Lipopolysaccharide Binding Protein and Cardiovascular Changes in Obese Children ()
1. Introduction
Childhood obesity is growing globally, with a significant public health burden [1]. Recently, many studies indicated that overweight and obesity of childhood frequently persist in adulthood and are linked to increased morbidity and mortality, such as the increased risk of cardiovascular diseases (CVD), which lead to premature death [2] [3] [4]. It has become apparent in recent years that there is a strong association between the increasing levels of various inflammatory biomarkers and childhood obesity [5] [6]. Nevertheless, the underlying mechanisms that induce the elevated inflammatory burden still unclear. The gut microbiome is one of the most important sources of pro-inflammatory factors [7] [8]. The increased microbial translocation, especially lipopolysaccharide (LPS), was reported to be linked with childhood obesity [9].
Lipopolysaccharide (LPS) is a well-defined pathogen-associated molecular model, which presents in the outer leaflet of most gram-negative bacteria in the outer membrane [10]. LPS may lead to an inflammatory response by activation of monocytes and endothelial cells when absorbed or translocated into intestinal capillaries [11]. The detection of LPS involves soluble proteins, such as Lipopolysaccharide binding protein (LBP) and soluble CD14 (sCD14), which generate and secrete a broad range of response mediators [12] [13]. Throughout the past two decades, several measures have been seen in the LPS-based inflammatory, antibacterial response.
Lipopolysaccharide binding protein (LBP) is produced mainly through hepatocytes and expressed and released through intestinal and visceral adipocytes [14] [15]. Plasma LBP levels are dramatically increased in response to inflammatory challenges [16]. LBP is well-known for binding to LPS substructures, for example, lipid IVa [17]. Plasma LBP significantly accelerates the connection between the LPS monomers released from aggregates and CD14 [18]. LBP counteracts the LPS effect by transferring LPS to lipoproteins.
Several articles were demonstrated an association between coronary artery disease prevalence and the serum of LBP [19] [20] [21] [22]. Furthermore, serum LBP levels were shown to be correlated with metabolic syndrome and type 2 diabetes [10] [18] [23].
Despite the growing evidence that LBP plays a major role in CVD pathophysiology and obesity, data regarding this association in children are rare. Therefore, our objectives were to assess whether there was a difference between overweight/obese and normal-weight children in plasma LBP levels and to assess the cardiovascular changes in both groups.
2. Material and Methods
2.1. Design and Population
In an observational, case-control, study, a total of 30 obese children were recruited from Pediatric and Endocrinology Outpatient Clinics of Al Zahraa University Hospital. Children were included if they aged between 8 - 16 years old and had a body mass index (BMI) above 95th percentile. In addition, 30 children, with a BMI of less than 95th percentile, were included as a control group. We excluded children with history of cardiovascular (CV) diseases or other chronic disorders (such as pulmonary or renal diseases), non-endocrinal or syndromic obesity, renal or hepatic failure, familial dyslipidemia, immunological diseases, and/or patients with acute or chronic infection. The study was conducted from September 2019 to February 2020 after the approval of local ethics committee (IRB No. 202005256) and written informed consents, signed by the parents, were a prerequisite for inclusion of eligible children.
2.2. Data Collection and Laboratory Investigations
After detailed history taking, eligible patients underwent physical examination to determine height, weight, BMI, and blood pressure. The laboratory investigation included complete blood count (CBC), hepatic function, lipid profile, blood glucose profile, thyroid functions, and serum LBP, while imaging investigations included carotid Duplex and echocardiography.
For laboratory investigations, 5 mL of venous blood were withdrawn from each child, and then centrifuged. The serum was separated and divided into two portions. The first portion used for blood cholesterol, high density lipoprotein (HDL), low density lipoprotein (LDL), and triglycerides (Cobas Integra 400 plus, Roche diagnostics, Germany). The second portion was stored at-20 until be used for LBP assessment, which was measured by enzyme-linked immunosorbent assay (ELISA) technique using the kitof Catalog (Number MBS704355 and Lot No. H1560E116. My Biosource comp; USA). The reference range was 0.625 - 40 μg/ml according to manufacturer instructions with the limit of detection 3.125 - 200 ng/ml and a sensitivity of 1.875 ng/ml.
