Prediagnostic Weight Loss Linked to Worse Colorectal Cancer Survival

A population-based case-control study with prospective follow-up found that the apparent survival advantage associated with overweight and obesity at colorectal cancer (CRC) diagnosis largely disappeared after investigators accounted for prediagnostic weight loss. The findings were published in JAMA Network Open.

The analysis included 5521 patients aged 30 years or older with histologically confirmed CRC who were recruited at more than 20 hospitals in southwest Germany from 2001 through 2016. Participants self-reported height and weight at diagnosis and past weight at each decennial age beginning at 20 years. Investigators assessed overall survival (OS), CRC-specific survival (CSS), and recurrence-free survival (RFS); follow-up ended September 9, 2018.

During a median follow-up of 5.2 years, 2429 deaths occurred, including 1381 CRC-related deaths. Compared with patients with normal weight at diagnosis, patients with overweight had modestly better OS (hazard ratio [HR], 0.89; 95% CI, 0.81-0.98). Associations for class I obesity (HR, 0.90; 95% CI, 0.79-1.02) and class II or III obesity (HR, 0.89; 95% CI, 0.71-1.11) did not reach statistical significance.

Compared with patients with no weight loss or weight gain, those with more than 10% prediagnostic weight loss had worse OS (HR, 1.55; 95% CI, 1.39-1.72), CSS (HR, 1.51; 95% CI, 1.31-1.74), and RFS (HR, 1.33; 95% CI, 1.17-1.52). After mutual adjustment for BMI and weight change, associations for overweight and both obesity categories were attenuated to null, while weight loss exceeding 10% remained associated with worse OS (HR, 1.54; 95% CI, 1.37-1.72).

The authors said the findings implicated reverse causation as a major contributor to the CRC obesity paradox and cautioned that apparent survival advantages at higher BMI should not be used to encourage weight gain or change survivorship guidance. Limitations included self-reported weight and height, nonuniform intervals for recalled prediagnostic weight, possible residual survivor bias, and BMI’s inability to distinguish fat from lean mass or capture their distribution.

Source: Mandic M, Wankhede D, Safizadeh F, Hoffmeister M, Brenner H. Body mass index at diagnosis, prediagnostic weight loss, and survival in colorectal cancer. JAMA Netw Open. 2026;9(8):e2629571. doi:10.1001/jamanetworkopen.2026.29571