To the Editor:

We read with interest an article from Shalaby et al.1 that showed that cardiac resynchronization therapy (CRT) failed to improve obstructive sleeprelated breathing disorders (SRBD) in patients with heart failure (HF). They reported that left ventricular ejection fraction (LVEF) only mildly improved with CRT compared to baseline (38.6 ± 11.1 vs 35.7 ± 11.2, P NS). This corresponded to a nonstatistically significant increase in the apneahypopnea index (AHI) of patients with obstructive SRBD after receiving CRT treatment, compared to baseline in the arm with no atrial overdrive pacing (31.4 ± 28.1 vs 20.3 ± 17.2, P NS). In the discussion, two studies were quoted; Stanchina et al.2 and Oldenburg et al.,3 which also evaluated the effects of CRT on AHI scores in obstructive sleep apnea (OSA). Stanchina et al. showed a significant improvement in LVEF in patients with CRT compared to baseline (LVEF: 33.6 ± 2.0 vs 22.5 ± 1.7, P <0.001), which corresponded to a significant improvement in AHI after CRT versus baseline (29.5 ± 21.3 vs 40.9 ± 6.4, P = 0.02). Additionally, Oldenburg et al. also saw a smaller but still significant improvement in LVEF with CRT compared to baseline (30.9 ± 6.7 vs 26.3 ± 5.7, P = 0.006). This corresponded to a nonsignificant reduction in AHI in patients after CRT implementation compared to baseline (14.6 ± 9.8 vs 18.2 ± 13.3, P NS). In our recently published meta-analysis,4 which looked at CRT and its effects on OSA as well as central sleep apnea (CSA), we showed that CRT significantly improved AHI scores in patients with CSA but not OSA. In our subanalysis of patients with OSA, all three of the studies discussed above were included.1–3 It is apparent that the results from Shalaby et al. contributed to shifting the results of the meta-analysis to be nonsignificant for the effects of CRT on severity of OSA. One of the most interesting and surprising findings in Shalaby et al. is the predominance