Abstract: Biological systems include a huge number of molecular entities which interact non-linearly. Even the simplest bacterial cell includes thousands of different molecular entities. Cellular pathways can be viewed as being evolved to reduce this complexity down to a lower dimensional space, leading to a defined cellular state.  However, we observe that when cells are exposed to conditions that drive them away from this defined state, their underlying complexity is revealed, leading to a disrupted cellular state. We characterize the disrupted state and show that universal features are observed under different conditions. By analogy with physical ageing in disordered systems, we show that the disrupted state’s dynamics can be reproduced with a statistical toy model that ignores most of the underlying details. Our understanding of the disrupted cellular state can account for various phenomena with clinical implications, such as the way bacteria respond to antibiotic treatments, or the recovery of cells after stress. More generally, the work identifies a cellular state, that may be common to all biological cells, which can be described by statistical physics rather than by biological regulation.