| Home | E-Submission | Sitemap | Contact us |  
Sleep Med Res > Volume 10(1); 2019 > Article
Saeseow, Chattakul, Khamsai, Limpawattana, Chindaprasirt, Chotmongkol, Silaruks, Senthong, and Sawanyawisuth: Predictors for Obesity Hypoventilation Syndrome in Thai Population
This article has been corrected. See "Predictors for Obesity Hypoventilation Syndrome in Thai Population" in Volume 12 on page 170.


Background and Objective

Obesity Hypoventilation Syndrome (OHS), is a condition with high morbidity and mortality. Body Mass Index (BMI) of more than 30 kg/m2 is used, to diagnose OHS. As BMI for obesity for Thais is 25 kg/m2, BMI more than 25 kg/m2 is used in our institution, to diagnose OHS. The purpose of this study was to evaluate if BMI of 25 kg/m2 is appropriate criterion for OHS in Thai patients.


This study was a retrospective study conducted at Khon Kaen University. Inclusion criteria were adult patients diagnosed with OHS in 2016. Patients diagnosed with obstructive sleep apnea (OSA), were randomly selected as control subjects. The ratio of OHS:OSA, was 1:4. Clinical factors associated with OHS were examined, using multivariate logistic regression analysis.


During the study period, there were 25 OHS and 108 OSA patients. The OHS group had a significantly higher average BMI (48.9 kg/m2 vs. 29.2 kg/m2), than the OSA group. The OHS group also had higher proportions of patients with pulmonary hypertension (50% vs. 2%), and heart failure (76% vs. 6.5%). There were two independent predictors for OHS, including BMI and serum bicarbonate levels. Adjusted odds ratio (95% CI) for each of these factors was 1.08 (1.01, 1.17) and 1.96 (1.15, 3.34), respectively. Body mass index greater than 25 kg/m2 and serum bicarbonate more than 25 mEq/L, yielded 100% sensitivity for OHS.


s Appropriate diagnostic criteria for OHS for the Thai population, may be different from those for populations in Western countries.


Obesity Hypoventilation Syndrome (OHS) is defined as the combination of obesity [Body Mass Index (BMI) ≥ 30 kg/m2] and daytime hypoventilation [arterial carbon dioxide pressure (PaCO2) ≥ 45 mm Hg], in absence of other possible causes of hypoventilation [1]. Although obstructive sleep apnea (OSA) and OHS are classified as sleep-related breathing disorders, OHS patients tend to have lower quality of life, greater health care expenses, greater risk of pulmonary hypertension, and higher mortality rate, compared with OSA patients [2].
Prevalence of OHS in the general population is unknown, but has been estimated between 0.15% and 0.30% in the United States adult population [3]. However, prevalence estimates for OHS, vary significantly across studies. One review article found the estimated prevalence of OHS in patients with OSA, was between 4–50% [4]. This considerable range, likely reflects variations in patient populations across studies.
Additionally, clinical features of OHS patients vary considerably by ethnicity. For example, Japanese OHS patients are younger, than those with pure OSA [5,6]. OHS was found in older patients, than in OSA patients in Saudi Arabia or the United Arab Emirates [3,7,8]. East Asian populations are also known to develop OHS at lower BMI, compared with non-Asian populations [4].
As mentioned above, criteria for OHS may differ among ethnic groups. Current diagnostic criteria for OHS, are BMI more than 30 kg/m2. However, BMI criterion may not be applicable in Thai or other Asian populations, as criterion for obesity used in Asian populations is BMI more than 25 kg/m2. So, BMI of more than 25 kg/m2 is used in our institution, to diagnose OHS. The purpose of this study was to evaluate if BMI of 25 kg/m2 is appropriate criterion, for OHS in Thai patients.


