High molecular weight adiponectin as a biomarker of hypertension in children and adolescents with obesity

Lower HMW (high molecular weight) adiponectin levels are associated with obesity, insulin resistance, and metabolic syndrome in children and adolescents. However, data on HMW levels in pediatric population with hypertension are lacking. This study aimed to examine the association and predictive capacity of HMW levels, HMW/HOMA-IR, and HMW/APN ratio with hypertension in obese children and adolescents. The 299 pediatric subjects were grouped in obese hypertensive (OH), obese normotensive (ON), and normal weight normotensive (NN). Plasma concentrations of HMW were investigated by ELISA. ANOVA was used to compare study groups, and a binary logistic regression analysis was used to verify if HMW, HMW/HOMA-IR, HMW/APN, APN, APN/HOMA-IR, and HOMA-IR are associated to hypertension regardless obesity in children and adolescents. To compare the strength and performance of each biomarker to classify individuals with and without hypertension, the receiver-operating characteristic (ROC) curve, area under the curve (AUC), and Youden index (J) were evaluated. Both HMW plasma levels and the HMW/HOMA-IR ratio were significantly lower in the OH group when compared to the ON group (HMW: 2.00 ± 1.33 µg/mL vs 2.48 ± 1.48 µg/mL; HMW/HOMA-IR ratio: 0.87 ± 0.95 vs 1.27 ± 1.2; P < 0.05) and NN weight groups (HMW: 2.00 ± 1.33 µg/mL vs 4.02 ± 1.99 µg/mL; HMW/HOMA-IR ratio: 0.87 ± 0.95 vs 2.62 ± 1.86; P < 0.05). Hypertension was associated with lowest HMW (OR = 4.50; 95% CI = 1.41–15.84) and HMW/HOMA-IR (OR = 12.13; 95% CI = 2.51–92.93) regardless of obesity. However, HOMA-IR or the HMW/APN was not significant (P > 0.05). In the ROC curve analyses, the HMW and HMW/HOM-IR were more sensitive to detect hypertension in children and adolescents with obesity. Conclusion: Low levels of HMW oligomer and HMW/HOM-IR are associated with hypertension in childhood obesity. Thus, these biomarkers could be clinically useful in identifying hypertension in childhood obesity. What is Known: • HMW has previously been reported as the most biologically active isoform of adiponectin, and lower HMW concentrations are associated with obesity, insulin resistance, and metabolic syndrome in children and adolescents. • HMW/HOMA-IR ratio is a sensitive predictor for metabolic syndrome in adults. What is New: • HMW levels are associated with hypertension in children and adolescents, independently of presence of obesity. • HMW was more sensitive to detect hypertension in children and adolescents with obesity when compared to HMW/HOMA-IR, HMW/APN, APN, APN/HOMA-IR, or HOMA-IR. What is Known: • HMW has previously been reported as the most biologically active isoform of adiponectin, and lower HMW concentrations are associated with obesity, insulin resistance, and metabolic syndrome in children and adolescents. • HMW/HOMA-IR ratio is a sensitive predictor for metabolic syndrome in adults. What is New: • HMW levels are associated with hypertension in children and adolescents, independently of presence of obesity. • HMW was more sensitive to detect hypertension in children and adolescents with obesity when compared to HMW/HOMA-IR, HMW/APN, APN, APN/HOMA-IR, or HOMA-IR.


