Frailty and Depression in Late Life: A High-Risk Comorbidity With Distinctive Clinical Presentation and Poor Antidepressant Response

Document Type

Journal Article

Publication Date



The journals of gerontology. Series A, Biological sciences and medical sciences








Antidepressant treatment; Depression; Frailty


BACKGROUND: To investigate the longitudinal relationship between physical frailty, the clinical representation of accelerated biological aging, and antidepressant medication response in older adults with depressive illness. METHODS: An 8-week randomized placebo-controlled trial (escitalopram or duloxetine) followed by 10 months of open antidepressant medication treatment (augmentation, switch strategies) was conducted in an outpatient research clinic. 121 adults aged 60 years or older with major depressive disorder (MDD) or persistent depressive disorder and a 24-item Hamilton Rating Scale for Depression (HRSD) ≥16 were enrolled. Primary measures assessed serially over 12 months include response (50% reduction from baseline HRSD score), remission (HRSD score <10), and frailty (non/intermediate frail [0-2 deficits] vs frail [≥3 deficits]); latent class analysis was used to classify longitudinal frailty trajectories. RESULTS: A 2-class model best fit the data, identifying a consistently low frailty risk (63% of the sample) and consistently high frailty risk (37% of the sample) trajectory. Response and remission rates (ps ≤ .002) for adults in the high-risk frailty class were at least 21 percentage points worse than those in the low-risk class over 12 months. Furthermore, subsequent frailty was associated with previous frailty (ps ≤ .01) but not previous response or remission (ps ≥ .10). CONCLUSIONS: Antidepressant medication is poorly effective for MDD occurring in the context of frailty in older adults. Furthermore, even when an antidepressant response is achieved, this response does little to improve their frailty. These data suggest that standard psychiatric assessment of depressed older adults should include frailty measures and that novel therapeutic strategies to address comorbid frailty and depression are needed.


Biostatistics and Bioinformatics