Eur J Pediatr DOI 10.1007/s00431-015-2577-6

ORIGINAL ARTICLE

Changes in whole-body fat distribution, intrahepatic lipids, and insulin resistance of obese adolescents during a low-level lifestyle intervention Fabian Springer 1 & Verena Ballweg 1 & Roland Schweizer 2 & Fritz Schick 1 & Michael B. Ranke 2 & Gerhard Binder 2 & Stefan Ehehalt 2,3 & for the DISKUS Study Group

Received: 6 January 2015 / Revised: 4 May 2015 / Accepted: 2 June 2015 # Springer-Verlag Berlin Heidelberg 2015

Abstract The aim of this study was to analyze changes in adipose tissue (AT) distribution, intrahepatic lipids (IHL), and insulin resistance (IR) among a group of obese adolescents undergoing a 7-months low-level lifestyle intervention. Thirty-nine obese Caucasian adolescents (mean age 13.9 years, body mass index standard deviation score (BMISDSLMS) 2.14) were included. AT and IHL were determined by T1-weighted magnetic resonance (MR) imaging and single-voxel MR spectroscopy; IR was estimated using the homeostatic model assessment (HOMA-IR). The lifestyle intervention led to a reduction of both BMI-SDSLMS (boys 2.27 to 2.17; girls 2.00 to 1.82) and HOMA-IR (boys 6.1 to 4.4 (p= 0.008); girls 6.2 to 4.7 (p=0.030)). IHL dropped in both genders (boys 7.5 to 4.3 %; girls 4.6 to 3.4 %) positively correlating with HOMA-IR (boys r=0.52; girls r=0.68), while in contrast visceral AT did not change significantly. Conclusions: Although the lifestyle intervention only slightly reduced BMI-SDSLMS, insulin sensitivity improved in both genders and came along with a marked reduction of

IHL. This suggests that IHL might play the dominant role regarding insulin resistance in the youth, especially if compared to other AT compartments such as visceral AT. What is Known: • MR imaging/spectroscopy can be used to evaluate body fat distribution and intrahepatic lipids in the youth. • The strength of associations between body fat compartments and insulin resistance is under scientific debate. What is New: • The study emphasizes that even a low-level lifestyle intervention has a beneficial effect. • The study suggests that intrahepatic lipids are an important factor in the development of insulin resistance.

Keywords Adolescence . Obesity . Insulin resistance . Intrahepatic lipids . Adipose tissue distribution

Communicated by Peter de Winter * Fabian Springer [email protected] Verena Ballweg [email protected]

Stefan Ehehalt [email protected] 1

Section on Experimental Radiology, Department of Diagnostic and Interventional Radiology, University Hospital Tübingen, Hoppe-Seyler-Str.3, 72076 Tübingen, Germany

2

Paediatric Endocrinology and Diabetes, University Children’s Hospital Tübingen, Hoppe-Seyler-Str.1, 72076 Tübingen, Germany

3

Public Health Department of Stuttgart, Department of Pediatrics, Dental Health Care, Health Promotion, and Social Services, Schloßstr.91, 70176 Stuttgart, Germany

Roland Schweizer [email protected] Fritz Schick [email protected] Michael B. Ranke [email protected] Gerhard Binder [email protected]

Eur J Pediatr

Abbreviations AT ATLE ATUE BMI BMI-SDSLMS CT DEXA DISKUS

HOMA-IR IHL IR MR MRI MRS NATLE NATUE SCAT SCATTR STEAM TAT TBV TE TNAT TM TR VAT WHR

Adipose tissue Adipose tissue of lower extremities Adipose tissue of upper extremities Body mass index Body mass index standard deviation score Computed tomography Dual-energy X-ray absorptiometry Abbreviation formed from the initial components of the German phrase “Dick Sein im Kindes- und Jugendalter Studie”, which translates as Bbeing obese during childhood and adolescence study^ in English Homeostatic model assessment of insulin resistance Intrahepatic lipids Insulin resistance Magnetic resonance Magnetic resonance imaging Single-voxel magnetic resonance 1 H-spectroscopy Non-adipose tissue of lower extremities Non-adipose tissue of upper extremities Subcutaneous adipose tissue Subcutaneous adipose tissue of the body trunk STimulated Echo Acquisition Mode in magnetic resonance spectroscopy Total adipose tissue Total body volume Echo time Total non-adipose tissue Mixing time Repetition time Visceral adipose tissue Waist-to-hip ratio

Introduction In the last decades, the worldwide increasing incidence of obesity and associated metabolic disturbances such as insulin resistance (IR) and type 2 diabetes became an important issue not only in adults but also in adolescents [12]. In this context, it has been shown that the distribution of adipose tissue (AT) is of great importance: visceral adipose tissue (VAT) and intrahepatic lipids (IHL) are more related to insulin resistance than subcutaneous adipose tissue (SCAT) [13, 20, 27]. Especially, the high prevalence of IHL in adolescents [18] and its tendency to induce hepatic inflammation, with consecutive hepatic fibrosis and cirrhosis, is alarming [15]. The causes for the increasing incidence of obesity are related to

