Evaluation of Left Ventricular Function in Obese Patients with Obstructive Sleep Apnea by Three-dimensional Speckle Tracking Echocardiography

doi:10.21203/rs.3.rs-1331917/v1

Purpose

Both obstructive sleep apnea (OSA)and obesity can cause myocardial remodeling and cardiac insufficiency via corresponding pathophysiological pathways. Therefore, it is speculated that the superposition of OSA and obesity may cause more severe impairment of cardiac function. The objective of our study was to evaluate the early changes of left ventricular systolic function in obese patients with OSA with three-dimensional speckle tracking echocardiography(3D-STE).

Methods

This study was conducted with33 obese OSA, 46 non-obese OSA, and 20 healthy subjects. Demographic, biochemical, and Polysomnography(PSG) data were collected, and their relation with the left ventricular strain was measured and analyzed with 3D-STE.

Results

The Left ventricular strain was significantly worse in the OSA group compared to the control group(P < 0.05). The global longitudinal strain(GLS) was significantly worse in the OSA obese group compared to non-obese OSA group (P < 0.05). The GLS value positively correlated with body mass index(BMI) (r = 0.406, P < 0.001),Apnea-hypopnea index(AHI)(r = 0.610, P < 0.001)and homeostasis model assessment of insulin resistance(HOME-IR) (r = 0.431, P < 0.001) in patients with OSA. Multiple linear regression analysis showed BMI as a predictor of GLS and global circumferential strain(GCS), AHI as a predictor of GLS, and HOME-IR as a predictor of global area strain(GAS) and global radial strain(GRS).

Conclusion

In OSA patients, the myocardial strain was impaired before the damages in left ventricular ejection fraction, suggesting that the left ventricular systolic function is damaged early. The coexistence of obesity and OSA can lead to severe impairment of cardiac function through hypoxia and insulin resistance.

Obesity

Obstructive sleep apnea

Left ventricular dysfunction

three-dimensional speckle tracking echocardiography

Obstructive sleep apnea (OSA) is a severe disorder in which an individual repeatedly starts and stops breathing during their sleep. It results from the complete or partial obstruction of the upper airway during sleep that causes intermittent hypoxia and carbon dioxide retention^{[1, 2]}. Obesity is one of the important pathogenic factors of OSA. Studies have shown that both OSA and obesity can cause myocardial remodeling and early impairment of cardiac function through sympathetic nervous system activation, insulin resistance, systemic inflammation, and oxidative stress^{[3, 4]}. When obesity and OSA coexist, it is speculated that the superposition of the pathophysiological effects of the two may lead to more severe impairment of cardiac function. However, the subclinically impaired cardiac functionare easily overlooked, especially in OSA patients with obesity. As the association of OSA or obesity with adverse cardiac remodeling is high, an early assessment of the cardiac function would aid in reducing cardiovascular morbidity and mortality. As such, it is imperative that the physicians identify such patients with early signs of changes in cardiac function. But the assessment of cardiac function over the early stage of cardiac insufficiencyin obese OSA patients has not yet been reported.Initial impairments in cardiac function in obesity or OSA may be subtle, such as alterations in the regional contractile strain due to aberrant pressure in the myocardium, to the extent that they are challenging to detect on routine echocardiography.

Three-dimensional (3D) speckle-tracking echocardiography (3DSTE) is an advanced imaging technique that tracks the movement of the myocardium based on ultrasonic speckles in full-volume 3D images. It provides a comprehensive assessment of the myocardial deformation from the perspective of the whole and segment. This enables early and sensitive detection of systolic and diastolic myocardial capacity changes compared to conventional echocardiography. 3DSTE can give parameters on myocardial deformation such as longitudinal strain, area strain, torsional strain for evaluating cardiac function^{[5, 6]}. In OSA patients with no apparent symptoms of cardiac dysfunction, the early stage of impaired cardiac function largely remains undetected by conventional echocardiography.This study proposes to identify subclinical left ventricular dysfunction in obese OSA patients by using 3DSTE and provide a reference for early treatment.

Study population

Seventy-nine patients with OSA who underwent sleep monitoring at the Sleep Medicine Center of Peking University Third Hospital from July 2018 to December 2019 were selected for the study. OSA was determined following the diagnostic criteria of obstructive sleep apnea underlined in the International Classification of Sleep – Third edition (ICSD-3) 2014^[7]. Exclusion criteria included previously diagnosed patients with heart failure, coronary artery disease, valvular disease, cardiomyopathy, arrhythmia, chronic obstructive/restrictive pulmonary disease, thyroid dysfunction (including hypothyroidism or hyperthyroidism), already treated with CPAP, and poor image quality of echocardiography. Twenty healthy subjects who underwent sleep monitoring at about the same period to exclude OSA were selected as the control group. The Ethics Committee of Peking University Third Hospital (Ethical approval number M2017152) approved the study. The study was performed following the ethical standards of the Declaration of Helsinki and its later amendments. Informed consent forms were obtained from all the individuals enrolled in the study.

Demographic data

Demographic characteristics of the participants was measured. Patients with BMI>28 kg/m²(Chinese BMI cutoff for obesity)were classified as obese ^[8], and those below were defined as the non-obese grouped likewise. Since metabolic abnormalities are common to obesity and sleep apnea,which may be mediators of cardiac insufficiency, we measured metabolic indices at the same time.For example,the levels of total cholesterol (TC), triglyceride (TG), low-density lipoprotein cholesterol (LDL-c), high-density lipoprotein (HDL), high sensitivity C-reactive protein (hs-CRP), glycated hemoglobin (HbA1C), fasting glucose, and fasting insulin of all participants were estimated. Homeostasis model assessment of insulin resistance (HOMA-IR) was calculated using the formula [fasting glucose (mmol/l)×insulin (µU/ml)]/22.5^[8]. Ambulatory blood pressure monitoring (ABPM) was performed in all patients. The patients who had 24-h ABP≥130/80 mmHg, awake ABP≥135/85 mmHg, and/or sleep ABP≥120/70 mmHg with ambulatory blood pressure monitoring were considered hypertensive.

