The Journal of Pediatrics
Volume 148, Issue 2 , Pages 149-151, February 2006

Pediatric metabolic syndrome: Smoke and mirrors or true magic?

  • Elizabeth Goodman, MD

      Affiliations

    • Corresponding Author InformationReprint requests: Elizabeth Goodman, MD, Brandeis University, Schneider Institute for Social Policy, Heller Graduate School, 415 South Street Mall, Stop 035, Waltham, MA 02454-9110.

Institute for Child, Youth, and Family Policy, Heller School for Social Policy and Management, Brandeis University, Waltham, MA 02454

Article Outline

AACE, American Association of Clinical Endocrinologists , ATP III, Adult Treatment Panel III , CI, Confidence interval , CV, Cardiovascular , HDL-C, High-density lipoprotein cholesterol , IRS, Insulin resistance syndrome , WHO, World Health Organization

 

Clustering of cardiovascular (CV) risk factors in adolescence is not a new phenomenon. However, interest in CV risk factor clustering has been renewed due to the recent surge in attention to the metabolic syndrome. In the past three years, research on this phenomenon has skyrocketed, with annual publications on the subject more than tripling between 2002 and 2004.

Despite this explosion of research, however, serious debate remains as to the definition of metabolic syndrome and even whether it actually exists.1, 2 Since Reaven’s description of insulin resistance syndrome (IRS) in 1988, this clustering of CV risks has gone by more than 15 different names.3 Hereafter, I will refer to it as metabolic syndrome.

Metabolic syndrome reflects a wide array of factors, including adiposity, dyslipidemia, hypertension, hyperinsulinemia, impaired glucose metabolism, microalbuminuria, abnormalities in fibrinolysis, and inflammation.3, 4 This abundance of factors makes defining metabolic syndrome a challenge. Three health organizations have created clinical criteria for defining this syndrome in adults. These include the World Health Organization (WHO),5 the National Cholesterol Education Panel’s Adult Treatment Panel III (ATP III),6 and the American Association of Clinical Endocrinologists (AACE).7 These definitions differ significantly. The AACE lists 12 clinical criteria for the diagnosis of what they specifically refer to as IRS. No required number of these risks is specified for diagnosis; rather, this is left to clinical judgment. With the development of diabetes, the diagnosis of IRS becomes inapplicable. In contrast, both the WHO and ATP III specify a minimum number of risk factors required for the diagnosis of metabolic syndrome and do not exclude those with type 2 diabetes mellitus. In fact, the WHO explicitly requires the presence of insulin resistance. In contrast, the ATP III does not include insulin resistance in its definition, focusing instead on 5 metabolic risk factors as a way to identify obese persons at greatest risk for medical complications.2 Both the WHO and ATP III definitions require the presence of at least 3 risk factors for a diagnosis of metabolic syndrome. Although there are shared risk factors between the WHO and ATP III definitions, the defining cutoff points differ. Moreover, the WHO definition, which was modified in 1999,8 condenses some risk factors into parameters, whereas each risk factor is independent in the ATP III definition.

The differences in these definitions have important implications for case identification. A person with hyperglycemia, hypertriglyceridemia, and low high-density lipoprotein cholesterol (HDL-C) levels would have metabolic syndrome by the ATP III criteria but not by the WHO crtieria. In contrast, a person with hyperinsulinemia, low HDL-C, and obesity would have metabolic syndrome by the WHO criteria, but not by the ATP III criteria. These concerns about differing specificity have been borne out by studies contrasting the WHO and ATP III definitions in adults9, 10 and adolescents11 that found important differences between definitions in case identification and in prognostic ability.