environmental changes with a more sedentary lifestyle and a nutrition comprising mainly saturated fatty acids and sugar. Furthermore, genetic and epigenetic factors play also an important role [9]. For evaluation of body fat content, various methods including body impedance measurement, underwater weighing, or dual-energy X-ray absorptiometry (DEXA) can be performed but provide only measures of total adipose tissue [3]. To visualize individual whole-body AT distribution, cross sectional imaging such as computed tomography (CT) or magnetic resonance imaging (MRI) has to be performed [13, 19, 20]. However, due to the high radiation exposure, CT imaging is mostly used as a single-slice modality and may not be applied in longitudinal studies during lifestyle intervention in children and adolescents. Thus, for reliable individual quantification of whole-body AT distribution, MRI is the method of choice, especially for children or adolescents. Furthermore, IHL can be accurately measured in a non-invasive way by single-voxel magnetic resonance 1H-spectroscopy (MRS) within the same magnetic resonance (MR) examination [13]. Outcome studies of lifestyle interventions including dietary regimes, physical exercise, education on nutrition, and habits previously showed that in adolescents the specific body fat compartments decrease differently and, thus, may also have a different influence on accompanying metabolic disorders [4, 16]. During lifestyle intervention, especially insulin resistance may be reversed even if body weight and/or total adipose tissue largely remain unaffected. For example, after a 12week exercise program, obese and lean Hispanic adolescents did not show a significant weight loss, but a reduction of IHL accompanied by an increase in insulin sensitivity [23]. In another study, this effect could also be observed after dietary nutrition [25]. Although some studies regarding the relation of body fat distribution and metabolic disturbances were published in the past, the relationship between individual body fat distribution and accumulation of ectopic fat within the liver is still unclear. In this context, we performed a detailed investigation on the changes of whole-body AT distribution (including AT of upper/lower extremities, body trunk, and visceral adipose tissue) and intrahepatic lipids during a low-level lifestyle intervention in adolescents using both MRI and MRS. Correlations between whole-body AT distribution as well as intrahepatic lipids to metabolic markers and insulin resistance were also evaluated.

Material and methods Study population Written informed consent was obtained from all participants and their legal guardians prior to inclusion in the study. The

Eur J Pediatr

study has been approved by the appropriate ethics committee and has therefore been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. In the performed observational, low-level lifestyle intervention study, a total of 39 adolescents (21 boys 13.8 (13.2, 14.5) years of age, median pubertal maturation stage: Tanner 3 (range 1 to 5); 18 girls 14.1 (13.6, 14.6) years of age, median pubertal maturation stage: Tanner 5 (range 3 to 5)) were examined at our obesity outpatient clinic. Each of them presented with a body mass index (BMI)>90th percentile for age [11]. Primary endpoint of this study was the age- and gender-matched BMI (BMI-SDS LMS ) after a period of 7 months low-level lifestyle intervention. Secondary endpoints were chosen to be body weight and BMI, waist and hip circumference, waist-to-hip ratio, total body volume (TBV) as well as the various (non-) adipose tissue compartments of the body, intrahepatic lipids (IHL), markers of insulin resistance, and the serum enzymes AST, ALT, and GGT. The low-level lifestyle intervention program consisted of (i) the advice to increase the mean self-reported physical activity by 1 to 2 h per day including all daily life activities (e.g., walking to school, taking the stairways instead of the lift) and fitness training as well as to decrease self-reported television viewing and excessive computer time to a total of 2 h or less per day; (ii) nutritional recommendations based on the “Optimized Mixed Diet for German Children and Adolescents” of the Research Institute of Child Nutrition, Dortmund, Germany [10]; and (iii) close personal surveillance by one physician (S.E.) with at least six telephone contacts and outpatient visits during the complete intervention program. Nutritionists were not specifically involved during the lifestyle intervention program, but a pediatric nurse assisted the physician during the outpatient visits. All participants were examined at a 1.5 T whole-body MR unit (Magnetom Sonata, Siemens Healthcare, Erlangen, Germany) equipped with MRS capabilities before and after approximately 7 months of low-level lifestyle intervention.

T1-weighted MRI for quantitative analysis of adipose tissue To obtain the distribution of whole-body fat axial T1weighted fast spin echo imaging (echo time (TE) 12 ms, repetition time (TR) 490 ms, slice thickness 10 mm, interslice gap 10 mm, matrix size 256×178) was applied within 12 s of breath hold. During each breath hold, five axially oriented slices were acquired. The participants were placed in the prone position with arms extended above the head. Data were acquired from fingertips to toes with one repositioning as described elsewhere [13]. For signal acquisition, the combined transmit/receive body coil was used.

