Characteristics of participators
In our research, all 173 OSA patients were selected and analyzed. OSA patients were divided into three groups in accord with the tertiles of serum S100A9. Demographic information and clinical characteristics were estimated in OSA patients with different subgroups. As shown in Table 1, the number of female patients was reduced with increased serum S100A9 in OSA patients. There was no difference of age, body mass index (BMI), neck circumference, nose score, snoring duration, systolic pressure, diastolic pressure, the counts of hypertension and diabetes mellitus in OSA patients with different subgroups. In addition, the levels of apnoea-hypopnoea index (AHI) score, STOP-BANG (snoring, tiredness, observed apnea, high blood pressure, BMI, age, neck circumference, and male gender) score, Epworth Sleepiness Scale (ESS) score, and average heart rate were gradually elevated in parallel with the elevated serum S100A9 concentration in OSA patients (Table 1). However, no obvious difference of apnea duration, hypopnea duration, minimum nocturnal oxygen saturation (Min SpO2), average SpO2, the indices of blood routine, liver function and renal function was observed in OSA patients with different serum S100A9 levels. Nevertheless, the contents of procalcitonin and interleukin-6 were risen with increased serum S100A9 in OSA patients (Table 1).
Table 1
Demographic characteristics of participators at baseline.
Characteristic | All participators | Tertile of serum S100A9 | P |
Tertile 1 (<11.34 pg/mL) | Tertile 2 (11.34 ~ 20.70 pg/mL) | Tertile 3 (>20.70 pg/mL) |
N | 173 | 58 | 58 | 57 | |
Age, years | 31.8 ± 0.53 | 31.7 ± 0.92 | 32.3 ± 0.93 | 31.6 ± 0.98 | 0.962 |
Female, n (%) | 124 (69.7) | 42 (72.4) | 46 (79.3) | 32 (56.1) | 0.022 |
BMI, kg/m2 | 40.0 ± 0.61 | 39.2 ± 0.81 | 39.7 ± 0.88 | 41.4 ± 1.45 | 0.180 |
Neck circumference, cm | 42.7 ± 0.39 | 42.2 ± 0.76 | 42.1 ± 0.52 | 43.9 ± 0.77 | 0.117 |
Nose score | 3.1 ± 0.29 | 3.1 ± 0.52 | 2.6 ± 0.51 | 3.76 ± 0.52 | 0.392 |
Snoring duration, years | 5.0 ± 0.35 | 4.6 ± 0.68 | 4.92 ± 0.51 | 5.67 ± 0.70 | 0.269 |
Systolic pressure (mmHg) | 134.3 ± 1.40 | 134.8 ± 2.54 | 128.2 ± 1.49 | 141.4 ± 2.87 | 0.089 |
Diastolic pressure (mmHg) | 81.7 ± 1.12 | 81.4 ± 1.52 | 78.0 ± 1.72 | 86.7 ± 2.50 | 0.066 |
Hypertension, n (%) | 27 (15.6) | 7 (12.1) | 9 (15.5) | 11 (19.3) | 0.565 |
Diabetes mellitus, n (%) | 5 (2.9) | 2 (3.4) | 1 (1.7) | 2 (3.5) | 0.809 |
AHI, /h | 16.5 ± 1.60 | 10.2 ± 1.98 | 14.1 ± 2.23 | 25.5 ± 3.64 | <0.001 |
STOP-BANG score | 3.4 ± 0.18 | 3.0 ± 0.29 | 3.4 ± 0.29 | 4.0 ± 0.43 | 0.049 |
ESS score | 6.2 ± 0.34 | 5.2 ± 0.43 | 6.3 ± 0.58 | 7.1 ± 0.71 | 0.022 |
Apnea duration, s | 17.4 ± 1.14 | 16.9 ± 1.94 | 15.6 ± 2.01 | 19.5 ± 2.03 | 0.358 |
Hypopnea duration, s | 21.3 ± 1.20 | 19.3 ± 1.30 | 22.6 ± 3.08 | 21.1 ± 1.16 | 0.310 |
Min SpO2, % | 74.0 ± 1.20 | 72.8 ± 2.76 | 76.0 ± 1.64 | 73.0 ± 1.79 | 0.947 |
