We estimated a non-Stationary dynamic factor model and used it to generate artificial episodes of disinflation (permanent changes in the mean inflationrate). These datasets were used to test the forecasting abilities of alternative underlying inflation indicators (i.e. measures that capture sustained movements in inflation extracted from information in a disaggregated set of price data).We found that the out of sample forecast errors of the benchmark underlying inflation measures (based on unobserved trend extraction) are more severely affected by disinflation than the alternative simpler methods (based on exclusionor re-weighting approaches). We also show that a non-stationary dynamic factor model may be employed for the extraction of the unobserved trend to be usedas an underlying inflation measure.