Reviews in Endocrine and Metabolic Disorders

, 10:189

The influence of fitness on insulin resistance in obese children


    • Department of Pediatrics, American Family Children’s HospitalUniversity of Wisconsin
    • Department of PediatricsUniversity of Wisconsin
  • David B. Allen
    • Department of Pediatrics, American Family Children’s HospitalUniversity of Wisconsin

DOI: 10.1007/s11154-009-9109-5

Cite this article as:
Carrel, A.L. & Allen, D.B. Rev Endocr Metab Disord (2009) 10: 189. doi:10.1007/s11154-009-9109-5


An increasingly pervasive environment of reduced activity and easy access to high caloric food is leading to an epidemic of poor cardiovascular fitness, obesity, insulin resistance and type 2 diabetes (T2DM) in children. Studies have shown that insulin resistance (IR) to be an independent predictor for morbidity as well as mortality. These serve as a strong stimulus for public health strategies to improve fitness in children and adolescents. Methods to assess IR, improve IR and understand complications are increasingly important in children.


Cardiovascular fitnessMax VO2Insulin sensitivity

1 Introduction

Childhood obesity and poor fitness increase the risk of insulin resistance (IR) and other complications including Type 2 diabetes mellitus (T2DM) [1, 2]. Increasing evidence indicates that IR plays a central role in the inflammatory cascade leading to cardiovascular disease and T2DM. Therefore, further understanding of the pathogenesis and prevention of IR is needed [3]. While obesity is clearly linked to IR and inflammation in children and adolescents, determination of obesity by body mass index (BMI) is a poor predictor of IR in children [4]. In the midst of interventions altering diet and exercise habits in obese children to reduce risk for IR and T2DM, time dedicated to physical activity continues to decrease for children [5]. Given the clear relationship between physical activity and IR in adults, declining levels of fitness may be an equal, if less visible, factor promoting IR and inflammation. The degree to which strategies for reducing the adverse health effects of obesity in children should emphasize improving fitness and/or reducing fatness remains unresolved [6]. In this chapter, we assert that effective public health strategies for assessment of and interventions for at-risk children should include measurement of fitness. While the “gold standard” for assessing cardiovascular fitness (CVF) is laboratory-based maximal VO2 testing for CVF [7], a reliable, practical, and relatively effort-independent method of assessing CVF in children is needed for large-scale childhood fitness evaluations. This test would optimally have the same predictive value for metabolic indicators of health risk (IR and inflammation) as laboratory-based measurements of maximal VO2.

2 Measurement of insulin resistance

Insulin resistance, the degree to which the tissues such as muscle and adipose respond to insulin levels, is strongly (but not exclusively) associated with obesity, and is considered to be the first step in the development of T2DM and other chronic conditions [2, 8, 9]. Conversely, greater insulin sensitivity is a protective factor against all of these clinical events. IR is increasing in childhood and adolescence, and both obesity and poor cardiovascular fitness (CVF) promote IR. A seminal study demonstrated IR is the strongest prospective predictor of the development of cardiovascular disease, hypertension, stroke, T2DM, and cancer in adults, and that the impact of IR on these outcomes is independent of BMI [10]. These data support the notion that IR with or without concomitant obesity plays a central role in the genesis of multiple health risks for all people, including children [4]. Importantly, improved CVF can attenuate the morbidity commonly associated with obesity, and active overweight individuals can have lower risk for T2DM than sedentary normal-weight individuals [11, 12].

Individuals with IR are at increased risk of developing multiple adverse clinical outcomes (48). Because of the severity of these clinical outcomes, there is considerable interest in accurately assessing IR. Since many techniques are viewed as being too time- and labor-intensive for routine use in studies, a variety of surrogate markers have been evaluated including fasting insulin alone, homeostasis model of insulin resistance (HOMA), and quantitative insulin sensitivity index (QUICKI). Correlations between HOMA and QUICKI have been strongest with insulin-mediated glucose uptake with correlations of HOMA at r = 0.83, and QUICKI at r = 0.89 [1315]. While fasting insulin levels may not be as diagnostic for IR as frequently sampled testing, the utility of using both body fat measurement and fasting insulin levels was found to have greater precision than HOMA alone in some studies [16].