This was a retrospective study conducted at Khon Kaen University’s Srinagarind Hospital. Inclusion criteria were adults (older than age 15), and had been diagnosed with OHS. Patients were excluded, if they had one of the following conditions: hypothyroidism, Chronic Obstructive Pulmonary Disease, neuromuscular diseases, opiate overdose, or severe electrolyte abnormalities such as hypophosphatemia, hypokalemia, or hypercalcemia. The study period was January–December 2016. All patients diagnosed as OHS, were included in the study. Patients diagnosed with OSA were randomly selected as controls, from the OSA database at the hospital. Ratio of OHS:OSA was 1:4. Study protocol was approved by the Ethics Committee in Human Research, Khon Kaen University, Thailand (HE541373).
Clinical data of all eligible patients in both groups were recorded including baseline characteristics, co-morbid diseases, physical signs, and laboratory results. All patients underwent polysomnography, and OSA was diagnosed according to the apnea-hypopnea index (AHI; evidence of five or more apnea or hypopnea events per hour).

Statistical Analyses

Descriptive statistics were used, to compare differences between OHS and OSA groups. Clinical predictive factors of OHS were calculated, using logistic regression analysis. Factors with a p value less than 0.20 by univariate logistic regression, were included in subsequent multivariate logistic regression analysis. If significant factors for OHS were expressed as numerical data, cutoff points to determine risk of OHS were computed, using a receiver operating characteristic (ROC) curve. All statistical analyses were performed, using STATA software version 10.0 (College Station, TX, USA).


There were 25 OHS and 108 OSA patients, included in this study. Median values of pCO2 and pO2 of the OHS group was 56 mm Hg (first to third interquartile range of 52–72), and 57 (first to third interquartile range of 50–65). The OHS group had significantly higher average BMI (48.9 kg/m2 vs. 29.2 kg/m2), than the OSA group. The OHS group also had higher proportions of patients with pulmonary hypertension (50% vs. 2%) and heart failure (76% vs. 6.5%), than the OSA group (Table 1).
There were two independent predictors for OHS, including BMI and serum bicarbonate level. Adjusted odds ratio (95% confidence interval) of these factors were 1.08 (1.01, 1.17) and 1.96 (1.15, 3.34), respectively (Table 2). BMI greater than 25 kg/m2 and serum bicarbonate more than 25 mEq/L yielded 100% sensitivity for OHS (Table 3). Areas under the ROC curves of body mass index and serum bicarbonate for OHS, were 88.37% and 93.88%, respectively (Figs. 1 and 2).


This study found that body mass index and serum bicarbonate, were two independent predictors for OHS among OSA patients (Table 2). This is in contrast to a previous study in the U.S., in which AHI was not found to be a significant factor for OHS (Table 3) [9]. Two other studies conducted in Turkey, also confirmed that serum bicarbonate was an independent predictor for OHS [10,11]. No previous study has evaluated if body mass index is a predictor for OHS, because obesity is a diagnostic criteria for the condition.
Although BMI criterion for OHS is 30 kg/m2, a report in Japan found that average BMI of OHS patients in the Japanese population was 37 [5]. This study also found that median BMI of OHS patients was also more than 30 kg/m2 (48.9 kg/m2) as shown in Table 1. However, the appropriate BMI cutoff point may not be 30 kg/m2 in this study. There were 8% of Thai OSA patients with OHS, who may not be diagnosed with BMI of 30 kg/m2 (Table 3). Findings may imply that Thai OSA patients could develop OHS, even if they have lower BMI than 30 kg/m2.
Likewise, setting the cutoff point of serum bicarbonate at 27 mEq/L or more may lead to under-diagnoses of OHS, in approximately 8% of Thai OSA patients (Table 3). This cutoff point has been shown to yield a sensitivity of 92% in the U.S. population, and 76–88% in the Turkish population [9-11]. However, according to our research, if the cutoff point for serum bicarbonate is set at 25 mEq/L, sensitivity increases to 100%, as it is when BMI is set at 25 kg/m2 (Table 3). Findings raise the concern that current diagnostic criteria for OHS, may not be suitable in Thai or other Asian populations.
There are limitations to this study. First, all participants had been diagnosed with OSA. However, OHS may develop without presence of OSA, in approximately 10% of patients. So, findings are applicable only to patients with OSA. Second, this study was conducted in only one university hospital setting in Thailand. Further studies in other settings in Asia are needed. Third, there are some missing values, such as serum HCO3 in the OSA group. Finally, the study population is small. However, other OHS studies also had study populations of fewer than 200 participants.
In conclusion, appropriate diagnostic criteria for OHS in Thai patients, may be different from those in Western populations.


