Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2014 May 1.
Published in final edited form as: J Abnorm Psychol. 2013 May;122(2):353–358. doi: 10.1037/a0032655

Is Liability to Recurrent Major Depressive Disorder Present Before First Episode Onset in Adolescence or Acquired After the Initial Episode?

Jeremy W Pettit 1, Chelsey Hartley 2, Peter M Lewinsohn 3, John R Seeley 4, Daniel N Klein 5
PMCID: PMC3777803  NIHMSID: NIHMS513239  PMID: 23713498

Abstract

Many individuals who experience a major depressive episode will subsequently develop recurrent episodes. Although numerous studies have investigated predictors of recurrent episodes, methodological limitations have made it difficult to determine the extent to which liability to recurrent major depressive disorder (rMDD) exists prior to first onset or develops after first onset. This study used a prospective design in a community sample of adolescents to examine variables before and after first onset MDD as predictors of rMDD over a 12 year follow-up. Among 59 adolescents who experienced first onset MDD, 72.88% developed rMDD during the follow-up period. Parental history of rMDD and lifetime history of minor depression prior to MDD onset significantly predicted rMDD. These two effects replicated in ancillary analyses in an expanded sample of N=205. Following MDD onset, a higher number of major life events significantly predicted rMDD. Liability to rMDD exists prior to MDD onset in the form of familial risk and less severe mood disturbances, whereas liability to rMDD in the form of elevated stress may develop following a first onset in adolescence.

Keywords: depression, onset, recurrence, risk factors, adolescence

An extensive literature documents the recurrent course of major depressive disorder (MDD), with recurrence defined as a new episode of MDD after full recovery from a prior episode (Frank et al., 1991). Estimates of recurrent episodes range from 35–85% depending on sample characteristics and follow-up length (e.g., Curry et al., 2011; Eaton et al., 2008; Mueller et al., 1999; Pettit, Lewinsohn, & Joiner, 2006). Theoretical and empirical work on recurrent MDD (rMDD) typically emphasizes stable liability models or scar models (Burcusa & Iacono, 2007; Monroe & Harkness, 2005; Wichers, Geschwind, van Os, & Peeters, 2010). Stable liability models assert that a trait-like susceptibility to depression exists prior to a first lifetime major depressive episode (FLED; Monroe & Harkness, 2011) and remains present after remission of the FLED. As long as the liability remains present, individuals remain susceptible to developing episodes of depression. Scar models posit that the experience of a FLED creates a liability that did not exist prior to the episode, which in turn increases risk for future episodes. It is possible that stable liabilities and scars each contribute to rMDD.

Numerous investigations have examined predictors of recurrent depressive episodes (see Burcusa & Iacono, 2007; Monroe & Harkness, 2005). Findings have been mostly unsupportive of scar models (e.g., Beevers, Rohde, Stice, & Nolen-Hoeksema, 2007; Ormel, Oldehinkel, & Vollebergh, 2004; Rohde, Lewinsohn, & Seeley, 1990), whereas findings that heritability estimates for rMDD exceed those for a single lifetime episode of depression (SLED; Monroe & Harkness, 2011) are consistent with a stable genetic liability to rMDD (e.g., Gershon, Weissman, Guroff, Prusoff, & Leckman, 1986; Kendler, Gardner, & Prescott, 1999). In a recent review, however, Monroe and Harkness (2011) delineated two primary methodological limitations that have impeded the field’s ability to test predictors of rMDD, and by extension, to draw conclusions about whether liability to rMDD exists prior to FLED onset or develops after FLED onset.

The first limitation is the failure to distinguish between variables that predict rMDD versus a SLED and variables that predict any recurrent episode of depression. In many studies, participants were assessed during or following an index episode and then tracked to examine which variables predicted a recurrent episode. Regrettably, no distinction was made between individuals whose index episode was a FLED and individuals whose index episode was itself a recurrence. It is possible that predictors of the transition from FLED to rMDD differ from predictors of one recurrent episode to another recurrent episode (Monroe & Harkness, 2011).

The second limitation is the failure to assess predictors of rMDD prior to onset of a FLED. To our knowledge, only one study has addressed this limitation (Eaton, et al., 2008). In a study of 92 adults from the Baltimore Epidemiologic Catchment Area Followup who experienced FLED onset during the follow-up, an early onset age was the only significant predictor of rMDD. Unfortunately, the breadth of risk factors investigated was minimal because identifying predictors of rMDD was not the main focus of the study.