Total adipose tissue (TAT), subcutaneous adipose tissue of the body trunk (SCATTR), adipose tissue (AT) of lower and upper extremities (ATLE and ATUE, respectively), visceral adipose tissue (VAT) as well as total non-adipose tissue (TNAT), NATLE and NATUE (i.e., non-adipose tissue of lower and upper extremities, respectively) were quantified by semiautomatic segmentation of AT and lean tissue from the wholebody data sets. Segmentation was performed using custombuild Matlab routines (The MathWorks Inc., Natick, MA, USA) [13]. The algorithm was mainly based on an intensity threshold separating bright fatty tissues from relatively dark lean tissues in a T1-weighted MR sequence. Since signal intensity of fatty bone marrow was also above the threshold but was thought to be relatively unchanged over the 7-month time period in hematological healthy adolescents, it was not manually excluded from the analysis and thus contributes to all adipose tissue compartments aside from VAT. The algorithm automatically excluded tissues without any measurable signal such as lung tissue or cortical bone. All adipose and nonadipose tissue compartments were additionally referenced to the total body volume in order to adjust for differences in height and weight. 1H-MRS of intrahepatic lipids The volunteers were positioned in the supine position on a sixchannel spine array coil. For topography imaging of the liver, T1-weighted gradient echo images were recorded during one breath hold. IHL were determined by single-voxel 1H MRS (Stimulated Echo Acquisition Mode (STEAM), TE 10 ms, mixing time (TM) 15 ms, TR 4 s, voxel of interest 30×30× 20 mm3, 40 acquisitions, total acquisition time 2 min 40 s) within the posterior part of segment 7 of the liver [13]. The signal was detected with one single element of the spine array coil chosen to be closest to the voxel of interest. To avoid spectral broadening, participants were asked to breathe in during each TR interval and to be in expiration during data acquisition. To train the volunteers before data acquisitions, the first eight breathing maneuvers were not used for data evaluation. The relative amount of intrahepatic lipids (combined integrals of methyl and methylene) were referenced to the water peak at 4.7 ppm and calculated by the sum of signal integrals of water and lipids obtained at fixed frequency intervals (water 3.1–6.2 ppm, lipids 0.5–1.8 ppm). Post-processing of the spectra included 4 k zero filling, multiplication with a Gaussian function (full width at half maximum 150 ms), Fourier transformation, and constant and linear phase correction. Anthropometric data Height, weight, and body mass index (BMI, calculated as weight/height2 in kg/m2) were assessed on the day of MR examination. Furthermore, waist and hip circumference were

Eur J Pediatr

measured (in cm). Measurements were taken by one of the authors (S.E.) or an assisting pediatric nurse during outpatient visits. Waist circumference was measured midway between the iliac crest and the rib cage. Hip circumference was measured at the maximum protuberance of the buttocks. Waist-tohip ratio was then calculated using these measures. To be able to compare BMI values as well as other anthropometric measures such as waist and hip circumference across different ages and by gender, the respective SDSLMS (standard deviation score) was calculated. For example, the calculation of BMI-SDSLMS is done according to the following equation: Lðt Þ

ðt ÞÞ BMI-SDSLMS = ðBMI=M Lðt Þ⋅S ðt Þ

−1

, where M(t), L(t), and S(t) de-

note the corresponding parameters for the age (t) and sex of the reference population [11, 5, 7]. Thus, BMI-SDSLMS is an expression measured in multiples of the standard deviation (zdistribution) corrected for age and gender. Metabolic data Due to the non-invasive character of the presented observational study, insulin sensitivity has only been estimated by homeostatic model assessment of insulin resistance (HOMA-IR) as in most pediatric studies rather than using more accurate invasive clamp techniques. However, the HOMA-IR can be regarded as a good estimate of insulin sensitivity and β-cell function, since it reflects the balance between hepatic glucose output and insulin secretion [22, 26] and shows a high correlation with fasting insulin levels (r= 0.99) in childhood and adolescence [17]. Thus, we have chosen HOMA-IR as a measure to estimate changes in insulin sensitivity during the performed low-level lifestyle intervention. For evaluation of liver integrity, serum enzyme levels (AST, ALT, GGT) were also analyzed. Statistical analysis For statistical analysis, the software package JMP (version 9.0.1; SAS Institute Inc. Cary, NC) was used. Actual data and differences were tested for normal distribution using the Shapiro-Wilk test. A logarithmic transformation of nonnormally distributed data was performed, but did not improve matters. Therefore, changes during lifestyle intervention were tested for statistical significance using either the two-sided Wilcoxon signed-rank test or the two-sided paired t test for non-normally or normally distributed data, respectively. To be consistent in the calculation of means and changes over time for all parameters (including various ratios and standard deviation scores), we have decided to use the geometric means and ratio of geometric means including their confidence intervals throughout the data as a homogenous descriptive measure for data presentation. Thus, all data at baseline and after lifestyle intervention are given as geometric mean and 95 %

confidence intervals. Correlation analysis between metabolic parameters and adipose tissue compartments at baseline was performed using Spearman’s rank correlation coefficient. In all tests, the significance level was chosen as 5 %, thus a p value <0.05 indicates statistical significance.

Results Before and after 7 months of lifestyle intervention, wholebody MR imaging could be successfully performed in all 39 adolescents. In three participants, MR spectroscopy did not lead to utilizable results due to insufficient breathing compliance: thus reliable data regarding IHL could be determined in 18/18 girls and 18/21 boys. Baseline analysis Baseline BMI-SDSLMS was 2.27 (2.14, 2.42) in boys and 2.00 (1.80, 2.22) in girls. The other anthropometric measures such as waist circumference and waist-to-hip ratio were also increased in both genders (Table 1). TAT was markedly increased while the biggest fat depots were SCATTR and ATLE. In contrast to these two subcutaneous fat compartments, the amount of VAT was relatively small in both genders as shown in Table 1. High percentages of IHL could be observed in both genders, but were more pronounced in boys, 7.5 % (5.1 %, 11.1 %) in boys vs. 4.6 % (2.9 %, 7.1 %) in girls. Furthermore, a markedly elevated HOMA-IR score was found in both genders (boys 6.1 (4.6, 8.2); girls 6.2 (4.7, 8.2)) and was associated with high serum levels of AST, ALT, and GGT—especially in boys. Evaluation of the complete study group (including boys and girls) regarding the relation between different body fat compartments and metabolic parameters revealed high correlation coefficients between IHL and HOMA-IR as well as between IHL and serum enzyme levels (see Table 2). These associations were also found if both genders were evaluated separately. Other body fat compartments showed either weak correlations with HOMA-IR score and serum liver enzyme levels (AST, ALT, and GGT) or marked differences of the correlation coefficients between both genders. Changes during lifestyle intervention During the 7-month low-level lifestyle intervention BMISDSLMS in boys and girls could be slightly reduced. The reduction of BMI-SDSLMS was significant in boys (p=0.045), but only approached significance in girls (p=0.054). Other anthropometric measures (waist circumference and waist-tohip ratio) did not show a significant change during lifestyle intervention. The most prominent changes could be noted for IHL, which decreased significantly in boys from 7.5 to 4.3 %

