- Top of page
Background and objective: Although it has been reported that pulmonary hypertension is more frequent in patients with obesity-hypoventilation syndrome than in patients with ‘pure’ obstructive sleep apnoea syndrome, little is known about the haemodynamic repercussions of this entity. The aim was to describe the haemodynamic status, as assessed by echocardiography and 6-min walk test (6MWT), of patients with a newly diagnosed, most severe form of obesity-hypoventilation syndrome, and to evaluate the impact of non-invasive ventilation in these patients.
Methods: A prospective, descriptive, and single-centre follow-up study was conducted. At baseline, patients underwent echocardiography, spirometry, static lung volume measurement, 6MWT, overnight pulse-oximetry and polygraphic recording. Changes in echocardiography and 6MWT were assessed after 6 months of non-invasive ventilation. Right ventricular overload was defined on the basis of right ventricular dilatation, hypokinesis, paradoxical septal motion and/or pulmonary hypertension.
Results: Thirty patients (20 women; mean age 69 ± 11) were tested. The percentage of patients with right ventricular overload did not change significantly after non-invasive ventilation (43.3–41.6%; P = 0.24). In patients with right ventricular overload at diagnosis, pulmonary artery systolic pressure decreased significantly at six months (58 ± 11 to 44 ± 12 mm Hg; P = 0.014), and mean distance on 6MWT increased from 350 ± 110 to 426 ± 78 m (P = 0.006), without significant changes in body mass index.
Conclusions: Right ventricular overload is a frequent finding in patients with the most severe form of obesity-hypoventilation syndrome. Treatment with non-invasive ventilation is associated with a decrease in pulmonary artery systolic pressure at six months and an increase in the distance covered during the 6MWT.
- Top of page
Obesity-hypoventilation syndrome (OHS) is characterized by the co-occurrence of obesity and hypercapnia during wakefulness.1 Hypoventilation is a frequent complication of morbid obesity in hospitalized adults, and it associates with excess morbidity and mortality.2,3 Obstructive sleep apnoea syndrome (OSAS) is associated to OHS in 90–95% of patients.3 The effects of obesity and upper airway obstruction combine to produce excessive daytime sleepiness (EDS), dyspnoea, and in some individuals, cor pulmonale.4 Because the pathogenic mechanisms underlying OHS remain largely unknown, it is difficult to quantify the relative contributions of obesity and OSAS in a particular patient.
Chronic respiratory diseases such as chronic obstructive pulmonary disease (COPD) and restrictive thoracic disorders are well-known causes of pulmonary hypertension (PH). Recent investigations demonstrated that OSAS alone might increase daytime pulmonary artery pressure (PAP) beyond the normal limit.5,6 Therapy with continuous positive airway pressure (CPAP) reduces PAP without any concomitant change in lung function, or daytime arterial blood gas (ABG) levels.7 Little is known about the haemodynamic repercussions of OHS, although PH seems more frequent among patients with OHS than among those with ‘pure’ OSAS8 or simple obesity. Besides, congestive heart failure is more frequent in patients with OHS than in those with ‘pure’ OSAS.9,10
The purpose of this study was to describe the haemodynamic status of patients diagnosed with the most severe form of OHS, and to evaluate the impact of non-invasive ventilation (NIV).
- Top of page
Thirty subjects (20 women; mean age 69 ± 11) participated in the study. Baseline demographic, functional and PR data are presented in Table 1. Three patients had normal pulmonary function tests, whereas the remaining patients showed a restrictive ventilatory pattern (mean FVC 59 ± 16%). OSAS was diagnosed in 28 patients (93%; mean AHI 48 ± 27).
