Premium
Eating disorders and the risk of esophageal and stomach cancers‐A nationwide, population‐based cohort study in Taiwan
Author(s) -
Yeh HuiWen,
Chien WuChien,
Chung ChiHsiang,
Chang HsinAn,
Kao YuChen,
Tzeng NianSheng
Publication year - 2021
Publication title -
international journal of eating disorders
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.785
H-Index - 138
eISSN - 1098-108X
pISSN - 0276-3478
DOI - 10.1002/eat.23367
Subject(s) - medicine , hazard ratio , proportional hazards model , cohort , cancer , confidence interval , cohort study , population , stomach cancer , demography , oncology , environmental health , sociology
Abstract Background Previous studies have shown elevated cancer risk in anorexia nervosa but the literature on other eating disorders (EDs) is scarce. This study aimed to investigate the association between all EDs and esophageal, stomach, and other cancers. Methods We used a retrospective cohort design, based on a two‐million randomized longitudinal health insurance dataset, a sub‐dataset of Taiwan's National Health Insurance Research Database. From all the potential participants aged 20 years or more, a total of 6,628 participants were enrolled, including 1,657 patients with EDs, with sex‐, age‐, and indexed date‐matched (1:3) 4,971 controls. Each participant was individually tracked from 2000 to 2015 to identify incident cases of cancers, including esophageal cancer (EC), stomach cancer (SC), and all other cancers (AOC). The multivariate Cox proportional hazards model was employed to estimate the hazard ratios (HRs) and 95% confidence intervals (CIs) for the association between EDs and cancer. Results Of the total 6,628 enrollees, 222 in 1,657 individuals with EDs and 810 in the 4,971 non‐ED control individuals developed cancer (1,262.40 vs. 1,472.15 per 100,000 person‐years), and the Kaplan–Meier survival analysis was not statistically significant (log‐rank, p = .324). However, after adjusting for covariates, the risk of EC and SC among the individuals with an ED was significantly higher, with adjusted HRs of 5.32 (95% CI: 1.07–26.49, p < .001) and 4.61 (95% CI: 1.91–11.14, p < .001), respectively. EDs were not associated with other cancers. Conclusions This study provides evidence for the association between EDs and the risk for EC and SC. Further research on mechanisms and prevention is therefore needed.