Publication Date:
2013-11-15
Description:
Introduction A number of risk factors for the development of acute myeloid leukemia (AML) have been previously described. The role of smoking in the development of AML has been postulated as a potential environmental risk factor. This association has been studied in previous observational reports; however, the dose-response relation between smoking and AML has not been evaluated. The primary objective of this meta-analysis is to evaluate the dose-relationship between smoking and the development of AML. Secondary objectives were to identify potential gender and/or geographical disparities. Methods A PubMed search from January 1, 1993 to December 31, 2012 was undertaken using the keywords: “(smoking OR tobacco OR cigarette) AND leukemia”. Prospective cohort and case-control studies reporting on the incidence of AML were included. Studies reporting on acute promyelocytic leukemia, acute leukemia or myeloid leukemia without specifying subtype and cross-sectional studies were excluded. Studies included in a previous meta-analysis by Brownson et al. (1993) were also excluded. The outcome of interest was the odds ratio (OR) with 95% confidence interval (CI) of developing AML in smokers compared to never smokers. Because the overall risk of AML in the general population is low, the relative risk mathematically approximates the OR, allowing the pooling of cohort and case-control studies (rare disease assumption). The random effects model (REM), which accounts for intra and inter-study heterogeneity, was used to estimate the combined outcome. Heterogeneity was also quantified by the I2statistic. Publication bias was assessed by the trim-and-fill analysis. Stratified analyses were performed in current smokers and ever smokers. Subset analyses were performed by sex, study design, geographical region, number of cigarettes smoked, number of years of smoking and cumulative smoking in pack-years. The quality of the studies was assessed by the Newcastle-Ottawa scale (NOS). Literature search, data gathering and quality assessment were performed independently by at least two of the investigators. All calculations and graphs were obtained using Comprehensive Meta-Analysis version 2.2.050 (Biostat, Englewood, NJ, USA). Results Our initial search yielded 573 articles, from which 16 case-control and 6 cohort studies were included in our final analysis; 10 studies were from Europe, 9 from North America and 3 from Asia. All the studies were considered of intermediate and high quality, based on the NOS score. Ever smokers had an OR of 1.26 (95% CI 1.15-1.38; p
Print ISSN:
0006-4971
Electronic ISSN:
1528-0020
Topics:
Biology
,
Medicine
Permalink