The authors would like to thank Mr. Dylan Southard for his English language editing and Sleep Apnea Research Group, and Research and Training Center for Enhancing Quality of Life of Working Age People, Khon Kaen University, Khon Kaen, Thailand.


Conflicts of Interest
The authors have no financial conflicts of interest.
Authors’ Contribution
Conceptualization: Sawanyawisuth K, Saeseow S, Chattakul P. Data curation: Saeseow S. Methodology: Chattakul P, Khamsai S, Limpawattana P, Chindaprasirt J. Supervision: Chotmongkol V, Silarucks S, Senthong V, Sawanyawisuth K. Writing—original draft: Sawanyawisuth K. Writing—review & editing: all authors.


1. Sateia MJ. International classification of sleep disorders-third edition: highlights and modifications. Chest 2014; 146: 1387-94.
crossref pmid
2. Berg G, Delaive K, Manfreda J, Walld R, Kryger MH. The use of health-care resources in obesity-hypoventilation syndrome. Chest 2001; 120: 377-83.
crossref pmid
3. Mokhlesi B. Obesity hypoventilation syndrome: a state-of-the-art review. Respir Care 2010; 55: 1347-62.
4. Balachandran JS, Masa JF, Mokhlesi B. Obesity hypoventilation syndrome: epidemiology and diagnosis. Sleep Med Clin 2014; 9: 341-7.
crossref pmid pmc
5. Akashiba T, Akahoshi T, Kawahara S, Uematsu A, Katsura K, Sakurai S, et al. Clinical characteristics of obesity-hypoventilation syndrome in Japan: a multi-center study. Intern Med 2006; 45: 1121-5.
crossref pmid
6. Harada Y, Chihara Y, Azuma M, Murase K, Toyama Y, Yoshimura C, et al. Obesity hypoventilation syndrome in Japan and independent determinants of arterial carbon dioxide levels. Respirology 2014; 19: 1233-40.
crossref pmid
7. Alzaabi A, Fizal S, Moilothkandy R, Mahboub B, Nagelkerke N. Obesity hypoventilation syndrome in obstructive sleep apnea patients in the United Arab Emirates: a retrospective cross-sectional study. JRSM Short Rep 2013; 4: 2042533313510156.
crossref pmid pmc
8. BaHammam AS. Prevalence, clinical characteristics, and predictors of obesity hypoventilation syndrome in a large sample of Saudi patients with obstructive sleep apnea. Saudi Med J 2015; 36: 181-9.
crossref pmid pmc
9. Mokhlesi B, Tulaimat A, Faibussowitsch I, Wang Y, Evans AT. Obesity hypoventilation syndrome: prevalence and predictors in patients with obstructive sleep apnea. Sleep Breath 2007; 11: 117-24.
crossref pmid
10. Bingol Z, Pıhtılı A, Cagatay P, Okumus G, Kıyan E. Clinical predictors of obesity hypoventilation syndrome in obese subjects with obstructive sleep apnea. Respir Care 2015; 60: 666-72.
crossref pmid
11. Basoglu OK, Tasbakan MS. Comparison of clinical characteristics in patients with obesity hypoventilation syndrome and obese obstructive sleep apnea syndrome: a case-control study. Clin Respir J 2014; 8: 167-74.
crossref pmid

