The conventional International Terrestrial Reference Frame (ITRF) is based on the combination of solutions from several space geodetic techniques, including observations until the end of 2020, and incorporates updated data and models. In contrast, the Celestial Reference Frame (CRF) is a VLBI-only solution based on data until 2015. These differing approaches could potentially lead to a gradual loss of consistency among EOP series. The recent release of ITRF2020 presents an opportunity to investigate this issue further. This study empirically evaluates the consistency among different terrestrial reference frames (TRF) and the newest celestial reference frame (i.e. ICRF3), and EOP using historical Very Long Baseline Interferometry (VLBI) data from 1984 to 2022. Different TRFs are used as alternative settings in the analysis, including ITRF2020, VTRF2020, DTRF2020 and ITRF2014. Additionally, Helmert transformations are computed to determine the extent to which any observed behavior can be attributed to orientation differences of the TRFs. The results evidence the robustness of the different analyzed products, but more effort needs to be made to achieve the stringent goals for the ITRF realization in terms of accuracy (1 mm) and precision (0.1 mm/year).