To the Editor McCullough et al1 reported that apparent mortality reductions following vertebroplasty and kyphoplasty likely result from selection bias. These findings are important to counter hasty conclusions from several recent population-based studies and reviews. Yet these new results are hardly surprising, given the a priori lack of evidence to assume a causal association between vertebral fractures and mortality, let alone mortality risk modifiable by targeting the spine. Reduced pulmonary function following vertebral fracture would be a plausible mechanistic explanation, but only small, short-term improvements in vital capacity following vertebral augmentation have been demonstrated. Contrary to the frequently cited increased risk of pulmonary deaths in older women with prevalent vertebral fractures in the Study of Osteoporotic Fractures, incident vertebral fractures (arguably the stronger outcome) in that cohort were no longer associated with mortality following adjustment for other determinants such as frailty.2 The same authors have further shown that hyperkyphosis and height loss are equally unfavorable, independent of vertebral fractures and their characteristics,3 arguing against a direct role of the latter.
Laurent M. Treatment of Osteoporotic Vertebral Fractures. JAMA Intern Med. 2014;174(4):641-642. doi:10.1001/jamainternmed.2013.13482