Reply to the Letter by Mascitelli et al.: Body Mass Index, Cholesterol Level and Poststroke Mortality

Abstract

Accessible online at: www.karger.com/ned We appreciate the thoughtful comments of Mascitelli et al. In our paper we did not adjust for serum cholesterol as this variable was not available in our database. However, for the purpose of our study adjustment for serum cholesterol is not necessary. Along with hypertension and diabetes, cholesterol is a mediator and not a confounder. Hence, these variables should not be adjusted for in the survival model, as they are part of the causal chain between BMI and death/survival. For this reason we presented two models, one including all the variables and one excluding mediators. The latter, we believe, is the most correct one. Both models, however, gave the same result: overweight persons with stroke had a better outcome than those who were underweight and normal weight. Moreover, compared to normal-weight individuals, survival was better in the obese and very obese individuals with stroke. Whether cholesterol, as suggested by Mascitelli et al., is responsible for the inverse relationship between BMI and survival is still not clarified, we believe. The possibility exists but other unknown elements in the causal chain are possibilities as well. Our study showed a clear association with BMI, diabetes and hypertension, both risk factors associated with an increased mortality rate. So, there is no doubt that the mortality rate is increased in many overweight and obese persons. However, our study shows that this is most likely due to hypertension and diabetes and not overweight and obesity per se. It can be hypothesized that overweight and obesity signal a lifestyle, which is leading to hypertension and diabetes and thereby decreased survival. If, however, an overweight or obese person has a ‘healthy’ lifestyle, survival is not endangered. We agree with Mascitelli et al. in their interpretation of our data. In patients with stroke the majority of poststroke deaths are caused by another stroke [1] . The risk of stroke is increased in overweight and obese persons but these strokes may be caused by small-vessel disease rather that large-vessel disease [1] . This might in part explain the lower mortality associated with overweight and obesity. Neuroepidemiology 2008;31:138 DOI: 10.1159/000151516