On the use of a continuous metabolic syndrome score in pediatric research

The constellation of adverse cardiovascular disease (CVD) and metabolic risk factors, including elevated abdominal obesity, blood pressure, glucose, and triglycerides (TG) and lowered high-density lipoprotein-cholesterol (HDL-C), has been termed the metabolic syndrome [1]. Based on the National Cholesterol Education Program Adult Treatment Program III (ATP II) criteria [1] and the International Diabetes Foundation criteria [2] approximately 35% and 39%, respectively, of U.S. adults possess the metabolic syndrome [3]. In terms of health outcomes, the metabolic syndrome is associated with an increased risk of all-cause mortality, CVD morbidity and mortality, and diabetes [4].

Given the current pediatric obesity epidemic [5], it is perhaps not surprising that recent reports indicate the emergence of the metabolic syndrome during childhood and adolescence [6‐10]. Data from the United States (U.S.) National Health and Nutrition Examination Survey (NHANES) III (1988–1994) showed that the prevalence rate of the metabolic syndrome was 4% in 12–19 yr old adolescents [10]. The prevalence rate increased to 6% in NHANES 1999–2000 [9]. Based on current estimates, > 2 million U.S. adolescents have the metabolic syndrome phenotype [9]. Another paper using the same data (i.e., NHANES III) but different criteria showed that the prevalence rate was nearly 10% [11]. The varying prevalence rates between studies reflects the lack of a universal definition and makes it difficult to draw conclusions. This issue is similar in the adult literature [12]. Prevalence rates from representative samples in other countries are limited. The metabolic syndrome in children and adolescents has been reported to be 6.5% in northern Mexico[13], 9% in Korea [14], 2% in Turkey [15], and 10% in Quebec, Canada [16]. Among obese children and adolescents the prevalence approximates 30–50% [10, 17, 18].

Since there is no universal definition of the metabolic syndrome in children or adolescence and the prevalence rate is relatively low, several authors [19‐28] have derived a continuous score representing a composite CVD risk factor profile or index (i.e., the metabolic syndrome score). The purpose of this paper is to provide an overview of the derivation and utility of the continuous metabolic syndrome score in pediatric epidemiologic research.

The rationale for creating a continuous metabolic syndrome score stems mainly from the fact that there is no clear definition of the metabolic syndrome in children or adolescence and the prevalence rate is relatively low. A low prevalence rate requires a large sample size in order to conduct association studies. For example, in a random sample of 500 U.S. adolescents approximately 15–50 subjects would possess the metabolic syndrome based on current estimated prevalence rates. Thus, the ability to show links between exposures (e.g., physical activity, diet, etc.) and the dichotomous outcome (i.e., metabolic syndrome) using logistic regression would limit the power to detect an association [29].

As mentioned, several authors have derived a continuous score to represent the clustering of components of the metabolic syndrome (Table 1). The inclusion of these key components (i.e., glucose, lipids, blood pressure, and adiposity) is supported by the results of factor analysis in children and adolescents [16, 30‐32] which shows the underlying patterns or structure among variables showing high degrees of inter-correlation. Thus, there are common variables that can be used to calculate a metabolic syndrome score. It has also been argued that a continuous metabolic syndrome score is statistically more sensitive and less error prone by comparison to the dichotomous approach [20, 29]. In the recent joint statement by the American Diabetes Association and the European Association for the Study of Diabetes [33], it was recommended that one area of necessary research was the definition of the metabolic syndrome based on continuous variables in a multivariate score system.

