Deviations from whiteness in the innovations of a Kalman filter indicate that the filter is not optimal for the given data. Lack of optimality can come from changes in the system properties but also from discrepancies between the noise statistics used to formulate the filter and the actual values. This paper examines the importance that changes in the noise part of the filter have on its ability to function as a damage detector. Analysis and numerical results show that the sensitivity is such that a realistic assessment of performance demands that potential fluctuations in the noise statistics be accounted for.