BackgroundChronic fatigue syndrome is likely to be a heterogeneous condition. Previous studies have empirically defined subgroups using combinations of clinical and biological variables. We aimed to explore the heterogeneity of chronic fatigue syndrome.MethodWe used baseline data from the PACE trial, which included 640 participants with chronic fatigue syndrome. Variable reduction, using a combination of clinical knowledge and principal component analyses, produced a final dataset of 26 variables for 541 patients. Latent class analysis was then used to empirically define subgroups.ResultsThe most statistically significant and clinically recognizable model comprised five subgroups. The largest, ‘core’ subgroup (33% of participants), had relatively low scores across all domains and good self-efficacy. A further three subgroups were defined by: the presence of mood disorders (21%); the presence of features of other functional somatic syndromes (such as fibromyalgia or irritable bowel syndrome) (21%); or by many symptoms – a group which combined features of both of the above (14%). The smallest ‘avoidant–inactive’ subgroup was characterized by physical inactivity, belief that symptoms were entirely physical in nature, and fear that they indicated harm (11%). Differences in the severity of fatigue and disability provided some discriminative validation of the subgroups.ConclusionsIn addition to providing further evidence for the heterogeneity of chronic fatigue syndrome, the subgroups identified may aid future research into the important aetiological factors of specific subtypes of chronic fatigue syndrome and the development of more personalized treatment approaches.
Latent class analysis of a heterogeneous international sample of patients with myalgic encephalomyelitis/chronic fatigue syndrome
