Baseline characteristics of the subjects
Of all participants, 77.6% (2585/3329) were diagnosed with OSA, the proportion of severity of mild, moderate and severe OSA was 37.9%(981/2585), 32.1%(829/2585) and 30.0% (775/2585), respectively. The overall mean age was 48.6 years, and 80.9% of patients were aged 30–60 years. The mean age, body mass index (BMI), serum estimated glomerular filtration rate (eGFR), low-density lipoprotein cholesterol (LDL)-c, and AHI of patients with OSA were significantly higher than in those without OSA, while the mean oxygen saturation (SaO2) and nadir SaO2 in the OSA group were much lower than that of patients without OSA. The constituent ratio of males, type 2 diabetes, coronary heart disease, and smokers in the OSA group was significantly higher than that in patients without OSA. Additionally, the proportion of ≥ 3 antihypertensive drugs, lipid-modifying agents, and antiplatelet drugs was greater in the OSA group. During the follow-up (median: 7.0 years), the crude incidence of overall outcomes was 18.8 per 1000 person-years in the total population, 20.6 per 1000 person-years in the OSA group, and 13.2 per 1000 person-years in patients without OSA (Table 1).
Table 1
Baseline characteristics of the subjects
|
Total subjects
(n = 3329)
|
OSA
(n = 2585)
|
without OSA
(n = 744)
|
P value
|
Age(yr)
|
48.6 ± 11.0
|
49.7 ± 10.8
|
44.8 ± 10.8
|
< 0.001
|
Gender(Male,%)
|
2184(65.6)
|
1775(68.7)
|
409(55.0)
|
< 0.001
|
BMI(kg/m2)
|
27.6(25.4–30.1)
|
28.1(25.9–30.6)
|
26.2(24.2–28.7)
|
< 0.001
|
Baseline office SBP(mmHg)
|
139.6 ± 19.5
|
140.0 ± 19.8
|
138.4 ± 18.5
|
0.046
|
Baseline office DBP(mmHg)
|
91.7 ± 13.8
|
91.7 ± 14.1
|
91.8 ± 12.9
|
0.820
|
Baseline MAP(mmHg)
|
107.7 ± 14.5
|
107.8 ± 14.7
|
107.3 ± 13.6
|
0.452
|
Smoker(n,%)
|
1099(33.0)
|
906(35.0)
|
193(25.9)
|
< 0.001
|
GGT (mmol/L)
|
28.4 ± 22.3
|
28.6 ± 20.4
|
27.5 ± 27.9
|
0.217
|
GOT(mmol/L)
|
22.2 ± 14.2
|
22.4 ± 14.5
|
21.6 ± 12.8
|
0.171
|
eGFR(ml/min/1.73m2)
|
96.8 ± 21.9
|
100.2 ± 21.8
|
95.9 ± 21.9
|
< 0.001
|
LDL-C(mmol/L)
|
2.61 ± 0.79
|
2.63 ± 0.80
|
2.54 ± 0.78
|
0.006
|
Total sleep time(mins)
|
390.2 ± 51.7
|
389.8 ± 50.8
|
391.4 ± 54.7
|
0.510
|
Sleep efficiency(%)
|
73.5 ± 9.6
|
73.6 ± 9.6
|
73.0 ± 9.5
|
0.169
|
AHI(event/h)
|
14.4(5.6–26.9)
|
19.7(10.7–32.2)
|
1.7(0.7–3.1)
|
< 0.001
|
Mean SaO2(%)
|
93.0(91.0–94.0)
|
92.0(91.0–94.0)
|
94.0(93.0–95.0)
|
< 0.001
|
Nadir SaO2(%)
|
82.0(77.0–86.0)
|
80.0(75.0–84.0)
|
88.0(87.0–90.0)
|
< 0.001
|
Diabetes(%)
|
550(16.5)
|
478(18.5)
|
72(9.7)
|
< 0.001
|
CHD(%)
|
373(11.2)
|
317(12.3)
|
56(7.5)
|
< 0.001
|
Antihypertensive regimen(n,%)
|
|
|
|
|
0–1 drug
|
1183(35.5)
|
845(32.7)
|
338(45.4)
|
< 0.001
|
2 drugs combination
|
1660(49.9)
|
1327(51.3)
|
333(44.8)
|
|
≥3 drugs combination
|
486(14.6)
|
413(15.8)
|
73(9.8)
|
|
Lipid-modifying agents (n,%)
|
2070(62.2)
|
1693(65.5)
|
377(50.7)
|
< 0.001
|
Antiplatelet drugs (n,%)
|
1639(49.2)
|
1348(52.1)
|
291(39.1)
|
< 0.001
|
Antidiabetic drugs use in patients with DM2(n,%)
|
442(80.4)
|
383(80.1)
|
59(81.9)
|
0.717
|
Regular CPAP treatment(%)
|
0
|
114(4.4%)
|
0
|
|
Follow-up, median(IQR),y
|
7.0(6.0–8.1)
|
6.9(6.0–8.0)
|
7.3(6.3–8.1)
|
|
Person-years followed,y
|
22016.15
|
16932.46
|
5083.69
|
-
|
Total primary endpoints(n)
|
415
|
348
|
67
|
-
|
Outcome per 1000 person-years
|
18.8
|
20.6
|
13.2
|
-
|
Abbreviations: BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; GGT, alanine transaminase; GOT, aspartate transaminase; eGFR, glomerular filtration rate; LDL-C, low-density lipoprotein cholesterol; AHI, apnea-hypopnea index; SaO2, saturation; CHD, coronary heart diseases.
