Dimensionality of the Fonseca Anamnestic Index and validation of its short-form derivative

Recently, the Short-Form Fonseca Anamnestic Index (SFAI) was shown to have high diagnostic accuracy when compared to the Diagnostic Criteria for Temporomandibular Disorders (DC/TMD) in patient samples. This study investigated the dimensionality of the parent instrument (Fonseca Anamnestic Index [FAI]) and validated its main component using Rasch analysis in non-patient populations.

FAI data from a total of 901 participants from Singapore and Indonesia with a mean age 19.30 ± 1.48 years (65.0% women) were examined. Of these, 53.8% were FAI positive and 46.2% were FAI negative. Principal Component Analysis (PCA) was performed to extract the main component of the FAI using an eigenvalue > 1.0 and direct oblimin rotation with item loading of > 0.40. Rasch analysis was subsequently carried out on the items of the main component.

The FAI was found to be multidimensional with the main component involving items F1, F2, F3, F6, and F7 which were the items of the SFAI. The SFAI had moderate internal consistency (Cronbach alpha = 0.63) and fitted the Rasch model with person and item infit/outfit mean square (MnSq) values of 0.98/0.96 and 1.00/0.96 logits respectively. The infit/outfit MnSq of the SFAI items ranged from 0.82 to 1.06 logits with Item F2 (side-movement difficulty) being the most difficult and item F3 (muscle pain) the easiest.

The FAI is multidimensional with the main component comprising the five items of the SFAI that fitted the Rasch model. With its good Rasch validity, separation, and reliability, the SFAI is a promising tool for TMD screening.