Trusting computerized data reduction too much: a critique of Anderson's ambulatory respiratory monitor

Recording respiration ambulatorily in order to study stress and anxiety is an important and interesting topic that has been widely neglected in the past, principally due to technical difficulties. In contrast to heart rate measurements, which are very reliable with current ambulatory devices, respiratory measurements are much more susceptible to artifacts. Unfortunately, measuring inspired and expired air directly from a mask covering the mouth and nose is not practical for most ambulatory applications. Instead, elastic bands containing coils of insulated wire worn around the chest and abdomen are widely used (Respitrace bandsTM, Ambulatory Monitoring, Ardsley, NY). Their main disadvantage is that they react not only to respiratory activity, but also to postural changes, bending of the trunk, and vibrations of abdomen, chest, or breasts due to accelerative forces on the body during physical activity. Anderson and Frank (1990) designed a portable device containing a microprocessor that analyzes the output of those bands on-line to retrieve respiratory rate and depth (in units of inductance change) for each breath. These measures are averaged