage of uneventful dives that included alcohol. (If, for example, 38% of dive accident victims drank the night before the dive, but 38% of divers with no accidents drank the night before their dives, clearly there would be no relationship of alcohol to DCS).

To date, no data on alcohol use in uneventful dives exist; indeed, the incidence of DCS in recreational scuba divers in general is difficult to estimate because neither the number of divers nor the number of dives is known. A few specialized databases exist: 1) records kept by the U.S. Navy of dives by Navy divers (Blood & Hoiberg, 1985); 2) similar records kept by the British Royal Navy (Murrison et al., 1991); and 3) records of over 77,000 sport dives from an extremely large (and short-lived) dive operation aboard a cruise ship (Gilliam, 1991). None of these databases include any information on alcohol consumption.

Although no data are available on the presence of alcohol in non-problematic dives, some suggestive findings can be derived from a reanalysis that I did of DAN statistics on different types of diving accidents. DAN's report on 1987 diving accidents ("Risk factors in diving accidents," 1991) includes a listing of the frequency with which various risk factors occurred in 270 accidents (out of 570 reported) that included information on these risk factors. Of the accidents, 218 were DCS and 52 were arterial gas embolism (AGE). DCS and AGE are quite different in their etiology and pathology. While DCS is caused by rapid off-gassing of nitrogen leading to nitrogen gas bubbles in blood and tissue, AGE is caused by air bubbles in arterial blood. As a diver ascends, the volume of air in the lungs increases as the air expands with the decreasing ambient pressure. Lung overinflation (due to breath-holding or lung congestion that traps air in the alveoli) can result in tears in the lung tissue, releasing air into the pulmonary veins and from there into arterial blood (Gorman 1989; Nichols 1990; Lippmann 1992).

Because of the differences in DCS and AGE, a comparison of the presence of alcohol in DCS and AGE accidents is one way to begin to examine the role of alcohol as a risk factor for DCS. It has never been suggested (to my knowledge) that alcohol is a risk factor for AGE, and the differences in the causes of DCS and AGE (nitrogen saturation vs. lung overinflation upon ascent) would suggest that the postulated mechanisms for alcohol-related DCS risk (i.e., dehydration, diuresis, vasodilation) would not be relevant in AGE. An analysis of DAN's 1987 statistics showed that alcohol was involved in similar proportions of AGE (15%) and DCS cases (18%; OR = 1.24, 95% CI .54-2.83). (It is unclear from the description of risk factors whether alcohol had been drunk directly before the dive, the night before the dive, post-dive, or a combination of these.) This null finding is probably not attributable to the uniqueness or lack of representativeness of this sample, because other known risk factors for AGE were significantly different in AGE and DCS groups. For example, rapid ascent was more likely in AGE than in DCS cases (46% vs. 23.3%; OR = 2.8, 95% CI 1.5, 5.3), as were low/no air (which would lead to a rapid ascent; 21% vs. 9.2%; OR = 2.7, 95% CI 1.2, 6.0) and buoyancy problems (27% vs. 12.8%). Thus, DAN's statistics suggest that alcohol is no more of a risk factor for DCS, a condition supposedly linked to alcohol use, than for AGE, a condition having nothing to do with drinking. (see here for a more recent analysis) What is still lacking is data on alcohol use before uneventful dives.

Interestingly, there have been two published studies suggesting that alcohol may be an effective treatment for decompression sickness. (Which goes to show there's nothing new under the sun: in 1841, in the first decription of "caisson disease," Triger described treating the disorder with alcohol applied externally and internally [Bookspan, 1995]). Zhang et al. (1989) reported a study in which 32 rabbits were compressed to 6 atm in a chamber, maintained at that pressure for 30 minutes, and then rapidly decompressed. Half of the rabbits were injected with an ethanol-saline solution immediately after decompression, and half the rabbits were injected with saline only. All of the rabbits that were administered ethanol survived the decompression; half of the rabbits in the control group died within an hour. Ethanol-treated rabbits also had fewer venous bubbles than control rabbits. In a case series of six marine fishery and salvage divers with DCS symptoms (Zhang et al., 1991), these symptoms improved following an alcohol dose equivalent to 50-75 ml pure ethanol. Four of the six divers were fully recovered in 60 minutes after drinking wine, with the remaining two recovered after two hours. The divers were also monitored for bubbles in the bloodstream with Doppler techniques, and bubbling decreased following alcohol administration.

These findings caused some consternation in the diving medicine community, with the editor of Undersea Biomedical Research calling the findings "without precedent in the literature" and cautioning against using ethanol in DCS treatment until the results were replicated (Eckenhoff, 1989). In a subsequent study of 34 male sport divers (Eckenhoff & Olstad, 1991), alcohol was administered to half the divers following an extended shallow dive (48 hours at 20 ft). Doppler monitoring for 24 hours after surfacing showed that the level of bubbles in the alcohol group did not differ from the level in the control group.