Thus, in spite of numerous investigations on predictors of recurrent depression, two important questions remain unanswered. First, does liability to rMDD exist before onset of a FLED? Second, is liability to rMDD acquired following onset of a FLED? Obtaining empirical data to answer these two questions may shed light on the identification of individuals at risk for rMDD and inform prevention strategies to alter that risk.

The purpose of the present study was to address these two questions with empirical data from the Oregon Adolescent Depression Project (OADP), a large, prospective study of psychological disorders among school-based adolescents. Potential predictors of rMDD were selected based on empirical research linking them to recurrent depression (Bockting, Spinhoven, Koeter, Wouters, & Schene, 2006; Dunn & Goodyer, 2006; Lewinsohn, Allen, Seeley, & Gotlib, 1999; Lewinsohn, Rohde, Seeley, Klein, & Gotlib, 2000; Mueller, et al., 1999). Based on evidence that heritability estimates for rMDD exceed those for a SLED, we hypothesized that parental history of rMDD would significantly predict participant rMDD. Such a pattern would be consistent with a stable liability model, in which some adolescents are at high familial risk for rMDD even prior to FLED onset. Because evidence in support of scar models has been weak (Wichers, et al., 2010), we did not expect liability variables following FLED onset to significantly predict rMDD.

Method

OADP participants were randomly selected from nine high schools in Oregon. A total of 1709 adolescents (mean age=16.6) completed an initial evaluation (T1). One year later, 1507 (88%) returned for a second evaluation (T2). Differences between participants and those who declined to participate or dropped out of the study before T2 were small (Lewinsohn, Hops, Roberts, Seeley, & Andrews, 1993). At age 24, all participants with a history of psychopathology by T2 (n=644) and a random sample with no history of psychopathology (n=457) were invited to a third (T3) evaluation. Of 1101 T2 participants selected for T3, 941 (85%) completed the evaluation. At age 30, all T3 participants were invited to a T4 evaluation; 816 (87%) completed the interview.

Near the time of the T3 evaluation, lifetime psychopathology in participants’ biological parents was assessed. Of 941 participants with T3 data, parental diagnostic data were available for 803 (85%).

For primary analyses, the sample included 59 participants with no T1 lifetime history of MDD who experienced FLED onset between T1–T2 and remained in the study through T4. The latter requirement ensured follow-up of sufficient length to classify participants as SLED or rMDD. On average, follow-up from T2 until T4 was 12.43 years (SD=1.09) and time from FLED offset to T4 was 12.54 years (SD=1.66). An additional 15 participants experienced FLED onset between T1–T2 but dropped out of the study prior to T4, and parental diagnostic data were not available for five participants (8.5%). No significant differences on measured variables were found between those who completed T4 and those who did not, or between those with and without parental diagnostic data (ps>.10).

Ancillary analysis on T1 predictor variables were run among 205 participants with no T1 lifetime history of MDD who experienced FLED onset between T1–T3. This approach provided greater statistical power to detect effects for a stable liability model. Among these 205, 59 participants described above had FLED onset between T1–T2 and an additional 146 participants had FLED onset between T2–T3. Mean time from FLED offset to T4 was 9.46 years (SD=2.77). We were unable to examine predictor variables measured after FLED onset among these 205 participants because the assessment battery was changed substantially after T2.

Diagnostic interviews

Participants were interviewed with the Schedule for Affective Disorders and Schizophrenia for School-Age Children (K-SADS; Orvaschel, Puigantich, Chambers, Tabrizi, & Johnson, 1982) at T1, T2, and T3. In conjunction with the K-SADS, the Longitudinal Interval Followup Evaluation (LIFE; Keller et al., 1987) was used to evaluate the presence and course of disorders since the previous interview. T4 diagnoses were based on the Structured Clinical Interview for Axis I DSM-IV Disorders–Non-Patient Edition (SCID-NP; First, Spitzer, Gibbon, & Williams, 1995), with the LIFE also used to evaluate disorder presence and course since T3.

Interviewers had advanced degrees in a mental health field, completed a 70-hour course in diagnostic interviewing, and completed a minimum of two supervised training interviews. Independent review of randomly selected cases indicated good-excellent interrater reliability for diagnoses used in this study (Lewinsohn, et al., 1993).

