Clone-censor weighting and dementia risk after statin initiation in diabetes: when the analyzed population is no longer the observed population
We read with interest the report by Ternhamar and colleagues about lower dementia risk with earlier statin initiation after diagnosis of type 2 diabetes in statin-naïve individuals.1 The authors should be commended for addressing one of the fundamental issues of observational studies, the prevalent-user bias, with the use of a new-user design. We highlight a less apparent issue for clinicians: the transition from real patients to cloned synthetic patients.
Authors
- Ali Ahmed (ORCID: https://orcid.org/0000-0002-6832-6424)
- Michael Böhm (ORCID: https://orcid.org/0000-0002-2976-2514)
- Andrew J. S. Coats
- Gregg C. Fonarow
- Stefan D. Anker
- John G.F. Cleland
Institutions
- University of California, Los Angeles (US)
- Georgetown University (US)
- George Washington University (US)
- Veterans Health Administration (US)
- The Heart Research Institute (AU)
- German Centre for Cardiovascular Research (DE)
- Washington DC VA Medical Center (US)
- Charité - Universitätsmedizin Berlin (DE)
- University of Glasgow (GB)
- Saarland University (DE)
Publication Details
- Journal
- The Lancet Regional Health - Europe
- Published
- 2026-11-14
- DOI
- https://doi.org/10.1016/j.lanepe.2026.101927
- Citations
- 1
- Primary Topic
- Statistical Methods in Epidemiology
- Type
- article
- Field-Weighted Citation Impact
- 1.46
Funders
- U.S. Department of Veterans Affairs
- National Institute on Aging