Does sleep apnea change over time in patients with acute myocardial infarction?

To the Editor: I read with great interest the article by Tsukamoto and Ohara who reported the temporal worsening of sleep-disordered breathing (SDB) in patients with acute coronary syndrome (ACS).1 The authors stated that SDB in the acute phase of myocardial infarction (MI) has not been well analyzed before and that there are no previous reports documenting changes in the severity of SDB during the acute phase of MI as analyzed by polysomnography (PSG). Additionally, I noticed that the authors did not exclude some of the conditions that may increase the severity of SDB in the acute phase of MI, such as being on sedation or narcotics or decreased level of consciousness, which may account for the temporary worsening in the acute phase of MI. In a recently published paper, we have already reported the prevalence and time-course of SDB in 50 consecutive patients with ACS using full overnight PSG within 3 days of the acute event and 6 months later after excluding the above conditions that may affect SDB.2 The body mass index of our group was comparable to that of Tsukamoto and Ohara’s patients (ie, 26.9±0.8kg/m2). 1,2 Fifty-six percent of the studied group had an apena/hypopnea index (AHI) >10h, 44% had an AHI >20 /h and 34% had an AHI >30h. AHI was 23.1±3.6 /h. AHI was divided into the obstructive apnea index (OAI) and central apnea index (CAI). OAI was 20.3±3.2 and CAI was 3.9±0.8. Cheyne-Stoke respiration (CSR) lasting more than 10% of total sleep time was documented in 6 patients. Ejection fraction was significantly lower in the group that had CSR compared with those who did not. Interestingly, AHI, OAI and the mean duration of obstructive apneas did not change significantly over the 6 months. On the other hand, CAI and central apnea duration were significantly lower in the second assessment. A few possible explanations may clarify the differences between our results and those of Tsukamoto and Ohara. A possible confounder is the administration of drugs that may affect respiration, such as narcotics and hypnotics. Although we used a drug-free period of at least 48 h, Tsukamoto and Ohara did not address this point in their methodology. Additionally, studies such as ours may have a risk of “regression towards the mean” phenomenon,3 which means that if a subgroup with a higher AHI is tested a second time, it is likely that the AHI may be lower at the second investigation. Obviously, more studies with larger number of patients and stringent control for possible confounders are required to clarify the relationship between ACS and SDB.