American Journal of Epidemiology Advance Access originally published online on April 12, 2007
American Journal of Epidemiology 2007 165(11):1247; doi:10.1093/aje/kwm077
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Tyas et al. Respond to "Predictors of Rate of Change in Disease Progression"
1 Department of Health Studies and Gerontology, University of Waterloo, Waterloo, Ontario, Canada
2 Division of Behavioural Neuroscience, Department of Psychology, University of Waterloo, Waterloo, Ontario, Canada
3 School of Statistics, National University of Colombia at Medellín, Medellín, Colombia
4 Department of Neurology, University of Kentucky, Lexington, KY
5 Sanders-Brown Center on Aging, University of Kentucky, Lexington, KY
6 Department of Preventive Medicine and Environmental Health, University of Kentucky, Lexington, KY
7 Department of Biostatistics, University of Kentucky, Lexington, KY
8 Department of Statistics, University of Kentucky, Lexington, KY
Received for publication January 31, 2007. Accepted for publication February 8, 2007.
In her commentary (1), Dr. Glymour describes four phenomena leading to a spurious association between risk factors for disease onset and the rate of disease progression. We agree with Dr. Glymour that the one that is relevant to our study (2) is the issue of beginning observations in the middle of a developing pathologic process. The inability to begin observation at the initiation of a pathologic process is, of course, not unique to this study but instead challenges all studies of dementia as well as many other conditions.
This issue is important to our study (2) because a number of our participants were diagnosed with dementia at their first assessment. Because we examined predictors of transitions in cognitive status across the trajectory from intact cognition to dementia, dementia was treated as an absorbing state (endpoint). Those women who already had dementia at the beginning of the study were thus excluded from our analyses. Contrary to Dr. Glymour's argument (1), however, we do not agree that this necessarily leads to spurious associations. Instead, our analyses found the opposite: adjusting for baseline status increased the odds ratios, indicating that our reported results of the effects of the covariates are, in fact, conservative.
Our study (2) presents a novel analytic approach to determine predictors of transitions in cognitive status, allowing for transitions in both directions: to poorer or to better cognitive status. Readers interested in designing prospective longitudinal studies will find the issues discussed in the commentary by Dr. Glymour (1) thought provoking; those interested in identifying a tool for the retrospective analysis of panel data will find our methodology worth considering.
 |
ACKNOWLEDGMENTS |
|---|
Conflict of interest: none declared.
|
References |
|---|
 TOP References |
|---|
- Tyas SL, Salazar JC, Snowdon DA, et al. Transitions to mild cognitive impairments, dementia, and death: findings from the Nun Study. Am J Epidemiol (2007) 165:1231–8.
[Abstract/Free Full Text] - Glymour MM. Invited commentary: When bad genes look good—APOE*E4, cognitive decline, and diagnostic thresholds. Am J Epidemiol (2007) 165:1239–46.
[Abstract/Free Full Text]
CiteULike 
Connotea 
Del.icio.us What's this?
Related articles in Am. J. Epidemiol.:
- Transitions to Mild Cognitive Impairments, Dementia, and Death: Findings from the Nun Study
- Suzanne L. Tyas, Juan Carlos Salazar, David A. Snowdon, Mark F. Desrosiers, Kathryn P. Riley, Marta S. Mendiondo, and Richard J. Kryscio
Am. J. Epidemiol. 2007 165: 1231-1238.[Abstract] [FREE Full Text]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||