The planar oxygen isotope effect on T_c observed in copper oxide superconductors is remarkable in that it increases from near nil at optimal doping to a value twice that derived from BCS theory in the underdoped region. This behavior is quantitatively followed by a formula proposed by Kresin and Wolf in 1994 for polarons along the c-axis. Herein it is revisited in a more transparent way, and it is pointed out that the heterogeneity of pairing is relevant and has to be taken into account to explain the unusual planar isotope effects on T_c in underdoped cuprates.