Several large-scale epidemiological studies that either focus on cardiovascular health or include CVD risk factors in children and adolescents have used various approaches to calculate a metabolic syndrome score (Table 1). As shown in Table 1, the variables included in the score and statistical approaches used to derive the score vary considerably. Indicators of obesity range from the body mass index (BMI) to skinfolds to waist circumference (WC), whereas others do not include a measure of adiposity. The same is true for the other components of the metabolic syndrome. To my knowledge, Eisenmann and colleagues are the only ones to use indicators aligned with the adult criteria. More specifically, WC, HDL-C, TG, mean arterial pressure (MAP) and some indicator of abnormal glucose metabolism are used. Besides differences in the variables included in the score, the statistical approaches used to calculate the score also vary considerably. Previous procedures include principal component analysis [23, 24], Z scores [19, 20, 23, 25‐27, 34], and centile rankings [21, 22]. In addition, a recent study proposed a categorical score referred to as the Metabolic Individual Risk-factor And Clustering Estimation (MIRACLE) score based on family history (early cardiovascular disease, type 2 diabetes, and hypertension), individual history (small for gestational age and ethnic origin), clinical features (BMI, waist circumference > 90th percentile and blood pressure > 95th percentile) and metabolic abnormalities (glucose intolerance or type 2 diabetes) [35].

In the past few years, we have used the Z score approach to calculate the metabolic syndrome score. In this section, some general considerations and the specific methodology of this approach are highlighted.

Several authors have shown significant differences or associations between various exposures and the metabolic syndrome score. For example, results from the Corpus Christi Child Heart Study indicated higher metabolic syndrome scores in Mexican-American boys and girls compared to White boys and girls [23]. Recent studies also show that habitual physical activity and aerobic fitness are inversely associated with the metabolic syndrome score [19, 20, 25‐27, 34]. Eisenmann and colleagues [25, 26, 34, 38] have also shown that aerobic fitness attenuates the metabolic syndrome score among high fat children and adolescents. The metabolic syndrome score has also been shown to track from childhood/adolescence into young adulthood [21, 22, 24, 27]. Other studies show an association between markers of the metabolic syndrome in childhood and type 2 diabetes and cardiovascular disease morbidities in adulthood [39, 40]. These latter findings hold importance to the predictive utility of the score on disease outcomes and confirms that the definition of the metabolic syndrome in youth and the calculation of the metabolic syndrome score should include the same variables as those in the adult criteria for the metabolic syndrome [37].

This paper provides an overview of the derivation and utility of the continuous metabolic syndrome score in pediatric epidemiology. The continuous score is important in epidemiological research since there is no universal definition of the metabolic syndrome for children or adolescents and the prevalence rate is relatively low. The score allows each subject to have a continuous value with lower values indicating a better metabolic syndrome profile and higher values indicating a poorer metabolic syndrome profile compared to the sample studied. Several variables and statistical procedures have been used to derive the metabolic syndrome score in previous research (Table 1). It is recommended that the five key metabolic syndrome variables be used in the calculation of the score in future research. These variables include 1) central obesity (as measured by WC – or BMI and/or skin fold thickness if WC is not available), 2) low HDL-C, 3) elevated TG, 4) elevated BP (systolic and/or diastolic and/or MAP), and 5) abnormal glucose metabolism (impaired fasting glucose, impaired glucose tolerance, and/or HOMA). Furthermore, the individual components should be age-standardized (and maturity-standardized, if available) given the influence of growth and maturation on the development of the metabolic risk factors.

Future research is needed to validate the current methodologies used on deriving the metabolic syndrome score in children and adolescents. Recently, Winjndaele et al [41] showed a continuous metabolic syndrome score derived from principal component analysis was higher in adult subjects with the metabolic syndrome and that the score increased progressively with increasing number of adverse risk factors. Although the metabolic syndrome score tracks from adolescence into adulthood [24, 42], association studies using established cohorts that link the metabolic syndrome score during childhood and adolescence with adult-diagnosed metabolic syndrome, type 2 diabetes, atherosclerosis and CVD mortality are required. If a simple and practical yet valid method utilizing criterion-reference standards can be developed, then the metabolic syndrome score could be used conventionally in pediatric epidemiological research, clinical medicine, and public health to better understand the prevention, diagnosis, and treatment of this emerging pediatric condition.