Limits...
Microstate connectivity alterations in patients with early Alzheimer's disease.

Hatz F, Hardmeier M, Benz N, Ehrensperger M, Gschwandtner U, Rüegg S, Schindler C, Monsch AU, Fuhr P - Alzheimers Res Ther (2015)

Bottom Line: Networks were reduced to 22 nodes for statistical analysis.The domain score for verbal learning and memory and the microstate segmented PLI between the left centro-lateral and parieto-occipital regions in the theta band at baseline differentiated significantly between the groups.Combining neuropsychological and quantitative EEG test results allows differentiation between subjects with aMCI remaining stable and subjects with aMCI deteriorating over 30 months.

View Article: PubMed Central - PubMed

Affiliation: Department of Neurology, University Hospital of Basel, Petersgraben 4, 4031, Basel, Switzerland. florian.hatz@usb.ch.

ABSTRACT

Introduction: Electroencephalography (EEG) microstates and brain network are altered in patients with Alzheimer's disease (AD) and discussed as potential biomarkers for AD. Microstates correspond to defined states of brain activity, and their connectivity patterns may change accordingly. Little is known about alteration of connectivity in microstates, especially in patients with amnestic mild cognitive impairment with stable or improving cognition within 30 months (aMCI).

Methods: Thirty-five outpatients with aMCI or mild dementia (mean age 77 ± 7 years, 47% male, Mini Mental State Examination score ≥24) had comprehensive neuropsychological and clinical examinations. Subjects with cognitive decline over 30 months were allocated to the AD group, subjects with stable or improving cognition to the MCI-stable group. Results of neuropsychological testing at baseline were summarized in six domain scores. Resting state EEG was recorded with 256 electrodes and analyzed using TAPEEG. Five microstates were defined and individual data fitted. After phase transformation, the phase lag index (PLI) was calculated for the five microstates in every subject. Networks were reduced to 22 nodes for statistical analysis.

Results: The domain score for verbal learning and memory and the microstate segmented PLI between the left centro-lateral and parieto-occipital regions in the theta band at baseline differentiated significantly between the groups. In the present sample, they separated in a logistic regression model with a 100% positive predictive value, 60% negative predictive value, 100% specificity and 77% sensitivity between AD and MCI-stable.

Conclusions: Combining neuropsychological and quantitative EEG test results allows differentiation between subjects with aMCI remaining stable and subjects with aMCI deteriorating over 30 months.

No MeSH data available.


Related in: MedlinePlus

Results of logistic regression comparing Alzheimer’s disease (AD) vs. patients with stable or improving cognition within 30 months (MCI-stable). First, a score was derived using a logistic regression model for the AD and healthy control (HC) groups, then the resulting score was applied to separate the AD group from the MCI-stable group. Score = (10.4 x msPLI) - (3.52 x VLM); a score greater than or equal to −13.2 had sensitivity 76.9 %, specificity 100 %, positive predictive value 100 %, and negative predictive value 60 %. AUC area under the curve, msPLI microstate segmented phase lag index log of link centrolateral to parieto-occipital left, VLM domain score for verbal learning and memory)
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC4697314&req=5

Fig7: Results of logistic regression comparing Alzheimer’s disease (AD) vs. patients with stable or improving cognition within 30 months (MCI-stable). First, a score was derived using a logistic regression model for the AD and healthy control (HC) groups, then the resulting score was applied to separate the AD group from the MCI-stable group. Score = (10.4 x msPLI) - (3.52 x VLM); a score greater than or equal to −13.2 had sensitivity 76.9 %, specificity 100 %, positive predictive value 100 %, and negative predictive value 60 %. AUC area under the curve, msPLI microstate segmented phase lag index log of link centrolateral to parieto-occipital left, VLM domain score for verbal learning and memory)

Mentions: A binary logistic regression with the most significant results of the six domain scores at baseline and the frequency and connectivity analysis as well as with AD vs. HC as outcomes was calculated. After stepwise backward elimination, only the connectivity between left centrolateral and parieto-occipital regions (theta band) and verbal learning and memory remained in the model and were used for calculating a score. This score was used in a second logistic regression model in which we compared AD with MCI-stable. The ROC curve showed an area under the curve (AUC) of 0.90 (maximal Youden index sensitivity 77 %, specificity 100 %, positive predictive value 100 %, negative predictive value 60 %) (Fig. 7). Relative theta power was not part of the final model. Adding relative theta power as a third variable increased the AUC only minutely.Fig. 7


Microstate connectivity alterations in patients with early Alzheimer's disease.