MACCE incidence
In total, 415 individuals experienced extended MACCE at the follow-up (median: 7.0 years). The incidence of extended MACCE was significantly higher in the OSA group than in the non-OSA group (hazard ratio [HR]: 1.59; 95% confidence interval [CI]: 1.27–1.99; P < 0.001, Fig. 1A), and the same applied to cardiac events (HR [95% CI]: 2.44 [1.80–3.29], P < 0.001, Fig. 1D). However, the incidence of all-cause death and stroke between the OSA and non-OSA groups was not significant.
OSA and extended MACCE and cardiac events
Table 2 presents the association between OSA and extended MACCE, as well as cardiac events, in the total population. In the crude model, OSA was shown as a risk factor for extended MACCE and cardiac events. Adjusted for confounders, no significant association between OSA and extended MACCE was observed. Still, OSA patients had a 57% increased risk of cardiac events compared with subjects without OSA (HR [95% CI]: 1.57 [1.04–2.39], P = 0.034), and the association did not change in a further sensitivity analysis. However, no association between OSA and stroke and all-cause death in the total population was observed (Table S1).
Table 2
Baseline characteristics of the subjects
|
Extended MACCE
|
Cardiac events
|
Model
|
HR(95%CI)
|
P value
|
|
HR(95%CI)
|
P value
|
Crude model
|
1.59(1.23–2.07)
|
0.001
|
|
2.44(1.63–3.68)
|
< 0.001
|
Partially adjusted model
|
1.24(0.95–1.61)
|
0.117
|
|
1.83(1.21–2.76)
|
0.004
|
Fully adjusted model
|
1.11(0.84–1.45)
|
0.461
|
|
1.57(1.04–2.39)
|
0.034
|
Sensitivity analysis
|
1.07(0.81–1.40)
|
0.645
|
|
1.53(1.01–2.33)
|
0.046
|
Abbreviations: HR, hazard ratio.
Notes:
Partially adjusted model: adjusted for age and sex.
Fully adjusted model: adjusted for age, sex, body mass index, baseline systolic blood pressure, diastolic blood pressure, low-density lipoprotein cholesterol, eGFR, smoking, type 2 diabetes, history of coronary heart diseases, lipid-lowering drugs, antidiabetic drugs, and antiplatelet drugs.
Sensitivity analysis was performed in 3201 hypertensive patients without OSA-specific treatment, ie, CPAP, oral appliance, surgery, etc. The confounders in the fully adjusted model were included.
OSA and BP status
The interaction between OSA status and BP level (≥ 140/90 mmHg or < 140/90 mmHg) was not significant for extended MACCE (HR [95% CI]: 1.12 [0.91–1.36], P = 0.287) or cardiac events (HR [95% CI]: 1.15 [0.88–1.50], P = 0.321) (Table S2), indicating that there was independent effect of OSA and BP on extended MACCE and cardiac events. In the uncontrolled hypertension population, OSA had a 1.93-fold increased risk of cardiac events compared to patients without OSA (P = 0.036) after adjustment for confounders. Further sensitivity analysis also showed OSA as a significant risk factor for cardiac events regardless of OSA-specific treatment. In contrast, in the controlled hypertension, OSA showed a 1.23-fold higher risk of cardiac events in the OSA group than in those without OSA, and as adjusted for confounding factors, the P value did not reach statistical significance (P = 0.489). However, no significant association between OSA and extended MACCE, stroke, and all-cause death was observed neither in the controlled hypertension nor in the uncontrolled hypertension group (Table 3, Table S3).
Table 3
Baseline characteristics of the subjects
|
BP controlled < 140/90 mmHg
|
|
BP controlled ≥ 140/90 mmHg
|
Model
|
adjusted HR(95%CI)
|
P value
|
|
adjusted HR(95%CI)
|
P value
|
Extended MACCE
|
|
|
|
|
|
Crude model
|
1.60(1.06–2.42)
|
0.027
|
|
1.56(1.08–2.25)
|
0.017
|
Partially adjusted model
|
1.21(0.79–1.84)
|
0.377
|
|
1.28(0.88–1.86)
|
0.200
|
Fully adjusted model
|
1.01(0.66–1.56)
|
0.957
|
|
1.22(0.83–1.78)
|
0.310
|
Sensitivity analysis
|
0.96(0.62–1.48)
|
0.853
|
|
1.17(0.80–1.72)
|
0.411
|
Cardiac events
|
|
|
|
|
|
Crude model
|
2.10(1.19–3.69)
|
0.010
|
|
2.50(1.38–4.54)
|
0.003
|
Partially adjusted model
|
1.52(0.86–2.67)
|
0.151
|
|
1.96(1.07–3.59)
|
0.029
|
Fully adjusted model
|
1.23(0.69–2.19)
|
0.489
|
|
1.93(1.04–3.57)
|
0.036
|
Sensitivity analysis
|
1.18(0.66–2.11)
|
0.580
|
|
1.89(1.02–3.49)
|
0.044
|
Abbreviations: HR, hazard ratio.
Notes:
Partially adjusted model: adjusted for age and sex.
Fully adjusted model: adjusted for age, sex, body mass index, baseline systolic blood pressure, diastolic blood pressure, low-density lipoprotein cholesterol, eGFR, smoking, type 2 diabetes, history of coronary heart diseases, lipid-lowering drugs, antidiabetic drugs, and antiplatelet drugs.
Sensitivity analysis was performed in 3139 hypertensive patients without OSA-specific treatment, ie, CPAP, oral appliance, and surgery. The confounders in the fully adjusted model were included.