Differences in Odor Identification in Early-Onset and Late-Onset Depression

Abstract

(1) Background: Odor identification (OI) dysfunction is a potential predictor of developing dementia in late life depression (LLD). However, it is not clear whether patients with early onset depression (EOD) and late onset depression (LOD) may exhibit different OI dysfunctions. The aim of this study was to compare OI between EOD patients and LOD patients and its relationship with cognitive function. (2) Methods: A total of 179 patients with LLD and 189 normal controls were recruited. Participants underwent clinical assessment, olfactory testing, and comprehensive neuropsychological assessment. The OI scores of EOD patients and LOD patients were compared, and correlation analyses and mediation analyses were used to explore the relationship between OI and cognition. (3) Result: LOD patients exhibited lower OI scores than EOD patients and normal controls (NCs). Additionally, the LOD patients exhibited a higher percentage of OI dysfunction than the EOD patients. Moreover, OI scores were associated with global cognition, memory, language, and visuospatial ability in the EOD group (p < 0.05) but were not associated with any cognitive score in the LOD patients (p > 0.05). Finally, the scores of the Auditory Verbal Learning Test Immediate recall and Boston Naming Test exhibited a partially mediating effect on the difference in OI scores between the EOD and LOD patients. (4) Conclusions: LOD patients exhibited worse OI than EOD patients, and their difference in OI was mediated by their memory and language function.

Keywords: cognitive impairment; late life depression; mediating effect; odor identification; olfactory dysfunction.

Figure 1

Comparison of the OI scores (A) and the percentage of OI dysfunction (B) among the EOD, LOD and NC groups. OI, odor identification; OI (%), percentage of OI dysfunction. EOD means patients with early onset depression; LOD means patients with late onset depression; NC means normal controls. * p < 0.05; ** p < 0.01.

Figure 2

Correlational analyses between cognitive function and OI in EOD, LOD and NC groups. (A) OI was associated with MMSE in EOD (r = 0.385, p = 0.001, q = 0.008) but not in NC and LOD. (B) OI was associated with AVLT N1–3 in EOD (r = 0.442, p < 0.001, q < 0.001) and NC (r = 0.286, p < 0.001, q < 0.001) but not in LOD. (C) OI was associated with AVLT N4 in EOD (r = 0.396, p = 0.002, q = 0.008) and NC (r = 0.299, p < 0.001, q < 0.001) but not in LOD. (D) OI was associated with AVLT N5 in NC (r = 0.203, p = 0.008, q = 0.032) but not in EOD and LOD. (E) OI was associated with AVLT N6 in EOD (r = 0.391, p = 0.002, q = 0.008) but not in LOD and NC. (F) OI was associated with BNT in EOD (r = 0.381, p = 0.002, q = 0.008) but not in LOD and NC. (G) OI was associated with SDMT in NC (r = 0.211, p = 0.006, q = 0.032) but not in EOD and LOD. (H) OI was associated with ROCF in EOD (r = 0.347, p = 0.006, q = 0.016) but not in LOD and NC. OI, odor identification; MMSE, Mini-Mental State Examination.

Figure 3

The mediating effect of cognitive function on the OI scores of EOD/LOD groups. (A) AVLT N1–3 mediated the difference in OI scores of EOD/LOD groups. (B) BNT mediated the difference in OI scores of EOD/LOD groups. * p < 0.05; ** p < 0.01; EOD means patients with early-onset depression; LOD means patients with late-onset depression. AVLT N1–3, Auditory Verbal Learning Test Immediate recall; BNT, Boston Naming Test; OI, odor identification.