Stroke in relation to use of raloxifene and other drugs against osteoporosis

Summary Prior studies have associated fatal stroke with raloxifene. In a cohort study, we found no excess risk of stroke with raloxifene; whereas, an excess risk of stroke and fatal stroke was seen with alendronate and etidronate. However, the excess risks were small. Purpose We aim to study the association between use of raloxifene and other drugs against osteoporosis and risk of stroke. Methods This is a nationwide cohort study from Denmark. All users of bisphosphonates and other drugs against osteoporosis between 1996 and 2006 (n=103,562) as exposed group and three age- and gender-matched controls from the general population (n=310,683). Results Before the drugs were started, patients later initiating alendronate or raloxifene had fewer strokes than the controls. In contrast, patients who later did start clodronate have more strokes. Among the later users of other bisphosphonates, strontium ranelate or parathyroid hormone, no change in the risk of stroke was present. Patients who started raloxifene neither had an excess risk of strokes nor of fatal strokes. No dose–response relationship was present. Among users of alendronate, a decreasing overall risk of stroke was seen with increasing dose. However, for fatal strokes, the risk increased with increasing dose of alendronate. Among users of etidronate, no trend with dose was present for overall stroke risk; whereas for fatal strokes, an increasing risk was seen with increasing dose of etidronate. Conclusions Raloxifene does not seem associated with an excess risk of strokes. The increase seen for alendronate did not seem to be causal as no classical dose–response relationship was present. The dose–response relationship for fatal strokes with alendronate and etidronate needs further examination. However, the excess risks were small and may be due to the underlying disease.