Relationship between olfaction and cognitive function in patients with type 2 diabetes mellitus: a follow-up study

丁群,张洲,于从从,倪文煜,缪应雯,徐湘,侯银娇,毕艳
DOI: https://doi.org/10.3760/cma.j.cn115791-20220508-00204
2022-01-01
Abstract:Objective:To investigate the evolution of cognitive function and its relationship with olfaction in patients with type 2 diabetes mellitus (T2DM) who were longitudinally followed up.Methods:The prospective study included T2DM patients with complete baseline and follow-up data who were hospitalized in the Endocrinology Department of Drum Tower Hospital from January 2016 to January 2022. Smell behavior tests were performed at baseline, and elaborate cognitive assessment were performed both at baseline and follow-up. The change value of cognitive function was defined as follow-up cognitive scores minus baseline cognitive scores, positive scores indicated increase and negative scores indicated decrease. Paired t-test was used to compare clinical indicators and cognitive test scores before and after the follow-up. Partial correlation was used to analyze the correlation between olfactory test scores and changes in cognitive domains. Participants were categorized into low-smell score group (<45.0 points), middle-smell score group (45.0-49.4 points) and high-smell score group (≥49.5 points) by tertile based on the total smell test scores at baseline, and the cognitive test scores of the three groups were compared by one-way analysis of variance (ANOVA). Results:Thirty-five T2DM patients completed the follow-up, the baseline age was (56.4±7.9) years, and the median follow-up time was 4.6 years. Compared with baseline, patients with T2DM showed statistically significant decrements in memory and verbal fluency at follow-up [(167.3±22.8) vs. (144.8±18.6) points, P<0.01; (22.1±4.4) vs . (19.1±4.8) points, P<0.01, respectively]. Memory and verbal fluency descent were separately found in 65.7% (23/35) and 37.2% (13/35) of the patients. Partial correlation analysis was performed after adjusting for baseline age, gender, years of education, duration of diabetes, glycated hemoglobin A 1c, and time of follow-up. The significant association of the changes in memory with the total scores of the smell test at baseline was observed ( r=0.400, P<0.05). There were separately 11, 12, 12 participants in the low-, middle- and high-smell scores group. Compared with patients in the high-smell scores group, patients with low-smell scores had greater memory decline ( P<0.05). Conclusions:After a median follow-up of 4.6 years, the memory and verbal fluency of T2DM patients significantly decreased. Patients with lower olfactory test scores had greater memory decline. Olfactory impairment might be a potential marker of memory decline in T2DM patients.
What problem does this paper attempt to address?