Objective. The present study sought to derive an algorithm using factor ana
lysis and structural equation modeling (SEM) to describe headache and orofa
cial pain patients using measures of behavioral and psychological functioni
ng. This investigation further examined whether the underlying factor struc
ture differed in 3 presumed distinct diagnostic categories: myofascial, neu
ropathic, and neurovascular pain.
Design. The Minnesota Multiphasic Personality Inventory-2 ((MMPI-2), Multid
imensional Pain Inventory (MPI), Beck Depression Inventory-II (BDI-II), and
visual analog scale for functional limitation (VAS-FL) were administered t
o the subjects. A split group design was used. Exploratory factor analysis
(EFA) was used to describe distinct factor domains in the first group. Conf
irmatory factor analysis (CFA) using SEM tested this structure in the secon
d group and described causal relationships between the revealed (latent) fa
ctors. Analysis of variance (ANOVA) was used to test for differences in dem
ographic variables and diagnostic group factor structure.
Setting. The Pain Center is a comprehensive, multidisciplinary pain medicin
e program at Cedars Sinai Medical Center, Los Angeles, California.
Subjects. Three hundred and ninety (N = 390) subjects were assigned to 1 of
3 diagnostic categories: myofascial pain syndrome, neuropathic pain, or ne
urovascular pain.
Results. EFA revealed a 3-factor solution. The factors were labeled Depress
ion, Illness Conviction, and Pain Impact, reflecting the content of their r
espective variables with highest loadings. CFA using SEM validated the 3-fa
ctor solution, and further revealed that Depression was a critical causal f
actor determining Illness Conviction and Pain Impact. No causal relationshi
p was observed between Illness Conviction and Pain Impact. ANOVA found no d
ifferences in demographics. No difference in factor structure emerged for t
he 3 diagnostic categories.
Conclusions. Analysis derived a 3-factor solution. The factors were Pain Im
pact, Illness Conviction, and Depression. SEM revealed the critical causal
pathway showing that Depression determined Illness Conviction and Pain Impa
ct. We conclude that the main target for pain treatment is depression. No d
ifferences in factor structure were found for the 3 diagnostic categories o
f myofascial, neuropathic, or neurovascular pain. This suggests that psycho
logical processes are similar in chronic headache and orofacial pain patien
ts despite their presumed distinct underlying pathophysiological mechanisms
. SME is a powerful methodology to construct causal models that has been un
derutilized in the pain literature.