Context. (Pre-)Transitional disks show gaps and cavities that can be related to ongoing planet formation. According to theory, young embedded planets can accrete material from the circumplanetary and circumstellar disks and can be detected using accretion tracers, such as the H_α emission line. Aims. We aim to detect accreting protoplanets within the cavities of five (pre-)transitional disks through adaptive-optics(AO)-assisted spectral angular differential imaging in the optical regime. Methods. We performed simultaneous AO observations in the H_α line and the adjacent continuum using the Spectro-Polarimetric High-contrast Exoplanet REsearch (SPHERE) with the Zurich Imaging Polarimeter (ZIMPOL) at the Very Large Telescope (VLT). We combined spectral and angular differential imaging techniques to increase the contrast in the innermost regions close to the star and search for the signature of young accreting protoplanets. Results. The reduced images show no clear H_α point source around any of the targets. We report the presence of faint H_α emission around TW Hya and HD163296: while the former is most probably an artifact related to a spike, the nature of the latter remains unclear. The spectral and angular differential images yield contrasts of 6–8 magnitudes at ~100 mas from the central stars, except in the case of LkCa15, with values of ~3 mag. We used the contrast curves to estimate average upper limits to the H_α line luminosity of L_Hα ~ 5 × 10^-6 L_⊙ at separations ≥200 mas for TW Hya, RXJ1615, and T Cha, while for HD163296 and LkCa15 we derive values of ~3 × 10⁻⁵ L_⊙. We estimated upper limits to the accretion luminosity of potential protoplanets, obtaining that planetary models provide an average value of L_acc ~ 10⁻⁴ L_⊙ at 200 mas, which is about two orders of magnitude higher than the L_acc estimated from the extrapolation of the L_Hα - L_acc stellar relationship. Conclusions. When considering all the objects observed with SPHERE/ZIMPOL in the H_α line, 5 in this work and 13 from the literature, we can explain the lack of protoplanet detections by a combination of factors, such as a majority of low-mass, low-accreting planets; potential episodic accretion; significant extinction from the circumstellar and circumplanetary disks; and the fact that the contrast is less favorable at separations of smaller than 100 mas, where giant planets are more likely to form.