326 STEIGER.ALLEMAND.ROBINS.AND FEND Adolescence as a Time of Change and Challenges with stable and global causes such as being gencrally inco Given the reorganization that takes place during adol in orde ion.negative the cks(967)cognitive theory of depre ck.De Raed& .08:Orth.Robins 2009)dem Thus ione deve ental traiectories in trated that low self-esteem a ctss risk factor fo 2005.d uations.specifi shown to b ved physical appearance and soci rg,2008:T 202 ski.&Do Self-Esteem Change and Depression together,it is important to consider adole ge for s the adole cnc (c.g s a predictor re that chan d n dual differences ntrindvidoalchaneiniicd加al may hin th onstructs.If self- n is malleabl durin nd dit iduals do not endure de of g t life acn as s study we the adaptation to age-specific developmental task oms as nge iod and a time of heightened self-exploration that laysthe wo dc es later ate it is Self-Esteem Level and Depression Several theoretical models suggest predictions about the longi ents the adoles ent years must be R s the onset and main ce of de ion (Beck.196 0 e-specific challenees adequately mieht be mor ative in ater ntial styles about th ther self-e nd thus dysfunctional attitudes toward themse es ar self see Rohde esteem are less likely to rch for also y Ziv,M after a fuilure del has been e nd th ential styles imply associating lack of success in a specific taskAdolescence as a Time of Change and Challenges Given the reorganization that takes place during adolescence, adolescents are prompted to show increased introspection in order to find out who they really are (and want to be), how they are perceived by their environment, and what they want to do and achieve in their lives (Steinberg, 2005). It can lead to later mal￾adjustment if this process of scrutinization is not successful and an unstable identity is being built up (Erikson, 1968; Harter, 2006). Thus, ignoring developmental trajectories in self-esteem during adolescence may neglect important aspects of an inherently dy￾namic construct (Greene & Way, 2005; Steinberg, 2005). Indeed, self-esteem has been shown to be especially malleable in adoles￾cence (Demo, 1992; Steinberg, 2008; Trzesniewski et al., 2003). Furthermore, the relatively lower consistency of self-esteem dur￾ing adolescence implies that it is more amenable to intervention during this developmental period (Robins, Trzesniewski, & Don￾nellan, 2012). Taken together, it is important to consider adolescent change for several reasons: First, self-esteem and other personality traits are not entirely stable constructs but systematically change across the life span, and especially during adolescence (e.g., Steinberg, 2008). Second, recent findings on the importance of studying change as a predictor revealed that change in certain personality domains influences important life outcomes. Third, malleability of self-esteem and personality traits allows for intervention programs within these constructs. If self-esteem is malleable during adoles￾cence, practical interventions aimed at improving low self-esteem should be considerably more effective than if we assume stability of this construct over time. Such findings would further highlight the importance of intervening early in the life span, so that indi￾viduals do not endure decades of greater risk for important life outcomes such as mental health problems. In this study, we there￾fore investigated the prospective effects of level and change in self-esteem across adolescence on depressive symptoms assessed two decades later, when study participants were 35 years old. Self-Esteem Level and Depression Several theoretical models suggest predictions about the longi￾tudinal association between personality characteristics such as self-esteem and depression (Klein, Kotov, & Bufferd, 2011; Orth & Robins, 2012). First, one prominent theoretical model assumes that level of self-esteem is predictive for depression. The basic idea of the vulnerability model is that low self-esteem causally influ￾ences the onset and maintenance of depression (Beck, 1967; Met￾alsky, Joiner, Hardin, & Abramson, 1993). This model assumes that individuals displaying dysfunctional attitudes or negative in￾ferential styles about themselves are at greater risk for developing depressive symptoms. For example, when individuals with low self-esteem and thus dysfunctional attitudes toward themselves are confronted with negative feedback at work, they may think that their self-esteem depends on others’ approval and thus feel worth￾less even after supportive criticism (Abramson, Metalsky, & Al￾loy, 1989; Beck, 1987). Another mechanism could be that adolescents with low self￾esteem are less likely to search for positive feedback from others (Cassidy, Ziv, Mehta, & Feeney, 2003). Furthermore, they may also use negative inferential styles after a failure. Negative infer￾ential styles imply associating lack of success in a specific task with stable and global causes such as being generally incompetent instead of assuming that one was not successful at a specific task in a specific situation (Abramson et al., 1989). Hence, according to Beck’s (1967) cognitive theory of depression, negative beliefs about oneself are a key cause in the etiology of depression. Several studies support the vulnerability model (e.g., Franck, De Raedt, & DeHouwer, 2007; Orth, Robins, & Roberts, 2008; Orth, Robins, Trzesniewski, Maes, & Schmitt, 2009). For example, Orth et al. (2009) demonstrated that low self-esteem acts as a risk factor for depression, but not vice versa. Finally, Bolognini, Plancherel, Bettschart, and Halfon (1996) showed that domain-specific self￾evaluations, specifically perceived physical appearance and social skills, at age 12 and 14 were related to subsequent depressive mood, albeit with weaker effects than global self-esteem.1 Self-Esteem Change and Depression One unique feature of the present study is that we were able to test whether change in self-esteem predicts depression. To date, almost no research has been conducted on self-esteem change as a predictor of depression. However, considering psychological changes in the study of development is a key component in life span development theory (e.g., Baltes, Lin￾denberger, & Staudinger, 2006). Because there exist interindi￾vidual differences in intraindividual change, individuals may increase, decrease, fluctuate, or remain stable in their intrain￾dividual development (Alwin, 1994; Baltes, Reese, & Nessel￾roade, 1977; Mroczek & Spiro, 2003). Thus, individuals can differ with respect to the degree and direction of change (Mroc￾zek & Spiro, 2005). We therefore assume that individual dif￾ferences in change reveal substantial information that are rele￾vant for later life outcomes because they reflect more or less successful adaptation to age-specific developmental tasks. Indeed, adolescence has been characterized as a transitional period and a time of heightened self-exploration that lays the foundation for later outcomes (Erikson, 1968). Furthermore, it is often described as a vulnerable age period and thus a time of “increased risk for the onset of a wide range of emotional and behavioral problems, including depression” (Steinberg, 2005, p. 69). Hence, due to the magnitude of change and the high number of challenges adolescents face, the adolescent years must be re￾garded not only as a transitional but also as a sensitive period that may lead to long-term consequences well beyond the adolescent years. In line with theory on adolescent identity formation as a prerequisite for later life adjustments, adolescents who are not able to process age-specific challenges adequately might be more prone to later health issues such as depressive symptoms in adulthood compared with individuals who develop a positive attitude toward themselves. Thus, although a positive prerequisite, we assume that it is not necessary to initially possess high self-esteem baseline level at the 1 Although the vulnerability model assumes that low self-esteem is a cause of depression, it is also possible that it is a consequence of depressive symptoms (see Rohde, Lewinsohn, & Seeley, 1990). This alternative model, typically referred to as the “scar” model, could not be tested in the present study because self-esteem was not assessed in adulthood. However, the scar model has been extensively tested in other studies, and the prospective effect of depression on self-esteem is typically small or non￾existent (Sowislo & Orth, 2013). This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. 326 STEIGER, ALLEMAND, ROBINS, AND FEND