| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Vestische Hospital for Children and Adolescents (T.R., G.d.S., W.A.), University of Witten/Herdecke, D-45711 Datteln, Germany; and Department of Pediatrics (C.L.R.), University of Bonn, D-53012 Bonn, Germany
Address all correspondence and requests for reprints to: Dr. Thomas Reinehr, Vestische Kinderklinik und Jugendklinik, University of Witten/Herdecke, Dr. F. Steiner Strasse 5, 45711 Datteln, Germany. E-mail: t.reinehr{at}kinderklinik-datteln.de.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Objective: The objective of this study was to compare androgens between obese and normal-weight children and to study the effect of weight loss on androgens.
Design: The design was a cross-sectional comparison between obese and normal-weight children separated according to pubertal stage and longitudinal 1-yr follow-up study in obese children participating in a weight-loss intervention.
Setting: The setting of this study was a primary care facility.
Patients: A total of 273 obese and 79 lean children (aged 4–14 yr) were studied, including a subgroup of 155 obese children for the longitudinal study.
Intervention: The intervention program was an outpatient 1-yr intervention program based on exercise, behavior, and nutrition therapy (high-carbohydrate low-fat diet).
Main Outcome Measures: The outcome measures included testosterone and dehydroepiandrosterone sulfate (DHEAS) at baseline and 1 yr later.
Results: The obese prepubertal children and the obese pubertal girls showed significantly (P < 0.01) higher testosterone and DHEAS levels, whereas obese pubertal boys did not significantly differ in androgens from their lean counterparts. Significant correlations with body mass index were demonstrated in multivariate regression analyses for DHEAS in all children and for testosterone in prepubertal children and in pubertal girls. The obese prepubertal children and obese girls losing substantial weight showed a significant (P < 0.05) decrease in their testosterone concentrations.
Conclusions: Moderately increased testosterone and DHEAS levels were found in obese prepubertal children and in obese pubertal girls, whereas androgen concentrations did not differ between obese and normal-weight pubertal boys. Weight loss induced a decrease in testosterone in obese prepubertal children and pubertal girls pointing to a reversible increase of androgens.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In children and adolescents, a high androgenic activity is discussed to be associated with polycystic ovarian syndrome, precocious puberty, and accelerated bone age with relatively tall stature (5, 6). All of these clinical features occur in obese children more frequently than in normal-weight children (5, 7). Furthermore, children with premature adrenarche may tend to obesity (8). Little and controversial information is available on androgens in obese children. Some investigators stated that hyperandrogenemia occurs only after menarche (9). One study reported lower testosterone levels in male and female obese prepubertal children compared with normal-weight children (10). Other studies revealed similar testosterone concentrations in obese and normal-weight prepubertal boys and girls (11, 12), whereas studies in obese female adolescents demonstrated increased testosterone levels compared with lean adolescents (13, 14). Conversely, DHEAS levels are reported to be increased in obese prepubertal children (6, 11, 12) and in obese female adolescents (13, 14), whereas studies in obese pubertal boys have not been performed yet.
Studies in obese women and obese adolescents after menarche have shown a normalization of hyperandrogenemia and metabolic disturbances after weight loss (14, 15). It is completely unknown whether obese prepubertal and pubertal children benefit from weight reduction in a similar manner. Only two longitudinal studies concerning androgen levels consisting of only very few subjects were performed in obese children. In one study, no change of testosterone could be demonstrated (16), whereas in the other study, a reduction of DHEAS could be measured in weight loss (11).
Because only very few and controversial data concerning androgen concentrations in prepubertal and pubertal obese children and their changes in weight loss exist, we analyzed testosterone and DHEAS levels in a large cohort of obese children with respect to their pubertal stage and their weight change over a 1-yr period.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Fasting testosterone and DHEAS serum concentrations of the 273 obese children were compared with fasting testosterone and DHEAS levels of 79 normal-weight [age-related BMI >10th and <90th BMI percentile (17)] healthy children aged 4–14 yr. Furthermore, SHBG was determined in a subgroup of these children (76 obese and 38 normal-weight subjects). The degree of overweight was quantified using Cole’s least mean square method, which normalizes the BMI skewed distribution and expresses BMI as a SD score (SDS-BMI) (18). Reference data for German children (17) were used.
