Obesity is increasing all over the world both in developed and underdeveloped countries [1, 2] and is known to be associated with inflammatory changes, insulin resistance and with a hyperlipidemic state [3]. The adipocyte is an important source of cytokines, namely plasminogen activator inhibitor type 1 (PAI)-1, leptin, interleukin (IL)-6 and tumor necrosis factor (TNF)-alpha, and their levels are significantly higher in the plasma of obese patients [4]. Moreover, some inflammatory markers, such as C-reactive protein (CRP), show an independent positive association with body mass index (BMI) [5].

Cardiovascular (CV) morbidity and mortality is associated with the classic risk factors, namely dyslipidemia, hypertension and impaired glucose metabolism. However, more recently, inflammatory mechanisms were recognized as playing a vital role in initiation, maintenance and progression of atherosclerotic vascular disease. Atherosclerosis is an inflammatory disease with the participation of leukocytes and several cytokines. Among several inflammatory markers that have been shown to predict CV events [namely IL-6, tumor necrosis factor-alpha and soluble intercellular adhesion molecule (ICAM)-1] CRP levels have emerged as the most powerful marker [6].

Regular physical activity is known to reduce obesity-related markers and risk of type 2 diabetes (physical activity is associated with improved insulin sensitivity) and to improve endothelial function, lipid profile (e.g., exercise decreases plasma triglycerides (TG) and increases high density lipoprotein (HDL), as well as several oxidative stress/inflammatory markers, all risk factors for CV disease (CVD) [7-10]. Diet restriction per se may reduce fat mass, total cholesterol (Chol) and low density lipoprotein cholesterol (LDLc) without apparent effect in HDL and limited effect on insulin resistance. The effect of both diet and physical exercise in obese individuals appear to be additive with greater impact in the improvement of lipid profile and insulin resistance [11]. However, few data exists in literature regarding the long-term effects of physical activity on young obese subjects. Moreover, in most studies, children obey to a restricted physical exercise program or a food intake program, but the effect of “natural” lifestyle modifications on CV risk factors has been poorly explored [12-14].

Higher cardiorespiratory fitness (CRF) has been associated to a decrease in CVD risk markers, CVD morbidity and mortality [15]. Besides physical activity, the initial CRF of an individual appears to be an independent predictor of weight variation. Individuals with lower CRF are more prone to gain weight when compared to those with higher CRF [16]. Actually, children with lower CRF usually present diminished levels of physical activity and increased time expended in sedentary behaviours. Despite the strong genetic influence, CRF status can be modulated through physical activity, which is a main determinant of CRF in pediatric populations [17].

The physical activity performed by children and adolescents in southern European countries is lower than in countries from centre and northern Europe. This fact may explain the geographical differences in obese prevalence found in Europe [17]. In fact, recent studies showed a very high prevalence of overweight/obesity (about 30%) in Portuguese children when compared to other European countries [18, 19]. Thus, the study of obesity is of particular importance in our country.

The aim of this study was to evaluate the short-term effects of lifestyle modifications, through a decrease in body mass index z-score (BMI z-sc), on CV risk factors (circulating levels of C-reactive protein, glucose, insulin and lipid profile), in Portuguese obese children and adolescents.

The characteristics of the participants in the cross-sectional study are presented in Table 1. Obese patients and controls were matched for age and gender. Compared with the lean children, obese patients demonstrated significantly higher insulin resistance index (HOMA), and TG, LDLc, apo B, insulin and CRP concentrations, whereas their HDLc and apo A levels were significantly lower (Table 2). The participants in the cross-sectional study had BMI which correlated positively and significantly to TG (r = 0.300, P < 0.001), LDLc (r = 0.235, P = 0.001), insulin (r = 0.652, P < 0.001), HOMA (r = 0.646, P < 0.001), and CRP (r = 0.451, P < 0.001) values, and negatively with HDLc (r = -0.462, P < 0.001). Adiponectin levels (n=153) correlated inversely and significantly with age (r = -0.299, P < 0.001), BMI (r = -0.220, P = 0.006), TG (r = -0.390, P < 0.001) and total cholesterol/HDLc (Chol/HDLc) ratio (r = -0.270, P = 0.001), insulin (r = -0.211, P = 0.009) and HOMA (r = -0.223, P = 0.006), and correlated positively and significantly with apo A (r = 0.214, P = 0.008) and HDLc (r = 0.270, P = 0.001).

