Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
May 2008

Intra-abdominal Adiposity and Individual Components of the Metabolic Syndrome in AdolescenceSex Differences and Underlying Mechanisms

Author Affiliations

Author Affiliations: Brain and Body Centre, University of Nottingham, Nottingham, England (Ms Syme, Drs Pitiot, Paus, and Pausova, and Messrs Totman and Qiu); and Department of Epidemiology and Biostatistics, McGill University (Drs Abrahamowicz, Leonard, and Paus and Ms Xiao) and Research Centre–CHUM (Centre hospitalier de l'Université de Montréal) (Dr Pausova), Montreal; Groupe ÉCOBES (d'étude des conditions de vie et des besoins de la population) (Drs Perron and Veillette), Université de Québec (Dr Richer), and Community Genomic Centre, Université de Montréal, Chicoutimi Hospital (Dr Gaudet), Chicoutimi, Quebec, Canada.


Copyright 2008 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2008

Arch Pediatr Adolesc Med. 2008;162(5):453-461. doi:10.1001/archpedi.162.5.453

Objective  To investigate the association between intra-abdominal adiposity and individual components of the metabolic syndrome (MS) in adolescent males and females.

Design  Cross-sectional study of a population-based cohort.

Setting  Saguenay Youth Study, Quebec, Canada.

Participants  A total of 324 adolescents, aged 12 to 18 years.

Intervention  Measures were compared between males and females with “high” or “low” intra-abdominal fat (IAF).

Main Outcome Measures  Intra-abdominal fat was quantified with magnetic resonance imaging. Primary outcome measures were blood pressure (BP) and fasting serum glucose, insulin, lipids, and C-reactive protein levels. Secondary mechanistic measures were cardiovascular variability indexes of autonomic nervous system function, pubertal development, and serum levels of cortisol, leptin, and sex hormones.

Results  The MS was completely absent in adolescents with low IAF and was present in 13.8% of males and 8.3% of females with high IAF. Excess IAF was associated with a higher homeostasis model assessment index (0.5 [95% confidence interval (CI), 0.3 to 0.8]; P < .001) and triglycerides level (17.7 mg/dL [to convert to millimoles per liter, multiply by 0.0113] [95% CI, 9.7 to 25.7 mg/dL]; P < .001), lower high-density lipoprotein cholesterol level (−3.9 mg/dL [to convert to millimoles per liter, multiply by 0.0259] [95% CI, −6.2 to −1.5 mg/dL]; P = .003), and higher C-reactive protein level (0.03 mg/L [to convert to nanomoles per liter, multiply by 9.524] [95% CI, 0.01 to 0.05 mg/L]; P = .003). High IAF was associated with elevations of BP and sympathetic activity in males only (higher systolic BP, 6 mm Hg [95% CI, 1 to 11 mm Hg]; P = .02 and low-frequency power of diastolic BP, 629 mm Hg2 [95% CI, 37 to 1222 mm Hg2]; P = .04).

Conclusions  Our results suggest that, already in adolescence, accumulation of IAF may promote development of the MS, affecting the metabolic and inflammatory components similarly in both sexes but influencing BP adversely only in males. The latter may be attributed, in part, to the augmentation of sympathetic activity also seen only in males.

Obesity is a leading cause of the metabolic syndrome (MS), defined by the co-occurrence of intra-abdominal obesity, atherogenic dyslipidemia, raised blood pressure (BP), insulin resistance and/or glucose intolerance, and a proinflammatory state.1 In the industrialized world, its prevalence is increasing. Based on criteria proposed by the World Health Organization2 and the Third Report of the National Cholesterol Education Program's Adult Treatment Panel,3 an estimated 47 million individuals in the United States have the syndrome.4 Moreover, the syndrome, typically regarded as a middle- to late-adulthood disorder, is now present in childhood and adolescence. Population-weighted estimates from a study conducted between 1988-1994 indicate that, in the United States, nearly 30% of obese adolescents have the MS.5

It has long been recognized that upper-body compared with lower-body obesity is more closely associated with cardiovascular and metabolic abnormalities of the MS.6,7 More recently, this difference has been related to the increased quantity of intra-abdominal fat (IAF) that is frequently found in individuals with upper-body obesity.8 Underlying mechanisms linking intra-abdominal obesity to development of the MS are not well understood. Accumulation of visceral fat, characterized by a relatively high lipid turnover, may result in higher levels of free fatty acids in the portal circulation.9 This, in turn, may contribute to the development of individual components of the MS via, for example, enhanced lipid synthesis, gluconeogenesis, and insulin resistance.8,10,11 Furthermore, IAF correlates positively with activation of the sympathetic nervous system,12 which may further enhance free fatty acid release into portal circulation.13 Sympathoactivation may also contribute to the elevation of BP through its effects on vasculature and renal handling of sodium and water.14,15

The aim of the present study was to investigate the impact of IAF, assessed with magnetic resonance imaging (MRI), on individual components of the MS in a population-based cohort of 324 adolescent males and females. We also examined whether the autonomic nervous system, assessed with power spectral analysis of beat-to-beat BP and interbeat interval (IBI),16 contributes to these effects.


