[Skip to Navigation]
Views 200
Citations 0
Comment & Response
November 25, 2020

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

Author Affiliations
  • 1Center for Behavioral Cardiovascular Health, Department of Medicine, Columbia University Medical Center, New York, New York
  • 2Department of Clinical Cancer Prevention, The University of Texas MD Anderson Cancer Center, Houston
  • 3Department of Cardiology, The University of Texas MD Anderson Cancer Center, Houston
JAMA Oncol. 2021;7(1):139-140. doi:10.1001/jamaoncol.2020.5877

In Reply We thank Rezende and colleagues for their comment on our recent Original Investigation examining the association of accelerometer-measured sedentary behavior with cancer mortality risk.1 We agree that reverse causation could have confounded the reported results. As the exclusion approach to addressing reverse causality (eg, excluding early deaths) used by the study cited by Rezende and colleagues can introduce bias,2,3 to evaluate potential for reverse causality, we conducted formal tests of heterogeneity to examine whether the hazard ratios (HRs) of later years differed from HRs in the first year by testing a model that allows for a change in the hazard after 1 year vs a model that does not (eg, HR unchanged over time). The P values comparing these models were .49 (comparing HRs in the first year vs later years; for example: tertile 3 vs tertile 1 in first year: 1.49; 95% CI, 0.99-2.23; tertile 3 vs tertile 1 in later years: 1.87; 95% CI, 0.98-3.56) and .43 (comparing HRs in the first year vs later years; for example: tertile 3 vs tertile 1 in first year: 1.33; 95% CI, 0.94-1.91; tertile 3 vs tertile 1 in later years: 1.61; 95% CI, 0.87-3.00) for total sedentary time and mean sedentary bout duration, respectively (eTable 7 in the Supplement),1 thus indicating that HRs in the first year did not differ from HRs in later years. As shown in eTable 7 of our article,1 HRs were actually higher in later years for the uppermost tertile relative to the first year; albeit the test of heterogeneity indicated the HRs did not significantly vary from that for the first year. As such, these data do not show evidence for reverse causality. Nonetheless, as Rezende and colleagues note, we cannot entirely rule out reverse causality given the length of follow-up in our study.

Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    ×