Cardiometabolic Risk in Overweight and Obese Children in Bangladesh

Introduction: Childhood obesity is increasing dramatically and represents an important public health issue due to associated metabolic and cardiovascular co-morbidities. Very limited data are available regarding cardiometabolic risk factors among this group in Bangladesh. Objective: To observe the cardiometabolic risk factors in overweight and obese children. Methods: This cross-sectional study was carried out in 88 overweight and obese children recruited consecutively by using CDC percentile chart for body mass index (BMI) in children between January 2017 and March 2018 in the Department of Endocrinology, Bangabandhu Sheikh Mujib Medical University, Dhaka, Bangladesh. After completing a questionnaire and relevant clinical examination, blood was collected for fasting plasma glucose (FPG), insulin, HbA1c, lipid profile and C-reactive protein (CRP). Homeostasis Model Assessment of Insulin Resistance (HOMA-IR) was used to determine insulin resistance. Results: Central obesity (100%), dyslipidaemia (88.6%), raised CRP (81.8%) and metabolic syndrome (69.3%) were the most common cardiometabolic risk factors. Children with grade 3 obesity had significantly higher systolic blood pressure (115.57 ± 11.60 vs 105.71 ± 8.84 mmHg, p = 0.043) and insulin resistance (7.15 ± 4.97 vs 3.53 ± 2.04, p = 0.017) than grade 1 obesity. Blood pressure, insulin resistance and CRP increased while high density lipoprotein (HDL) decreased with increasing severity of obesity. BMI z score was a significant predictor of systolic blood pressure; waist circumference was an independent predictor of diastolic blood pressure and HDL; waist height ratio best predicted insulin resistance, CRP and total cholesterol in over-weight/obese children. Conclusions: We have observed a high frequency of cardiometabolic risk factors in overweight and obese children and they increased worsened with increasing grade of obesity.

tipsychotics), chronic infection or inflammation, diabetes and diseases or medications that alter blood pressure or lipid metabolism were excluded. Cardiometabolic risk factors (central obesity, hypertension, raised CRP, insulin resistance, impaired fasting glucose or diabetes mellitus, dyslipidemia) and presence of metabolic syndrome were determined from clinical examination and laboratory investigation.
The CDC age-and sex-specific growth chart was used to classify participants as overweight and obese. Overweight was defined as BMI at or greater than 85th to less than 95th percentile and obesity as BMI at or greater than 95th percentile for age and sex [12]. Obesity was further divided into 3 grades: Grade I obesity -BMI at or above 95th percentile to less than 120% of the 95th percentile, Grade II obesity -BMI at or above 120% to less than 140% of the 95th percentile, or BMI at or above 35 kg/m 2 and Grade III obesity -BMI at or above 140% of the 95th percentile, or BMI at or above 40 kg/m 2 [13]. Because the body-mass index varies according to age, we standardized the value for age and sex with the use of conversion to a z score from the website [14]. Central obesity was defined as waist circumference (WC) ≥ 90 percentile and/or waist height ratio (WHtR) ≥ 0.5 [15]. Hypertension was taken as systolic and/or a diastolic blood pressure ≥ 95th percentile for age, gender, and height according to the "Fourth Report on Diagnosis, Evaluation, and Treatment of High Blood Pressure in Children and Adolescents" [16] [17]. The cut point of raised CRP for increased cardiovascular risk was taken at 2 mg/l [18] [19]. Homeostasis model assessment of insulin resistance (HOMA-IR) value above 3 (corresponds to the 95th percentile healthy reference children) was regarded as presence of insulin resistance [20]. Impaired fasting glucose (IFG) was defined as fasting plasma glucose (FPG) levels between 5.6 and 6.9 mmol/l and diabetes when FPG ≥ 7 mmol/l [21]. Dyslipidemia was defined as at least one abnormal value for High Density Lipoprotein (HDL), Low Density Lipoprotein (LDL), total cholesterol (TC) or triglyceride (TG) [22]. Table 1 below shows abnormal cutoffs of individual blood lipids in children.
The definition for metabolic syndrome in children was taken from National Cholesterol Education program (NCEP) [15] in which children must have at least three of the given criteria: 1) Serum Triglyceride ≥ 110 mg/dL, 2) Serum HDL-C ≤ 40 mg/dL, 3) Fasting plasma glucose ≥ 100 mg/dL, 4) Waist circumference ≥ 90th percentile for age and gender and 5) systolic or diastolic blood pressure ≥ 90th percentile for age and sex [15].
Weight was measured using an electronic digital weighing machine to the nearest 0.1 kg, with the participant wearing light clothes and without shoes. Height was measured by a portable wall-mounted stadiometer to the nearest 0.1 cm with the participant without shoes in the erect position, back against the wall with his/her head held in Frankfurt horizontal plane with a right-angled triangle resting on the scalp and against the wall. WC was measured midway between the lowest rib and the superior border of the iliac crest by using a non-extensible and non-elastic measuring tape in mid respiration and inferences were drawn in percentiles WHtR was calculated by the formula WC in centimeters divided by Open Journal of Endocrine and Metabolic Diseases body height in centimeters [23]. Blood pressure was measured according to method described by the Seventh Report of the Joint National Committee [24]. It was measured three times by the same individual with aneroid sphygmomanometer (Yamasu) after calibration and standardization and mean value was recorded. Ten ml of venous blood was collected after a 12 hour fast for fasting glu-