Hatz F, Hardmeier M, Benz N, Ehrensperger M, Gschwandtner U, Rüegg S, Schindler C, Monsch AU, Fuhr P - Alzheimers Res Ther (2015)

Results of logistic regression comparing Alzheimer’s disease (AD) vs. patients with stable or improving cognition within 30 months (MCI-stable). First, a score was derived using a logistic regression model for the AD and healthy control (HC) groups, then the resulting score was applied to separate the AD group from the MCI-stable group. Score = (10.4 x msPLI) - (3.52 x VLM); a score greater than or equal to −13.2 had sensitivity 76.9 %, specificity 100 %, positive predictive value 100 %, and negative predictive value 60 %. AUC area under the curve, msPLI microstate segmented phase lag index log of link centrolateral to parieto-occipital left, VLM domain score for verbal learning and memory)
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC4697314&req=5

Fig7: Results of logistic regression comparing Alzheimer’s disease (AD) vs. patients with stable or improving cognition within 30 months (MCI-stable). First, a score was derived using a logistic regression model for the AD and healthy control (HC) groups, then the resulting score was applied to separate the AD group from the MCI-stable group. Score = (10.4 x msPLI) - (3.52 x VLM); a score greater than or equal to −13.2 had sensitivity 76.9 %, specificity 100 %, positive predictive value 100 %, and negative predictive value 60 %. AUC area under the curve, msPLI microstate segmented phase lag index log of link centrolateral to parieto-occipital left, VLM domain score for verbal learning and memory)
Mentions: A binary logistic regression with the most significant results of the six domain scores at baseline and the frequency and connectivity analysis as well as with AD vs. HC as outcomes was calculated. After stepwise backward elimination, only the connectivity between left centrolateral and parieto-occipital regions (theta band) and verbal learning and memory remained in the model and were used for calculating a score. This score was used in a second logistic regression model in which we compared AD with MCI-stable. The ROC curve showed an area under the curve (AUC) of 0.90 (maximal Youden index sensitivity 77 %, specificity 100 %, positive predictive value 100 %, negative predictive value 60 %) (Fig. 7). Relative theta power was not part of the final model. Adding relative theta power as a third variable increased the AUC only minutely.Fig. 7

Bottom Line: Networks were reduced to 22 nodes for statistical analysis.The domain score for verbal learning and memory and the microstate segmented PLI between the left centro-lateral and parieto-occipital regions in the theta band at baseline differentiated significantly between the groups.Combining neuropsychological and quantitative EEG test results allows differentiation between subjects with aMCI remaining stable and subjects with aMCI deteriorating over 30 months.

View Article: PubMed Central - PubMed

Affiliation: Department of Neurology, University Hospital of Basel, Petersgraben 4, 4031, Basel, Switzerland. florian.hatz@usb.ch.

ABSTRACT

Introduction: Electroencephalography (EEG) microstates and brain network are altered in patients with Alzheimer's disease (AD) and discussed as potential biomarkers for AD. Microstates correspond to defined states of brain activity, and their connectivity patterns may change accordingly. Little is known about alteration of connectivity in microstates, especially in patients with amnestic mild cognitive impairment with stable or improving cognition within 30 months (aMCI).

Methods: Thirty-five outpatients with aMCI or mild dementia (mean age 77 ± 7 years, 47% male, Mini Mental State Examination score ≥24) had comprehensive neuropsychological and clinical examinations. Subjects with cognitive decline over 30 months were allocated to the AD group, subjects with stable or improving cognition to the MCI-stable group. Results of neuropsychological testing at baseline were summarized in six domain scores. Resting state EEG was recorded with 256 electrodes and analyzed using TAPEEG. Five microstates were defined and individual data fitted. After phase transformation, the phase lag index (PLI) was calculated for the five microstates in every subject. Networks were reduced to 22 nodes for statistical analysis.

Results: The domain score for verbal learning and memory and the microstate segmented PLI between the left centro-lateral and parieto-occipital regions in the theta band at baseline differentiated significantly between the groups. In the present sample, they separated in a logistic regression model with a 100% positive predictive value, 60% negative predictive value, 100% specificity and 77% sensitivity between AD and MCI-stable.

Conclusions: Combining neuropsychological and quantitative EEG test results allows differentiation between subjects with aMCI remaining stable and subjects with aMCI deteriorating over 30 months.

No MeSH data available.


Related in: MedlinePlus