A Prediction Nomogram for Severe Obstructive Sleep Apnea in Snoring Patients: A Retrospective Study

Figures & data

Table 1 Baseline Clinical Features of the Study Participants [M (P25, P75)/N (%)]

Variable	Total (N=674)	Training Cohort (N=464)	Validation Cohort (N=210)	P
Gender				0.006
Male	546 (81.0%)	363 (78.2%)	183 (87.1%)
Female	128 (19.0%)	101 (21.8%)	27 (12.9%)
Age (years)	47.0 (35.0, 58.0)	50.0 (38.8, 59.3)	40.0 (31.3, 48.8)	0.000
BMI (kg/m^2)	27.0 (24.4, 29.4)	27.3 (24.4, 29.7)	26.4 (24.2, 28.4)	0.011
ESS total score	5.0 (3.0, 8.0)	5.0 (3.0, 9.0)	4.5 (1.0, 6.0)	0.000
Smoking history				0.884
Yes	282 (41.8%)	195 (42.0%)	87 (41.4%)
No	392 (58.2%)	269 (58.0%)	123 (58.6%)
Drinking history				0.931
Yes	313 (46.4%)	216 (46.6%)	97 (46.2%)
No	361 (53.6%)	248 (53.4%)	113 (53.8%)
Dizzy				0.145
Yes	199 (29.5%)	145 (31.3%)	54 (25.7%)
No	475 (70.5%)	319 (68.7%)	156 (74.3%)
Chest tightness				0.146
Yes	162 (24.0%)	119 (25.6%)	43 (20.5%)
No	512 (76.0%)	345 (74.4%)	167 (79.5%)
Morning dry mouth				0.000
Yes	258 (38.3%)	135 (29.1%)	123 (58.6%)
No	416 (61.7%)	329 (70.9%)	87 (41.4%)
Dream recall				0.000
Yes	157 (23.3%)	86 (18.5%)	71 (33.8%)
No	517 (76.7%)	378 (81.5%)	139 (66.2%)
Hypertension				0.497
Yes	289 (42.9%)	203 (43.8%)	86 (41.0%)
No	385 (57.1%)	261 (56.2%)	124 (59.0%)
Diabetes				0.000
Yes	105 (15.6%)	90 (19.4%)	15 (7.1%)
No	569 (84.4%)	374 (80.6%)	195 (92.9%)
GERD				0.113
Yes	44 (6.5%)	35 (7.5%)	9 (4.3%)
No	630 (93.5%)	429 (92.5%)	201 (95.7%)
AHI	25.5 (9.2, 47.8)	30.3 (11.7, 49.9)	19.2 (4.8, 39.2)	0.000

Notes: Numerical information is presented as mean ± standard deviation or median (interquartile range); categorical data are presented as numbers (percentages). N is the number of participants. The training cohort consisted of 464 participants with 60 simple snoring, 75 with mild OSA, 95 with moderate OSA, and 234 with severe OSA; the validation cohort consisted of 210 participants with 55 simple snoring, 33 with mild OSA, 45 with moderate OSA, and 77 with severe OSA.

Abbreviations: BMI, body mass index; ESS, Epworth sleepiness scale; GERD, Gastroesophageal Reflux Disease; AHI, apnea-hypopnea index.

Figure 1 Process diagram for patient selection.

Abbreviations: OSA, obstructive sleep apnea; ROC, receiver operating characteristic curve; DCA, decision curve analysis.