6.1 (5.4, 7.0)

19.3 (17.2, 21.6)

10.4 (9.6, 11.3)

ATUE (liter)

NATLE (liter)

NATUE (liter)

14.5 (13.8, 15.2)

20.2 (16.6, 24.5)

34.8 (28.7, 42.3)

26.6 (23.3, 30.3)

4.4 (3.2, 6.0)

4.3 (2.6, 7.1)

10.7 (9.8, 11.6)

19.8 (17.7, 22.2)

5.7 (4.9, 6.6)

19.3 (17.0, 21.9)

15.6 (12.9, 18.9)

2.4 (2.0, 2.9)

51.5 (46.4, 57.1)

43.3 (37.4, 50.2)

95.5 (85.5, 106.5)

2.835 (2.367, 3.395)

0.976 (0.933, 1.021)

4.83 (4.22, 5.53)

108.8 (102.4, 115.5)

2.17 (2.00, 2.37)

31.9 (29.8, 34.2)

96.9 (87.1, 107.8)

174.1 (169.3, 179.1)

1.045 (1.042, 1.048)

0.870 (0.780, 0.970)

0.958 (0.808, 1.136)

0.890 (0.778, 1.017)

0.713 (0.560, 0.908)

0.499 (0.340, 0.730)

1.025 (0.996, 1.054)

1.030 (1.001, 1.059)

0.929 (0.852, 1.013)

0.965 (0.912, 1.021)

0.929 (0.868, 0.995)

0.958 (0.871, 1.054)

1.013 (0.986, 1.040)

0.946 (0.891, 1.005)

0.983 (0.950, 1.018)

1.118 (0.932, 1.342)

0.991 (0.959, 1.025)

0.956 (0.900, 1.015)

0.994 (0.963, 1.026)

0.957 (0.917, 0.998)

0.983 (0.958, 1.009)

1.018 (0.989, 1.047)

1.017 (1.012, 1.023)

<0.001*

0.021*

NS°

NS°

0.008°

0.003°

NS* 0.037**

NS* 0.004**

NS* NS**

NS* NS**

0.043* 0.0140**

NS* NS**

NS* 0.028**

NS* 0.028**

NS*

NS°

NS°

NS°

NS*

0.045*

NS*

NS*

<0.001*

14.1 (13.6–14.6)

18.3 (14.7, 22.7)

24.6 (20.4, 29.7)

22.2 (20.4, 24.2)

6.2 (4.7, 8.2)

4.6 (2.9, 7.1)

8.5 (8.0, 9.0)

15.3 (14.4, 16.3)

5.2 (4.4, 6.1)

17.7 (15.1, 20.7)

14.8 (12.3, 17.7)

1.7 (1.4, 2.2)

40.8 (38.4, 43.4)

39.6 (33.7, 46.6)

81.2 (73.7, 89.5)

2.014 (1.463, 2.775)

0.933 (0.879, 0.991)

4.03 (3.40, 4.79)

101.4 (94.9, 108.3)

2.00 (1.80, 2.22)

31.7 (29.4, 34.2)

82.1 (75.7, 89.0)

161.0 (157.9, 164.1)

18.9 (14.6, 24.5)

24.1 (19.3, 30.0)

20.3 (17.7, 23.3)

4.7 (3.7, 5.9)

3.4 (2.2, 5.3)

8.6 (8.2, 9.1)

15.5 (14.6, 16.4)

5.1 (4.4, 6.0)

17.2 (14.6, 20.3)

14.6 (12.0, 17.8)

1.7 (1.4, 2.1)

40.8 (38.6, 43.2)

38.9 (32.8, 46.1)

80.5 (73.0, 88.9)

2.213 (1.831, 2.673)

0.926 (0.950, 1.108)

3.95 (3.41, 4.58)

101.4 (96.1, 107.0)

1.82 (1.44, 2.30)

31.6 (28.9, 34.5)

82.7 (75.9, 90.0)

161.7 (158.9, 164.6)

14.7 (14.2, 15.2)

After intervention

1.034 (0.901, 1.186)

0.977 (0.814, 1.174)

0.915 (0.812, 1.032)

0.752 (0.598, 0.944)

0.746 (0.519, 1.071)

1.020 (0.987, 1.053)

1.011 (0.984, 1.040)

0.989 (0.926, 1.058)

0.973 (0.935, 1.014)

0.990 (0.941, 1.042)

0.968 (0.895, 1.047)

1.001 (0.972, 1.030)

0.981 (0.939, 1.025)

0.991 (1.022, 0.962)

1.098 (0.822, 1.469)

0.992 (0.937, 1.050)

0.979 (0.887, 1.080)

0.999 (0.970, 1.030)

0.911 (0.790, 1.050)

0.997 (0.976, 1.020)

1.007 (0.982, 1.032)

1.005 (1.001, 1.009)

1.043 (1.040, 1.047)

(95 % CI)

