High prevalence of undiagnosed comorbidities among adolescents with obesity

Metabolic diseases are increasing among adolescents with obesity. Although the reported prevalence of metabolic syndrome is approximately 30% worldwide, its prevalence is largely unknown among New Zealand adolescents. Therefore, we assessed the health of adolescents with obesity (BMI ≥ 30 kg/m2) enrolled in a randomised clinical trial (Gut Bugs Trial), to identify the prevalence of undiagnosed comorbidities. Assessments included anthropometry, 24-h ambulatory blood pressure monitoring, and insulin sensitivity. We report on baseline data (pre-randomisation) on 87 participants (14–18 years; 59% females), with mean BMI 36.9 ± 5.3 kg/m2 (BMI SDS 3.33 ± 0.79). Approximately 40% of participants had undiagnosed metabolic syndrome, which was twice as common among males. Half (53%) had pre-diabetes and 92% a reduction in insulin sensitivity. Moreover, 31% had pre-hypertension/hypertension, 69% dyslipidaemia, and 25% abnormal liver function. Participants with class III obesity had a greater risk of metabolic syndrome than those with classes I/II [relative risk 1.99 (95% CI 1.19, 3.34)]. Risks for pre-hypertension/hypertension and inflammation were also greater among those with class III obesity. We identified a high prevalence of undiagnosed comorbidities among adolescents with obesity in New Zealand. As adolescent obesity tracks into adulthood, early interventions are needed to prevent progression to overt cardiometabolic diseases.

www.nature.com/scientificreports/ normal changes observed in puberty can be magnified, leading to greater weight gain and metabolic dysfunction, including persistence of insulin resistance 14 . As 90% of adolescents with obesity continue to have obesity as adults 15 , early intervention is crucial.
Consistent with the rise in obesity, cardiometabolic comorbidities such as metabolic syndrome and type 2 diabetes mellitus (T2DM) are increasing in children and adolescents 16 . Increasing body mass index (BMI) is associated with an increased risk of metabolic syndrome 17 , which includes increased abdominal obesity, hypertension, impaired fasting glycaemia, dyslipidaemia, and is associated with insulin resistance 16 . The prevalence of metabolic syndrome among adolescents with obesity has been reported to be as high as 60% 18 , and it is associated with the development of T2DM 19 , cardiovascular diseases 20 , and a two-fold increase in the risk of coronary artery disease and stroke, and a 1.5-fold increase in the risk of all-cause mortality 21 . In the US, a national cross-sectional study reported that T2DM is increasingly diagnosed among adolescents and accounted for 40% of adolescent diabetes, with more than a third of T2DM cases undiagnosed prior to the study 22 . In New Zealand, the number of children with T2DM is increasing at approximately 5% per year, and this disease disproportionally affects highrisk ethnic groups (Māori and Pacific) 23 . Apart from serious cardiometabolic complications, paediatric obesity has been associated with increased mortality even in early adulthood 24 . This is likely due to increased systemic inflammation, insulin resistance, impaired cardiovascular function, and the development of non-alcoholic fatty liver disease 25,26 . Moreover, many of these children face bullying 27 and social isolation 28 , as well as increased rates of depression 29 , suicide and self-harm 24 .
Overall, there are limited data on the prevalence of obesity-related comorbidities among adolescents with obesity in New Zealand 30,31 . Due to the numerous complications associated with obesity, early identification particularly in high-risk populations is necessary so that targeted interventions can be implemented. Therefore, we aimed to assess the metabolic health of a group of adolescents with obesity enrolled in a clinical trial and identify the prevalence of undiagnosed metabolic syndrome and other obesity-related cardiometabolic comorbidities.

