Abstract
Objective
This study aims to utilize quantitative computed tomography (QCT) to compare body composition parameters and bone mineral density (BMD) between two distinct patient groups: individuals with obesity and obstructive sleep apnea (OSA) and those with obesity alone (non-OSA group). By systematically evaluating QCT-derived metrics, we seek to identify specific imaging biomarkers that may aid in distinguishing OSA among obese individuals and to assess their potential diagnostic utility.
Methods
A total of 77 patients from Yan'an Hospital Affiliated with Kunming Medical University were enrolled and categorized into groups based on the presence of OSA. QCT was employed to collect data on fat-related and BMD parameters. The correlation between these parameters and OSA was analyzed, influencing factors were identified, and the diagnostic utility was assessed through receiver operating characteristic (ROC) curve analysis.
Results
The OSA group exhibited greater subcutaneous fat area, visceral fat area, total fat area, total fat at the inferior margin of L1-L4 vertebrae, visceral fat, subcutaneous fat weight, area, and volume compared to the non-OSA group. Additionally, the OSA group had a lower BMD at the L1 vertebra compared to the non-OSA group, and all of these differences were statistically significant (P < 0.05). ROC curve analysis indicated that the area under the ROC curve (AUC) for body weight and subcutaneous fat area at the inferior margin of L1 vertebra was 0.751 and 0.726, respectively.
Conclusion
The QCT-derived measurements of intra-abdominal fat, subcutaneous fat, and BMD exhibit variations between patients with obesity and OSA compared to those with obesity alone. These differences serve as a potential foundation for identifying OSA in the obese demographic. Notably, the weight and subcutaneous fat area at the lower boundary of the L1 vertebra demonstrate enhanced diagnostic efficacy.
Keywords: Abdominal fat, Bone mineral density (BMD), Obstructive sleep apnea (OSA), Quantitative CT (QCT), Visceral fat
Background
Obstructive sleep apnea (OSA) is a prevalent disorder characterized by recurrent collapse of the upper airway during sleep, resulting in obstructive apnea and hypopnea, and frequently accompanied by snoring, sleep fragmentation, and intermittent hypoxia. These disturbances not only impair sleep quality but also contribute significantly to the development and exacerbation of various metabolic, cardiovascular, and cerebrovascular conditions.Obesity represents one of the most prominent risk factors for OSA, with epidemiological studies consistently demonstrating a dose-response relationship between increasing body weight and OSA incidence [1, 2].
Emerging evidence further underscores the role of specific adiposity indicators in this context. For instance, the Chinese visceral adiposity index has been significantly associated with an elevated risk of new-onset myocardial infarction among patients with OSA [3], while the body roundness index and weight-adjusted-waist index have also been correlated with cardiovascular disease risk in hypertensive OSA patients—the latter exhibiting a J-shaped relationship [4, 5].
The situation is particularly concerning in China, where obesity has emerged as a major public health challenge. More than 50% of adults are now classified as overweight or obese [6], yet public awareness of OSA remains strikingly low. This lack of awareness, combined with the high prevalence of obesity, contributes to a substantial rate of underdiagnosis [7].
Although polysomnography (PSA) remains the gold standard for OSA diagnosis, its practical utility is limited by complexity, cost, and accessibility [8], highlighting an urgent need for simpler, more scalable screening tools. In this context, biomarkers such as visfatin have attracted interest for their potential role in the pathophysiology of obstructive sleep apnea hypopnea syndrome (OSAHS) [9]. Simultaneously, quantitative computed tomography (QCT) has emerged as a promising modality for body composition analysis, providing precise measures of visceral and subcutaneous adipose tissue, hepatic and muscular fat deposition, and bone mineral density (BMD). Although QCT shows considerable potential in chronic disease imaging, its utility as a predictor of OSA in obese populations remains inadequately explored [9].
To address this knowledge gap, the present study aims to assess obesity-related indices and sleep parameters among snoring obese individuals. Specifically, we seek to evaluate the correlation between QCT-derived adiposity measures, BMD, and OSA presence and severity. Furthermore, we will employ receiver operating characteristic (ROC) curve analysis to determine the predictive capacity of these indices for OSA. Ultimately, our goal is to establish an accessible and efficient screening framework capable of facilitating early OSA detection in high-risk obese populations, thereby addressing the limitations inherent in current diagnostic approaches.
Data and methods
Participants
The information of obese patients admitted to Yan’an Hospital Affiliated to Kunming Medical University, specifically in the Weight Loss Surgery Department and General Surgery Department, during the period of January to June 2021 was retrospectively collected. These patients were scheduled for metabolic weight loss surgery. The inclusion criteria were as follows: (1) age 18 years or older; (2) underwent portable sleep monitoring for at least 7 h overnight; (3) had a body mass index (BMI) ≥ 25 kg/m2; (4) complete clinical data was available. Exclusion criteria included: (1) prior treatment for OSA; (2) individuals with severe cardiac, hepatic, or renal insufficiency. The retrospective nature of the study led to an exemption from the requirement of an informed consent form, in compliance with the regulations set forth by the Ethics Committee of Yan’an Hospital of Kunming City.
All patients underwent QCT of the upper abdomen.
Grouping
A total of 77 participants, comprising 69 females and 8 males, with an average age of (31.49 ± 8.37), were included in the study. The participants were categorized into two groups based on their clinical diagnosis of OSA: the OSA group and the non-OSA group.
Methods
General data collection
The height and weight of the enrolled participants were collected, and subsequently the BMI value was calculated; BMI = body mass (kg)/height (m)2. The outcomes of liver and kidney functions in participants were collected two hours post-meal.
