Increased Abdominal Adiposity in Adolescents and Young Adults With Classical Congenital Adrenal Hyperplasia due to 21-Hydroxylase Deficiency

Abstract

Context:

Childhood obesity rates in congenital adrenal hyperplasia (CAH) exceed the high rates seen in normal children, potentially increasing their risk of cardiovascular disease (CVD). Abdominal adiposity, in particular visceral adipose tissue (VAT), is strongly associated with metabolic syndrome and CVD. However, it remains unknown whether VAT is increased in CAH.

Objective:

The objective of the study was to determine whether adolescents and young adults with classical CAH have more VAT and sc adipose tissue (SAT) than matched controls and whether VAT and SAT are associated with biomarkers of metabolic syndrome, inflammation, and hyperandrogenism in CAH.

Design/Setting:

This was a cross-sectional study at a tertiary center.

Participants:

CAH subjects (n = 28; 15.6 ± 3.2 y; 15 females) were matched for age, sex, ethnicity, and body mass index to healthy controls (n = 28; 16.7 ± 2.3 y; 15 females).

Main Outcome Measures:

VAT and SAT, using computed tomography imaging and serum biomarkers associated with CVD risk, were measured. Data are reported as mean ± SD.

Results:

Both VAT (43.8 ± 45.5 cm²) and SAT (288.1 ± 206.5 cm²) were higher in CAH subjects than controls (VAT 26.4 ± 29.6 cm² and SAT 226.3 ± 157.5 cm²; both P < .001). The VAT to SAT ratio was also higher in CAH subjects (0.15 ± 0.07) than controls (0.12 ± 0.06; P < .05). Within CAH, measures of obesity (waist to height ratio, fat mass) and inflammation (plasminogen activator inhibitor-1, high-sensitivity C-reactive protein, leptin) correlated strongly with VAT and SAT. In addition, homeostasis model assessment of insulin resistance, and low-density lipoprotein correlated with abdominal adiposity. There were no sex differences for VAT or SAT in CAH subjects.

Conclusions:

CAH adolescents and young adults have increased abdominal adiposity, with a higher proportion of proinflammatory VAT than SAT. An improved understanding of the mechanism of obesity in CAH may lead to targeted prevention and therapeutics in this high-risk population.

The prevalence of obesity in youth with classical congenital adrenal hyperplasia (CAH) due to 21-hydroxylase deficiency exceeds the alarmingly high rates seen in children and adolescents today (1–4). Because obesity is a key risk factor for cardiovascular disease (CVD), its heightened prevalence in CAH youth could result in increased rates of metabolic sequelae and CVD in adulthood in this population. Studies have suggested specific hormonal imbalances inherent in classical CAH and their treatment as likely contributors to obesity and CVD risk, including hyperandrogenism and life-long glucocorticoid therapy (1, 5, 6). In addition, circulating concentrations of adipokines such as leptin are abnormally elevated in both CAH and obese individuals and are important in the modulation of food intake, insulin sensitivity, and energy homeostasis. It is thus important to better understand the role of these factors in the development of obesity in CAH. In addition, CAH could serve as a model of hyperandrogenism to better understand classic sex differences in the prevalence of CVD in the general population.

Studies thus far have found increased fat mass in CAH youth, using simple skin-fold thickness and bioelectrical impedance analysis measures (4, 7). This increase in fat develops progressively during childhood (4), even in children under good hormonal control, and has been found in young adults with CAH by using whole-body dual-energy x-ray absorptiometry (WB DXA) (8). These fundamental studies of fat mass allow us to now further explore the role fat could play in the development of CVD risk in CAH.

Once thought of simply as a storage depot for energy, adipose tissue is now seen as a compartmentalized endocrine organ with multiple metabolic functions (9). We now also know that abdominal adiposity is positively associated with risk for metabolic disease, independent of total body adiposity (10). Visceral adipose tissue (VAT) is of particular concern because it is highly inflamed in individuals with obesity and metabolic syndrome (11) and produces inflammatory substances associated with CVD. Little is known about the metabolic role of abdominal adipose tissue in CAH patients, and imaging studies of specific fat depots have yet to be performed in this at-risk cohort. Therefore, we sought to study abdominal adiposity in adolescents and young adults with classical CAH, with a focus on quantifying visceral and sc adipose tissue (SAT) compartments with computed tomography (CT) imaging. We hypothesized that abdominal adiposity would be increased in subjects with CAH. We also measured metabolic and inflammatory markers and quantified androgens and hormonal control in these same individuals.

Participants and Methods

The study was cross-sectional and was approved by the Children's Hospital Los Angeles Institutional Review Board. Written consent was obtained from all parents and/or participants, and all minors up to 14 years of age gave assent.

We studied 28 adolescents and young adults with CAH recruited from the pediatric endocrinology clinic and rigorously matched for age, sex, ethnicity, and body mass index (BMI) to controls who were healthy other than overweight/obesity. Controls were recruited from the general pediatric and endocrinology clinics at our center. Anthropometric measures of height (centimeters) and weight (kilograms) were obtained in all participants. Waist circumference (WC) was measured from CT images using Synedra View (Synedra Information Technologies). Abnormal WC was defined using the Centers for Disease Control and Prevention criteria of greater than the 90th percentile if younger than 20 years old (12), greater than 102 cm for adult males, and greater than 88 cm for adult females (13). BMI (kilograms per square meter) and waist to height ratio (WHtR) were calculated. BMI z-score and Height z-score were calculated using Epi Info 7.1.4 (Centers for Disease Control and Prevention). WHtR is emerging as an important measure of abdominal obesity, with a strong predictive value for metabolic syndrome and CVD risk (14), and an abnormal WHtR was defined as greater than 0.5 (15).

