Effects of stabilization splint combined with manual therapy on condylar morphology and skeletal parameters in patients with temporomandibular disorders

Purpose To evaluate the effects of stabilization splint combined with manual therapy on condylar morphology and skeletal parameters in patients with TMD using CBCT. Methods This retrospective cohort study included 200 TMD patients treated between January 2022 and December 2025. Condylar volume (CV), condylar height (CH), degenerative changes, joint space parameters, cephalometric skeletal parameters (SNA, SNB, ANB, AFH, PFH, ramus height, mandibular body length, mandibular plane angle), visual analog scale (VAS) pain scores (rest, opening, chewing), maximum mouth opening (MMO), and Mandibular Function Impairment Questionnaire (MFIQ) scores were assessed. Results Patients were divided into a combined therapy group (n = 123) receiving stabilization splint plus manual therapy and a stabilization splint group (n = 77) receiving splint alone. CBCT scans were performed at baseline and at 6-month follow-up. At 6 months, the combined therapy group showed significantly greater CV (P = 0.040) and CH (P = 0.036), lower rates of flattening (P < 0.001) and erosion (P = 0.002), a lower SNB (P = 0.017), higher ANB (P < 0.001), larger condylar angle (P = 0.015), greater AFH (P = 0.023), lower VAS-rest (P = 0.007), VAS-open (P < 0.001), VAS-chew (P < 0.001), greater MMO (P < 0.001), and lower MFIQ scores (P < 0.001) compared with the splint alone group. Adjusted between-group differences were 29.8 mm 3 (95% CI 1.2–58.4) for condylar volume and 0.31 mm (95% CI 0.02–0.60) for condylar height. The primary outcomes were condylar volume and condylar height; all other endpoints were secondary and exploratory. No multiplicity adjustment was applied. Conclusion Combined stabilization splint and manual therapy was associated with greater improvements in condylar morphology, skeletal parameters, pain, and mandibular function than splint alone. However, given the non-randomized retrospective design and lack of multiplicity adjustment, these findings are exploratory associations rather than causal effects.