On 2017 Nov 21, Ralph Brinks commented:

The authors state that Hoyer and Brinks set up "a hypothetical (micro-simulated) population" with "a linear increase of diabetes incidence to age 70 years". It is not true that we performed a micro-simulation. Instead, we used analytical relations between incidence and prevalence as cited, e.g. Brinks R, 2014. The linear increase of the age-specific incidence is an approximation to the values reported in the original paper. We agree that this an approximation only.

The key point we made is that ignoring mortality yields to biased estimates of the incidence. This point has not been - and cannot be - undermined by the authors. Thus, we strongly discourage the use of the Styblo method whenever mortality matters.

On 2017 Nov 21, Ralph Brinks commented:

The authors state that Hoyer and Brinks set up "a hypothetical (micro-simulated) population" with "a linear increase of diabetes incidence to age 70 years". It is not true that we performed a micro-simulation. Instead, we used analytical relations between incidence and prevalence as cited, e.g. Brinks R, 2014. The linear increase of the age-specific incidence is an approximation to the values reported in the original paper. We agree that this an approximation only.

The key point we made is that ignoring mortality yields to biased estimates of the incidence. This point has not been - and cannot be - undermined by the authors. Thus, we strongly discourage the use of the Styblo method whenever mortality matters.