T1 diagnostic status

Dichotomous variables represented lifetime diagnosis of any anxiety disorder, alcohol or substance use disorder, disruptive behavior disorder, and minor depressive disorder (mDep).

T1– T2 FLED clinical characteristics

FLED duration and number of DSM criteria symptoms present during the FLED were obtained from the K-SADS/LIFE. Duration was defined as the length of time from first meeting MDE criteria to the point at which episode recovery criteria were first met (see below). A dichotomous variable represented mental health treatment utilization for the FLED. Any psychosocial or pharmacological treatment for the FLED was coded as positive for mental health treatment utilization.

MDD recurrence

A dichotomous variable represented the recurrence of MDD by T4. Recurrence was operationalized as a new episode of MDD after full recovery (i.e., a minimum of eight consecutive weeks with no more than 1–2 mild depressive symptoms).

Parental diagnosis

Effort was made to obtain two sources of diagnostic data for each parent: 76% of parents were directly interviewed using the SCID-NP; data via informants (i.e., participants and/or another first degree relative) were collected on all parents using the Family Informant Schedule and Criteria (FISC; Mannuzza & Fyer, 1990), supplemented with items necessary to derive DSM-IV diagnoses. Lifetime best-estimate DSM-IV diagnoses were derived independently by two senior diagnosticians from all available information. Interrater reliabilities of diagnoses from direct and family history interviews and best-estimates were good-excellent (Klein, Lewinsohn, Seeley, & Rohde, 2001). Five dichotomous (0=neither parent met criteria; 1=one or both parents met criteria) variables represented parental lifetime diagnostic history: MDD, SLED, rMDD, anxiety disorder, and substance use disorder.

T1 and T2 psychosocial measures

The following psychosocial measures with established reliability and validity in the OADP (see Andrews, Lewinsohn, Hops, & Roberts, 1993; Mathew, Pettit, Lewinsohn, Seeley, & Roberts, 2011) were assessed at T1 and again at T2.

Academic problems

This construct was assessed with seven items (α = .68), such as “Were you satisfied with your grades?”

Coping

Coping skills assessed the ways in which individuals cope with stressful situations, using 17 items (α= .76).

Dysfunctional attitudes

Intensity of dysfunctional attitudes was assessed with nine items (α= .74) from the Dysfunctional Attitudes Scale (Weissman & Beck, 1978).

Family social support

This construct was measured with eight items (α=.77), such as “How well do you get along with your parents?”

Friend social support

This construct was assessed with seven items (α=.72), such as “How many close friends do you have?”

Health problems

This construct was assessed with six items (α = .61), such as “In the past year, have you been unable to work and/or participate in school activities because of some illness or injury?”

Interpersonal dependency

The extent to which participants desired more support and approval from others, were anxious about being alone or abandoned, and were interpersonally sensitive was assessed with the 10-item emotional reliance subscale of the Interpersonal Dependency Inventory (Hirschfeld et al., 1977) (α= .83).

Major Life Events

This measure assessed the occurrence of 14 low frequency major life events in the preceding year.

Analytic Strategy

Separate logistic regression models were used to examine the association between each predictor and rMDD. To test whether liability to rMDD existed prior to FLED onset, T1 participant variables and parental history of psychopathology were examined as predictors of recurrence status. To test whether liability to rMDD developed after FLED onset, T2 participant variables were examined as predictors of recurrence status. Because the statistical power available to detect effects for T2 participant variables was low, we focused on examination of effect sizes in addition to tests of statistical significance. Full Information Maximum Likelihood was used to estimate missing values for the five cases without parental diagnostic data.

Results

All 59 participants experienced remission of the FLED. Mean FLED duration was 21.44 weeks. Two participants (3.39%) met DSM-IV criteria for the chronic episode specifier. Forty-three participants (72.88%) experienced a recurrence and were classified as rMDD: 25 experienced two episodes, eight experienced three episodes, and 10 experienced 4–7 episodes. Mean length of time from FLED offset to onset of a recurrent episode was 60.09 months (SD=42.28). The other 16 participants (27.12%) did not experience a recurrence and were classified as SLED. Mean length of time from FLED offset to T4 evaluation did not significantly differ across SLED and rMDD cases, t(58)=0.91, p=ns.

Is liability to rMDD present prior to a FLED onset?

