How should clinicians manage patients with mild cognitive impairment?

Mild cognitive impairment (MCI) follows a variable course: not all patients progress to dementia. The data on pharmacologic interventions are limited and, so far, not very encouraging. Of the nonpharmacologic interventions, the best, albeit limited, data support exercise. (LOE = 5)

Petersen RC, Lopez O, Armstrong MJ, et al. Practice guideline update summary: Mild cognitive impairment: Report of the Guideline Development, Dissemination, and Implementation Subcommittee of the American Academy of Neurology. Neurology 2018;90(3):126-135.  [PMID:29282327]

Practice guideline

Foundation

Outpatient (any)

This guideline development panel was convened by the American Academy of Neurology (AAN). The panel used a systematic review of the literature to address 4 broad areas important to managing patients with MCI: prevalence, prognosis, pharmacologic management, and nonpharmacologic management. Overall, 62 studies informed the panel's recommendations. One of the challenges the panel faced is the variability in definition of MCI. Their best guess is that MCI becomes more prevalent with advancing age (approximately 7% in adults aged 60 years to 64 years; 8% for 65- to 69-year olds; 10% for 70- to 74-year olds; 15% for 75- to 79-year olds; and 25% for those 80 and older). Additionally, for adults older than 65 years with MCI, approximately 15% will develop dementia after 2 years. Among the few studies that report the natural history of MCI, between 15% and 38% of patients with MCI regress to normal. The panel identified 3 studies that assessed donepezil (Aricept), but could not pool data because of variable duration and variable outcome measures. These studies found little likelihood that donepezil prevents progression from MCI to Alzheimer dementia. Two studies assessed galantamine (Razadyne), and similarly failed to provide evidence that it slows progression to dementia. Only one trial evaluated rivastigmine (Exelon). After 4years the rate of progression from MCI to dementia was similar for placebo-treated patients. A single trial evaluated a flavonoid-containing beverage, but that study duration was only 8 weeks, so we can't really conclude much. Additional studies with limited data included homocysteine-lowering vitamin B, nicotine patches, piribedil (an anti-Parkinson agent sold under many trade names), rofecoxib (Vioxx), growth hormone–releasing hormones, and various other vitamins. Most studies either showed the treatments were ineffective, or were too short or too limited to have much faith in their results. The panel identified 7 studies of nonpharmacologic interventions. Only exercise seemed to show any short-term improvements. Cognitive interventions were too variable in scope, quality, and generalizability to recommend widespread adoption. Although this guideline was funded by the AAN, most of the panelists had financial ties to industry.