The filtration effect improves the impedance change through specific binding of target molecules in plasma, and decreases this change by nonspecific binding of matrix factors in plasma (i.e., matrix effect). A difference in dielectrophoresis (DEP) forces applied to target molecules and matrix factors causes the filtration effect. An optimized DEP force affects target molecules, which remain in the reaction region of an interdigitated microelectrode (IME) sensor. Various matrix factors, which are larger than the target molecules, are influenced by a strong DEP force and are filtered out of the reaction region. To demonstrate the filtration effect, the matrix effect was confirmed in standard plasma and in phosphate-buffered saline, based on the detection of amyloid beta (Aβ), an Alzheimer's disease (AD)-associated peptide. The filtration effect was verified using the matrix effect factor (MEF), which was calculated from the impedance change values in different detection environments. In standard plasma, the MEF value decreased by approximately 78.12%, and in buffer with heterogeneous Aβ by approximately 75.43%. Plasma from patients with AD and normal controls (NCs) was analyzed using the value of the impedance change by the filtration effect. The impedance change was enhanced approximately 1.52 ± 0.03-fold in AD plasma, but declined approximately 0.90 ± 0.03-fold in NC plasma. This difference tendency by the filtration effect was the disease evaluation index and used as an important criterion that distinguished between the AD and NC plasma. Plasma-based AD diagnosis may be possible, based on the filtration effect.