Progression to Dementia or Reversion to Normal Cognition in Mild Cognitive Impairment as a Function of Late-Onset Neuropsychiatric Symptoms

BACKGROUND AND OBJECTIVES: Mild cognitive impairment (MCI) is an at-risk state for dementia; however, not all individuals with MCI transition to dementia, and some revert to normal cognition (NC). Here, we investigate whether mild behavioral impairment (MBI), the late-life onset of persistent neuropsychiatric symptoms (NPS), improves the prognostic specificity of MCI. METHODS: Participants with MCI from the National Alzheimer's Coordinating Center Uniform Data Set were included. NPS were operationalized with the Neuropsychiatric Inventory Questionnaire to identify participants without NPS and those with MBI (persistent, late-onset NPS). Individuals with late-onset NPS not meeting the MBI persistence criterion (NPS_NOT_MBI) were retained for secondary analyses. Progression to dementia, stable MCI, and reversion to NC after 3 years of follow-up were defined per National Institute on Aging–Alzheimer’s Association and Petersen criteria. RESULTS: The primary sample consisted of 739 participants (NPS− n = 409 and MBI+ n = 330; 75.16 ± 8.6 years old, 40.5% female). After 3 years, 238 participants (33.6%) progressed to dementia, and 90 (12.2%) reverted to NC. Compared to participants without NPS, participants with MBI were significantly more likely to progress to dementia (adjusted odds ratio [AOR] 2.13, 95% CI 1.52–2.99), with an annual progression rate of 14.7% (vs 8.3% for participants with MCI without NPS). Compared to participants without NPS, participants with MBI were less likely to revert to NC (AOR 0.48, 95% CI 0.28–0.83, 2.5% vs 5.3% annual reversion rate). The NPS_NOT_MBI group (n = 331, 76.5 ± 8.6 years old, 45.9% female) were more likely to progress to dementia (AOR 2.18, 95% CI 1.56–3.03, 14.3% annual progression rate) but not less likely to revert to NC than those without NPS. Accordingly, both NPS_NOT_MBI and MBI+ participants had lower Mini-Mental State Examination scores than NPS− participants after 3 years. DISCUSSION: Late-onset NPS improve the specificity of MCI as an at-risk state for progression to dementia. However, only persistent late-onset NPS are associated with a lower likelihood of reversion to NC, with transient NPS (i.e., NPS_NOT_MBI) not differing from the NPS− group. Clinical prognostication can be improved by incorporating late-onset NPS, especially those that persist (i.e., MBI), into risk assessments. Clinical trials may benefit from enrichment with these higher-risk participants with MCI.