QCT examination
A Canon (USA) 320 CT scanner was utilized for imaging, and monthly quality control checks were conducted using a phantom before each scan. Participants were positioned supine, holding their head with both hands, aligning the centerline with the xiphoid process. Scans were performed with breath-holding after inhalation. The scanning parameters included a tube voltage of 120 kV, 100 mA, a reconstruction layer thickness of 1.25 mm, and a scanning range focused on the upper abdomen. The following HU threshold were used for fat segmentation: visceral fat (VAT): − 150 to − 50 HU; subcutaneous fat (SAT): − 190 to − 30 HU.
Image analysis
The outcomes of abdominal QCT were transferred to the QCT processing software workstation (Beijing Guangda Hongshun) for further analysis.
QCT was employed for the measurement of single-level body composition. The body composition analysis module was utilized, and CT threshold segmentation technology was applied to distinguish between soft tissue and fat. This allowed for the separate coloration of areas, enabling the assessment of area, volume, and weight for each segment to evaluate overall body composition. Various parameters such as total fat weight (g), total fat area, total fat volume, intra-abdominal fat weight (g), intra-abdominal fat area, intra-abdominal fat volume, subcutaneous fat area, subcutaneous fat volume, and subcutaneous fat weight (g) at the inferior margin of L1-L4 vertebrae were measured. Additionally, mean liver fat content, mean muscle fat content, subcutaneous fat area, visceral fat area, total fat area, T12 BMD, L1 BMD, L2 BMD, and mean BMD were assessed (refer to Figs. 1 and 2).
Fig. 1.
Female, 22 years old, with OSA. A-I shows the regions of interest of T12-L2 vertebrae located for QCT measurement of BMD; J-K shows the intra-abdominal fat (within the green circle) and subcutaneous fat (outside the green circle) delineated by QCT. A-C shows a schematic transverse view of T12-L2 vertebrae, D-F shows a schematic sagittal view of T12-L2 vertebrae, and G-I shows a schematic coronal view of T12-L2 vertebrae; Female, 22 years old, with OSA. A-I shows the regions of interest of T12-L2 vertebrae located for QCT measurement of BMD. A-C shows a schematic transverse view of T12-L2 vertebrae, D-F shows a schematic sagittal view of T12-L2 vertebrae, and G-I shows a schematic coronal view of T12-L2 vertebrae; J-K shows the intra-abdominal fat (within the green circle) and subcutaneous fat (outside the green circle) delineated by QCT, with (J) being a bone window transverse view and (K) being a pseudo-color image, where blue represents adipose tissue and yellow represents non-fat tissue
Fig. 2.
Box plot comparing the fat areas at the L1 vertebral levelbetween the two groups: L1-TFW(g) is the abbreviation of L1 vertebra total fat weight (g); L1-TFV(mm3) is the abbreviation of L1 vertebra total fat volume(mm3) ;L1-IAFW(g) is the abbreviation of L1 vertebra intra-abdominal fat weight (g); L1-IAFV(mm3) is the abbreviation of L1 vertebra intra-abdominal fat volume(mm3) ;L1-subFW(g) is the abbreviation of L1 vertebra subcutaneous fat weight (g); L1-subFV(mm3) is the abbreviation of L1 vertebra subcutaneous fat volume(mm3)
Statistical methods
The data were processed using SPSS 19.0 statistical software. Measurement data following a normal distribution are presented as mean ± standard deviation, and independent samples t-tests were employed for data comparison. Enumeration data are presented as the number of cases (%) and subjected to inter-group comparison using theoretical frequency lattice, with distribution comparison tested through chi-squared and Fisher tests. Measurement data not adhering to a normal distribution are expressed as median [25th percentile, 75th percentile], and inter-group comparisons were conducted using the nonparametric Kruskal–Wallis rank sum test. Statistical significance was defined as P < 0.05. Multiple imputation was performed using the Fully Conditional Specification method with 20 imputed datasets. To enhance imputation accuracy, all analysis variables were included along with auxiliary covariates. Convergence was confirmed via trace plots, and estimates were pooled using Rubin’s rules. Sensitivity analyses comparing multiple imputation with complete-case analysis showed consistent effect estimates (Δβ < 10%).
Results
General data
This study comprised 77 participants, whose ages ranged from 18 to 52 years, with an average age of (31.51 ± 8.34) years. The group consisted of 69 females and 8 males. Among them, 29 patients (37.7%) were classified in the OSA group, while 48 participants (62.3%) were categorized into the non-OSA group.
The weights and BMI of the OSA group were found to be higher than those of the non-OSA group, and the levels of high-density lipoprotein (HDL) and apolipoprotein a1, which are related to liver function, were observed to be lower than those of the non-OSA group. These differences were determined to be statistically significant (P < 0.05). Additionally, there were no statistically significant differences (P > 0.05) observed in age, height, cholinesterase, triglyceride, low-density lipoprotein (LDL), and apolipoprotein b between the two groups, as indicated in Table 1.
Table 1.