2.3. Echocardiography
Children were instructed to lie in the left lateral recumbent position to perform the 2D, transthoracic echocardiographic examination using Vivid E9 (GE Ultrasound, Horten, Norway) ultrasound machine with a multifrequency probe (2.5 MHz). Both apical and parasternal axial views were used for comprehensive assessment, which included Doppler and color flow mapping parameters. The LV mass was calculated using modified Penn formula. The pulsed Doppler parameters included early (e) and late (a) diastolic velocity across the mitral valve, E/A ratio, and deceleration time. The tissue Doppler imaging (TDI) parameters for LV (averaged across four mitral annuli) and RV (at RV free wall) assessment included Sm, Em, and Am. Additionally the Speckle-tracking echocardiography (STE) analysis was used to obtain the LV and RV global longitudinal strain via Echo PAC version 210. In addition, the left atrium volume index (LAVI) was calculated through obtained the maximum and minimum volumes the end of T wave and at the onset of QRS wave, followed by the calculation of LAVI by incorporating the body surface area [24] [25]. Other parameters for RV systolic function included tricuspid annular plane systolic excursion (TAPSE) and fractional area change (FAC) were assessed [26]. All measurements were obtained using two-to-three consecutive cardiac cycles.
2.4. Carotid Duplex
In all recruited participants, the carotid Duplex was performed using General Electric (GE) system 7 machines, with 10 MHz transducer. Patient was placed in supine position with chin extended and the head was rotated away from side to be examined, it is helpful to put a pillow below patient’s shoulder for hyperextension of the neck as recommended by the 2007 Mannheim Consensus. Image was focused on the posterior wall of each common carotid artery, in a segment 1 cm proximal to the carotid bifurcation on each side. We solely assessed the intima (echogenic line) and the media (hypoechoic line) for assessment of intima-media thickness (IMT). IMT for common carotid artery of both sides were assessed [27].
2.5. Statistical Analysis
The SPSS version 20.0 (SPSS Inc., Chicago, Illinois, USA) was used for data analysis. The continuous and dichotomous data were summarized in the form of the quantitative mean ± standard deviation (SD) and frequency (percentages), respectively. The association between continuous and dichotomous data was examined by independent-samples t-test or Mann-Whitney test. Chi squared test was applied to test the hypotheses in categorical variables. Null hypothesis was rejected when p value at level less than 0.05.
3. Results
In this case-control study, 30 obese children (mean age 10.48 ± 1.99 years; females = 56.7%) and 30 matched controls (mean age 11.42 ± 2.03 years; females = 50%). As expected, the patients’ weight, BMI, and waist circumference were significantly higher than the control group (p < 0.001). Likewise, the mean arterial blood pressure, serum LDL, HLD, and cholesterol were significantly higher in obese children than the control group (p < 0.001; Table 1).
In terms of echocardiographic findings, the mean aortic root and LAVI were significantly higher in the obese children than the control group (p < 0.001). Obese children had significantly higher interventricular septum (p < 0.001), posterior wall diameter (p < 0.001), LV end diastolic diameter (p < 0.001), stroke volume (p < 0.001), LV mass index (p < 0.001), relative wall thickness (p = 0.035), mitral E/A ratio (p = 0.011), mitral annular systolic velocity (Sm) (p < 0.001), mitral annular early diastolic velocity Em (p = 0.005), and lower ejection fraction (p < 0.001) and LVGL strain (p = 0.02) than the control group. On the other hand, the obese children had apparent RV dysfunction as evident by the lower TAPSE, RV global strain, and FAC (p < 0.001; Table 2).
In terms of carotid Dupplex findings, we demonstrated higher CIMT in patients than control group (0.08 ± 0.01 versus 0.04 ± 0.01 mm; P < 0.001; Figure 1).
