Common and uncommon manifestations of anxiety across the life span

The present study examined the latent structure of a common set of items appearing on the original and scaled versions of the Revised Children's Manifest Anxiety Scale (RCMAS), Adult Manifest Anxiety Scale-College Version (AMAS-C), Adult Manifest Anxiety Scale-Adult Version (AMAS-A), and Adult Manifest Anxiety Scale-Elderly Version (AMAS-E) across age and across gender to determine whether anxiety manifests itself differently across the life span and to determine whether relations exist between anxiety and age, and anxiety and gender. Subjects included 7,224 individuals, 3,802 females and 3,422 males, ranging in age from 6 to 100 years with 0 to 17 or more years of education. Results of factor analyses and item content analyses within and across measures as well as scaling of two new measures, the AMAS-A and AMAS-C, suggested that anxiety does manifest itself somewhat differently across the life span. Although some similarities along the dimensions across measures were identified, the dimensions across measures were found to be different. Results of factor analyses produced a three-factor structure of anxiety on each measure, and these results provide support for the multidimensional nature of the construct of anxiety. Moreover, results of curve estimation regression analyses revealed a curvilinear relation between anxiety and aging, with children and adolescents, college students, and older adults obtaining higher scores on the common set of items appearing on the scaled versions of the four measures of anxiety than young and middle-aged adults. In addition, children and adolescents obtained higher scores in comparison to older adults on the common set of items appearing on the scaled versions of the four measures of anxiety. Results of logistic regression analyses with gender as the criterion categorical variable indicated that anxiety and age but not the interaction between anxiety and age were significant in predicting gender. However, these predictor variables were able to explain only a small portion of the variance in gender. Implications of the findings for the design and development of measures of anxiety for specific age groups and for the diagnosis and treatment of anxiety across the life span are discussed.