Letter by Rutten-Jacobs and de Leeuw Regarding Article, “Long-Term Mortality After First-Ever and Recurrent Stroke in Young Adults”
To the Editor:
We read with great interest the recent article by Aarnio et al1 on long-term mortality after young stroke. The authors studied long-term mortality after young stroke and compared mortality in the young stroke population with that in the general population, matched with age, sex, calendar year, and geographical area. Similar to a previous study from our group,2 they observed an increased long-term mortality compared with the general population, expressed as the standardized mortality ratio (SMR). Aarnio et al1 commented that the SMR observed in their study was remarkably 2× higher than in ours and that the reasons for this difference are unclear. In the following, we try to clarify this difference.
The expected number of deaths that is needed to calculate the SMR, is generally calculated using indirect standardization.3 In this method, the age, sex, and calendar-year–specific mortality rates in the reference population (general population) are weighted by the age, sex, and calendar-year structure of the study population. Therefore, when using indirect standardization, it can be problematic to compare SMRs between different study populations as each study population’s SMR is based on its own sets of weights determined by the age, sex, and calendar-year structure of the study population.3
Furthermore, there seems to be a difference between the 2 studies in the used calculation of the expected number of deaths. In our study,2 we matched the reference population to the study population on patient level, ensuring that the weighting is consistent, irrespectively of the subsequent division in subgroups. The data in Table 3 from the article by Aarnio et al suggest that the authors might have taken a somewhat different approach, as the sum of the expected number of deaths for subgroups of the population is not equal to the reported expected number of deaths for the total population, in contrast to the number of observed deaths. For example, the sum of the reported number of deaths for the 2 age categories, patients aged 15–39 years and patients aged 40–49 years is substantially lower than the reported number of expected deaths in the total group of patients aged 15–49 years (4.1+29.3=33.4 versus 21.9). As a result, the SMR for the total group of patients is higher than the SMRs for the 2 subgroups. Recalculating the SMR for the overall population based on the sum of the expected number of deaths in the 2 age subgroups would yield an SMR of 4.6 (rather than 6.9), which is more in the same magnitude as the SMR shown in our study (3.9).2
In the past, publications of both studies consistently showed much agreement in results on mortality and recurrent vascular events,2,4,5 emphasizing the similarities and meritorious robustness of both study designs. Also in the present article,1 the crude cumulative mortality at 16 years is similar to that observed in our study,2 whereas the difference in the magnitude of the SMR seems to be merely caused by a different standardization approach.
Loes C.A. Rutten-Jacobs, PhD
Department of Clinical Neurosciences, Neurology Unit
University of Cambridge, Cambridge, UK
Frank-Erik de Leeuw, MD, PhD
Donders Institute for Brain, Cognition and Behaviour
Department of Neurology, Radboud University Medical Center
Nijmegen, the Netherlands
Stroke welcomes Letters to the Editor and will publish them, if suitable, as space permits. Letters must reference a Stroke published-ahead-of-print article or an article printed within the past 3 weeks. The maximum length is 750 words including no more than 5 references and 3 authors. Please submit letters typed double-spaced. Letters may be shortened or edited.
- © 2014 American Heart Association, Inc.