Table 5.

Datasets, data analysis methods, and key findings for selected studies in research foci 3 - stratifying risks for AD.

Study	Datasets	Primary data analysis methods	Key findings
Gracia- Garcia et al. (34,36)	3,864 participants from ZARADEMP	Multivariate model	Severe depression increases the risk of AD
Li et al. (50)	Clinical data from EHR (212 AD, 15040 HC)	Statistical analysis (Chi-square test and Mann-Whitney U test)	Erythrocyte sedimentation rate (ESR) is a significantly associated with AD.
Chang et al. (51)	879 asymptomatic higher risk persons (with parental family history of AD) and 355 asymptomatic lower risk persons (without parental family history of AD) from WRAP	Aggregate measure using Euclidean distance	Finer differences in memory strategy measured by machine learning method can be used as a potential AD risk factor.
Rosenberg et al. (52)	1821 MCI (527 PMCI, 454 SMCI) from National Alzheimer’s Coordinating Center database	Cox proportionality hazard model	Neuropsychiatric symptoms in MCI are associated with significantly increase of incident dementia and AD.
Yasar et al. (53)	320 MCI and 1928 HC from Ginkgo Evaluation of Memory study	Cox proportional hazard model	The use of diuretic, angiotensin-1 receptor blocker (ARB), and angiotensin-converting enzyme inhibitors (ACE-I) was associated with reduced AD risk for healthy normal. Diuretic use associated with reduced AD risk in MCI patients.

Note: ZARADEMP, Zaragoza Dementia and Depression Project; WRAP, Wisconsin Registry for Alzheimer’s Prevention; MCI, mild-cognitive impairment; SMCI, stable mild cognitive impairment; PMCI, progressive mild cognitive impairment; HC, healthy control.