Even setting aside differences in definitional criteria, there remain differences in how each risk factor within a definition is operationalized. Although true for both adult and pediatric definitions, these differences are especially problematic for pediatric populations, for whom no accepted definition of metabolic syndrome exists. Almost all of the pediatric metabolic syndrome studies use a definition based on age- and sex-specific percentiles for the various cutoff points of the 5 ATP III risk factors. However, these cutoff points range from 75% to 97.5% (5% to 25% for HDL-C), depending on the study and the risk factor of interest.12, 13, 14, 15, 16, 17 A major problem with this approach is that the pediatric percentiles do not adjust for the transition to adulthood at which point the adult criteria, which are not based on percentiles of distribution, will be applied. For example, the pediatric definitions often use the 90% cutoff point for waist circumference. For 18-year-olds, the adult cutoff points (102 cm for men and 88 cm for women) fall somewhere between the 75th and 90th percentiles for boys and slightly above the 75th percentile for girls.18 Thus an 18-year-old could be classified as having central obesity based on adult definitions of metabolic syndrome but not be considered to have central obesity by the pediatric definition. Such differences in criteria between adolescents and adults create difficulties in understanding the developmental trajectory of cardiovascular risk as adolescents age into the adulthood.

The use of different cutoff points can also have striking effects on reported prevalence rates. Two studies assessing metabolic syndrome prevalence in NHANES III among 12- to 19-year-olds, published about a year apart, differed in the cutoff points used to define the metabolic syndrome components.13, 19 The differences in the reported prevalences between the 2 studies were dramatic. Mexican-Americans had the highest prevalence in both studies, but the first study reported a prevalence of 5.6% (95% confidence interval [CI], 3.6% to 7.5%), whereas the second study reported a prevalence more than twice as high (12.9%; 95% CI, 10.4% to 15.4%). These CIs do not overlap, suggesting that the difference in reported rates is significant. Similar differences and nonoverlapping confidence intervals were seen for European-American adolescents (4.8% vs 10.9%). The prevalences were much closer for African-American youth (2.0% vs 2.5%).

The article by Retnakaran et al in this issue of The Journal provides an assessment of metabolic syndrome in Native Canadian adolescents taking part in the Sandy Lake Health and Diabetes Project.20 This is the first report of metabolic syndrome in a First Nations population, a group that has been experiencing alarming increases in obesity and is at increased risk for type 2 diabetes mellitus and CV disease.21 Retnakaran et al use a definition of metabolic syndrome derived from the NHANES II study discussed earlier.19 Their reported prevalence is 18.6%. As the authors note, this rate is high for a population-based study. In addition to the definitional issues noted earlier, a higher prevalence of obesity in this ethnic group may be influencing the prevalence in this population;21 metabolic syndrome has consistently been shown to be much higher in obese youth.11, 13, 17

Obesity clearly plays an important role in the pathogenesis of metabolic syndrome, but the underlying “cause” of this risk factor clustering remains elusive. Investigators have used factor analysis over the past 7 years as a means to assess the underlying structure and cause of metabolic syndrome. Both pediatric and adult studies generally find 2 to 4 factors when assessing the traditional risks associated with metabolic syndrome. However, when the number of risks included in the analysis expands, the constellation of factors changes. Retnakarnan et al’s factor analysis, which included several nontraditional CV risk factors, revealed 5 factors. This is a larger number than has been reported in previous pediatric factor analytic studies, including a population-based study and a clinic-based study, each of which used expanded biomarkers of CV risk in their factor analysis.14, 22

Making sense of this widely divergent number of factors and changing constellation of included risks is a daunting task. How much do these analyses really tell us about the underlying cause or structure of the metabolic syndrome? Very little, I believe. Why? Because the method of data analysis does not support making such claims. Principal components analysis (the type of exploratory factor analysis used in these studies) is an empirical, atheoretical method of data reduction. This technique merely uses the linear relationships among variables to create a smaller number of summary factors that maximize the explained variance in the observed variables. Although investigators have used factor-loading patterns to explain the structure underlying the metabolic syndrome and CV risk clustering, this statistical technique does not provide a test of a hypothetical causal model. The factors are simply a mathematical transformation of the measured variables, and thus no latent meaning can be ascribed to them.23 Moreover, the relationships of many of these biomarkers of CV risk, although correlated, are often not highly linear, which may account for the low cumulative variance explained by some of the factor analyses. Explained variance in the factor analyses by Retnakarnan et al was only 50.7% for traditional metabolic syndrome risks and 53.9% for both traditional and nontraditional risks.

