ERS3016. Brainstem Atrophy Across the AD Spectrum and its Correlation with Markers of Cognitive Decline, Sleep, and Mood Disturbance
Authors * Denotes Presenting Author
  1. Deepali Bhalla *; Medical College of Wisconsin-Milwaukee
  2. Barbara Bendlin; Wisconsin Alzheimer's Disease Research Center
  3. Elias Granadillo; Medical College of Wisconsin-Milwaukee
Alzheimer’s disease (AD) is the most common dementia that causes progressive problems with memory, critical thinking, and behavior. Studies have shown the brainstem displays the earliest signs of AD progression. Midbrain structural changes are linked to AD symptoms i.e., memory impairment, sleep disorders, and emotional disturbances. Neuroreader (NR), an MRI automated volumetric assessment, is emerging as an inexpensive, and easily accessible alternative to imaging techniques used previously such as neuromelanin-sensitive MRI. The purpose of this study is to investigate changes in brainstem volumes (BS-V) and elucidate the differences in BS atrophy patterns across the AD spectrum by utilizing NR, which to our knowledge is the first study evaluating such changes. Potential correlations between BS-V and biomarkers of cognition, sleep, and mood disturbances were also evaluated.

Materials and Methods:
We conducted a retrospective chart review on 60 AD spectrum patients who had NR scans done after 2015 and who also underwent a sleep study procedure. We ran preliminary assessments to test for differences in BS-V and Hippocampal volumes (Hip-V) among the 3 groups with 20 patients each: Subjective Cognitive impairment (SCI), Mild Cognitive Impairment (MCI), and AD, and collected data on three separate markers: sleep, mood, and cognition. We ran multiple regression models, corrected for age and gender, to determine group-wise differences in BS-V and Hip-V with SCI set as the comparison group. We also performed Pearson correlations between BS-V, sleep biomarkers, and cognitive test scores.

No difference in raw BS-V or BS-V corrected for total intracranial volume (BS-VM) across AD spectrum compared to SCI was found. Hip-V difference between AD and SCI remained significant after adjustment for age (p=0.014), and trended toward significance after adjusting for age and gender (p=0.081) and total intracranial volume (p=0.061); confirming the prior utility of NR for the measurement of hippocampi in AD. Female gender was a significant predictor of lower raw brainstem volumes (ß =-0.382, p=0.004) but when corrected for total intracranial volume (BS-VM), male gender was associated with lower brainstem volumes (ß =-0.356, p=0.009). Similarly, female gender was also a significant predictor of lower raw hippocampal volumes (ß =-0.403, p=0.002). This is consistent with AD literature suggesting a differential susceptibility to AD pathology based on gender and needs to be further explored (2). Interestingly, the inverse correlation between sleep efficiency percentages and BS-V seems to go contrary to the prevailing literature and needs further investigation.

Significant results obtained from this study can be a step towards standardizing the brainstem atrophy rate in AD patients, thus serving as a non-invasive clinical biomarker to provide early detection signs of AD progression and of common comorbidities like insomnia and mood disorders; leading to earlier diagnosis, appropriate change in care, and possible treatments of patients.