|
A Pilot Study of a Parent-Education Group for Families Affected by Depression
Procedures
Participants were assessed in their home by trained bachelor’s-level research
assistants. After they obtained informed consent, the research assistants
administered the baseline structured interviews and questionnaires to the
participants (and partners who agreed to participate). Selected outcome
measures were readministered 2 weeks later. Randomization was carried out
after the baseline assessment, by a person independent of the study team.
The procedure employed a random numbers table using blocks of 4 (study
personnel were blind to blocking). The resulting assignments were placed
in numbered, sealed, opaque envelopes that were opened consecutively as
participants completed assessment. The group program was mounted within
4 weeks of the baseline assessment. At completion of the group program,
experimental and control families completed the questionnaires. Eight weeks
later, all participants were again assessed in their home, with measures
identical to those used at posttreatment. At the conclusion of the project,
an independent research assistant obtained the treatment dropouts’ stated
reasons for failure to attend. Treatment groups were run during a 1-year
period by leaders (4 nurses, 1 social worker, and 2 leaders with bachelor’s
degrees in psychology) who received 2 half-days’ training. Training emphasized
close adherence to the manual; however, we did not assess adherence to
the treatment model.
Figure 1 provides the sampling results. Of the 65 patients referred to
the study, 14 refused to participate and 7 were excluded, leaving 44 participants
who were assessed at baseline and randomly assigned to experimental (n
= 21) or control (n = 23) conditions. At posttreatment, 32 participants
(73%) completed assessment (15 treatment and 17 control subjects), while
12 were lost to assessment. By follow-up, 25 participants (57%) completed
assessment, and 19 were missing. Owing to the high loss of participants
at follow-up, we did not use data from this assessment. Among the 21 participants
assigned to the experimental condition, 7 attended fewer than 2 of 8 sessions.
Six parents provided reasons for their failure to attend (2 stated inconvenient
timing or location, 2 were attending groups elsewhere, 1 didn’t feel like
attending, and 1 stated “mind was racing . . . too much on the go”). To
assess whether there was selective loss of participants, we compared those
completing the posttreatment assessment with those missed, both in the
total sample and separately for the experimental and control groups. For
the whole sample, baseline CES-D scores for proband and partner were higher
in those lost to assessment than in those assessed (proband mean 38.5,
SD 12.4 and mean 29.3, SD 13.6, respectively; partner mean 21.5, SD 16.2
and mean 9.3, SD 9.4, respectively) (t = –1.9, df 40, P = 0.07; and t =
–2.3, df 24, P = 0.03, for proband and partner, respectively). There were
more single parents among those missed (58% vs 22% in those completing
the assessment) (c2 = 5.3, df 1, P = 0.03). When we contrasted experimental and control groups, baseline proband CES-D scores were higher in those
lost to assessment than in those assessed in the experimental group (mean
41.0, SD 3.5 vs mean 25.8, SD 13.8) (t = –2.6, df 19, P = 0.02) but not
in the control group (mean 35.5, SD 18.6 vs mean 32.7, SD 12.9) (t = –0.4,
df 19, P = ns). A similar pattern was seen for partner CES-D. There were
more single parents lost at posttreatment in both the experimental and
the control groups. These findings show a selective loss of participants
with higher depression scores in the experimental group, but not in the
control group. Because depressive symptom severity is often associated
with a poor response to treatment, this could lead to bias in favour of
the experimental group (see below).
Figure 1 Flow Chart of Study Participation
Instruments and Measures
Demographics Interview. We interviewed probands and partners for information
on income, education level, employment, family structure, and marital status.
University of Michigan CIDI. We used the short-form CIDI to assess whether
parents met DSM-IV criteria for major depression in the past year or in
their lifetime. In both clinic and community samples, the short-form CIDI
is highly sensitive for major depression but has a high false-positive
rate (27.3%) (20,21).
The CES-D. This 20-item questionnaire measures depressive symptoms. It
is reliable in clinical and population samples, with a = 0.90 and 0.85, respectively, and it discriminates among individuals with depression, normal
individuals, and those with other psychiatric disorders (22). Test–retest
reliability is higher in clinic than in community samples (estimates range
from r = 0.32 to r = 0.67) (23).
