To the Editor The study by Greenspan et al1 is laudable and significant in that it focuses on a neglected but expanding cohort of patients. The group evaluated the efficacy of zolendronic acid. The study included patients with previous fragility fractures. Greenspan et al also permitted patients on bisphosphonate and/or alternative antiresorptive therapy coincident with the time of original fragility fractures. Patients were only excluded if medicated on bisphosphonate at the time of recruitment or if they had been taking a bisphosphonate for over 1 year within 2 years of trial recruitment. Unsurprisingly, in my view, they found zolendronate not to be protective against fragility fractures in this group in this context. The selection process involved the accumulation of bisphosphonate and/or antiresorptive failures. This is index event bias (also known as collider stratification bias), also articulated in JAMA.2 Those with a fragility fracture while receiving bisphosphonate or other antiresorptive treatment are likely to find a substitute or new concurrent bisphosphonate ineffective or weakly protective.
Uzoigwe CE. Dosing Strategies of Bone-Targeting Agents. JAMA Intern Med. 2015;175(11):1865. doi:10.1001/jamainternmed.2015.4796