Gender-related differences in cardiometabolic risk factors and lifestyle behaviors in treatment-seeking adolescents with severe obesity

The participants’ lifestyle behaviors including physical activity, screen time, breakfast eating frequency and intake of sugar-sweetened soda, fruit and vegetables were reported using a self-administered food frequency questionnaire (FFQ) including questions about physical activity. The food and activity questionnaire included questions about habitual daily consumption of specified meals and 23 food items, daily physical activity outside school and daily screen time outside school (time in front of the TV- or PC screen). For each of the questions regarding screen time, physical activity level, breakfast eating, intake of fruit, vegetables and sugar-sweetened soda there were six to eight frequency categories to choose from as shown in Additional file 1: Table S1. The questionnaire was developed by the Department of Nutrition at the University of Oslo, Norway for use in a nationwide study of diet among fourth- and eighth-grade children (9- and 13 years old) in year 2000; the UNGKOST study [30]. The food frequency part of the questionnaire in the UNGKOST study has been validated against a four-day precoded food diary also used in the same study by Lillegaard and colleagues, and they demonstrated that the FFQ can be used to distinguish between low and high consumers of various food items [31]. Furthermore, they demonstrated that self-reported data from the FFQ were more consistent with data obtained from the food diary in the UNGKOST study considering intake of drinks, fruits and vegetables compared to infrequently eaten food items such as chocolate, sweets, savory snacks and pizza [31].

In the present analysis the highest frequencies of intake of fruits, vegetables and soda intake were rarely reported, therefore the number of original response categories were reduced and recoded into 3: low, moderate or high intake of fruits and berries and vegetables and low, moderate or high intake of sugar-sweetened soda. For simplicity, the response categories for physical activity level, screen time and breakfast eating were also reduced and recoded into 3: low, moderate or high physical activity level, low, moderate or high screen-time and skipping breakfast regularly, sometimes eating breakfast or regularly eating breakfast (Additional file 1: Table S1). Low physical activity, skipping breakfast regularly, high screen time, low intake of fruits and vegetables and/or high intake of sugar-sweetened soda were considered as unhealthy lifestyle behaviors and CVD risk factors.

Height was measured using Heightronic Digital Stadiometer® to the nearest 0.1 cm. Weight (kg) and body fat percentage was measured using Tanita bioimpedance body composition analyzer (BC-418, 8-polar, TANITA Corp., Tokyo, Japan). Waist circumference (WC) was measured using a standardized anthropometric tape, measuring the circumference at the midpoint between the top of the iliac crest and the lower part of the lateral rib cage to the nearest 0.1 cm. Waist to height ratio (WHtR) was calculated as WC (in cm) divided by height (in cm). BMI was converted to SDS by means of the current Norwegian growth references [32].

Systolic and diastolic blood pressure (SBP and DBP) were measured using a digital oscillimetric device, Dinamap ProCare (GE Healthcare). BP measurements were taken in the sitting position using appropriate sized cuffs (the mid-upper arm circumference was measured) four times on the dominant arm, and the average between the last three measurements was calculated.

Blood samples were obtained after an overnight fast by venipuncture in vacutainer gel tubes, and serum was separated from cells by centrifugation. Analyses of serum glucose and blood lipids were performed using dry reagent slide technology on the Vitros FS 5.1 (Ortho-Clinical Diagnostics, NY, USA). HbA1c was analyzed by high-performance liquid chromatography (HLPC) using Tosoh HLC-723 G7 (Tosoh Corporation, Tokyo, Japan). Insulin was analyzed by radioimmunossay (RIA) (Millipore Corporation, Billerica, MA, USA) until March 2012, thereafter the electrochemiluminescence immunoassay method (ECLIA) (ECLIA kit, Roche Diagnostics, Mannheim, Germany) was used. Intra- and inter-assay coefficient of variation were < 10% for all assays. Insulin resistance was estimated using the homeostasis model assessment for insulin resistance (HOMA-IR): (insulin (pmol/L) x fasting blood glucose (mmol/L)/135) [33].