Parenting Practices Scale (Modified for Use in the Canadian National Longitudinal
Study of Children and Youth [NLSCY]). This 18-item scale measures parenting
behaviours. Preliminary factor analyses from NLSCY data for the age group
used in this study revealed 3 factors (positive practices, hostile-ineffective
practices, and consistency factors) (24). Therefore, in this study, we
derived subscale scores for these 3 factors.
Sense of Parenting Competence Scale (Modified for Use in NLSCY) (24). This
11-item scale measures respondents’ sense of competence as a parent.
Family Assessment Device (FAD) (25). The 12-item general functioning subscale
of the FAD is a reliable measure of family functioning.
Family Conflict Scale (NLSCY) (24). This 6-item scale measures the level
of perceived conflict within the family.
Parent Disagreements Scale (Child Development Supplement, Panel Study of
Income) (26). This 9-item scale measures agreement between parents with
respect to their performance on frequent parenting practices.
Depression Facts Quiz. This scale assesses knowledge in areas covered by
the educational component of the program. The scale from which it was derived
detected increases in knowledge about depression in parents of adolescents
with depression after they had participated in a psychoeducation program
(27).
Health Behaviours Questionnaire and Activities Scale (Ontario Child Health
Study) (28). This 15-item scale measures the child’s participation in coached
and noncoached activities. There are 3 subscales: competence in sports
and arts; level of participation in sports, clubs, arts, and groups; and
participation in outside activities with adult leadership.
Peer-Relationships and School-Problems Scales. These 10-item scales were
derived from a structured-interview version of the Social Adjustment Inventory
for Children and Adolescents (29). The Peer Relationship Scale covers such
aspects of peer relationships as having problems in maintaining relationships
or having a friend to confide in. The School Problems Scale assesses the
presence or absence of school-related problems, such as difficulty completing
classwork or difficulty in getting along with teachers. Intraclass correlation
coefficients (ICCs) indicated that the test–retest reliability of the parent
version was high for the Peer Relationships Scale (ICC = 0.95) and medium
for the School Problems Scale (ICC = 0.67) (30).
Children’s Depression Inventory (CDI). This 27-item scale measures depressive
symptoms. The scale is reliable, with high test–retest reliability and
high internal consistency, and is content-valid (31). The parent-informant
version also has high test–retest reliability (32).
Several of the scales had good face validity but lacked adequate psychometric
data. Therefore, we assessed internal consistency and 2-week test–retest
reliability in this sample. For instruments in which parents rate the same
behaviours, we also assessed the level of agreement between parents. Test–
retest reliability was medium to high (ICCs ranged from 0.59 for the Parent
Disagreements Scale to 0.82 for the Positive Parenting Practices Scale).
Similarly, internal consistency was high (that is, a > 0.74) for all scales except for the Depression Facts Quiz, in which a = 0.42. In contrast, the level of agreement between parents was low to medium. The greatest discrepancies occurred when parents rated family functioning (ICC = 0.33) and child
behaviour (ICCs for CDI scores, competence in sports and arts, participation
in activities, and involvement in adult-led activities were 0.2, 0.15,
0.41, and 0.26, respectively).
Statistical Analyses
We employed analysis of covariance (ANCOVA) using “intention-to-treat”
analysis to test differences between experimental and control groups in
posttreatment outcome measures. This included all participants randomized
to treatment or control groups for whom data at baseline and posttreatment
were available (n = 32 for these analyses). The baseline value of the dependent
variable was entered as a covariate to introduce a control for baseline
differences, because there were small (nonsignificant) group differences.
We calculated standardized ES when there were significant between-group
differences at posttreatment or when differences approached significance
(P < 0.10). A second set of ANCOVAs were conducted for variables wherein
the significance of the F-value was P < 0.10. In these analyses, proband
baseline CES-D score was a second covariate to control for the confounding
effect of parental depression severity. We selected the proband CES-D score
because there was selective loss of more subjects with depression in the
experimental group and because baseline data were complete for this variable.
Covariates were restricted to 2, owing to the small sample size.