Adolescents, aged 12 to 18 years, were recruited in a remote, French-Canadian population as part of the Saguenay Youth Study; all subjects were white. This is an ongoing, cross-sectional, family-based (adolescent sibships) investigation of the long-term consequences of prenatal exposure to maternal cigarette smoking (PEMCS) on cardiovascular and metabolic health and on the brain and behavior in adolescence; details on subject ascertainment are described elsewhere.17 Adolescence was chosen as a period when initial stages of cardiovascular and metabolic abnormalities may become apparent and are not yet altered by confounding variables such as medication. The outcomes studied in this article were not prespecified but were included in the Saguenay Youth Study protocol from its outset. Because PEMCS has been implicated in increasing the risk for obesity,18,19 we included PEMCS as a covariate in all analyses. The Research Ethics Committee of the Chicoutimi Hospital approved the study protocol.


The current sample consists of 324 subjects recruited and tested between November 2003 and December 2006. At the time of analysis, 408 subjects had undergone the study protocol, but 63 subjects were excluded because of technical issues, 17 subjects did not or could not complete the protocol (eg, following dizziness due to postural hypotension or need to urinate), and 2 subjects had poor-quality recordings throughout the protocol.


Quantity and distribution of body fat were assessed with (1) MRI, (2) anthropometry, and (3) bioelectrical impedance. (1) Ten axial slices, 10-mm thick, were acquired in a Phillips 1.0-T magnetic resonance scanner. Adipose tissue was imaged with a heavily T1-weighted, single breath-hold spin-echo sequence. A single slice at the level of the umbilicus was selected for quantification of abdominal fat. Images were smoothed using an adaptive bilateral filter to remove image noise while preserving edge information.20 An initial fat classification map was obtained using a standard region-growing algorithm. An iterative refinement procedure corrected false positives and false negatives using a battery of morphological operators, including hysteresis, thresholding over small neighborhoods,21 and median filtering to remove salt and pepper noise. The resulting classification map was manually segmented into subcutaneous abdominal fat (SAF) and IAF. Subcutaneous abdominal fat was defined as areas of adipose tissue lying between the skin surface and the outer aspect of the musculature of the abdominal cavity, and IAF was defined as adipose tissue lying within the innermost aspect of the abdominal cavity and not contained within other abdominal organs or muscles. A histogram counting algorithm computed the total number of voxels for each type of fat. This semiautomated method was validated against manual segmentation in 20 randomly selected subjects (SAF, r2 = 0.99; IAF, r2 = 0.97). (2) Trained nurses measured weight (0.1-kg precision), height (0.1-cm precision), waist and hip circumferences (0.1-cm precision), and suprailiac skinfold (1-mm precision) 3 times. Body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared. (3) Multifrequency bioimpedance analysis was used to estimate total body fat (Xitron Technologies, Inc, San Diego, California). Subjects were asked to refrain from caffeine, alcohol, and vigorous activity 24 hours before the measurement.


To evaluate energy intake, we used a standard 24-hour food recall, previously validated for Quebec adolescents.22 The physical activity questionnaire was used to assess the number of physical activity sessions (at least 20 minutes in duration) per week. The Puberty Development Scale used is an 8-item self-report measure of physical development based on Tanner stages with separate forms for males and females.23 Information on household income as an index of socioeconomic status (SES) was obtained from parents. Information on current smoking and life history of smoking was obtained from the adolescents. However, since the prevalence (< 15%) and dose (2.6 cigarettes/d) of cigarette smoking were rather low, and since subjects with “high” vs “low” IAF did not differ in this respect (P = .35), information on current smoking was not included in the analyses.


A fasting blood sample was drawn between 8 AM and 9 AM. The sample was evaluated for glucose and lipid metabolism (glucose [to convert milligrams per deciliter to millimoles per liter, multiply by 0.0555], insulin [to convert micro–international units to picomoles per liter, multiply by 6.945], triglycerides [to convert milligrams per deciliter to millimoles per liter, multiply by 0.0113], and high-density lipoprotein [HDL] cholesterol [to convert milligrams per deciliter to millimoles per liter, multiply by 0.0259] levels), a proinflammatory state (C-reactive protein [CRP] level [to convert milligrams per liter to nanomoles per liter, multiply by 9.524), hypothalamic-pituitary-adrenal axis activity (cortisol level [to convert micrograms per deciliter to nanomoles per liter, multiply by 27.588), sexual maturation (total testosterone [to convert nanograms per deciliter to nanomoles per liter, multiply by 0.0347], bioavailable testosterone [to convert nanograms per deciliter to nanomoles per liter, multiply by 0.0347],24 and estradiol [to convert picograms per milliliter to picomoles per liter, multiply by 3.671] levels), and leptin level (Human Leptin RIA kit; Linco Research, Inc, St Charles, Missouri). Homeostasis model assessment (HOMA), an index of insulin resistance, was calculated.25


All subjects underwent a 53-minute cardiovascular protocol involving simple physical and mental challenges intended to evoke cardiovascular responses. The physical challenge was a change in posture from supine (10 minutes) to standing (10 minutes) and from standing to sitting (10 minutes). The mental challenge was a sequence of 46 simple arithmetic problems to be solved aloud. The level of difficulty progressively increased to ensure some failure for all subjects.26 Throughout the protocol, conducted in a hospital setting on Saturdays between 8 AM and 12 PM, a noninvasive hemodynamic monitor (Finometer; FMS Finapres, Amsterdam, the Netherlands) was used to continuously record finger blood flow. The Finometer derives beat-to-beat brachial systolic BP (SBP) and diastolic BP (DBP), IBI, and its inverse, heart rate (HR).