Polysomnography (PSG)

All subjects were monitored using the Polywin Polysomnography (PSG) system (Philips, USA). Caffeine, sedatives, hypnotic drugs, and alcohol were not allowed on the day of the examination. The sleep state was determined by three-channel electroencephalography, two-channel electrooculography, one-channel submental electromyography, and bilateral tibia electromyography. To detect OSA events, oral and nasal airflow, thoracic cage and abdominal respiratory motion, nasal air pressure, and arterial oxyhemoglobin saturation were used. Respiratory movements were measured with chest and abdominal belts, nasal pressure with a pressure sensor and oxygen saturation was determined using an oximeter probe. Sleep stages, respiratory events, thermistor, electrocardiography were scored by trained sleep technicians following the criteria laid down by the American Academy of Sleep Medicine (AASM) in 2012^[10]. A reduction in the airflow by at least 90% lasting≥10 sec was defined as an apnea event. Hypopnea was defined as cessation of airflow by at least 30% with at least 3% oxygen desaturation in the preceding 30s, and a reduction in chest wall movement and/or arousal. The average number of apnea plus hypopnea events/hour of sleep defined the apnea-hypopnea index (AHI). Individuals with an AHI≥5 were classified as OSA, whereas those with an AHI<5 were classified as control subjects. Apnea severity was classified as mild OSA(AHI 5-14), moderate OSA (AHI 15-29), and severe OSA (AHI≥30).

Three-dimensional speckle-tracking echocardiography(3D-STE)

We used GE Vivid E9 color Doppler echocardiography diagnostic instrument with M5Sc and 4V-D probe of 2-4 Hz. It was additionally equipped with an EchoPAC PC workstation and 4D Auto LVQ strain image analysis software to capture images. Standard echocardiographic parameters recorded were left atrial diameter (LAD), left atrial area (LAA), left ventricular end-diastolic diameter (LVEDD), E/A, and left ventricular ejection fraction (LVEF) to evaluate the left ventricular systolic function. The European Association of Echocardiography/ American Society of Echocardiography (EAE/ ASE) criteria was followed for all standard 2D and Doppler images and parameters. We also calculated LV strain parameters including three-dimensional echocardiographic global longitudinal strain (3DE-GLS), three-dimensional echocardiographic global circumferential strain (3DE-GCS), three-dimensional echocardiographic global radial strain (3DE-GRS), three-dimensional echocardiographic global area strain (3DE-GAS). The GLS, GAS, and GCS measurements had negative values, and GRS had positive values. The negative values of GLS, GAS, and GCS indicate the myocardium's shortening, thinning, or counterclockwise rotation. The positive value of GRS represents the lengthening, thickening, or clockwise rotation of the myocardium. Thus, lower negative strain values numerically represent the greater degree of cardiac deformation^[5].To ensure the stability and accuracy of the three-dimensional strain parameters, we took three readings of each parameter and calculated the average.

Statistical analysis

Statistical analysis was performed using SPSS software version 20.0. Normality test was performed by Kolmogorov-Smirnov test. Measurement data with normal distribution were expressed as mean±standard error. A t-test was performed for comparison between two groups, while ANOVA was used to compare multiple groups with LSD for post hoc analysis. The measurement data of skewed distribution were represented by M (P25, P75). The Mann-Whitney U test was used to compare two groups, and the Kruskal-Wallis test was used to compare multiple groups. Relative numbers expressed count data, and comparison was performed by χ2 test. Pearson’s Correlation analysis was performed for data with normal distribution and Spearman correlation analysis for skewed distribution. Multiple linear regression was used to analyze the predictive factors of three-dimensional strain. P<0.05 was considered statistically significant.

Clinical and demographic characteristics in different groups

This study was conducted with a total of 99 subjects. Of them, 79 subjects had mild to severe OSA, and 20 were healthy and without OSA. They were considered as the control group. The OSA subjects consisted of 12 cases of mild OSA, 29 cases of moderate OSA, and 38 cases of severe OSA. The OSA group was further divided into 33 obese patients and 46 non-obese patients. the BMI, T-G, fasting blood glucose, and HbA1Cin the OSA group was significantly higher than that of the control group, while the HDL was lower (P < 0.05). Within the OSA group, the hsCRP, fasting blood glucose, HbA1C, and insulin, and HOME-IR were higher in obese OSA patients compared to that of the non-obese OSA patients (P < 0.05). The PSG results showed that the total recorded time spent below 90% oxygen saturation (Ts90%) was significantly longer in the obese group compared to the non-obese OSA group. Demographic characteristics of patients between different subgroups were presented in Table 1–4.

Table 1

Clinical and demographic characteristics in control and OSA groups
Variables	Control group(n = 20)	OSA group(n = 79)	P-value
Sex(men: women)	6: 14	67:12	0.001
Age	37.5(33.0,42.0)	40.0(34.3,48.0)	0.110
BMI(kg/m² )	24.49(21.94,24.97)	27.17(25.35,29.66)	0.006^bc
TC(mmol/l)	4.64(4.34,4.75)	4.72(4.12,5.21)	0.153
TG(mmol/l)	1.03(0.81,1.07)	1.60(1.25,2.39)	0.002^bc
HDL(mmol/l)	1.34(1.10,1.54)	1.02(0.90,1.17)	0.001^ab
LDL-c(mmol/l)	2.77(2.50,3.02)	3.14(2.52,3.62)	0.057
Hs-CRP(mg/l)	0.87(0.52,2.13)	1.02(0.60,2.09)	0.815
Fasting glucose(mmol/l)	5.15(5.00,5.30)	5.40(5.10,6.00)	0.003
HbA1C(%)	5.50(5.27,5.60)	5.70(5.50,6.00)	0.010 ^c
Fasting insulin(µU/ml)	8.50(7.50,15.00)	12.15(8.48,16.48)	0.199
HOME-IR	1.89(1.73,4.54)	3.14(2.03,4.36)	0.112
24hour mean SBP(mmHg)	119.18 ± 12.75	120.89 ± 11.59	0.655
24hour mean DBP(mmHg)	78(70,88)	78(73,86)	0.886
Daytime mean SBP(mmHg)	125(111,130)	123(116,133)	0.984
Daytime mean DBP(mmHg)	80(74,91)	81(76,89)	0.902
Night-time mean SBP(mmHg)	109.27 ± 14.74	111.73 ± 12.16	0.547
Night-time mean DBP(mmHg)	67(62,80)	70(64,78)	0.709