NS°

NS°

NS*

0.030*

NS*

NS° NS**

NS* NS**

NS° NS**

NS* NS**

NS° NS**

NS° NS°°

NS* NS**

NS* NS**

NS*

NS°

NS°

NS*

NS*

0.054°

NS*

NS*

0.018*

<0.001*

P value

Ratio of geometric means (before/after)

*Two-sided paired t test; **Two-sided paired t test after adjustment for total body volume; °Wilcoxon signed-rank test; °°Wilcoxon signed-rank test after adjustment for total body volume

BMI body mass index; BMI SDSLMS BMI standard deviation score; WHR waist-to-hip ratio; WHR SDSLMS WHR standard deviation score; TBV total body volume; TAT total adipose tissue; TNAT total nonadipose (lean) tissue; VAT visceral adipose tissue; SCATTR subcutaneous adipose tissue of the body trunk; AT adipose tissue; NAT non-adipose tissue; LE lower extremities; UE upper extremities; IHL intrahepatic lipids; HOMA homeostasis model assessment score; AST, ALT, and GGT liver enzymes; NS not significant (P value>0.05)

23.2 (19.3, 28.0)

20.0 (17.4, 23.1)

ATLE (liter)

GGT

16.8 (14.0, 20.1)

SCATTR (liter)

36.3 (27.6, 47.9)

2.5 (2.1, 2.9)

VAT (liter)

29.9 (25.3, 35.2)

50.8 (45.5, 56.7)

TNAT (liter)

ALT

45.8 (39.6, 53.0)

TAT (liter)

AST

97.1 (86.3, 109.2)

TBV (liter)

7.5 (5.1, 11.1)

2.535 (1.967, 3.268)

WHR SDSLMS

6.1 (4.6, 8.2)

0.985 (0.953, 1.017)

WHR

HOMA

5.06 (4.43, 5.77)

Waist SDSLMS

IHL (%)

2.27 (2.14, 2.42)

109.4 (103.0, 116.2)

32.5 (30.5, 34.6)

BMI (kg/m2)

BMI SDSLMS

95.2 (85.4, 106.1)

Weight (kg)

Waist (cm)

13.8 (13.2, 14.5)

171.2 (166.0, 176.5)

Height (cm)

Before intervention

(95 % CI)

Before intervention

P value

Geometric means (95 % CI)

Ratio of geometric means (before/after)

Geometric means (95 % CI) After intervention

Girls (N=18)

Boys (N=21)

Anthropometric and metabolic data as well as size of AT compartments before and after lifestyle intervention

Age (years)

Table 1

Eur J Pediatr

Eur J Pediatr Table 2

Correlations between baseline values of AT compartments, insulin sensitivity, and liver enzymes

TAT (liter) VAT (liter) SCATTR (liter) IHL (%) ATLE (liter) ATUE (liter)

HOMA

AST

ALT

GGT

0.37* (0.53*/0.15) 0.34* (0.48*/0.29) 0.49** (0.66**/0.29) 0.57*** (0.52*/0.68**) 0.22 (0.36/0.04) 0.35* (0.51*/0.21)

0.31 (0.27/−0.02) 0.42** (0.14/0.52*) 0.35* (0.44*/−0.03) 0.67*** (0.63**/0.68**) 0.18 (0.12/−0.08) 0.34* (0.35/0.12)

0.39* (0.34/0.18) 0.46** (0.25/0.52*) 0.43** (0.51*/0.19) 0.70*** (0.64**/0.73***) 0.23 (0.17/0.09) 0.37* (0.38/0.20)

0.26 (0.32/0.03) 0.45** (0.43/0.35) 0.32 (0.49*/0.04) 0.52*** (0.48*/0.46) 0.13 (0.15/−0.08) 0.26 (0.24/0.11)

Spearman correlation coefficient is given for both complete group and in parentheses separately for males/females TAT total adipose tissue; VAT visceral adipose tissue; SCATTR subcutaneous adipose tissue of the body trunk; IHL intrahepatic lipids; AT adipose tissue, LE lower extremities, UE upper extremities; HOMA homeostasis model assessment score; AST, ALT, and GGT liver enzymes *P<0.05; **P<0.01; ***P<0.001

(p=0.003). Starting at lower baseline values, the data derived from girls also showed a slight reduction of IHL from 4.6 to 3.4 %, but this did not reach statistical significance. Results are exemplarily shown in Figs. 1 and 2. The HOMA-IR score decreased from 6.1 to 4.4 in boys (p=0.008) and from 6.2 to 4.7 in girls (p=0.030). In boys, AST, ALT, and GGT were also reduced after lifestyle intervention, but reached only significance for GGT (p<0.021). Similar to the observed lower IHL levels, the serum enzyme levels at baseline were also lower in girls than in boys and did not change significantly during lifestyle intervention. Regarding body fat distribution, all AT compartments were slightly reduced in size after lifestyle intervention, but even after adjustment for total body volume only the decrease of TAT and SCATTR in boys became significant (p=0.028 and p=0.014, respectively). Regarding the non-adipose tissue compartments (NAT)—which more or less represent the lean tissue comprising bones and muscle mass—only a slight increase of NAT of the lower and upper extremities in boys could be found (p=0.004 and p=0.037, respectively). In girls, neither adipose nor non-adipose tissue compartments showed significant changes after lifestyle intervention.

