R Q Hintzen, D Buljevac, W C J Hop, A C J W Janssens
Erasmus MC, Postbox 2040, 3000 CA Rotterdam, Netherlands
We agree with Galea et al that recall bias could be an alternative explanation for our finding that stressful events are associated with the risk of exacerbation in multiple sclerosis.
To minimise the possibility of recall bias we chose a high (weekly) sampling frequency. Nevertheless, when patients experienced an event and still had to complete the questions for the preceding week, their perception of stress may have been influenced by the relapse. In such a situation a high frequency of stress within or just before the week of the relapse would also be expected. This was not the case in our study: we found no significant increase in the number of reported stress events two weeks before an exacerbation.
Galea et al are right that increases in stressful events preceding exacerbations do not directly prove a causal relationówe do not claim that. They suggest that subclinical disease processes may underlie the experience of stress in the weeks before the onset of a relapse. In that case, the experience of stress does not precede exacerbations but is the consequence of subclinical disease activity.
The occurrence of events that are not related to multiple sclerosis is independent of subclinical disease processes, but patients may experience these as more stressful when they do not feel well. Yet, if this were true, one would also expect heightened report of stress at the time of infections. We did not find evidence for a relation between stress report and clinically manifest infections. These findings argue against the possibility of experiencing stress as a consequence of subclinical disease activity.
Copyright © 2004, BMJ