Differential measurement of depression by racial group: Findings from an Alzheimer’s Disease Center cohort

Background Both researchers and clinicians require risk screeners that are psychometrically‐sound and culturally‐appropriate in addition to being efficient and low‐burden. Brief versions of the Geriatric Depression Scale (GDS) exist, but there is limited study on the performance of these screeners across historically underserved and underrepresented racial groups. We examined the ability of 10 brief versions of the GDS to detect group differences in depressive symptom endorsement and to predict cognitive status in White, Black, and American Indian/Alaska Native (AI/AN) participants in an ADC‐based cohort. Method Participants included 555 cognitively‐normal middle‐aged and older‐adults (Mage =60.85) enrolled in the Wisconsin ADRC. The majority of participants were female (66.5%). 17.7% of participants were Black and 3.1% were AI/AN. For each GDS version, we used Kruskal‐Wallis tests with post‐hoc‐pairwise comparisons with Bonferroni corrections to compare scores across the three groups. Result Using nine of 10 GDS versions, there were significant group differences in total score. Post hoc pairwise comparisons revealed that AI/AN endorsed significantly fewer symptoms of depression than Blacks (p < 0.05) when using the GDS‐4a. Whites and Blacks were not significantly different in GDS‐4a. However, Whites and Blacks were different in eight of 10 versions (i.e., GDS‐1a; GDS‐4b; GDS‐5; GDS‐7; GDS‐8; GDS‐10a; GDS‐10b; GDS‐15). No group differences were found between Whites and AI/AN and Blacks and AI/AN in most versions of the GDS. In the combined sample, all GDS versions showed a significant, moderate positive correlation with CDR score. Conclusion Rigorous assessment and detection of depression is an important step to reducing dementia morbidity and possibly risk, but many brief screeners were evaluated in racial homogenous groups. Most GDS short versions were able to detect significant differences in depression score between Whites and Blacks. Interestingly, the GDS‐4a (4 non‐behavioral items) showed significant difference between AI/AN and Blacks, not detected in other versions. These early findings suggest that more research must address culturally‐valid assessment of depressive symptoms in cognitive aging research and geriatric clinic settings. Acknowledgement: Data collection and contributions by author Carey E. Gleason are supported by funding from NIH‐NIA for African Americans Fighting Alzheimer’s at Midlife (AA‐FAIM, R01 AG054059) and the Wisconsin Alzhe