Discussion In a cohort of obese adolescents, whole-body fat distribution, IHL, and metabolic markers were assessed before and after 7 months of a low-level lifestyle intervention. The increase of physical activity combined with nutritional recommendations led to a slight reduction of BMI-SDSLMS in boys and girls, which represents a successful outcome of the intervention. As assessed by MRS, the successful low-level lifestyle intervention resulted in a marked decrease of liver lipid storage (IHL) by about 50 % in boys. In contrast, the decrease of IHL in girls was smaller (only about 25 %) and did not reach statistical significance. However, this might be due to an already lower baseline value in girls (IHL in %, 4.6 (2.9, 7.1) in girls vs. 7.5

(5.1, 11.1) in boys), which is consistent with previously published data on gender differences in hepatic lipid accumulation. In accordance with other intervention studies, the observed marked decrease of IHL emphasizes the beneficial effect even of a low-level lifestyle intervention therapy on the unfavorable ectopic fat deposition in the liver [9, 25, 24]. Accompanying the reduction in IHL, the HOMA-IR score—as a non-invasive estimate of insulin resistance—significantly improved in both genders. This is also in accordance with previous studies in which a decrease of IHL was often accompanied by an increase in insulin sensitivity and a loss in total fat mass [9, 25]. However, contrary data have also been published: in case of exercising adolescents Bell et al. reported an improvement in insulin sensitivity, but a constant fat mass [1]. In our study, the observed correlations between IHL and serum liver enzyme levels/HOMA-IR score in both genders might indicate both a reduction of hepatic injury due to inflammation processes as well as a beneficial metabolic impact of reduced intrahepatic lipid storage on insulin sensitivity, which is consistent with other studies [24, 6]. In this context, the marked effect of the performed low-level lifestyle intervention is promising and may help to counteract the development of metabolic disturbances including chronic hepatic steatosis and its possible progress to liver inflammation and cirrhosis in the youth. Besides the IHL, various other whole-body fat compartments were also analyzed in our study regarding their impact on the investigated metabolic markers. At baseline, but also after lifestyle intervention, boys showed a greater amount of VAT than girls, whereas total amount of VAT in both genders was relatively small compared to adults [14]. The genderspecific differences in body composition have been previously reported by others and are already developed at the age of three [2, 21]. However, during pubertal maturation, body fat distribution further develops and differences increase with age, even in adults [14]. During the performed low-level lifestyle intervention, most AT compartments showed a slight

Eur J Pediatr Fig. 1 T1-weighted axial images showing fatty tissue as bright areas: at the a head of the humerus, b umbiculus, and c head of femur in a 15-year-old boy before (left column) and after (right column) successful lifestyle intervention. d MRS spectrum of the liver segment 7 before (left column) and after (right column) successful lifestyle intervention

decrease in both genders. After adjustment for total body volume, TAT (total adipose tissue) and its biggest subpart SCATTR (subcutaneous adipose tissue of the body trunk) were the only ones that could be significantly reduced in boys. In contrast, none of the AT compartments in girls significantly changed during lifestyle intervention. It has been published that without any change in lifestyle, both SCAT and VAT of adolescents increase with age. Thus, a stable or even a slight decrease may already be a successful result. The higher growth velocity in boys during the intervention period (∼2.9 vs. ∼0.7 cm per 7 months for boys and girls, respectively) during the intervention was accompanied by a slight increase of non-adipose tissue compartments of the upper and lower extremities after adjustment for body volume. This may be most likely attributed to the physiological increase of muscle mass in pubertal boys rather than to an increase in

physical activity. Since in adults the amount of VAT has a marked effect on insulin sensitivity [13], various whole-body AT compartments were evaluated and especially the VAT was of great interest. Although other studies revealed a relation between insulin sensitivity and VAT, SCAT, and percentage of fat mass in both genders [8, 24], this could only be found for boys in our study. As a limitation of our study, the examined obese adolescents—although examined at a comparable age of about 14 years—presented with different stages of pubertal maturation from Tanner stage 1 to Tanner stage 5 with a trend towards lower stages in boys. Thus, substantial endocrine differences (sex steroids, pituitary hormones) might exist between genders, within each gender group and also intra-individually differences during the lifestyle intervention may have occurred. All of these might have affected the results of our study, since

Eur J Pediatr Fig. 2 Comparison of intrahepatic lipids (IHL) and insulin resistance (homeostatic model assessment of insulin resistance—HOMA-IR) before and after lifestyle intervention separately for boys and girls. Each bold line represents the respective geometric mean with its 95 % confidence interval (dotted lines)

it has already been shown that the stage of maturation significantly influences the AT distribution in adolescents and that gender-specific differences in AT distribution already exist in adolescence [2]. However, the primary endpoint of the study was not to compare AT compartments between genders or inter-individually, but more to evaluate intra-individual changes during lifestyle intervention. Although only small changes in body fat compartments could be observed during 7 months, the performed low-level lifestyle intervention was nevertheless successful in terms of decreasing the overall BMISDS LMS and improving the insulin sensitivity as assessed by the HOMA-IR. Thus, evaluation of lifestyle interventions in children and adolescents should not only rely on simple anthropometric measurements, since the reduction of liver fat and/or beneficial effects on metabolic balance may not be well reflected by the absolute body weight, waist circumference, or waist-to-hip ratio. Especially in adolescence, one has to keep in mind that the body weight may be influenced by several variables: for example, pubertal growth may be combined with a physiologic muscle gain induced by increasing sex steroid levels and accompanied by a decrease in AT without any obvious changes in body weight. However, the serum enzyme levels AST, ALT,

and GGT might act as indirect indicators especially for both the reduction of IHL and a possible increase in insulin sensitivity, but also as markers for hepatic injury due to inflammation processes.

