ÌÇÐÄvlog

[Skip to Navigation]
Sign In
Views 526
Comment & Response
November 25, 2020

Possible Reverse Causation and Confounding in Study of the Association of Sedentary Behavior With Cancer Mortality

Author Affiliations
  • 1Departamento de Medicina Preventiva, Escola Paulista de Medicina, Universidade Federal de São Paulo, São Paulo, São Paulo, Brazil
  • 2Department of Nutrition, Harvard T.H. Chan School of Public Health, Boston, Massachusetts
  • 3Departments of Nutrition and Epidemiology, Harvard T.H. Chan School of Public Health, Boston, Massachusetts
JAMA Oncol. 2021;7(1):138-139. doi:10.1001/jamaoncol.2020.5874

To the Editor In their Original Investigation, Gilchrist et al1 estimated the association of sedentary behavior with cancer mortality using data from 8002 middle-aged US adults from the Reasons for Geographic and Racial Differences in Stroke (REGARDS) cohort study. The use of accelerometer data for 7 consecutive days was an important improvement toward a more accurate measurement of sedentary behavior compared with studies using self-reported sedentary behavior. While reducing measurement error is laudable, reverse causation bias and residual confounding may have exaggerated the magnitude of their observed association.1 Cancer frequently persists for years or decades. The often-prolonged trajectory toward disability and illness, strong predictors of death, is likely to increase sedentary behavior (reverse causation). Individuals receiving active treatment for cancer were excluded at baseline, but it is unclear if all patients with cancer were excluded. Moreover, deaths that occurred in the first several years may have been latent cancers at baseline.

Add or change institution
×