Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Editor,—I admired Mulholland et al's study1 into whether stress is a trigger for recurrences of acute anterior uveitis (RAAU). They conclude that it does not, but I wonder if this conclusion would have been strengthened by the use of a more appropriate control group.
A more suitable control group to answer the question of whether stress is a trigger for RAAU would consist of people with RAAU whose disease was quiescent. Instead, they used other eye casualties who may well have been under a different level of stress from people with inactive RAAU going about their daily business. Indeed, the trend that they highlight for 19–39 year old men with active RAAU to have higher Spielberger state-trait anxiety inventory (STAI) scores than published normals, is also present in the trait score of similar controls.
Casualty controls had higher life event scores (average 80.1) than patients with active RAAU (average 67.54). The authors state that the RAAU score was not higher than controls, but with no statistical comparison. Is this because the control's life event score was in fact significantly higher? I would be interested in their response.
Editor,—I was pleased to see that Mr Beare admired our paper. The patients with RAAU attended the casualty department within 1–3 days of their attack and the events assessed in the questionnaires were, over the previous month, looking for a significantly increased occurrence of stressful events that could be involved in triggering a recurrent attack. They were therefore walking around quiescent for most of the relevant time looked at in the study.
There was no significant difference (p>0.05) in our controls compared with those from other published studies matched for age and sex as detailed in Tables 3 and 4 and Figures 1 and 2 in our paper. All studies of this kind report large standard deviations, as does this one, and it is important not to overinterpret results that are, in fact, non-significant, however interesting a trend might be.