Anthropometric factors and risk of myeloid leukaemias and myelodysplastic syndromes: a prospective study and meta‐analysis

Summary There is insufficient evidence linking excess body weight to risk of myeloid malignancies. We investigated this association using data from the Cancer Prevention Study‐ II ( CPS ‐ II ), and a meta‐analysis of published cohort studies. Among 152 090 CPS ‐ II participants, 387 acute myeloid leukaemias ( AML ), 100 chronic myeloid leukaemias ( CML ) and 170 MDS were identified over 21 years of follow‐up. In CPS ‐ II , body mass index ( BMI ) was weakly associated with risk of CML (hazard ratio [ HR ] = 1·04, 95% confidence interval [ CI ]: 0·99–1·09 per 1 unit increase in BMI ), AML ( HR  = 1·01, 95% CI : 0·98–1·03) and MDS ( HR  = 1·03, 95% CI : 0·99–1·07). After controlling for other anthropometric factors, no clear association was observed for height, BMI at age 18 years or weight change. In the meta‐analysis ( n  = 7117 myeloid leukaemias), BMI 25–29·9 kg/m 2 ( HR pooled  = 1·36, 95% CI : 1·12–1·59) and BMI ≥30 kg/m 2 ( HR pooled  = 1·43, 95% CI : 1·18–1·69) were associated with higher risk of myeloid leukaemia overall, compared to a BMI <25 kg/m 2 . Likewise, BMI ≥25 kg/m 2 was positively associated with both AML and CML risk individually in the meta‐analysis. These results underscore the need for large studies to detect associations with rare cancers, and show a modest, but positive association between excess body weight and myeloid malignancy risk.