Table 1. Baseline characteristics of the patients with obesity-hypoventilation syndrome
|Characteristic|| n = 30||Patients with RVO||Patients without RVO|| P-value|
| n = 13|| n = 17|
|Age, years||69.0 ± 11.0||72.6 ± 10.1||66.8 ± 11.2||0.15|
|BMI, kg/m2||42.2 ± 8.3||43.3 ± 9.8||41.3 ± 7.0||0.51|
|ESS||14.0 ± 4.0||13.4 ± 2.9||14.6 ± 4.8||0.46|
|SBP, mm Hg||140.0 ± 13.3||144.7 ± 15.7||136.3 ± 10.0||0.08|
|DPB, mm Hg||76.8 ± 10.9||77.9 ± 11.4||76.0 ± 10.9||0.64|
|Diagnosis of SAS||28 (93.3%)||11 (84%)||17 (100%)||0.35|
|AHI||48.0 ± 27.0||37.1 ± 23.0||56.4 ± 28.0||0.053|
|mnSaO2, %||77.7 ± 7.6||75.9 ± 8.8||79.1 ± 6.4||0.26|
|CT90%, %||84.8 ± 19.2||90.0 ± 12.1||80.6 ± 23.1||0.32|
|PaO2||50.0 ± 8.0||48.3 ± 7.3||52.0 ± 8.5||0.22|
|PaCO2||57.0 ± 8.0||58.9 ± 9.7||56.6 ± 8.0||0.9|
|pH||7.38 ± 0.03||7.39 ± 0.04||7.37 ± 0.03||0.17|
|FEV1, % predicted||61.9 ± 15.5||56.7 ± 12.0||66.3 ± 17.0||0.10|
|FVC, % predicted||58.9 ± 16.3||55.1 ± 11.6||62.2 ± 19.4||0.26|
|FEV1/FVC, %||78.3 ± 6.3||75.2 ± 5.4||81.0 ± 5.9||0.01|
Thirteen patients (43.3%) presented with RVO. The mean PASP for all patients was 44 ± 15 mm Hg, whereas the mean PASP for patients with RVO was 58 ± 11 mm Hg. Ten (76.9%) of the patients with RVO had left heart abnormalities: LV hypertrophy in five patients, mitral valve regurgitation in four patients (three mild and one moderate) and LV diastolic dysfunction in six patients. Several of these abnormalities occurred in combination in most patients.
Seventeen patients did not have RVO. Mean PASP for these patients was 32 ± 5 mm Hg, and left heart abnormalities were identified in 12 patients (70.5%): LV dilatation in one patient, LV hypertrophy in four patients, LV diastolic dysfunction in nine patients and LV apical hypokinesis in one patient. Several of these defects combined in most patients. The percentage of patients with left heart anomalies among those with or without RVO did not differ significantly (P = 0.97).
Table 1 shows the baseline characteristics of patients with or without RVO at diagnosis. Table 2 shows the baseline echocardiography results for patients with or without RVO.
Table 2. Baseline echocardiographic parameters for patients with or without right ventricular overload
|—||Patients with RVO||Patients without RVO|| P-value|
|E wave, cm/s||78.1 ± 30.4||66.3 ± 20.6||0.50|
|A wave, cm/s||78.4 ± 15.0||97.7 ± 32.5||0.35|
|E/A ratio||0.89 ± 0.48||0.68 ± 0.08||0.35|
|Deceleration time, ms||225.0 ± 97.1||222.5 ± 35.0||0.36|
|Left atrial diameter, cm||4.2 ± 0.7||4.1 ± 0.6||0.87|
|LV diastolic diameter, cm||5.0 ± 0.4||5.0 ± 0.7||0.65|
|LV systolic diameter, cm||3.1 ± 0.4||3.3 ± 0.6||0.35|
|LV shortening fraction, %||39.5 ± 9.2||36.1 ± 7.9||0.43|
|LVEF, %||65.5 ± 8.5||65.7 ± 7.5||0.96|
|Interventricular septum, cm||1.1 ± 0.2||1.0 ± 0.2||0.67|
|Posterior LV wall, cm||1.0 ± 0.2||1.1 ± 0.2||0.57|
|LV mass, g||201.3 ± 61.9||209.7 ± 77.4||0.79|
|PASP, mm Hg||58 ± 11||32 ± 5||<0.001|
The mean distance on the 6MWT was 350 ± 110 m, and was shorter for patients with than for those without RVO, although the difference was not significant (299 ± 129 m vs 378 ± 95 m; P = 0.06). Average SaO2 during the 6MWT was lower in patients with than in those without RVO (80.9 ± 5.4% vs 85.3 ± 5.3%; P = 0.033).
Results at 1-month follow-up
Three patients rejected NIV, and another three were lost to follow-up before the 1-month visit (Fig. 1). Data for these patients were excluded from the analysis. Dyspnoea improved in all but four patients. ESS score decreased from 14 ± 4 at baseline to 6 ± 2 at follow-up (P < 0.0001). PaO2 increased from 50 ± 8 to 67 ± 8 mm Hg (P < 0.0001). PaCO2 decreased from 57 ± 8 to 46 ± 5 mm Hg (P < 0.0001). Distance walked on the 6MWT increased from 350 ± 110 to 416 ± 74 m (P = 0.0015).