Introduction
The prevalence of obesity has reached epidemic proportions worldwide in recent years [1]. The Brazilian population is a high-risk group for developing obesity, especially in children and adolescents [2]. This is worrisome because childhood obesity increases the risk of developing hypertension, persisting into adulthood [3,4].
Adiponectin (APN) is one of the main adipokines produced by adipose tissue and is widely known for its proangiogenic and anti-apoptotic effects on vascular endothelium and heart [5]. Mechanistically, many of the effects of APN against hypertension are realized through its ability to increase nitric oxide (NO) production through activation of endothelial nitric oxide synthase [6][7][8]. The authors have previously reported that obese hypertensive children had the lowest levels of APN, and NO bioavailability was positively associated by APN levels [8].
APN circulates in the blood as oligomers of various molecular weights: low molecular weight (LMW), medium molecular weight (MMW), and high molecular weight (HMW) [9]. HMW has previously been reported as the most biologically active isoform, and it has been suggested that lower HMW concentrations are associated with obesity, insulin resistance, and metabolic syndrome in children and adolescents [5,10,11]. In addition, the study suggested that the HMWA/HOMA-IR ratio was a sensitive predictor for metabolic disorders in adults with metabolic syndrome [12]. However, to our knowledge, there is no study that has evaluated the association of HMW with hypertension regardless of obesity in children and adolescents.
Therefore, this study was designed to investigate the association of HMW, HMW/HOMA-IR, and HMW/APN ratio with hypertension in obese children and adolescents.

Subjects and anthropometric and biochemical measurements
The study population consisted of 106 obese normotensive (ON) patients and 50 obese hypertensive (OH) patients recruited as an outpatient from the Endocrinology Outpatient Clinic of Juiz de Fora and from the Child Endocrinology Outpatient Clinic of the IMEPEN Foundation in Juiz de Fora (MG, Brazil). The exclusion criteria for our study subjects comprised a history of diagnosed hypertension, ongoing use of antihypertensive medications, adulthood, pregnancy, diabetes, and secondary hypertension. The normal weight normotensive (NN) group consisted of 143 healthy children and adolescents recruited from the local community and unrelated to the obese patients. All children underwent physical examination. Body weight was measured with a digital scale to the nearest 0.1 kg. The measurement of waist circumference (WC) was taken at the midpoint between the last rib and the iliac crest, while the measurement of hip circumference (HP) was taken at the maximum circumference around the gluteal region with a simple measuring tape and height with a stadiometer accurate to 0.1 cm. Obesity was defined as a body mass index greater than the 95th percentile matched for age and sex [13]. Systolic (SBP) and diastolic (DBP) blood pressure were measured at least 3 times on separate visits, with each measurement taken after at least 15 min of rest. Hypertension was defined as SBP and/or DBP above the 95th percentile accordingly to sex, age, and height [14]. The measurements were taken using a sphygmomanometer mercury column. At the time of clinical care, venous blood samples were collected overnight fasting (8-12 h), immediately centrifuged at 2000 × g for 10 min at room temperature, and plasma samples were stored at − 70 °C until analyzed. Glucose and uric acid concentration, and lipid parameters (triglycerides and high-density lipoprotein [HDL] cholesterol) were determined in plasma and serum, respectively, with routine enzymatic methods using commercial kits (Labtest Diagnostic, SA, Lagoa Santa, Brazil).

Enzyme immunoassays of APN and its HMW oligomer, and insulin
APN and its HMW oligomer were measured in EDTAplasma using commercially available enzyme-linked immunosorbent assay (ELISA) kits (R&D Systems, Minneapolis, MN, USA) according to the manufacturer's instructions. Insulin concentrations were measured in EDTA-plasma using a kit (Genese Produtos Diagnosticos, São Paulo, Brazil). The estimate of insulin resistance obtained by the Homeostatic Model Assessment of Insulin Resistance (HOMA-IR) was calculated as described in Matthews et al. [15].