Results
Table 2 provides means and SDs of family outcome measures, together with
the results of ANCOVAs. The F-values of ANCOVAs are displayed in the far
right column, as are the ES and F-values for the ANCOVAs that included
proband baseline CES-D as a covariate. For proband report, significant
between-group differences were found for FAD score (F = 7.6, df 31, P =
0.01). The Family Conflict Scale (F = 3.5, df 31, P = 0.07), Parent Disagreements
Scale (F = 3.7, df 23, P = 0.07), and Sense of Parenting Competence Scale
(F = 3.7, df 30, P = 0.06) showed trends toward significance (P < 0.1).
For these variables, the experimental group showed greater change toward
positive functioning. The ES were medium (0.6, 0.6, 0.4, and 0.5 for FAD,
Parent Disagreement, Family Conflict, and Sense of Parenting Competence
scale scores, respectively). There were no significant differences for
the CES-D, Depression Facts Quiz, or Parenting Practices subscale scores.
For partner report, there was a trend toward significance for CES-D score
(with a lower score in the experimental group) (F = 4.4, df 17, P = 0.05).
The ES for this difference was large (ES = 1.3). There were no between-group
differences for the remaining variables. The ANCOVAs with proband CES-D
as a second covariate consistently reduced effects attributable to treatment
grouping, especially for Parent Disagreement and Sense of Parenting Competence.
The F-value for the FAD was only marginally lowered (F = 6.6, df 31, P
= 0.02).
The between-group differences on the FAD general-functioning scale provide
an opportunity to assess the clinical significance of positive findings.
This scale has a validated clinical cut-off of 20. At baseline, the mean
score by proband for the experimental group was 1 full SD above this cut-off
(indicating that 84% exceeded cut-off). By posttreatment, approximately
60% exceeded the cut-off.
Table 3 provides the means and SDs of child outcome measures, as well as
the results of ANCOVAs to assess statistical significance of between-group
differences at posttreatment. Only the School Functioning Scale score showed
a significant between-groups difference (by proband report). Because this
was an isolated positive result, no further analyses are reported for these
data.
|
Table 2 Group statistics for parent outcome measures (Mean, SDs, F-values,
and significance level)
|
|
|
Experimental group
n = 21
|
Control group
n = 23
|
F, (ES)a
[F controlling for CES-D]
|
|
Proband report
|
Baseline
(n = 21)
|
Posttreatment
(n = 15)
|
Baseline
(n = 23)
|
Posttreatment
(n = 17)
|
Experimental vs control group at posttreatment
|
|
Parenting practices:
Positive
Hostile-ineffective
Consistency
Sense of parenting competence
FADb
Family conflict
Parent conflict
Depression facts
CES-Db
|
11.4 (4.2)
13.9 (4.5)
11.2 (3.2)
25.9 (5.4)
27.6 (7.3)
14.6 (3.4)
25.3 (3.6)
19.2 (2.4)
30.1 (13.7)
|
12.1 (3.7)
16.3 (4.4)
14.1 (4.3)
29.9 (5.0)
22.8 (5.0)
16.3 (3.9)
27.1 (3.6)
20.2 (1.9)
22.5 (14.2)
|
11.8 (3.7)
14.5 (5.3)
12.8 (4.7)
26.2 (5.8)
29.7 (8.0)
14.6 (4.0)
22.1 (5.6)
20.0 (2.7)
33.2 (14.0)
|
12.2 (4.2)
16.8 (5.3)
13.4 (4.7)
27.2 (6.4)
29.3 (7.5)
14.7 (3.4)
21.0 (6.5)
20.1 (2.2)
