Validity of a 12-item version of the CES-D used in the National Longitudinal Study of Children and Youth

Department of Community Health and Epidemiology, Dalhousie University, 5790 University Avenue, Halifax, Nova Scotia B3H 1V7, Canada.
Chronic diseases in Canada (Impact Factor: 1.6). 03/2005; 26(2-3):65-72.
Source: PubMed

ABSTRACT This validation study assessed the degree of confidence that can be placed on inferences from depressive symptoms among adolescents, based on a 12-item version of the Centre for Epidemiological Studies Depression scale (CES-D). This short version of the scale had been developed for application in the National Longitudinal Study of Children and Youth and we refer to it as the CES-D-12-NLSCY. The major data source for the present validation study was a 2002/2003 survey of 12,990 students in junior and senior high school in the Atlantic provinces of Canada. Receiver operating characteristic curve analyses for two different proxy gold standards yielded adequate areas under the curve (AUCs) of .84 and .80, allowing us to establish cut points for three categories of depressive symptoms in the general adolescent population: Minimal (CES-D-12-NLSCY total score 0 to 11), Somewhat Elevated (total score 12 to 20) and Very Elevated (total score 21 to 36). The CES-D-12-NLSCY was found to have acceptable internal consistency (Cronbach=s alpha .85). All but one of the 12 items of the CES-D-12-NLSCY were found to have acceptable discrimination ability. The prevalence of Minimal, Somewhat and Very Elevated depressive symptoms in the adolescent student population of the Atlantic provinces was estimated to be 72.3, 19.5 and 5.5 percent, respectively. A further 2.6 percent of students who responded to fewer than 11 items of the scale were classified as Indeterminate with regards to depressive symptom category. The major threat to the accuracy of the CES-D-12-NLSCY is its lack of inquiry about irritability, which is a key symptom of depression in youth.