Blood sampling was performed at 0800 h. The children were fasted for 10–12 h before blood sampling. Serum total testosterone concentrations were determined by high-specific chemiluminescence immunoassay (ADVIA) (intraassay and interassay coefficient of variation, <5%; sensitivity, 0.1 nmol/liter). The normal values ranges were as follows: for prepubertal girls, less than 0.4 nmol/liter; for pubertal girls, less than 1.0 nmol/liter; for prepubertal boys, less than 0.8 nmol/liter; and for pubertal boys, less than 9.7 nmol/liter. Serum DHEAS concentrations were determined by high-specific competitive chemiluminescence immunoassay (Immulite) (intraassay and interassay coefficient of variation, <10%; sensitivity, 0.05 µmol/liter). The normal value ranges were as follows: for prepubertal girls, less than 1.8 µmol/liter; for pubertal girls, less than 4.7 µmol/liter; for prepubertal boys, less than 2.4 µmol/liter; and for pubertal boys, less than 4.1 µmol/liter. Serum SHBG concentrations were determined by chemiluminescence immunoassay (Immulite) (intraassay and interassay coefficient of variation, <5%; sensitivity, 0.02 nmol/liter).
The pubertal developmental stage was determined according to Marshall and Tanner and categorized into two groups: prepubertal, boys with pubic hair and gonadal stage I, girls with pubic hair stage and breast stage I; pubertal, boys with pubic hair and gonadal stage of at least II and girls with pubic hair stage and breast stage of at least II. Boys with change of voice and girls with menarche were excluded from the study. All results were separately analyzed in the groups of prepubertal and pubertal children and in boys and girls (81 obese prepubertal boys, 24 normal-weight prepubertal boys, 60 obese pubertal boys, 18 normal-weight pubertal boys, 72 obese prepubertal girls, 24 normal-weight prepubertal girls, 60 obese pubertal girls, and 13 normal-weight pubertal girls) because pubertal stage and gender influence androgen levels. Delayed puberty was defined by lack of physical manifestations of sexual maturation in boys at an age of 14 yr and in girls at an age of 13 yr. Sexual precocity was defined by appearance of secondary sexual maturation at an age of less than 9 yr in boys and at an age of less than 8 yr in girls.
Furthermore, testosterone and DHEAS concentrations were measured at baseline and 1 yr later in the 155 of the 273 obese children who took part in the Obeldicks intervention program. Additionally, SHBG levels were determined at baseline and 1 yr later in a subgroup of 43 children participating in the Obeldicks intervention program. The weight reduction in this program was uniformly achieved in this program through a 1-yr therapy with behavioral components, physical exercise, and nutrition education (based on high-carbohydrate low-fat diet), including individual psychological care of the child and his or her family. A detailed description of the program has been published previously (19, 20). A total of 74% of participants reduced their degree of overweight (intention-to-treat analysis), leading to an improvement of cardiovascular risk factors such as hypertension, dyslipidemia, and impaired glucose metabolism (21, 22). Four years after the end of treatment, a significant weight reduction was measurable in 71% of the participants (23).
The children were separately analyzed with respect to gender and pubertal stage. In the longitudinal study, the children were also distributed according to their weight change in the 1-yr period. The following classification was used because, in a reduction of less than 0.5 SDS-BMI, no improvement of insulin sensitivity and cardiovascular risk factors could be measured, and hormones such as leptin, cortisol, or adiponectin did not tend to normalize below this cutoff point in obese German children (21, 22, 24, 25): 1) substantial reduction of overweight was a decrease in SDS-BMI of more than 0.5; 2) minimal reduction of overweight was a decrease in SDS-BMI between 0.1 and 0.5; and 3) no reduction of overweight was a decrease in SDS-BMI of less than 0.1. The 22 obese children, who entered into puberty or had a change of voice or began menarche during the study period, were excluded from this analysis.