Methods
Ethics. This study reported on baseline data (pre-randomisation) from a randomised placebo-controlled trial (Gut Bugs Trial) to evaluate the effectiveness of faecal microbiome transfer for treatment of adolescent obesity in Auckland, New Zealand 32 . The trial was registered with the Australian New Zealand Clinical Trials Registry (ACTRN12615001351505); ethics approval was granted by the Northern A Health and Disability Ethics Committee (16/NTA/172). Participants provided verbal and written informed consents. All procedures in this study were conducted according to the ethical principles and guidelines laid down in the Declaration of Helsinki 33 .
Recruitment. Participants were recruited from social media through Facebook advertisements between 2017-2018. All were post-pubertal, aged 14-18 years, with BMI ≥ 30 kg/m 2 , who were not diagnosed with diabetes or chronic diseases that could affect weight or metabolism 32 .
Clinical assessments. Clinical assessments included medical and physical examinations previously described in the trial's protocol 32 , and briefly summarised here.
Height, weight, and waist and hip circumferences were measured 32 . BMI values were converted into standard deviation score (SDS) using WHO standards 34 . For comparison within our study population, BMI was stratified using the adult criteria for obesity: Class I (BMI ≥ 30 but < 35 kg/m 2 ); Class II (≥ 35 but < 40 kg/m 2 ); and Class III (≥ 40 kg/m 2 ) 35 . Body composition was assessed using whole-body dual-energy X-ray absorptiometry (DXA; Lunar Prodigy and Lunar iDXA; GE Medical Systems, Chicago, Illinois, USA).
Participants underwent a 75-g oral glucose tolerance test (OGTT) 32 . Insulin sensitivity was assessed by homeostatic model assessment of insulin resistance (HOMA-IR) 36 and Matsuda index 37 , as previously described 32 .
Other key markers of glucose metabolism measured were fasting insulin and fasting glucose, 2-h glucose, and glycated haemoglobin (HbA1c). From fasting blood samples, uric acid, high-sensitivity C-reactive protein (hsCRP), lipid profile, and liver function were measured 32 .
Health outcomes in this study were cardiometabolic comorbidities as defined in Table 1
In addition, 13% of participants had pre-hypertension and 8% had hypertension from clinic BP. From 24hABPM data, 17% were pre-hypertensive and 14% hypertensive with nocturnal pre-hypertension recorded in 18% and nocturnal hypertension in 10% of participants (Table 3).
Pre-diabetes was common, affecting approximately half of participants (52%): 29% females and 83% males (p < 0.0001). Fasting insulin was elevated in 94% of participants including all the males (Table 3). Most participants displayed a reduction in insulin sensitivity, as 92% had a high HOMA-IR when compared to a cohort of adolescents 39 , and 87% had a low Matsuda index when compared to healthy adults 40 (Table 4).

BMI classes.
There were marked differences in the prevalence of cardiometabolic comorbidities between obesity classes ( Table 4). The risk of metabolic syndrome increased among those with class III obesity compared to those with a lesser degree of obesity [RR 1.99 (95% CI 1.19, 3.34); p = 0.0091] ( Table 4). The prevalence of BP abnormalities was markedly higher in participants with class III obesity, with the relative risks of prehypertension/hypertension and loss of the nocturnal dipping BP status more than 3 times greater in this group (Table 4). A higher BMI was associated with an increased likelihood of inflammation, with the relative risk of hyperuricaemia and elevated hsCRP being 1.4 and 2.1 times greater in participants with class III obesity, respectively (Table 4).

Discussion
We identified a high prevalence of undiagnosed comorbidities amongst our cohort of adolescents with obesity. Notably, more than a third were diagnosed with metabolic syndrome, which was twice as common in males than in females. More than half (52%) of our cohort had pre-diabetes and more than 90% had fasting hyperinsulinaemia, with higher rates of these complications in males. In addition, almost all had a reduction in insulin sensitivity. Moreover, increased levels of adiposity were associated with a higher risk of metabolic syndrome, hypertension, and inflammation. The presence of these adverse cardiometabolic outcomes at a relatively young age is alarming, and along with published data documenting the tracking of weight-related comorbidities from childhood into adulthood 41 , further reaffirms that obesity in adolescence is far from a benign condition.