Table 1

(continued). Clinical and demographic characteristics in control and OSA groups.
Variables	Control group(n = 20)	OSA group(n = 79)	P-value
AHI	3.00(1.10,4.60)	29.2(19.98,46.95)	< 0.001^bcef
LSaO2	92%(89%,93%)	80%(70%,80%)	< 0.001^abcdef
Ts90%	0.0%(0.0%,0.0%)	5.3%(1.0%,19.0%)	< 0.001^abcdef
LAD(mm)	30.41 ± 3.78	33.93 ± 3.89	0.006^b
LAA(mm²)	16(13,17)	17(15,19)	0.030
LVEDD(mm)	46.40(43.90,47.50)	47.45(44.85,50.22)	0.095
E/A	1.31(0.90,1.80)	1.17(0.92,1.44)	0.067
LAP(mmHg)	8.0(6.0,8.3)	8.0(6.0,9.0)	0.343
LVEF(%)	73.0(68.0,74.0)	69.5(66.0,73.0)	0.117
GLS	-16.24 ± 4.60	-12.78 ± 3.83	0.035^bcde
GCS	-15.19 ± 3.67	-13.14 ± 3.46	0.032^bcde
GAS	-27.15 ± 5.83	-22.57 ± 5.35	0.035^bcde
GRS	38.34(35.63,53.08)	32.83(26.75,41.75)	0.039^bcde
3D-LVEF(%)	57.86 ± 4.55	56.12 ± 4.58	0.139
Abbreviations:. BMI: body mass index; TC: total cholesterol; TG: triglyceride; HDL-C: high density lipoprotein cholesterol; LDL-c:low-density lipoprotein cholesterol;hs-CRP: high sensitivity C-reactive protein;HOMA-IR:homeostasis model assessment of insulin resistance;ABPM:ambulatory blood pressure monitoring; SBP:systolic blood pressure;DBP:diastolic blood pressure; AHI:apnea-hypopnea index;LSaO2:minimum oxygen saturation;Ts90%: the total recorded time spent below 90% oxygen saturation;LAD:left atrial diameter;LAD:left atrial area;LVEDD:left ventricular end diastolic diameter;LAP: left atrial pressure; LVEF: left ventricular ejection fraction; 3D-STE:3D speckle tracking echocardiography; GLS: global longitudinal strain; GCS: global circumferential strain; GAS: global area strain; GRS: global radial strain.a:p < 0.05 control versus mild OSA,b:p < 0.05 control versus moderate OSA, c:p < 0.05control versus severe OSA,d:p < 0.05 mild OSA versus moderate OSA,e:p < 0.05 mild OSA versus severe OSA,f:p < 0.05 moderate OSA versus severe OSA

Table 2

Clinical and demographic characteristics in obese and non-obese OSA groups
Variables	Non-obese(n = 46)	Obese(n = 33)	P-value
Sex(man: woman)	37:9	30:3	0.350
Age	40(34,47)	40(36,48)	0.624
BMI(kg/m²)	25.7(24.3,26.9)	30.0(28.7,31.1)	< 0.001
TC(mmol/l)	4.71 ± 0.83	4.73 ± 0.73	0.894
TG(mmol/l)	1.40(1.15,2.44)	1.69(1.44,2.34)	0.124
HDL(mmol/l)	1.04(0.96,1.22)	0.97(0.87,1.09)	0.038
LDL-c(mmol/l)	3.04 ± 0.77	3.17 ± 0.76	0.455
Hs-CRP(mg/l)	0.76(0.48,1.44)	1.69(0.83,3.72)	0.006
Fasting glucose(mmol/l)	5.30(5.00,5.70)	5.80(5.35,6.45)	0.006
HbA1C(%)	5.7(5.5,5.9)	5.9(5.6,6.3)	0.044
Fasting insulin(µU/ml)	10.70(8.20,14.10)	16.40(12.03,20.30)	< 0.001
HOME-IR	2.55(1.87,3.66)	4.10(2.97,5.60)	0.001
24hour mean SBP (mmHg)	116.0(111.5,121.8)	122.0(120.7,132.0)	0.002
24hour mean DBP(mmHg)	77.0(72.0,80.0)	79.3(76.5,86.5)	0.015
Daytime mean SBP(mmHg)	120.0(115.0,126.5)	126.0(123.4,137.5)	0.004
Daytime mean DBP(mmHg)	80.83 ± 7.97	85.29 ± 9.24	0.032
Night-time mean SBP(mmHg)	108.0(102.0,111.5)	114.0(111.0,121.5)	0.003

Table 2(continued) .Clinical and demographic characteristics in obese and non-obese OSA groups

Variables	Non-obese(n=46)	Obese(n=33)	P-value
Night-time mean DBP(mmHg)	69.0(63.5,72.5)	74.0(70.0,79.0)	0.004
AHI	29.30(16.95,38.90)	28.70(20.60,50.95)	0.327
LSaO2	81%(72%,88%)	79%(70%,86%)	0.416
Ts90%	3.0%(0.1%,11.5%)	5.0%(1.0%,28.5%)	0.032
LAD(mm)	32.8±3.7	35.4±3.7	0.004
LAA(mm²)	16(15,18)	18(16,20)	0.002
LVEDD(mm)	46.63±3.13	49.26±3.84	0.002
E/A	1.17(0.93,1.48)	1.16(0.89,1.28)	0.426
LAP(mmHg)	8(6,9)	9(6,10)	0.062
LVEF(%)	69.45±4.72	69.81±5.15	0.756
GLS	-13.86±3.53	-11.12±3.73	0.002
GCS	-13.01±3.35	-13.34±3.66	0.686
GAS	-23.16±5.18	-21.68±5.58	0.242
GRS	35.13±9.59	32.96±11.08	0.377
3D-LVEF	54.98±5.69	55.61±5.85	0.630
Abbreviations:. BMI: body mass index; TC: total cholesterol; TG: triglyceride; HDL-C: high density lipoprotein cholesterol; LDL-c:low-density lipoprotein cholesterol;hs-CRP: high sensitivity C-reactive protein;HOMA-IR:homeostasis model assessment of insulin resistance;ABPM:ambulatory blood pressure monitoring; SBP:systolic blood pressure;DBP:diastolic blood pressure; AHI: apnea- hypopnea index;LSaO2:minimum oxygen saturation;Ts90%: the total recorded time spent below 90% oxygen saturation;LAD:left atrial diameter;LAD:left atrial area;LVEDD:left ventricular end diastolic diameter;LAP: left atrial pressure;LVEF: left ventricular ejection fraction; 3D-STE:3D speckle tracking echocardiography; GLS: global longitudinal strain; GCS: global circumferential strain; GAS: global area strain; GRS: global radial strain.a:p<0.05 control versus mild OSA,b:p<0.05 control versus moderate OSA, c:p<0.05control versus severe OSA,d:p<0.05 mild OSA versus moderate OSA,e:p<0.05 mild OSA versus severe OSA,f:p<0.05 moderate OSA versus severe OSA

