*4.7. Behavioral Correlates Mental Health and Cardiovascular Measurements*

Correlation analysis through all the components suggests functional interactions between NPS and cognitive impairment with amygdala and HPA axis. The vascular function correlated with activity, while the left and right hemispheres' asymmetry of rCBF with NPS and cognitive impairments. Increased activation of the amygdala in 3xTg-AD mice [55] could explain the specific correlation between amygdala CBF and body weight. Corticosterone level correlations agree with our first report on the increase in glucocorticoid levels concomitantly to increased anxiety and peripheral immune dysfunction [12] and recent work of other laboratories [77]. Systolic blood pressure correlated positively with the number of half-buried marbles and negatively correlated with latency to cross the T-intersection and TM test, which are indicators of a worse neuro-immunoendocrine function, accelerated aging, and premature death in mice [53]. These correlates agree with mice chronically subjected to high blood pressure being more active in the openfield and faster in a spontaneous alternation test [78]. Recently, we demonstrated the correlation between peripheral small vessel properties and anxiety in both NTg and 3xTg-AD mice, mostly in females [21]. Here, we show that the animals with lower endothelialindependent vasodilatations to sodium nitroprusside (i.e., lower muscle relaxation capacity) were the most active in the open-field test, in both non-goal- (horizontal activity) and goaldirected (rearing) behaviors [79,80]. The correlation was consistent, as also observed in the repeated test**.** Regarding relative cerebral blood flow and neuropsychiatric and cognitive impairment, the cortex was the better area correlated with behavior. Cortex hemisphere asymmetry correlated with risk assessment (stretch attendance), neophobia in the corner test, and worse memory performance in the T-maze test. Correlation between asymmetry and neophobia was also observed in the hippocampus and amygdala. Since testing the pairwise correlations between dozens of variables without multiple comparison correction may involve a high risk of type-1 error, only those which are meaningful, obtained a maximum statistical significance (p < 0.001), and could be verified with closerelated variables were considered. Despite our aim to highlight and explore the sex-based differences, the statistical power of the sample size per sex did not allow to compare whether the correlations between these variables were different between male and female mice and thus modulated by the sex. Overall, this correlation analysis allows for finding meaningful functional correlations between behavioral responses related to the levels of anxiety, cognition and locomotor activity, and cardiovascular measurements. Especially, the present work provides pieces of evidence of the brain regions' asymmetry of both males and females with normal and AD-neurodegenerative aging correlation with neuropsychiatric symptoms and cognitive deficits.