Statistical analysis was performed by the Winstat software package. Kolmogorov-Smirnov test revealed normal distribution in all continuous variables at baseline and in 1-yr follow-up. Statistically significant differences were tested for quantitative items by the Student’s t test, for paired observation by Student’s t tests for paired observations, and for qualitative items by 
2 test. Multiple variables were tested by Kruskal-Wallis test. Direct multivariate linear regression analyses were conducted for the dependent variables DHEAS and testosterone concentrations, including weight status (BMI), age, gender, and pubertal stage as independent variables in the 273 obese children and in the summarized collective of obese and normal-weight children (n = 352) separately. Gender (0, female; 1, male) and pubertal stage (0, prepubertal; 1, pubertal) were used as classification variables in each model. In the 155 obese children with follow-up, changes of DHEAS and changes of testosterone concentrations were correlated to changes of weight status (SDS-BMI) using partial regression adjusted for changes of pubertal stage. Because changes of testosterone, changes of DHEAS, and changes of SDS-BMI were not normally distributed, they were log transformed. P < 0.05 was considered significant. The study was approved by the local Ethics Committee of the University of Witten/Herdecke. Informed consent was obtained from all children and their parents.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
and 2). The DHEAS and testosterone concentrations were only very moderately increased in the obese children compared with the normal values. SHBG concentrations were significantly lower in the obese children compared with the normal-weight children (Table 3). None of the obese children demonstrated precocious or delayed puberty.
|
|
|
DHEAS concentrations were significantly associated with age (coefficient, 20; 95% confidence interval, 13–28; P < 0.001), gender (coefficient, 54; 95% confidence interval, 25–83; P < 0.001), pubertal stage (coefficient, 42; 95% confidence interval, 6–8; P = 0.022), and weight status (BMI) (coefficient, 7; 95% confidence interval, 5–10; P < 0.001) using a multivariate linear regression (r2 = 0.37) in the collective of 352 obese and normal-weight children. Performing this multiple regressions analysis in the collective of obese children excluding the normal-weight children did not demonstrate different results (data not shown).
In the 155 obese children of the follow-up study, changes of testosterone concentrations were significantly related to changes of weight status (r = 0.22; P = 0.004), whereas the changes of DHEAS concentrations were not significantly related to changes of weight status (r = 0.11; P = 0.091) using partial correlation adjusted to changes of pubertal stages.
Obese prepubertal boys losing substantial weight showed a significant decrease in testosterone levels, whereas obese prepubertal boys with minimal or without weight loss did not change their testosterone concentrations significantly (Table 4). Obese prepubertal boys demonstrated a significant increase of DHEAS concentrations regardless of their weight change over the 1-yr study period (Table 4).
|
).
|
). Obese prepubertal girls with substantial and minimal weight loss demonstrated no significant change of their DHEAS concentrations, whereas obese prepubertal girls without weight loss showed a significant increase of their DHEAS concentrations (Table 6).
|
). Obese pubertal girls demonstrated an increase of their DHEAS concentrations regardless of their weight change over the 1-yr study period, which was significant in the group of obese pubertal girls without weight loss (Table 7).
|
SHBG concentrations significantly increased in 19 obese children with substantial weight loss over the 1-yr period (at baseline: median, 43 [interquartile range (IQR), 32–60; 1-yr follow-up, 47 (IQR 42–71) nmol/liter; P = 0.037], whereas SHBG levels did not significantly change in 24 obese children without substantial weight loss [at baseline, median, 27 (IQR 19–55); 1-yr follow-up, 27 (IQR 19–55) nmol/liter; P = 0.929]. Obese children with and without substantial weight loss did not differ in terms of age [median, 9 yr (IQR 7–11) in both groups], gender (42 vs. 44% boys), or pubertal stage (63% prepubertal in both groups).
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Compared with normal-weight children of the same age, gender, and pubertal stage, obese prepubertal children and obese pubertal girls demonstrated significantly increased testosterone and DHEAS levels, which were usually within the normal range. Indices of extent of adiposity (BMI) correlated to androgens in multiple regression analyses. Our finding of increased DHEAS levels in prepubertal obese children are in concordance with the three other studies consisting of only small number of prepubertal girls (6, 12) and boys (11). Studies in obese pubertal children have not been performed until now. The increased testosterone concentrations in our study are in contrast to two studies in prepubertal obese boys (10, 11) and one study in prepubertal and pubertal obese girls (12). The different findings in these studies could probably be explained by the small sample size of only six obese children in these studies (11, 12). A tendency but without significance of increased testosterone levels was measured in these studies. Therefore, obese pubertal girls and obese prepubertal children regardless of their gender demonstrated increased androgens.
