Cross‐Sectional Versus Longitudinal Estimates of Cognitive Change in Nondemented Older People: A CERAD Study

E-ISSN： 1532-5415|47|5|559-563

ISSN： 0002-8614

Source： JOURNAL OF AMERICAN GERIATRICS SOCIETY, Vol.47, Iss.5, 1999-05, pp. : 559-563

Disclaimer: Any content in publications that violate the sovereignty, the constitution or regulations of the PRC is not accepted or approved by CNPIEC.

Previous Menu Next

Abstract

OBJECTIVES: To investigate the effect of increasing age on cognition in nondemented older people.DESIGN: A cross‐sectional and longitudinal analysis.PARTICIPANTS: A total of 454 control subjects for Alzheimer's cases from the cohort assembled by the Consortium to Establish a Registry for Alzheimer's Disease (CERAD).MEASUREMENT: The Mini‐Mental State Examination (MMSE) to assess cognitive function.RESULTS: Cross‐sectional estimates were derived by generalized linear models and longitudinal estimates by generalized estimating equations. The cross‐sectional model indicated a small but significant decline in MMSE of ‐.4 points per 10 years. The longitudinal model indicated a small but significant increase in MMSE of about +.6 points per 10 years. Evidence of an early learning effect and nonrandom dropout exists.CONCLUSIONS: The question of “normal” aging can be approached by considering cross‐sectional information and, usually separately, longitudinal information. This study does both using recently developed statistical methods. We conclude that there is a small but significant decline in scores on the MMSE with increasing population age. The effect can be masked in longitudinal cohorts by a learning effect (especially early in follow‐up) and other factors associated with repeated testing. J Am Geriatr Soc 47:559–563, 1999.