Results at 6-month follow-up
Only one patient who commenced BiPAP treatment was non-compliant, while the other three patients were lost to follow-up. There were no significant differences between patients who did or did not finish the study. Mean BMI at the end of the study was not significantly different from that at baseline (41.2 ± 7.8 vs 42.2 ± 8.2 kg/m2; P = 0.09). Echocardiographic and 6MWT results at the 6-month follow-up are summarized at Tables 3 and 4. Ten of the 24 patients had RVO at the end of the study. The percentage of patients with RVO at baseline did not change (43.3–41.6%; P = 0.24). Mean PASP in patients with RVO at diagnosis decreased significantly at the 6-month follow-up (58 ± 11 to 44 ± 12 mm Hg; P = 0.014).
Table 3. Echocardiographic and 6-min walk test results at follow-up
|—||Baseline (1)||One-month follow-up (2)|| P-value (1 vs 2)||Six-month follow-up (3)|| P-value (1 vs 3)|
|PASP (mm Hg)|| || || || || |
| All patients||44 ± 15||—||—||43 ± 15||0.52|
| RVO at diagnosis||58 ± 11||44 ± 12||0.014|
| No RVO at diagnosis||32 ± 5||42 ± 19||0.17|
|Patients with RVO||13 (43.3%)||—||—||10 (41.6%)||0.24|
|Distance on 6MWT, m||350 ± 110||416 ± 74||0.0015||426 ± 78||0.006|
|mSaO2 during 6MWT, %||83.8 ± 5.6||87.3 ± 2.9||0.06||86.7 ± 4.2||0.049|
Table 4. Echocardiographic findings at baseline and at six-month follow-up
|—||Baseline||Six–month follow-up|| P-value|
|E wave, cm/s||68.4 ± 26.4||70.1 ± 15.3||0.76|
|A wave, cm/s||86.7 ± 27.2||88.8 ± 25.6||0.84|
|E/A ratio||0.82 ± 0.41||0.86 ± 0.42||0.57|
|Deceleration time, ms||236.6 ± 89.3||212.2 ± 72.6||0.46|
|Left atrial diameter, cm||4.1 ± 0.7||4.3 ± 0.6||0.20|
|LV diastolic diameter, cm||5.0 ± 0.6||4.5 ± 0.5||0.01|
|LV systolic diameter, cm||3.2 ± 0.4||2.9 ± 0.4||0.02|
|LV shortening fraction, %||38.6 ± 7.6||37.3 ± 4.5||0.50|
|LVEF, %||66.3 ± 8.0||67.8 ± 6.1||0.44|
|Interventricular septum, cm||1.1 ± 0.2||1.2 ± 0.1||0.07|
|Posterior LV wall, cm||1.1 ± 0.1||1.1 ± 0.2||0.86|
|LV mass, g||217.8 ± 70.2||205.5 ± 63.9||0.40|
Left heart abnormalities (LV hypertrophy, LV diastolic dysfunction and mitral valve regurgitation) were identified in eight (80%) patients with and 11 (78%) patients without RVO. Mean LV diameters decreased significantly at follow-up (Table 4).
The average 6MWT distance at the 6-month follow-up (426 ± 78 m) was significantly greater than that at diagnosis, although it had not increased further compared with that at 1 month after discharge. Mean SaO2 during 6MWT increased slightly from baseline (86.7 ± 4.2% vs 83.8 ± 5.6%, P = 0.049). There was no correlation between changes in PASP and distance covered during the 6MWT (r = 0.0217; P = 0.92).
- Top of page
In this study, we evaluated the incidence of RVO in patients with OHS and examined the effects of NIV on echocardiographic results and performance in the 6MWT. Using a cut-off value for PASP >40 mm Hg to define PH, the prevalence of RVO was 43.3%. In the only published analysis of the prevalence of PH in OHS, Kessler et al.8 reported a prevalence of 58% among 34 patients, with an average mean PAP (mPAP) of 23 ± 10 mm Hg, as measured by right heart catheterization (RHC). In that study, PH was defined as mPAP ≥20 mm Hg, which is lower than the currently accepted value of 25 mm Hg.18
The incidence of RVO in patients with OHS was identical to that in patients with ‘pure OSAS’,16 although Arias et al. defined PH as a PASP >30 mm Hg.19 As a more restrictive threshold of 40 mm Hg was adopted in the present study, it seems likely that the haemodynamic status of patients with OHS is worse than that of those with ‘pure OSAS’. It should be noted that the elevation of PASP in the present study was generally mild and of uncertain clinical relevance.
