Associations Between Depression Symptom Burden and Delirium Risk: A Prospective Cohort Study.

Abstract

Background and objectives: Delirium and depression are prevalent in aging. There is considerable clinical overlap, including shared symptoms and comorbid conditions, including Alzheimer's disease, functional decline, and mortality. Despite this, the long-term relationship between depression and delirium remains unclear. This study assessed the associations of depression symptom burden and its trajectory with delirium risk in a 12-year prospective study of older hospitalized individuals.

Research design and methods: A total of 319 141 UK Biobank participants between 2006 and 2010 (mean age 58 years [range 37-74, SD = 8], 54% women) reported frequency (0-3) of 4 depressive symptoms (mood, disinterest, tenseness, or lethargy) in the preceding 2 weeks prior to initial assessment visit and aggregated into a depressive symptom burden score (0-12). New-onset delirium was obtained from hospitalization records during 12 years of median follow-up. 40 451 (mean age 57 ± 8; range 40-74 years) had repeat assessment on average 8 years after their first visit. Cox proportional hazard models examined whether depression symptom burden and trajectory predicted incident delirium.

Results: A total of 5 753 (15 per 1 000) newly developed delirium during follow-up. Increased risk for delirium was seen for mild (aggregated scores 1-2, hazards ratio, HR = 1.16, [95% confidence interval (CI): 1.08-1.25], p < .001), modest (scores 3-5, 1.30 [CI: 1.19-1.43], p < .001), and severe (scores ≥ 5, 1.38 [CI: 1.24-1.55], p < .001) depressive symptoms, versus none in the fully adjusted model. These findings were independent of the number of hospitalizations and consistent across settings (eg, surgical, medical, or critical care) and specialty (eg, neuropsychiatric, cardiorespiratory, or other). Worsening depression symptoms (≥1 point increase), compared to no change/improved score, were associated with an additional 39% increased risk (1.39 [1.03-1.88], p = .03) independent of baseline depression burden. The association was strongest in those over 65 years at baseline (p for interaction <.001).

Discussion and implications: Depression symptom burden and worsening trajectory predicted delirium risk during hospitalization. Increased awareness of subclinical depression symptoms may aid delirium prevention.