Zhang and colleagues (1991) also report that in a study of 440 divers, incidence of DCS in those who frequently drank wine was lower than in those who did not. Recent research on mortality from heart disease has suggested that drinking red wine may protect against heart disease by way of vasodilation. It may be, then, that such vasodilation may improve the movement of nitrogen bubbles through the circulatory system, leading to faster elimination of small bubbles.

In an article in Alert Diver, a magazine distributed to DAN members, Dr. Ernest Campbell discussed the effects of alcohol on diving. (This information is also available on Dr. Campbell's web site (link)). The article cites a study that purportedly showed that drinking affected scuba diving performance. However, that study had nothing to do with scuba diving: in fact, it was about performing surface dives off the edge of a swimming pool (link).


Although conventional wisdom holds that alcohol use, either before or after diving, contributes to the development of DCS, there are no data supporting this contention. The two laboratory studies that have been published have equivocal results, and, due to the lack of information on non-accident dives, there are currently no data available to enable a statistical analysis of alcohol-related risk for DCS in sport divers. There are several theoretical reasons to expect that alcohol might be related to DCS risk, but these remain speculative and unsupported by data.


Aharon-Peretz, J., Adir, Y., Gordon, C.R., Kol, S., Gal, N., & Melamed, Y. (1993). Spinal cord decompression sickness in sport diving. Archives of Neurology, 50(7), 753-756.

Bennett, P.B., Moon, R.E., Dovenbarger, J.A., Wachholz, C.J. (1989, Spring). DAN position statement: Injury risk in sport diving. Alert Diver, 5(1), p. 8

Bennett, P.B., Dovenbarger, J.A., Bond, B.G., Wachholz, C.J. (1989, Spring). Flying after diving: 1987 accidents. Alert Diver, 5(1), p. 1.

Blood, C., & Hoiberg, A. (1985). Analyses of variables underlying U.S. Navy diving accidents. Undersea Biomedical Research, 12(3), 351-360.

Bookspan, Jolie (1995). Diving physiology in plain English. Undersea and Hyperbaric Medical Society, Kensington, MD.

Brylske, A. (1994, January). Here's to your health: The responsible diver's guide to drinking and diving. Dive Training, pp. 22-26.

Dovenbarger, J. (1988, Fall). Managing risk factors of decompression sickness, embolism, and diving injuries. Alert Diver, 4(3), pp. 3-5

Dovenbarger, J. (1994, January). DAN's perspective on alcohol and diving. Dive Training, pp. 23, 26.

Eckenhoff, R.G. (1989). Alcohol and bends (editorial). Undersea Biomedical Research, 16(4), 269.

Eckenhoff, R.G., & Olstad, C.S. (1991). Ethanol and venous bubbles after decompression in humans. Undersea Biomedical Research, 18(1), 47-51.

Gilliam, B. (1991, June). 77,680 computer and table dives: Who got bent and why. Undercurrent, 16(6), pp. 8-12.

Gilliam, B. (1994, March/April). Why? Because it's the rules! Scuba Times, 15(2), pp. 76-78.

Gilliam, B., & von Maier, R.(1992). Deep diving: An advanced guide to physiology procedures and systems. Watersport Publishing, San Diego.

Gorman, D.F. (1989). Decompression sickness and arterial gas embolism in sports scuba divers. Sports Medicine, 8(1), 32-42.

PADI (1984). PADI Advanced Diver Manual. Santa Ana, CA: PADI.

Lippmann, John (1992). The essentials of deeper sport diving. Locust Valley, NY: Aqua Quest Publishing.

Murrison, A.W., Lacey, E.J., Restler, M., Martinique, J., & Francis, T.J. (1991). Ten years of diving-related illness in the Royal Navy. Journal of Social and Occupational Medicine, 41(2), 89-93.

Nichols, G. (1990). A survey of diving illnesses. Alert Diver, 1990 special edition; pp. 1, 28.

Risk factors in diving accidents (1989, Spring). Alert Diver, 5(1), p. 10.

Smith, K. (1993, Jan/Feb). An overview of drugs and diving. Alert Diver, pp. 8-9.

Zhang, L-D., Kang, J-F., & Xue, H-L. (1989). Ethanol treatment for acute decompression sickness in rabbits. Undersea Biomedical Research, 16(4), 271-274.

Zhang, L-D., Kang, J-F., & Xue, H-L. (1991). Ethanol treatment of 6 acute decompression sickness cases (letter). Undersea Biomedical Research, 18(1), 64-67.