The Factor Structure of the Twelve Items General Health Questionnaire (GHQ-12): a Population Based Study

Mental health is a key component of general health based on world health organization (WHO) definition of health. A short form of the general health questionnaire (GHQ-12) has been widely used as an assessment tool of general and mental health. This study was designed to evaluate the properties of the Persian version of this questionnaire. Of 6140 subjects aged 10–90 years of a cohort study of Northern Iran, the relevant data of 5074 individuals aged ≥18 years were considered for analysis. Confirmatory factor analysis was conducted to evaluate the six models including one-dimensional model, one-dimensional model with correlated errors on negative items, two-dimensional model based on negative and positive statement items, two-dimensional model based on social dysfunction and psychological distress, three-dimensional model based on anxiety/depression, social dysfunction and loss of confidence and finally one-dimensional model with an additional factor for the negatively phrased items. The best fit indices belonged to one-dimensional model with correlated errors on negative items with χ2 = 686.8, df = 39, CFI = 0.953, RMSEA = 0.057 (95%CI: 0.054–0.061), SRMR = 0.0317, AGFI = 0.952, TLI = 0.920 and AIC = 764.8 and also three-dimensional model with χ2 = 860.7, df = 51, CFI = 0.941, RMSEA = 0.056 (95%CI: 0.053–0.059), SRMR = 0.0343, AGFI = 0.955, TLI = 0.923 and AIC = 915.0. In the comparison between these two best models we had ?χ2 = 174.2, ?df = 12, p-value < 0.001 in favor of the one-dimensional model. One-dimensional model with correlated errors on negative items showed the best fitting indices with our population.