Single-photon emission computed tomography (SPECT) perfusion and neuropsychological performance in mild cognitive...

Single-photon emission computed tomography (SPECT) is an affordable neuroimaging technique that measures cerebral perfusion and has been utilized repeatedly in aging populations. However, we are aware of no studies to date examining relationships between SPECT imaging and comprehensive neuropsychological evaluations in a clinical sample of patients with mild cognitive impairment (MCI). Participants were 124 older adults with MCI (age, M = 75.07 years, SD = 7.65; years of education, M = 14.03, SD = 3.09; 60.2% female) who underwent neuropsychological evaluations and brain SPECT scans as part of their routine clinical care. Based on SPECT interpretations, participants were grouped by suspected etiology (i.e., the neuroradiologists noted that hypoperfusion patterns were most consistent with Alzheimer's disease, AD; frontotemporal lobar degeneration, FTLD; or other disease processes) and regional hypoperfusion (e.g., frontal, temporal, right/left hemisphere). Neuropsychological tests were grouped into domain scores (i.e., attention/processing speed, language, visuospatial, memory, executive; verbal/nonverbal). Consistent with a priori predictions, patients with an AD pattern of hypoperfusion scored lower than comparison groups on the attention/processing speed (partial χ² = 0.10) and memory (partial χ² = 0.07) composites. More patients with the AD-hypoperfusion signal met criteria for amnestic MCI (82%) than did those with a non-AD pattern (70%); this result approached statistical significance (p = .07). Contrary to hypotheses, patients whose SPECT scans were most consistent with FTLD did not underperform on the executive composite, and most regional analyses were nonsignificant. When integrating SPECT data into their clinical conceptualizations of MCI, neuropsychologists should place more weight on AD patterns of hypoperfusion, while de-emphasizing data suggestive of FTLD or regional pathology. Alternative neurodiagnostic markers may be more informative in these instances.