The effects of T1 variables on recurrence status are presented in Table 1. The odds of rMDD were significantly higher among participants with a T1 history of mDep and among participants with a parental history of rMDD. No other variable was a significant predictor of recurrence status. When mDep and parental rMDD were entered simultaneously as predictors of rMDD, the effect of parental rMDD remained significant, odds ratio (OR)=4.80, 95% CI=1.03, 22.35, p=.04, and the effect of mDep decreased to a trend level, OR=6.88, 95% CI=0.78, 60.80, p=.08.

Table 1.

T1 Predictors of Recurrence

Predictors SLED (n=16) rMDD (n=43) Odds Ratio (95% CI) Nagelkerke R2
Female Gender 68.75% 67.44% 1.06 (0.31, 3.61) .001
Lifetime Psychological Disorder
 Anxiety Disorder 12.5% 9.30% 0.72 (0.12, 4.36) .003
 Substance Use Disorder 0.0% 2.33% -- .016
 Disruptive Behavior Disorder 12.50% 4.65% 0.34 (0.04, 2.66) .025
 mDep 6.25% 37.21% 8.89 (1.07, 73.80) .154
Psychosocial Variablesa
 Family Support 9.74 (6.81) 8.18 (6.99) 0.97 (0.89, 1.05) .014
 Friend Support 30.27 (3.60) 29.84 (3.84) 0.97 (0.83, 1.13) .004
 Health Problems 1.00 (1.25) 1.33 (1.40) 1.21 (0.75, 1.95) .017
 Academic Impairment 7.88 (2.06) 8.86 (3.35) 1.12 (0.92, 1.37) .031
 Major Event Stress 1.00 (1.37) 1.21 (1.26) 1.15 (0.71, 1.84) .008
 Coping Skills 49.31 (9.49) 45.79 (9.25) 0.96 (0.90, 1.02) .041
 Dysfunctional Attitudes 31.94 (4.09) 30.91 (6.85) 0.97 (.089, 1.07) .008
 Interpersonal Dependency 25.88 (7.47) 23.60 (6.65) 0.95 (0.87, 1.04) .031
Parental Lifetime
Psychological Disorder
 MDD 40.0% 69.23% 3.38 (0.98, 11.62)† .100
 SLED 26.67% 30.77% 1.22 (0.32, 4.63) .002
 rMDD 13.33% 48.71% 6.18 (1.23, 31.07) .160
 Anxiety Disorder 31.25% 38.71% 0.85 (0.24, 2.97) .002
 Substance Use Disorder 56.25% 51.16% 0.82 (0.26, 2.59) .003
Open in a new tab

Note. N=59.

a

Mean scores (standard deviations) are presented for Psychosocial Variables. SLED=single lifetime major depressive episode; rMDD=recurrent major depressive disorder; mDep=minor depression. Bold font indicates statistical significance at p<.05;

†

p=.054.

To examine whether the combination of mDep and parental rMDD predicted the development of rMDD with higher accuracy than either risk factor alone, predictive power was calculated using an “and” decision rule and an “or” decision rule with these two risk factors. The former indicates that only adolescents who met both criteria were predicted to develop rMDD; the latter indicates that only one of the two criteria must be met. As shown in Table 2, the “and” rule produced perfect specificity and positive predictive value, but low sensitivity, negative predictive value, and overall accuracy. The “or” rule also produced high specificity and positive predictive value, and moderate sensitivity, negative predictive value, and overall accuracy. Overall accuracy of the “or” rule exceeded that of either predictor alone and the “and” rule.

Table 2.

Screening Properties for T1 Risk Factor Combinations

Predictor of rMDD Sensitivity Specificity PPV NPV Overall accuracy
mDep .37 .94 .94 .36 .43
Parental rMDD .44 .88 .90 .37 .56
mDep OR parental rMDD .63 .81 .90 .45 .67
mDep AND parental rMDD .19 1.00 1.00 .31 .41
Open in a new tab

Note. mDep=minor depression; rMDD=recurrent major depressive disorder; PPV=positive predictive value; NPV=negative predictive value.

Ancillary analyses

Analyses were repeated among 205 participants who experienced FLED onset between T1 and T3. Mean FLED duration was 21.64 weeks and all participants experienced remission. Eight participants (3.90%) met DSM-IV criteria for the chronic episode specifier. One-hundred fifteen participants (56.09%) experienced a recurrence and were classified as rMDD: 75 experienced two episodes, 22 experienced three episodes, and 18 experienced 4–7 episodes. Mean length of time from FLED offset to onset of a recurrent episode was 57.70 months (SD=35.83).