30.2 (11.4)
|
0.3
0.0
2.6
3.7c (0.5) [0.7]
7.6d (0.6) [6.6e]
3.5c (0.4) [0.4]
3.7c (0.6) [1.5]
0.2
0.9
|
|
Partner report
|
Baseline
(n = 14)
|
Posttreatment
(n = 10)
|
Baseline
(n =14)
|
Posttreatment
(n = 9)
|
Experimental vs control group at posttreatment
|
|
Parenting practices:
Positive
Hostile-ineffective
Consistency
Sense of parenting
Competence
FADb
Family conflict
Parent conflict
Depression facts
CES-Db
|
9.6 (3.7)
16.6 (4.6)
13.6 (4.4)
30.2 (4.2)
25.9 (5.4)
14.7 (3.7)
23.4 (6.1)
19.6 (2.7)
11.5 (14.4)
|
12.2 (3.1)
19.4 (2.2)
15.1 (3.3)
33.1 (4.1)
22.1 (5.1)
17.4 (2.3)
25.7 (6.3)
19.9 (2.3)
2.6 (3.7)
|
12.2 (5.0)
15.4 (5.9)
12.4 (4.8)
30.9 (5.2)
26.3 (6.1)
14.8 (4.7)
23.1 (5.4)
20.0 (3.5)
12.5 (9.5)
|
10.7 (4.5)
17.2 (5.9)
15.4 (3.2)
31.6 (6.6)
26.5 (8.2)
14.7 (3.9)
23.9 (5.2)
19.3 (2.1)
18.5 (14.6)
|
1.2
0.2
0.7
0.1
0.3
1.3
0.1
0.6
4.4c (1.3)
|
|
a Standardized effect size (ES) calculated as: difference score experimental
group – difference score control group
SD of all baseline scores
bAll scales higher score represents better functioning except for CES-D
and Family Functioning
cP < 0.1; dP < 0.01; eP < 0.05
|
|
Table 3 Group statistics for child outcome measures (mean, SDs, F-values,
and significance level)
|
|
|
Experimental group
|
Control group
|
F value
|
|
Proband report
|
Baseline
(n = 21)
|
Posttreatment
(n = 15)
|
Baseline
(n = 23)
|
Posttreatment
(n = 17)
|
Experimental vs control group at posttreatment
|
|
Competence in sports and arts
Participation
Adult-led activity
School functioning
Peer relationships
CDI - Parent Report
|
6.7 (1.5)
2.6 (1.7)
1.5 (0.7)
7.6 (2.1)
9.5 (5.5)
11.7 (8.7)
|
6.1 (1.4)
2.1 (1.0)
1.5 (0.7)
8.6 (1.9)
9.8 (5.2)
9.2 (6.3)
|
6.4 (1.0)
2.5 (2.6)
1.5 (1.1)
7.2 (2.7)
9.0 (4.7)
11.0 (6.9)
|
6.5 (1.2)
3.0 (4.9)
1.3 (0.6)
7.2 (3.2)
8.7 (4.6)
8.4 (6.3)
|
1.4
0.9
2.1
1.7
0.1
0.1
|
|
Partner report
|
Baseline
(n = 14)
|
Posttreatment
(n = 10)
|
Baseline
(n = 14)
|
Posttreatment
(n = 9)
|
Experimental vs control group at Posttreatment
|
|
Competence in sports and arts
Participation
Adult-led activity
School functioning
Peer relationships
CDI–parent report
|
6.2 (0.8)
1.8 (1.1)
1.2 (0.7)
7.1 (2.4)
9.3 (6.1)
8.5 (6.3)
|
6.4 (1.0)
2.4 (1.5)
1.8 (0.9)
8.4 (2.0)
7.4 (8.1)
6.7 (7.2)
|
6.6 (1.5)
2.2 (1.9)
1.3 (0.9)
8.1 (2.3)
7.7 (4.7)
6.6 (4.8)
|
6.4 (1.0)
1.9 (1.8)
1.4 (0.7)
8.3 (2.2)
6.9 (2.2)
5.2 (3.4)
|
0.1
0.7
0.4
0.1
0.0
0.4
|
Discussion
General Findings
Consistent with earlier work (16), this study successfully recruited parents
with depression who had children in the target age range. All expressed
interest in the program, indicating that they thought an intervention related
to parenting issues was relevant. Baseline characteristics showed that
the participants suffered clinically significant mood disorders with many
current depressive symptoms, high rates of psychiatric hospitalizations,
and recurrent illness and treatment by psychiatrists. The CIDI interview
did not identify all subjects as having depression in the prior 12 months,
which probably reflects lower-than-expected instrument sensitivity. Despite
participant interest, there was a high dropout rate beyond randomization.
