AIM:Toinvestigatethelatentstructureofanirritablebowelsyndrome(IBS)symptomseverityscaleinapopulationofhealthyadults.METHODS:TheBirminghamIBSsymptomquestionnairewhichconsistsofthreesymptomspecificscales(diarrhea,constipation,pain)wasevaluatedbymeansofstructuralequationmodeling.Wecomparedtheoriginal3-factorsolutiontoageneralfactormodelandabifactorsolutioninalargeinternetsampleofcollegestudents(n=875).Statisticalcomparisonsofcompetingmodelswereconductedbymeansofχ2differencetests.Regardingtheevaluationofmodelfit,weexaminedthecomparativefitindex(CFI)andtheRootMeanSquareErrorofApproximation(RMSEA).RESULTS:ResultsclearlyfavoredabifactormodelofIBSsymptomseverity(CFI=0.99,RMSEA=0.05)whichconsistedofastronggeneralIBSsomatizationfactorandthreesymptomspecificfactors(diarrhea,constipation,pain)basedonthesubscalesoftheBirminghamIBSsymptomquestionnaire.Thefitindicesofthecompetingonefactormodel(CFI=0.85,RMSEA=0.17)andthreefactormodel(CFI=0.97,RMSEA=0.08)wereclearlyinferior.χ2differencetestsshowedthatthedifferencesbetweenthemodelswereindeedsignificantinfavorofthebifactormodel(P<0.001).Correlationsofthefourlatentfactorswithmeasuresofpainsensitivity,somatoformdissociation,fatigueseverity,anddemographicvariablessupportthevalidityofourbifactormodelofIBSspecificsymptomseverity.CONCLUSION:ThefindingssuggestthatIBSsymptomseveritymightbestbeunderstoodasacontinuousandmultidimensionalconstructwhichcanbereliablyandvalidlyassessedwiththeB-IBS.