Fig. 1.
A receiver operating characteristic (ROC) curve, of Body Mass Index and Obesity Hypoventilation Syndrome.
Fig. 2.
A receiver operating characteristic (ROC) curve, of serum bicarbonate and Obesity Hypoventilation Syndrome.
Table 1.
Baseline clinical features of OHS and OSA
Factors OHS (n = 25) OSA (n = 108) p-value
Age (yr) 48 (35–59) 49 (37–59) 0.167
Male sex (%) 10 (40) 57 (52.8) 0.250
Weight (kg) 83 (67–106) 78.4 (65–95) -
Height (cm) 161 (155–170) 162 (155–170) -
BMI (kg/m2) 48.9 (39.5–57.5) 29.2 (25.7–35.4) < 0.001
Mallampati (%)
 Class 1 1 (7.1) 4 (5.4)
 Class 2 1 (7.1) 27 (36.5)
 Class 3 3 (21.4) 31 (41.89) 0.001
 Class 4 9 (64.3) 12 (15.22)
Hypertension (%) 13 (52) 84 (77.8) 0.009
Diabetes mellitus (%) 12 (48) 27 (25) 0.023
Stroke (%) 0 (0) 7 (6.5) 0.191
Coronary artery disease (%) 1 (4) 8 (7.4) 0.541
Heart failure (%) 19 (76) 7 (6.5) < 0.001
Atrial fibrillation (%) 2 (8) 1 (0.9) 0.032
SBP (mm Hg) 140 (127.5–150.5) 140 (129–151) 0.087
DBP (mm Hg) 85 (76–94) 85 (78–94) 0.223
Hct (%) 40.0 (37.1–44.8) 39.7 (36.3–44.0) 0.024
WBC (cell/mm3) 7700 (6400–9900) 7450 (5700–8850) 0.050
Plt (cell/mm3) 241000 (198000–285000) 245000 (198000–280000) 0.547
BUN 12.7 (9.5–16.3) 12.8 (9.7–16.2) 0.891
Cr 0.8 (0.7–1.0) 0.85 (0.7–1.1) 0.413
Serum HCO3 (mg/dL) 49 (37–59) 24.8 (22.7–26.8) < 0.001
AHI 21.5 (12–43) 20 (10–39) 0.110
Lowest spO2 (%) 83.5 (71–88) 83.5 (72–89.5) 0.948
EF 60 (56–65) 67 (64–72) 0.001
PHT (%) 9 (50) 2 (2.11) 0.000

Data presented as median (interquartile range) or number (percentage).

OHS: Obesity Hypoventilation Syndrome, OSA: obstructive sleep apnea, BMI: Body Mass Index, SBP: systolic blood pressure, DBP: diastolic blood pressure, Hct: hematocrit, WBC: white blood cells, Plt: platelet, BUN: blood urea nitrogen, Cr: creatinine, HCO3: bicarbonate, AHI: apnea hypopnea index, spO2: oxygen saturation from polysomnography, EF: ejection fraction, PHT: pulmonary hypertension.

Table 2.
Factors associated with obesity hypoventilation syndrome by multivariate logistic regression analysis
Factors Adjusted odds ratio 95% confidence interval
Male sex 5.95 0.30–118.17
Systolic blood pressure 0.95 0.87–1.04
Apnea hypopnea index 1.01 0.97–1.06
Body Mass Index 1.09 1.01–1.17
Serum bicarbonate 1.96 1.15–3.33
Table 3.
Various cutoff points and its diagnostic properties of body mass index and serum bicarbonate for obesity hypoventilation syndrome
Cutoff points Sensitivity Specificity
Body mass index
 25 100 22
 28 100 43
 30 92 53
Serum bicarbonate
 25 100 52
 27 92 76
 29 76 98
PDF Links  PDF Links
PubReader  PubReader
ePub Link  ePub Link
XML Download  XML Download
Full text via DOI  Full text via DOI
Download Citation  Download Citation
Similar articles within SMR
Predictors for Obesity Hypoventilation Syndrome in Thai Population  2021 December;12(2)