This indicates that barriers to participation need to be understood if
the intervention is to be useful to a greater proportion of those in need
and to ensure that participants are exposed to an adequate intervention
dosage.
A priori, we assumed that, owing to the small sample size, there would
be few, if any, statistically significant differences between the treatment
and control groups and that medium ES could be taken as preliminary evidence
of effectiveness. For the probands, 4 of 9 measures had medium ES at posttreatment,
and for 1 measure (FAD), the between-group differences were statistically
significant. However, comparison of those missing with those included revealed
selective loss of those participants in the experimental group who had
higher depressive symptom scores. ANCOVAs controlling for baseline depressive
symptoms resulted in reduced F-values for all variables. The FAD scores
remained significantly different between groups. These analyses suggest
that selective loss of participants with more severe depression contributed
to, but did not explain entirely, the superior gains of the experimental
group. For partners, the only difference between experimental and control
groups at posttreatment was their lower depressive symptom scores. Fewer
partners were involved and this could account for the absence of other
positive findings for these respondents.
The assessment and treatment-retention problems encountered in the study
demonstrate the significant obstacles to undertaking research in this population.
It was evident that many of the participating families were highly stressed
and that, for these families, there were many competing treatment and social
service needs. Even though efforts were made to select a convenient time
and location and to provide child care, it was still not possible for some
parents to attend. These difficulties will interest those planning trials
or clinical interventions for this population; they must be addressed if
effective prevention that extends to all affected individuals is to be
realized. In this study, participants had lower depressive symptom scores
than did dropouts, suggesting that more severe depression acted as a barrier
to the intervention. In future, individuals with more severe depression
could be offered entry to the program when their depression has improved,
although this will not be feasible within the constraints of a clinical
trial. Therefore, efficacy trials should exclude referred patients whose
depression is simply too severe for them to attend. On a positive note,
if the group intervention brings about sustained improvements in family
functioning among those who can attend, it will be an important addition
to the therapeutic and preventive interventions currently available. Given
the different needs of individual families, it is likely that a selection
of interventions will be required. Further research is required to develop
measures that accurately identify which interventions should be offered
according to each family’s specific needs.
Study Limitations
The study had 4 main limitations. First, estimated treatment effects are
unreliable, owing to the small sample size and lack of statistical power,
which limits the possibilities for hypothesis testing. Second, some of
the instruments had limited published psychometric data. This situation
arose because of the lack of established measures for many of the dimensions
studied. While we addressed this limitation in part by obtaining reliability
data within the study, the validity of these measures is not established.
Further, the study included no observational measures. This was by choice,
given limited staff resources and a lack of readily applied observational
tools. These measurement issues affect interpretation of findings: it is
impossible to state unequivocally whether one might expect a preventive
impact, based on positive changes on study measures. Further work should
be directed to identifying observational and self-report measures of parenting
and marital and family relations. Some of the instruments used in this
study are promising, as they appear to be reliable and sensitive to change.
Third, failure to assess fidelity to the intervention means that one cannot
be fully confident that the intervention and its described ingredients
led to positive family changes. However, the intervention was manualized
and had a clear structure, which reduces the likelihood of major variation
from the intended program. Finally, data loss owing to dropouts and missed
assessments was the most serious study limitation. As discussed already,
there is no guarantee that missed experimental subjects were not poor responders
to the intervention. If this were the case, it would bias findings toward
inflated treatment effects. Readers should assess the study findings keeping
in mind that this was an exploratory study with the foregoing methodological
limitations.
Conclusions
Findings of positive differences between the treatment and control groups
in several areas of family functioning and relationships are promising
and indicate that further development and testing of the intervention is
warranted. However, the high rate of participant loss, especially among
participants with more severe depression, may have contributed to these
observed differences. The program could offer advantages for these families
over existing preventive programs. Specifically, the program should be
more efficient, because several families receive therapy at once. Further,
it is structured and manualized, making it readily transferrable to other
settings. However, more testing is needed to demonstrate that, with modification,
a high rate of participation can be achieved; that it consistently improves
family functioning; and that this contributes to better child outcomes,
alone or in tandem with child-directed interventions. The program is not
suitable for use in clinical settings until these problems and questions
are resolved.
