Introduction
The ongoing global rise in the prevalence of overweight and obesity among all age and ethnic groups is accompanied by a higher incidence in serious health risks, such as type 2 diabetes (T2DM), and the development of cardiovascular damage, which is already potentially present at an early age.[
1‐
3] In both children and adults, detection of the metabolic syndrome, which encloses a clustering of cardiometabolic disorders (i.e. central obesity, disorders in glucose regulation, dyslipidemia and hypertension) is a tool to identify individuals with high risk on future T2DM and cardiovascular disease (CVD).[
4‐
6] However, no validated definition with standardized cut-off values of the metabolic syndrome is available in children, and as a consequence, many derivatives from adult the metabolic syndrome-definitions are used to attempt to implement the metabolic syndrome in pediatrics.[
7‐
10] Therefore, a wide range of prevalence rates of the metabolic syndrome among pediatric cohorts has been reported, making it impossible to draw conclusions with respect to differences between cohorts from different countries. To overcome this issue in the present review, description and comparison of the prevalence of single cardiometabolic risk factors (the metabolic syndrome-criteria; impaired fasting glucose (IFG)/impaired glucose tolerance (IGT), low HDL-cholesterol, high triglycerides and hypertension) was performed.
Although it is difficult to standardize data from different studies regarding age and BMI, we embarked on this study to identify potential early trends with respect to cardiometabolic variables between children from different countries. The premise of this review was to estimate CVD risk by the number of the metabolic syndrome features, assuming that the risk for future CVD conferred by the metabolic syndrome in adults is proportioned to the number of the metabolic syndrome features present. Moreover, mean values of aforementioned variables between different countries can be compared and used to estimate cardiometabolic risk. In the present review, we performed a systematic literature search, and aimed to provide an overview of available studies reporting data regarding cardiometabolic risk factors in pediatric cohorts with different ethnicities and nationalities.
Discussion
The present review showed a large variation for prevalence rates and mean values of cardiometabolic risk parameters among children according to country of origin, and reveals several interesting findings. In summary, children from Norway, China, Belgium, France and the Dominican Republic constitute the tertile with the most favorable values of cardiometabolic risk parameters. In contrast, children from Central and Eastern Europe (Germany and Poland), together with South Eastern Europe (Hungary, Greece) and Turkey constitute the third of children with the most adverse values of cardiometabolic risk parameters. The former is in line with the finding that mortality rates from CVD are generally higher in Central and Eastern Europe than in Northern, Southern and Western Europe, as concluded by a report concerning cardiovascular disease statistics.[
21]
Some of the comparisons between cohorts from different countries have been studied before, however, most studies originate from the United States, with the focus on differences between the white, black and Hispanic population. Therefore, the cardiometabolic differences between black and white children living in the USA are well-known (i.e. a more favorable lipid profile and slightly higher prevalence of hypertension.[
7,
22‐
26] Our review supports this finding, and adds that this difference is also found between black children and other cohorts, however, children from Greece and Hungary present with similar blood pressure values as the cohort of children of mainly African-American descent. Although some differences may be explained by characteristics of the cohort (i.e. more obese), others cannot. In example, high triglycerides and high blood pressure were quite low in an Italian, extremely obese cohort, while they were frequently present in a moderately obese German cohort, and in a cohort from Central Europe (Germany/Switzerland/Austria).[
14,
27] Therefore, it is presumed that genetic profile and dietary habits play a substantial role outside the degree of obesity in affecting cardiometabolic risk parameters.
The relation between mean values in relation to prevalence rates remains under debate. Despite varying mean values amongst populations from different origins or ethnic backgrounds, reference values suggested by international guidelines are often applied, assuming a linear relationship between mean values of cardiometabolic parameters and actual cardiometabolic risk. However, it has been shown that not all populations have similar baseline risk for CVD and therefore, ethnic-specific reference values are warranted. To illustrate, in Mexico, an unusual high prevalence of low HDL-cholesterol among non-overweight subjects was found, (NCEP reference values), namely 83%.[
28] Surprisingly, mortality from CVD among the Mexican population is lower than among non-Hispanic whites.[
29] Therefore, one might argue that lower HDL-cholesterol levels among Hispanics (as compared to whites) are acceptable and do not lead to a higher mortality rate. In contrast, one might hypothesize that interventions improving lipid metabolism may lead to an even lower prevalence of death from CVD.
