Resumo:
Introduction – Temporomandibular dysfunction (TMD) is a condition affecting the
temporomandibular joint (TMJ) and is associated with symptoms such as facial pain,
headaches, and limited mandibular movement. Studies suggest that TMD may impact postural
balance (PB), a crucial function maintained by the complex interaction between the body's
sensory and motor systems. Understanding the relationship between TMD and PB is essential
for improving the assessment and treatment of patients with TMD. The objective of this thesis
was to explore the relationship between TMD and PB from both linear and machine-learning
perspectives. Specifically, it aimed to compare PB between individuals with and without TMD
and develop a decision tree representing the interaction between TMD and PB. Methods –
This is an observational, cross-sectional study conducted with a non-probabilistic sample of
50 women divided into groups with (37) and without (13) TMD. Linear measures were used
to assess PB, including the area and velocity of the centre of pressure (COP) sway, as well as
circular variables such as Rho (line size) and Theta (angle). The conditions tested included
open/closed eyes, open/closed mouth, and a base of support at hip width and semi-tandem.
Machine learning algorithms were applied to explore the underlying mechanisms of TMD,
such as neuromuscular and sensory integration and psychosocial variables. Results –
Participants with TMD showed greater COP sway area and velocity, particularly under
conditions of closed eyes and a base of support at hip width. Analyses of Rho and Theta
variables revealed significant differences between the groups under certain conditions,
indicating greater magnitude in sway and less directional variation of the COP in participants
with TMD. Article 1 confirmed the significant influence of TMD on PB, while Article 2
highlighted the relevance of decision trees for diagnosing TMD, emphasising the importance
of joint sounds and postural balance variables. Conclusion – The presence of TMD can affect
postural stability, and the use of decision trees indicates that the key variables for diagnosing
TMD are joint sounds and postural balance. Therefore, the results of this study support the
notion that both in clinical practice and future studies, the effectiveness of therapeutic
interventions for TMD should be assessed with particular attention to these two variables.