Conclusion In a cohort of obese adolescents, a low-level lifestyle intervention over a period of 7 months slightly reduced the BMI-SDSLMS. Furthermore, intrahepatic lipid storage (IHL) decreased markedly, while other AT compartments including visceral AT were only slightly affected. The reduction in IHL came along with a significantly improved HOMA-IR and reduced serum enzyme levels, indicating a beneficial effect on both insulin sensitivity and hepatic injury due to intrahepatic lipid storage. In contrast to previous studies in adults, visceral AT seems to have only a slight impact on insulin sensitivity in adolescents. These results emphasize the positive effect even of low-level lifestyle interventions in the youth, although further studies are warranted to fully investigate the effects of age, sexual maturation, and genderspecific differences on the development of body fat distribution and their relation to insulin resistance.

Eur J Pediatr Acknowledgment Members of the DISKUS Study Group are Stefan Ehehalt, Roland Schweizer, Nicole Schurr, Coya Pfaff, Andreas Neu, Hans Peter Haber, Michael B. Ranke, and Gerhard Binder, Department of Paediatrics, University Children’s Hospital, Tübingen, Germany; Perikles Simon, Institute of Sport Science, Department of Sports Medicine, Johannes Gutenberg University, Mainz, Germany; Andre Lacroix, Jochen Hansel, and Andreas Nieß, University Department of Medicine, Department of Sports Medicine, Tübingen, Germany; Katrin Giel, Markus Schrauth, Paul Enck, and Stephan Zipfel, University Department of Medicine, Department of Psychosomatic Medicine and Psychotherapy, Tübingen, Germany; Jürgen Machann, Fabian Springer, Verena Ballweg, and Fritz Schick, Section on Experimental Radiology, Department of Diagnostic and Interventional Radiology, Tübingen, Germany; Pablo Brockmann and Michael Urschitz, Department of Neonatology, University Children’s Hospital, Tübingen, Germany; and Huu Phuc Nguyen, Institute of Human Genetics, Department of Medical Genetics, Tübingen, Germany

7.

8.

9.

10.

Conflict of interest The authors declare that they have no conflict of interest. Ethical standards Written informed consent was obtained from all participants and their legal guardians prior to their inclusion in the study. The study has been approved by the appropriate ethics committee and has therefore been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. Authors’ contributions F.S., V.B., R.S., F.S., M.B.R., G.B., and S.E. prepared the manuscript and performed literature research. F.S. and V.B. performed whole-body MRI/MRS examinations and evaluated the data under the supervision of F.S.; S.E. and R.S. examined the patients during lifestyle intervention and acquired anthropometric/metabolic data. F.S., M.B.R., and G.B. supervised the study and helped in data analysis performed by F.S. V.B., and S.E. All authors reviewed and approved the manuscript on behalf of the DISKUS Study Group.

11.

12.

13.

14.

References 15. 1.

2.

3.

4.

5. 6.

Bell LM, Watts K, Siafarikas A, Thompson A, Ratnam N, Bulsara M, Finn J, O'Driscoll G, Green DJ, Jones TW, Davis EA (2007) Exercise alone reduces insulin resistance in obese children independently of changes in body composition. J Clin Endocrinol Metab 92:4230–4235. doi:10.1210/jc.2007-0779 Benfield LL, Fox KR, Peters DM, Blake H, Rogers I, Grant C, Ness A (2008) Magnetic resonance imaging of abdominal adiposity in a large cohort of British children. Int J Obes (Lond) 32:91–99. doi:10. 1038/sj.ijo.0803780 Brodie DA, Stewart AD (1999) Body composition measurement: a hierarchy of methods. J Pediatr Endocrinol Metab 12:801–816. doi: 10.1515/JPEM.1999.12.6.801 Caranti DA, de Mello MT, Prado WL, Tock L, Siqueira KO, de Piano A, Lofrano MC, Cristofalo DM, Lederman H, Tufik S, Damaso AR (2007) Short- and long-term beneficial effects of a multidisciplinary therapy for the control of metabolic syndrome in obese adolescents. Metabolism 56:1293–1300. doi:10.1016/j. metabol.2007.05.004 Cole TJ, Freeman JV, Preece MA (1995) Body mass index reference curves for the UK, 1990. Arch Dis Child 73:25–29 Fishbein MH, Miner M, Mogren C, Chalekson J (2003) The spectrum of fatty liver in obese children and the relationship of serum aminotransferases to severity of steatosis. J Pediatr Gastroenterol Nutr 36:54–61. doi:10.1097/00005176-200301000-00012

16.

17.

18.

19.

20.

