Thyroid cancer risk and smoking status: a meta-analysis
A partir des données de 31 études (6 260 cas et 32 935 témoins issus des études cas-témoins ; 2 715 participants issus des études de cohorte), cette méta-analyse évalue l'association entre le tabagisme et le risque de cancer de la thyroïde
Purpose : Previous researchers have reported an inverse association between cigarette smoking and thyroid cancer risk. To summarize the role of smoking in relation to thyroid cancer occurrence, we conducted a meta-analysis. Methods : We performed a meta-analysis of 31 eligible studies to summarize the data describing the association between thyroid cancer occurrence and smoking. The case–control studies consisted of 6,260 thyroid cancer cases and 32,935 controls. Cohort studies contained 2,715 thyroid cancer patients that participated from recruitment to follow-up. Q-statistic and I 2 statistic were calculated to examine heterogeneity. Summary relative risks (RRs) and 95 % confidence intervals (95 % CIs) were calculated using a random effects model. Potential sources of heterogeneity were investigated via subgroup and sensitivity analyses, and publication biases were estimated. Results : Thyroid cancer risk was reduced in persons who had ever-smoked (RR = 0.79; 95 % CI 0.70–0.88) compared with never-smokers. However, strong evidence of heterogeneity was found among the investigated studies; therefore, subgroup analyses were conducted according to study type, smoking status, study location, source of controls, sex, and histological type of thyroid cancer. When the data were stratified by smoking status, an inverse association was observed only among current smokers (RR = 0.74; 95 % CI 0.64–0.86), not former smokers (RR = 1.01; 95 % CI 0.92–1.10). An inverse association was observed only in case–control studies (RR = 0.75; 95 % CI 0.66–0.85). Conclusions : This meta-analysis of geographically diverse epidemiological data suggests that smoking, particularly current smoking, may influence susceptibility to thyroid cancer. Further well-designed studies with larger sample sizes should be conducted