Eight cohort studies with a total of 11,236 cases (case numbers ranged from 95 fatal strokes to 2,781 incident events) were included in the review. Study follow-up ranged from eight to 19 years.
Total fibre intake was associated with a statistically significant reduction in stroke risk (RR per 7g/day increase 0.93, 95% CI 0.88 to 0.98; seven studies); there was evidence of heterogeneity for this result (I²=59%). Risk steadily reduced with increasing total fibre intake. There was evidence of publication bias. Estimates were generally unchanged when subgroups were pooled using meta-regression.
There was no statistically significant association between soluble fibre intake and stroke risk (RR per 4g/day increase 0.94, 95% CI 0.88 to 1.01; three studies).
There was significant heterogeneity in the meta-analysis of cereal fibre intake and stroke risk (I²=90%) so the pooled results were not reported and individual study results were presented.
Individual study results were presented for ischaemic stroke risk, haemorrhagic stroke risk, the association between insoluble fibre and stroke risk and the association between fruit and vegetable fibre and stroke risk.