Comparison of general conditions between the OSA and non-OSA groups (± s)
| Item | OSA group | Non-OSA group | t/χ2 value | p value |
|---|---|---|---|---|
| Age | 31.52 ± 8.40 | 31.51 ± 8.34 | 0.019 | 0.985 |
| Height | 164.07 ± 7.49 | 163.58 ± 6.47 | 0.299 | 0.766 |
| Weight | 112.00 ± 26.07 | 92.18 ± 15.22 | 3.728 | 0.01* |
| BMI | 42.03 ± 8.43 | 34.18 ± 4.46 | 4.570 | < 0.01* |
| Cholinesterase | 9003.10 ± 1515.62 | 8625.98 ± 1926.43 | 0.899 | 0.372 |
| Triglycerides | 1.88 ± 0.86 | 1.97 ± 1.23 | -0.349 | 0.728 |
| HDL | 1.06 ± 0.21 | 1.25 ± 0.44 | -2.164 | 0.034* |
| LDL | 3.13 ± 0.89 | 3.08 ± 0.73 | 0.306 | 0.760 |
| Apolipoprotein a1 | 1.19 ± 0.18 | 1.30 ± 0.26 | -2.290 | 0.025* |
| Apolipoprotein b | 1.19 ± 0.26 | 3.43 ± 15.42 | -0.781 | 0.437 |
*P < 0.05
QCT body composition analysis
Higher values for total fat, visceral fat, subcutaneous fat weight, area, and volume were observed in the OSA group compared to the non-OSA group. The total subcutaneous fat area, visceral fat area, and total fat area at the inferior margin of L1-L4 vertebrae were measured, and statistically significant differences (P < 0.05) were noted, as depicted in Table 2; Figs. 1, 2, 3, 4, 5 and 6.
Table 2.
Assessing body composition differences between the OSA and non-OSA groups (± s)
| Item | OSA group | Non-OSA group | p value | |
|---|---|---|---|---|
| L1 vertebra | Inferior margin of vertebra: total fat weight (g) | 58.23 ± 23.37 | 43.59 ± 11.78 | 0.003 |
| Inferior margin of vertebra: total fat area | 630.89 ± 253.31 | 472.15 ± 127.81 | 0.003 | |
| Inferior margin of vertebra: total fat volume (TAFV) | 63.09 ± 25.33 | 47.22 ± 12.76 | 0.003 | |
| Inferior margin of vertebra: intra-abdominal fat weight (g) | 22.42 ± 12.02 | 16.64 ± 5.54 | 0.020 | |
| Inferior margin of vertebra: intra-abdominal fat area | 242.86 ± 130.25 | 180.25 ± 60.01 | 0.020 | |
| Inferior margin of vertebra: intra-abdominal fat volume (TAFV) | 24.28 ± 13.03 | 18.03 ± 6 | 0.020 | |
| Inferior margin of vertebra: subcutaneous fat weight (g) | 35.81 ± 13.55 | 26.95 ± 9.84 | 0.001 | |
| Inferior margin of vertebra: subcutaneous fat area | 388.03 ± 146.8 | 291.9 ± 106.87 | 0.001 | |
| Inferior margin of vertebra: subcutaneous fat volume (TAFV) | 38.81 ± 14.69 | 29.19 ± 10.68 | 0.001 | |
| L2 vertebra | Inferior margin of vertebra: total fat weight (g) | 63.06 ± 22.41 | 48.08 ± 13.16 | 0.002 |
| Inferior margin of vertebra: total fat area | 683.25 ± 242.8 | 520.86 ± 142.61 | 0.002 | |
| Inferior margin of vertebra: total fat volume (TAFV) | 67.28 ± 25.69 | 52.09 ± 14.26 | 0.006 | |
| Inferior margin of vertebra: intra-abdominal fat weight (g) | 24.76 ± 12.2 | 17.96 ± 6.03 | 0.008 | |
| Inferior margin of vertebra: intra-abdominal fat area | 267.22 ± 130.95 | 194.5 ± 65.35 | 0.008 | |
| Inferior margin of vertebra: intra-abdominal fat volume (TAFV) | 26.62 ± 13.01 | 19.45 ± 6.53 | 0.009 | |
| Inferior margin of vertebra: subcutaneous fat weight (g) | 38.3 ± 13.08 | 30.11 ± 11.11 | 0.004 | |
| Inferior margin of vertebra: subcutaneous fat area | 416.03 ± 141.77 | 326.36 ± 120.38 | 0.004 | |
| Inferior margin of vertebra: subcutaneous fat volume (TAFV) | 40.66 ± 16.26 | 32.64 ± 12.04 | 0.016 | |
| L3 vertebra | Inferior margin of vertebra: total fat weight (g) | 67.25 ± 24.89 | 51.82 ± 14.48 | 0.004 |
| Inferior margin of vertebra: total fat area | 728.6 ± 269.7 | 561.33 ± 156.89 | 0.004 | |
| Inferior margin of vertebra: total fat volume (TAFV) | 72.86 ± 26.98 | 56.13 ± 15.69 | 0.004 | |
| Inferior margin of vertebra: intra-abdominal fat weight (g) | 24.45 ± 12.47 | 17.84 ± 6.25 | 0.011 | |
| Inferior margin of vertebra: intra-abdominal fat area | 264.95 ± 135.08 | 193.26 ± 67.76 | 0.011 | |
| Inferior margin of vertebra: intra-abdominal fat volume (TAFV) | 26.5 ± 13.51 | 19.31 ± 6.77 | 0.011 | |
| Inferior margin of vertebra: subcutaneous fat weight (g) | 42.8 ± 15.29 | 33.98 ± 12.23 | 0.007 | |
| Inferior margin of vertebra: subcutaneous fat area | 463.66 ± 165.62 | 368.07 ± 132.49 | 0.007 | |
| Inferior margin of vertebra: subcutaneous fat volume (TAFV) | 46.36 ± 16.58 | 36.82 ± 13.25 | 0.007 | |
| L4 vertebra | Inferior margin of vertebra: total fat weight (g) | 72.49 ± 28.02 | 55.32 ± 15.56 | 0.004 |
| Inferior margin of vertebra: total fat area | 785.47 ± 303.61 | 582.59 ± 177.98 | 0.002 | |
| Inferior margin of vertebra: total fat volume (TAFV) | 78.55 ± 30.36 | 59.94 ± 16.85 | 0.004 | |
| Inferior margin of vertebra: intra-abdominal fat weight (g) | 24.07 ± 13.29 | 16.5 ± 5.71 | 0.006 | |
| Inferior margin of vertebra: intra-abdominal fat area | 260.73 ± 143.88 | 178.78 ± 61.81 | 0.006 | |
| Inferior margin of vertebra: intra-abdominal fat volume (TAFV) | 26.08 ± 14.39 | 17.88 ± 6.18 | 0.006 | |
| Inferior margin of vertebra: subcutaneous fat weight (g) | 48.42 ± 17.14 | 38.82 ± 12.55 | 0.006 | |
| Inferior margin of vertebra: subcutaneous fat area | 524.74 ± 185.69 | 403.81 ± 154.25 | 0.003 | |
| Inferior margin of vertebra: subcutaneous fat volume (TAFV) | 52.48 ± 18.57 | 42.06 ± 13.59 | 0.006 | |
| Total | Total soft tissue area | 1004.05 ± 275.47 | 808.21 ± 168.09 | 0.001 |
| Total muscle area | 317.77 ± 92.21 | 290.35 ± 44.79 | 0.143 | |
| SAT subcutaneous fat area | 415.77 ± 141.72 | 322.71 ± 126.03 | 0.004 | |
| VAT visceral fat area | 274.55 ± 132.85 | 197.47 ± 66.94 | 0.006 | |
| Total fat area | 690.32 ± 244.06 | 520.18 ± 150.57 | 0.002 |
Fig. 3.
