If the inflammatory response to inhalation of cigarette smoke causes chroni
c obstructive pulmonary disease (COPD), suppression of that natural respons
e might be beneficial. We hypothesized that a smoker's risk of developing C
OPD is inversely related to physiologic levels of two fatty acids that have
antiinflammatory properties: eicosapentaenoic acid (EPA, C20:5) and docosa
hexaenoic acid (DHA, C22:6). The proportion of each fatty acid in plasma li
pids was measured in 2,349 current or former smokers. COPD was identified a
nd defined by clinical symptoms and/or spirometry. After adjustment for smo
king exposure and other possible confounders, the prevalence odds of COPD w
ere inversely related to the DHA (but not to the EPA) content of plasma lip
id components in most of the models. For example, as compared with the firs
t quartile of the DHA distribution, the prevalence odds ratios (ORs) for ch
ronic bronchitis were 0.98, 0.88, and 0.69 for the second, third, and fourt
h quartiles, respectively (p for linear trend = 0.09). The corresponding OR
s for COPD as defined spirometrically, were 0.65, 0.51, and 0.48 (p < 0.001
). Among 543 current heavy smokers, adjusted mean values of FEV1 (lowest to
highest DHA quartile) were 2,706, 2,785, 2,801, and 2,854 ml. DHA may have
a role in preventing or treating COPD and other chronic inflammatory condi
tions of the lung. Pilot testing of that hypothesis in experimental models
seems warranted.