Component Loadings
To begin to interpret the two dimensions of your solution, look at the component loadings. All variables have a positive component loading in the first dimension, which means that there is a common factor that correlates positively with all of the variables.
The second dimension separates the variables. The variables Binge eating, Vomiting, and Purging form a bundle having large positive loadings in the second dimension. These symptoms are typically considered to be representative of bulimic behavior.
The variables Emancipation from family, School/employment record, Sexual attitude, Body weight, and Menstruation form another bundle, and you can include Restriction of food intake (fasting) and Family relations in this bundle, because their vectors are close to the main cluster, and these variables are considered to be anorectic symptoms (fasting, weight, menstruation) or are psychosocial in nature (emancipation, school/work record, sexual attitude, family relations). The vectors of this bundle are orthogonal (perpendicular) to the vectors of binge, vomit, and purge, which means that this set of variables is uncorrelated with the set of bulimic variables.
The variables Friends, Mental state (mood), and Hyperactivity do not appear to fit very well into the solution. You can see this in the plot by observing the lengths of each vector. The length of a given variable’s vector corresponds to its fit, and these variables have the shortest vectors. Based on a two-component solution, you would probably drop these variables from a proposed symptomatology for eating disorders. They may, however, fit better in a higher dimensional solution.
The variables Sexual behavior, Preoccupation with food and weight, and Body perception form another theoretic group of symptoms, pertaining to how the patient experiences his or her body. While correlated with the two orthogonal bundles of variables, these variables have fairly long vectors and are strongly associated with the first dimension and therefore may provide some useful information about the “common” factor.