Power spectral analysis of IBI and DBP is a well-established means of estimating cardiovascular autonomic nervous system activity.16,27 Power spectral analysis of IBI and DBP was carried out for 8 two-minute periods of the cardiovascular protocol: supine, standing 1, standing 2, sitting 1, sitting 2, sitting 3 (pre–math test), math test, and post–math test. For each period, beat-to-beat time series of IBI and DBP were interpolated using a piecewise cubic spline method, resampled at a frequency of 5 Hz and detrended before being transformed by a 1024-point fast Fourier transform,28 using standard Matlab functions (Matlab 7.3.0; MathWorks, Inc, Natick, Massachusetts). Low-frequency (LF) and high-frequency (HF) spectral powers were determined by integrating the power spectrum between 0.04 and 0.15 Hz and between 0.15 and 0.4 Hz, respectively; HFIBI and LFDBP were considered proxies of parasympathetic27 and sympathetic nervous system activity,29,30 respectively.


Subjects were classified as having the MS if they had 3 or more of the following 5 conditions: (1) SBP and/or DBP in the 95th (age-, sex-, and height-specific) percentile or higher,31 using the US Centers for Disease Control and Prevention growth charts (; (2) BMI in the 95th (age- and sex-specific) percentile or higher of the Centers for Disease Control and Prevention BMI curves5; (3) fasting glucose level of 109.9 mg/dL32 and/or HOMA index of 4.39 or higher33; (4) triglycerides level of 109.7 mg/dL or higher5; and (5) HDL cholesterol level of 39.8 mg/dL or lower.5

In the 324 subjects, less than 1% of collected phenotypes were missing, except for food recall and percentage of body fat determined by bioelectrical impedance, which were missing with frequencies of 5.9% and 4.6%, respectively.


Subjects were divided into low and high IAF groups based on the sex-specific median of IAF. The median was used because, to our knowledge, there are no established cutoffs for high or low IAF. For single-valued outcomes, we used multivariable mixed linear models, which extend the conventional multiple linear regression by accounting for correlations due to clustering of 2 or 3 siblings within the same family and assume the exchangeable structure of the covariance matrix of residuals. Each mixed model included binary indicators of IAF, sex, and their interaction and a priori–selected potentially confounding variables: age, height, Tanner stage, PEMCS status, and household income as an indicator of SES. Because we considered the possibility that the impact of high IAF on some outcomes may depend on sex, in the case of statistically significant IAF × sex interactions, we performed mixed-model analyses stratified by sex to estimate sex-specific adjusted effects of IAF. Otherwise, the interaction was removed from the mixed model, and pooled effects of both IAF (adjusted for sex and all covariates) and sex (adjusted for IAF and all covariates) were estimated. For multivalued outcomes (53 one-minute means of SBP, DBP, and HR), a similar approach was used, but the mixed model additionally adjusted for time and accounted for potential correlation between consecutive observations on the same subject, assuming an autoregressive order 1 covariance structure. In all mixed-model analyses, hypotheses were tested using a 2-tailed Wald test, with 1 df, at the .05 significance level. Intra-abdominal fat and sex effects are reported as adjusted differences (and 95% confidence interval [CI]) between mean values in subjects with high vs low IAF and males vs females.

Our analysis of LFDBP and HFIBI, each calculated for 8 selected 2-minute periods, was based on an a priori assumption that IAF effects may change in a nonsystematic way from one period to another and/or may be limited to only a few periods. This assumption stems from the fact that sympathetic and parasympathetic nervous system activities change rapidly to meet the demands made on the heart and vasculature in response to physical and mental challenges, such as those presented in our protocol.34 Accordingly, LFDBP and HFIBI were analyzed separately for each period. We used a multiple linear regression model including IAF, sex, their interaction, and the covariates mentioned earlier. In sex-stratified and sex- or IAF-pooled analyses, the statistical significance of the adjusted IAF effect was tested using a model-based partial F test (with 1 df in the numerator and the denominator df corresponding to N−k−1, where N = sample size and k = number of independent variables).

In addition, we performed a posteriori power estimations to obtain the mean differences detectable with 80% power by a 2-tailed, independent-groups t test. Using the median split, we had sufficient but not excessive power to detect clinically meaningful differences, such as a reduction in SBP of 4 mm Hg.35 Mean detectable differences for other outcomes are listed in the Table legends. Statistical analyses were performed with the SAS statistical software package (SAS Institute Inc, Cary, North Carolina).


Males and females were classified as having high or low IAF based on the median IAF: 14 050 mm3 in males and 19 507 mm3 in females. Subjects with high vs low IAF (both males and females) were significantly older, taller, and more advanced in terms of sexual maturation (Table 1) but did not differ by SES, PEMCS, physical activity, or food intake (Table 1).

