The theoretical modeling of the relaxation behavior of polymers in the glass transition region, advocated by Moynihan and co-workers, has been used to analyze the heat flow and the relaxation of polymer systems during isothermal modulated DSC experiments in the glass transition region. An analytic solution for the frequency dependent fictive temperature is obtained, which takes a particularly simple form in the high-frequency region. The maximal phase lag of the fictive temperature T-f is beta pi/2, where the exponent of the stretched-exponential characterizing the enthalpy relaxation, beta, is on the order of 0.1-0.7. The corresponding maximal phase lag in the heat flow is much smaller, on the order of 2-5 deg. It is once more iterated that, as observed long ago by Birge and Nagel, the loss heat capacity corresponds to the entropy production due to a redistribution of energy over the heat baths. The possibility of using specific-heat spectroscopy as a tool to determine miscibility in polymer blends whose constituents possess similar glass transition temperatures is discussed. Compared to conventional differential scanning calorimetry, the resolution is enhanced. However, in many cases an unambiguous conclusion still requires additional enthalpy relaxation of the blend induced by physical aging in the glassy state.