Cigarette smoking and decompression illness severity: a retrospective study in recreational divers

Aviat Space Environ Med. 2003 Dec;74(12):1271-4.


Background: Severe decompression illness (DCI) could be more likely in cigarette smokers because of airway obstruction or vascular disease. The present study evaluated the severity of DCI as a function of cigarette smoking in recreational divers.

Methods: We examined all DCI reports recorded in the Divers Alert Network (DAN) database from 1989 through 1997. Smoking history was quantified as heavy (>15 pack-years), light (0 to 15 pack-years), and never smoked. DCI symptoms were classified as severe (alteration in consciousness, balance or bladder/bowel control, motor weakness, visual symptoms, convulsions), moderate (other neurological symptoms), or mild (pain, skin, or nonspecific symptoms). The proportional odds model and generalized logits were used for the adjusted analysis when accounting for other covariates.

Results: There were 4,350 patients included in the analysis. After adjustment for confounding variables, heavy smokers were more likely to have severe vs. mild symptoms than nonsmokers (OR = 1.88) (95% CI 1.36, 2.60) or light smokers (OR = 1.56) (95% CI 1.09, 2.23). Heavy smokers and light smokers were more likely to have severe vs. moderate symptoms than nonsmokers (OR = 1.36) (95% CI 1.06, 1.74) and (1.22) (1.02, 1.46), respectively. Although these data do not reveal whether smoking predisposes to DCI, the results are consistent with a tendency, when DCI occurs, for cigarette smoking to trigger more severe symptoms.

Conclusions: The data suggest that when DCI occurs in recreational divers, smoking is a risk factor for increased severity of symptoms.

MeSH terms

  • Decompression Sickness / classification
  • Decompression Sickness / etiology*
  • Decompression Sickness / pathology*
  • Diving / adverse effects*
  • Humans
  • Odds Ratio
  • Recreation*
  • Retrospective Studies
  • Risk Factors
  • Severity of Illness Index
  • Smoking / adverse effects*