Average SpO2, % | 96.2 ± 4.89 | 89.3 ± 2.78 | 107.6 ± 1.47 | 91.8 ± 0.90 | 0.397 |
Average heart rate, /min | 73.7 ± 0.71 | 71.5 ± 1.12 | 73.3 ± 1.33 | 76.5 ± 1.24 | 0.004 |
WBC (109/L) | 9.2 ± 0.23 | 9.2 ± 0.40 | 9.2 ± 0.37 | 9.5 ± 0.43 | 0.650 |
Neutrophil (109/L) | 5.9 ± 0.22 | 6.1 ± 0.39 | 5.8 ± 0.31 | 6.0 ± 0.48 | 0.837 |
Lymphocyte (109/L) | 2.5 ± 0.08 | 2.4 ± 0.12 | 2.5 ± 0.12 | 2.7 ± 0.14 | 0.087 |
ALT (U/L) | 51.7 ± 3.94 | 53.8 ± 8.63 | 42.4 ± 5.17 | 57.1 ± 6.72 | 0.760 |
AST (U/L) | 35.6 ± 2.34 | 38.5 ± 5.14 | 29.8 ± 2.73 | 37.7 ± 4.09 | 0.904 |
Urea nitrogen (mmol/L) | 4.7 (4.0, 5.6) | 4.7 (3.9, 5.5) | 4.6 (3.8, 5.6) | 4.9 (4.2, 6.1) | 0.466 |
Creatinine (µmol/L) | 49.0 (41.0, 57.3) | 48.0 (40.0, 59.3) | 49.5 (41.0, 55.3) | 53.0 (43.5, 60.0) | 0.428 |
Uric acid (µmol/L) | 417.5 ± 9.64 | 408.6 ± 15.59 | 406.1 ± 15.32 | 443.2 ± 19.91 | 0.181 |
LDH (U/L) | 1.8 (1.5, 2.1) | 1.8 (1.6, 2.0) | 1.8 (1.5, 2.1) | 1.9 (1.7, 2.3) | 0.085 |
Procalcitonin (ng/L) | 0.03 (0.01, 0.05) | 0.03 (0.01, 0.04) | 0.03 (0.01, 0.04) | 0.04 (0.01, 0.06) | 0.035 |
D-dimer (mg/L) | 0.3 ± 0.02 | 0.3 ± 0.02 | 0.3 ± 0.03 | 0.3 ± 0.03 | 0.337 |
C-reactive protein (mg/L) | 8.6 ± 0.06 | 7.6 ± 0.97 | 8.6 ± 1.15 | 9.3 ± 1.23 | 0.077 |
Interleukin-6 (pg/mL) | 3.2 ± 0.26 | 2.5 ± 0.34 | 2.9 ± 0.38 | 3.9 ± 0.49 | 0.030 |
Data in bold denote statistically significant results. |
Levels of serum S100A9 in OSA patients
The levels of serum S100A8 were detected and compared in OSA patients with different severity scores. As shown in Fig. 1A, the level of serum S100A9 was higher in OSA patients with >30 AHI scores than those with other AHI scores. According to ESS score, the level of serum S100A9 was significantly elevated in OSA patients with ≥ 11 scores than those with <11 scores (Fig. 1B). Based on STOP-BANG score, serum S100A9 was evidently increased in OSA patients with ≥ 3 scores than those with <3 scores (Fig. 1C). Furthermore, the level of serum S100A9 was substantially increased in OSA patients with higher apnea duration than those in lower apnea duration (Fig. 1H). Additionally, no difference of serum S100A9 was found in OSA patient with different average heart rates, Min SpO2, average SpO2, and hypopnea durations (Fig. 1D-1G).
Associations between serum S100A9 and clinical characteristics in OSA patients
The associations between serum S100A9 and clinical characteristics were evaluated by Spearman correlation coefficient or Pearson rank correlation. As shown in Figure 2, the level of serum S100A9 was positively associated with BMI (r=0.21, P<0.01), neck circumference (r=0.30, P<0.001), snoring duration (r=0.23, P<0.001), systolic pressure (r=0.28, P<0.001), diastolic pressure (r=0.24, P<0.01), alanine aminotransferase (ALT) (r=0.34, P<0.0001), aspartate aminotransferase (AST) (r=0.35, P<0.0001), lactate dehydrogenase (LDH) (r=0.32, P<0.001). In addition, serum S100A9 was inversely related with the count of lymphocyte (r=-0.18, P<0.05) in OSA patients (Figure 2).
