ABSTRACT Changes in emotion and social behavior are early yet overlooked features of Alzheimer's disease (AD). In AD, there is progressive accumulation of beta amyloid (Aβ) and tau proteins, a pathological process that leads to neurodegeneration and functional connectivity alterations in distributed brain networks. Episodic memory decline is a hallmark feature of typical amnestic AD presentations, but AD can also cause atypical dysexecutive/amnestic, language (i.e., logopenic variant primary progressive aphasia [lvPPA]), and visuospatial (i.e., posterior cortical atrophy [PCA]) syndromes. In typical AD, default mode network dysfunction is accompanied by elevated functional connectivity in the salience network, a system that supports emotion generation and interoception. Enhanced salience network connectivity in AD is associated with greater affective symptoms such as anxiety. Our previous studies have found that specific types of emotional empathy, a rudimentary form of affect-sharing, are elevated in typical AD and relate to default mode network atrophy. Whether emotion elevations are present in atypical AD syndromes is unknown but are suggested by clinical and neuroimaging data. There is emerging evidence that gains in emotional empathy are evident even in preclinical AD, a prodromal phase in which people have signs of AD pathology on biomarker testing but lack cognitive symptoms. A central hypothesis of this proposal is that early neuropathology and neurodegeneration in AD-vulnerable networks disinhibits the salience network and elevates emotion functioning across AD syndromes. A more detailed understanding of emotion in AD will help to improve diagnosis by broadening current conceptualizations of each syndrome, to identify emotion measures that suggest the presence of early AD, and to uncover new biomarkers that change as neuropathology affects emotion-relevant brain networks. In the proposed studies, we will conduct laboratory-based assessments of emotion (i.e., autonomic nervous system activity, facial behavior, and subjective experience) in 200 people with AD, as indicated by elevated Aβ (Aβ+) and tau deposition on molecular PET scans (110 amnestic/dysexecutive AD, 45 lvPPA, and 45 PCA), and 150 older healthy controls with and without AD pathology (75 Aβ+ and 75 Aβ-) with varying levels of tau pathology. By relating emotion measures to clinical data, structural and functional connectivity, Aβ and tau burden, and affective symptoms, we will address three key aims. In Aim 1, we will quantify emotion, empathy, and social behavior in AD syndromes and Aβ+ healthy controls. In Aim 2, we will delineate the structural and functional neural networks underlying emotion alterations across the AD spectrum. In Aim 3, we will elucidate associations among Aβ and tau pathology, emotion system functioning, and affective symptoms. By forging links among measures of emotion, neural network dysfunction, affective symptoms, and Aβ and tau deposition, the propo...