Are symptoms clustered into latent classes in pediatric functional neurological disorder?

DSM-5-TR and ICD-11 classify functional/dissociative seizures (FDS, formerly known as psychogenic nonepileptic seizures [PNES]) and functional movement disorder (FMD) under the broad category of functional neurological disorder (FND). However, substantial heterogeneity within FND has led some to hypothesize that FDS and FMD may represent distinct disease entities. To test this hypothesis, we aimed to determine whether the variation in FND symptoms among pediatric patients is best explained by either a single disease entity or two or more distinct disease entities.

We analyzed data from 155 Australian children and adolescents with FND who were recruited between 2006 and 2023 and who had one or more symptoms across four categories: negative motor symptoms (e.g., limb weakness), positive motor symptoms (e.g., tremor), sensory deficit symptoms (e.g., loss of touch), and functional seizures. We performed latent class analysis (LCA) with binary indicators (e.g., symptom present/absent in each category) to identify latent classes (subgroups) based on non-random patterns in symptom co-occurrence.

LCA supported a two-class model rather than a single-class, three-class, or four-class model. The Bayesian Information Criterion (BIC) was lowest for the two-class model (839.2), compared with 866.2 for the single-class model, 850.9 for the three-class model, and 871.3 for the four-class model. In the optimal two-class solution, the first class (70.7 % of the sample) frequently exhibits negative motor symptoms (76.2 %) as well as positive motor symptoms (64.3%), along with occasional functional seizures (37.9 %) and sensory deficit symptoms (49.3%). The second class (29.3 % of the sample) is characterized by functional seizures in all cases (100 %), occasional sensory deficit symptoms (10.9 %) and positive motor symptoms (29.9 %), with almost no negative motor symptoms (1.1 %).

Our latent class analysis of symptom profiles suggests that the Australian pediatric FND cohort is best explained by a two-class model, in which the two classes have a substantial overlap and can produce the same set of symptoms.