3 Insulin resistance is related to fitness in children

Recently, it has been shown that both higher percent body fat and lower CVF were independently associated with higher insulin concentrations in adolescents [17, 18] Adjustment for BMI and other potential confounders did not eliminate the significant inverse association between physical activity and diabetes incidence [19]. The European Youth Heart Study group demonstrated strong association between CVF and the clustering of cardiovascular risk factors [20]. In this study, the odds ratios for clustering in each quartile of fitness, using the quartile with the highest fitness as reference, were 13.0, 4.8, and 2.5 respectively, after adjusting for age, socioeconomic status, family history of CVD and diabetes. The Bogalusa Heart Study has shown that childhood obesity is a powerful predictor of IR later in life [21]. One limitation of this study is that it did not consider the role of CVF. Because of the strong inverse correlation between fitness and percent body fat, it is possible that negative health consequences ascribed to obesity may be partially due to poor CVF levels [22, 23].

Thus, it is important to consider the role of fitness and other non-weight parameters, as well as fatness, in the development of IR and T2DM. The Diabetes Prevention Program (DPP) demonstrated a reduction in the incidence of diabetes in high-risk individuals with lifestyle intervention [24]. Since measurement of IR appears to serve as a reliable “marker” for associated risks of obesity [13, 25], targeting IR may provide a more effective population health strategy than weight alone in an environment which promotes childhood obesity [26]. Further, both the US DPP and the Finnish DPP demonstrate clearly that increased fitness can reduce the risk of IR or overt T2DM [24, 27]. In older adolescents (ages 14–18 years), both higher fatness and lower CVF are independently associated with higher insulin concentrations [9, 10]. Further, school-based interventions in adolescent females have been shown to increase physical activity and prevent an anticipated decline in CVF [28]. Expanded clarification of these associations remains incomplete, but ability to assess CVF in children remains central to these interventions.

In adults, level of fitness more accurately predicts cardiovascular and all-cause mortality than weight status [12]. Physical activity also improves insulin sensitivity, independent of changes in weight and body composition in adults [29]. Cross-sectional studies in adults also demonstrate that CVF (VO2 max) is inversely associated with the metabolic syndrome [30]. In children, physical activity not only improves insulin sensitivity, but also increases the effects on hormones that promote lean mass accretion, such as insulin-like growth factor-1 (IGF-1), and growth hormone [31, 32]. In one cohort of older teenage women, max VO2 was a more critical determinant of IR than body fat [33]. Thus, beneficial effects of fitness training on insulin sensitivity in children likely result from a combined increase in lean mass and reduction in fat mass, in contrast to reductions in both lean and fat mass observed during caloric restriction alone [34]. Thus, efforts to improve insulin sensitivity in children may be best focused on increasing physical activity rather than simply restricting calories for weight control [35].

Extensive evidence links obesity to various health problems, and these associations are dose-related, temporally consistent, and biologically plausible, yet the same can be shown for physical inactivity [36]. However, the majority of studies examining obesity and health have not accurately measured or documented physical activity [3]. Particularly in studies involving children, physical activity is often “measured” by questionnaire, or self-report, rather than by direct measurement [37]. Use of maximal exercise tests (max VO2) to quantify cardiovascular fitness is an objective measurement that is stronger than self-reported physical activity as a predictor of many health outcomes [7]. Studies of obese adults with at least moderate CVF have lower mortality rates than their normal-weight but unfit peers [38, 39]. Our group has shown that in middle school children, body fat and CVF were significant independent predictors of fasting insulin, with CVF a better predictor of insulin sensitivity than %BF [40]. Thus, in children like adults, CVF is an equal, perhaps even more important predictor of insulin sensitivity than fatness. Further, the ability to document changes in CVF using direct measurements in children outside of the laboratory would significantly enhance the detection of children at risk for IR.