Table 3

Clinical and demographic characteristics in obese patients with different severity of OSA
Variables	Mild-moderateOSA (n = 17)	Severe OSA (n = 16)	P-value
Sex(man: woman)	15:2	15:1	0.582
Age	42.35 ± 11.90	41.88 ± 8.61	0.451
BMI(kg/m²)	30.37 ± 2.47	31.23 ± 3.48	0.419
T-C(mmol/l)	4.65 ± 0.72	4.92 ± 0.77	0.297
T-G(mmol/l)	1.71(1.43,2.76)	1.69(1.39,2.37)	0.843
HDL(mmol/l)	1.00 ± 0.22	1.00 ± 0.14	0.947
LDL-C(mmol/l)	3.04 ± 0.83	3.32 ± 0.75	0.302
Hs-CRP(mg/l)	1.24(0.65,2.47)	2.85(1.07,4.00)	0.105
Fasting glucose(mmol/l)	5.65 ± 0.70	6.28 ± 1.21	0.086
HbA1c(%)	5.73 ± 0.46	6.27 ± 0.64	0.010
Fasting insulin(µU/ml)	14.69 ± 6.01	18.11 ± 6.82	0.143
HOME-IR	3.66 ± 1.57	5.13 ± 2.33	0.040
24hour mean SBP(mmHg)	123.64 ± 8.94	127.00 ± 14.25	0.452
24hour mean DBP(mmHg)	79.07 ± 7.30	85.80 ± 10.86	0.062
Daytime mean SBP(mmHg)	126.93 ± 9.48	131.93 ± 13.69	0.261
Daytime mean DBP(mmHg)	81.79 ± 7.13	89.33 ± 10.80	0.036
Night-time mean SBP(mmHg)	116.50 ± 9.37	116.53 ± 16.69	0.995

Table 3(continued). Clinical and demographic characteristics in obese patients with different severity of OSA

Variables	Mild-moderateOSA (n=17)	Severe OSA (n=16)	P-value
Night-time mean DBP(mmHg)	73.21±8.29	78.27±11.81	0.197
AHI	20.27±5.92	55.07±20.22	0.010
LSaO2	83.0%±6.4%	72.2%±8.9%	0.107
Ts90%	5.0%(1.0%,28.5%)	28.0%(4.7%,38.7%)	0.073
LAD(mm)	34.70(33.25,37.75)	35.20(32.73,37.70)	0.971
LAA(mm²)	19.00(15.50,21.00)	18.00(16.25,19.00)	0.562
LVEDD(mm)	49.05±3.39	49.30±4.20	0.853
E/A	1.24(1.04,1.53)	1.09(0.85,1.23)	0.105
LAP(mmHg)	9.0(6.0,10.0)	8.5(6.0,9.8)	0.373
LVEF	69.0%±5.8%	70.4%±4.0%	0.419
GLS	-13.88±2.97	-8.18±1.57	<0.001
GCS	-13.78±3.84	-12.96±3.30	0.513
GAS	-22.63±6.05	-20.77±4.71	0.335
GRS	35.07±11.38	30.92±10.04	0.276
3d-LVEF(%)	57.77±5.80	53.31±5.12	0.026
Abbreviations:. BMI: body mass index; TC: total cholesterol; TG: triglyceride; HDL-C: high density lipoprotein cholesterol; LDL-c:low-density lipoprotein cholesterol;hs-CRP: high sensitivity C-reactive protein;HOMA-IR:homeostasis model assessment of insulin resistance;ABPM:ambulatory blood pressure monitoring; SBP:systolic blood pressure;DBP:diastolic blood pressure; AHI:apnea-hypopnea index;LSaO2:minimum oxygen saturation;Ts90%: the total recorded time spent below 90% oxygen saturation;LAD:left atrial diameter;LAD:left atrial area;LVEDD:left ventricular end diastolic diameter;LAP: left atrial pressure;LVEF: left ventricular ejection fraction; 3D-STE:3D speckle tracking echocardiography; GLS: global longitudinal strain; GCS: global circumferential strain; GAS: global area strain; GRS: global radial strain.a:p<0.05 control versus mild OSA,b:p<0.05 control versus moderate OSA, c:p<0.05control versus severe OSA,d:p<0.05 mild OSA versus moderate OSA,e:p<0.05 mild OSA versus severe OSA,f:p<0.05 moderate OSA versus severe OSA

Table 4 Clinical and demographic characteristics in non-obese patients with different severity of OSA

Variables	Mild-moderateOSA (n=24)	Severe OSA (n=22)	P-value
Sex(man：woman)	18:6	19:3	0.332
Age	40.54±10.03	41.82±11.22	0.686
BMI(kg/m²)	25.19±2.22	25.65±1.36	0.404
T-C(mmol/l)	4.48±0.71	4.91±0.89	0.072
T-G(mmol/l)	1.71(1.19,2.47)	1.35(0.98,2.41)	0.538
HDL(mmol/l)	1.01(0.84,1.17)	1.17(1.04,1.32)	0.009
LDL-C(mmol/l)	2.84±0.74	3.26±0.71	0.056
Fasting glucose(mmol/l)	5.25(4.93,5.90)	5.30(5.10,5.73)	0.749
Hs-CRP(mg/l)	0.70(0.31,1.45)	0.95(0.51,2.62)	0.267
HbA1c(%)	5.6(5.4,5.9)	5.7(5.6,5.9)	0.154
Fasting insulin(µU/ml)	10.51±4.38	12.04±4.91	0.270
HOME-IR	2.61±0.24	2.90±1.14	0.406
24hour mean SBP(mmHg)	115.00±8.56	120.57±11.54	0.083
24hour mean DBP(mmHg)	74.76±6.82	79.48±8.02	0.047
Daytime mean SBP(mmHg)	118.90±9.57	124.95±12.22	0.082
Daytime mean DBP(mmHg)	78.14±7.55	82.86±8.10	0.058
Night-time mean SBP(mmHg)	105.80±8.55	110.76±10.92	0.097