Box plot comparing the fat areas at the L2 vertebral levelbetween the two groups: L2-TFW(g) is the abbreviation of L2 vertebra total fat weight (g); L2-TFV(mm3) is the abbreviation of L2 vertebra total fat volume(mm3) ;L2-IAFW(g) is the abbreviation of L2 vertebra intra-abdominal fat weight (g); L2-IAFV(mm3) is the abbreviation of L2 vertebra intra-abdominal fat volume(mm3) ;L2-subFW(g) is the abbreviation of L2 vertebra subcutaneous fat weight (g); L2-subFV(mm3) is the abbreviation of L2 vertebra subcutaneous fat volume(mm3)
Fig. 4.
Box plot comparing the fat areas at the L3 vertebral levelbetween the two groups: L3-TFW(g) is the abbreviation of L3 vertebra total fat weight (g); L3-TFV(mm3) is the abbreviation of L3 vertebra total fat volume(mm3) ;L3-IAFW(g) is the abbreviation of L3 vertebra intra-abdominal fat weight (g); L3-IAFV(mm3) is the abbreviation of L3 vertebra intra-abdominal fat volume(mm3) ;L3-subFW(g) is the abbreviation of L3 vertebra subcutaneous fat weight (g); L3-subFV(mm3) is the abbreviation of L3 vertebra subcutaneous fat volume(mm3)
Fig. 5.
Box plot comparing the fat areas at the L4 vertebral levelbetween the two groups:L4-TFW(g) is the abbreviation of L4 vertebra total fat weight (g);L4-TFV(mm3) is the abbreviation of L4 vertebra total fat volume(mm3) ;L4-IAFW(g) is the abbreviation of L4 vertebra intra-abdominal fat weight (g);L4-IAFV(mm3) is the abbreviation of L4 vertebra intra-abdominal fat volume(mm3) ;L4-subFW(g) is the abbreviation of L4 vertebra subcutaneous fat weight (g); L4-subFV(mm3) is the abbreviation of L4 vertebra subcutaneous fat volume(mm3)
Fig. 6.
Box plot comparing the abdominal fat volume corresponding to each vertebral body between the two groups
QCT BMD analysis
In this study, a statistically significant difference was observed in the BMD at the L1 vertebra level between the OSA group and the non-OSA group, with the BMD of the OSA group being lower than that of the non-OSA group (P < 0.05). The BMD differences at the levels of T12 and L2 vertebrae and the mean BMD between the two groups were not found to be statistically significant (P > 0.05), as presented in Table 3.
Table 3.
Contrast in bone mineral density (BMD) as well as fat content in the liver and muscle between OSA and non-OSA groups (± s)
| Item | OSA group | Non-OSA group | t/χ2 value | p value |
|---|---|---|---|---|
| T12 BMD | 156.64 ± 24.84 | 168.46 ± 32.65 | -1.676 | 0.098 |
| L1 BMD | 150.15 ± 27.12 | 166.52 ± 32.21 | -2.289 | 0.025 |
| L2 BMD | 148.21 ± 28.93 | 162.3 ± 33.25 | -1.888 | 0.063 |
| Mean BMD | 151.63 ± 26.74 | 165.68 ± 32.2 | -1.974 | 0.052 |
| Mean liver fat content (%) | 21.8 ± 10.84 | 16.76 ± 8.3 | 2.299 | 0.024 |
| Mean muscle fat content (%) | 7.63 ± 4.75 | 7.03 ± 4.17 | 0.575 | 0.567 |
Analysis of QCT liver fat content and muscle fat content
The mean liver fat content (%) was found to be higher in the OSA group compared to the non-OSA group, and a statistically significant difference was observed (P < 0.05). The mean muscle fat difference between the two groups was not found to be statistically significant (P > 0.05), as illustrated in Table 3.
ROC curve analysis of statistically different parameters between the two groups
ROC curve analysis was conducted, utilizing parameters that exhibited statistical differences between the two groups. Among these parameters, 15 displayed an area under the curve (AUC) ranging from 0.700 to 0.751. The specific parameters used for analysis included body weight, subcutaneous fat area at the inferior margin of the L1 vertebra, subcutaneous fat volume (TAFV), subcutaneous fat weight (g), total fat weight (g), total fat volume (TAFV), total fat area, and total soft tissue area. Additionally, subcutaneous fat area and volume at the inferior margins of the L2 and L4 vertebrae were considered.
