Agreement between CES-D and CIDI-SF scales of depression among older adults: a cross-sectional comparative study based on the longitudinal aging study in India, 2017-19.

Background: Understanding the differences between various instruments for assessing depression will help researchers and health practitioners to choose a more appropriate tool and develop a framework to enhance resilience to mental health problems in the older population. The current study aimed to compare the 10-item Center for Epidemiological Studies-Depression (CES-D) scale with the Composite International Diagnostic Interview-Short Form (CIDI-SF) for measuring depressive symptoms in older Indians.

Methods: Data came from the first wave of the Longitudinal Aging Study in India (LASI) which was conducted during 2017-19. The final sample included 30,368 older individuals aged 60 years and above (15,824 women and 14,544 men). The level of agreement between the CES-D and CIIDI-SF instruments for depression classification was assessed using Kappa coefficients at various cut-off values. Multivariable logistic regression models were used to examine the associations between background characteristics and depressive symptoms assessed by both instruments.

Results: The prevalence of depressive symptoms based on the CES-D instrument was higher than that of CIDI-SF (30.2% vs. 8.3%). The level of agreement between the CES-D and CIDI-SF for depression classification was 'none' to 'minimal' (κ = 0.04-0.24). Assuming the CIDI-SF scale as the "gold standard", with rising threshold values, sensitivity of the CES-D scale decreased while specificity increased. The CES-D scale yielded a sensitivity of 29-82% and specificity of 39-92% across cut-off values of 3 to 6. The Pearson correlation between the CIDI-SF and CES-D total scores was significant but weak (r =.20). Multivariable analysis showed that depressive symptoms were more prevalent among women, those with higher education, individuals living alone, those with diagnosed psychiatric disorders, and individuals with two or more chronic conditions, compared to men, those who were uneducated, those not living alone and healthier peers, when assessed using both CES-D and CIDI-SF scales. The associations of marital status, religion and wealth quintiles were significant only for depressive symptoms assessed using the CES-D scale.

Conclusions: Depression cases identified by the CES-D showed poor agreement with those identified by the CIDI-SF. Therefore, the prevalence of depressive symptoms measured by the CESD is not interchangeable with that measured by the CIDI-SF. These findings suggest the importance of using both the CES-D and CIDI-SF in large population-based cohorts and surveillance surveys to obtain more accurate and nuanced understanding of depressive disorders across various subgroups of the older population.