Noise and noise propagation are vital in various biological processes. In a phenotypic transition cascade of colonic cells, there are three compartments: stem cells (SCs), transit-amplifying cells (TACs), and fully differentiated cells (FDCs). Instead of pure linear feedback (L) or pure saturation feedback (S) in regulating SCs and TACs differentiation, four coupled feedbacks including LL, LS, SL, and SS are considered to compare the impacts of different feedbacks on noise and noise propagation of cells. Rather than LL feedback and LS feedback, SL feedback and SS feedback have obvious advantages: Firstly, the characteristics of cells steady states changing with the parameter are more consistent with the tumor development. Secondly, as long as the key parameter is adjusted reasonably, the strong correlation between any two fluctuations of cells can be effectively utilized or avoided. Finally, although there is no direct interaction between SCs and FDCs, the noise in SCs can propagate to FDCs by TACs. And the transmitted noise from upstream cells can cause the large total noise of downstream cells even the number of downstream cells is large. The strong correlation and noise propagation between upstream and downstream cells may be the theoretical basis of the targeted therapy, which can achieve dual or even triple-drug targeted anti-tumor therapy by acting on different targets in the upstream or downstream pathways. Therefore, SL feedback or SS feedback may be a better choice for SCs and TACs to adjust their differentiation rather than LL feedback and LS feedback. This work is an extension and application of the elementary fluctuation theory of statistical physics in life system. •We discuss the influence of different feedbacks on noise and noise propagation.•The elementary fluctuation theory and stochastic dynamics theory are used.•Noise can propagate from upstream cells to downstream cells due to the interaction between cells.•Even if the total number of downstream cells is not small, their noise will be large due to the propagation of noise.