Table 1. 
Image not available
Comparison of the Characteristics of Adolescent Males and Females With “High” vs “Low” IAFa

As expected, subjects with high vs low IAF demonstrated higher values of all indexes of adiposity (Table 2). The difference in IAF only was greater in males than females (37 183 mm3 [95% CI, 29 370 to 44 995 mm3]; P < .001 and 28 076 mm3 [95% CI, 23 047 to 33 105 mm3]; P < .001, respectively; IAF × sex interaction, P = .04).

Table 2. 
Image not available
Comparison of Body-Fat Quantity and Distribution in Adolescent Males and Females With “High” vs “Low” IAFa

In both males and females, subjects with high vs low IAF had higher levels of fasting insulin and triglycerides and HOMA index and lower levels of HDL cholesterol (Table 3). The values of insulin, HOMA index, and triglycerides were higher by 20% to 30% and those of HDL cholesterol were lower by 7%. When analyzing SBP measured at rest while seated, we found a significant IAF × sex interaction (P = .01). Males with high vs low IAF had significantly higher SBP (6 mm Hg [95% CI, 1 to 11 mm Hg]; P = .02), but there was no difference in females (−1 mm Hg [95% CI, −5 to 2 mm Hg]; P = .45) (Table 3). Intra-abdominal fat showed no significant effect on fasting glucose level or DBP (measured at rest while seated) in either males or females (Table 3). With respect to the MS-related variables, in both males and females, subjects with high vs low IAF showed higher CRP levels and no significant difference in morning cortisol, total testosterone, and estradiol levels (Table 3). In addition, subjects with high vs low IAF demonstrated higher leptin levels, but the difference was significantly greater (IAF × sex interaction, P < .001) in females (8.7 ng/mL [95% CI, 6.2 to 11.1 ng/mL]; P < .001) than in males (2.8 ng/mL [95% CI, 1.9 to 3.7 ng/mL]; P < .001) (Table 3). Subjects with high vs low IAF also exhibited higher bioavailable testosterone levels, and this difference (IAF × sex interaction, P = .008) was greater in males (34.7 ng/dL [95% CI, 0 to 72.3 ng/dL]; P = .06) than in females (2.9 ng/dL [95% CI, 0 to 5.8 ng/dL]; P = .06) (Table 3).

Table 3. 
Image not available
Impact of Intra-abdominal Adiposity on Individual Components of the Metabolic Syndrome and Related Variables in Adolescent Males and Femalesa

The Figure shows 53 one-minute means of SBP, DBP, and HR calculated throughout the entire protocol and 8 two-minute calculations of LFDBP and HFIBI computed for specific periods. When analyzing these measures, SBP was significantly higher in males with high vs low IAF but not females (IAF × sex interaction, P < .001) (Figure). In males, the mean estimated difference between the 2 groups over the entire protocol was 5 mm Hg (95% CI, 3 to 8 mm Hg; P < .001), reaching up to 9 mm Hg during both physical and mental challenges (Figure). A similar but less pronounced pattern was also observed for DBP (IAF × sex interaction, P = .009); in males, the estimated mean difference was 2 mm Hg (95% CI, 0 to 4 mm Hg; P = .02), exceeding 3 mm Hg during and in between the challenges (Figure). No significant effect of intra-abdominal adiposity on HR was observed in either males or females (Figure).

Image not available

Cardiovascular and autonomic nervous system at rest and in response to postural and mental challenges in males and females with “high” or “low” intra-abdominal fat (IAF). One-minute means of beat-to-beat systolic blood pressure (SBP), diastolic blood pressure (DBP), and heart rate (HR) during a 53-minute cardiovascular protocol, including a posture test (minutes 1-30) and math stress test (minutes 31-53), are shown. In addition, indexes of sympathetic and parasympathetic activity (low-frequency power of DBP [LFDBP] and high-frequency power of interbeat interval [HFIBI], respectively) are presented for 8 two-minute periods: supine (minutes 9-10), standing 1 (minutes 11-12), standing 2 (minutes 17-18), sitting 1 (minutes 21-22), sitting 2 (minutes 27-28), sitting 3 (pre–math test, minutes 34-35), math test (minutes 42-43), and post–math test (minutes 49-50). The data are all adjusted for potential confounding variables, including age, height, Tanner stage, household income, and prenatal exposure to maternal cigarette smoking. They are presented as mean (SEM); solid squares indicate the mean for individuals with high IAF and open squares show the mean for individuals with low IAF. The dashed vertical line indicates the end of the posture test. The 53 one-minute means of SBP, DBP, and HR were analyzed with mixed-model regression analyses. Statistics show the P value of the IAF × sex interaction and the estimated difference (95% confidence interval [CI]) and P value for high vs low IAF in males (A) and in females (B). The 8 estimates of sympathetic (LFDBP) and parasympathetic (HFIBI) nervous system functions, calculated for specific 2-minute periods, were analyzed using multiple linear regression analysis. Statistical significance for a difference between males with high and low IAF is indicated with * (P = .04).