When these parameters were individually employed as indexes, the resulting AUCs were as follows: 0.751, 0.726, 0.726, 0.725, 0.724, 0.724, 0.723, 0.723, 0.718, 0.717, 0.715, 0.711, 0.707, 0.704, and 0.700, respectively. The optimal cut-off values, along with sensitivity and specificity, were determined for each parameter: 96.50 (0.759, 0.729), 325.35 (0.724, 0.667), 32.55 (0.724, 0.667), 30.20 (0.724, 0.687), 45.20 (0.759, 0.667), 48.95 (0.759, 0.667), 490.05 (0.759, 0.667), 521.05 (0.759, 0.646), 49.55 (0.724, 0.667), 536.80 (0.724, 0.667), 339.65 (0.69, 0.646), 815.80 (0.724, 0.646), 344.30 (0.69, 0.646), 31.90 (0.69, 0.667), and 433.95 (0.69, 0.646).
Notably, the diagnostic efficiency of the remaining parameters was below 0.700, as detailed in Table 4; Fig. 7.
Table 4.
Diagnostic effectiveness of parameters exhibiting both statistical differences between the two groups and an AUC surpassing 0.7 in OSA
| Item | AUC | P value | Critical value | Youden’s index | Sensitivity | Specificity |
|---|---|---|---|---|---|---|
| Weight | 0.751 | < 0.001 | 96.50 | 0.488 | 75.90% | 72.90% |
| Inferior margin of L1 vertebra: subcutaneous fat area | 0.726 | 0.001 | 325.35 | 0.391 | 72.40% | 66.70% |
| Inferior margin of L1 vertebra: subcutaneous fat volume (TAFV) | 0.726 | 0.001 | 32.55 | 0.391 | 72.40% | 66.70% |
| Inferior margin of L1 vertebra: subcutaneous fat weight (g) | 0.725 | 0.001 | 30.20 | 0.411 | 72.40% | 68.70% |
| Inferior margin of L1 vertebra: total fat weight (g) | 0.724 | 0.001 | 45.20 | 0.426 | 75.90% | 66.70% |
| Inferior margin of L1 vertebra: total fat volume (TAFV) | 0.724 | 0.001 | 48.95 | 0.426 | 75.90% | 66.70% |
| Inferior margin of L1 vertebra: total fat area | 0.723 | 0.001 | 490.05 | 0.426 | 75.90% | 66.70% |
| Total fat area | 0.723 | 0.001 | 521.05 | 0.405 | 75.90% | 64.60% |
| Inferior margin of L2 vertebra: total fat weight (g) | 0.718 | 0.001 | 49.55 | 0.391 | 72.40% | 66.70% |
| Inferior margin of L2 vertebra: total fat area | 0.717 | 0.001 | 536.80 | 0.391 | 72.40% | 66.70% |
| SAT subcutaneous fat area | 0.715 | 0.002 | 339.65 | 0.336 | 69.00% | 64.60% |
| Total soft tissue area | 0.711 | 0.002 | 815.80 | 0.370 | 72.40% | 64.60% |
| Inferior margin of L2 vertebra: subcutaneous fat area | 0.707 | 0.002 | 344.30 | 0.336 | 69.00% | 64.60% |
| Inferior margin of L2 vertebra: subcutaneous fat volume (TAFV) | 0.704 | 0.003 | 31.90 | 0.357 | 69.00% | 66.70% |
| Inferior margin of L4 vertebra: subcutaneous fat area | 0.700 | 0.003 | 433.95 | 0.336 | 69.00% | 64.60% |
Fig. 7.
ROC curve analysis of statistically different parameters between the two groups: The figure visually summarizes the performance of multiple parameters, with area under the curve (AUC) values ranging from 0.700 to 0.751. It specifically highlights parameters such as body weight and subcutaneous fat area at the L1 vertebra inferior margin, which demonstrated the highest diagnostic efficacy (AUC 0.751 and 0.726, respectively), as emphasized in the Results and Conclusion
Discussion
General condition and differences in blood fat parameters in individuals with obesity and OSA
OSA, a common sleep respiratory disorder, holds the first position globally in terms of patient numbers in China [10]. Mutual influencing factors exist between OSA and obesity, as evidenced by a study indicating that 50%~70% of individuals with OSA are found to be obese, and 40%~90% of individuals with obesity concurrently experience OSA. The severity of OSA tends to escalate with increasing BMI. In this study, individuals with OSA and obesity exhibited higher weight and BMI compared to those with obesity alone. Notably, weight emerged as a highly effective diagnostic parameter, with a critical value of 96.5 kg. Consistent with numerous studies [11–13], the prevalence of OSA is observed to rise alongside increasing body weight and BMI. This correlation may be attributed to a higher obesity index leading to an increase in neck adipose tissue, accumulation of peripharyngeal soft tissues, muscle relaxation, and pharyngeal lumen narrowing, thereby causing OSA. The persistent hypoxia induced by OSA contributes to fatigue, reduced body mobility, decreased lipolysis, and exacerbated obesity, creating a vicious cycle.
A blood lipid analysis revealed lower levels of HDL and apolipoprotein a1 in individuals with OSA and obesity compared to those with obesity alone. This phenomenon is attributed to the impact of OSA on lipid metabolism [13]. The underlying mechanism remains unclear, but prolonged hypoxia may activate the sterol-regulating element binding protein-1 pathway, leading to hyperlipidemia [14]. Elevated body weight and related blood lipid parameters render obese individuals more susceptible to OSA.
