Background: Depression is under-diagnosed and under-treated in older adults. The purposes of this study were to (a) evaluate the psychometric properties of the Collateral Source Geriatric Depression Scale (CS-GDS), (b) compare collateral source scores on the CS-GDS with patient scores on the GDS, and (c) examine factors associated with any discrepancies between the CS-GDS and the GDS.

Methods: This secondary analysis used data from 132 older adults and their collateral sources attending a geriatric assessment program over a 15-month period. Scores on the 30-, 15-, and 5-item CS-GDS were compared to clinician diagnoses of depression using DSM-IV-TR criteria and patient GDS scores.

Results: The three forms of the CS-GDS had acceptable internal consistency, sensitivity and specificity with recommended cut-off scores of 18, 9 and 3, respectively. Collateral sources reported more depressive symptoms than patients did themselves. Simple regression analysis showed that caregiver burden significantly influenced the discrepancy between CS-GDS scores and GDS scores (β = 0.147; p = 0.004). Functional limitations and collateral relations to the patient were not associated with these discrepancies.

Conclusion: All three collateral versions had acceptable psychometric properties, which supports the use of the CS-GDS to assess depression in older adults. The CS-GDS provides an important alternative for depression screening with older adults who cannot complete screening tools themselves; however, alternative cut-off scores must be used for high sensitivity and specificity.