Multiple linear regression analyses of LFDBP, a surrogate for sympathetic activity,34,35 revealed significant IAF × sex interactions during the first and second sitting periods (P = .04 for both), when males with high vs low IAF demonstrated higher LFDBP (Figure). The adjusted difference was significant for the second sitting period (629 mm Hg2 [95% CI, 37 to 1222 mm Hg2]; P = .04) and marginally nonsignificant for the first sitting period (310 mm Hg2 [95% CI, −127 to 1098 mm Hg2]; P = .12). In contrast, the LFDBP differences between females with high and low IAF during both periods were nonsignificant (sitting 1, −194 mm Hg2 [95% CI, −567 to 178 mm Hg2]; P = .30; sitting 2, −132 mm Hg2 [95% CI, −652 to 389 mm Hg2]; P = .62). For other periods, the main effect of IAF on LFDBP was nonsignificant (P values ≥ .10) (Figure). High-frequency IBI, a proxy for parasympathetic activity,32 did not show any significant IAF × sex interactions (P values ≥ .30) (Figure) or differences between subjects with high and low IAF (P values ≥ .10) (Figure).


The MS was not present in any subjects with low IAF. Among subjects with high IAF, the MS (defined by the presence of at least 3 of 5 components) was found in 13.8% of males and 8.3% of females; this sex difference was not statistically significant. Since obesity is a component of the MS and, to a certain degree, was used to categorize subjects as having high or low IAF (BMI and IAF were correlated [males with high IAF, r2 = 0.56; females with high IAF, r2 = 0.56]), we also assessed the prevalence of the MS without including obesity as a component. This analysis showed that the MS (defined by the presence of at least 3 of 4 components) was present in 7.5% of males with high IAF and 5.9% of females with high IAF; again, the sex difference was not significant.


The present study is one of the first to examine the relationship between intra-abdominal obesity and the MS directly using MRI. With this method, we found that the MS was completely absent in adolescents with low IAF while it was present in 13.8% of males and 8.3% of females with high IAF. Overall, these results are consistent with recent population-based reports indicating that the MS, traditionally regarded as an adult disorder, is now emerging during adolescence.5,36 Our results suggest that intra-abdominal obesity may be a driving force behind its development.

In the present investigation, SBP and DBP were elevated in males with high vs low IAF. A similar pattern was also observed for LFDBP, an index of sympathetic activity, suggesting that sympathoactivation may be involved in the observed BP elevations. Consistently, IAF, but not SAF, has been previously related to increased sympathetic activity in adult males.12,37 Sympathoactivation is thought to be a key underlying mechanism of obesity-related hypertension38; augmented sympathetic outflow to the kidneys and vasculature can increase sodium and water reabsorption and peripheral vascular resistance, respectively, and hence BP. Previous research suggests that obesity-induced hyperleptinemia and hyperinsulinemia may play a role in sympathoactivation.3942 In our study, leptin and insulin serum levels were higher in subjects with high vs low IAF, with the former difference being greater in females than males and the latter difference being similar in both sexes. These results do not support the involvement of leptin and insulin in high IAF–associated sympathoactivation (and possibly BP) observed only in males.

Adolescence is a period of sexual development when a number of physiological differences between males and females emerge; one such difference is an elevation of BP in males vs females.43 The mechanisms of this difference are not clear at present, though a large body of research implicates sex hormones44 that change dramatically during adolescence. Males increase production of testosterone and females increase generation of estradiol prior to menarche, after which estradiol cycles with the menstrual cycle at levels constantly higher than in males. Long-term exposure to testosterone is thought to promote BP elevation; it may compromise renal function and impair vascular reactivity.45 Multiple abnormalities, including sympathoactivation, have been implicated in these effects. Testosterone enhances vasoconstriction in response to adrenergic agonists.46 In addition, levels of plasma norepinephrine, which is primarily derived from sympathetic nerve endings, increase in males with advancing puberty and testosterone levels.47 In the current study, we observed that bioavailable testosterone levels significantly increased in subjects with high vs low IAF, with the difference being greater in males than females. These results suggest that testosterone could play a role in sympathoactivation and BP elevation observed in males with high compared with low IAF. No significant differences between subjects with high and low IAF were identified in estradiol, which is thought to be cardioprotective.44

In males, we saw that excess IAF was associated with increased BP throughout the entire cardiovascular protocol, whereas augmented sympathetic activity was observed only during recovery from a physical challenge (10-minute standing). These results suggest that excess IAF affects the sympathetic nervous system mainly by diminishing its expected withdrawal after a physical challenge (as seen in females) (Figure). This finding may be important, as clinical studies have found that delayed cardiovascular recovery from exercise is a significant predictor of mortality in subjects with cardiovascular disease.48,49

Based on our current data, we cannot determine the exact causes of the differences in IAF quantity between subjects with high and low IAF, as they did not differ in food intake or physical activity. We speculate they may vary in metabolic rate and/or genetic predisposition for body fat distribution (eg, intra-abdominal rather than subcutaneous). It is also possible that subjects with higher IAF were less active and/or had higher energy intake in the past or that a false-negative finding occurred because of a measurement error.

