Marijuana Use and Hemoglobin Concentrations in NHANES 2009-2018: Implications for Subclinical Hypoxemia.

RATIONALE:Cannabis use is rapidly growing in the United States, but its health implications are poorly understood, particularly when compared with cigarette smoking. Previous research conducted on animal models or non-representative populations with small sample sizes has yielded mixed results on the impact of marijuana use on hemoglobin levels, which may reflect subclinical hypoxemia and/or carbon monoxide exposure. OBJECTIVES:We evaluated the association between marijuana use and hemoglobin levels in a nationally representative sample of U.S. adults. METHODS:This cross-sectional study included 16,038 individuals aged 18 to 59 years enrolled in the National Health and Nutrition Examination Survey (NHANES) from 2009 to 2018. We related current and former marijuana use with measured hemoglobin levels, with adjustment for demographics, education, housing, and cigarette smoking status in multivariable analyses that incorporated complex survey weights. As candidate positive and negative control exposures, we used similar methods to relate cigarette smoking and benzodiazepine use, respectively, with hemoglobin concentrations. RESULTS:Current marijuana use was associated with significantly higher hemoglobin concentrations. After multivariable adjustment, compared with never use, current marijuana use was associated with a 0.111, 95% CI [0.021,0.201] g/dL higher hemoglobin concentration, whereas former use was associated with a 0.047, 95% CI [-0.018,0.113] g/dL higher concentration (linear trend p=0.01). As hypothesized, cigarette smoking was also associated with higher hemoglobin concentrations, while benzodiazepine use was not. CONCLUSIONS:Among American adults, current marijuana use was associated with higher hemoglobin concentrations, as is cigarette smoking but not benzodiazepine use. These results suggest the possibility that marijuana smoking induces subclinical hypoxemia stimulating hemoglobin production. Further confirmation of this observational finding is needed, in light of the increasing medical and recreational use of smoked marijuana products.