Sixty obese patients participated in the longitudinal study and, after one year of follow up, 17 (28.3%) achieved a substantial BMI z-sc reduction (mean ΔBMI z-sc = -0.46 ± 0.16), while 43 individuals did not reach the cut-off value (mean ΔBMI z-sc = 0.01 ± 0.26) (Table 3). At baseline (T1), no significant differences were observed in age, sex, BMI, BMI z-sc, and in any markers of the metabolic syndrome or inflammatory markers between obese children who presented with and without substantial BMI z-sc reduction, at the end of the study (T2).

A substantial decrease in BMI z-sc in 17 children led to a significant reduction in TG, cholesterol, LDLc, apo B, glucose, insulin levels and in HOMA (Table 4). In the 43 children without change of BMI z-sc, a reduction in adiponectin and in glucose values occurred as well as a significantly increase in Lp(a) values (Table 4).

The ΔBMI values over the studied period (longitudinal study, n=60) correlated inversely and significantly with BMI at baseline (r = -0.487, P<0.001, Fig. 1A), and HOMA at baseline (r = -0.291, P=0.026, Fig. 1B). In multiple linear regression analysis, BMI values at baseline remained associated to changes in BMI over the studied period (standardised Beta: -0.271, P=0.05).

Fig. (1). Associations between change in body mass index (BMI) over the studied period (one year of follow-up) and BMI and insulin resistance index (Homeostasis model assessment; HOMA) values presented by participants at baseline. Spearman’s rank correlation coefficient values are shown.

The present study was able to demonstrate in young obese patients that small reductions in BMI z-sc are associated with significant improvements in lipid profile and in insulin resistance index.

In a first analysis (cross-sectional study) we compared obese patients with a lean control group. As some of the studied CV risk factors change with age and are influenced by gender, we matched controls and patients for gender and age. The results obtained confirm obesity as an inflammatory disease. Indeed, the median CRP value observed in obese children and adolescents was about 6 times higher than that observed in controls, demonstrating an inflammatory process in such young patients (Table 2). We also observed that obese patients presented with an altered lipid profile and raised markers of insulin resistance when compared with controls (Table 2).

There are few studies assessing the long-term changes of lifestyle modifications in obese children and adolescents, particularly on CV risk factors. In the present longitudinal study, obese patients (children and adolescents) were asked to improve their lifestyle habits (conventional weight loss programs based on dietary counselling and encouragement to exercise) and an objective measure (BMI z-sc reduction) was used to observe which children achieved the objective after one year of follow-up. We observed that 17 patients (28.3%) presented a significant reduction in BMI z-sc (0.3 or more (ΔBMI ≤ -0.3)) over the studied period and that changes in BMI over the studied period were independently associated with BMI values presented at baseline. It seems that patients with higher BMI experience a greater relative reduction in their BMI (Fig. 1). Also, the correlation between HOMA in the beginning of the study and relative BMI reduction points to a predisposition for children who are more insulin resistant to lose more weight. These associations can be related to a change in body physiology towards weight loss.

In our longitudinal study, children did not obey to a restricted physical exercise program or a food intake programme. Also, it was not controlled the initial CRF of our population. Thus, it is impossible to know the possible impact of CRF in BMI z-sc reduction. It is true that the lack of standardisation on plans on intervention may mitigate the interpretation of the results that we observed. On the other hand, the counseling to improve lifestyle habits may be a “more natural” approach, and is useful in clinical practice, as participants can directly observe the real benefits of changing their lifestyle habits. Furthermore, the reduction in BMI-zc that we observed is likely to express a real change in lifestyle habits.

It was shown that 3 months of moderate lifestyle intervention (45 min of physical activity 3 times per week) in obese adolescents seems to attenuate the inflammatory state associated with obesity, as observed by a reduction in elevated circulating concentrations of CRP, fibrinogen and IL-6, non-traditional risk factors for CVD [12, 14]. Reinehr et al., by evaluating 16 [14] and 10 [13] obese children who lost weight over a 1-year period, found a significant decrease in CRP but no significant changes in TNF-alpha levels. In the studies performed by Reinehr et al., obese children participated in a 1-year obesity intervention program and the cut-off value used for BMI z-sc reduction was 0.5. As we used a smaller cut-off value (0.3), this may explain why we did not observe significant changes in CRP levels. Indeed, CRP shows a positive association with BMI, and our cut-off value may be insufficient to observe clear improvements in CRP. Even thus, it is important to emphasize that in our study, involving a larger number of participants, 64.3% of the patients that reduced their BMI z-sc presented a decrease in circulating CRP levels, in contrast with the group without change of BMI z-sc (47.6%). Furthermore, we were able to demonstrate that smaller reductions in BMI z-sc are associated with significant improvements in lipid profile and in insulin resistance index HOMA. Previous studies suggested that the improvement of insulin sensitivity and CV risk factors in obese children only occurs if BMI z-sc decreases by at least 0.5 over a 1-year period [21], but our data suggests that such improvements may be observable with smaller reductions in BMI-zc. Again, our results are probably explained by the fact that the reductions that we observed in BMI z-sc are likely to express a continuous daily motivation in changing lifestyle habits. Nevertheless, although there was no change in HDLc, the lower values of HOMA together with the reduction in TG seen in the group that reached the cut-off values leads to the probable influence of concomitant changes in diet and in the levels of physical exercise in this group to be linked with the improvement seen in the global CV risk profile.