This study has certain limitations, including (1) a cross-sectional design, (2) the use of power spectral analysis of cardiovascular variability to assess autonomic nervous system function, (3) a lack of dose-response analyses, and (4) a potential for type I error inflation. (1) A longitudinal design would facilitate examination of a causal relationship between, eg, age-related changes in IAF, SBP, and LFDBP. Usefulness of cross-sectional studies, however, should not be underestimated, as they have generated many clinically highly relevant findings (eg, the National Health and Nutrition Examination Survey III5 and the initial stages of the Framingham Study50). (2) Power spectral analysis of cardiovascular variability is a method that assesses autonomic nervous system function in a noninvasive albeit indirect fashion. Direct methods, such as muscle sympathetic nerve activity recording and ganglionic blockade, are invasive and thus much less suitable for population-based studies of adolescents. Importantly, power spectral analysis of cardiovascular variability has been validated against these methods16,51,52 and its use in clinical practice is being debated.51,53 (3) Further studies should assess continuous dose-response relationships between IAF and different components of the MS. These studies will have to be carried out in larger samples and, if possible, designed specifically for this type of investigation. (4) The total number of results reported in Table 2, Table 3, and the Figure (primary and secondary mechanistic outcomes) equals 80. About 4 tests would be expected to yield P < .05 by chance alone (80 × 0.05 = 4), but we found 18 significant results. Although these calculations are only approximate, as they assume independence of the 80 tests, they clearly indicate that a vast majority of the significant associations reported reflect genuine effects of IAF rather than type I errors.

In summary, our results suggest that intra-abdominal obesity is associated with adverse cardiovascular and metabolic consequences in male and female adolescents. Already in this age category, it appears to increase the risk for the MS, affecting insulin resistance, dyslipidemia, and a proinflammatory state similarly in males and females but influencing BP adversely only in males. The latter may be related in part to an intra-abdominal obesity–induced augmentation in sympathetic activity also seen only in males.

Back to top
Article Information

Correspondence: Zdenka Pausova, MD, Brain and Body Centre, University of Nottingham, University Park, Nottingham NG7 2RD, England (

Accepted for Publication: October 11, 2007.

Author Contributions: Dr Pausova had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Leonard, Veillette, Gaudet, Paus, and Pausova. Acquisition of data: Syme, Leonard, Perron, Richer, Gaudet, and Pausova. Analysis and interpretation of data: Syme, Abrahamowicz, Pitiot, Qiu, Totman, Xiao, and Pausova. Drafting of the manuscript: Syme, Abrahamowicz, Pitiot, and Pausova. Critical revision of the manuscript for important intellectual content: Abrahamowicz, Leonard, Perron, Qiu, Richer, Totman, Veillette, Xiao, Gaudet, Paus, and Pausova. Statistical analysis: Syme, Abrahamowicz, Xiao, and Pausova. Obtained funding: Paus and Pausova. Administrative, technical, and material support: Syme, Leonard, Perron, Pitiot, Qiu, Richer, Totman, Veillette, Paus, and Pausova. Study supervision: Syme and Pausova.

Financial Disclosure: None reported.

Funding/Support: The Saguenay Youth Study project is funded by the Canadian Institutes of Health Research (Drs Pausova and Paus), Heart and Stroke Foundation of Quebec (Dr Pausova), the Canadian Foundation for Innovation (Dr Pausova), and the University of Nottingham (Drs Pausova and Paus). Dr Abrahamowicz is a James McGill Professor of Biostatistics at McGill University.

Additional Contributions: The following individuals helped design the protocol and acquire and analyze the data (* indicates paid staff): database architect: Manon Bernard, MSc,* Research Centre–CHUM; MR team: Michel Bérubé, MD,* and radiographers Sylvie Masson, Suzanne Castonguay, Julien Grandisson, and Marie-Josée Morin, Saguenay Hospital; cardiovascular nurses: Jessica Blackburn, RN,* Mélanie Gagné, RN,* Jeannine Landry, RN,* Catherine Lavoie, RN,* Lisa Pageau, RN,* Réjean Savard, RN,* France Tremblay, RN,* and Jacynthe Tremblay, RN,* Saguenay Hospital; psychometricians: Chantale Belleau, MA,* Mélanie Drolet, MA,* Catherine Harvey, MA,* Stéphane Jean, MA,* Hélène Simard, MA,* and Mélanie Tremblay, MA,* ÉCOBES; ÉCOBES team: Nadine Arbour, MEd,* Julie Auclair, MSc,* Marie-Ève Blackburn, BA,* Marie-Ève Bouchard,* Annie Houde, BSc,* Catherine Lavoie,* and Luc Laberge, PhD*; nutritionists: Caroline Benoit, RN,* and Henriette Langlais, RN,*Saguenay Hospital; and laboratory technicians: Denise Morin, MSc,* and Nadia Mior, MSc,* Saguenay Hospital. In addition, we would like to acknowledge the contributions of Jean Mathieu, MD, Saguenay Hospital, Julie Bérubé, MSc,* and Celine Bourdon, MSc,* Research Centre–CHUM, Bruce Pike, PhD, Rosanne Aleong, MSc, Jennifer Barrett, PhD, Candice Cartier, BA,* and Valerie Legge, BA,* McGill University, Dale Einarson, BSc,* and François Morvillier, MSc, University of Nottingham, and Helena Jelicic, PhD, Tufts University. Ian Macdonald, PhD, University of Nottingham, critically read the manuscript.