Altering physical activity and improving fitness is, of course, dependent on lifestyle interventions that, while challenging, can be successful. The Diabetes Prevention Program (DPP) showed a reduction in the incidence of diabetes in high-risk individuals with lifestyle intervention [24]. The Danish Youth and Sport Study demonstrated a strong relationship with physical fitness and the tracking of IR from adolescence to young adulthood, justifying for fitness intervention during childhood [41, 42]. Thus, both the US and the Finnish DPP demonstrate that increased fitness can reduce the risk of IR or overt T2DM [24, 27, 43]. Since measurement of IR appears to serve as a reliable “marker” for associated risks of obesity [13, 25], targeting the identification and treatment of IR may provide a more effective population health strategy than weight alone for childhood obesity [21].

4 Fitness is linked to metabolic health in children

Although obesity is clearly linked with an increased risk for T2DM, CHD, and other chronic illnesses, there remains controversy over the relationship between BMI and mortality [44]. Many researchers claim that fitness is not taken into account in this relationship. The role that CVF plays on morbidity and mortality is evaluated in studies that evaluate physical activity (PA), cardiovascular fitness (CVF) or both. The European Youth Heart Study has tried to separate the independent roles that PA, CVF, body fat, and gender play on metabolic risk [45]. A strong relationship was also shown between CVF, waist circumference, and systolic blood pressure; with the greatest relation between CVF and anthropometric measurements [46]. Of note, these same authors report that the relation of PA and CVF is not as strong in children, as is seen in adults. In a review of 37 studies, there was only a moderate relation (median correlation r = 0.17) between CVF and PA in the adolescent population, unlike adults [47]. CVF is thought to be an integrated measure of multiple body systems, and is thus considered an important health marker for CVD [48]. Cardiovascular disease events most often occur during the fifth decade of life or later. However, there is evidence that the precursors of CVD have their origins in childhood. Risk factors such as lipid profiles, insulin resistance, blood pressure, and body fat have been shown to track into adulthood [49]. Each of these risk factors can be converted to a z-score, and then summed to compute an overall “metabolic risk score”. Multiple groups have documented negative associations in children with low CVF and increased clustering of CVD risk scores, including markers of inflammation [20, 45, 50]. Our group and others have shown that cardiovascular fitness is as important as body fat with respect to predicting degree of insulin resistance [40]. The concept of insulin sensitivity (or resistance) can be seen as a “balance” depicted in the figure below. (Fig. 1)
Fig. 1

This depicts the spectrum of insulin sensitivity/resistance and the role that fitness and visceral body fat play upon insulin and health

5 School-based programs can improve fitness

Schools play a critical role in promoting general fitness for children, and the Institute of Medicine recommends that schools conduct annual assessments of students’ BMI [51, 52]. The Surgeon General’s Report recommends 60 min of moderate physical activity on most days for children. Unfortunately, 85% of American adults do not engage in moderate activity 5 or more days per week [53]. Thus, childhood has been an identified as a critical period for nurturing lifetime activity behavior [54] and school physical education becomes a primary site for promoting active lifestyles. Physical education classes have the potential to provide 97% of children in the US with regular physical activity [55, 56]. For example, SPARK (Sports, Play, and Active Recreation for Kids), a research-based organization developed in 1989, which has developed age-appropriate PE programs designed to promote lifelong wellness all of which have proven to increase physical activity substantially during school physical education class time [56].

During the past two decades, in spite of the knowledge that physical activity leads to improvement in fitness and self-esteem, even without weight loss, the requirement for and time allotted to physical education classes has decreased [57, 58]. For children, supplementary exercise and healthy nutritional changes can be successfully promoted in varying environments [5963]. The most successful programs incorporate consistent dietary guidelines, reduced television viewing time, and increased physical activity into children’s daily routine, including the school environment [64] (e.g. “Planet Health”) While promoting increased physical activity centered at school can demonstrate positive changes, decreasing sedentary time such as television viewing has also demonstrated success [65]. In most of the school-based studies to date, endpoints included weight, body fat or BMI. When obesity alone is the endpoint, however, other beneficial outcomes may be obscured: e.g. the Pathways project failed to reduce obesity in Native American children, but did succeed at improving dietary intake and activity [66, 67]. These suggest that interventional studies should optimally incorporate measures of fitness as well as obesity when assessing outcomes. In a recent review of school-based interventions, the value of longer term interventions and outcomes including fitness, not just BMI or obesity were emphasized [68].

