Background: Studies of cognitive impairment (CI) in Amish communities have identified sibships containing multiple CI and cognitively unimpaired (CU; unaffected after age 75) individuals. We hypothesize that these CU individuals may carry protective alleles delaying age at onset (AAO) of CI, preserving cognition in older age despite increased genetic risk. As well, the genetic and cultural isolation in the Amish since the early 1800s may have reduced the complexity of the genetic architecture of CI, increasing the power to detect protective alleles in this population. With this in mind we conducted a genome-wide study (GWAS) to identify loci associated with AAO of CI in a sample of Amish adults over age 75. Methods: 1,522 individuals aged 43-99 (mean age 73.1, 42% men) screened at least once for CI using the Modified Mini-Mental State exam (3MS) were genotyped using Illumina chipsets. Genotypes were imputed for 7,815,951 single nucleotide variants (SNV) with minor allele frequency (MAF) > 1%. The outcome studied was age, defined as 1) age at the first 3MS result indicating impairment (AAO; 3MS <87; 362 CI individuals) or 2) age at last normal exam (3MS >=87, 1,160 CU individuals). Cox mixed-effects models examined association between age and each SNV, adjusting for sex and familial relationships. To replicate genome-wide significant findings, SNVs in a 1 Megabase region centered on the peak SNV were examined for association with age using these same methods in the NIA-LOAD family study dataset (1,785 AD cases, 1,565 unaffected controls, mean age 73.5. Results: Three SNV were significantly associated (p<5 x 10-8) with AAO in the Amish, on chromosomes 6 (rs14538074; HR=3.35), 9 (rs534551495; HR=2.82), and 17 (rs146729640; Hazard Ratio (HR)=6.38). Each region found the common allele associated with later AAO. Replication analysis detected association at rs146729640, with nominal statistical significance (HR=1.49, p=0.02). Conclusions: The replicated genome-wide significant association with AAO on chromosome 17 suggest this may be novel locus associated with delayed onset of AD. The associated SNP is located in the SHISA6 gene, which is involved in post-synaptic transmission in the hippocampus and is a biologically plausible candidate gene for AD.

BackgroundWhile studying cognition in the Old Order Amish (OOA), we have observed strong performance on the constructional praxis delayed recall (CPDR) as compared to other cognitive tests, independent of overall cognitive status. This may indicate a preferential preservation of visuospatial memory in this population. Here, we investigate this by comparing the CPDR to the word list delayed recall (WLDR) within the OOA, as well as by comparing these results to a non-Amish cohort.Method420 OOA individuals in Indiana/Ohio age 66-95 who had complete data for the CPDR and WLDR were included. From the non-Amish CERAD cohort, 401 individuals age 60-96 with the same tests were included. For both cohorts, education-adjusted Z-scores were calculated for the CPDR and WLDR. The difference between the CPDR Z-score and the WLDR Z-score was calculated as a measure of the preservation of visuospatial memory over verbal memory. T-tests were first used to compare the tests within both cohorts and then stratified by case/control status. Linear regression was then used to investigate the effects of age, sex, cognitive status, and cohort on the Z-scores and difference between Z-scores. Additional t-tests and regressions were then performed to further investigate the effect of sex and its interaction with cohort.ResultWe found a significantly better performance on CPDR over WLDR in every cognitive status group in the OOA, but not in all groups of the CERAD cohort. After controlling for age, sex, and cognitive status, this preferential preservation remains significantly higher in the Amish, with being in the Amish cohort increasing the difference between Z-scores by an average of 0.615 units when compared to being in the CERAD cohort. When adjusting for age, sex, cognitive status, and cohort, the interaction between cohort and sex is significant, with the Amish males exhibiting a greater difference between Z-scores compared to other groups, with a significant interaction value of 0.676.DiscussionOverall, these findings suggest that the OOA preferentially preserve visuospatial memory over verbal memory, regardless of cognitive status. This effect is particularly strong in OOA males. In summary, this study gives additional evidence that the Amish exhibit unique patterns of memory loss and aging, with a preferential preservation of visuospatial memory over verbal memory. Additional studies are needed to further explain this phenomenon.