Intended for healthcare professionals

CCBYNC Open access

Rapid response to:


Sugary drink consumption and risk of cancer: results from NutriNet-Santé prospective cohort

BMJ 2019; 366 doi: (Published 10 July 2019) Cite this as: BMJ 2019;366:l2408

Related BMJ Opinion

“Sin taxes”—the language is wrong, but the evidence is clear

Rapid Response:

Zero incident cases represent higher than expected number of cases?

In the last 2 rows of Table 2 under Quartile 3, it shows Incident Cases = 0, which corresponds to sHR = 1.41.
Doesn't that imply that the expected number of cancer cases per your risk model for Quartile 3 is less than zero.
A negative expected number of cancer cases is impossible.
That is, how can zero cases of cancer represent a higher than expected risk of cancer?

Competing interests: No competing interests

20 July 2019
Shawn Phelps
Statistical Programmer
Oregon, U.S.A.