Abdominal obesity and low physical activity are associated with insulin resistance in overweight adolescents: a cross-sectional study

The number of adolescents who are overweight is increasing [1, 2], which is a public health problem. This status is associated with the globalisation and technological advances that affect lifestyle [3, 4]. In Colombia, the prevalence of overweight adolescents was 10.3% in 2005 [5], and this figure increased to 17.5% in 2010 [6]. The current study examines adolescents from the city of Medellin, where an overweight prevalence of 20.8% was reported in 2010 [7].

Obesity, specifically abdominal obesity, triggers insulin resistance (IR) because excessive free fatty acids and inflammatory substances alter insulin receptor signalling in different organs [8, 9]. Furthermore, IR causes the metabolic alterations that comprise metabolic syndrome (MS) [8–12]. The prevalence of MS increases with obesity [2, 13, 14].

MS has been detected in younger people at an increasing rate [10, 15, 16]. A recent study in Medellin that included 851 adolescents between 10 and 18 years old revealed rates of 25%, 4.1%, and 4.9% for overweight, MS, and IR, respectively [11]. The biological findings associated with this disease suggest that the β-pancreatic cells of these adolescents are forced to produce more insulin to maintain normoglycaemia, which predisposes them to hyperglycaemia and Diabetes Mellitus II (DM2) [12, 17, 18]. Early IR detection amongst obese adolescents might enable the application of preventive measures to decrease chronic disease development [19].

According to a recent publication, biochemical markers such as high-sensitivity C-reactive protein (hsCRP) and fasting insulin might enable early IR detection; the authors of this study suggested using hsCRP as an inflammation marker for risk stratification and treatment initiation amongst adults with moderate cardiovascular disease (CVD) risk [1]. However, this marker cannot be used for adolescents because little evidence supports the associations amongst hsCRP, the long-term risk of IR, and the development of chronic diseases during adulthood. Moreover, the authors suggested that daily hsCRP or insulin measurements are not feasible for the healthcare systems of developing countries because of the large number of adolescents [1]. Nevertheless, efforts are needed to discover anthropometric, clinical, or biochemical markers of IR for adolescents. These markers must be able to predict the progression of CVD during adulthood.

Waist circumference (WC) can be used to evaluate visceral fat, particularly amongst overweight adolescents. This measurement is correlated with obesity-related metabolic changes [20–22] and might serve as an early marker of chronic disease. However, the health institutions in Colombia do not account instruments to measure the WC and the health personnel does not register this information at the clinic history, that is why the health system just measures weight and height to classify overweight by body mass index (BMI).

Defining overweight as the consequence of a positive energy balance [23, 24], WC might not be sufficient for use as a unique predictor of chronic disease. Therefore, identifying the type of lifestyle associated with IR development amongst overweight adolescents is essential.

Many studies have characterized the metabolic and inflammatory alterations, as well as the lifestyles, of overweight adolescents [25–34]; however, not all overweight individuals are at the same risk of developing future disease. This study examined adolescents who have already developed IR because of their weight status and assessed the associations between IR and the anthropometric, metabolic, inflammatory, food consumption, and physical activity (PA) variables in overweight adolescents.

All adolescents in the O + IR group had a WC >75p. Significantly more participants (25%) had high WCs (>90p) in the O + IR group compared with the O-IR and NW groups (p = 0.00001). Overweight adolescents who also presented high WCs were 1.9 times more likely to develop IR (OR = 1.9, CIs = 1.3-2.7, p = 0.013). WC was correlated with HOMA (r = 0.67, p = 0.00001).

Inflammatory status, as assessed by hsCRP (Table 3), was significantly higher in the O + IR group than in the other groups (p = 0.020); this group also showed the highest percentage of adolescents with high cardiovascular risk (22.5%; hsCRP >3 mg/L). Furthermore, these participants were 1.6 times more likely to have inflammation (hsCRP ≥1 mg/L) than the O-IR group (OR = 1.6, CIs = 1.0-2.7, p = 0.028). hsCRP was significantly correlated with HOMA (r = 0.35, p = 0.0001).

TG and HDL-c significantly differed between the O + IR and other groups (p = 0.015 and p = 0.001, respectively). Approximately 60% of the adolescents in the O + IR group simultaneously presented low HDL-c and high TG values. HOMA was negatively correlated with HDL-c but positively correlated with TG (r = -0.31, p = 0.0010 and r = 0.45, p = 0.0001, respectively). The O + IR group was 2.1 times more likely to present low HDL-c (OR = 2.1, CIs = 1.4-3.3, p = 0.001) and 1.7 times more likely to present high TG (OR = 1.7, CIs = 1.1-2.7, p = 0.012) than the O-IR group.