A description of comorbidities among 239 children and adolescents with obesity in New Zealand was provided by Anderson et al. in 2016 30 . In that study, 1 in 10 had elevated blood pressure, 1 in 4 had increased inflammation, and nearly half had dyslipidaemia and abnormal liver function 30 . While their reported prevalence of obesity-related comorbidities were relatively high, they were lower than those observed in the present study 30 , probably because their study population was younger (mean age 10.7 vs 17.2 years in our study), leaner (mean BMI 3.09 vs 3.33 SDS), and had a different ethnic make-up with a much lower representation from those of Pacific descent (3% vs 23%) than ours.
Worldwide, the reported prevalence of metabolic syndrome among children and adolescents with obesity varied between 10 to 66% 18,31,42-50 . In New Zealand, Grant et al. reported a lower rate of metabolic syndrome among 29 Pacific adolescents with obesity aged 15-18 years 31 -21% vs 36% in our study. In comparison, reported rates of metabolic syndrome in adolescents with obesity vary widely across the world: 15% to 50% in the US 18,50 , 23% to 60% in Latin America 18,43,46,49 , 12% to 42% in Asia 18 , and 14% to 44% in Europe 18,44,45,47 . The marked differences in prevalence among these studies could be attributed to variations in age distribution and ethnic composition of the respective study populations, as well as the definitions of metabolic syndrome used. Nonetheless, the findings from two systematic reviews clearly show increasing BMI as an important risk factor associated with the development of metabolic syndrome 18,42 , with this relationship also shown to occur at the upper end of the BMI spectrum by our stratified analyses. www.nature.com/scientificreports/ Reduction in insulin sensitivity as well as impaired glucose metabolism were common complications among our study population. Insulin resistance as assessed from the HOMA-IR values among our adolescents was more than 1.5 times higher when compared with adolescents with obesity in the US 51 and Europe 52 . In addition, more Table 2. Demographic and clinical characteristics of participants enrolled into the Gut Bugs Trial. Age data are median [quartile 1, quartile 3]; other data are n (%) or means ± SD, as appropriate. BMI body mass index, HOMA-IR homeostatic model assessment of insulin resistance, SDS standard deviation score. Obesity classes were defined as: Class I (BMI ≥ 30 kg/m 2 but < 35 kg/m 2 ); Class II (BMI ≥ 35 kg/m 2 but < 40 kg/m 2 ); and Class III (BMI ≥ 40 kg/m 2 ). a Higher maternal/paternal education status refer to university degree or post-high-school vocational qualification. b Socioeconomic deprivation was estimated using the New Zealand Indices of Multiple Deprivation 73 . www.nature.com/scientificreports/ than half of our participants had pre-diabetes (i.e. impaired fasting glycaemia, impaired glucose tolerance, and/or elevated glycated haemoglobin). It could be argued that our high rate of pre-diabetes could be attributed, at least in part, to our lower cut-off value for impaired fasting glycaemia (i.e. ≥ 5.6 mmol/L as recommended by the ADA 53 and ISPAD 54 , rather than the WHO value ≥ 6.1 mmol/L 55 ), as using the higher WHO cut-off, our pre-diabetes rate would have dropped from 52 to 38%. Nonetheless, when compared to previous studies in US and Europe that used the same cut-off values as ours, the prevalence of pre-diabetes in our study was still 4 times greater 52,56 . Moreover, due to the high risk of diabetes in our vulnerable study population and our aim to prevent worsening of their metabolic health through early identification and intervention, we contend that the lower threshold for abnormal fasting glycaemia was justified. Nichols et al. reported that without appropriate intervention, nearly one in ten adults with pre-diabetes will develop T2DM within 3.5 years, and the progression to T2DM could be accelerated by risk factors such as increased BMI, elevated blood pressure and triglyceride levels, and lower HDL levels, all of which were present in our participants 57 . As improvement in insulin sensitivity and reversal of pre-diabetes have been reported with therapeutic