Funding Support
This study was completed with the support of a grant from The Hamilton
Community Foundation.
Acknowledgements
Special thanks go to Dr Michael Boyle, Canadian Centre for Studies of Children
at Risk and Department of Psychiatry and Behavioural Neurosciences, McMaster
University, for his review of the paper and comments. The authors also
acknowledge the generous support of the families who participated and of
the Chedoke Child and Family Centre, Hamilton Health Sciences, for releasing
staff to run the groups.
References
1. Hammen C, Burge D, Burney E, Adrian C. Longitudinal study of diagnoses
in children of women with unipolar and bipolar affective disorder. Arch
Gen Psychiatry 1990;47:1112–7.
2. Birmaher B, Ryan ND, Williamson DE, Brent DA, Kaufman J, Dahl RE, and
others. Childhood and adolescent depression: a review of the past 10 years.
Part 1. J Am Acad Child Adolesc Psychiatry 1996;35:1427–39.
3. Beardslee WR, Versage E, Gladstone TR. Children of affectively ill parents:
a review of the past 10 years. J Am Acad Child Adolesc Psychiatry 1998;37:1134–41.
4. Weissman MM, Warner V, Wickramaratne P, Moreau D, Olfson M. Offspring
of depressed parents: 10 years later. Arch Gen Psychiatry 1997;54:932–40.
5. Zahn-Waxler C, Mayfield A, Radke-Yarrow M, McKnew D, Cytryn L, Davenport
Y. A follow-up investigation of offspring of parents with bipolar disorder.
Am J Psychiatry 1988;145:506–9.
6. Rutter M, Sandberg S. Psychosocial stressors: concepts, causes and effects.
European Journal of Child and Adolescent Psychiatry 1986;1:3–13.
7. Byrne C, Browne G, Roberts J, Ewart B. Surviving social assistance:
12 month prevalence of depression in sole support parents receiving social
assistance. CMAJ 1997;58:881–8.
8. Plomin R. The Emanuel Miller Memorial Lecture, 1993: genetic research
and identification of environmental influences. J Child Psychol Psychiatry
1994;35:817–34.
9. Garber J, Hilsman R. Cognition, stress and depression in young adolescents.
Child Adolesc Psychiatr Clin N Am 1992;1:129–67.
10. Radke-Yarrow M, Cummings EM, Kuczynski L, Chapman M. Patterns of attachment
in two- and three-year olds in normal families and families with parental
depression. Child Dev 1985;56:884–93.
11. Field T, Healy B, Goldstein S, Perry S, Bendell D, Schanberg S, and
others. Infants of depressed mothers show “depressed” behaviour even with
non-depressed adults. Child Dev 1988;59:1569–79.
12. Field T, Healy B, Goldstein S, Guthertz M. Behavioral-state matching
and synchrony in mother-infant interactions of non-depressed versus depressed
dyads. Dev Psychol 1990;26:7–14.
13. Hammen C, Burge D, Adrian C. Timing of mother and child depression
in a longitudinal study of children at risk. J Consult Clin Psychol 1991;59:341–5.
14. Clarke GN, Hornbrook M, Lynch F, Polen M, Gale J, Beardslee W, and
others. A randomized controlled trial of a group cognitive intervention
for preventing adolescent depression in adolescent offspring of depressed
parents. Arch Gen Psychiatry 2001;58:1127–34.
15. Beardslee WR, Hoke L, Wheelock B, Rothberg P, van der Velde P, Swatling
S. Initial findings on preventive intervention for families with parental
affective disorders. Am J Psychiatry 1992;149:1335–40.
16. Beardslee WR, Wright EJ, Salt P, Drezner K, Gladstone TR, Versage EM,
Rothberg PC. Examination of children’s responses to two preventive intervention
strategies over time. J Am Acad Child Adolesc Psychiatry 1997;36:196–204.
17. Kuipers L, Leff J, Lam D. Family work for schizophrenia: a practical
guide. London: Gaskell; 1992.
18. Goldstein MJ. Psychoeducation and relapse prevention. Int Clin Psychopharmacol
1995;9 (Suppl 5):59–69.