With regard to serum LBP, the mean LBP was significantly higher in obese children than the control group (52.74 ± 17.25 versus 12.34 ± 2.67 µg/mL, respectively; p < 0.001; Figure 2). The ROC curve showed that the serum LBP, at a cutoff value of >19 µg/mL, was significant discriminator of obesity with a sensitivity of 96.67% and specificity of 100% (Figure 3). The regression analysis showed that BMI was independent predictor of serum LBP (B coefficient = 0.684; p = 0.024).
The serum LBP correlated significantly with age (r = 0.58; p = 0.001), BMI (r = 0.834; p = 0.001), and LV longitudinal strain (r = 0.362; p = 0.05; Table 3).
Figure 1. Comparison between obese and control group as regard carotid intma-media thickness.
Table 1. Baseline demographic, clinical data and lipid profile of study groups.
Abbreviation: BSA = body surface area; BMI = body mass index; cholest = cholesterol; LDL = low density lipoprotein; HDL = high density lipoprotein; TG = triglyceride.
Table 2. Echocardiographic parameters in the study population.
Abbreviation: Ao, aortic; LAD, left atrial dimensions; IVSd, interventricular septal thickness in diastole; LVIDd, left ventricular internal dimension in diastole; LVIDs, left ventricular internal dimension in systole; LVPWd, left ventricular posterior wall thickness in diastole; LVDD, left ventricular dimension in diastole; LVDS, left ventricular dimension in systole; LVVD, left ventricular volume in diastole; LVVS, left ventricular volume in systole; SV, stroke volume; FS, fractional of shorting; EF, ejection fraction; LVMI, left ventricle mass index; RWT, Relative wall thickness; MV, mitral valve; E vel, early diastolic velocity; A vel, late diastolic or atrial velocity; TDI, tissue Doppler imaging; LVSm, myocardial systolic excursion velocity; LVEm, myocardial early diastolic excursion velocity; LV Am, myocardial late diastolic velocity; E/Ea, early diastolic velocity measured by pulsed-Doppler/myocardial early diastolic excursion velocity measured by tissue Doppler echocardiography; LVGLS TDI, left ventricular global longitudinal stain by tissue Doppler; LVGLS 2STE, left ventricular global longitudinal strain by 2D speckle tracking; TV, tricuspid valve; FAC, fractional area change; TAPSE, tricuspid annular plane systolic excursion; LAVI, left atrial volume index.
Table 3. Correlation between LBP level and different variables.
Figure 2. Comparison between obese and control group as regard LBP level.
Figure 3. Roc analysis of LBP and best cut off value to discriminate obesity.
4. Discussion
To the best of our knowledge, the correlation between the serum LBP levels and the presence of atherosclerosis and CVD was reported in several epidemiological studies [19] [20] [28]. Moreover, a significant positive association between the serum level of LBP and atherosclerosis was reported by two cross-sectional studies as assessed by aortic pulse wave velocity and carotid intima-media thickness [13] [29]. Furthermore, Lepper’s study showed that patients with higher serum LBP levels were significantly associated with increased all-cause and cardiovascular mortality risks [20].
The Hisayama study has shown that higher LBP serum levels, even after adjustment to conventional risk factors, are significantly correlated with the development of CVD, insulin resistance index, and HOMA-IR. For the CVD subtypes, participants with greater serum levels of LBP were at significantly more risk of stroke, especially ischemic stroke, while hemorrhagic stroke was not statistically significant. The CHD incidence showed a tendency to grow, but this correlation was not statistically significant with high serum LBP levels [30].
Regarding the interventricular septum (IVS), our findings indicated that obese children had a thicker IVS than control (<0.001). Similarly, Schusterova and his colleagues demonstrated a significant association between overweight/obesity and IVS compared to normal individuals (p < 0.01) [31]. In contrast, Balaji et al. showed that the IVS was comparable in both obese children and normal group (p = 0.11) [32]. Regarding the association between IVS and the serum level of BLP, we could not find any significant association (r = 0.20, p = 0.271).
