Prevalence of Overweight, Obesity and Metabolic Abnormalities Among 12-15 Year Age Group in an Urban City in Sri-Lanka

Objectives To determine the prevalence of childhood overweight, obesity and metabolic abnormalities among children aged 12-15 years within the schools in Kandy Municipality area, Sri Lanka. Design Cross- sectional observational study. Setting Methods The anthropometric measurements of 1766 school children were taken and those who were overweight or obese were recruited for further evaluation of metabolic abnormalities.


Introduction
Childhood obesity has been increasing world-wide over the past few decades [1]. A study on North-Indian urban children reported prevalence of overweight and obesity at 9.7% and 3.3% respectively [2]. According to a similar study conducted in Pakistan, 12% of school children were obese and 8% were overweight [3].
A study on primary school children of urban Nepal showed that 18.6% of children were overweight and 7.1% obese [4]. Prevalence of obesity and overweight among school-aged children in Bangladesh was respectively 3.5% and 9.5% [5].
A survey conducted in the Colombo district of Sri-Lanka during 2004 to 2005 showed that the prevalence of obesity among boys and girls aged 5-15 years were 5.7% and 6.4% respectively [6]. A cross sectional descriptive study from Negombo, revealed that 27% of girls and 19% of boys aged 10-15 years were obese [7]. Prevalence of overweight and obesity were respectively 9.4% and 5.5% among teenage-girls in Batticaloa which in addition showed a 21.6% prevalence of central obesity [8]. Collectively, literature acknowledged that the prevalence of childhood obesity has been increasing over the time and shows a regional variation [6,9].
The growing prevalence of obesity among Sri-Lankan children, particularly central obesity has led to the emergence of a constellation of metabolic derangements including dyslipidaemia, insulin resistance and non-alcoholic fatty liver disease (NAFLD) [8,10,11]. The link between central obesity, metabolic syndrome and type 2 diabetes is increasingly recognized in children [12]. The predominant dyslipidaemia pattern in childhood obesity, a combination of hypertriglyceridaemia and low high density lipoprotein cholesterol, together with central obesity, hyperglycaemia and hypertension represent a clustering of atherogenic risk factors described as metabolic syndrome [13]. The prevalence of metabolic syndrome in Colombo, Sri Lanka was 1.6% and 22.1% of them were obese [14]. NAFLD is a potential cause of chronic liver disease [15] and its prevalence was recently studied in Ragama Medical O cer of Health area reporting 8.4% of adolescents suffer with NAFLD [10].
Although some literature is available regarding the prevalence of childhood obesity in Sri-Lanka, there are no published studies in Kandy which is one of the highly populated urban cities in the country. Moreover, a very few authors have reported the prevalence of metabolic syndrome and NAFLD in childhood. Thus, our objective was not only to estimate the prevalence of overweight, obesity and central obesity among children aged 12-15 years attending the schools in Kandy Municipality area, but also to explore the prevalence of metabolic abnormalities and non-alcoholic fatty liver disease among overweight and obese.

Sampling
Multistage-cluster sampling method was used to recruit a total of 1766 school children aged 12-15 years.
Minimum sample size required for the study was calculated [16] considering the prevalence of obesity as 5% (precision 0.05, con dence-level 95%) and doubling the amount to minimize the clustering effect. The size of the sample was further in ated by 10% for contingencies. A 20% sample from type 1AB, 1C and type 2 schools in the Kandy municipality area was selected using the strati ed random sampling method proportionate to the population. All students from grade 8, 9 and 10 classes from the selected schools were listed separately for each grade where each class was a cluster. The cluster size was decided as 20 therefore a minimum of 50 clusters were required.

Study design:
This cross-sectional observational study was conducted in selected schools from November 2018 to June 2019. Body weight was measured to the nearest 0.1 kg using a pre-calibrated OMRON (HBF 510W) body composition monitor and height was quanti ed to the 0.1 cm using the GIMA portable stadiometer.
Waist circumference was measured with a exible measuring tape with the subject standing upright, feet slightly apart and abdomen relaxed. Waist circumference was taken at the narrowest point of the torso above the umbilicus and below the rib cage. Body fat percentage was measured by OMRON (HBF 510W) body composition monitor using the Bioelectrical Impedance Analysis technique.
Participants whose Body Mass Index (BMI) was above 85 th percentile for age and sex were recruited to evaluate for metabolic abnormalities [17,18]. Their blood samples were analysed using fully automated Biochemical analyser (HumaStar 150 SR-Human) for fasting lipid pro le, fasting blood sugar, and liver enzymes (Alanine Amino Transaminase and Aspartate Amino Transaminase). Real-time ultrasonography of abdomen was done utilizing Toshiba TUS-A 300 Ultrasound Machine with 3.5 MHz convex transducer.
Metabolic syndrome was diagnosed according to International Diabetes Federation criteria [19]. Overweight and obesity were diagnosed based on BMI cut-offs of World Health Organization (WHO) [18].
Waist Circumference cut-offs by British growth standards were used [20]. Percentages total body fat of 35% for girls and 25% for boys were considered as cut-off values for obesity related morbidity [21,22].