There are also methodological challenges to the use of exploratory factor analysis. The technique is not standardized, making comparisons across studies difficult. There are different means for defining factors, including several methods of factor extraction, inconsistent use of standardization of observed variables, and no consensus on the type of rotation or strength of the rotated factor loading used for interpretation. Factor loadings are measures of shared variance between variables and summary factors. The higher the factor loading, the more shared variance exists. A factor loading of 0.4 indicates at least 15% shared variance between the variable and summary factor. Retnakarnan et al used a cutoff point of 0.3 for interpreting factors. Although this cutoff point has been used in other pediatric factor analytic studies, in this instance it seems generous. A review of these authors’ factor loadings indicates that most variables have low factor loadings. Insulin, shown in most studies to strongly load onto at least 1 factor (and often onto 2 factors), loads only weakly (0.34) onto the glucose tolerance factor. Diastolic blood pressure does not make the 0.3 cutoff point for any factor, even the blood pressure factor. This suggests that a more parsimonious model may fit the data better and that the conclusion that 5 core traits underlie early development of CV risk may be premature.

What does drive development of CV risk early in life? Scientific investigations seek answers, and questions such as this one regarding causality are often the most compelling. For epidemiologic studies, causality is more a philosophical state than a tangible reality. Complex statistical techniques, such as exploratory factor analysis, appear to substantiate causal explanations. Yet exploratory factor analysis on cross-sectional data does not meet most of the basic standards for assessing causality used in epidemiology today. Take time sequencing, for instance. For X to be a cause of Y, X must precede Y. Does insulin resistance precede obesity? Probably not. Obesity likely precedes insulin resistance for most individuals. Yet the etiologic explanation promoted by many factor analytic studies, including the study by Retnakaran et al, is that metabolic syndrome is a consequence of insulin resistance. This idea is the most widely accepted hypothesis describing the pathophysiology underlying metabolic syndrome.24 Principal components analysis, by its nature, is not able to test this hypothesis.

Given the complexity of defining metabolic syndrome, the lack of consensus regarding definition, and the questions regarding its very existence and meaning, it is not surprising that there is no general agreement regarding clinical assessment of or potential treatment for pediatric metabolic syndrome. However, both the American Diabetes Association and the American Heart Association agree that obesity prevention and treatment in childhood should be the first-line approach to this problem and that a fasting glucose or oral glucose tolerance test should be obtained on children and adolescents considered at risk for development of type 2 diabetes mellitus.25 Continued efforts to prevent and treat obesity in children and adolescents, and vigilant attention to the early diagnosis of diabetes, provide the pediatrician with the most evidence-based methods for addressing metabolic syndrome and the clustering of CV risks that it represents in childhood and adolescence. As to the fate of metabolic syndrome, whether it will continue to dominate our thinking or will be laid to “requiescat in pace,”1 only time will tell.

