Differential longitudinal decline on the Mini-Mental State Examination in frontotemporal lobar degeneration and Alzheimer disease

Alzheimer Dis Assoc Disord. 2013 Oct-Dec;27(4):310-5. doi: 10.1097/WAD.0b013e31827bdc6f.


Objective: To examine how phenotype affects longitudinal decline on the Mini-Mental State Examination (MMSE) in patients with frontotemporal lobar degeneration (FTLD) and Alzheimer disease (AD).

Background: The MMSE is the most commonly administered assessment for dementia severity; however, the effects of phenotype on longitudinal MMSE performance in FTLD and AD have not been extensively studied.

Methods: Data from 185 patients diagnosed with AD (n=106) and 3 FTLD (n=79) phenotypes [behavioral variant frontotemporal dementia (bvFTD), nonfluent agrammatic variant of primary progressive aphasia (nfaPPA), and semantic variant PPA (svPPA)] were collected for up to 52 months since initial evaluation.

Results: Differential rates of decline were noted in that MMSE scores declined more precipitously for AD and svPPA compared with bvFTD and nfaPPA patients (P=0.001). The absolute 4-year MMSE decline given median baseline MMSE for bvFTD [14.67; 95% confidence interval (CI), 14.63-14.71] and nfaPPA (11.02; 95% CI, 10.98-11.06) were lower than svPPA (22.32; 95% CI, 22.29-22.34) or AD (22.24; 95% CI, 22.22-22.26).

Conclusions: These data suggest that within-group AD and FTLD phenotypes present distinct patterns of longitudinal decline on the MMSE. MMSE may not be adequately sensitive to track disease progression in some phenotypes of FTLD.

Publication types

  • Research Support, N.I.H., Extramural

MeSH terms

  • Aged
  • Alzheimer Disease / diagnosis*
  • Alzheimer Disease / psychology*
  • Brief Psychiatric Rating Scale*
  • Cohort Studies
  • Disease Progression*
  • Female
  • Follow-Up Studies
  • Frontotemporal Lobar Degeneration / diagnosis*
  • Frontotemporal Lobar Degeneration / psychology*
  • Humans
  • Longitudinal Studies
  • Male
  • Middle Aged
  • Neuropsychological Tests
  • Single-Blind Method