The AI and CI were significantly higher for the O + IR group than for the other groups (p = 0.015 and p = 0.005, respectively; Table 3); both indices were significantly correlated with HOMA (AI: r = 0.36 p = 0.00001, CI: r = 0.32 p = 0.0001). The probability of presenting high AI and CI values was increased by 2 (OR = 2, CI = 1.4-2.8, p = 0.006) and 1.6 (OR = 1.6, CIs = 1.1-2.4, p = 0.026) for the O + IR group compared with the O-IR group.

The average caloric consumption was 2197 kcal/day, with a macronutrient distribution of 13% protein, 55% carbohydrates (CHO), and 32% fat. Of the total fat, 13% was saturated, 11% was monounsaturated, and 7% was polyunsaturated, with no differences between groups. Both overweight groups consumed more fast food; the O + IR group consumed the fewest fruits and vegetables.

The current study showed that overweight adolescents with IR differ from those without IR with regard to their bodily dimensions and lifestyles. Overweight adolescents with WCs > p90 and less PA (METs/day) are more likely to have IR.

The O + IR group had higher BMI, WC, and %BF values than the O-IR and NW groups (Table 2). These adolescents also presented higher HOMA index values (Me = 3.85), higher inflammatory statuses (hsCRP = 1.32), and more metabolic alterations (lower HDL-c, higher TG) and higher indices of arterial risk than those in the O-IR and NW groups. The above findings were statistically significant. Such findings confirmed that MS alterations are triggered amongst overweight adolescents who also present IR.

Not all overweight individuals are at the same risk of developing future disease. This subgroup has been called the “healthy obese” [1]. The current study did not find significant differences with regard to metabolic alterations, lipid profile (HDL-c and TG), glycaemia, insulin, the HOMA index, arterial indices, or inflammatory variables (hsCRP) between the O-IR and NW groups. These results suggest that the disease risk for these adolescents is not directly associated with BMI. The connection between overweight status and metabolic alterations seems to be in the development of IR due to excess visceral fat.

Within the overweight groups, IR development was directly associated with upper segment fat, based on WC and SFFs values. These values were significantly higher in the O + IR group than in the O-IR group; moreover, according to the multiple regression model, WC was the only marker that (together with sedentarism) explained 43.3% of the occurrence of IR amongst overweight adolescents. Kotlyarevska K et al. reported similar results with regard to 12- to 18-year-old adolescents. These results showed that BMI and WC were associated with IR as measured by the HOMA index [22]; the same results were previously reported for adults [20]. This association can be explained by the proinflammatory adipokines and the release of non-esterified (free) fatty acids from visceral fat to the portal system. These compounds induce lipotoxicity and insufficient phosphorylation (serine phosphorylation) in the insulin type 1 receptor (IRS-1) substrate present in adipocytes and myocytes [8, 59]. This effect results in insulin signalling alteration and IR, which in advanced states, leads to DM2 due to β-cells failure [12].

Proinflammatory adipokine production also triggers a mild and chronic inflammatory state that leads to CVD [8]. In the current study, the O + IR group was 1.6 times more likely to develop CVD risk than the O-IR group, as measured by hsCRP. In accordance with these results, Saíto E et al. found that an increase in the WCs of 10- to 13-year-olds was significantly associated with IR in both genders and with an increase in hsCRP amongst male adolescents [60]. Adiposity and IR increased the activation and aggregation of thrombocytes, promoted smooth muscle cell proliferation, increased adhesion molecule expression, and decreased nitric oxide bioavailability in the endothelium. All of these effects produce pro-atherosclerotic alterations in the arterial wall that affect cardiovascular health [61].

The O + IR group showed higher plasma TG and lower HDL-c concentrations than the other groups. Koike T et al. found similar results amongst young obese adults with IR [20]. These lipid alterations are related to an excess release of fatty acids from adipocytes and hepatic IR. Both situations increase TG synthesis [62]. At the same time, hepatic lipase (HL) and cholesterol ester transfer protein (CETP) increase, which increase HDL₂-c hydrolysis and the formation of much smaller HDL₃-c particles. The latter are excreted by the kidney, thereby decreasing their concentration [63]. Lipid profile, arterial, and CI alterations, as well as high hsCRP concentrations, in the O + IR group indicated a higher future risk of CVD amongst young people [64].

