Dear Sir,

We greatly appreciate the interest in our recent pharmacoepidemiological study of heart failure in patients treated with bisphosphonates [1].The main finding of this nationwide cohort study was an increased risk of heart failure in users of bisphosphonates compared to age- and sex-matched control subjects, even after adjusting for differences in comorbidities and comedication. As mentioned by Canpolat et al., two other interesting observations were made in this context. First, raloxifene, which has a different mechanism of action but is used for the same indication as bisphosphonates, was not associated with increased risk of heart failure. Second, significant trends in the risk of heart failure were observed across refill compliance strata, when the two most commonly used bisphosphonates were analysed separately. Accordingly, the risk of heart failure increased significantly with increasing refill compliance for etidronate (P for trend <0.01), whereas it decreased for alendronate (P < 0.01).

Canpolat et al. point out important study limitations, some of which are inherent to the design of our study. Although several precautions were taken regarding study design, data analyses and interpretation of results, residual confounding can never be excluded in a registry-based study. Regarding age-associated physiologic changes and presence of hypertension and other risk factors predisposing to development of heart failure, our risk estimates were adjusted for a long list of potential confounders (please see Tables 1 and 2), including age and use of antihypertensive drugs [1]. The authors also mention the lack of echocardiographic data, however, we believe that this is not a major drawback, as the diagnosis of heart failure in Danish national registers, although underreported (sensitivity 29%), has a specificity of 99% with a positive predictive value of 81% and a negative predictive value of 90% [2]. Any undiagnosed heart failure would bias the risk estimates towards null.

Canpolat et al. also comment that, owing to the well-known association between atrial fibrillation and heart failure, data on atrial fibrillation would have been valuable. We fully agree but would like to add that such information would also have added value to the study because of the possibility of investigating the hypothesis that atrial fibrillation might represent a mechanism explaining at least part of the increased heart failure risk observed in users of bisphosphonates [3, 4].

We believe that the findings of our study are important and should be further explored, as bisphosphonates are widely used, and heart failure constitutes a major health issue.

Conflict of interest statement

There is no conflict of interest to declare.


  1. Top of page
  • 1
    Grove EL, Abrahamsen B, Vestergaard P. Heart failure in patients treated with bisphosphonates. J Intern Med 2013; 274: 34250.
  • 2
    Kumler T, Gislason GH, Kirk V et al. Accuracy of a heart failure diagnosis in administrative registers. Eur J Heart Fail 2008; 10: 65860.
  • 3
    Heckbert SR, Li G, Cummings SR, Smith NL, Psaty BM. Use of alendronate and risk of incident atrial fibrillation in women. Arch Intern Med 2008; 168: 82631.
  • 4
    Abrahamsen B, Eiken P, Brixen K. Atrial fibrillation in fracture patients treated with oral bisphosphonates. J Intern Med 2009; 265: 58192.