Despite numerous studies conducted by various research teams, predicting long-term outcomes (known as Post-COVID-19 Syndrome, PCS) that may result from Coronavirus Disease 2019 (COVID-19) remains challenging. PCS affects over a million people, primarily those with comorbid conditions. Therefore, it is crucial to undertake research aimed at developing a predictive model for early diagnosis of PCS, which in turn would enable faster preventive actions. The aim of this study was to assess the value of measuring and attempt a quantitative evaluation using Enzyme-Linked Immunosorbent Assay (ELISA) tests of three non-serum proteins, whose presence in the blood during COVID-19 was associated with severe disease progression: neutrophil elastase (NE), calcium-binding protein S100B, and neuron-specific enolase (NSE). The concentrations of these proteins were measured in blood serum samples collected before the COVID-19 pandemic from (1) patients with type 2 diabetes (T2DM); (2) advanced stage diabetic nephropathy (NfT2DM); (3) a healthy group; and in blood serum samples collected two years after recovering from COVID-19 from patients with (4) T2DM and (5) NfT2DM. It was found that elevated levels of NE and NSE were significantly more common (p < 0.05) in patients with NfT2DM after recovering from COVID-19 compared to the other groups, while elevated levels of S100B were significantly more frequently observed in patients with T2DM after recovering from COVID-19 (p < 0.05). Demonstrating differences in the prevalence of NE, NSE, and S100B in individuals who recovered from COVID-19 with T2DM and NfT2DM makes these proteins important components of the developing predictive model for early detection of PCS. To our knowledge, this is the first study showing the significance of NE, NSE, and S100B in PCS in the context of T2DM and NfT2DM.