We, and others, have shown that a school-based fitness intervention can improve CVF, decrease body fat, and reduce fasting insulin levels in obese children. It is important to point out, however, that most school-based interventions use BMI as an outcome, rather than fitness, or insulin sensitivity. Although BMI is a screening tool, and does not provide as direct an assessment on childhood health, unlike assessment of fitness or insulin sensitivity. Two methodology concerns arise from these types of studies—lack of ability to measure individual activity longitudinally, and the time, effort, and expense involved in laboratory-based measurements. Reliable, inexpensive, in-school (or other community setting) measurements would provide tools for physical education teachers and others to assess children’s capacity for exercise. The next step in the development of such tools is to establish the accuracy of measurement systems that can be used in a school compared to this laboratory-based evaluation.

6 New methods are needed to measure fitness in the school and community setting

It is becoming increasingly clear that physical activity in children provides an important counterbalance to the negative health effects simplistically blamed on obesity. Yet, the ability of investigators to assess CVF outside of the laboratory relies on surrogate tests, sub-maximal tests, or tests that are effort-dependent and thus less reliable. Evaluating children within their environment (individual, interpersonal, organizational/school, community)—is an increasingly recommended approach in public health, acknowledging the social ecological model of behavior. (Fig. 2) We have demonstrated that improved CVF, body composition, and insulin sensitivity are attainable with a school based intervention [62].
Fig. 2

The social ecological model is a theoretical model that acknowledges the importance of studying how children interact in their natural environment

The ability to accurately measure and track IR outside the laboratory is limited. Historically, precise measurement of IR requires blood studies of insulin levels required to maintain euglycemia after glucose administration (“euglycemic clamp study”), a procedure clearly impractical in school or community settings. It has been shown, that precise laboratory assessment of fitness in children (utilizing laboratory maximum oxygen consumption treadmill testing, or VO2 max) has a highly significant correlation with IR. Thus, it is reasonable to ask whether feasible field-based assessments of childhood fitness, body composition, and genetic predisposition could provide a valid prediction of childhood IR.

Additionally, a major challenge for any intervention involving an enhancement of physical activity is the ability to objectively and precisely document (rather than relying on self-reporting) what physical activity is actually taking place, and to measure the changes in work and energy expenditure (power) associated with this physical activity in the free-living environment. There are currently no available methods for the measurement power and fitness (i.e. outside of the exercise laboratory) that are affordable, accurate, effort-independent, and accessible to schools or individuals. Several devices based on combinations of three-dimensional accelerometers, heart rate monitors, breathing sensors and other technologies are commercially available to estimate total energy expenditure over hours or days. However, these are not designed to measure fitness or the capacity to produce power, nor are they designed or calibrated for children.

To scientifically validate effectiveness of interventions and to facilitate their acceptance and implementation, there remains an important and pressing need to develop technology that will directly measure CVF and correlate it with outcome. What has been needed is a new way of monitoring energy expenditures that provide specific, replicable metrics—metrics that relate to short-term and total energy expenditures under varying environmental conditions. Short-term measures of power is one example that can provide the basis for assessment of physical condition as it relates to CVF and IR diseases. If this is rapid and inexpensive, it could be used for screening large numbers of children for health risks [69]. Use of the PowerTap device, which measures power generation during cycling, provides reliable assessment of power in the exercise lab and school settings, and could potentially be widely used to measure childhood aerobic fitness [7072].

7 Power and PACER: translating laboratory assessment of CVF into the real world

The current gold standard for assessing CVF, maximum oxygen consumption (VO2 max), is highly specialized and cannot be performed in larger school or community settings. Serious efforts to improve CVF are being implemented in an effort to stem the epidemic increase in the incidence and progression of obesity, cardiovascular disease, and T2DM. However, accurate “measurement of fitness” is difficult to accomplish in schools. Because CVF is an independent predictor of multiple cardiovascular and metabolic health problems, an appropriate public health goal should be to improve the CVF levels of all children, obese and non-obese alike. These changes could be used in school, after-school, or summer programs to avoid any loss during the 3-month summer break [73].

