Which Sleep Apnea Phenotype Leads to Obesity Hypoventilation?

Aaron B. Holley, MD


May 05, 2016

Approximately 90% of patient with obesity hypoventilation syndrome (OHS) will have obstructive sleep apnea (OSA).[1] Most OSA patients won't develop OHS, though, and the apnea-hypopnea index (AHI) doesn't correlate well with the presence of OHS.[2] There's clearly a relationship between the two diseases, but why does one OSA patient retain CO2 while another does not?

In the February issue of the Annals of the American Thoracic Society, investigators studied 14 patients who had serial polysomnography over time.[3] Some of the patients were treated with surgery or an oral appliance in the interim, but none had used continuous positive airway pressure. All patients had an AHI greater than 15 (moderate-to-severe disease). The researchers found that the end-tidal carbon dioxide (ETCO2) increased over time. They measured post- to pre-event breath amplitude using a pressure transducer and the ratio of breathing event (apnea or hypopnea) duration to interevent duration. Both of these ratios provide an assessment of ventilatory loading during events. The lower the post- to pre-event breath amplitude ratio and the higher the event to interevent duration, the more CO2 retention could theoretically occur. As the investigators expected, the AHI did not correlate with ETCO2, but post- to pre-event amplitude, event to interevent duration, and body mass index (BMI) accounted for 53% of the ETCO2 change in their model. The magnitude of these variables did not change over time, but serum bicarbonate levels increased between the first polysomnography and the second.

Several factors deserve note. First, in the researchers' model, the average increase in ETCO2 over 886 ± 307 days was 35.9 ± 4.2 to 39.4 ± 3.0 (P<.001). Their patients had an average BMI of approximately 34 kg/m2 and a mean AHI of 20. We see patients similar to these all the time—obese, with moderate OSA, who have gone untreated for years. We rarely find that they have CO2 retention, though. I'm surprised that the average increase in retention was that high, as it implies on average that all of these patients would meet criteria for OHS at the second study.

The increase in serum bicarbonate is also interesting. In their discussion, the authors used this phenomenon to propose the same hypothesis espoused by the authors of a recent CHEST publication.[4] In short, the idea is that as CO2 accumulates overnight in patients with OSA, serum bicarbonate increases to adjust acid-base abnormalities. At wake, partial pressure of CO2 (pCO2) quickly normalizes, but bicarbonate does not. Over time, the bicarbonate level slowly increases and changes the pCO2 set-point required to maintain acid-base neutrality. Patients end up retaining, and OHS is diagnosed. While the bicarbonate seems to have increased in all of the patients in the Annals of the American Thoracic Society study,[3] we know that this doesn't happen in actual practice. So why do some patients adjust their bicarbonate and ultimately develop OHS?