Table 4

(continued) .Clinical and demographic characteristics in non-obese patients with different severity of OSA
Variables	Mild-moderateOSA (n = 24)	Severe OSA (n = 22)	P-value
Night-time mean DBP(mmHg)	65.0(62.3,69.0)	71.0(65.0,77.5)	0.029
AHI	18.23 ± 7.82	47.06 ± 14.60	< 0.001
LSaO2	90.5%(87.0%,92.5%)	84.0%(81.0%,94.0%)	< 0.001
Ts90%	5.0%(0.8%,13.5%)	14.7%(5.2%,30.4%)	0.007
LAD(mm)	32.90 ± 4.00	32.85 ± 3.38	0.961
LAA(mm²)	16.0(15.0,18.0)	15.0(14.8,18.0)	0.704
LVEDD(mm)	47.05 ± 3.40	45.67 ± 3.28	0.169
E/A	1.23 ± 0.29	1.14 ± 0.31	0.293
LAP(mmHg)	7.5(5.3,9.0)	8.0(5.0,9.0)	0.991
LVEF(%)	69.6 ± 4.8	69.3 ± 4.8	0.829
GLS	-15.80 ± 3.56	-11.75 ± 1.96	< 0.001
GCS	-12.74 ± 3.26	-13.71 ± 2.96	0.601
GAS	-22.54 ± 5.31	-24.36 ± 4.54	0.422
GRS	34.35 ± 10.28	37.75 ± 8.30	0.604
3d-LVEF(%)	54.78 ± 5.84	55.20 ± 5.66	0.829
Abbreviations:. BMI: body mass index; TC: total cholesterol; TG: triglyceride; HDL-C: high density lipoprotein cholesterol; LDL-c:low-density lipoprotein cholesterol;hs-CRP: high sensitivity C-reactive protein;HOMA-IR:homeostasis model assessment of insulin resistance;ABPM:ambulatory blood pressure monitoring; SBP:systolic blood pressure;DBP:diastolic blood pressure; AHI:apnea-hypopnea index;LSaO2:minimum oxygen saturation;Ts90%: the total recorded time spent below 90% oxygen saturation;LAD:left atrial diameter;LAD:left atrial area;LVEDD:left ventricular end diastolic diameter;LAP: left atrial pressure;LVEF: left ventricular ejection fraction; 3D-STE:3D speckle tracking echocardiography; GLS: global longitudinal strain; GCS: global circumferential strain; GAS: global area strain; GRS: global radial strain.a:p < 0.05 control versus mild OSA,b:p < 0.05 control versus moderate OSA, c:p < 0.05control versus severe OSA,d:p < 0.05 mild OSA versus moderate OSA,e:p < 0.05 mild OSA versus severe OSA,f:p < 0.05 moderate OSA versus severe OSA.

In 3D speckle tracking echocardiography, OSA patients' GLS, GCS, GAS, and GRS were more damaged than the control group (P < 0.05). Intra-group analysis between groups showed that the strain of patients with moderate to severe OSA was worse than the control group (P < 0.05). The GLS of the obese OSA group was worse than that of the non-obese group (P < 0.05). In the same weight group, the more severe the OSA, the worse the strain.All the data are listed in Table 1–4.A example of strain image was shown in Fig. 1.

Correlations Analysis Of 3d-ste Strain Parameters In Osa Patients

Negative global longitudinal strain (GLS), an indicator of impaired longitudinal strain was positively correlated with BMI (r = 0.406, P < 0.001) ,AHI(r = 0.610, P < 0.001)and HOME-IR (r = 0.431, P < 0.001) in patients with OSA. There were no linear correlations (P > 0.05) with age, gender, and blood pressure(including 24hour mean SBP and DBP). This indicates that the larger the BMI,the more deteriorated the OSA condition,and the more severe the insulin resistance, the more severe was the longitudinal strain damaged. There were no linear correlations between GCS, GAS, GRS and BMI, AHI,HOME-IR, age, gender,bloodpressureseparately (P > 0.05) (data not shown). The positive results of correlation analysis are shown in Fig. 2.

Multiple Linear Regression Of 3d-ste Strain Parameters In Osa Patients

Multiple linear regression analysis was performed using GLS, GCS, GAS, GRS as dependent variables respectively in patients with OSA, and BMI, HOME-IR, AHI, age, gender, and blood pressure as independent variables. The results showed that BMI [β = 0.245, 95%CI (0.017, 0.473)] was a predictor of GLS in OSA patients,as well asAHI[β = 0.096, 95%CI (0.055, 0.137)].HOME-IR was, however, the predictor of GAS [β = 1.114, 95%CI (0.306, 1.923)] and GRS [β=-2.14, 95%CI(-3.694, -0.586)]. BMI was also the predictor of GCS [β = 0.548, 95%CI (0.003, 1.092)] The results of multiple linear regression analysis are given in Table 5.

Table 5

Multiple linear regression of 3D-STE strain parameters in OSA patients.
Variables	β	95%CI	t	P-value
GLS
Constants	-19.230	(-29.755,-8.706)	-3.651	0.001
BMI	0.245	(0.017,0.473)	2.151	0.035
HOME-IR	0.376	(0.197,1.230)	1.633	0.108
AHI	0.096	(-0.133,0.019)	4.704	< 0.001
Age	0.051	(-0.024,0.127)	1.355	0.180
Gender	-1.475	(-3.921,0.971)	-1.205	0.233
SBP	-0.011	(-0.131,0.109)	-0.183	0.856
DBP	-0.049	(-0.212,0.113)	-0.606	0.547
GAS
Constants	-10.756	(-29.254, 7.742)	-1.162	0.250
BMI	0.003	(-0.398, 0.403)	0.014	0.989
HOME-IR	1.114	(0.306, 1.923)	2.756	0.008
AHI	-0.069	(-0.14, 0.003)	-1.921	0.059
Age	-0.022	(-0.154, 0.111)	-0.325	0.746
Gender	-3.946	(-8.246, 0.353)	-1.834	0.071
SBP	-0.101	(-0.311, 0.11)	-0.957	0.342
DBP	0.056	(-0.229, 0.341)	0.393	0.695
GCS
Constants	-0.107	(-0.377,0.162)	-0.796	0.429
BMI	0.548	(0.003,1.092)	2.01	0.049
HOME-IR	-0.039	(-0.087,0.009)	-1.614	0.112
AHI	-0.034	(-0.123,0.056)	-0.754	0.454
Age	-1.215	(-4.111,1.682)	-0.838	0.405

