Approximately 30 million people currently suffer from late-onset Alzheimer's disease (LOAD) worldwide. Twin studies demonstrated that 60 to 80% of LOAD is genetically determined, 20% of which remaining unassigned. This case-control study included 118 cognitively healthy controls, 52 patients with mild cognitive impairment (MCI; the pre-stage of LOAD) and 71 LOAD patients. The participants were genotyped for the genetic LOAD marker apolipoprotein E4 (APOE4) and the single-nucleotide polymorphism rs4925 in glutathione S-transferase omega-1 (GSTO1). Additive logistic regression showed a novel, statistically significant association of the major allele GSTO1*C with MCI (OR1.9; p = 0.032). However, identification of significant SNP-disease relations required well-defined study groups. When classifying participants solely by the short Mini Mental State examination (MMSE), the associations of GSTO1*C and the reference marker APOE4 with MCI were cancelled. Moreover, even identifying only the control group by MMSE nullified a statistically significant association (OR1.8; p = 0.045) between GSTO1*C and LOAD. In contrast, these statistical relations were retained when the detailed Consortium to Establish a Registry for Alzheimer's Disease (CERAD-Plus) test battery was used. Hence, besides proposing rs4925 as a genetic marker for cognitive impairment, this work also emphasized the importance of carefully characterized controls in addition to well-diagnosed patients in case-control studies.
Keywords: Alzheimer’s disease; CERAD-Plus; Gerotarget; MCI; MMSE; well-defined control group.