Metabolic syndrome was identified using the definition from the International Diabetes Federation (IDF) Consensus Group Criteria: Abdominal (central) obesity (waist circumference ≥ 90th percentile for age and sex for ages < 16 years, and ≥ 94 cm for males and ≥80 cm for female for ages ≥ 16 years) and the presence of two or more of the following characteristics: 1) triglycerides ≥ 1.7 mmol/L, 2) HDL cholesterol < 1.03 mmol/L for age 10- < 16 years, and < 1.03 mmol/L in males and < 1.29 mmol/L in females for age ≥ 16 years, 3) SBP ≥ 130 mmHg or DBP ≥ 85 mmHg and 4) fasting plasma glucose ≥ 5.6 mmol/L or known type 2 diabetes [34]. Waist circumference was missing for 10 boys and 6 girls, and for these we assumed abdominal adiposity when BMI was above the International Obesity Task Force threshold for obesity [28].

All procedures in the study were performed in accordance with the ethical standards of the institutional and/or national research committee and with the Helsinki declaration and its later amendments or comparable ethical standards. The research protocol for the Vestfold Register of Obese Children has been approved by the Regional Committees for Medical and Health Research Ethics (S-08742c 2008/19081), Norwegian Centre for Research Data (NSD) and by the Norwegian Data Inspectorate (20,789 grh/rh). The research protocol for the current study has been approved by the Regional Committees for Medical and Health Research Ethics, REK sør-øst (2016/2039). Written informed consent was obtained from all participants and the parents/guardians of participants under the age of 16 in the study.

The cardiometabolic characteristics presented in Table 1 demonstrate that the mean waist circumference, triglyceride levels and SBP were significantly higher in boys compared with girls. In addition, boys had lower body fat percentage and HDL cholesterol levels than girls. The standardized mean differences between genders (Cohen’s d) are shown in Additional file 2: Table S2. Three of the participants (all boys) did not have central adiposity as defined by the IDF guidelines [34]. The prevalence of metabolic syndrome and components of the metabolic syndrome such as high DBP, dyslipidemia and dysglycemia did not differ significantly between genders (Table 2). However, the prevalence of high systolic blood pressure was higher in boys than in girls (19% vs. 9%, p = 0.021).

Gender was significantly associated with a number of lifestyle habits, as larger proportions of boys had higher screen time, more frequent breakfast eating, higher intake of sugar sweetened soda, and lower intake of vegetables than girls (Table 3). By contrast, physical activity level and intake of fruit and berries did not differ between boys and girls. The percentages of all adolescents reporting unhealthy lifestyle habits were 34% for low physical activity level, 55% for high screen time, 28% for not eating breakfast regularly, 25% for high intake of sugar-sweetened soda, and 18% for low intake of vegetables, fruits and berries. In a sub-analysis, those with unhealthy lifestyle behaviors (low physical activity, skipping breakfast regularly, high screen time, low intake of fruits and vegetables and/or high intake of sugar-sweetened soda) were compared with those reporting healthier lifestyle behaviors. As compared with girls, a larger proportion of boys reported high screen time (p = 0.010) and high intake of sugar-sweetened soda (p = 0.041). By contrast, the percentage of girls skipping breakfast regularly was higher than for boys (p = 0.037). There were no significant differences in the proportions of boys and girls with low physical activity (p = 0.436) or low intake of fruit (p = 0.223) and vegetables (p = 0.133). Figure 1 shows the prevalence of CVD risk factors (both cardiometabolic and lifestyle habits) according to gender.