Results from logistic regression models among the 205 participants were identical to those among the 59 participants in terms of statistical significance: mDep (OR=2.89, 95% CI=1.29, 6.51, p<.01) and parental rMDD (OR=1.85, 95% CI=1.06, 3.24, p<.05) were the only variables that significantly predicted rMDD. When mDep and parental rMDD were entered simultaneously as predictors of rMDD, the effect of mDep remained significant, OR=2.84, 95% CI=1.20, 6.76, p=.02, and the effect of parental rMDD decreased to a trend level, OR=1.72, 95% CI=0.98, 3.02, p=.06.

Is liability to rMDD acquired after a FLED onset?

The effects of T2 variables on recurrence status are presented in Table 3. Seventeen of 59 participants (28.80%) were currently in the FLED at the time of the T2 evaluation. Current T2 FLED status did not significantly predict recurrence status, OR=1.30, 95% CI=0.35, 4.79, p=.69.

Table 3.

T2 Predictors of Recurrence

Predictors SLED (n=16) rMDD (n=43) Odds Ratio (95% CI) Nagelkerke R2
T2 Psychosocial Measures
 Family Support 13.05 (8.94) 8.59 (7.40) 0.93 (0.86, 1.01) .084
 Friend Support 28.93 (3.49) 30.17 (3.85) 1.09 (0.93, 1.29) .031
 Health Problems 1.27 (1.01) 1.37 (1.02) 1.10 (0.56, 2.17) .002
 Academic Impairment 8.93 (4.04) 9.13 (3.43) 1.02 (0.85, 1.20) .001
 Major Event Stress 0.80 (0.94) 1.72 (1.67) 1.66 (1.01, 2.76) .111
 Coping Skills 48.88 (8.41) 46.60 (7.96) 0.93 (0.86, 1.01) .082
 Dysfunctional Attitudes 32.50 (5.69) 31.84 (7.52) 0.99 (0.91, 1.07) .003
 Interpersonal Dependency 24.38 (5.48) 24.58 (7.33) 1.01 (0.92, 1.09) .001
FLED Characteristics
 Duration in Weeks 14.69 (21.67) 6.79 (5.71) 1.00 (0.99, 1.01) .012
 Number of DSM Symptoms 5.56 (2.03) 6.24 (1.25) 1.35 (0.91, 2.02) .057
 Treatment Utilization 12.50% 25.58% 2.41 (0.47, 12.31) .022
Open in a new tab

Note. N=59. Mean scores (standard deviations) are presented for all variables except Treatment Utilization. Percentages are presented for Treatment Utilization. SLED=single lifetime major depressive episode; rMDD=recurrent major depressive disorder; FLED=first lifetime major depressive episode. Bold font indicates statistical significance at p<.05.

Inspection of ORs in Table 2 revealed consistently small effect sizes, with the exception of a medium yet nonsignificant effect of treatment utilization. Findings did not differ when psychological treatment and pharmacological treatment were examined separately. Major event stress was the only T2 variable that significantly predicted recurrence status. To determine whether the significant effect was accounted for by current depression, we included T2 MDD status as a covariate; the predictive effect of major event stress remained statistically significant, OR=1.65, 95% CI=1.00, 2.75, p < .05. When T2 major event stress, mDep, and parental rMDD were all entered simultaneously as predictors, none was a statistically significant predictor of rMDD (all ps>.09).

Discussion

This study examined whether liability variables for rMDD could be identified prior to and shortly following onset of a FLED in adolescence. Over 70% of 59 adolescents who experienced a FLED developed rMDD over a 12 year follow-up. The rate of rMDD was 56% among 205 participants in ancillary analyses using a shorter follow-up period (mean=9 years). These recurrence rates are similar to those reported in some previous investigations that used follow-ups of a decade or more (Fombonne, Wostear, Cooper, Harrington, & Rutter, 2001; Mueller, et al., 1999), but are considerably higher than the 35% rate reported in the only other study that prospectively tracked participants in a FLED over a lengthy period (Eaton, et al., 2008). The higher rate found in the present study may be due to the young age at which participants experienced a FLED, as an early onset of MDD may portend a more recurrent course (Eaton, et al., 2008; Pettit, Lewinsohn, Roberts, Seeley, & Monteith, 2009).

