Causal effect of blood osteocalcin on the risk of Alzheimer's disease and the mediating role of energy metabolism

Xingzhi Guo,Yu-ying Yang,Rong Zhou,Ge Tian,Chang Shan,Jian-min Liu,Rui Li

DOI: https://doi.org/10.1038/s41398-024-02924-w

2024-05-21

Translational Psychiatry

Abstract:Growing evidence suggests an association between osteocalcin (OCN), a peptide derived from bone and involved in regulating glucose and lipid metabolism, and the risk of Alzheimer's disease (AD). However, the causality of these associations and the underlying mechanisms remain uncertain. We utilized a Mendelian randomization (MR) approach to investigate the causal effects of blood OCN levels on AD and to assess the potential involvement of glucose and lipid metabolism. Independent instrumental variables strongly associated ( P < 5E-08) with blood OCN levels were obtained from three independent genome-wide association studies (GWAS) on the human blood proteome ( N = 3301 to 35,892). Two distinct summary statistics datasets on AD from the International Genomics of Alzheimer's Project (IGAP, N = 63,926) and a recent study including familial-proxy AD patients (FPAD, N = 472,868) were used. Summary-level data for fasting glucose (FG), 2h-glucose post-challenge, fasting insulin, HbA1c, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, total cholesterol (TC), and triglycerides were incorporated to evaluate the potential role of glucose and lipid metabolism in mediating the impact of OCN on AD risk. Our findings consistently demonstrate a significantly negative correlation between genetically determined blood OCN levels and the risk of AD (IGAP: odds ratio [OR, 95%CI] = 0.83[0.72–0.96], P = 0.013; FPAD: OR = 0.81 [0.70–0.93], P = 0.002). Similar estimates with the same trend direction were obtained using other statistical approaches. Furthermore, employing multivariable MR analysis, we found that the causal relationship between OCN levels and AD was disappeared after adjustment of FG and TC (IGAP: OR = 0.97[0.80–1.17], P = 0.753; FPAD: OR = 0.98 [0.84–1.15], P = 0.831). There were no apparent instances of horizontal pleiotropy, and leave-one-out analysis showed good stability of the estimates. Our study provides evidence supporting a protective effect of blood OCN levels on AD, which is primarily mediated through regulating FG and TC levels. Further studies are warranted to elucidate the underlying physio-pathological mechanisms.

psychiatry