Differences in body composition parameters in individuals with obesity and OSA based on QCT
Abdominal obesity has been identified as a high-risk factor for OSA [15, 16], and the potential role of visceral fat in the development of OSA has been suggested in relevant research [17]. The distribution of body fat is inadequately reflected by BMI and body weight. Abdominal circumference solely indicates the presence of abdominal obesity and lacks the capacity to distinguish between subcutaneous and visceral fat content. QCT technology has evolved as a method for measuring abdominal fat. The measurement involves utilizing data obtained from routine clinical CT scans, the QCT quality control phantom, and the corresponding QCT post-processing software analysis system. This comprehensive approach enables the quantitative assessment of BMD, abdominal visceral fat content, subcutaneous fat content, liver fat content, and muscle fat content in a single scan. QCT has been extensively applied in BMD measurement [18–21]. Its use in abdominal fat measurement has increased, with a study proposing QCT as a potential “gold standard” for such measurements [22].
Limited studies currently employ QCT to evaluate the distribution of abdominal fat in individuals with OSA. The results of this study revealed that individuals with OSA and obesity exhibit higher weight, area, and volume of total fat/visceral fat/subcutaneous fat at the inferior margin of L1-L4 vertebrae, mean muscle fat, and overall subcutaneous visceral fat area, visceral fat area, and total fat area compared to those without OSA. Subcutaneous fat and total fat indices demonstrate high diagnostic efficiency, while total visceral fat and visceral fat indices at each vertebra level display moderate diagnostic efficiency.
The assessment of OSA based on visceral fat remains contentious. Some studies suggest that visceral fat can be used to gauge OSA incidence rates, possibly due to metabolic disorders and insulin resistance associated with obesity [23–25]. Conversely, some studies posit that visceral fat cannot reflect the severity of OSA [26]. The optimal cut-off value varies among studies, necessitating large-sample, multicenter studies for consensus.
In this study, subcutaneous fat and total fat content exhibit higher diagnostic efficiency. Subcutaneous fat, considered the body’s energy storage repository, maintains a close association with insulin sensitivity and leptin [27]. Previous studies remain inconclusive about the role of subcutaneous fat, however, the results of this study suggest that higher subcutaneous fat may elucidate a higher incidence of OSA, providing data support for future investigations.
Based on the results of this study, QCT appears to provide valuable insights into the distribution of subcutaneous and visceral fat in the abdominal region. The findings illustrate that utilizing T1, T2, and T4 may offer improved assessments of abdominal fat in individuals with OSA. Currently, various studies uphold the notion that evaluating fat at the L1 and L2 levels is more indicative of overall body fat distribution [28–30]. Notably, given the prevalence of abdominal obesity in all participants in this study, the downward position of the L4 vertebra, influenced by the gravitational effect in the participants while standing, could also be employed to some extent for assessing whole-body fat distribution.
The presence of heightened subcutaneous fat, visceral fat, and total fat at the L1-2 and L4 levels suggests a potential susceptibility to OSA among obese individuals.
Differences in BMD parameters in individuals with obesity and OSA based on QCT
Based on the results of this study, lower BMD was observed in the L1 vertebra of individuals with OSA compared to the participants in the non-OSA group. This result is consistent with other studies [31, 32]. The involvement of OSA in bone metabolism, possibly through intermittent hypoxia, has been proposed as the mechanism [33]. The prolonged chronic inflammatory state of the body due to oxidative stress has also been implicated. RANKL signal activation in OSAHS was identified as a factor, leading to a reduction in the OPG/RANKL ratio, the activation of osteoclast differentiation, and the inhibition of osteoblast activation, ultimately resulting in osteoporosis [34]. Currently, the bone content of vertebrae can be accurately detected through QCT [35–37]. In this study, the diagnostic efficacy of using the L1 vertebra was found to be higher than that of the other vertebrae. This difference may be attributed to the gradual increase in the CT value of vertebral BMD from the top to the bottom, making the reduction of BMD most detectable at L1. The significant decrease in BMD at the L1 vertebra, as detected by QCT, may suggest the susceptibility of obese individuals to OSA.
Limitations of the study
This study involved a small sample size, and a higher proportion of female patients was observed, potentially introducing data bias. Further data collection is deemed necessary for subsequent validation. Additionally, a comparison with healthy individuals was not conducted in this study, highlighting the need for additional data from the validation group in the future. Thirdly, portable sleep monitoring was selected in this study for its practicality in large-scale epidemiological studies, offering advantages such as data collection in participants’ natural sleep environments, higher recruitment rates, and cost reduction for longitudinal cohorts. However, given PSG’s superior accuracy, future validation efforts should include PSG-confirmed studies. Lastly, our study is primarily observational and did not directly measure hypoxia or inflammatory biomarkers. While we identified an association between OSA and low L1 BMD, the underlying mechanisms remain to be validated through hypoxia biomarker assays (e.g., nocturnal SpO2 nadir) and direct assessments of osteoclast/osteoblast activity.
Conclusion
In conclusion, the intra-abdominal fat/subcutaneous fat content, as determined by QCT in individuals with obesity and OSA, was found to be higher than that in individuals with obesity alone. Furthermore, the BMD at the L1 vertebra level in individuals with obesity and OSA was determined to be lower than that in individuals with obesity alone. The aforementioned indices can serve as a basis for screening OSA in the obese population.
Acknowledgements
We would like to acknowledge the hard and dedicated work of all the staff who implemented the intervention and evaluation components of the study.