Back to Article Outline

References 

  1. Reaven GM . The metabolic syndrome (requiescat in pace) . Clin Chem . 2005;51:931–938
  2. Grundy SM . Point (the metabolic syndrome still lives) . Clin Chem . 2005;51:1352–1354
  3. Reaven GM . Role of insulin resistance in human disease . Diabetes . 1988;37:1595–1607
  4. Meigs JB . Invited commentary: Insulin resistance syndrome? Syndrome X? Multiple metabolic syndrome? A syndrome at all? Factor analysis reveals patterns in the fabric of correlated metabolic risk factors . Am J Epidemiol . 2000;152:908–911
  5. Alberti KG , Zimmet PZ . Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus provisional report of a WHO consultation . Diabetes Med . 1998;15:539–553
  6. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive summary of the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) . JAMA . 2001;285:2486–2497
  7. Einhorn D , Reaven GM , Cobin RH , Ford E , Ganda OP , Handelsman Y , et al.   American College of Endocrinology position statement on the insulin resistance syndrome . Endocr Pract . 2003;9:237–252
  8. Balkau B , Charles MA . Comment on the provisional report from the WHO consultation. European Group for the Study of Insulin Resistance (EGIR) . Diabetes Med . 1999;16:442–443
  9. Ford ES , Giles WH . A comparison of the prevalence of the metabolic syndrome using two proposed definitions . Diabetes Care . 2003;26:575–581
  10. Laaksonen DE , Lakka HM , Niskanen LK , Kaplan GA , Salonen JT , Lakka TA . Metabolic syndrome and development of diabetes mellitus (application and validation of recently suggested definitions of the metabolic syndrome in a prospective cohort study) . Am J Epidemiol . 2002;156:1070–1077
  11. Goodman E , Daniels SR , Morrison JA , Huang B , Dolan LM . Contrasting prevalence of and demographic disparities in the World Health Organization and National Cholesterol Education Program Adult Treatment Panel III definitions of metabolic syndrome among adolescents . J Pediatr . 2004;145:445–451
  12. Raikkonen K , Matthews KA , Salomon K . Hostility predicts metabolic syndrome risk factors in children and adolescents . Health Psychol . 2003;22:279–286
  13. Cook S , Weitzman M , Auinger P , Nguyen M , Dietz WH . Prevalence of a metabolic syndrome phenotype in adolescents (findings from the Third National Health and Nutrition Examination Survey, 1988-1994) . Arch Pediatr Adolesc Med . 2003;157:821–827
  14. Weiss R , Dziura J , Burgert TS , Tamborlane WV , Taksali SE , Yeckel CW , et al.   Obesity and the metabolic syndrome in children and adolescents . N Engl J Med . 2004;350:2362–2374
  15. Cruz ML , Weigensberg MJ , Huang TT , Ball G , Shaibi GQ , Goran MI . The metabolic syndrome in overweight Hispanic youth and the role of insulin sensitivity . J Clin Endocrinol Metab . 2004;89:108–113
  16. Frontini MG , Srinivasan SR , Berenson GS . Longitudinal changes in risk variables underlying metabolic syndrome X from childhood to young adulthood in female subjects with a history of early menarche (the Bogalusa Heart Study) . Int J Obes Relat Metab Disord . 2003;27:1398–1404
  17. Duncan GE , Li SM , Zhou X-H . Prevalence and trends of a metabolic syndrome phenotype among U.S. adolescents, 1999-2000 . Diabetes Care . 2004;27:2438–2443
  18. Fernandez JR , Redden DT , Pietrobelli A , Allison DB . Waist circumference percentiles in nationally representative samples of African-American, European-American, and Mexican-American children and adolescents . J Pediatr . 2004;145:439–444
  19. de Ferranti SD , Gauvreau K , Ludwig DS , Neufeld EJ , Newburger JW , Rifai N . Prevalence of the metabolic syndrome in American adolescents (findings from the Third National Health and Nutrition Examination Survey) . Circulation . 2004;110:2494–2497
  20. Retnakaran R , Zinman B , Connelly PW , Harris SB , Hanley A . Non-traditional cardiovascular risk factors in pediatric metabolic syndrome . J Pediatr . 2006;148:176–182
  21. Hanley AJ , Harris SB , Gittelsohn J , Wolever TM , Saksvig B , Zinman B . Overweight among children and adolescents in a native Canadian community (prevalence and associated factors) . Am J Clin Nutr . 2000;71:693–700
  22. Goodman E , Dolan LM , Morrison JA , Daniels SR . Factor analysis of clustered cardiovascular risks in adolescence (obesity is the predominant correlate of risk among youth) . Circulation . 2005;111:1970–1977
  23. Nunnally JM , Bernstein IH . Psychometric theory . 3rd ed.. New York: McGraw-Hill; 1994;
  24. Eckel RH , Grundy SM , Zimmet PZ . The metabolic syndrome . Lancet . 2005;365:1415–1428
  25. Steinberger J , Daniels SR . Obesity, insulin resistance, diabetes, and cardiovascular risk in children (an American Heart Association Scientific Statement from the Atherosclerosis, Hypertension, and Obesity in the Young Committee (Council on Cardiovascular Disease in the Young) and the Diabetes Committee (Council on Nutrition, Physical Activity, and Metabolism)) . Circulation . 2003;107:1448–1453

PII: S0022-3476(05)00824-3

doi:10.1016/j.jpeds.2005.08.057

Refers to article:

  • Nontraditional cardiovascular risk factors in pediatric metabolic syndrome

    Ravi Retnakaran, Bernard Zinman, Philip W. Connelly, Stewart B. Harris, Anthony J.G. Hanley
    The Journal of Pediatrics February 2006 (Vol. 148, Issue 2, Pages 176-182)

The Journal of Pediatrics
Volume 148, Issue 2 , Pages 149-151, February 2006

Article Tools