Motter, Jeffrey N.

AbstractBackgroundInexpensive, non‐invasive tests may improve the identification of persons at increased risk for cognitive decline and dementia. We compared impairment in odor identification and global cognition with neuro‐imaging biomarkers to predict cognitive decline and dementia in the population‐based Mayo Clinic Study of Aging (MCSA).MethodAt the 2008 assessment, 647 participants who were ≥ 55 years old with at least one follow‐up had the following procedures: modified Blessed Information‐Memory‐Concentration Test (BIMCT), 12‐item Brief Smell Identification Test (BSIT), brain magnetic resonance imaging (MRI), and Positron Emission Tomography (PET) amyloid imaging with 11C‐Pittsburgh compound B (11C‐PiB). A subsample (n=556) also had 18F‐fluorodeoxyglucose (18F‐FDG) PET scans. The utility of these baseline measures to predict cognitive decline (cognitively unimpaired to mild cognitive impairment [MCI] or dementia, or MCI to dementia) was examined during follow‐up (mean 8.1 SD 3.4 years). Transition to dementia was a secondary outcome.ResultAmong 647 participants without dementia (mean age 74.0 SD 8.1 years; 54.4% male) at baseline, 102 participants manifested cognitive decline and 34 participants transitioned to dementia. In Cox regression survival analyses for cognitive decline, PiB PET showed the strongest predictive utility (hazard ratio HR 5.05, 95% CI = 3.32‐7.69, p<.001). Impaired BSIT (HR 3.46, 95% CI = 2.34‐5.11, p<.001), impaired BIMCT (HR 3.63, 95% CI 2.46‐5.36, p<.001), MRI and FDG measures were also significant predictors. After adjusting for age, sex and education, these predictors remained significant with attenuated effects (PiB HR 3.89, 95% CI 2.54‐5.95, p<.001; BSIT HR 2.77, 95% CI 1.85‐4.14, p<.001; BIMCT HR 3.3, 95% CI 2.21‐4.95, p<.001). After adding these demographic variables in the model, the combination of BSIT and BIMCT showed strong predictive utility (C‐index=0.85), similar to PiB PET (C‐index=0.84). Combining all predictors showed a C‐index of 0.89. These measures showed similar but stronger results for dementia transition.ConclusionTests of global cognition and odor identification provide a brief, non‐invasive, inexpensive method comparable to PET amyloid imaging for identifying community‐dwelling persons likely to decline cognitively or transition to dementia. Other potential uses for this combination include identification of persons likely to benefit from disease‐modifying medications and determination of eligibility for prevention trials.

AbstractBackgroundPrevious studies have linked impaired odor identification and global cognition with increased risk of cognitive decline and transition to dementia. However, the reverse question remains: if individuals have intact performance on these measures, are they at reduced risk for transition? We aimed to examine the accuracy of intact odor identification and global cognition for identifying lack of transition to dementia/cognitive decline using the population‐based Mayo Clinic Study of Aging and compare their accuracy against and in combination with amyloid PET (Positron Emission Tomography).Methodn = 647 participants age≥55 without dementia completed at baseline the Brief Smell Identification Test (BSIT; ‘Intact’ = 9‐12), Blessed Information‐Memory‐Concentration Test (BIMCT; ‘Intact’ = 18‐20,), and amyloid PET (‘Normal/Intact’ SUVR<1.48). We calculated sensitivity, specificity, positive predictive value, and negative predictive value for these measures, with “lack of transition to dementia” as the gold standard/target. “Lack of cognitive decline” (cognitively unimpaired to mild cognitive impairment [MCI]; cognitively unimpaired to dementia; or MCI to dementia) was a secondary target.Result94.7% of participants (613/647) did not transition to dementia and 84.2% (545/647) did not cognitively decline over 11.25‐year‐follow‐up. The combination of intact BSIT+BIMCT had high rates of lack of transition to dementia (98.3%, 355/361) and lack of cognitive decline (93.1%, 336/361). This combination yielded mixed sensitivity (0.579) and specificity (0.824) though had high positive predictive value (0.983) for identifying lack of transition to dementia. Intact amyloid PET alone exhibited comparable rates of lack of transition to dementia (99.0%, 418/423) and lack of cognitive decline (92.4%, 391/423). When evaluated conjointly, the combination of all three intact measures also had mixed sensitivity (0.421) and specificity (0.962) though higher positive predictive value (PPV = 0.996) for lack of transition to dementia. 99.6% (258/259) of participants with intact/normal status on all three did not transition to dementia and 96.5% (250/259) did not cognitively decline.ConclusionThis replicates prior reports that intact olfaction/global cognition have strong utility for identifying individuals unlikely to develop MCI/dementia. Intact/normal amyloid PET further enhances the positive predictive value. Clinically, identifying individuals at low risk of transition can shorten diagnostic workups, reduce need for early follow‐up visits, and screen patients in‐or‐out of clinical trials.

AbstractINTRODUCTIONWe evaluated impaired odor identification and global cognition as simple, cost‐effective alternatives to neuroimaging biomarkers to predict cognitive decline and dementia in the Mayo Clinic Study of Aging.METHODSSix hundred forty‐seven participants (mean 8.1, standard deviation 3.4 years’ follow‐up) had the following baseline procedures: modified Blessed Information Memory Concentration Test (BIMCT), 12‐item Brief Smell Identification Test (BSIT), structural brain magnetic resonance imaging (MRI), and positron emission tomography (PET) imaging with 11C‐Pittsburgh compound B (11C‐PiB) and fluorodeoxyglucose (FDG; subset).RESULTSCognitive decline developed in 102 participants and dementia in 34 participants. In survival analyses, PiB PET showed robust prediction for cognitive decline. Impaired BSIT, impaired BIMCT, MRI, and FDG measures were also significant predictors. The combination of demographics + BSIT + BIMCT showed strong predictive utility (C‐index 0.81), similar to demographics + PiB PET (C‐index 0.80). Similar but stronger results were obtained for prediction of dementia.DISCUSSIONImpairment in both odor identification test and global cognition was comparable to PiB PET for predicting cognitive decline and dementia.Highlights In 647 participants in the population‐based Mayo Clinic Study of Aging, several clinical markers and biomarkers each predicted cognitive decline or dementia during an average 8 years of follow‐up. The combination of the demographic variables of age, sex, and education with a brief odor identification test (BSIT) and a global cognitive test (Blessed Information Memory Concentration Test) showed strong predictive utility (C‐index 0.81) for cognitive decline that was similar to the demographic variables combined with Pittsburgh Compound B amyloid imaging (C‐index 0.80). Combining a brief odor identification test with a brief cognitive test needs consideration as a simple, cost‐effective option in the clinical assessment of individuals at risk of cognitive decline and dementia, as well as a potential tool to identify individuals who may benefit from disease‐modifying treatments and to screen participants for prevention trials.