Association of past obesity and BMI trajectories with cancer mortality: a prospective cohort study.

OBJECTIVE

This study aimed to assess the association of past obesity and past BMI trajectories with cancer mortality in National Health and Nutrition Examination Survey.

METHODS

Past obesity was identified based on past maximum body weight, and trajectories of past BMI change were determined by latent class trajectory modeling (LCTM). Cox regression was used to assess the association of past obesity and past BMI trajectories with cancer mortality.

RESULTS

A total of 4,058 cancer patients participated in this study, of which 46.3% were past obesity, resulting in a significantly lower risk of cancer mortality compared to participants who were not past obesity (HR = 0.92, 95% CI: 0.92-0.93, P < 0.01). The LCTM identified five trajectories of past BMI, and compared with participants whose BMI remained in the normal range, the risk of death was 17% and 23% lower for participants in the "Long-term overweight" and "Long-term obesity" trajectory groups, respectively. In contrast, participants in the "Recent weight gain" and "Recent weight loss" trajectory groups had an increased risk of cancer death of 19% and 40%, respectively.

CONCLUSIONS

This study found that past obesity is consistent with the "obesity paradox." In furtherance, a moderately elevated and stable BMI might be associated with lower cancer mortality.