There were no significant differences in the percentage of patients with RVO before and after 6 months of NIV, possibly due to the small sample size. However, NIV resulted in a significant decrease in PASP and an increase in 6MWT distance in patients with RVO at diagnosis. This effect could not be attributed to a reduction in body weight, as BMI was unchanged. Several mechanisms could explain these results, including reversal of hypoxia, abolition of apnoeas, and consequently, mechanical overload due to increased respiratory effort, suppression of reflex mechanisms that could influence the pulmonary vasculature and improvement in LV function. To date, no studies have examined the haemodynamic effects of NIV in OHS. However, the prevalence of PH in ‘pure’ OSAS is 20–40%.5,6 There is evidence that some patients with OSAS have an increased incidence of LV diastolic dysfunction and/or left atrial enlargement that may lead to post-capillary PH.20 In addition, CPAP may reduce PASP and improve LV diastolic function.19 The frequent LV echocardiographic alterations and the decrease of LV diameters after NIV in our population suggest that consideration of a post-capillary mechanism as one of the causes of PH is warranted. This study demonstrated an increase in the functional capacity of patients with OHS who were treated with NIV. Distance in the 6MWT increased by 76 m, on average, after NIV. Obesity is associated with reduction in 6MWT distance;21 however, the improvement in the distance walked cannot be attributed to a reduction in body weight, as this was unchanged. It should be noted that the distance was related to the PASP: the lower the PASP, the longer the patients walked. This finding suggests that haemodynamic status may ultimately contribute to the limitation of exercise performance in OHS. It should be noted that desaturation during exercise only improved slightly, despite the correction of respiratory failure at rest. Although several studies have suggested that exertional desaturation may portend a poor prognosis for patients with COPD,22 the clinical significance of this in OHS remains unknown.
This study had limitations including a small sample size. Because the study lacked a control group, the effects of confounding factors, such as obesity and age, that might have contributed to the high incidence of PH could not be definitely ruled out. Recently, findings from a PH database in the United States23 indicated a higher prevalence of obesity among individuals with idiopathic forms of pulmonary arterial hypertension. Importantly, this association appears to be independent of conditions associated with the development of PH, such as diastolic dysfunction and OSAS. Nevertheless, based on the findings of McQuillan et al.,15 who measured PASP >40 mm Hg in 6% of subjects >50 years and in 5% of subjects with BMI >30 kg/m2, we would expect a false-positive diagnosis of PH in only a small proportion of the present population. Therefore, obesity per se is unlikely to be responsible for the high prevalence of PH among patients with OHS.
We decided to use a restrictive definition of OHS, which excluded patients with mild hypercapnia. This conceptual restriction made it difficult to recruit patients; on the other hand, it allowed certainty that we were dealing with ‘real pickwickian patients’. We believe that there is a wide range of severity of OHS; therefore, the haemodynamic repercussions, prognosis and treatment could differ depending on the particular patient. Another limitation relates to the method used for evaluation of PASP because echocardiographic assessment is inaccurate in patients with PH.24 However, we aimed to reflect the real-life setting, and the decision to perform RHC must be made on a case-by-case basis because it is not feasible in all patients. We adopted a conservative approach and labelled as abnormal only those patients with PASP >40 mm Hg, to avoid false-positive results.15 On the other hand, from a study of the accuracy of echocardiography in patients with suspected PH, Rich et al.25 concluded that it tended to underestimate PASP. The fact that RHC excluded PH in only a small minority of the patients previously diagnosed by echocardiography suggests that this technique can be used to reliably identify patients with true PH. Also, most patients with PASP >40 mm Hg showed other signs of RVO, such as RV dilatation. These observations give us confidence that the findings provided a good indication of the haemodynamic status of the study population.
In conclusion, this study showed that RVO is a frequent finding in patients with the most severe form of OHS. Treatment with NIV was associated with a decrease in PASP and an increase in the distance covered during the 6MWT. These findings suggest that NIV has a beneficial effect on the haemodynamic status of patients with OHS.