The improvement in insulin sensitivity with BMI z-sc reduction observed in our study is of particular importance, as insulin resistance represents a major underlying abnormality driving CVD [22]. It is important to emphasize that insulin resistance associates with increased TG and that one of the most striking findings in the longitudinal analysis of the present study was the reduction in TG levels. Although hypercholesterolemia is a major risk factor for the initiation and progression of atherosclerosis [23], hypertriglyceridemia and elevated levels of triglyceride (TG)-rich lipoproteins have recently re-emerged as risk factors for atherosclerosis [24]. TG-rich particles may directly damage the endothelium principally via oxidative mechanisms. Also, an important consequence of hypertriglyceridemia, relating to increased atherosclerotic risk, is a shift in the spectrum of LDL subfractions towards smaller, denser species, which are believed to be more atherogenic [25]. Moreover, raised TG levels are frequently associated with reduced HDL levels. In our study, we only observed a slight (non-significant) increase in HDLc levels, which seems to have a protective role in the development of atherosclerosis [26-28]. Reinehr et al. observed a trend to increase HDLc levels in patients who achieved a substantial BMI z-sc reduction [14].

Plasma Lp(a) levels are known to be mainly genetically determined [29, 30]. However, it appears that plasma Lp(a) concentration increases with age. In our study, Lp(a) levels were similar betweens controls and obese individuals but, in the longitudinal study an increase in this lipoprotein was observed (Table 4). In fact, this increase achieved statistically significance in the group without a substantial reduction in BMI z-sc. Thus, our data is in agreement with the idea that ageing elevates plasma Lp(a) concentration, and that this increment might be modulated be lifestyle modifications.

Adiponectin is an adipocytokine that has been noted as an important antiantherogenic, antidiabetic and as an anti-inflammatory protein [31]. Adiponectin seems to be inversely related to systolic blood pressure, waist circumference, triglycerides, and 2-hour glucose levels, while powsitively related to HDL [32]. These associations are in agreement with the results obtained in our cross-sectional study. However, and although it was suggested that adolescent obesity is associated with low plasma adiponectin levels [33], we were unable to find statistically significant differences in plasma levels of adiponectin between obese children and controls. This may be explained by the fact that we defined controls as having a BMI lower than the 85^th percentile adjusted for sex and age, in contrast with other work where control group was defined as a group of adolescents who had a BMI between the 50^th and 75^th percentiles [33].

Regarding our longitudinal study, it is important to highlight that adiponectin levels decreased significantly in the group without BMI z-sc reduction, in contrast with a moderate (non-significant) increase in the group that presented a significant decrease in BMI z-sc. It was previously reported that a significant weight loss in children, defined as BMI z-sc reduction of 0.5 or more, is associated with a significant increase in adiponectin [34]. Considering the protective effects of adiponectin, it seems reasonable to hypothesize that the increase in adiponectin is likely to explain, at least in part, some of modifications in the lipid profile (namely the reduction in triglycerides) and in glucose metabolism associated with BMI z-sc reduction.

The concern with childhood obesity must begin in overweight children and not only in severe obese children as it is easier to intervene in an early phase. Besides, there are studies that point that the majority of the diseases appear in individuals with moderated elevated risk factors [35].

Further studies involving a standardized physical activity and diet programs together with the assessment of important variables as CRF, at baseline and through the study period, could help to better understand the process involved in weight loss differences found among children and, also, the relation between the changes in adiposity and variations in CVD risk markers.

In summary, our data demonstrates that lifestyle modifications in obese children and adolescents, even associated with small reductions in BMI z-sc, improve the CV risk profile, namely lipid profile and insulin resistance. This may be important in motivating patients as health changes can appear earlier than the esthetical ones. Furthermore, patients with higher BMI and/or insulin resistance seem to experience a greater relative reduction in their BMI after lifestyle improvements, probably due to a physiologic adaptation favoring weight loss.

None declared.