We observed a positive correlation of adiposity (to the extent indicated by BMI) to androgens not only in girls but also in prepubertal boys. This contrasts with the findings of decreased androgens in obese men (2). In accordance with these findings, obese pubertal boys did not demonstrate increased androgen levels in our study. Probably, the different origin of androgen production could explain the gender difference and the difference between prepubertal and pubertal obese boys. In men, circulating androgen levels are mainly of gonadal origin and may mask secretion and regulation of adrenal androgens. In children, adrenal androgen production starts before gonadal production at the age of 6–8 yr, with a maximum in puberty (26). However, because we did not measure ACTH levels and did not perform function tests of the adrenal glands, we were not able to estimate the relative contribution of adrenal and gonadal androgen production. Conversely, the concentrations of DHEAS, which is largely of adrenal origin, tended to be slightly higher in the obese children, indicating that an increased adrenal production could contribute to the high androgenic activity.
The cause of increased androgens in obesity is still under debate. CRH, which is activated in obesity, has been suggested recently to increase adrenal androgen secretion directly (27). The fat-derived hormone leptin has a specific, dose-dependent role to promote the formation of adrenal androgens, stimulating 17,20-lyase activity of cytochrome P450 (28). Furthermore, leptin influences the synchronization of the LHRH-pulse generator (29, 30). In addition, leptin acts in concert with other growth-derived signals to regulate the onset of puberty in humans (31). Longitudinal studies have revealed that leptin increases before steroid hormones are secreted and might provide the brain with information on body composition and size to start pituitary, adrenal, and gonadal maturation (31, 32). According to cross-sectional and longitudinal observations, adrenarche begins at about the same time as preadolescent rise in BMI (26).
Furthermore, insulin is hypothesized to stimulate androgen secretion (33). This may explain the link between hyperandrogenism and metabolic syndrome in obese women, which is based on insulin resistance. Insulin may represent a regulatory factor for ovarian steroidgenesis by amplifying the stimulatory effect of LH (34) and adrenal activation by stimulating the 17-hydroxylase activity (35). Accordingly, a correlation between insulin levels and testosterone levels has been reported in obese children (16), and the pubertal increase of DHEAS is parallel to the increase of insulin (26, 36).
The decrease of testosterone serum levels after weight loss in prepubertal children and pubertal girls points to a reversible increase of testosterone in obesity. This change in testosterone serum levels is not caused by other influencing factors because, in a collective of obese prepubertal children and obese girls of similar age, gender, and degree of overweight without weight loss, there were no changes in testosterone concentrations over the same time period, whereas in the group of obese pubertal girls without weight loss, testosterone levels increased. The findings in obese pubertal boys are difficult to interpret, because testosterone levels increase in the progress of puberty. Because the follow-up period was 1 yr, the physiological increase of androgens may have covered the effect of weight loss.
Decreasing testosterone levels in weight loss are in line with studies in obese women and female adolescents after menarche (14, 15, 37) and in contrast to the only existing study in childhood demonstrating stable testosterone levels after weight loss over a time period of 3 wk (16). The facts that (1) the degree of weight loss was only very moderate in this study, (2), the study period was very short, (3) the small study cohort consisted of prepubertal and pubertal children, and (4) the weight loss was based on a hypocaloric diet may explain the different findings.
DHEAS levels increased during follow-up in all obese children regardless of their weight changes. Because the majority of our children were older than 6 yr, when DHEAS concentrations start to increase due to adrenal "puberty" (26, 36), we cannot examine the effect of weight loss over a long study period without this covering effect. An indirect hint that weight loss may lead to decreasing DHEAS levels could be the fact that children with substantial weight loss demonstrated a lower degree of increase in DHEAS concentrations compared with their counterparts without weight loss. Studies in adults and short-term studies in children and adolescents revealed decreasing DHEAS concentrations in weight loss (11, 14, 15, 26, 37).
