Predictive Factors for Verbal Memory Performance Over Decades of Aging: Data from the Women's Healthy Ageing Project

Am J Geriatr Psychiatry. 2016 Oct;24(10):857-67. doi: 10.1016/j.jagp.2016.05.008. Epub 2016 May 18.


Background: Abnormalities in brain structure and function can occur several decades prior to the onset of cognitive decline. It is in the preceding decades that an intervention is most likely to be effective, when informed by an understanding of factors contributing to the disease prodrome. Few studies, however, have sufficient longitudinal data on relevant risks to determine the optimum targets for interventions to improve cognition in aging. In this article we examine the timing and exposure of factors contributing to verbal memory performance in later life.

Methods: 387 participants from the population-based Women's Healthy Ageing Project, mean age at baseline of 49.6 years (range: 45-55 years), had complete neuropsychiatric assessments, clinical information, physical measures, and biomarkers collected at baseline, with at least three follow-up visits that included at least one cognitive reassessment. Mixed linear models were conducted to assess the significance of risk factors on later-life verbal memory. We explored the influence of early, contemporaneous, and cumulative exposures.

Results: Younger age and better education were associated with baseline memory test performance (CERAD). Over the 20 years of study follow-up, cumulative mid- to late-life physical activity had the strongest effect on better later life verbal memory (0.136 [0.058, 0.214]). The next most likely contributors to verbal memory in late life were the negative effect of cumulative hypertension (-0.033 [-0.047, -0.0.18] and the beneficial effect of HDL cholesterol (0.818 [0.042, 1.593]).

Conclusions: Findings suggest that midlife interventions focused on physical activity, hypertension control, and achieving optimal levels of HDL cholesterol will help maintain later-life verbal memory skills.

Keywords: cognition; life course research; longitudinal prospective; memory; modifiable risk factors; older persons.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Age Factors
  • Aged
  • Aging / psychology*
  • Cholesterol, HDL / blood*
  • Educational Status
  • Exercise*
  • Female
  • Humans
  • Hypertension / epidemiology*
  • Linear Models
  • Longitudinal Studies
  • Memory*
  • Middle Aged
  • Neuropsychological Tests
  • Protective Factors
  • Risk Factors


  • Cholesterol, HDL