Association between serum S100A9 and the severity in OSA patients
The associations between serum S100A9 and OSA severity scores were estimated via linear regression models and logistic regression models. In order to control confounding factor, gender and BMI were adjusted. Multivariable linear regression analyses found that each 1 pg/mL elevation of serum S100A9 was associated with 0.395 score (95%CI: 0.234 ~ 0.557), 0.021 score (95%CI: 0.002 ~ 0.041), 0.039 score (95%CI: 0.002 ~ 0.076), and 0.832 second (95%CI: 0.514 ~ 0.914) changes for AHI score, STOP-BANG score and ESS score, and apnea duration, respectively (Table 2). Moreover, multivariable logistical regression analyses revealed that the tertiles of serum S100A9 were positively linked with the odd ratios (ORs) of AHI score (Tertile 2, OR = 2.680, 95%CI: 1.067 ~ 6.735; Tertile 3, OR = 6.135, 95%CI: 2.471 ~ 15.232) and STOP-BANG score (Tertile 3, OR = 2.508, 95%CI: 1.735 ~ 8.554) in OSA patients (Table 2).
Table 2
Associations between serum S100A9 and OSA severity.
Variables | Estimated changes continues serum S100A9 | Estimated changes (95% CI) by tertiles of serum S100A9 | P trend |
Tertile 1 | Tertile 2 | Tertile 3 |
N | 173 | 58 | 58 | 57 | |
AHI, /h | 0.395 (0.234, 0.557) | 0 (Ref) | 2.680 (1.067, 6.735) | 6.135 (2.471, 15.232) | 0.010 |
STOP-BANG score | 0.021 (0.002, 0.041) | 0 (Ref) | 1.307 (0.416, 4.099) | 2.508 (1.735, 8.554) | 0.036 |
ESS score | 0.039 (0.002, 0.076) | 0 (Ref) | 1.352 (0.585, 3.127) | 1.199 (0.524, 3.742) | 0.101 |
Apnea duration, s | 0.832 (0.514, 0.921) | 0 (Ref) | 0.696 (0.297, 1.628) | 1.819 (0.794, 4.164) | 0.055 |
Hypopnea duration, s | 0.020 (-0.109, 0.148) | 0 (Ref) | 0.779 (0.346, 1.758) | 0.952 (0.427, 2.123) | 0.920 |
Min SpO2, % | 0.033 (-0.107, 0.172) | 0 (Ref) | 0.955 (0.427, 2.135) | 0.712 (0.325, 1.556) | 0.145 |
Average SpO2, % | -0.058 (-0.640, 0.524) | 0 (Ref) | 0.859 (0.374, 1.972) | 0.676 (0.299, 1.526) | 0.245 |
Average heart rate, /min | 0.085 (0.003, 0.167) | 0 (Ref) | 1.587 (0.695, 3.621) | 3.023 (1.327, 6.888) | 0.059 |
Data in bold denote statistically significant results. |
Gender and BMI were adjusted. |
Predictive power for severity in OSA patients
OSA patient were divided into two subgroups, included mild or moderate patients (AHI<30) and severe patients (AHI ≥ 30). The predictive powers for severe patients of serum S100A9 and severity scores were estimated by ROC curve and area under the curve (AUC) in OSA patients. As shown in Fig. 3, the predictive power (AUC) for severe patients of serum S100A9 (0.769) was lower than those in STOP-BANG score (0.817), but higher compared with ESS (0.687), apnea duration (0.705), average heart rate (0.643), hypopnea duration (0.505), average SpO2 (0.139), and Min SpO2 (0.114) in OSA patients. The cut-off concentration of serum S100A9 was 32.12 pg/mL. The sensitivity was 92.2% and the specificity was 76.3% in OSA patients. In addition, serum S100A9 combination with severity scores further elevated the predictive capacities for severe patients compared with single serum S100A9 or STOP-BANG score in OSA patients (Fig. 3).