Table 1. Resampling algorithm.

for k = 1 to K

for s = 1 to 2

sample $m_1^{(s)}$, i (s), p (s) from the KORA data at ages a = 60 and a = 70

make log-parabola fit for i (0,s)

make log-linear fit for R (s)

sample relative risk reduction h and set i (1,s) ← (1 − h)i (0,s)

calculate p (ℓ,s), ℓ = 0, 1, by solving the PDE (3)

calculate the outcomes y k by applying Eqs (4) and (5), (7)

end

end

The algorithm simulates the effects of the increased active travel scenario compared to business as usual scenario.