Data analyses
Clinical and biochemical characteristics were compared by ANOVA followed by Tukey's test (variables with normal distribution) or the Kruskall-Wallis test followed by Dunn's Multiple Comparison test (variables without normal distribution). Categorical variables were compared between groups using χ 2 tests (StatView, Cary, NC, USA).
A binary logistic regression analysis was used to determine the odds ratios for the association of hypertension with biomarkers grouped into quartiles adjusted for biochemical and sociodemographic explanatory variables of the development of hypertension in children and adolescents such as obesity, age, sex, waist circumference/hip circumference, ethnicity, cholesterol, HDL cholesterol, triglycerides and uric acid. Plasma concentrations of HMW, HMW/HOMA-IR, HMW/APN, APN, APN/HOMA-IR, and HOMA-IR were divided into interquartile ranges Q1 (25th), Q2 (50th), and Q3 (75th) percentiles. Interquartile range Q1 corresponds to lower levels, Q2 corresponds to intermediate levels, and Q3 corresponds to higher levels. All requirements of the logistic regression model were carefully inspected and none was violated. To compare the strength and measure the diagnostic test performance of each biomarker to classify individuals with and without hypertension, the receiveroperating characteristic (ROC) curve, area under the curve (AUC), and Youden index (J) were evaluated. The authors chose the cut-off point of 0.2 where it presented the best balance between sensitivity and specificity for all variables. Table 1 shows the clinical and biochemical characteristics of NN, ON, and OH groups. BMI and WC were higher in the ON and OH groups versus the NN group (P < 0.05). Furthermore, SBP, DBP, and Mean Arterial Pressure (MAP) were higher in the OH group when compared to the ON and NN groups (P < 0.05). Regarding the biochemical parameters, the OH group had the lowest levels of APN between the groups (P < 0.05). Higher HOMA-IR index and lower HDL cholesterol were found in OH group in comparison to the NN group (P < 0.05). No significant differences were found in triglyceride levels between groups (P > 0.05).

HMW levels plasma, HMW/HOMA-IR ratio, and HMW/APN ratio in children with obesity and hypertension
As shown in Fig. 1A, B, lower HMW levels and the HMW/HOMA-IR ratio were associated with the presence of hypertension in obese children and adolescents. Both HMW plasma levels and the HMW/HOMA-IR ratio were significantly lower in the OH group when compared to the ON group (HMW: 2.00 ± 1.33 µg/mL vs 2.48 ± 1.48 µg/ mL; HMW/HOMA-IR ratio: 0.87 ± 0.95 vs 1.27 ± 1.2; P < 0.05) and NN weight groups (HMW: 2.00 ± 1.33 µg/mL vs 4.02 ± 1.99 µg/mL; HMW/HOMA-IR ratio: 0.87 ± 0.95 vs 2.62 ± 1.86; P < 0.05). In addition, the ON group had lower HMW levels, HMW/HOMA-IR ratio than the NN group (P < 0.05). A reduction in the HMW/APN ratio was found in the OH group (0.22 ± 0.06) and ON group (0.21 ± 0.06) when compared to NN group (0.29 ± 0.09) (P < 0.05), but no difference was found between OH and ON groups (P > 0.05).

Association between HWM, HMW/HOMA-IR ratio, and HMW/APN ratio levels with hypertension
To determine the association and strength between the biomarkers with hypertension, odds ratio for associations between hypertension and the quartiles of HMW, HMW/HOMA-IR, HMW/APN, APN, APN/HOMA-IR, or HOMA-IR were calculated using logistic regression analysis adjusted for variables that could contribute the hypertension outcome (

Comparison of detection powers between HMW, HMW/HOMA-IR ratio, HMW/APN ratio, APN, APN/ HOMA-IR ratio, and HOMA-IR for hypertension
To compare the detection powers of models including HMW, HMW/HOMA-IR, HMW/APN, APN, APN/HOMA-IR, or HOMA-IR for hypertension, ROC curves were plotted, and the sensitivity, specificity, AUC, and J index of each model are shown in Table 3. The methods that showed the highest sensitivity and specificity were HMW (80.3% and 74.2%, respectively) and HMW/HOMA-IR (74.4% and 76.7%) with adequate balance between sensitivity and specificity to discriminate between subjects with or without hypertension. In addition, they had the highest AUC (0.841 and 0.839, respectively). ROC curves for diagnoses of hypertension are presented in Fig. 2.

