How can researchers study developmental constructs over time when age-appropriate measures change as their sample ages?
Studying change in depressive symptoms in youth over time poses specific challenges for researchers related to both change in symptom manifestation and change in age-validated measurement.
Assessing change in mental health, such as depressive symptoms, across development is particularly challenging for two related reasons. First, the symptoms of depression look different at different ages; for example, in childhood, depression often manifests as angry mood, but as youth age, depression manifests as sadness and suicidal ideation. Second, and accordingly, the way clinicians and researchers measure mental health symptoms also changes across childhood, adolescence, and adulthood. To examine depressive symptoms, children are often assessed using tools like the Children’s Depression Inventory (CDI; validated for use with children age 8-17 years), while adults are assessed using measures like the Beck Depression Inventory (BDI-II; validated for use with adolescents and adults age 13 and older). Although both tools are reliable, valid, and age-appropriate, they include different items and response options. This makes it challenging to track how individuals’ level of depression changes with age. If different measures are used at different times, it is not possible to know whether the observed changes in depression are indicative of an individual’s symptoms changing over time or if they are a by-product of change in the measurement instrument. Tracking and answering questions about changes in depressive symptoms when different measurement tools are used requires some creative linking of the different tools.
Utility of a conversion formula in studying age-related change across time with two age-appropriate measurement tools. Both the CDI and BDI-II are validated for use in adolescents ages 13-17. By measuring depressive symptoms with both measures in a sample of adolescents, we developed a conversion formula that can be used to link or translate BDI-II scores into CDI scores. Once the scores are all in the same metrics, we can obtain and study in a meaningful way how individuals’ depressive symptoms change from childhood, through adolescence, and into adulthood. Making use of unique data collected from 4 groups of adolescent females (ages 11, 13, 15, and 17 at study entry; N = 262) and followed for 4 years, we created a conversion formula for linking the CDI and BDI-II, and use that formula to study age-related changes in depressive symptoms.
Are there age-related differences in depression symptomology? We found that when change in depressive symptoms is tracked using the CDI measurement tool, depressive symptoms appear to decrease with age across adolescence, on average. And, we found that when change in depressive symptoms is tracked using the BDI-II, depressive symptoms appear to increase with age across adolescence, on average. Same girls, same study. Using our conversion formula to combine the two measurement tools, we found – using both the CDI and the BDI-II – that depressive symptoms decreased across adolescence (ages 11-21), on average. This finding – which arguably uses the best measurement tools available at all ages – is inconsistent with prior findings and much of the extant literature. Why? We are not sure. It may be an artifact of our study design, or our sample. But it may also be that the prior literature, mostly studies that used only one measurement tool and only obtained a few assessments during a small age span, is not an accurate representation of how girls’ depressive symptoms actually develop with age.
Future directions for developmental scientists. The study suggests that researchers should continue to evaluate the links between the CDI and BDI-II with larger and more diverse samples in order to develop a more precise conversion formula for widespread use. Publishers of the CDI and BDI-II should consider collecting these data and replicating these findings in order to produce a measurement instrument that is not threatened by change in symptomatology across age. To build upon what was accomplished in this study, researchers should include use of the CDI and BDI-II at overlapping measurement occasions (i.e., concurrent use of both measures). We also encourage researchers to adopt and extend this approach for other constructs that face similar measurement challenges related to developmental change over time (e.g., internalizing symptoms, aggression, self-regulation, etc.).
Implications for adolescent development research. This study’s efforts will help developmental scientists and mental health practitioners gain more insight into the progression of depressive symptoms across key developmental risk periods. By understanding how depressive symptoms change in a normative population across the lifespan, we can improve the identification and developmentally appropriate treatment of depressive symptoms and mechanisms that drive risk and resilience.
Read the original study by Emily LoBraico, Nathan Lutz, Sarah Beal, Lorah Dorn, and Nilam Ram here.
Emily LoBraico is a PhD candidate in Human Development and Family Studies at Penn State University. Her research interests include adolescent development in the family context and effective use of family-based preventive interventions for adolescent problem behaviors.
Sarah Beal is a developmental psychologist and research faculty member at Cincinnati Children’s Hospital. She is interested in research addressing the transition to adulthood, and what aspects of involvement in foster care contribute to poor developmental outcomes for adolescents.
Image by Adobestock/Pavel Iarunichev