Skip to main content
. 2021 Jun 1;7:11. doi: 10.1038/s41514-021-00064-9

Table 2.

Pooled associations between metabolite levels and disease status/severity of AD pathology.

Disease group CERAD Braak
ITG MFG ITG MFG ITG MFG
Metabolite name β P value β P value β P value β P value β P value β P value
De novo cholesterol biosynthesis
Cholesterol . . −0.062 0.170 . . −0.037 0.292 . . 0.007 0.806
Lanosterol . . −0.278 <0.001 . . −0.139 0.012 −0.090 0.050 −0.180 <0.001
24,25-dihydrolanosterol . . . . . . . . −0.032 0.333 . .
7-dehydrocholesterol . . . . . . . . 0.160 0.147 . .
Desmosterol
Cholesterol catabolism (enzymatic)
27-hydroxycholesterol 0.210 0.170 . . 0.066 0.555 . . 0.051 0.641 −0.120 0.223
4β-hydroxycholesterol 0.119 0.234 . . 0.030 0.688 . . . . . .
24S-hydroxycholesterol −0.113 0.011 . . −0.074 0.008 . . −0.094 <0.001 −0.087 0.042
7α-hydroxycholesterol 0.154 0.068 . . 0.135 0.023 . . −0.002 0.977 −0.109 0.066
Cholesterol catabolism (non-enzymatic)
5α,6α-epoxycholesterol 0.199 0.040 0.115 0.139 0.118 0.116 0.038 0.508 . . . .
5α,6β-dihydroxycholestanol 0.213 0.017 −0.155 0.088 0.105 0.096 −0.111 0.101 . . −0.106 0.062
5β,6β-epoxycholesterol 0.196 0.013 0.071 0.327 0.112 0.069 . . 0.085 0.152 . .
7-ketocholesterol 0.304 0.001 . . 0.179 0.008 . . 0.102 0.155 −0.035 0.537
7β-hydroxycholesterol 0.237 0.005 . . 0.157 0.010 . . 0.050 0.448 −0.089 0.284

ITG Inferior Temporal Gyrus, MFG Middle Frontal Gyrus; P < 0.05 in bold.

Negative coefficients indicate that lower metabolite concentration is significantly associated with AD, higher neuritic plaque burden (CERAD score), or higher neurofibrillary tangle pathology (Braak score). Positive coefficients indicate that higher metabolite concentration is significantly associated with AD, higher neuritic plaque burden (CERAD score), or higher neurofibrillary tangle pathology (Braak score). Blank cells indicate that results were not pooled; these are included in cohort-specific secondary analyses in Supplementary Table 1. Significant associations (P < 0.05) are indicated in bold. Baltimore Longitudinal Study on Aging (BLSA) sample size: AD (n = 15), CN (n = 8); Religious Orders Study (ROS) sample size: AD (n = 31), CN (n = 22); note: samples pooled in analyses.