In addition, we observed that the Ao, LAD, LVVD, and SV were significantly increased in obese children (p < 0.001). These findings were in agreement with the previous literature [31] [32] [33]. However, all of these parameters were not significantly associated with LBP level, which requires further studies with larger samples to explore this association. Similarly, LV mass, LV mass index, and I were significantly increased in the obese groups without any significant association with LBP serum levels. Mangner et al. compared obese children and non-obese children regarding the echocardiographic parameters and showed a significant increase in the following parameters in obese children: IVSd, PWd, LVEDD, LVM, LVMI, SV, and LA diameter. In terms of LVEF, both groups were comparable (p = 0.4). Obesity leads to greater overall blood volume, cardiac output, heart and peripheral resistance. Increases in pressure and volume lead to dilation of cardiac chambers because of the increased filling [34]. It is proved that adverse cardiac events and worse prognosis are associated with LV dilation and LV hypertrophy, as well as LA enlargement. In addition, increased LA size is also a feature of left ventricle impaired diastolic function [35].
Concerning the LV E/Em and LAVI we found a significant increase in obese children compared to the normal group (p < 0.001), indicating an impairment in the LV diastolic function. The impact of obesity on the LV diastolic function can be attributed to the workload exerted by obesity on cardiac function, which, in return, leads to dilatation and hypertrophy [36]. However, in terms of LVGLS, we could not find any significant difference between both groups (p = 0.293). In agreement, Yang et al., showed that LVGLS could be used as an independent predictor in adults more than in children [37]. On the other hand, we found a significant decrease in TAPSE, RVGLS and FAC among patients’ group (p < 0.001), indicating a global impairment in RV systolic function. Sleep disorders are prevalent among obese individuals; alongside excessive cardiac output, these sleep disorders can significantly increase the pulmonary artery pressure leading to subsequent dysfunction in RV. In previous study by Sokmen et al. [36], subclinical RV dysfunction was noted among young obese patients as noted in this study. However, according to Zeller and his colleagues, the found that parameters of RV function and dimension of right heart chambers were comparable to non-obese subjects even extremely obese subjects (with BMI > 40 kg/m2) had no greater impairment of RV function according to TAPSE [38]. Thus, a future study investigating the RV function in obese patients is warranted.
CIMT is a well-established marker for early atherosclerosis, as well as other vascular abnormalities, in children and adults. Previous reported indicated higher thickness of CIM in pediatric population with hypertension [39] and other cardiovascular diseases [40]. Moreover, the current body of evidence suggests that CIMT is significantly positively correlated with the degree of obesity and other metabolic abnormalities in children [41]. In the present study, we demonstrated higher CIMT in patients than control group (0.08 ± 0.01 versus 0.04 ± 0.01 mm; P < 0.001). These findings are line with previous reports showing higher CIMT values in obese children than normal weight controls [42] [43]. However, in the present study, the serum LBP did not correlate significantly with CIMT. On the contrary, Serrano et al. [13], demonstrated positive correlation between serum LBP and CIMT. Thus, further studies are warranted to characterize the association between serum LBP and markers of atherosclerosis, like CIMT.
On the other hand, LBP serum level was significantly associated with age (r = 0.58, p = 0.001), weight (r = 0.71, p < 0.001), height (r = 0.449, p = 0.013), BSA (r = 0.652, p < 0.001), BMI (r = 0.834, p < 0.001), and WC (r = 0.459, p = 0.01). These findings confirm the association between LBP level and obesity. Furthermore, the LV strain by tissue-Doppler imaging was found to be significantly associated with LBP serum level (r = 0.362, p = 0.050). In multivariate analysis, only BMI was observed to be significantly associated with LBP (B = 2.021, p = 0.024).
5. Limitations
One of the major limitations in our study is the small number of included children. Another factor is that we considered both overweight children and obese children as one group. Also it is better to measure the bacterial endotoxin lipopolysaccharide itself rather than measuring its binding protein. However it is more costly.
In conclusion, our findings showed that obesity was associated with worse lipid profile and cardiovascular function. LBP is a promising predictor of obesity in children. Further studies are required for more exploration about the association between the LBP as a biomarker for cardiovascular disease in obese children.