102 Reads
  • Source
    • "The reliability coefficient for this study was 0.70. A 12-item version of the Centre for Epidemiological Studies Depression Scale, (CES-D-12-NLSCY; α=0.85; Poulin et al. 2005) was included to measure risk levels for depression. Items such as " I had crying spells " , " I had trouble keeping my mind on what I was doing " , and " I felt like I was too tired to do things " are rated on a 4-point 1 Prioritised ethnicity was used in the study "
    [Show abstract] [Hide abstract]
    ABSTRACT: As the fields of resilience and positive psychology grow, there is increased focus on understanding the resources that facilitate wellbeing and positive development rather than factors preventing negative outcomes. As these developments occur, there is an increasing need for measures validated on high risk and marginalised populations of youth that assess the resources that potentiate wellbeing. Using data from a cross-sectional study of 1,477 youth (12-to-17 years old) in New Zealand, who were in either high risk or low risk environments, this paper presents the validation of questions developed by Hektner (Journal of Research in Rural Education, 11, 3–14, 1995) as a scale for use with youth, assessing their feelings about their own futures. A bi-dimensional construct composed of positive (three items) and negative (four items) emotions has been identified and confirmed through the use of exploratory factor analysis and confirmatory factor analysis that also tested factorial invariance across the two groups. Inter-item correlations ranged from 0.357 to 0.517 (negative emotions) and 0.295 to 0.455 (positive emotions). The two sub-scales show good internal reliability. Convergent and divergent validity testing of the measure is assessed using scales of risk (conduct problems and risk of depression), resilience, and satisfaction with life. Patterns in the responses of youth across the two groups are also compared.
    Child Indicators Research 10/2015; 8(3). DOI:10.1007/s12187-014-9269-z. · 0.96 Impact Factor
  • Source
    • "Participants were considered to have experienced a SLE if the PMK reported any of 13 specific events, as well as 'other traumatic events' (Table 2). Maternal depression was assessed at each time point using a 12-item version of the Centre for Epidemiological Studies-Depression scale (CES-D-12) (Poulin et al. 2005). Children were considered to have been exposed to maternal depression if "
    [Show abstract] [Hide abstract]
    ABSTRACT: The objective of this study was to examine associations between trajectories of childhood neighbourhood social cohesion and adolescent mental health and behaviour. This study used data from the National Longitudinal Survey of Children and Youth, a nationally representative sample of Canadian children. The sample included 5577 children aged 0-3 years in 1994-1995, prospectively followed until age 12-15 years. Parental perceived neighbourhood cohesion was assessed every 2 years. Latent growth class modelling was used to identify trajectories of neighbourhood cohesion. Mental health and behavioural outcomes were self-reported at age 12-15 years. Logistic regression was used to examine associations between neighbourhood cohesion trajectories and outcomes, adjusting for potential confounders. Five distinct trajectories were identified: 'stable low' (4.2%); 'moderate increasing' (9.1%); 'stable moderate' (68.5%); 'high falling' (8.9%); and 'stable high' (9.3%). Relative to those living in stable moderately cohesive neighbourhoods, those in stable low cohesive neighbourhoods were more likely to experience symptoms of anxiety/depression [odds ratio (OR) = 1.73, 95% confidence interval (CI) 1.04-2.90] and engage in indirect aggression (OR = 1.62, 95% CI 1.07-2.45). Those with improvements in neighbourhood cohesion had significantly lower odds of hyperactivity (OR = 0.67, 95% CI 0.46-0.98) and indirect aggression (OR = 0.69, 95% CI 0.49-0.96). In contrast, those with a decline in neighbourhood cohesion had increased odds of hyperactivity (OR = 1.67, 95% CI 1.21-2.29). Those in highly cohesive neighbourhoods in early childhood were more likely to engage in prosocial behaviour ('high falling': OR = 1.93, 95% CI 1.38-2.69; 'stable high': OR = 1.89, 95% CI 1.35-2.63). These results suggest that neighbourhood cohesion in childhood may have time-sensitive effects on several domains of adolescent mental health and behaviour.
    Psychological Medicine 07/2015; -1:1-10. DOI:10.1017/S0033291715001245 · 5.94 Impact Factor
    • "The interview also assessed depression in the PMK using an abbreviated 12-item version of the CES-D rating scale for depressive symptoms (Radloff, 1977). Here, a score of 21+ was regarded as evidence of depression in the PMK (Poulin et al. 2005). Excessive alcohol consumption by the PMK was also recorded, specifically whether the PMK consumed five or more drinks or more at least once per month in the 12 months preceding the interview. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Aims. Accumulating evidence links childhood adversity to negative health outcomes in adulthood. However, most of the available evidence is retrospective and subject to recall bias. Published reports have sometimes focused on specific childhood exposures (e.g. abuse) and/or specific outcomes (e.g. major depression). Other studies have linked childhood adversity to a large and diverse number of adult risk factors and health outcomes such as cardiovascular disease. To advance this literature, we undertook a broad examination of data from two linked surveys. The goal was to avoid retrospective distortion and to provide a descriptive overview of patterns of association. Methods. A baseline interview for the Canadian National Longitudinal Study of Children and Youth collected information about childhood adversities affecting children aged 0-11 in 1994. The sampling procedures employed in a subsequent study called the National Population Health Survey (NPHS) made it possible to link n = 1977 of these respondents to follow-up data collected later when respondents were between the ages of 14 and 27. Outcomes included major depressive episodes (MDE), some risk factors and educational attainment. Cross-tabulations were used to examine these associations and adjusted estimates were made using the regression models. As the NPHS was a longitudinal study with multiple interviews, for most analyses generalized estimating equations (GEE) were used. As there were multiple exposures and outcomes, a statistical procedure to control the false discovery rate (Benjamini-Hochberg) was employed. Results. Childhood adversities were consistently associated with a cluster of potentially related outcomes: MDE, psychotropic medication use and smoking. These outcomes may be related to one another since psychotropic medications are used in the treatment of major depression, and smoking is strongly associated with major depression. However, no consistent associations were observed for other outcomes examined: physical inactivity, excessive alcohol consumption, binge drinking or educational attainment. Conclusions. The conditions found to be the most strongly associated with childhood adversities were a cluster of outcomes that potentially share pathophysiological connections. Although prior literature has suggested that a very large number of adult outcomes, including physical inactivity and alcohol-related outcomes follow childhood adversity, this analysis suggests a degree of specificity with outcomes potentially related to depression. Some of the other reported adverse outcomes (e.g. those related to alcohol use, physical inactivity or more distal outcomes such as obesity and cardiovascular disease) may emerge later in life and in some cases may be secondary to depression, psychotropic medication use and smoking.
    Epidemiology and Psychiatric Sciences 02/2015; DOI:10.1017/S2045796015000104 · 3.91 Impact Factor
Show more