Response by Falkstedt et al to Letters Regarding Article, “Cannabis, Tobacco, Alcohol Use, and the Risk of Early Stroke: A Population-Based Cohort Study Among 45 000 Swedish Men”
First, we express our appreciation of the comments on our study by Kawada and by Behrouz and Aachi. We seem to agree on the value of studies on this matter. In both letters, the authors highlight that cannabis use in our study (as well as tobacco and alcohol use) could be measured only in late adolescence and, as a result, little is known about its prevalence later, when cases of stroke occur. Therefore, the authors suspect that our analysis could have missed an association between cannabis use in adulthood and stroke events. This cannot be ruled out, which we touched on in the discussion of our results. However, some prevalence of cannabis use in adulthood is still expected, and we think that in such cases, it was generally preceded by teenage use. With regard to tobacco smoking, >90% of those who were smokers later in life already smoked before the conscription examination (as shown in a previous study of 649 participants from the cohort).1 Thus, if cannabis use contributed to cause several stroke events during the follow-up, an elevation of the risk among the early cannabis users should remain after adjustment for other risk factors.
The main motivation for our study, however, was the nearly total lack of population-based epidemiological studies regarding cannabis and early stroke. In addition, an important advantage of our study as compared with previous ones was the possibility of multivariable analyses and, in particular, to examine associations between cannabis use and stroke before and after adjustment for other known risk factors. These analyses showed that high tobacco consumption did in fact explain the elevated risk of stroke among the men who reported use of cannabis in late adolescence, as well as among those who reported high consumption of alcohol.
In our study, the association between tobacco smoking reported in late adolescence and subsequent strokes was considerably stronger before the age of 45 years than later in the follow-up period. Kawada suggests this could be interpreted as an increasing importance of cardio-metabolic components in relation to stroke, which is an important remark. It is also supported by the fact that cardiorespiratory fitness and body mass index measured in late adolescence in our study were more consistent in their associations with stroke across the follow-up period than was smoking (not shown in the article). Nevertheless, we think that another reason for the attenuation of the hazard ratio is smoking cessation. That is, some of the men who were smokers as young adults eventually did not smoke and, therefore, were to some extent misclassified as smokers. Consequently, the true effect of tobacco smoking may have been underestimated in relation to strokes occurring later in the follow-up period.
Behrouz and Aachi direct our attention to a study showing that use of cocaine, similarly to cannabis use, has been found to be unexpectedly common among cases of early stroke, and they suggest cocaine use could have been included in our study. Unfortunately, no such data were possible to collect. Moreover, the extent to which cocaine was used among young adults in Sweden during this time is unknown. All the same, without the possibility to adequately control for confounders, which applies also to the findings on cocaine by Cheng et al 2016,2 results remain ambiguous. If the aim is to isolate the effects of cannabis use and cocaine, then cases of stroke need to occur among exclusive users of these substances and not only among concurrent users of tobacco.
The main contribution of our study was the long-term follow-up of stroke risk in a nearly unselected population of Swedish men who in late adolescence reported their consumption of cannabis, tobacco, and alcohol. We did not find cannabis use to be an evident risk factor for early stroke after adjustment for tobacco smoking. Nevertheless, there are currently limited opportunities to confirm or reject the hypothesis of (acute) effects of cannabis on stroke, and it is well justified to collect data for improved studies of this, especially given the trend toward increased use of cannabis worldwide.
Daniel Falkstedt, PhD
Anna-Karin Danielsson, PhD
Department of Public Health Sciences
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 4 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.
- © 2017 American Heart Association, Inc.
- Hemmingsson T,
- Kriebel D,
- Melin B,
- Allebeck P,
- Lundberg I
- Cheng YC,
- Ryan KA,
- Qadwai SA,
- Shah J,
- Sparks MJ,
- Wozniak MA,
- et al