Population-based case-control study of body mass index at diagnosis, prediagnostic weight loss, and survival in colorectal cancer

Evidence on the association of excess adiposity with colorectal cancer (CRC) prognosis is conflicting, with some studies reporting better survival in patients with elevated body mass index (BMI), often referred to as the obesity paradox. This paradox may be explained by prediagnostic cancer-associated weight loss. A population-based case-control study at more than 20 hospitals in southwest Germany included 5521 patients aged 30 year or older with CRC between 2001 and 2016, with survival follow-up through September 2018. During a median follow-up of 5.2 years (IQR 3.4-9.9), 2429 deaths occurred, including 1381 CRC-related deaths. Compared with patients with normal weight at diagnosis (BMI 18.5 to <25.0), those who had overweight (BMI 25.0 to <30.0) had higher overall survival (HR 0.89; 95% CI 0.81-0.98), while no association was observed for obesity (BMI ≥30.0). Weight loss of more than 10% during 5 to 14 years before diagnosis was associated with worse survival (HR 1.55; 95% CI 1.39-1.72). Upon mutual adjustment for at-diagnosis BMI and prediagnostic weight change, association for overweight was attenuated to null. The authors conclude that in this case-control study with prospective follow-up, the obesity paradox appeared to be largely attributable to prediagnostic cancer-associated weight loss. The apparent survival advantage associated with overweight at diagnosis should not be interpreted as evidence to encourage weight gain for patients with healthy weight.