Cognitive impairment, anxiety, and depression are prevalent among older adults during prolonged hospital admissions, yet their interplay remains insufficiently understood despite its relevance to patient-centred care. Emotional distress is often presumed to influence cognitive performance in inpatient settings, although empirical evidence remains inconsistent. Clarifying whether global cognition and emotional symptomatology function as overlapping or independent clinical domains is therefore essential for designing responsive care pathways.

This study analysed the associations between global cognition and emotional symptoms in 61 Spanish adults aged 65 years and older who experienced prolonged hospital stays on general medical wards. Standardised assessments included the Spanish-adapted Mini-Mental State Examination (MMSE) and the Goldberg Anxiety and Depression Scale (GADS). Spearman correlations indicated no association between MMSE scores and anxiety (ρ = –.010, p = .937) or depression (ρ = .066, p = .614). A multivariable linear regression model incorporating both emotional measures confirmed the absence of a relationship, accounting for only 1.1% of the variance in global cognition (R² = .011; p = .721). In contrast, anxiety and depression showed a robust correlation (ρ = .613, p < .001), suggesting a coherent emotional profile that may be amenable to joint screening.

These findings indicate that cognitive performance and emotional distress operate as differentiated yet complementary domains in older adults admitted for prolonged stays. The absence of association between MMSE scores and GADS anxiety or depression suggests that cognitive status cannot be inferred from emotional symptomatology, nor vice versa. Clinically, these results challenge the assumption that higher emotional distress predicts poorer cognitive functioning. From a patient-centred perspective, the findings highlight the need for dual screening pathways: an integrated circuit for emotional distress and a distinct circuit for cognitive evaluation. Separating these domains may prevent unsupported clinical inferences, improve early identification of vulnerability, and support more precise models of inpatient care for older adults.