Predictors of Post-Infectious Chronic Fatigue Syndrome in Adolescents, 2014, Jason et al

This 2014 paper is open access and pretty interesting.

On rates of CFS following glandular fever - 3 studies quoted finding about 10% of people have CFS at 6 months after diagnosis of glandular fever

This study: 301 adolescents recruited in Chicago; a baseline home-visit on average 2 months after diagnosis:

At six-months, a telephone screening to check for CFS. Complete medical and psychiatric workup from adolescents with self-reported CFS and screened-negative controls. 39 cases and 50 controls.

 

Of the 39 adolescents with CFS, 90% were female. The Jason et al revision of Fukuda was used to diagnose CFS - I don't know if that includes PEM or not.

Finding: (ASC = Autonomic Symptoms Checklist)

The lack of association between psychological factors and illness onset accords with my experience and that of my children. However, there do some to be some items quantified in this study associated with stress of various kinds that had some association with CFS onset. (See Table 2). The Perceived Stress Scale measures stress over the last month and could easily have been affected by being sick and so unable to meet obligations and enjoy life as normal. But the Life Events score also seemed to have a good P value, and that measures events over the preceding 12 months. The authors seemed to wave this away a bit, noting that the timing of the baseline survey meant that it probably included about 3 months of the 12 month period assessed for stressful events.

I guess what they tried to do was say, if we explain the likelihood of CFS onset as much as possible by this measure of illness severity (the Autonomic Symptoms Checklist), what other factors strengthen the predictive power of an equation? And they report that the only other predictive factor was days in bed. But there is mention of a checklist of of infectious symptoms that sounds like it might have been a better, more direct measure of the severity of IM symptoms:

Why, I wonder was a quantification of this checklist not used, rather than just the number of days spent in bed? I don't get a clear idea about the number of possible measures that could have been used to find associations.

I'm left wondering, what if someone said, 'I think the stressful events before and just after infection are the most important factor and ANS symptoms and number of days in bed are a result of that stress'? If the stepwise regression was done with the Life Events score first, what would happen?

I'm also not sure about the Jason et al version of Fukuda - does it include PEM?

I'm left feeling rather uncertain about whether this study proves the authors' hypothesis that severity of the IM illness is the primary predictor of CFS. I look forward to others' comments.