Lee, Sang Jun;Kim, Tae Hyung;Huh, Lyang;Choi, Seung Eun;Lee, Bong Ju;Kim, Gyung Mee;Lee, Jung Goo;Kim, Hong Dae;Mun, Chi Woong;Kim, Young Hoon
223
Objectives In this study, the authors evaluated the correlation between levels of serum lipid, homocysteine, and folate with volumes of hippocampus, amygdala, corpus callosum, and in patients with amnestic mild cognitive impairment (aMCI) or Alzheimer's disease (AD) type. Methods The study recruited patients who visited the dementia clinic of Haeundae Paik Hospital in Korea between March 2010 and June 2014. Among those, patients who had taken the neurocognitive test, brain magnetic resonance imaing, tests for serum lipid, homocysteine, folate, and apolipoprotein E (APOE) genotyping and diagnosed with aMCI or AD were included for analysis. Bilateral hippocampus, entorhinal cortex, amygdala and corpus callosum were selected for region of interest (ROI). The cross-sectional relationships between serum lipid, homocysteine, folate and ROI were assessed by partial correlation analysis and multiple linear regression analysis. Results In patients with aMCI, old age (> 80) and APOE ${\varepsilon}4$ carrier were associated with AD [odds ration (OR) : 12.80 ; 95% confidence interval (CI) : 2.25-72.98 and OR : 4.48 ; 95% CI : 1.58-12.67, respectively]. In patients with aMCI or AD, volumes and thickness of ROI were inversely correlated with levels of serum lipid and homocysteine. In multiple linear regression analyses, higher total cholesterol level was related to lower left, right hippocampus volume and left amygdala volume ; higher low-density lipoprotein cholesterol was related to lower right entorhinal cortex thickness ; higher homocysteine level was related to lower corpus callosum volume. Conclusions Higher serum lipid and homocysteine levels are associated with decreased volume of hippocampus, amygdala, corpus callosum and entorhinal cortex thickness in patients with aMCI or AD. These findings suggest that serum lipid and homocysteine levels are associated with AD as a modifiable risk factor.