Sample Size Estimation
Sample size (n) was determined by the formula used in cross sectional studies (n = Z 2 pq/d 2 ) with 95% confidence interval (value of standard normal distribution (Z) = 1.96) and 10% margin of error (d). The prevalence (p) of metabolic syndrome in childhood obesity was taken to be 0.307 [26]. Taking 10% drop out, sample size was calculated to be 88.

Statistical Analysis
All values were expressed as means ± SD or frequencies.

Results
In this study, 88 overweight and obese children with a mean age of 11.48 ± 2.72 years were enrolled. The male to female ratio was 1.4:1. Their mean BMI was 29.31 ± 5.11 kg/m 2 , with a range from 21.1 to 46.9 kg/m 2 . The mean BMI z score was 2.19 ± 0.36. Table 2 shows the frequency of overweight and different degree of obesity, where maximum number of children had grade 1 followed by grade 2 obesity.

Cardiometabolic Risks
The cardiometabolic risk factors of the study population are depicted in Table 3 and

Cardiometabolic Risk and Obesity
There was significant difference in WC and WHtR among overweight and different grades of obesity. There was significantly higher SBP (115.57 ± 11.60 vs 105.71 ± 8.84 mmHg, P 0.043) and insulin resistance (7.15 ± 4.97 vs 3.53 ± 2.04, P 0.017) in children with grade 3 compared to grade 1 obesity. Trend of values increased but was not statistically significant in case of diastolic blood pressure and FPG (Table 5). There was also a significant association between systolic hypertension and severity of obesity (P 0.015) ( Table 6).
The association between cardiometabolic risk factors and different measures of obesity were assessed using a multiple linear regression model, where covariates which were not statistically significant were removed from the model (Table 8). BMI z score was linearly related to systolic blood pressure. When BMI z score increased by one unit, systolic blood pressure increased by 0.446 units (β

