A major challenge in network science is to determine whether an observed network property reveals some non-trivial behavior of the network's nodes, or if it is a consequence of the network's elementary properties. Statistical null models serve this purpose by producing random networks whilst keeping chosen network's properties fixed. While there is increasing interest in networks that evolve in time, we still lack a robust time-aware framework to assess the statistical significance of the observed structural properties of growing networks. We fill this gap by introducing a null model that, differently from static models, preserves both the network's degree sequence and the individual nodes' degree time-series. By preserving the temporal linking patterns of the real systems it is applied to, the proposed model disentangles structural from temporal patterns. As a consequence, it allows us to properly assess the significance of structural properties in settings where temporal patterns heavily impact the outcomes of structural measurements, which is the case for a wide range of real networks. We apply the model to real citation networks to explore the significance of widely studied network properties such as degree-degree correlations and the relations between popular node centrality metrics. The proposed model can be used to assess the significance of structural properties for any growing network.