A recent review reported a lower prevalence of lipid disorders in black children as compared to white children, and were less frequently diagnosed with the metabolic syndrome as a consequence.[
30] This finding is contradictory to the finding that blacks are having more frequently diabetes and CVD as compared to whites. To overcome this discrepancy, the review suggested that lipid reference values in black children should be more strict, in order to achieve a prevalence of the metabolic syndrome which would correspond better with actual cardiometabolic outcome. Although the adjustment of lipid levels reference values might in part accomplish forementioned goal, probably hypertension and other factors such as ox-LDL, hs-CRP and IL-6 (which are not accounted for in the metabolic syndrome) may cause the higher incidence of diabetes and CVD among blacks.[
31]
,[
32] Therefore, in case of African-American children, the use of stricter cut-off values for lipid levels will probably not induce identification of those most at risk for CVD. Further studies are needed to determine the optimal reference values for cardiometabolic risk parameters in specific populations.
Also within countries, large differences between means for cardiometabolic risk parameters were found, which limits extension of the outcome to the general population. In example, cohorts from the two studies from Italy (Milano, mean BMI 33.9 ± 5.5 and Rome, mean BMI 26.8 ± 3.6, respectively), showed opposite cardiometabolic characteristics. While the cohort from Milano (despite the higher mean BMI), was in the tertile with the most favorable cardiometabolic risk profile, the cohort from Rome was in the terile with the most adverse cardiometabolic risk profile. Selection bias and geographical location of both institutions (ie outskirts vs. city center) may have accounted for the difference in outcomes.
The difference between cardiometabolic risk profile between the Turkish cohorts may be explained by differences in age range (2-19 vs. 7-18 yrs), confirming a more adverse cardiometabolic risk profile in older children. Interestingly, Dutch-Turkish children, (age range 3-18 yrs), presented with a similar risk profile as the older cohort (7-18 yrs) from Turkey. Forementioned findings thus indicate that Dutch-Turkish children have a cardiometabolic risk profile which is even less favorable as compared to children living in Turkey.
Next to age and geographical location, varying BMI may account for some of the differences found between the studies. It would be expected that the cardiometabolic risk parameters would be more favorable in similar samples with lower BMI values. Although, in the present review, the former statement was true for blood pressure (highest among extremely obese subjects), this did not apply to glucose levels, as they were lowest among the extremely obese cohorts. Since we do not have data from moderately obese children from the same country, the meaning of this finding remains unclear, however it might be hypothesised that in some cohorts, a favorable genetic profile may preponderate over increasing BMI in glucose regulation.
Few studies have been performed with regard to cardiometabolic risk factors among children of ethnic descent within countries. In a Norwegian study which compared immigrants (Pakistani, Tamil and Turkish) to native Norwegian children, higher rates of insulin resistance and the metabolic syndrome were found among immigrant subjects.[
33] In a study performed among 516 multi-ethnic children in the Netherlands, the cardiometabolic risk profile appeared more favorable in children of Moroccan descent, and more adverse in children of Turkish origin, as compared to Dutch native children. This previous finding is partly confirmed by the results of the present study, in which Turkish cohorts appear to have high mean glucose and triglyceride levels (but also high HDL-cholesterol).[
34] It may be postulated that immigration of Turkish children to the Netherlands has an additional adverse effect on cardiometabolic risk profile. This may be attributed to several reasons, among which change of dietary habits and less physical exercise.
One of the limitations of this review is the inclusion of mainly studies of clinical nature, with small sample sizes, which makes it hard to extend the individual study outcomes to the general population of a specific country. Despite careful selection of the studies, there was still some heterogeneity among them, and small differences between methods of blood pressure measurements and measuring height could have in part attributed to some the differences found. Moreover, data with respect to the values of cardiometabolic parameters in childhood and the actual cardiometabolic risk in adulthood are scarce. A single study was performed using mortality as primary end-point, and reported that glucose intolerance, and hypertension (not high total cholesterol) in childhood were strongly associated with increased rates of premature death from endogenous causes.[
35] Due to aformentioned reasons, the outcome of this review should be interpreted with caution; however, we believe that this review gives some insight and a good overview of available studies on the subject of cardiometabolic risk in childhood obesity.
In conclusion, among pediatric cohorts, consisting of both pubertal and prepubertal children, a wide variation of mean values for different cardiometabolic parameters were found, with most favorable values in cohorts of children from Norway, China, Belgium, France and the Dominican Republic, and least favorable values in a cohort of Dutch-Turkish children, as well as children from cohorts from Germany, Poland, Hungary, Greece and Turkey. These results should be taken with caution, given the heterogeneity of the relatively small, mostly clinical cohorts and the lack of information concerning the influence of the values of risk parameters on true cardiometabolic outcome measures in comparable cohorts.
Competing interests
The authors declare that they have no competing interests
Authors' contributions
MvV wrote the first draft of the manuscript. MvV and MD contributed in concept and design, data collection, analysis and interpretation, drafting the article and revision. MwH contributed in the concept and design, analysis and interpretation of data, and revision. IvR, DPMB and JB contributed to the design, data collection, interpretation of data, and revision. All authors have read and approved the final manuscript.