Grundy  SMBrewer  HB  JrCleeman  JISmith  SC  JrLenfant  C Definition of metabolic syndrome: report of the National Heart, Lung, and Blood Institute/American Heart Association conference on scientific issues related to definition. Arterioscler Thromb Vasc Biol 2004;24 (2) e13- e18
Alberti  KGZimmet  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. Diabet Med 1998;15 (7) 539- 553
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 (19) 2486- 2497
Ford  ESGiles  WHDietz  WH Prevalence of the metabolic syndrome among US adults: findings from the third National Health and Nutrition Examination Survey. JAMA 2002;287 (3) 356- 359
Cook  SWeitzman  MAuinger  PNguyen  MDietz  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 (8) 821- 827
Vague  J The degree of masculine differentiation of obesities: a factor determining predisposition to diabetes, atherosclerosis, gout, and uric calculous disease. Am J Clin Nutr 1956;4 (1) 20- 34
Kaplan  NM The deadly quartet: upper-body obesity, glucose intolerance, hypertriglyceridemia, and hypertension. Arch Intern Med 1989;149 (7) 1514- 1520
Kanai  HMatsuzawa  YKotani  K  et al.  Close correlation of intra-abdominal fat accumulation to hypertension in obese women. Hypertension 1990;16 (5) 484- 490
Björntorp  P The regulation of adipose tissue distribution in humans. Int J Obes Relat Metab Disord 1996;20 (4) 291- 302
Goodfriend  TLKelley  DEGoodpaster  BHWinters  SJ Visceral obesity and insulin resistance are associated with plasma aldosterone levels in women. Obes Res 1999;7 (4) 355- 362
Ritchie  SAConnell  JM The link between abdominal obesity, metabolic syndrome and cardiovascular disease. Nutr Metab Cardiovasc Dis 2007;17 (4) 319- 326
Alvarez  GEBeske  SDBallard  TPDavy  KP Sympathetic neural activation in visceral obesity. Circulation 2002;106 (20) 2533- 2536
Lönnqvist  FThorne  ALarge  VArner  P Sex differences in visceral fat lipolysis and metabolic complications of obesity. Arterioscler Thromb Vasc Biol 1997;17 (7) 1472- 1480
Hall  JE Pathophysiology of obesity hypertension. Curr Hypertens Rep 2000;2 (2) 139- 147
Huggett  RJBurns  JMackintosh  AFMary  DA Sympathetic neural activation in nondiabetic metabolic syndrome and its further augmentation by hypertension. Hypertension 2004;44 (6) 847- 852
Pagani  MMontano  NPorta  A  et al.  Relationship between spectral components of cardiovascular variabilities and direct measures of muscle sympathetic nerve activity in humans. Circulation 1997;95 (6) 1441- 1448
Pausova  ZPaus  TAbrahamowicz  M  et al.  Genes, maternal smoking, and the offspring brain and body during adolescence: design of the Saguenay Youth Study. Hum Brain Mapp 2007;28 (6) 502- 518
Al Mamun  ALawlor  DAAlati  RO'Callaghan  MJWilliams  GMNajman  JM Does maternal smoking during pregnancy have a direct effect on future offspring obesity? Evidence from a prospective birth cohort study. Am J Epidemiol 2006;164 (4) 317- 325
Power  CJefferis  BJ Fetal environment and subsequent obesity: a study of maternal smoking. Int J Epidemiol 2002;31 (2) 413- 419
Tomasi  CManduchi  R Bilateral filtering for gray and color images.  Proceedings of the Sixth International Conference on Computer Vision January 4-7, 1998 Bombay, India
Ilea  DGhita  ORobinson  K  et al.  Identification of body fat tissues in MRI data.  Paper presented at: Optimization of Electrical and Electronic Equipment May 13-14, 2004 Brasov, Romania
Berthiaume  PLavalee  CVilleneuve  MVigneaut  M Enquête sociale et de santé auprès des enfants et des adolescents québécois: volet nutrition.  Quebec, QC, Canada Les Publications du Québec2004;19- 33
Peterson  ACrockett  LRichards  MBoxer  A A self-report measure of pubertal status. J Youth Adolesc 1988;17 (2) 117- 133Article
Södergård  RBackstrom  TShanbhag  VCarstensen  H Calculation of free and bound fractions of testosterone and estradiol-17 beta to human plasma proteins at body temperature. J Steroid Biochem 1982;16 (6) 801- 810
Matthews  DRHosker  JPRudenski  ASNaylor  BATreacher  DFTurner  RC Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985;28 (7) 412- 419
McAdoo  WGWeinberger  MHMiller  JZFineberg  NSGrim  CE Race and gender influence hemodynamic responses to psychological and physical stimuli. J Hypertens 1990;8 (10) 961- 967
 Heart rate variability: standards of measurement, physiological interpretation, and clinical use. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Eur Heart J 1996;17 (3) 354- 381
