Shakar Kutal, Nilufer Yesilot, Tomi Sarkanen, Petra Redfors, Annika Nordanstig, Pauli Ylikotila, Marialuisa Zedde, Ulla Junttola, Lauri Tulkki, Kristina Ryliskiene, Radim Líčeník, Phillip Ferdinand, Dalius Jatužis, Liisa Kõrv, J Korv, Alessandro Pezzini, Juha Sinisalo, Mika Lehto, Eva Gerdts, Jaana Autere, Ana Catarina Fonseca, De Paula A, Ulrike Waje‐Andreassen, Bettina von Sarnowski, Tiina Sairanen, Turgut Tatlisumak, Juha Huhtakangas, Pekka Jäkälä, J Putaala, Nicolas Martinez‐Majander
BACKGROUND: Data on the association between physical activity (PA) and cryptogenic ischaemic stroke (CIS) in young adults are scarce. We investigated the relationship between PA levels and early-onset CIS, stratified by the presence of a high-risk patent foramen ovale (PFO). METHODS: Patients aged 18-49 years with first-ever CIS and sex- and age-matched stroke-free controls were recruited from 19 European centres. PA was assessed using the short International Physical Activity Questionnaire and expressed as Metabolic Equivalents, categorized into percentiles (bottom 10%, 10%-25%, 25%-75% [reference], 75%-90%, top 10%). Associations between PA and CIS were analyzed using conditional logistic regression adjusted for age, education level, traditional risk factors, and non-traditional risk factors. RESULTS: Altogether, 533 patients (median age 41 [interquartile range 34-46]; 47.3% women) and 533 controls were included. Scoring in the top 10% of PA was independently associated with CIS: adjusted odds ratio 2.07; 95% confidence interval 1.22-3.51. Comparing patients without high-risk PFO to all controls, the top 10% PA category (1.78; 1.07-2.94) and the bottom 10% category (1.76; 1.06-2.92) were associated with CIS. No independent association was observed in patients with high-risk PFO. CONCLUSIONS: PA in the top 10% was associated with an increased risk of early-onset CIS in the overall study population. Among patients without a high-risk PFO, both the bottom 10% and top 10% of PA were consistently associated with elevated CIS risk. These findings highlight the need to better understand the mechanisms through which both low and high activity levels may predispose to stroke.