Discussions
Childhood overweight and obesity have a strong association with different cardiometabolic risk factors. The frequency of risk factors and their association with severity of obesity were explored in this study. Majority of children in this study had grade 1 followed by grade 2 obesity. Central obesity, dyslipidaemia, raised CRP and metabolic syndrome were the most common cardiometabolic    risk factors. Systolic blood pressure and insulin resistance were significantly associated with grade of obesity. Blood pressure, insulin resistance and CRP increased while HDL decreased with increasing severity of obesity. Moreover, among measures of obesity, BMI z score was a significant predictor of systolic blood pressure while waist circumference was an independent predictor of diastolic blood pressure and HDL. Waist height ratio best predicted insulin resistance, CRP and total cholesterol in overweight/obese children.
Our findings suggest that metabolic syndrome is far more common (69.3%) among overweight and obese children and adolescents than previously reported.
The prevalence of the metabolic syndrome was 6.8% among overweight and 28.7% among obese adolescents in a study conducted from 1988 to 1994 in USA [27]. The rate was 38.7% in moderately obese and in 49.7% of severely obese North American children in 1999 [9]. Various studies among obese children and adolescents show that the prevalence of metabolic syndrome varies from 28.7% to 50% [27] [28]. However, meta-analysis of Tailor et al. shows that rate of me-Open Journal of Endocrine and Metabolic Diseases tabolic syndrome can be up to 60% in the overweight and obese which is not markedly different from the finding of our study [29]. Metabolic syndrome was identified in 36.6% of obese children and adolescent (6 -18 y) attending paediatric endocrine OPD of BIRDEM, Dhaka which is also much lower than our study population [30]. There are several definitions for metabolic syndrome in children. Therefore, the criteria used to define metabolic syndrome may influence its prevalence in this group. In addition, cut-off points used to define other cardiometabolic risk factors, ethnicity and eating behavior may also be contributing factors for the different prevalence of metabolic syndrome [9] [27] [28].
In our study, the high rate of metabolic syndrome may be due to the fact that all of the participants had central obesity, were mostly from urban area and were of high and middle socioeconomic condition. We found a high prevalence of dysli- Among all the cardiometabolic risk factors, only systolic blood pressure and insulin resistance were significantly associated with grade of obesity (classified using BMI). In other words, children with grade 3 obesity had higher blood pressure and abnormal glucose metabolism. This indicates that grade 3 obesity represents a higher risk group among obese children. Since all subjects had central obesity, participants could not be compared on the basis of waist circumference or waist height ratio.
Systolic blood pressure correlated with different measures of obesity (mostly WC). However, BMI z score was the main predictor of systolic blood pressure in overweight/obese children. Similar to this study, a study done in Australian children showed that BMI was the best predictor for systolic blood pressure, where blood pressure increased by 1.05 mmHg for every one unit increase of BMI. [34] Furthermore, in a multivariate analysis done by Moser et al. on 1441, 10 -16 year old Brazilian students only BMI was a predictor of high blood pressure (Odds ratio = 2.9), whereas WC and WHtR were not associated with risk of high blood pressure [35]. In contrast to our finding, WHtR was better than BMI for predicting hypertension [36]. On the other hand, a cross sectional study done in 1044 overweight and obese Italian children showed both WC and WHtR predicted high blood pressure in obese boys and girls [37]. WC was a significant predictor of both systolic and diastolic blood pressure in pre pubertal Chinese boys and girls [38].
HDL-C decreased with increasing WC, which was an independent predictor.
In accordance with our finding, a meta-analysis stated that WHtR, closely followed by WC was an indicator of dyslipidaemia [37].
This study showed that insulin resistance and CRP correlated most with WHtR. In addition, WHtR independently predicted development of insulin re-Open Journal of Endocrine and Metabolic Diseases sistance and CRP. In accordance with our study, Kondaki et al. showed greatest correlation of insulin resistance with WHtR [39]. On the other hand, a study from India observed strongest correlation of insulin resistance with WC [40]. Other studies also show significant increment of HOMA-IR with BMI, WC and WHtR mimicking our experience [41]. Our finding is in accordance with other studies which demonstrated that WHtR was a better anthropometric measure of obesity than WC and BMI for predicting cardiometabolic risk factors in adults and children [37] [42]. A study in pediatric population however stated that WC rather than BMI was the main predictor of cardiovascular disease [43].
Waist circumference and waist height ratio are measures of central obesity and correlate with intra-abdominal visceral fat, which is implicated in the pathogenesis of cardiometabolic disease [44]. BMI does not give us information about the distribution of fat and cannot differentiate between muscle and fat mass [45] [46]. Therefore, it is understandable and expected that insulin resistance, inflammatory markers and lipids correlate most with WHtR. However, SBP was related with BMI z score, a marker of generalized obesity. Therefore, according to this study, no single anthropometric measure can predict cardiometabolic risk in children.
A limitation of the study is that we could not measure insulin resistance with euglycaemic clamp. Future research on the cellular mechanism linking obesity and cardiovasvular risk can be undertaken to better understand the pathophysiology of these disorders.
There was a high rate of cardiometabolic risk factors in Bangladeshi overweight and obese children. Children with grade 3 obesity had worse cardiometabolic risk. Worsening of systolic blood pressure, insulin resistance, CRP and HDL were associated with increasing obesity. Although measures of central obesity best predicted insulin resistance, inflammation and lipids, no single anthropometric measure can predict cardiometabolic risk in children.