Given the difficulty of applying maximal oxygen consumption (max VO2) measurements to large populations, a variety of power assessments at sub-maximal work have been developed, among the more widely used are the (sub-maximal) physical work capacity at a heart rate of 170 (PWC170), and the Astrand-Ryhming test [74]. Sub-maximal power has been used as an assessment of CVF in epidemiological and population-based studies [75], while assessment of maximal power has been used in elite athletes, especially in cycling and military recruits [76]. We conducted one of very few trials testing this method in children, and found that some modifications were needed. Some are unable to attain sustained heart rates of 170, and the ability to generate power in absolute terms varies considerably by the type of test (e.g., running versus cycling). Nevertheless, the children felt comfortable on bicycles, and could successfully perform maximal power testing at the same time that other assessments of CVF were occurring [72]. Power, then, can be reliably and fairly easily measured in non-laboratory settings, and provides a useful predictor of CVR and IR. On the other hand, the requirement for specialized (albeit portable) equipment, and one-child-at-a-time testing limitation would likely limit widespread applicability to school settings.

A more attractive option is a valid, cost-effective, and feasible field-based test of childhood CVF which can be shown to be predictive of IR. We studied children who performed the Progressive Aerobic Cardiovascular Endurance Run (PACER), a school-based CVF test using Fitnessgram, and then underwent evaluation of maximal oxygen consumption treadmill testing (VO2 max), body composition (percent body fat and BMI Z-score), and IR (derived homeostasis model assessment index-HOMAIR). Results showed the PACER to be a feasible and cost-effective field-based test that correlated very closely (r > 0.8) with CVF in both normal and obese kids and was predictive of IR with a high degree of statistical significance [77]. In this group of children, simple measures of both childhood fitness (PACER) and fatness (BMI) together accurately predicted IR to a greater degree than using BMI alone. This approach is practical and could help schools and communities evaluate the need for interventions, track progress, identify best practices, and target limited resources.

8 Practical approach to increasing fitness

There is a growing understanding of how physical activity improves health, and reduces illness. In the last decade, clinicians have accepted the concept of enhanced physical activity as a preventive strategy. Although the vast majority of individuals can initiate and sustain an increased personal fitness program on their own, clinicians should not underestimate their important role in promoting fitness. Children and families do respond to physician advice, and therefore it is important that we encourage everyone to be physically active on a regular basis. We can promote incorporating activity (“exercise” may sound too rigorous) into daily routines as a simple and inexpensive means to achieve this goal. This “lifestyle” approach is aimed at decreasing sedentary habits and incorporating more activity into our daily lives, such as taking stairs rather than elevators, walking and biking to school, if we are driven—park further away and spend a few minutes walking, use a exercise ball to sit on (rather than a chair) at a desk or other sedentary times. The literature supports that physician counseling is effective [78]. This can lower the barrier to participation, and promote a new important dialogue in children’s health.

9 Summary

Poor cardiovascular fitness is increasingly common in today’s children and increases the risk of mortality, cardiovascular disease, hypertension, and Type 2 diabetes. A key factor linking poor fitness, obesity, and risk for early onset of these adverse health outcomes is the development of IR. Efforts to address the morbidity of childhood obesity should therefore focus on multiple pathways toward reducing IR in children, must be multidisciplinary in approach (i.e. increasing fitness and physical activity and reducing obesity), and be pervasive in multiple layers of children’s environment and persistent throughout the year. In other words, the public health interventions needed to curb, and ultimately reverse the childhood obesity and poor fitness epidemic will include changes to the built environment, the nutritional environment, and the social-behavioral environment. Practical methods for fitness assessment outside of the laboratory will be critical in addressing this challenge.

Copyright information

© Springer Science+Business Media, LLC 2009