Findings concerning the lifestyle characteristics that affect IR occurrence and its complications were of great importance with regard to the objectives of the present study [3, 12]. Diet composition is undoubtedly related to health risk factors [26]. Our study showed that the groups had similar caloric intakes (2197 kcal/day); however, these intakes did not have the same effect on each group because of the differences in PA levels. This finding might explain the nutritional status differences observed amongst groups because the same caloric intake might be recommended for active adolescents (NW) but not for sedentary adolescents (O + IR) for whom the same caloric intake might represent an energy surplus.

Energy imbalances that lead to overweight individuals are directly correlated with insulin metabolism alterations [65]. Androutsos O et al. and Mirza N et al. evaluated lifestyle and IR associations in children younger than 15 years old in Greece and the United States. They showed that a higher consumption of sweetened beverages and more time watching television were positively associated with IR [25, 28]. The present study also found that adolescents with O + IR had significantly less PA (measured as METs) and they watched more television than the other groups (p = 0.043 and p = 0.035, respectively). In addition, an inverse relationship was found between the HOMA index and METs/day; furthermore, the overweight adolescents who watched TV ≥3 hours/day were 1.7 times more likely to develop IR.

Although adolescents with O-IR had significantly higher BMIs,%BF, and WCs than NW adolescents, they had similar PA (METs/day) levels and watched a similar number of hours of TV. However, they had more PA and watched less TV than O + IR adolescents. These findings suggest that PA protects adolescents from IR development, regardless of their overweight status. In accordance with this result, the regression analysis showed that for each MET, the HOMA index decreased by 0.026.

A recent meta-analysis [66] showed that exercise/training had a small-to-moderate effect on fasting insulin and improved IR in youth (Hedges effect size = 0.48 [95% CIs: 0.22–0.74], p = 0.001, and 0.31 [95% CIs: 0.06–0.56], p = 0.05, respectively). Our group performed another study showing that a PA intervention amongst 10- to 18-year-old adolescents, together with the daily consumption of 5 portions of fruits and vegetables, significant decreased WC, glycaemia, insulinaemia, and HOMA-IR [29]. In summary, these and other studies concluded that lifestyle changes including PA (regardless of its intensity, duration, and type) hamper IR progress in youth, primarily amongst those who are overweight and have adiposity, regardless of pubertal maturation and gender [67–74].

Healthy lifestyle habits should be favoured to prevent and treat these complications. PA models should be included in schools that meet the recommendations of 1 hour/day of MVPA to maintain active adolescents who are in good health [75]. At an individual level, adequate dietary habits and active lifestyles [65] should be adopted. If these habits are maintained from childhood through adulthood, then they might decrease organs damage and DM2 as well as CVD morbidity and mortality rates [75].

This study has certain methodological limitations. For example, although the 24-hr food consumption questionnaire properly reflects population food consumption, it does not enable the establishment of individual associations amongst biochemical variables due to the high intra-individual variability of this variable. On the other hand, PA was assessed in young people using the method 3DPAR, because the study did not have a better instrument like accelerometer. However, Cheryl B [74], in their validation study of the method PDPAR, demonstrated moderate correlations between this questionnaire and the accelerometer MTI, similar or higher to the correlations found with another self-report measures.

The results of the present study suggest that a WC > p90 and sedentary behaviour are associated with IR, lipid profile alterations, higher inflammatory status, and CVD risk amongst overweight adolescents. This two variables associated with HOMA, could be considered in the assessment of adolescents with overweight during the process of attention in health to detect those that required confirmatory laboratory tests of IR and to perform multidisciplinary interventions that modify the risk factors in adolescents with overweight, based on the eating habits and healthy lifestyles.

Future studies should also consider looking at a broad age range – from preschool through young adulthood - to determine how a proper screening including BMI, WC and PA evaluations might enable the early detection of features that promotes the development of chronic diseases like diabetes and CVD. This evaluation could have effects not only on the individual but on the health economy with the completion of paraclinical only young people who are at risk.

¹ ND, Mg Basic Biomedical Sciences. Professor, Universidad de Antioquia, Human Nutrition and Food Research Group Leader, Universidad de Antioquia, Medellin, Colombia

² ND Human Nutrition and Food Research Group, Universidad de Antioquia, Medellin, Colombia

³ ND Human Nutrition and Food Research Group, Universidad de Antioquia, Medellin, Colombia

⁴ ND, Mg Food Science and Human Nutrition. Human Nutrition and Food Research Group, Universidad de Antioquia, Medellin, Colombia