Objective/hypothesis: The significance of gastroesophageal reflux disease (GERD) in laryngeal cancer is controversial due to disparate studies. To evaluate the overall strength of the association of GERD with laryngeal cancer, we performed meta-analysis of the original studies in literature.
Study design: Meta-analysis.
Methods: All studies cited on Ovid Medline (1966-June 2004), EMBASE (1980-June 2004), and Cochrane database describing GERD and laryngeal cancer were eligible for inclusion. The inclusion criteria for the study included original controlled study design and a clear documentation of the reflux prevalence in cases and controls. Statistical analysis was performed by NCSS software.
Results: Fifteen original studies were identified. Eight studies did not have control groups, two studies did not clearly document GERD prevalence in controls, and two studies were published using the same data, one of which was included in this meta-analysis. Thus, four studies qualified for inclusion for the meta-analysis. The methodologic quality in the studies was heterogeneous, not only in the evaluation of confounding risk factors such as smoking and alcohol but also in the mode of GERD diagnosis. There was also significant heterogeneity of effect of reflux among the studies (P = .001). The pooled odds ratio on the basis of fixed-effects model was 2.86 (95% CI, 2.73-2.99), and on the basis of random-effects model was 2.37 (95% CI 1.38-4.08).
Conclusion: Our meta-analysis suggests that GERD may be a significant risk factor for laryngeal cancer. However, given the heterogeneity of the published data, future prospective controlled studies are needed.