19. Cunningham C, Bremner P, Boyle M. Large group community-based parenting
programs for families of pre-schoolers at risk for disruptive behaviour
disorders: utilization, cost-effectiveness and outcome. J Child Psychol
Psychiatry 1995;36:1141–59.
20. Patten SB. Performance of the Composite International Diagnostic Interview
Short Form for major depression in community and clinical samples. Chronic
Dis Can 1997;18:109–12.
21. Patten SB, Brandon-Christie J, Devji J, Sedmak B. Performance of the
Composite International Diagnostic Interview Short Form for major depression
in a community sample. Chronic Dis Can 2000;21:68–72.
22.Weissman MM, Sholomskas D, Pottenger M, Prusoff BA, Locke BZ. Assessing
depressive symptoms in five psychiatric populations: a validation study.
Am J Epidemiol 1977;06:203–14.
23. Radloff LS. The CES-D Scale: a self-report depression scale for research
in the general population. Appl Psychol Measurement 1977;1:385–401.
24. Statistics Canada and Human Resources Development Canada. National
Longitudinal Survey of Children, overview of survey instruments for 1994-95,
data collection cycle 1. Ottawa: Statistics Canada and Human Resources
Development Canada; 1995.
25. Byles J, Byrne C, Boyle M, Offord D. Ontario Child Health Study: reliability
and validity of the general functioning subscale of the McMaster Family
Assessment Device. Fam Process 1988;27:97–104.
26. Hofferth S, Davis-Kean P, Davis J, Finkelstein J. The Child Development
Supplement of the Panel Study of Income Dynamics, user guide. Ann Arbour
(MI): University of Michigan; 1998.
27. Brent DA, Poling K, McKain B, Baugher M. A psychoeducational program
for families of affectively ill children and adolescents. J Am Acad Child
Adolesc Psychiatry 1993;32:770–4.
28. Boyle M, Offord D, Hofmann H, Catlin G, Byles J, Cadman D, and others.
Ontario Child Health Study: I. methodology. Arch Gen Psychiatry 1986;44:826–31.
29. John K, Gammon G, Prusoff BA, Warner V. The Social Adjustment Inventory
for Children and Adolescents (SAICA): testing of a new semi-structured
interview. J Am Acad Child Adolesc Psychiatry 1987;26:898–911.
30. McCleary L, Sanford M. Parental expressed emotion in depressed adolescents:
prediction of clinical course and relationship to comorbid disorders and
social functioning. J Child Psychol Psychiatry 2002. Forthcoming.
31. Kovacs M. The Children’s Depression Inventory (CDI). Psychpharmacol
Bull 1985; 21:995–8.1987;
32. Wierzbicki M. A parental form of the Children’s Depression Inventory:
reliability and validity in non-clinical populations. J Clin Psychol 1987;43:390–7.
Manuscript received May 2002, revised, and accepted July 2002.
1. Formerly, Associate Professor, Department of Psychiatry and Behavioural
Neurosciences, McMaster University, Hamilton, Ontario; now, Associate Professor,
Department of Psychiatry, University of Toronto, Toronto, Ontario.
2. Formerly, Associate Professor, School of Nursing, McMaster University,
Hamilton, Ontario; now, Dean, School of Health Sciences, University of
Ontario Institute of Technology, Oshawa, Ontario.
3. Assistant Professor, Department of Psychiatry and Behavioural Neurosciencese,
McMaster University, Hamilton, Ontario.
4. Facilitator, Mood Disorders Service, Chedoke Child and Family Centre,
Hamilton Health Sciences, Hamilton, Ontario.
5. Facilitator, Behaviour Problems Service, Chedoke Child and Family Centre,
Hamilton Health Sciences, Hamilton, Ontario.
6. Nurse, Chedoke Child and Family Centre, Hamilton Health Sciences, Hamilton,
Ontario.
7. Research Assistant, Chedoke Child and Family Centre, Hamilton Health Sciences,
Hamilton, Ontario.
Address for correspondence: Dr M Sanford, Child Psychiatry Program, Centre
for Addiction and Mental Health, 250 College Street, Toronto, ON M5T 1R8
1 | 2 | 3
|
