(S-014) Survival Modeling of Dementia Risk Using Metabolic Syndrome-Related Genetic Variants

Seonghwa Song – College of Pharmacy – Chungnam National University; Sungwoo Goo – Department of Bio-AI convergence – Chungnam National University; Soyoung Lee – College of Pharmacy – Chungnam National University; Hwi-yeol Yun – College of Pharmacy – Chungnam National University; Jung-woo Chae – College of Pharmacy – Chungnam National University

Objectives: With the increasing prevalence of aging populations, there is growing interest in dementia, metabolic disorders, and obesity. This study was motivated by previous findings that evaluated the relationship between dementia severity, metabolic status, and obesity, with particular attention to the observation that individuals categorized as having Metabolically Healthy Obesity (MHO) had a lower incidence of dementia. In this context, we aimed to investigate the genetic association between APOE, genes involved in dementia risk and specific variants related to metabolic syndrome. This study investigated how inherited differences in metabolic profiles may contribute to differential vulnerability to dementia by characterizing inter-individual variability at the genetic level.

Methods: A genetic survival analysis was performed using cohort data from the Clinical Omics Data Archive (CODA). Metabolic syndrome-associated single nucleotide polymorphisms (SNPs) were identified via literature review, with variant details obtained from the NCBI dbSNP database. The impact of each SNP on dementia severity was evaluated using Weibull survival analysis [1]. To account for multiple SNPs concurrently, principal component analysis (PCA) [2] was applied to the genetic data, followed by Weibull survival analysis on the derived principal components.

Results: The Weibull survival model to individual SNPs identified two significant associations with dementia onset time. The rs9939609 variant (chr16:53786615)[3], known for its association with Type 2 Diabetes(T2D), was found to significantly accelerate the onset of dementia (p = 0.006). Conversely, the rs7756992 variant (chr6:20679478), also linked to T2D, showed a significant association with delayed dementia onset (p = 0.0418). To assess the combined impact of multiple SNPs, we employed PCA followed by Weibull survival analysis on the derived principal components (PCs). This approach revealed that only the first principal component (PC1) had a statistically significant effect on dementia onset time (p < 0.05). Further investigation showed that the rs9939609 variant strongly contributed to PC1, confirming its role in accelerating dementia onset within the context of broader genetic covariation captured by PCA. In contrast, the rs7756992 variant exhibited minimal influence on PC1, suggesting its individual effect might be less pronounced when considering the combined genetic landscape represented by the primary principal component.

Conclusions: Our survival analysis of Type 2 Diabetes (T2D)-associated SNPs within the CODA cohort demonstrated their influence on dementia onset timing. the rs9939609 variant significantly accelerated onset, confirmed via PCA highlighting its strong contribution to the primary genetic component affecting timing. This study moves beyond traditional risk association, using survival analysis to establish that T2D genetic factors not only associate with dementia presence but critically impact when the disease develops.

Citations: [1]Carroll, K. J. (2003). On the use and utility of the Weibull model in the analysis of survival data. Controlled clinical trials, 24(6), 682-701.
[2]Abdi, H., & Williams, L. J. (2010). Principal component analysis. Wiley interdisciplinary reviews: computational statistics, 2(4), 433-459.
[3]Fawwad, A., Siddiqui, I. A., Zeeshan, N. F., Shahid, S. M., & Basit, A. (2015). Association of SNP rs9939609 in FTO gene with metabolic syndrome in type 2 diabetic subjects, rectruited from a tertiary care unit of Karachi, Pakistan. Pakistan Journal of Medical Sciences, 31(1), 140.

Keywords: survival analisys,single nucleotide polymorphisms