Table 5

(continued). Multiple linear regression of 3D-STE strain parameters in OSA patients.
Variables	β	95%CI	t	P-value
Gender	-0.107	(-0.377,0.162)	-0.796	0.429
SBP	-0.1	(-0.241,0.042)	-1.408	0.164
DBP	0.101	(-0.091, 0.293)	1.052	0.297
GRS
Constants	12.576	(-22.989, 48.14)	0.707	0.482
BMI	0.118	(-0.652, 0.888)	0.306	0.761
HOME-IR	-2.14	(-3.694, -0.586)	-2.752	0.008
AHI	0.098	(-0.04, 0.235)	1.416	0.162
Age	0.022	(-0.233, 0.278)	0.176	0.861
Gender	9.179	(0.913, 17.445)	2.219	0.030
SBP	0.114	(-0.29, 0.518)	0.564	0.575
DBP	-0.04	(-0.588, 0.509)	-0.145	0.885
Multiple linear regression was used to analyze the predictive factors of three-dimensional strain. P < 0.05 was considered as statistically significant.BMI: body mass index; HOMA-IR:homeostasis model assessment of insulin resistance;AHI:apnea-hypopnea index;Gender:(assignment: male = 1, female = 2) ; GLS: global longitudinal strain; GCS: global circumferential strain; GAS: global area strain; GRS: global radial strain.

Cardiac magnetic resonance is the gold standard for evaluating the global or local myocardial function. The 3D echocardiography correlates well with cardiac magnetic resonance in measuring myocardial strain.In addition, it has the advantage of being more sensitive, easier to handle, and cheaper ^{[6, 11]}. Torrent-Guasp et al. ^[12] found that various arrangements of myocardial fibers in anatomy determine the form of myocardial movement in different directions corresponds to GLS,GRS,GCS and GAS.. Previous studies have confirmed that patients with OSA or obesity generally exhibit impaired LV diastolic function with normal LVEF values, which can reliably reflect the normal systolic function of the left ventricle^[13]. In our study we found that OSA patients' left ventricular strain was worse than the control group. Within the OSA group, the GLS of obese patients was significantly damaged compared to the non-obese patients. Consistent with the results of previous studies^[14], the impaired LV myocardial strain observed in our study points out that though the LV ejection fraction is still normal, subclinical LV systolic insufficiency is already developing in patients with OSA. In addition, we found that the subclinical left ventricular systolic function was more significantly impaired in OSA patients who also suffered from obesity.

Studies have shown that OSA increases the risk of hypertension, arrhythmia, coronary heart disease, and heart failure^[14]. The ARIC-SHHS study showed that residents with OSA already had subclinical myocardial damage without any discernible cardiovascular and cerebrovascular diseases^[16].Recurrent apnea and hypoxemia at night in OSA patients result in sympathetic hyperactivity and decreased parasympathetic excitability. This results in oxidative stress, systemic inflammation, and vascular endothelial dysfunction that leads to insulin resistance, abnormal lipolysis, and myocardial metabolic disorders. The significantly higher values of the biochemical parameters in the OSA group compared to the control and that of the OSA obese group compared to the non-obese measured in the present study reflect severe metabolic abnormalities and systemic inflammatory in OSA patients, particularly in those with obesity.

The OSA is also accompanied by cardiac hypertrophy and myocardial fibrosis, which are essential in myocardial remodeling^[17]. The elevated blood pressure level in patients with OSA increases the left ventricular afterload. Hypoxia and carbon dioxide retention exaggerate swings in intrathoracic pressure during the obstructive episode, which leads to an increase in return blood volume and left ventricular preload^[18]. At the molecular level, the contraction and relaxation of the heart are highly energy-dependent, 90% of which is provided by the oxidative phosphorylation in mitochondria. Intermittent hypoxia can lead to mitochondrial DNA dysfunction, mutation, or reduction, which results in cardiac insufficiency^[19]. In our study, the left ventricular strain was significantly damaged in moderate to severe OSA compared to the normal subjects. The larger the AHI, the more powerful is the strain damage. The underlying mechanism leading to cardiac insufficiency in OSA obese patients may be the aggravated oxidative stress and mitochondrial dysfunction due to apnea and hypoxemia.

In this study, the left ventricular global longitudinal strain was worse in obese OSA than non-obese OSA patients, and the higher the BMI, the more severe the damage to the left ventricular strain was. Studies have shown that the risk association between obesity and cardiovascular disease is independent of other factors like hypertension, diabetes, and hyperlipidemia, and so on^[20]. The duration and severity of obesity have been recognized as important determinants of the development of heart failure. Large-scale clinical trials and basic experiments have proved that obesity leads to left ventricular hypertrophy, left atrium volume increase, and diastolic dysfunction.^[21]. Although the left ventricular ejection fraction tends to be normal in obese patients, several speckle-tracking echocardiography studies have shown that the radial or longitudinal strain is impaired in asymptomatic obese patients, both children and adults. In the heart of obese people, there is an accumulation of intracellular triglycerides and lipids. The myocardial substrate selection due to obesity favors fatty acid oxidation, leading to lipotoxicity and myocardial dysfunction^{[22, 23]}. The thickening of epicardial adipose tissue (EAT) leads to excessive activation of macrophages, releasing pro-inflammatory cytokines such as TNF-α, IL-1β, and IL-6. This results in activated pro-inflammatory signaling pathways that further aggravates organ dysfunction^[25]. In addition, the activation of the renin-angiotensin-aldosterone system with stimulation of the sympathetic nervous system in obese patients increases left ventricular afterload even in normotensive-obese patients, leading to cardiac remodeling and myocardial fibrosis.