Our study presents some potential limitations. First, BMI percentiles were used to classify overweight. Although BMI is a good measure for overweight, one needs to be aware of its limitations as an indirect measurement of fat mass. Second, testosterone is banded to SHBG, which is influenced by many factors. Children taking any medication or with a medical condition, which can affect SHBG, were excluded from our study. SHBG is decreased in obese children regardless of pubertal stage or gender (10, 38), as demonstrated in a subgroup of our children. Therefore, a measurement of increased total testosterone in obese children likewise points to an increased free testosterone, which was also measured in obese adolescence and adults (2, 13). Additionally, SHBG levels increased in obese children losing weight, as demonstrated in a subgroup of our children. This underlines the decrease of testosterone activity in obese children losing weight. Third, testosterone and DHEAS concentrations correlate with fat distribution in adults and adolescents (14). This probably explains the different findings concerning androgens in childhood obesity. Because no established method exists in childhood to validate fat distribution (39), this confounding factor cannot be examined in our collective. Fourth, consideration of pubertal stage instead of binary dividing children in prepubertal and pubertal stage would be ideal to analyze the effect of weight status on androgens. This could probably explain why no different changes of testosterone levels were measured in pubertal boys losing and not losing weight. However, performing such a study needs a great sample because children have to be divided not only according to different pubertal stage but also according to gender and different degree of weight loss. Additionally, the division of pubic hair stages partially depends on the investigator. Fifth, weight loss may be too small to measure effects on DHEAS levels or testosterone levels in pubertal boys. Conversely, a reduction of more than 0.5 SDS-BMI, as achieved in our children with substantial weight loss, is reported to normalize many hormonal and metabolic changes in childhood obesity, such as insulin sensitivity (22), the cardiovascular risk factor profile (21), hormones (25), and adipocytokines (24). Finally, increased androgens could be a consequence of precocious pubarche, which is frequent at least in obese girls (5). Because androgens were analyzed according to pubertal stage and not to age, this confounder effect is excluded.
In summary, moderately increased testosterone and DHEAS serum levels were found in obese prepubertal children regardless of their gender and in obese pubertal girls, whereas androgen concentrations did not differ between obese and normal-weight pubertal boys. Weight loss induced a decrease in testosterone serum concentrations in obese prepubertal children and pubertal girls, pointing to a reversible increase of testosterone levels in obesity. There is still a great need for additional research to determine the cause-and-effect nature between obesity and hyperandrogenemia.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: BMI, Body mass index; DHEAS, dehydroepiandrosterone sulfate; IQR, interquartile range; SDS-BMI, BMI expressed as a SD score.
Received February 28, 2005.
Accepted July 5, 2005.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  C. R. McCartney, K. A. Prendergast, S. K. Blank, K. D. Helm, S. Chhabra, and J. C. Marshall Maturation of Luteinizing Hormone (Gonadotropin-Releasing Hormone) Secretion across Puberty: Evidence for Altered Regulation in Obese Peripubertal Girls J. Clin. Endocrinol. Metab., January 1, 2009; 94(1): 56 - 66. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Barat, D. E. W. Livingstone, C. M. C. Elferink, C. R. McDonnell, B. R. Walker, and R. Andrew Effects of Gonadectomy on Glucocorticoid Metabolism in Obese Zucker Rats Endocrinology, October 1, 2007; 148(10): 4836 - 4843. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. L. Rosenfield Identifying Children at Risk for Polycystic Ovary Syndrome J. Clin. Endocrinol. Metab., March 1, 2007; 92(3): 787 - 796. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. R. McCartney, S. K. Blank, K. A. Prendergast, S. Chhabra, C. A. Eagleson, K. D. Helm, R. Yoo, R. J. Chang, C. M. Foster, S. Caprio, et al. Obesity and Sex Steroid Changes across Puberty: Evidence for Marked Hyperandrogenemia in Pre- and Early Pubertal Obese Girls J. Clin. Endocrinol. Metab., February 1, 2007; 92(2): 430 - 436. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Poretsky Polycystic Ovary Syndrome--Increased or Preserved Ovarian Sensitivity to Insulin? J. Clin. Endocrinol. Metab., August 1, 2006; 91(8): 2859 - 2860. [Full Text] [PDF]
|
|
|
|
|
|
C. R. McCartney, K. A. Prendergast, S. Chhabra, C. A. Eagleson, R. Yoo, R. J. Chang, C. M. Foster, and J. C. Marshall The Association of Obesity and Hyperandrogenemia during the Pubertal Transition in Girls: Obesity as a Potential Factor in the Genesis of Postpubertal Hyperandrogenism J. Clin. Endocrinol. Metab., May 1, 2006; 91(5): 1714 - 1722. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