Discussion
In this study, we found a significant association of HMW, HMW/HOMA-IR ratio, and HMW/APN ratio with hypertension in children and adolescents with obesity. These results demonstrate that HMW oligomer and HMW/ HOMA-IR ratio decrease in ON and these reductions are more pronounced in OH children and adolescents.
The relationship between APN and blood pressure (BP) is well described in the literature for the elderly, adults, and recently for children and adolescents where plasma APN values were negatively correlated with SBP and DBP [16,   [8,18]. At the time of writing this report, there were no studies evaluating the relationship between HMW and the presence of hypertension in obese children and adolescents. In a study that investigated the relationship between metabolic parameters and endothelial function in children, the HMW levels were associated with cardiovascular risk in overweight and obese children [19]. One study showed that serum levels of HMW were significantly lower in obese children than in controls and this reduction is more pronounced in obese children with greater insulin resistance [20]. Additionally, they reported that both APN and its HMW oligomers were inversely correlated with insulin resistance. In the present study, after adjustment for obesity among other variables, we confirmed the association between lower HMW and HMW/HOMA-IR ratio levels with hypertension, independent of obesity.
APN and its HMW oligomers, which is the most biologically relevant APN isoform, play an important protective role in cardiovascular diseases by modulating the cross-link between endothelial cells and platelets, by stimulating NO production, and by inhibiting vascular endothelial growth factor (VEGF) [21]. A third receptor (AdipoR3) specific for the HMW form of APN has been identified on endothelial cells and smooth muscle cells [22] and T-cadherin has also been shown to be a receptor for hexamers and HMW [23] indicating the relevance of HMW oligomers.
Another study that conducted a meta-analysis of observational studies including 10,230 hypertension cases demonstrates that individuals with increased levels of insulin resistance are independently associated with increased risk of incident hypertension [24]. Together, these data corroborate that both HMW oligomer and HMW/HOMA-IR ratio are closely correlated with obesity and hypertension in children and adolescents.
In the present study, it was constructed ROC curves to examine the ability of HMW, HMW/HOMA-IR, HMW/ APN, APN, APN/HOMA-IR, or HOMA-IR to identify obesity-related hypertension. These data suggest that HMW and HWM/HOMA-IR were more sensitive to detect hypertension in children and adolescents with obesity. Another study evaluated the HMW/HOMA-IR ratio as a biomarker of metabolic syndrome in adults showed that this specific biomarker was more strongly associated with metabolic syndrome than HMW, and had a better detection power for metabolic syndrome, when compared with HMW and HOMA-IR alone [12]. Notably, in the present work, the results confirmed that HMW and HMW/HOMA-IR are more explanatory than other biomarkers for hypertension, thus providing a better tool for hypertension detection in obese children and adolescents.
Additionally, Makni et al. and Cândido et al. demonstrated the ratio of APN and HOMA-IR was more suitable for detecting metabolic syndrome than HOMA-IR, indicating the utility of these biomarkers as a tool to screen for insulin resistance in obese children [25,26].
Childhood obesity is increasing, and it strongly predisposes to hypertension that is one of the most severe public health problems [27,28]. The concomitant presence of obesity and hypertension is difficult to revert and are independent risk factors for cardiovascular morbidity and mortality later in life [29,30]. Therefore, early markers of hypertension in obese children are required in order to initiate personalized therapeutic and monitoring strategies. In this study, it was demonstrated that HMW levels and HMW/ HOMA-IR ratio can be useful to screen hypertension in obese children and adolescents. A limitation of this study is that it was not able to establish that low levels of HMW precede the onset of hypertension. Several suggestions for future research can be made based on our findings. Firstly, a cohort study that follows individuals over time could provide better insight into the temporal sequence of events, specifically whether low levels of biomarkers precede the development of hypertension, thus establishing a causal relationship. This would allow for more accurate prediction and prevention of hypertension in at-risk individuals. Secondly, increasing the sample size of the study would allow for more precise and reliable results. Finally, conducting similar studies in different populations would increase the generalizability and reproducibility of our findings, enabling us to better understand the underlying mechanisms of hypertension in diverse populations. By addressing these gaps in knowledge, we can further advance our understanding of hypertension and ultimately improve clinical care and patient outcomes.