Our study shows that the LV strain is associated with the degree of insulin resistance in addition to OSA and obesity. Previous studies have shown that insulin resistance is one of the major causes of myocardial injury, and both obesity and OSA can cause subclinical myocardial dysfunction by insulin resistance^[26]. The degree of OSA correlates with the severity of insulin resistance. Intermittent hypoxemia and oxidative stress in OSA patients have been critical factors leading to insulin resistance. It has been suggested that ATP synthesis in pancreatic islet β cells and inhibit insulin secretion is affected by hypoxemia. In addition, it reduces the phosphorylation of insulin receptor tyrosine kinases and thereby reduces the effects and sensitivity of the insulin receptors^[27], resulting in insulin resistance. A high HOMA-IR index is one of the characteristics of metabolic disorders in obese patients. Insulin resistance increases cardiac toxicity by altering myocardial metabolism. The toxicity and the increased release of insulin-related growth factors activate the sympathetic and RAAS systems that result in myocardial remodeling^[28]. Studies have shown that hyperinsulinemia can further deteriorate myocardial structure in patients suffering from heart failure with preserved ejection fraction (HFpEF). In addition, insulin resistance is independently associated with impaired GLS, leading to LV hypertrophy and damaged myocardial function over time^[29].

Various mechanisms closely related to obesity, such as insulin resistance, oxidative stress, systemic inflammation, visceral fat accumulation, and dyslipidemia, may also occur as OSA-associated manifestations. Obesity and OSAS interlink mutually and exist as complex, interleaved vicious cycles. This eventually leads to overlapping and potential effects of organ crosstalk^[30]. It is widely known that both obesity and OSAS can independently lead to cardiovascular complications. Our study further elucidates that the coexistence of obesity and OSA might have a synergistic effect on myocardial strain compared to the presence of OSA alone. In patients with OSA, if combined with obesity, the myocardial strain damage will be further exacerbated. This is also valid for obese patients sufferingfrom sleep apnea.As discussed previously, although obesity contributes to the development of cardiac insufficiency through several mechanisms, many of them also overlap with those of OSA. Obesity is also characterized by an increase in systemic blood volume and cardiac output, unlike OSA, which alters hemodynamics, increasing LV pressure and volume, and changes the strain of the myocardium. Obesity tends to coexist with OSA. And based on 3DSTE results of this study, it might be suggested that OSA and obesity have a superimposed effect on the early impairment of left ventricular systolic function. Therefore, for the treatment of OSA and improving the cardiovascular prognosis, it is crucial to focus on the treatment and the weight management of OSA patients.

Our study has several limitations. A comparison should also be drawn between obese but not OSA patients with the obese and OSA patient. None of the controls were even overweight. Most of the patients with OSA are obese, which may affect the accuracy of speckle-tracking and data analysis due to unclear image quality restricted by ultrasonic transmission conditions. As an observational study, the results may be affected by confounding factors such as blood pressure levels. Definitive conclusions regarding LV myocardial deformation could not be drawn as our sample size is small. As such, the reliability and accuracy of our research results need to be further verified by expanding the sample size. In addition, the observed substantial improvement in LV strain after interventional treatment of obese and OSA patients needs to be confirmed in future studies with a larger cohort of patients having OSA with or without obesity.

Our study thus shows that the myocardial strain of OSA patients is impaired before the left ventricular ejection fraction damages. This indicates that the left ventricular systolic function of patients with OSA is damaged at the early stage. When obesity coexists with OSA, the pathophysiological effects of both superimpose, possibly leading to more severe impairment of cardiac function through mechanisms such as hypoxia and insulin resistance.

Funding (Not applicable)

Conflicts of interest/Competing interests (Not applicable)

Availability of data and material

All data generated or analyzed during this study are included in this published article and its supplementary information files

Code availability (Not applicable)

Authors' contributions

Liqiang Zhang and Yongzhen Zhangput forwarded the idea for this article and guided the implementation of the whole project. Weihong Li and Jianli Wang assisted with project implementation and patient enrollment, and Dr.Li performed echocardiographic measurements. Zixuan Hu and Yongwei Huang were responsible for baseline data collection, PSG monitoring, and reporting analysis. Jingwen Zhao was responsible for data collection and analysis and wrote the article. Dr. Liqiang Zhang critically revised the work. All authors read and approved the final manuscript.

Ethics approval

The Ethics Committee approved this study of Peking University Third Hospital (M2017152).

Consent to participate: All the subjects had given their written informed consent.

Consent for publication: All the authors agreed to the publication of the manuscript.