Abbreviations
- OSA
Obstructive sleep apnea
- PSG
Polysomnography
- QCT
Quantitative computed tomography
- ROC
Receiver operating characteristic curve
- BMI
Body mass index
- BMD
Bone mineral density
Authors’ contributions
Conception and design of the research: Cheng-De Liao, Yan-Kun Zhu, Rui ZhaoAcquisition of data: Song Qi, Xue-Min Dong, Peng-Cheng Ma, Jin Wang, Rui ZhaoAnalysis and interpretation of the data: Zheng-Yun Sun, Song Qi, Jin WangStatistical analysis: Zheng-Yun Sun, Peng-Cheng Ma, Xue-Min Dong, Rui ZhaoObtaining financing: Yu-Sen FengWriting of the manuscript: Yu-Sen FengCritical revision of the manuscript for intellectual content: Cheng-De Liao, Yan-Kun ZhuAll authors read and approved the final draft.
Funding
Yunnan Provincial Department of Science and Technology - Kunming Medical University joint special plan (No.202401AY070001-364); Special Fund for High level Health Technology Talents in Yunnan Province, Yan’an Hospital Affiliated to Kunming Medical University (H-2024078);Kunming Health Science and Technology Talent Training Project (2024-SW (Leading) -11);Yunnan Provincial Department of Science and Technology Major Science and Technology Special Program - Biomedical Special Project (No. 202102AA310003-5).
Data availability
The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.
Declarations
Ethics approval and consent to participate
The retrospective study was approved by ethics committee of the Kunming Yan’an Hospital, the signed informed consent requirement was waved. This study was conducted in accordance with the declaration of Helsinki.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Yu-Sen Feng and Zheng-Yun Sun contributed equally to this work.
References
- 1.DE SOUSA A G, CERCATO C, MANCINI M C, et al. Obesity and obstructive sleep apnea-hypopnea syndrome. Obes Reviews: Official J Int Association Study Obes. 2008;9(4):340–54. [DOI] [PubMed] [Google Scholar]
- 2.Zhang P, Zhang H, Hu YX, et al. [Correlation and predictive value of visceral fat index and obesity combined with obstructive sleep apnea in Chinese]. ShiyongYixue Zahzi. 2023;39(17):2210–4. [Google Scholar]
- 3.Cai X, Li N, Hu J et al. Nonlinear relationship between Chinese visceral adiposity index and New-Onset myocardial infarction in patients with hypertension and obstructive sleep apnoea: insights from a cohort study. J Inflamm Res. 2022;15:687–700. [DOI] [PMC free article] [PubMed]
- 4.Cai X, Song S, Hu J, Zhu Q, Yang W, Hong J, Luo Q, Yao X, Li N. Body roundness index improves the predictive value of cardiovascular disease risk in hypertensive patients with obstructive sleep apnea: a cohort study. Clin Exp Hypertens. 2023;45(1):2259132. [DOI] [PubMed] [Google Scholar]
- 5.Zhao J, Cai X, Hu J. ,et al.J-Shaped relationship between Weight-Adjusted-Waist index and cardiovascular disease risk in hypertensive patients with obstructive sleep apnea: A cohort Study[J].Diabetes. Metabolic Syndrome & Obesity: Targets & Therapy; 2024. p. 17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.PAN X F, WANG L. PAN A. Epidemiology and determinants of obesity in China [J]. The lancet Diabetes & endocrinology. 2021;(9)6:373–92. [DOI] [PubMed]
- 7.GOTTLIEB D J, PUNJABI NM. Diagnosis and management of obstructive sleep apnea: A review [J]. JAMA. 2020;323(14):1389–400. [DOI] [PubMed] [Google Scholar]
- 8.Sleep respiratory Disorders Group, Respiratory Society of Chinese Medical Association, and Sleep respiratory Equipment Group, Respiratory Equipment Technical Committee of Chinese Medical Equipment Association. Expert consensus on screening and management of adults at high risk for obstructive sleep apnea. Chin J Health Manage. 2022;16(8):520–8. [Google Scholar]
- 9.Zhang MM, Pan ZL, Lv WF. [Correlation between abdominal fat and cardiometabolic indexes based on QCT]. Chin J CT MRI. 2022;20(11):101–3. [Google Scholar]
- 10.BENJAFIELD A V, AYAS N T, EASTWOOD P R, et al. Estimation of the global prevalence and burden of obstructive sleep apnoea: a literature-based analysis [J]. Lancet Respiratory Med. 2019;7(8):687–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.DRAGER L F, TOGEIRO S M, POLOTSKY V Y, et al. Obstructive sleep apnea: a cardiometabolic risk in obesity and the metabolic syndrome [J]. J Am Coll Cardiol. 2013;62(7):569–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.KAWADA T, OTSUKA T, NAKAMURA T, et al. Relationship between sleep-disordered breathing and metabolic syndrome after adjustment with cardiovascular risk factors [J]. Diabetes Metabolic Syndrome. 2016;10(2):92–5. [DOI] [PubMed] [Google Scholar]
- 13.Li J. The research of lipid metabolism and its correlative factors in patients with obstructive sleep apnea [PhD Dissertation]. Soochow University; 2016.
