RSNA 2012 

Abstract Archives of the RSNA, 2012


SSA16-07

The Value of Structural MRI Measurements of Cerebral Atrophy in Predicting the Rate of Cognitive Decline in the Non-demented Elderly. A GEE Analysis Based on Data from the Vienna Transdanube Aging (VITA) Study

Scientific Formal (Paper) Presentations

Presented on November 25, 2012
Presented as part of SSA16: ISP: Neuroradiology (Aging and Cognition)

Participants

Martin Roesel MD, Presenter: Nothing to Disclose
Manuel Gmeiner MD, Abstract Co-Author: Nothing to Disclose
Andreas Nachbagauer BSc, Abstract Co-Author: Nothing to Disclose
Wolfgang Krampla MD, Abstract Co-Author: Nothing to Disclose
Peter Fischer, Abstract Co-Author: Nothing to Disclose
Walter F. Hruby MD, Abstract Co-Author: Nothing to Disclose
Karl-Heinz Tragl, Abstract Co-Author: Nothing to Disclose

PURPOSE

To determine if baseline measurements of cerebral atrophy predict the rate of future cognitive decline in the non-demented elderly.  

METHOD AND MATERIALS

Data were drawn from the Vienna Transdanube Aging (VITA) study. Magnetic resonance images of 532 non-demented subjects aged 75-76 years at baseline were analyzed to assess 7 cerebral atrophy markers, including hippocampal head and tail measurements for both sides and three ventricular indices. A population averaged model with the corresponding analytical technique of generalized estimating equations (GEE) were applied to the birth-cohort to determine whether baseline MRI measurements predicted the rate of cognitive decline at three follow-up investigations after 30, 60 and 90 months. Cognitive tests were comprised of the Mini Mental State Examination (MMSE), Boston Naming Test (BNT), Fuld Object Memory Evaluation (FOME) and the Trail Making Tests A and B (TMTA, TMTB). Vascular risk factors and several other covariates that were available in the VITA database were included as additional predictors in the longitudinal data analysis. 

RESULTS

Severity of right hippocampal head atrophy predicted the rate of cognitive decline in all 5 test scores. (MMSE: β = −0.310; p = 0.033, BNT: β = −0.219; p = 0.001, FOME: β = −1.354; p < 0.001, TMTA: Exp(β) = 0.185; p = 0.009, TMTB: Exp(β) = 1.149; p < 0.001); severity of left hippocampal head atrophy followed closely, predicting decline in 4 out of 5 test scores. The degree of ventricular atrophy assessed using the nucleus caudatus index predicted cognitive decline in MMSE scores (β = −0.287; p < 0.001) and FOME scores (β = −0.614; p < 0.001); the cella media index predicted the rate of cognitive decline in MMSE scores (β = −0.374; p < 0.001) and FOME scores (β = −0.732; p = 0.003).

CONCLUSION

The degree of hippocampal head atrophy is the strongest predictor of cognitive decline in the VITA birth-cohort showing statistically significant associations in 9 out of 10 hippocampal models. Moreover, generalized estimating equation models demonstrated, that lower educational level, presence of the Apo E ε4 allele, being male, positive history of depression, higher scores on the short geriatric depression scale and the degree of atrophy assessed with described markers have a synergistic effect on future cognitive decline.

CLINICAL RELEVANCE/APPLICATION

The degree of cerebral atrophy is associated with the rapidity of cognitive decline in the non-demented elderly.

Cite This Abstract

Roesel, M, Gmeiner, M, Nachbagauer, A, Krampla, W, Fischer, P, Hruby, W, Tragl, K, The Value of Structural MRI Measurements of Cerebral Atrophy in Predicting the Rate of Cognitive Decline in the Non-demented Elderly. A GEE Analysis Based on Data from the Vienna Transdanube Aging (VITA) Study.  Radiological Society of North America 2012 Scientific Assembly and Annual Meeting, November 25 - November 30, 2012 ,Chicago IL. http://archive.rsna.org/2012/12030882.html