Impact of pre-diagnosis behavior on risk of death from esophageal cancer : a systematic review and meta-analysis

Purpose: Most people diagnosed with esophageal cancer will die from their disease, but it is not known whether survival is influenced by pre-morbid behavior. We undertook a systematic review and meta-analysis to investigate the impact of pre-diagnosis behavior on risk of death for esophageal cancer. Methods: We performed a systematic review of studies reporting on the relationship between pre-diagnosis smoking, alcohol consumption, overweight and obesity, physical activity and regular consumption of nonsteroidal anti-inflammatory drugs, and risk of death from esophageal squamous cell carcinoma (ESCC) and adenocarcinomas (EACs). Study characteristics are presented and aggregate results are compiled using meta-analysis. Results: From an initial pool of 644 non-duplicate records, 13 articles arising from 12 studies met the inclusion criteria. Considerable variation was observed between studies in location, measurement categories, adjustment for other risks, and results. Pooled estimates suggested that for ESCC pre-diagnosis smoking was associated with a 1.19 times [95 % confidence interval (CI) 1.04-1.36] increased risk of death and pre-diagnosis alcohol consumption with a 1.36 times increased risk of death (95 % CI 1.15-1.61). No significant effects were observed for EAC. We observed a lower risk of death for both ESCC and EAC associated with high pre-diagnosis body mass index (BMI) ≥25 kg/m2 (ESCC hazard ratio 0.80, 95 % CI 0.67-0.95; EAC 0.80, 95 % CI 0.68-0.95), although there was significant heterogeneity across studies. Conclusions: Our findings suggest that a number of modifiable pre-diagnosis risk factors have a carryover effect on the risk of death from esophageal cancer. These include smoking, drinking alcohol, and BMI.