Prior history of mDep and parental history of rMDD were the only variables that significantly predicted development of rMDD. This was the case in primary analyses and in ancillary analyses with an expanded sample. The finding that adolescents with a past history of mDep were at elevated odds of rMDD suggests a chronic life course of mood disturbance, beginning with subthreshold depressive symptoms before mid-adolescence, escalating to syndromal FLED during adolescence, and proceeding to rMDD by early adulthood. This is consistent with a model of stable liability to experience mood disturbances of varying intensity, although the possibility that the experience of mDep left a “scar” that increased odds of rMDD cannot be ruled out (Wichers, et al., 2010).

The finding that a positive parental history of rMDD was associated with elevated odds of offspring rMDD is consistent with a stable familial risk liability model (Gershon, et al., 1986; Kendler, et al., 1999). A previous study (Eaton, et al., 2008) did not find a significant association between parental depression and offspring rMDD. However, parental rMDD was not assessed. The present findings indicate that odds of offspring rMDD are elevated only in the presence of parental rMDD, not parental SLED.

Examination of combinations of mDep and parental rMDD using “and” and “or” decision rules revealed that both rules led to high specificity and positive predictive power, but modest to low sensitivity and negative predictive power. In addition, multivariate models indicated that parental rMDD and mDep remained significant or trend level predictors of rMDD when entered jointly. As such, the presence of either mDep or parental history of rMDD provides a useful index for clinicians and researchers to identify adolescents in a FLED who are at very high risk for rMDD.

After adolescents had experienced a FLED, the number of major life events in the past year was the only significant predictor of rMDD. The finding that a higher number of major life events after, but not before, the onset of a FLED significantly predicted rMDD is consistent with the possibility of a scarring effect of a FLED.

This study’s findings should be viewed in light of limitations. The small sample constrained statistical power available to detect effects for predictors of rMDD measured after FLED onset. Given the early age of MDD onset, caution should be used in generalizing the findings beyond adolescent onset MDD. It is possible that some participants classified as SLED will subsequently transition to rMDD. It also is possible that recurrent episodes may have been missed due to memory bias given the six-year intervals between T2, T3, and T4. Finally, several of the psychosocial measures have relatively low temporal stability and their impact on depression may be constrained to short time intervals (e.g., the impact of major events on depression is strongest in the first few months following events). Future research is encouraged to include variables that may have more distal, enduring effects.

In summary, lifetime history of mDep and parental history of rMDD predicted the development of rMDD among adolescents who had not yet experienced a major depressive episode. Following onset of a first major depressive episode, a higher level of major life events predicted the development of rMDD. These findings suggest some liabilities to rMDD may be present before FLED onset and others may be acquired following FLED onset. Among adolescents who are experiencing a FLED, a history of less severe mood disturbances such as mDep or a parental history of rMDD indicate a high risk for recurrence. Such adolescents may represent ideal candidates for maintenance treatments following response to acute phase treatments (Jarrett et al., 1998).

Acknowledgments

This project was supported by National Institute of Health Grants MH40501, MH50522, MH52858, and MH75744.

Contributor Information

Jeremy W. Pettit, Florida International University

Chelsey Hartley, Florida International University.