Prakash  ESMadanmohanSethuraman  KRNarayan  SK Cardiovascular autonomic regulation in subjects with normal blood pressure, high-normal blood pressure and recent-onset hypertension. Clin Exp Pharmacol Physiol 2005;32 (5-6) 488- 494
Pagani  MLombardi  FGuzzetti  S  et al.  Power spectral analysis of heart rate and arterial pressure variabilities as a marker of sympatho-vagal interaction in man and conscious dog. Circ Res 1986;59 (2) 178- 193
Shin  KMinamitani  HOnishi  SYamazaki  HLee  M Assessment of training-induced autonomic adaptations in athletes with spectral analysis of cardiovascular variability signals. Jpn J Physiol 1995;45 (6) 1053- 1069
 The fourth report on the diagnosis, evaluation, and treatment of high blood pressure in children and adolescents. Pediatrics 2004;114 (2) ((suppl 4th report)) 555- 576
Fagot-Campagna  ASaaddine  JBFlegal  KMBeckles  GL Diabetes, impaired fasting glucose, and elevated HbA1c in US adolescents: the Third National Health and Nutrition Examination Survey. Diabetes Care 2001;24 (5) 834- 837
Lee  JMOkumura  MJDavis  MMHerman  WHGurney  JG Prevalence and determinants of insulin resistance among US adolescents: a population-based study. Diabetes Care 2006;29 (11) 2427- 2432
Ravits  JM AAEM minimonograph #48: autonomic nervous system testing. Muscle Nerve 1997;20 (8) 919- 937
Food and Drug Administration, Clinical evaluation of antihypertensive drugs: clinical medical draft guidelines.  Rockville, MD FDA1998;
Weiss  RDziura  JBurgert  TS  et al.  Obesity and the metabolic syndrome in children and adolescents. N Engl J Med 2004;350 (23) 2362- 2374
Alvarez  GEBallard  TPBeske  SDDavy  KP Subcutaneous obesity is not associated with sympathetic neural activation. Am J Physiol Heart Circ Physiol 2004;287 (1) H414- H418
Esler  MStraznicky  NEikelis  NMasuo  KLambert  GLambert  E Mechanisms of sympathetic activation in obesity-related hypertension. Hypertension 2006;48 (5) 787- 796
Correia  MLHaynes  WGRahmouni  KMorgan  DASivitz  WIMark  AL The concept of selective leptin resistance. Diabetes 2002;51 (2) 439- 442
Muntzel  MBeltz  TMark  ALJohnson  AK Anteroventral third ventricle lesions abolish lumbar sympathetic responses to insulin. Hypertension 1994;23 (6, pt 2) 1059- 1062
Muntzel  MSMorgan  DAMark  ALJohnson  AK Intracerebroventricular insulin produces nonuniform regional increases in sympathetic nerve activity. Am J Physiol 1994;267 (5, pt 2) R1350- R1355
Rahmouni  KCorreia  MLHaynes  WGMark  AL Obesity-associated hypertension: new insights into mechanisms. Hypertension 2005;45 (1) 9- 14
Dasgupta  KO'Loughlin  JChen  S  et al.  Emergence of sex differences in prevalence of high systolic blood pressure. Circulation 2006;114 (24) 2663- 2670
Reckelhoff  JF Sex steroids, cardiovascular disease, and hypertension: unanswered questions and some speculations. Hypertension 2005;45 (2) 170- 174
Iliescu  RReckelhoff  JF Testosterone and vascular reactivity. Clin Sci (Lond) 2006;111 (4) 251- 252
Malkin  CJJones  RDJones  THChanner  KS Effect of testosterone on ex vivo vascular reactivity in man. Clin Sci (Lond) 2006;111 (4) 265- 274
Weise  MEisenhofer  GMerke  DP Pubertal and gender-related changes in the sympathoadrenal system in healthy children. J Clin Endocrinol Metab 2002;87 (11) 5038- 5043
Cole  CRBlackstone  EHPashkow  FJSnader  CELauer  MS Heart-rate recovery immediately after exercise as a predictor of mortality. N Engl J Med 1999;341 (18) 1351- 1357
Nishime  EOCole  CRBlackstone  EHPashkow  FJLauer  MS Heart rate recovery and treadmill exercise score as predictors of mortality in patients referred for exercise ECG. JAMA 2000;284 (11) 1392- 1398
Kannel  WBBrand  NSkinner  JJ  JrDawber  TR McNamara  PM The relation of adiposity to blood pressure and development of hypertension: the Framingham study. Ann Intern Med 1967;67 (1) 48- 59
Parati  GMancia  GDi Rienzo  MCastiglioni  P Point: cardiovascular variability is/is not an index of autonomic control of circulation. J Appl Physiol 2006;101 (2) 676- 678, discussion 681-672
Zhang  RIwasaki  KZuckerman  JHBehbehani  KCrandall  CGLevine  BD Mechanism of blood pressure and R-R variability: insights from ganglion blockade in humans. J Physiol 2002;543 (pt 1) 337- 348
Taylor  JAStudinger  P Counterpoint: cardiovascular variability is not an index of autonomic control of the circulation. J Appl Physiol 2006;101 (2) 678- 681, discussion 681