Portal de Periódicos da CAPES

RESPONSE

2009; Lippincott Williams & Wilkins; Volume: 41; Issue: 4 Linguagem: Inglês

10.1249/mss.0b013e31819af6ad

1530-0315

Sofia Beloka, Philippe van de Borne, Robert Naeije,

Inflammatory mediators and NSAID effects

Dear Editor-in-Chief Gläser et al. (1,2) question the relevance of our report that β-adrenergic blockade decreases aerobic exercise capacity in healthy subjects. No effect of β-blocker intake on exercise capacity emerged from their database of healthy volunteers recruited within the frame of a population-based survey in the northeast of Germany, the Study of Health in Northern Pomerania (SHIP) (2). We agree that this apparent contradiction deserves a discussion. Our own study was a prospective randomized controlled trial with cross-over design in 14 carefully selected healthy volunteers (1). The SHIP study was a public health registry, which sampled 7008 representative subjects from a population of 212,157 inhabitants with unusually high prevalence of poor cardiovascular health (3). A step-by-step procedure allowed for the inclusion of 1708 subjects who volunteered for an incremental cardiopulmonary exercise test (CPET). Exclusion of cardiac, pulmonary and neuromuscular conditions, and intake of any drug except β-blockers, led to a subselection of 1076 healthy volunteers. A proportion of these subjects presented with an increased body mass index (BMI), hypertension, diastolic dysfunction, a history of smoking and β-blocker use, and the practice of exercise was variable. Only a BMI >30 kg·m−2 and smoking habits were found to affect peak or maximum oxygen uptake and lactic threshold (3). The SHIP study is a remarkable large-scale effort that has produced valuable data of major public health relevance. However, we do not think that its methodology was adequate for the study of the effects of any pharmacological intervention on exercise capacity. There was no prespecified question asked at the start of the study. The subgroup tested for the impact of β-blockers resulted from a sequence of selections, all heavily exposed to biases. Essentially, all of the analysis of the results of the SHIP study rested on a posteriori consideration of subgroups, which is fraught with a statistical multiplicity problem (5) and particularly exposed to false-positive as well as false-negative results (4). All the SHIP study allows to state is that effects of β-blockers on exercise capacity do not emerge as a signal because of excessive noise in a registry that was not built to address this specific question. The same could be said for the amount of weekly exercise that surprisingly did not impact the fitness of Pomeranians (3). Large-scale population studies are of immense public health interest and relevance. The SHIP study showed a high prevalence of hypertension, increased body weight, smoking habits, and β-blocker use in Pomerania (3). Its results will be very helpful for improved health care strategies in that particular region of Germany. Registries may be useful by drawing attention on particular pathophysiological or therapeutic questions to be addressed afterward by properly designed trials. Our small but randomized controlled trial showed that a 1-week β-blocker therapy decreases aerobic exercise capacity by an average of 8%. Normal subjects feel it as a cause of decreased comfort during aerobic exercise, even if, for methodological reasons, it may not show up in registries. The SHIP registry and our study are not contradictory, but complementary. Sofia Beloka, MSc Philippe van de Borne, MD, PhD Robert Naeije, MD, PhD Department of Cardiology Erasme University Hospital Brussels, Belgium