Peter M. Lewinsohn, Oregon Research Institute

John R. Seeley, Oregon Research Institute

Daniel N. Klein, Stony Brook University

References

  1. Andrews JA, Lewinsohn PM, Hops H, Roberts RE. Psychometric properties of scales for the measurement of psychosocial variables associated with depression in adolescence. Psychological Reports. 1993;73(3):1019–1046. doi: 10.2466/pr0.1993.73.3.1019. [DOI] [PubMed] [Google Scholar]
  2. Beevers CG, Rohde P, Stice E, Nolen-Hoeksema S. Recovery from major depressive disorder among female adolescents: A prospective test of the scar hypothesis. Journal of Consulting and Clinical Psychology. 2007;75(6):888–900. doi: 10.1037/0022-006x.75.6.888. [DOI] [PubMed] [Google Scholar]
  3. Bockting CL, Spinhoven P, Koeter MW, Wouters LF, Schene AH. Prediction of recurrence in recurrent depression and the influence of consecutive episodes on vulnerability for depression: a 2-year prospective study. Journal of Clinical Psychiatry. 2006;67(5):747–755. [PubMed] [Google Scholar]
  4. Burcusa SL, Iacono WG. Risk for recurrence in depression. Clinical Psychology Review. 2007;27(8):959–985. doi: 10.1016/j.cpr.2007.02.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Curry J, Silva S, Rohde P, Ginsburg G, Kratochvil C, Simons A, March J. Recovery and recurrence following treatment for adolescent major depression. Archives of General Psychiatry. 2011;68(3):263–269. doi: 10.1001/archgenpsychiatry.2010.150. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Dunn V, Goodyer IM. Longitudinal investigation into childhood- and adolescence-onset depression: psychiatric outcome in early adulthood. British Journal of Psychiatry. 2006;188:216–222. doi: 10.1192/bjp.188.3.216. [DOI] [PubMed] [Google Scholar]
  7. Eaton WW, Shao H, Nestadt G, Lee HB, Bienvenu OJ, Zandi P. Population-based study of first onset and chronicity in major depressive disorder. Archives of General Psychiatry. 2008;65(5):513–520. doi: 10.1001/archpsyc.65.5.513. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. First MB, Spitzer RL, Gibbon M, Williams JBW. Structured Clinical Interview for DSM-IV Axis I Disorders – Non-Patient Edition (SCID-I/NP, Version 2.0) Biometrics Research Department, New York State Psychiatric Institute; New York: 1995. [Google Scholar]
  9. Fombonne E, Wostear G, Cooper V, Harrington R, Rutter M. The Maudsley long-term follow-up of child and adolescent depression. 1. Psychiatric outcomes in adulthood. British Journal of Psychiatry. 2001;179:210–217. doi: 10.1192/bjp.179.3.210. [DOI] [PubMed] [Google Scholar]
  10. Frank E, Prien RF, Jarrett RB, Keller MB, Kupfer DJ, Lavori PW, Weissman MM. Conceptualization and rationale for consensus definitions of terms in major depressive disorder. Remission, recovery, relapse, and recurrence. Archives of General Psychiatry. 1991;48(9):851–855. doi: 10.1001/archpsyc.1991.01810330075011. [DOI] [PubMed] [Google Scholar]
  11. Gershon ES, Weissman MM, Guroff JJ, Prusoff BA, Leckman JF. Validation of criteria for major depression through controlled family study. Journal of Affective Disorders. 1986;11(2):125–131. doi: 10.1016/0165-0327(86)90018-2. [DOI] [PubMed] [Google Scholar]
  12. Hirschfeld RM, Klerman GL, Gough HG, Barrett J, Korchin SJ, Chodoff P. A measure of interpersonal dependency. J Pers Assess. 1977;41(6):610–618. doi: 10.1207/s15327752jpa4106_6. [DOI] [PubMed] [Google Scholar]
  13. Jarrett RB, Basco MR, Risser R, Ramanan J, Marwill M, Kraft D, Rush AJ. Is there a role for continuation phase cognitive therapy for depressed outpatients? Journal of Consulting and Clinical Psychology. 1998;66(6):1036–1040. doi: 10.1037//0022-006x.66.6.1036. [DOI] [PubMed] [Google Scholar]
  14. Keller MB, Lavori PW, Friedman B, Nielsen E, Endicott J, McDonald-Scott P, Andreasen NC. The Longitudinal Interval Follow-up Evaluation. A comprehensive method for assessing outcome in prospective longitudinal studies. Arch Gen Psychiatry. 1987;44(6):540–548. doi: 10.1001/archpsyc.1987.01800180050009. [DOI] [PubMed] [Google Scholar]
  15. Kendler KS, Gardner CO, Prescott CA. Clinical characteristics of major depression that predict risk of depression in relatives. Archives of General Psychiatry. 1999;56(4):322–327. doi: 10.1001/archpsyc.56.4.322. [DOI] [PubMed] [Google Scholar]