Data analysis:
Data was analysed using Microsoft Excel and SPSS version 20. Descriptive statistics and means with standard deviation were reported. Chi square test was used to determine the association of the presence of fatty liver with elevated liver enzymes.
Out of the 258 children who were overweight or obese, only 85 children reported for blood investigations. Prevalence of metabolic syndrome of 11.76% (5 girls and 5 boys) was found in this group of children.
Acceptable, borderline and abnormal levels in the fasting lipid pro les were de ned as per cut-offs provided by American College of Cardiology [23] and proportions of children in each category are shown in Table 02. A considerable number of children in our sample had high levels of cholesterol, triglycerides, low density lipoprotein and low levels of high density lipoprotein. Elevated Alanine Amino Transaminase and Aspartate Amino Transaminase levels were reported in 33 (38.82%) and 7 (8.24%) respectively from the 85 children who underwent blood investigations [24].
Only 49 children reported for ultrasound scan of abdomen. Grade 1 fatty liver was seen in 14 (28.57%) while grade 2 fatty liver was seen in 3 (6.12%). From these children, 19 (38.78%) had elevated Alanine Amino Transaminase levels. All who had grade 2 fatty liver reported to have elevated Alanine Amino Transaminase. There was a correlation between elevated Alanine Amino Transaminase and the presence of fatty liver (p=0.04).

Discussion
The study found that nearly one in seven school children (14.6%) aged 12-15 years in this urban Sri Lankan setting are either overweight or obese. The prevalence of overweight and obesity which were 7.81% and 6.8% respectively are in congruence with ndings from other parts of the country [8]. However, neighbouring countries like Pakistan and Nepal report a higher prevalence of overweight [3,4]. This contrast in the ndings could be attributed to the slight variation in the age range of the study populations.
The study revealed a prevalence of metabolic syndrome among overweight and obese children at 11.76% which is congruent with the ndings from other urban settings of the country [14]. The prevalence of metabolic syndrome among Sri-Lankan adolescents is signi cantly lower compared to obese adolescents in India [25].
Waist circumference above 90th percentile was found at a prevalence of 86.43% of the overweight and obese population surfacing the increased metabolic risk in this age group. The waist circumference was interpreted according to the British standards since country speci c nomograms are unavailable. Waist circumference could vary among diverse ethnic groups [26,27]. International Diabetes Federation, recommends using ethnic speci c waist circumference charts where available. The validity and reliability of the British standards for waist circumference have not been adequately examined for Sri Lankan children which could be a limitation of our study. Waist circumference cut-offs outlined in a cohort of urban Sri Lankan children, identi ed 43.88% of central obesity in our total study population while British waist circumference cut-offs detected only 16.93% [21]. Hence there is a potential for both cut-offs to either over estimate or under estimate central obesity in Sri Lankan children. Therefore, further studies from different areas of the country would be useful to determine waist circumference standards for Sri-Lankan children.
Hypertriglyceridaemia and low high density lipoprotein cholesterol levels in the abnormal range were found in one in three overweight children. However, fasting blood sugar was abnormal only in 9.41% and none were hypertensive. These ndings, while demonstrating a contrasting difference in the prevalence of different metabolic risk factors, highlights the abundance of central obesity with dyslipidaemia in this urban child population.
The lipid pattern in this overweight & obese cohort, when borderline and abnormal cut-offs taken together revealed adiposity related dyslipidaemia pattern (high triglycerides with low high density lipoprotein cholesterol) in more than two third. The prevalence of high cholesterol levels and high low density lipoprotein cholesterol levels were found in nearly 50% of this population thus surfacing the possibility of the existence of familial hyperlipidaemia in addition to adiposity related dyslipidaemia. Further studies are needed to identify the lipid pattern in Sri Lankan child population. Even though NAFLD was detected in 34.69% it is di cult to derive a signi cance due to under reporting for ultra sound scan procedure which could be attributed to lack of awareness among general public regarding obesity and its implications.

Conclusion
Overweight and obesity prevails at 14.6% with metabolic syndrome at 11.67% among the children aged 12-15 years in this urban area which are comparable with the prevalence rates reported from similar urban settings in the island. Presence of central obesity in 5.04% among children with normal BMI implies that the prevailing BMI cut-offs may under estimate adiposity in Asian ethnicities. Presence of elevated low density lipoprotein cholesterol in 27.06% is an eye-opener as to the existence of familial hyperlipidaemia in addition to adiposity related dyslipidaemia.
Findings of this study contribute to the national data base of the non-communicable disease prevention programme and assimilation of data from different parts of the country would prompt the concerned stakeholders to develop a strategic preventive programme. Further studies are recommended to determine the prevalence of central obesity, NAFLD and the dyslipidaemia pattern among Sri-Lankan children.