Fredriks AM, van Buuren S, Fekkes M, Verloove-Vanhorick SP, Wit JM (2005) Are age references for waist circumference, hip circumference and waist-hip ratio in Dutch children useful in clinical practice? Eur J Pediatr 164:216–222. doi:10.1007/s00431-0041586-7 Gutin B, Owens S (1999) Role of exercise intervention in improving body fat distribution and risk profile in children. Am J Hum Biol 11:237–247. doi:10.1002/(SICI)1520-6300(1999) 11:2<237::AID-AJHB11>3.0.CO;2-9 Hasson RE, Adam TC, Davis JN, Kelly LA, Ventura EE, ByrdWilliams CE, Toledo-Corral CM, Roberts CK, Lane CJ, Azen SP, Chou CP, Spruijt-Metz D, Weigensberg MJ, Berhane K, Goran MI (2012) Randomized controlled trial to improve adiposity, inflammation, and insulin resistance in obese African-American and Latino youth. Obesity (Silver Spring) 20:811–818. doi:10.1038/ oby.2010.343 Kersting M, Alexy U, Clausen K (2005) Using the concept of food based dietary guidelines to develop an optimized mixed diet (OMD) for German children and adolescents. J Pediatr Gastroenterol Nutr 40:301–308. doi:10.1097/01.MPG. 0000153887.19429.70 Kromeyer-Hauschild K, Wabitsch M, Kunze D, Geller F, Geiß HC (2001) Perzentile für den Body-mass-Index für das Kindes- und Jugendalter unter Heranziehung verschiedener deutscher Stichproben. Monatsschr Kinderheilkd 149:807–818. doi:10.1007/ s001120170107 Lobstein T, Frelut ML (2003) Prevalence of overweight among children in Europe. Obes Rev 4:195–200. doi:10.1046/j.1467789X.2003.00116.x Machann J, Thamer C, Schnoedt B, Stefan N, Stumvoll M, Haring HU, Claussen CD, Fritsche A, Schick F (2005) Age and gender related effects on adipose tissue compartments of subjects with increased risk for type 2 diabetes: a whole body MRI/MRS study. Magn Reson Mater Phy 18:128–137. doi:10.1007/s10334-0050104-x Machann J, Thamer C, Stefan N, Schwenzer NF, Kantartzis K, Haring HU, Claussen CD, Fritsche A, Schick F (2010) Follow-up whole-body assessment of adipose tissue compartments during a lifestyle intervention in a large cohort at increased risk for type 2 diabetes. Radiology 257:353–363. doi:10.1148/radiol.10092284 Molleston JP, White F, Teckman J, Fitzgerald JF (2002) Obese children with steatohepatitis can develop cirrhosis in childhood. Am J Gastroenterol 97:2460–2462. doi:10.1111/j.1572-0241. 2002.06003.x Savoye M, Shaw M, Dziura J, Tamborlane WV, Rose P, Guandalini C, Goldberg-Gell R, Burgert TS, Cali AM, Weiss R, Caprio S (2007) Effects of a weight management program on body composition and metabolic parameters in overweight children: a randomized controlled trial. JAMA 297:2697–2704. doi:10.1001/jama. 297.24.2697 Schwartz B, Jacobs DR Jr, Moran A, Steinberger J, Hong CP, Sinaiko AR (2008) Measurement of insulin sensitivity in children: comparison between the euglycemic-hyperinsulinemic clamp and surrogate measures. Diabetes Care 31:783–788. doi:10.2337/dc071376 Schwimmer JB, Deutsch R, Kahen T, Lavine JE, Stanley C, Behling C (2006) Prevalence of fatty liver in children and adolescents. Pediatrics 118:1388–1393. doi:10.1542/peds.2006-1212 Shen W, Wang Z, Punyanita M, Lei J, Sinav A, Kral JG, Imielinska C, Ross R, Heymsfield SB (2003) Adipose tissue quantification by imaging methods: a proposed classification. Obes Res 11:5–16. doi: 10.1038/oby.2003.3 Siegel MJ, Hildebolt CF, Bae KT, Hong C, White NH (2007) Total and intraabdominal fat distribution in preadolescents and adolescents: measurement with MR imaging. Radiology 242:846–856. doi:10.1148/radiol.2423060111

Eur J Pediatr 21.

22.

23.

24.

Taylor RW, Gold E, Manning P, Goulding A (1997) Gender differences in body fat content are present well before puberty. Int J Obes Relat Metab Disord 21:1082–1084. doi:10.1038/sj.ijo.0800522 Turner RC, Holman RR, Matthews D, Hockaday TD, Peto J (1979) Insulin deficiency and insulin resistance interaction in diabetes: estimation of their relative contribution by feedback analysis from basal plasma insulin and glucose concentrations. Metabolism 28:1086–1096. doi:10.1016/0026-0495(79) 90146-X van der Heijden GJ, Toffolo G, Manesso E, Sauer PJ, Sunehag AL (2009) Aerobic exercise increases peripheral and hepatic insulin sensitivity in sedentary adolescents. J Clin Endocrinol Metab 94: 4292–4299. doi:10.1210/jc.2009-1379 van der Heijden GJ, Wang ZJ, Chu ZD, Sauer PJ, Haymond MW, Rodriguez LM, Sunehag AL (2010) A 12-week aerobic exercise program reduces hepatic fat accumulation and insulin resistance in

obese, Hispanic adolescents. Obesity (Silver Spring) 18:384–390. doi:10.1038/oby.2009.274 25. Vitola BE, Deivanayagam S, Stein RI, Mohammed BS, Magkos F, Kirk EP, Klein S (2009) Weight loss reduces liver fat and improves hepatic and skeletal muscle insulin sensitivity in obese adolescents. Obesity (Silver Spring) 17:1744–1748. doi:10.1038/oby.2009.171 26. Wallace TM, Levy JC, Matthews DR (2004) Use and abuse of HOMA modeling. Diabetes Care 27:1487–1495. doi:10.2337/ diacare.27.6.1487 27. Weiss R, Dufour S, Taksali SE, Tamborlane WV, Petersen KF, Bonadonna RC, Boselli L, Barbetta G, Allen K, Rife F, Savoye M, Dziura J, Sherwin R, Shulman GI, Caprio S (2003) Prediabetes in obese youth: a syndrome of impaired glucose tolerance, severe insulin resistance, and altered myocellular and abdominal fat partitioning. Lancet 362:951–957. doi:10.1016/S01406736(03)14364-4