Self-esteem lability (SEL), defined as daily event-related variability in state self-esteem, and low trait self-esteem (TSE) were assessed among 205 male and female undergraduates who were currently depressed (CD), previously depressed (PD), and never depressed (ND). SEL scores were derived for the effect of positive, negative, and combined events on state self-esteem over 30 days. Consistent with psychodynamic and cognitive theories, SEL was found to be a better index of depression proneness than TSE. PD Ss showed higher lability on all SEL scores than ND controls but did not differ from controls on TSE. Ss were reassessed 5 months later, and new cases showed higher premorbid SEL than ND controls but did not differ from controls on premorbid TSE. SEL at Time 1 was found to increase risk for depression at Time 2 among Ss reporting high life stress at Time 2. Theoretical and methodological implications are discussed.

The present study was designed to examine responses to depressive interpersonal behavior. Undergraduate subjects interacted with a same-sex confederate for 7 min. in the context of waiting together for an experiment to begin. Confederates employed either a depressed role (depressive interpersonal behavior and reporting serious deficits in functioning), a normal role (normal interpersonal behavior and reporting minimal deficits in functioning), or a physically ill role (normal interpersonal behavior and reporting serious deficits in functioning). Subjects who interacted with a "depressive" responded with a higher rate of silences and directly negative comments and a lower rate of overall verbal responding. Their expressions of direct support were equivalent to those made to the "physically ill" confederates and greater than those in the normal condition. Subjects also were more rejecting of partners who behaved in a depressed manner and described them in more negative terms and as having greater interpersonal impact than confederates in other roles. There were no induced mood differences.A steadily growing body of literature has begun to delineate interpersonal processes in depression. Several investigators have reported evidence concerning the social behavior of the depressive, ranging from fewer interpersonal behaviors of any type (Lewinsohn & Shaffer, 1971;Libet & Lewinsohn, 1973) to more self-punitive interpersonal behavior (Forrest & Hokanson, 1975).Despite their theoretical importance, data describing the part played by the social partner in the interaction have, until recently, been absent from the literature. Coyne, in his 1976a study, examined the reactions of undergraduate females to telephone conversations with depressed female outpatients. He found that his subjects were themselves more depressed, anxious, and hostile after talking with a depressive than were subjects who spoke with a nondepressed patient or a nonpatient control. Further, his subjects were

Three groups of subjects (depressed, nondepressed/other psychological problems, and normal) interacted with a same-sex normal person in a modified Prisoner's Dilemma procedure in which each player's relative power was manipulated. Dyads also had several opportunities to exchange communications during the game. The results indicated that when depressed individuals were in the high-power role, the interactive pattern in the Prisoner's Dilemma procedure was relatively exploitive and noncooperative. High-power depressed persons also displayed elevation in communications of self-devaluation-sadness and helplessness. This array of behaviors elicited noncooperativeness, extrapunitiveness, and expressions of helplessness in their normal partners. Depressives in the low-power role exhibited no unique game behaviors but communicated self-devaluation and helpless messages; and in addition, they tended to blame their partner for their devalued condition-a pattern that elicited more ingratiating behaviors in their normal partners. These findings are discussed in relation to social-interactional approaches to depression. This report is an attempt to extend the literature on the interpersonal behaviors of depressed individuals and to study further the reactions of others to those behaviors. The growing body of data on these issues (cf.

Depressed college students were compared with other-psychopathology and normal controls regarding the relationship they developed with dormitory roommates during a 9-month period. Diagnostic status was periodically assessed via SADS interviews, thus also permitting identification of new cases of depression during the year. Psychosocial characteristics found to be uniquely associated with current depression were: (a) low social contact with roommates, (b) low enjoyability of these contacts, and (c) high life-event stress. Roommates of depressives reported low enjoyability of the relationship and high levels of aggressive behavior towards the depressive. No features were found to be uniquely associated with new cases before they became depressed; however, several antecedents of general psychopathology were identified.

No abstract