Early adolescence has been referred to as a stage of transition from childhood into the increasingly

1
Stability of Depressed Mood in Early
Adolescence A Longitudinal Study.Lynae A.
Johnsen, M.A., Susan M. Swearer, Ph.D., Jami E.
Givens, B.A.The University of Nebraska-Lincoln

• INSTRUMENT
• The Childrens Depression Inventory (CDI
  Kovacs, 1992). This instrument is the most
  commonly used self-report measure of depression
  for children 7 to 17 years of age. The CDI
  consists of 27 items designed to assess the overt
  symptoms of childhood depression. The CDI
  measures five highly-correlated factors
  Negative Mood, Interpersonal Problems,
  Ineffectiveness, Anhedonia, and Negative
  Self-Esteem. These five factors are combined to
  yield one higher-order factor of childhood
  depression. Participants are asked to rate the
  severity of each item on a three-point scale of 0
  to 2 during the two weeks prior to testing.
  Total scores of 19 or greater are considered to
  indicate potential depression (Stark, 1990). The
  CDI has demonstrated acceptable internal
  consistency and test-retest as well as convergent
  validity (Kovacs, 1992).
• PROCEDURES
• Active parental consent and youth assents were
  obtained for each participant.
• All students with parental consent to
  participate in the study were administered the
  CDI
• during 3 consecutive years during the Spring of
  1999 to the Spring of 2003.
• These data are from of a five-year longitudinal
  study.
• RESULTS
• Results show that normal levels of depressed
  mood in early adolescence appears to be
• stable from Time 1 to Time 3.
• Due to the limited use of advanced statistical
  methods used in previous longitudinal
• studies examining stability of depressed mood
  over time, hierarchal linear modeling
  
• (HLM) was used to analyze the data.
• HLM is a recently developed statistical model
  that examines change of individual
• growth trajectories over time (Raudenbush
  Bryk, 2002).

• INTRODUCTION
• Early adolescence has been referred to as a stage
  of transition from childhood into the
  increasingly complex time of adolescence (Heath
  Camarena, 2002).
• Normative stressors associated with the
  transition from elementary to middle school add
  to the heightened risk for the development of
  depressed mood during adolescence (Clarizio,
  1994).
• Problems in meeting basic developmental
  challenges during adolescence have been
  empirically linked to chronically depressed mood
  (Petersen, Compas, Brooks-Gunn, 1992).
• Less than 3 of preadolescent children experience
  major depressive disorder, whereas 6.4 of
  adolescents present with clinical levels of
  depression (Fleming Offord, 1990).
• Stability, or persistence in symptomatology has
  been described as one of the three criteria which
  permit clinicians to distinguish between a
  depressive mood state and an isolated depressive
  episode (Rutter, 1988).
• Research suggests that incidence of depressed
  mood increases across adolescence, with many
  adolescents reporting symptoms of depression at
  some point during that period of time (Fleming
  Offord, 1990).
• Researchers have found that formerly depressed
  adolescents who developed psychiatric problems
  during young adulthood were characterized during
  adolescence by more severe depressive episodes
  (e.g., longer episode duration, multiple
  episodes, greater number of symptoms Lewinsohn,
  Rohde, Seeley, Klein, Gotlib, 2000).
• Lewinsohn, Rohde, Klein, Seeley (1999) found
  that major depressive disorder (MDD) in young
  adulthood was significantly more common in an
  adolescent MDD group than nonaffective and no
  disorder groups.
• It is widely recognized that increased knowledge
  of the antecedents and processes underlying
  depressed mood in early life may contribute to an
  improved understanding of the nature and course
  of depression in later life (Lewinsohn et al.,
  1997).
• Several longitudinal studies have examined
  stability of depressed mood across varying units
  of time in adolescence and have found relatively
  stable trends in depression scores (Devine,
  Kempton, Forehand, 1994 Dubois, Felner,
  Bartels, Silverman, 1995 Holsen, Kraft,
  Vitterso, 2000 Heath Camarena, 2002).
• In contrast, other researchers have concluded
  that depressive symptoms in adolescents seem to
  show limited over-time stability (Garrison et
  al., 1990).
• Previous research on stability of depressed mood
  during adolescence has employed basic statistical
  methods, including examining mean level
  differences and correlations.
• The purpose of this study was to examine the
  stability of depressed mood during early
  adolescence longitudinally using hierarchical
  linear modeling.
• PARTICIPANTS

• Covariance Parameters (see Figures 2, 3).
• Intercept There is significant variability at
  Time 1, meaning not all participants have the
  same initial score of 8.91.
• Slope There is significant variability in the
  growth trajectory across the three time points.
• Covariance Indicates the relationship between
  initial CDI scores and final CDI scores. There is
  a significant negative correlation between
  intercept and slope.
• Students with higher CDI scores initially at
  Time 1 tend to decrease over time. Students with
  lower scores at Time 1 tend to slightly increase
  over time.

• Table 1.
• Estimates of Fixed Effects
• __________________________________________________
  _______________
• Parameter Estimate
  Standard Error Significance
• __________________________________________________
  _______________
• Intercept 8.912
  .510
  .000
• Time -.007
  .239
  .997
• __________________________________________________
  _______________
• Table 2.
• Estimates of Covariance Parameters
• __________________________________________________
  ________________
• Parameter Estimate
  Standard Error Significance
• __________________________________________________
  ________________
• Residual 21.146 1.84
  .000
• Intercept Time
• Intercept 58.545 8.951
  .000
• Covariance -12.471 3.692
  .001
• Slope 5.550 1.840
  .003

Figure 2.
Negative slope, negative slope/intercept
covariance
1 2 3
Time
Figure 3.
Variability of Random Sample Data
Figure 1.
Expected CDI Scores Across Time
Score of 19 or greater is clinically
significant CDI Total Score Range from 0-54.
Poster presented at the Kansas Conference in
Clinical Child and Adolescent Psychology, October
22, 2004, Lawrence, KS.