Depression is remarkably persistent and recurrent.Why? Although several predictors of depression chron-icity have been identified, a conceptual framework re-garding mechanisms whereby depression chronicity oc-curs is lacking. The goal of this article is to explaindepression chronicity, at least in part, with referenceto processes, mostly interpersonal in nature, that serveto extend and reestablish depression. Several such pro-cesses are described, and available empirical evidenceregarding each is reviewed. Clinical and research impli-cations of the present conceptualization are elucidated.To the degree that these processes receive continuedempirical support as mechanisms whereby depressionpersists, they represent leverage points to combat thevexing problem of depression chronicity.Key words:depression chronicity, interpersonal ex-planation, self-propagatory factors.[Clin Psychol SciPrac 7:203–218, 2000]Depression persists and recurs. Average length of majordepressive episodes is approximately 8 months in adults(Shapiro & Keller, 1981), and 9 months in children(Kovacs, Obrosky, Gatsonis, & Richards, 1997). Incredi-bly, mean length of dysthymic episodes may be as much as30 years in adults (Shelton, Davidson, Yonkers, & Koran,1997); the corresponding figure for children is 4 years(Kovacs et al., 1997). Emslie, Rush, Weinberg, and Guil-lon (1997) found that 61% of depressed children experi-enced recurrence of depression within 2 years; similarAddress correspondence to Thomas E. Joiner, Department ofPsychology, Florida State University, Tallahassee, FL 32306-1270. Electronic mail may be sent to [email protected]’s Vicious Scree: Self-Propagating and ErosiveProcesses in Depression ChronicityThomas E. Joiner, Jr.Florida State University2-year relapse rates have been reported among adults(Belsher & Costello, 1988; Bothwell & Scott, 1997). Instudies with follow-ups of 10 years or more, Coryell andWinokur (1992) found that 70% of people with onedepressive episode subsequently experienced at least onemore. In the DSM-IV Mood Disorders Field Trial (Kelleret al., 1995), the most frequent course among severalhundred patients with current major depression was “re-current, with antecedent dysthymia, without full inter-episode recovery.” Depression is thus persistent withinacute episodes, and recurrent across substantial portions ofpeople’s lives.Why is depression so remarkably persistent and recur-rent? Several clinical-descriptive predictors of depressionchronicity have been identified (e.g., longer duration ofprevious episodes, severity of previous episodes, youngerage at first episode, neuroticism; Bothwell & Scott, 1997;Maier, 1996; Miller, Ingham, Kreitman, & Surtees, 1987).Unless couched in a broader conceptual framework, how-ever, such predictors elucidate the “who” but not the“why” of depression chronicity.THE EROSIVE PERSPECTIVEOne attempt to explain depression chronicity derivesfrom conceptual and empirical work on the “scarring”effects of a depressive episode. The central idea of this per-spective is that a depressive episode erodes personal andpsychological resources, such that episodes may belengthened and, upon recovery, the formerly depressedindividual is left with fewer buffers to protect againstfuture depression. This view—referred to here as the“erosive” perspective—has received some attention, withvarying degrees of support.Personality has received the most empirical attention,and findings in this domain are unsupportive of the erosive 2000 AMERICAN PSYCHOLOGICAL ASSOCIATION D12 203shown, one set of processes may lay the groundwork fordevelopment of the other. Depression may not only erodeexisting resources, then, but may also build up processesthat serve to maintain depression vulnerability.It is the purpose of this article to argue (a) that a set ofsuch processes—referred to here as self-propagatory pro-cesses—exists, and (b) that these processes explain aspectsof depression chronicity not fully accounted for by theerosive perspective. Erosive and self-propagatory factors,however, are not viewed as exclusive, but rather as com-plementary, working in tandem to encourage depressionchronicity.It should be noted that parts of the view adopted hereare speculative and extend beyond the facts. The goal isto present a broad view that will stimulate interest andresearch, rather than to articulate a fully developed inter-personal theory of depression chronicity; as a result, some“reaching” and “selective ignoring” is unavoidable (e.g.,it is beyond the scope of the present article to reviewdepression theory in general and compare and contrast itwith the view presented here). However, I feel that theseextensions are rational, and an attempt is made to desig-nate when points should be regarded cautiously.THREE TYPES OF DEPRESSION CHRONICITYIn general, three forms of depression chronicity are distin-guished. First, consistent with research from clinical (e.g.,Brown & Moran, 1994) and interpersonal perspectives(e.g., Miller et al., 1987), episode duration is emphasized.This dimension of depression chronicity is of obviousimportance, in that long episodes involve prolongedexperience of the most acute, painful, and debilitatingaspects of depression. Second and third, also consistentwith past work from clinical (e.g., Lara et al., 1997) andinterpersonal perspectives (e.g., Hooley & Teasdale,1989), recurrence—of two types—is emphasized. Recur-rence, too, is of considerable importance, in that it canaffect substantial portions of people’s lives (cf. Bothwell &Scott, 1997). Recurrence may take two forms: relapse—the resumption of symptoms in the vulnerable time-framejust following remission, and strictly defined recurrence—theresumption of symptoms after a substantial period ofsymptom-free functioning. Where appropriate, applica-tion of self-propagatory processes to episode duration,relapse, and strictly defined recurrence is noted.These three forms of depression chronicity (episodeduration, relapse, and recurrence) correspond well withperspective. For example, Shea, Leon, Mueller, and Solo-mon (1996) conducted a 6-year study of an originallydepression-free group of participants. Those who experi-enced their first-episode of depression during the 6-yearstudy were compared to those who remained well. Per-sonality change did not differ between the groups (see alsoRohde, Lewinsohn, & Seeley, 1990; Zeiss &