Myth: The Incidence of Dementia is Declining

Here is another example of how bad science and bad journalism can provide you with a great learning opportunity. Why the “prestigious” scientific journal, Neurology, could publish such garbage is beyond me.

The conclusions drawn by the researchers, based on statistical analysis, would not even pass a first-year college course in statistics. I give this article an “F” grade. Read on. I highlighted the important conclusion in red.

~ Jennifer

Neurology. 2012 May 8;78(19):1456-63. Epub 2012 May 2.

Is dementia incidence declining?: Trends in dementia incidence since 1990 in the Rotterdam Study.

Schrijvers EM, Verhaaren BF, Koudstaal PJ, Hofman A, Ikram MA, Breteler MM.


Department of Epidemiology, Erasmus MC University Medical Center, Rotterdam, The Netherlands.



To investigate whether dementia incidence has changed over the last 2 decades.


We compared dementia incidence in 2 independent subcohorts of persons aged 60-90 years from the Rotterdam Study, a population-based cohort study. The first subcohort started in 1990 (n = 5,727), the second in 2000 (n = 1,769). Participants were dementia-free at baseline and followed for at maximum 5 years. We calculated age-adjusted dementia incidence rates for the 2 subcohorts in total, in 10-year age strata, and for men and women separately. We also compared mortality rates, differences in prevalence of vascular risk factors, and medication use. Finally, we compared brain volumes and the extent of cerebral small vessel disease in participants who underwent brain imaging 5 years after the baseline examinations.


In the 1990 subcohort (25,696 person-years), 286 persons developed dementia, and in the 2000 subcohort (8,384 person-years), 49 persons. Age-adjusted dementia incidence rates were consistently, yet nonsignificantly, lower in the 2000 subcohort in all strata, reaching borderline significance in the overall analysis (incidence rate ratio 0.75, 95% confidence interval [CI] 0.56-1.02). Mortality rates were also lower in the 2000 subcohort (rate ratio 0.63, 95% CI 0.52-0.77). The prevalence of hypertension and obesity significantly increased between 1990 and 2000. This was paralleled by a strong increase in use of antithrombotics and lipid-lowering drugs. Participants in 2005-2006 had larger total brain volumes (p < 0.001) and less cerebral small vessel disease (although nonsignificant in men) than participants in 1995-1996.


Although the differences in dementia incidence were nonsignificant, our study suggests that dementia incidence has decreased between 1990 and 2005.


Neurology. 2012
PMID: 2255173 PubMed – indexed for MEDLINE