Table 2 General Characteristics of Patients in the Training Cohort

Variable	Total (N=464)	Non-Severe OSA (N=230)	Severe OSA (N=234)	P
Gender				0.007
Male	363 (78.2%)	168 (73.0%)	195 (83.3%)
Female	101 (21.8%)	62 (27.0%)	39V(16.7%)
Age (years)	50.0 (38.8, 59.3)	51.0 (40.0, 60.3)	49.0 (37.8, 59.0)	0.392
BMI (kg/m^2)	27.3 (24.4, 29.7)	26.3 (23.9, 28.7)	28.3 (25.6, 31.1)	0.000
ESS total score	5.0 (3.0, 9.0)	3.0 (2.0, 6.0)	7.0 (4.0, 11.0)	0.000
Smoking history				0.000
Yes	195 (42.0%)	71 (30.9%)	124 (53.0%)
No	269 (58.0%)	159 (69.1%)	110 (47.0%)
Drinking history				0.001
Yes	216 (46.6%)	89 (38.7%)	127 (54.3%)
No	248 (53.4%)	141 (61.3%)	107 (45.7%)
Dizzy				0.670
Yes	145 (31.3%)	74 (32.2%)	71(30.3%)
No	319 (68.7%)	156 (67.8%)	163 (69.7%)
Chest tightness				0.394
Yes	119 (25.6%)	63 (27.4%)	56 (23.9%)
No	345 (74.4%)	167 (72.6%)	178 (76.1%)
Morning dry mouth				0.026
Yes	135 (29.1%)	56 (24.3%)	79 (33.8%)
No	329 (70.9%)	174 (75.7%)	155 (66.2%)
Dream recall				0.039
Yes	86 (18.5%)	34 (14.8%)	52 (22.2%)
No	378 (81.5%)	196 (85.2%)	182 (77.8%)
Hypertension				0.000
Yes	203 (43.8%)	76 (33.0%)	127 (54.3%)
No	261 (56.2%)	154 (67.0%)	107 (45.7%)
Diabetes				0.074
Yes	90 (19.4%)	37 (16.1%)	53 (22.6%)
No	374 (80.6%)	193 (83.9%)	181 (77.4%)
GERD				0.239
Yes	35 (7.5%)	14 (6.1%)	21 (9.0%)
No	429 (92.5%)	216 (93.9%)	213 (91.0%)
AHI	30.3 (11.7, 49.9)	11.6 (4.8, 18.6)	49.7 (36.7, 65.0)	0.000

Notes: Numerical information is presented as mean ± standard deviation or median (interquartile range); categorical data are presented as numbers (percentages). N is the number of participants.

Abbreviations: BMI, body mass index; ESS, Epworth sleepiness scale; GERD, Gastroesophageal Reflux Disease; AHI, apnea-hypopnea index.

Table 3 Covariance Diagnosis and Binary Logistic Regression Analysis of Screening Predictors

Variable	Collinearity Statistics		Multivariate Analysis
	Tolerance	VIF	B	SE	Wald	OR	95% CI	P
Gender	0.960	1.042	−0.176	0.301	0.340	0.839	0.465–1.514	0.560
BMI (kg/m^2)	0.980	1.021	0.146	0.029	24.639	1.158	1.093–1.226	0.000
ESS total score	0.835	1.197	0.237	0.032	56.431	1.268	1.192–1.349	0.000
Smoking history	0.731	1.369	0.801	0.228	12.366	2.228	1.426–3.482	0.000
Drinking history	0.777	1.287	0.364	0.250	2.119	1.439	0.882–2.349	0.145
Morning dry mouth	0.984	1.016	0.491	0.249	3.867	1.633	1.002–2.663	0.049
Dream recall	0.992	1.008	0.680	0.292	5.431	1.973	1.114–3.495	0.020
Hypertension	0.962	1.039	0.873	0.228	14.711	2.394	1.533–3.740	0.000

Abbreviations: BMI, body mass index; ESS, Epworth sleepiness scale; VIF, variance inflation factor; SE, standard error; OR, odds ratio; CI, confidence interval.

Figure 2 Nomogram for the prediction of the occurrence of severe OSA in snoring patients.

Notes: To illustrate the use of the nomogram, here we provide an example of a patient with the following clinical characteristics: BMI=36 kg/m², total ESS score=6, and a history of smoking, hypertension, dry mouth in the morning, and dream recall. Applying the above values to this nomogram, where the BMI reading is approximately 56.5 points, the ESS total score reading is approximately 25 points, and the presence of smoking, hypertension, dry mouth in the morning, and dream recall reads 14.5, 15.5, 8.5, and 12 points, respectively. Therefore, this patient had a total score reading of approximately 132 and a corresponding risk of severe OSA of >0.95 (95%).

Figure 3 ROC curves. (A) Training cohort. (B) Validation cohort. ROC=receiver operating characteristic; AUC=area under the ROC curve.

Figure 4 Calibration curves predicting the probability of severe OSA in snoring patients. (A) Training cohort. (B) Validation cohort.

Figure 5 Decision curve analysis for predicting the occurrence of severe OSA in snoring patients. (A) Training cohort. (B) Validation cohort.

Data Sharing Statement

The data that support the findings of this study are available from the corresponding authors upon reasonable request.