interventions 58 , early identification of pre-diabetes among adolescents with obesity becomes increasingly important. Although small, our study population was likely representative of Auckland's ethnic and socioeconomic make-up, with relatively similar demographics when compared to national census data 59 . Both ethnicity and socioeconomic status are factors known to be associated with an increased risk of obesity and obesity-related diseases 60 . As we were able to recruit adolescents with obesity but not with any pre-diagnosed chronic conditions Table 3. Baseline cardiometabolic comorbidities of adolescents with obesity enrolled into the Gut Bugs Trial. Data are n (%). a For the full definitions of all comorbidities please refer Table 1. b All diastolic non-dippers were also systolic non-dippers. 24hABPM 24-h ambulatory blood pressure monitoring, ALT alanine transaminase, AST aspartate transaminase, BP blood pressure, GGT gamma-glutamyl transferase, HbA1c haemoglobin A1c, HDL high-density lipoprotein cholesterol, HOMA-IR homeostatic model assessment of insulin resistance, hsCRP high-sensitivity C-reactive protein, LDL low-density lipoprotein cholesterol, OGTT oral glucosetolerance test. www.nature.com/scientificreports/ from the general population, our findings may be extrapolated to describe the health of adolescents with obesity in Auckland. A strength of our study was our robust clinical assessments. In particular, accurate measurement of clinic BP is challenging, with wide variations due to many environmental factors 61 . 24hABPM, is a far more robust method to identify BP abnormalities compared to commonly used clinic devices 62 . Notably, pre-hypertension/hypertension was underdiagnosed when measured using the clinic BP monitor; only one in five was diagnosed to have elevated BP whereas with a 24hABPM, more than a third were reported to have elevated BP. Moreover, nearly a third of participants were diagnosed to have nocturnal prehypertension/hypertension which would have been undetected during daytime clinic BP measurements, and further emphasized the importance of undertaking BP monitoring over a 24-h period. Participants also underwent an OGTT which provided a more comprehensive assessment of glucose homeostasis and insulin sensitivity 53 .
In conclusion, we identified a high prevalence of undiagnosed comorbidities among adolescents with obesity. Of note, the high prevalence of metabolic syndrome in our study population emphasises the importance of screening adolescents with obesity for these metabolic complications. Obesity is a complex chronic condition that once established is not only difficult to treat, but requires life-long support 41 . As a result, it is undeniable that prevention of obesity should be the primary focus in this health crisis. However, for adolescents with established obesity, early identification of individuals with poor metabolic health and implementation of early targeted interventions are important, with the aim of preventing the development of overt cardiometabolic disease.

Data availability
The clinical data cannot be made available in a public repository according to the strict conditions of the study's ethics approval. Nonetheless, anonymized and de-identified data could be made available to other investigators upon bona fide request, and following all the necessary approvals (including ethics) of the detailed study proposal and statistical analyses plan. Any queries should be directed to Prof Wayne Cutfield (w.cutfield@auckland.ac.nz).
Received: 13 July 2020; Accepted: 7 October 2020 Table 4. Relative risks of comorbidities among participants according to their obesity class. Data are n (%), or relative risks (adjusted for sex) and respective 95% confidence intervals. P-values for statistically significant differences are shown in bold. Obesity classes were defined as: Class I (BMI ≥ 30 but < 35 kg/m 2 ); Class II (≥ 35 but < 40 kg/m 2 ); and Class III (≥ 40 kg/m 2 ). All blood pressure parameters were derived from 24-h ambulatory blood pressure monitoring. hsCRP high-sensitivity C-reactive protein. a For the full definitions of comorbidities please refer to Table 1. www.nature.com/scientificreports/