  16. Klein DN, Lewinsohn PM, Seeley JR, Rohde P. A family study of major depressive disorder in a community sample of adolescents. Arch Gen Psychiatry. 2001;58(1):13–20. doi: 10.1001/archpsyc.58.1.13. yoa20078 [pii] [DOI] [PubMed] [Google Scholar]
  17. Lewinsohn PM, Allen NB, Seeley JR, Gotlib IH. First onset versus recurrence of depression: differential processes of psychosocial risk. Journal of Abnormal Psychology. 1999;108(3):483–489. doi: 10.1037//0021-843x.108.3.483. [DOI] [PubMed] [Google Scholar]
  18. Lewinsohn PM, Hops H, Roberts RE, Seeley JR, Andrews JA. Adolescent psychopathology: I. Prevalence and incidence of depression and other DSM-III-R disorders in high school students. Journal of Abnormal Psychology. 1993;102(1):133–144. doi: 10.1037//0021-843x.102.1.133. [DOI] [PubMed] [Google Scholar]
  19. Lewinsohn PM, Rohde P, Seeley JR, Klein DN, Gotlib IH. Natural course of adolescent major depressive disorder in a community sample: predictors of recurrence in young adults. American Journal of Psychiatry. 2000;157(10):1584–1591. doi: 10.1176/appi.ajp.157.10.1584. [DOI] [PubMed] [Google Scholar]
  20. Mannuzza S, Fyer AJ. Family Informant Schedule and Criteria (FISC), July 1990 Revision. New York: Anxiety Disorders Clinic, New York State Psychiatric Institute; 1990. [Google Scholar]
  21. Mathew AR, Pettit JW, Lewinsohn PM, Seeley JR, Roberts RE. Co-morbidity between major depressive disorder and anxiety disorders: shared etiology or direct causation? Psychological Medicine. 2011;41(10):2023–2034. doi: 10.1017/S0033291711000407. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Monroe SM, Harkness KL. Life stress, the “kindling” hypothesis, and the recurrence of depression: considerations from a life stress perspective. Psychological Review. 2005;112(2):417–445. doi: 10.1037/0033-295X.112.2.417. [DOI] [PubMed] [Google Scholar]
  23. Monroe SM, Harkness KL. Recurrence in major depression: a conceptual analysis. Psychological Review. 2011;118(4):655–674. doi: 10.1037/a0025190. [DOI] [PubMed] [Google Scholar]
  24. Mueller TI, Leon AC, Keller MB, Solomon DA, Endicott J, Coryell W, Maser JD. Recurrence after recovery from major depressive disorder during 15 years of observational follow-up. American Journal of Psychiatry. 1999;156(7):1000–1006. doi: 10.1176/ajp.156.7.1000. [DOI] [PubMed] [Google Scholar]
  25. Ormel J, Oldehinkel AJ, Vollebergh W. Vulnerability before, during, and after a major depressive episode - A 3-wave population-based study. Archives of General Psychiatry. 2004;61(10):990–996. doi: 10.1001/archpsyc.61.10.990. [DOI] [PubMed] [Google Scholar]
  26. Orvaschel H, Puigantich J, Chambers W, Tabrizi MA, Johnson R. Retrospective assessment of prepubertal major depression with the Kiddie-Sads-E. Journal of the American Academy of Child and Adolescent Psychiatry. 1982;21(4):392–397. doi: 10.1016/S0002-7138(09)60944-4. [DOI] [PubMed] [Google Scholar]
  27. Pettit JW, Lewinsohn PM, Joiner TE., Jr Propagation of major depressive disorder: relationship between first episode symptoms and recurrence. Psychiatry Research. 2006;141(3):271–278. doi: 10.1016/j.psychres.2005.07.022. [DOI] [PubMed] [Google Scholar]
  28. Pettit JW, Lewinsohn PM, Roberts RE, Seeley JR, Monteith L. The long-term course of depression: development of an empirical index and identification of early adult outcomes. Psychological Medicine. 2009;39(3):403–412. doi: 10.1017/S0033291708003851. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Rohde P, Lewinsohn PM, Seeley JR. Are people changed by the experience of having an episode of depression - a further test of the scar hypothesis. Journal of Abnormal Psychology. 1990;99(3):264–271. doi: 10.1037//0021-843x.99.3.264. [DOI] [PubMed] [Google Scholar]
  30. Weissman AN, Beck AT. Development and validation of the Dysfunctional Attitude Scale. Paper presented at the Annual meeting of the American Educational Research Association; Toronto. 1978. [Google Scholar]
  31. Wichers M, Geschwind N, van Os J, Peeters F. Scars in depression: is a conceptual shift necessary to solve the puzzle? Psychological Medicine. 2010;40(3):359–365. doi: 10.1017/s0033291709990420. [DOI] [PubMed] [Google Scholar]

RESOURCES