- 14.LI J, THORNE L N, PUNJABI NM, et al. Intermittent hypoxia induces hyperlipidemia in lean mice [J]. Circul Res. 2005;97(7):698–706. [DOI] [PubMed] [Google Scholar]
- 15.KRITIKOU I, BASTA M, TAPPOUNI R, et al. Sleep Apnoea and visceral adiposity in middle-aged male and female subjects [J]. Eur Respir J. 2013;41(3):601–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.ZHAO X, XU H, QIAN Y, et al. Abdominal obesity is more strongly correlated with obstructive sleep apnea than general obesity in china: results from two separated observational and longitudinal studies [J]. Obes Surg. 2019;29(8):2535–47. [DOI] [PubMed] [Google Scholar]
- 17.Zhou B, Jiang XZ, Chen RH, et al. [Changes of serum metabolic indexes and adipokines in male patients with metabolic syndrome and sleep apnea]. Shiyong Yixue Zazhi. 2020;36(1):79–83. [Google Scholar]
- 18.Wang CY, Su LL, Qi X, et al. [Quantitative CT technique to evaluate the value of chemotherapy on bone mineral density in patients with lung cancer]. J Chin Practical Diagnosis Therapy. 2022;36(3):268–70. [Google Scholar]
- 19.Zhang Y, Cheng XG, Tian W, et al. [Quantitative CT study of cervical and lumbar volume bone mineral density in healthy Chinese]. Chin J Anat Clin. 2020;25(1):1–7. [Google Scholar]
- 20.JAWORSKI M, KOBYLIŃSKA M. Peripheral quantitative computed tomography of the lower leg in children and adolescents: bone densities, cross-sectional sizes and muscle distribution reference data [J]. J Musculoskel Neuronal Interact. 2021;21(2):215–36. [PMC free article] [PubMed] [Google Scholar]
- 21.TEZOL O, BALCı Y, ALAKAYA M, et al. Bone densitometry measurements in children with neurofibromatosis type 1 using quantitative computed tomography [J]. Singapore Med J. 2022;63(9):520–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Sheu Y, Marshall L M, Holton K F et al. Abdominal body composition measured by quantitative computed tomography and risk of non-spine fractures: the osteoporotic fractures in men (MrOS) study [J]. osteoporosis international: a journal established as result of Cooperation between the European foundation for osteoporosis and the National osteoporosis foundation of the USA. 2013;24(8):2231–41. [DOI] [PMC free article] [PubMed]
- 23.XIA M F, CHEN Y, LIN HD, et al. A indicator of visceral adipose dysfunction to evaluate metabolic health in adult Chinese [J]. Sci Rep. 2016;6:38214. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Yang HH. Correlation between abdominal obesity and type 2 diabetes mellitus with obstructive sleep apnea hypopnea syndrome . [PhD Dissertation]. Chengde Medical College; 2022.
- 25.ZHAO L, ZHU L, SU Z, et al. The role of visceral adipose tissue on improvement in insulin sensitivity following Roux-en-Y gastric bypass: a study in Chinese diabetic patients with mild and central obesity [J]. Gastroenterol Rep. 2018;6(4):298–303. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.MAZZUCA E, BATTAGLIA S, MARRONE O, et al. Gender-specific anthropometric markers of adiposity, metabolic syndrome and visceral adiposity index (VAI) in patients with obstructive sleep apnea [J]. J Sleep Res. 2014;23(1):13–21. [DOI] [PubMed] [Google Scholar]
- 27.HAMDY O, AL-OZAIRI PORRAMATIKULS, E. Metabolic obesity: the paradox between visceral and subcutaneous fat [J]. Curr Diabetes Rev. 2006;2(4):367–73. [DOI] [PubMed] [Google Scholar]
- 28.SCHWEITZER L, GEISLER C, POURHASSAN M, et al. What is the best reference site for a single MRI slice to assess whole-body skeletal muscle and adipose tissue volumes in healthy adults? [J]. Am J Clin Nutr. 2015;102(1):58–65. [DOI] [PubMed] [Google Scholar]
- 29.LV H, LI M, LIU Y, et al. The clinical value and appropriateness criteria of upper abdominal magnetic resonance examinations in patients before and after bariatric surgery: a study of 837 images [J]. Obes Surg. 2020;30(10):3784–91. [DOI] [PubMed] [Google Scholar]
- 30.Zhao L, Lv H, Liu Y et al. [Study on the correlation between abdominal fat distribution and obstructive sleep apnea in patients undergoing bariatric metabolic surgery]. Chin J Obes Metabolic Diseases(Electronic Edition). 2021;(01): 13–8.
- 31.EIMAR H, SALTAJI H. GHORASHI S, Association between sleep apnea and low bone mass in adults: a systematic review and meta-analysis [J]. osteoporosis international: a journal established as result of Cooperation between the European foundation for osteoporosis and the National osteoporosis foundation of the USA. 2017;28(6):1835–52. [DOI] [PubMed]
- 32.Mu YP, Wei XX, Sajidan K et al. [A systematic review and meta-analysis of the relationship between obstructive sleep apnea hypopnea syndrome and bone metabolism]. Chin Gen Pract. 2022;(30):3825–33.
- 33.RUCHAŁA M, BROMIŃSKA B, CYRAŃSKA-CHYREK E, et al. Obstructive sleep apnea and hormones - a novel insight [J]. Archives Med Science: AMS. 2017;13(4):875–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.DOMAZETOVIC V, MARCUCCI G, IANTOMASI T et al. Oxidative stress in bone remodeling: role of antioxidants [J]. Clinical cases in mineral and bone metabolism: the official journal of the Italian Society of Osteoporosis, Mineral Metabolism, and Skeletal Diseases. 2017;14(2):209 – 16. [DOI] [PMC free article] [PubMed]
- 35.MAO SS, LI D, SYED YS, et al. Thoracic quantitative computed tomography (QCT) can sensitively monitor bone mineral metabolism: comparison of thoracic QCT vs lumbar QCT and Dual-energy X-ray absorptiometry in detection of Age-relative change in bone mineral density [J]. Acad Radiol. 2017;24(12):1582–7. [DOI] [PubMed] [Google Scholar]
- 36.Gao B. Influence of the selection method of interest area on bone mineral density measurement by QCT and gem energy spectrum CT [PhD Dissertation]. Dalian Medical University; 2021.
- 37.Li K, Chen J, Zhao LF, et al. Establishment of normal reference value of spinal bone mineral density by quantitative CT(QCT) in Chinese population and validation of QCT diagnostic criteria for osteoporosis. Chinese J Osteoporos. 2019;25(09):1257– 62+72.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.