Somers VK, White DP, Amin R, American Heart Association/american College Of Cardiology Foundation Scientific Statement from the American Heart Association Council for High Blood Pressure Research Professional Education Committee, Council on Clinical Cardiology, Blood Institute National Center on Sleep Disorders Research (National Institutes of Health)[J]. Circulation (2008) Sleep apnea and cardiovascular disease: an, Stroke Council, and Council On Cardiovascular Nursing. In collaboration with the National Heart, Lung, and 118(10):1080–1111.DOI:10.1161/CIRCULATIONAHA.107.189375
Bauters F, Rietzschel ER, Hertegonne KB et al (2016) The Link Between Obstructive Sleep Apnea and Cardiovascular Disease[J]. Curr Atheroscler Rep 18(1):1. DOI:10.1007/s11883-015-0556-z
Carneiro G, Zanella MT (2018) Obesity metabolic and hormonal disorders associated with obstructive sleep apnea and their impact on the risk of cardiovascular events[J]. Metabolism 84:76–84DOI. 10.1016/j.metabol.2018.03.008
Korcarz CE, Peppard PE, Young TB et al (2016) Effects of Obstructive Sleep Apnea and Obesity on Cardiac Remodeling: The Wisconsin Sleep Cohort Study[J]. Sleep 39(6):1187–1195. DOI:10.5665/sleep.5828
Collier P, Phelan D, Klein A (2017) A Test in Context: Myocardial Strain Measured by Speckle-Tracking Echocardiography[J]. J Am Coll Cardiol 69(8):1043–1056. DOI:10.1016/j.jacc.2016.12.012
Seo Y, Ishizu T, Atsumi A et al (2014) Three-dimensional speckle tracking echocardiography[J]. Circ J 78(6):1290–1301. DOI:10.1253/circj.cj-14-0360
Sateia MJ (2014) International classification of sleep disorders-third edition: highlights and modifications[J]. Chest 146(5):1387–1394. DOI:10.1378/chest.14-0970
Flegal KM (2021) BMI and obesity trends in Chinese national survey data[J]. Lancet 398(10294):5–7. DOI:10.1016/S0140-6736(21)00892-8
Tang Q, Li X, Song P et al (2015) Optimal cut-off values for the homeostasis model assessment of insulin resistance (HOMA-IR) and pre-diabetes screening: Developments in research and prospects for the future[J]. Drug Discov Ther 9(6):380–385. DOI:10.5582/ddt.2015.01207
Berry RB, Budhiraja R, Gottlieb DJ et al (2012) Rules for scoring respiratory events in sleep: update of the 2007 AASM Manual for the Scoring of Sleep and Associated Events. Deliberations of the Sleep Apnea Definitions Task Force of the American Academy of Sleep Medicine[J]. J Clin Sleep Med 8(5):597–619. DOI:10.5664/jcsm.2172
Brown J, Jenkins C, Marwick TH (2009) Use of myocardial strain to assess global left ventricular function: a comparison with cardiac magnetic resonance and 3-dimensional echocardiography[J]. Am Heart J 157(1):101–102. DOI:10.1016/j.ahj.2008.08.032
Buckberg GD, Coghlan HC, Hoffman JI et al (2001) The structure and function of the helical heart and its buttress wrapping. VII. Critical importance of septum for right ventricular function[J]. Semin Thorac Cardiovasc Surg 13(4):402–416. DOI:10.1053/stcs.2001.29961
Zhou NW, Shu XH, Liu YL et al (2016) A Novel Method for Sensitive Determination of Subclinical Left-Ventricular Systolic Dysfunction in Subjects With Obstructive Sleep Apnea[J]. Respir Care 61(3):366–375. DOI:10.4187/respcare.04381
Wang D, Ma GS, Wang XY et al (2016) Left ventricular subclinical dysfunction associated with myocardial deformation changes in obstructive sleep apnea patients estimated by real-time 3D speckle-tracking echocardiography[J]. Sleep Breath 20(1):135–144. DOI:10.1007/s11325-015-1197-8
Bradley TD, Floras JS (2009) Obstructive sleep apnoea and its cardiovascular consequences[J]. Lancet 373(9657):82–93. DOI:10.1016/S0140-6736(08)61622-0
Querejeta RG, Redline S, Punjabi N et al (2013) Sleep apnea is associated with subclinical myocardial injury in the community. The ARIC-SHHS study[J]. Am J Respir Crit Care Med 188(12):1460–1465. DOI:10.1164/rccm.201309-1572OC
Drager LF, Togeiro SM, Polotsky VY et al (2013) Obstructive sleep apnea: a cardiometabolic risk in obesity and the metabolic syndrome[J]. J Am Coll Cardiol 62(7):569–576. DOI:10.1016/j.jacc.2013.05.045
Niroumand M, Kuperstein R, Sasson Z et al (2001) Impact of obstructive sleep apnea on left ventricular mass and diastolic function[J]. Am J Respir Crit Care Med 163(7):1632–1636. DOI:10.1164/ajrccm.163.7.2007014
Kim YS, Kwak JW, Lee KE et al (2014) Can mitochondrial dysfunction be a predictive factor for oxidative stress in patients with obstructive sleep apnea?[J]. Antioxid Redox Signal 21(9):1285–1288. DOI:10.1089/ars.2014.5955
Ndumele CE, Cobb L, Lazo M et al (2018) Weight History and Subclinical Myocardial Damage[J]. Clin Chem 64(1):201–209. DOI:10.1373/clinchem.2017.282798
Turkbey EB, McClelland RL, Kronmal RA et al (2010) The impact of obesity on the left ventricle: the Multi-Ethnic Study of Atherosclerosis (MESA)[J]. JACC Cardiovasc Imaging 3(3):266–274. DOI:10.1016/j.jcmg.2009.10.012
Sanchez AA, Singh GK (2014) Early ventricular remodeling and dysfunction in obese children and adolescents[J]. Curr Treat Options Cardiovasc Med 16(10):340. DOI:10.1007/s11936-014-0340-3
Blomstrand P, Sjöblom P, Nilsson M et al (2018) Overweight and obesity impair left ventricular systolic function as measured by left ventricular ejection fraction and global longitudinal strain[J]. Cardiovasc Diabetol 17(1):113. DOI:10.1186/s12933-018-0756-2
Rider OJ, Cox P, Tyler D et al (2013) Myocardial substrate metabolism in obesity[J]. Int J Obes (Lond) 37(7):972–979. DOI:10.1038/ijo.2012.170
Packer M (2018) Epicardial Adipose Tissue May Mediate Deleterious Effects of Obesity and Inflammation on the Myocardium[J]. J Am Coll Cardiol 71(20):2360–2372. DOI:10.1016/j.jacc.2018.03.509
Morgenstern M, Wang J, Beatty N et al (2014) Obstructive sleep apnea: an unexpected cause of insulin resistance and diabetes[J]. Endocrinol Metab Clin North Am 43(1):187–204. DOI:10.1016/j.ecl.2013.09.002
Li M, Li X, Lu Y (2018) Obstructive Sleep Apnea Syndrome and Metabolic Diseases[J]. Endocrinology 159(7):2670–2675. DOI:10.1210/en.2018-00248
Alpert MA, Karthikeyan K, Abdullah O et al (2018) Obesity and Cardiac Remodeling in Adults: Mechanisms and Clinical Implications[J]. Prog Cardiovasc Dis 61(2):114–123. DOI:10.1016/j.pcad.2018.07.012
Kucukseymen S, Neisius U, Rodriguez J et al (2020) Negative synergism of diabetes mellitus and obesity in patients with heart failure with preserved ejection fraction: a cardiovascular magnetic resonance study[J]. Int J Cardiovasc Imaging 36(10):2027–2038. DOI:10.1007/s10554-020-01915-4
Newmarch W, Weiler M, Casserly B (2019) Obesity cardiomyopathy: the role of obstructive sleep apnea and obesity hypoventilation syndrome[J]. Ir J Med Sci 188(3):783–790. DOI:10.1007/s11845-018-01959-5

Evaluation of Left Ventricular Function in Obese Patients with Obstructive Sleep Apnea by Three-dimensional Speckle Tracking Echocardiography

Status:

Version 1

Abstract

Purpose

Methods

Results

Conclusion

Figures

Introduction

Methods

Results

Clinical and demographic characteristics in different groups

Correlations Analysis Of 3d-ste Strain Parameters In Osa Patients

Multiple Linear Regression Of 3d-ste Strain Parameters In Osa Patients

Discussion

Study Limitations

Conclusion

Declarations

References

Status:

Version 1