Is Smoking a Predictor for Acute Mountain Sickness? Findings From a Meta-Analysis

Studies of the potential association between cigarette smoking and acute mountain sickness (AMS) have reached contradictory conclusions. Our aim was to perform a meta-analysis of studies across a range of populations to ascertain better the true relationship between cigarette smoking and AMS.

We used the PRISMA protocol to identify and screen eligible studies of smoking and AMS. Databases including Pubmed and Google Scholar were searched, using the terms "smoking" and "acute mountain sickness." We conducted a meta-analysis of the selected studies in order to evaluate causal inference, evaluate potential biases, and investigate possible sources of heterogeneity across studies.

We identified 3907 publications, of which 29 were eligible for inclusion by reporting smoking status and AMS. Of these, eight publications were excluded because they were duplicative or were lacking quantitative data. The 21 studies analyzed included 16 566 subjects. These fell into two groups: occupational/military (n = 8) or volunteers/trekkers/mixed (n = 13). Study heterogeneity was high (X (2) = 55.5, P < .001). Smoking was not statistically associated with increased risk of AMS: pooled OR = 0.88 (95% CI = 0.74-1.05). Stratification yielded similar risk estimates among the occupational/military studies versus all others and studies at relatively higher and lower altitudes.

Overall, smoking was not statistically significantly associated with AMS: there is no consistent effect of cigarette smoking acting as either a protective factor against or a risk factor for AMS.

This is the first quantitative assessment of published studies on smoking and AMS, which shows smoking to be neither a risk, nor protective. Studies specifically focusing on smoking as a risk factor, should guide further research on this issue. Although all smokers should be strongly advised to quit, studies on risk factors for AMS focusing on other exposures could shed light on the full range of risks for AMS.