Nathalie Michels1, Fien De Backer2, Myrto Dimakopoulou3, Katerina Mane3, Iciar Indave3, Inge Huybrechts2,3. 1. Department of Public Health and Primary Care, Ghent University, 4K3, Corneel Heymanslaan 10, 9000, Gent, Belgium. nathalie.michels@ugent.be. 2. Department of Public Health and Primary Care, Ghent University, 4K3, Corneel Heymanslaan 10, 9000, Gent, Belgium. 3. International Agency for Research on Cancer, Lyon, France.
Abstract
PURPOSE: Evidence concerning eating disorders as risk toward developing cancer is sparse. Energy restriction might be cancer protective, while malnutrition, vomiting, laxative and substance use might stimulate cancer development. We examined whether individuals with an eating disorder (not restricted to anorexia nervosa) had a different risk of developing cancer. METHODS: A systematic search on Medline and Embase until 28th April 2020 identified relevant human original research publications, including all populations and all cancer types. RESULTS: From 990 records, 6 case reports and 9 cohorts were included. Some cohorts found a decreased breast (3/5 studies) or cervical (1/2) cancer risk, while an increased esophageal (2/3), liver (1/1), brain (1/1 in men) and respiratory (2/4) cancer risk, but other cancer risks were non-significant, and an increased mortality overall (1/2), from breast (1/1), female genital (1/1) and skin (1/1) cancer in eating disorder patients. The case reports further described esophageal cancer and leukemia. No clear statistical differences in cancer risk were found depending on eating disorder type, perhaps due to the small sample size (n = 1783 for other than anorexia nervosa). CONCLUSIONS: The literature on eating disorders and cancer risk is sparse with many gaps. Hormonal changes, sexual activity, nutritional status, vomiting and concomitant tobacco/alcohol abuse may explain increased/decreased cancer risk. Future large studies (now 1-366 cancer cases) that also include men (now 4.7%), bulimia nervosa (now 3.8%) and several cancer sites (now mainly breast cancer) are needed and should foresee longer follow-up time (now 5.4-15.2 years) and extensive confounder adjustment (now only age and sex). LEVEL OF EVIDENCE: Level I, systematic review.
PURPOSE: Evidence concerning eating disorders as risk toward developing cancer is sparse. Energy restriction might be cancer protective, while malnutrition, vomiting, laxative and substance use might stimulate cancer development. We examined whether individuals with an eating disorder (not restricted to anorexia nervosa) had a different risk of developing cancer. METHODS: A systematic search on Medline and Embase until 28th April 2020 identified relevant human original research publications, including all populations and all cancer types. RESULTS: From 990 records, 6 case reports and 9 cohorts were included. Some cohorts found a decreased breast (3/5 studies) or cervical (1/2) cancer risk, while an increased esophageal (2/3), liver (1/1), brain (1/1 in men) and respiratory (2/4) cancer risk, but other cancer risks were non-significant, and an increased mortality overall (1/2), from breast (1/1), female genital (1/1) and skin (1/1) cancer in eating disorderpatients. The case reports further described esophageal cancer and leukemia. No clear statistical differences in cancer risk were found depending on eating disorder type, perhaps due to the small sample size (n = 1783 for other than anorexia nervosa). CONCLUSIONS: The literature on eating disorders and cancer risk is sparse with many gaps. Hormonal changes, sexual activity, nutritional status, vomiting and concomitant tobacco/alcohol abuse may explain increased/decreased cancer risk. Future large studies (now 1-366 cancer cases) that also include men (now 4.7%), bulimia nervosa (now 3.8%) and several cancer sites (now mainly breast cancer) are needed and should foresee longer follow-up time (now 5.4-15.2 years) and extensive confounder adjustment (now only age and sex). LEVEL OF EVIDENCE: Level I, systematic review.