Arjen Lindenholz

116 CHAPTER 5 score (SMART risk score) was calculated for each participant to act as a proxy for a multivariable analysis, which included 14 variables. Age and sex were included as covariates. The GEE Poisson regression models, with a robust variance estimator, were corrected for overdispersion using the Pearson chi-square method. For the additional analysis of the total burden of vessel wall lesions in the anterior and posterior circulation, the same method was used. Poisson regression generally provides rate ratios, which are interpretable as relative risks (RRs). Associations are therefore presented in more commonly used RRs with 95% confidence intervals (CIs). A Bonferroni correction for multiple testing was performed, and a two-sided p-value of 0.0013 (0.05 divided by 39) was indicative of a statistically significant difference. A two-sided p-value < 0.05 was considered to indicate trend. Statistical analyses were performed using SPSS version 21.0 (IBM SPSS Statistics, IBM Corp., Armonk, NY). Results Participants characteristics Participant characteristics are summarized in Table 1 . Of the 105 patients included in this study, 90 met the inclusion criteria. Fifteen patients were excluded because of a final diagnosis other than TIA or ischemic stroke ( n = 5), aborted MRI examination due to claustrophobia or participant discomfort ( n = 7), or withdrawn informed consent ( n = 3) ( Figure 1 ). Of the 90 study participants, the average age was 60 years (range 27–85 years) and 52 were man (58%). Forty-six of the 90 participants had hypertension, 44 had hyperlipidemia and 11 had diabetes mellitus at the moment of the ischemic event. Twenty-eight of the 90 were current smokers and 29 had a history of smoking.