The associations between lifestyle behaviors (physical activity level, breakfast eating frequency, frequency of intake of fruits, vegetables and sugar-sweetened soda) and cardiometabolic markers (total cholesterol, HDL-and LDL cholesterol, waist circumference, triglycerides, glucose, SBP and DBP) were significant only in terms of screen time and DBP, r = 0.17, p = 0.005). In the girl cohort HDL cholesterol was significantly related to breakfast intake frequency (r = 0.21, p = 0.020), whilst in boys DBP was related to screen time and fruit intake frequency (r = 0.26, p = 0.002, and r = − 0.22, p = 0.016, respectively). The mean levels of HDL cholesterol, LDL cholesterol, triglycerides, SBP and DBP did not differ significantly between those with unhealthy and healthier lifestyle behaviors (Tables 4 and 5). However, fasting blood sugar was significantly higher in those with high physical activity compared with those with low physical activity (Table 4).

One out of four boys (27%) and one out of five girls (19%) had metabolic syndrome. These figures are in accordance with those from an Italian cohort of adolescents with severe obesity (30 and 19%, respectively), but lower than demonstrated in a German cohort (46 and 37%, respectively) [12]. Although in the current study the average serum triglyceride levels were higher and HDL cholesterol levels lower in boys than in girls, the percentage with high triglycerides or low HDL cholesterol did not differ significantly between genders. Furthermore, high blood pressure was diagnosed in 19% of the boys, which was significantly higher than in the girls (9%). In two other cross-sectional studies prehypertension was present in 25% of adolescents with obesity and hypertension was present in 5%, but these studies did not assess gender differences [9, 10]. However, we did not find significant differences in cardiometabolic risk factors between adolescents with unhealthy or healthier lifestyle behaviors, except for blood glucose which was higher in those with high physical activity. Thus, we could not confirm our hypothesis that unhealthy lifestyle habits were associated with more unfavorable levels of cardiometabolic risk factors in our cohort of adolescents with severe obesity.

In the current study, 55, 34 and 22% of treatment seeking adolescents with severe obesity had high screen time, low physical activity and skipped breakfast regularly. These figures are considerably higher than those from a representative population of Norwegian 13-year olds taken from a national food and activity survey; the corresponding percentages were 18, 10 and 10% [30]. In the current study more boys than girls had high screen time as well as high intake of sugar-sweetened beverages. High screen time has been associated with high total cholesterol to HDL cholesterol ratio, high triglycerides and obesity in children [15, 17, 35]. In addition, Stamatakis et al. showed that screen time of 4 h or more per day in adults doubled the risk of CVD (HR:2.25 (95% CI:1.30–3.89)) [20]. A study by Shang et al. found a higher energy intake of 136 kcal and lower intake of fruits and vegetables of 0.2 servings in overweight children with high screen time in comparison with those with low screen time [36]. Data from the first National Health and Nutrition Examination Survey Epidemiologic Follow-up Study showed that adults who reported a high intake of fruits and vegetables had a 27% lower CVD mortality than those who had a low intake of fruits and vegetables [22]. Furthermore, sugar-sweetened beverages have been associated with weight gain, metabolic syndrome and type 2 diabetes [23, 37].

Furthermore, in the current study, a high proportion of boys and girls did not eat breakfast regularly, although the percentage of girls skipping breakfast regularly was higher than among boys. This is in accordance with findings of Boutelle et al. who suggest that girls may skip meals more often than boys as a method of losing weight [38]. Not eating breakfast regularly has been associated with increased waist circumference, high fasting insulin, high glucose and obesity in both children and adolescents [19, 39]. Skipping breakfast is associated with a poor diet quality, with low intake of whole-grain products and low micronutrient intake [18, 40].

One strength of this study is the relatively large cohort of treatment-seeking adolescents with severe obesity. Furthermore, in the Vestfold Register of Obese Children a wide range of cardiometabolic markers with clinical relevance have been registered. There are, however, limitations to our study which need to be addressed. The questionnaire was based on self-reported data about physical activity and dietary habits. In a population classified as obese, under-reporting of unhealthy choices and over-reporting of healthy choices can be expected [41].

The apparent lack of association between lifestyle habits and cardiometabolic risk factors might be explained by either the relatively low number of participants in the extreme categories (lowest, highest), or that severe adiposity has a larger impact on cardiometabolic risk factors than lifestyle behavior.