CT imaging

A single, cross-sectional CT slice (CT HiLight Advantage; GE Medical Systems) was used to quantify VAT and SAT at the level of the umbilicus. Studies were obtained using the following technical factors: 80 kVp, 70 mA, 2 seconds, and 1-cm slice thickness. Analyses were performed using a graphical interface created with Matlab (The Mathworks). VAT and SAT were defined as previously described (16) and were determined by measuring voxels in the area with negative Hounsfield units after manually excluding the bowel for VAT analyses. Intra- and intercoefficients of variation for repeated measurements ranged from 1% to 3% (17). The VAT to SAT ratio was calculated.

CAH adolescents: additional measures

After an overnight fast (12 h) and prior to routine morning CAH medications, adolescents with CAH had their blood drawn at our Clinical Trials Unit for genotyping and measurement of analytes. Analytes were measured at Quest Diagnostics Nichols Institute as follows: leptin by electrochemiluminescence; insulin, SHBG, and plasminogen activator inhibitor-1 (PAI-1) by RIA; 17-hydroxyprogesterone (17-OHP), androstenedione, plasma renin activity, and total and free T by liquid chromatography tandem mass spectrometry; high-sensitivity C-reactive protein (hs-CRP) by nephelometry; and plasma catecholamines by HPLC. Glucose and lipids [total cholesterol, low density lipoprotein (LDL), high density lipoprotein (HDL), very low density lipoprotein (VLDL), and triglycerides] were analyzed at the Children's Hospital Los Angeles Laboratory (Los Angeles, California). CYP21A2 mutations were determined by multiplex ligation probe analysis (Esoterix Laboratory Services). Hormonal control in CAH patients was assessed by their early-morning 17-OHP. Individuals were considered suppressed if the 17-OHP was less than 100 ng/dL, acceptable between 100 and 1200 ng/dL, intermediate control if greater than 1200–5000 ng/dL, and under poor control if 17-OHP greater than 5000 ng/dL. Insulin sensitivity was quantified by calculated homeostasis model assessment of insulin resistance [HOMA-IR = (fasting insulin [microinternational units per milliliter] × fasting glucose [millimoles per liter])/22.5]. Elevated HOMA-IR index was defined as greater than 3.16 in adolescents (18) and greater than 2.5 in adults.

Blood pressure was determined from the average of three separate measurements. For youth younger than 18 years old, prehypertensive was classified as systolic blood pressure (SBP) or diastolic blood pressure (DBP) at the 90th to 94th percentile and hypertensive as SBP or DBP at the 95th percentile or greater (19). For young adults, prehypertensive was classified as SBP of 120–140 mm Hg or DBP of 80–90 mm Hg and hypertensive as SBP of 140 or greater or DBP of 90 mm Hg or greater (20). WB DXA (Hologic Delphi W) was used to assess the total fat mass and the regional body fat distribution (eg, trunk fat). Metabolic syndrome was assessed using International Diabetes Federation criteria (21).

Glucocorticoid dose equivalencies for longer-acting glucocorticoids were calculated based on their growth-suppressing effects compared with hydrocortisone: the prednisone dose was multiplied by 5 and the dexamethasone dose was multiplied by 80 (1). A medical and family history pertaining to CVD, hypertension, type 2 diabetes, and smoking was obtained.

Statistical analyses

Paired-samples t tests and Wilcoxon signed rank tests were used to determine whether there was a statistically significant difference between CAH and matched controls. Bivariate associations (Pearson's product moment correlations and Spearman's rank order correlations) were performed to assess various relationships within the CAH group. Data are reported as mean ± SD. Analyses were performed using SPSS (version 21; IBM Corp). All P values have been adjusted using the Benjamini-Hochberg false discovery rate controlling procedure to address false-positive results that may arise from performing multiple statistical comparisons (22, 23). Nominal P values associated with the entire set of comparisons were ordered, adjusted critical values based on the ordered position of the test were computed, and the nominal P values were compared with the adjusted critical values to ascertain statistical significance. Using this procedure, one comparison previously considered to be significant at P < .05 (correlation between VAT and HDL cholesterol levels) was excluded after adjustment for multiple comparisons. Adjusted significance levels are presented in the text and tables.

Results

Baseline characteristics

Baseline characteristics of CAH and matched controls are shown in Table 1. The CAH group was confirmed to have 21-hydroxylase deficiency by genotype. It was comprised of 71% salt-wasting type and 29% simple-virilizing type, as determined by genotype for 90.5% of the cohort; six subjects with multiple heterozygous mutations identified on PCR analysis were classified based on clinical phenotype. Of the CAH subjects, 60.7% were overweight or obese and 53% had an abnormal WHtR. Seventy-five percent of CAH youth younger than 18 years old were normotensive, 11% were prehypertensive, and 14% were hypertensive. The four young adults with CAH (≥ 18 y old) had normal blood pressures. Only one CAH patient had metabolic syndrome.

Table 1.

Clinical Characteristics of Adolescents and Young Adults With CAH due to 21-Hydroxylase Deficiency and Matched Controls

	CAH (n = 28)	Control (n = 28)
Sex, femalea	15	15
Age, ya	15.6 ± 3.2 (11.8–24.1)	16.7 ± 2.3 (10.8–22.8)
Ethnicitya	19 Hispanic (68%)	19 Hispanic (68%)
	5 non-Hispanic white (18%)	5 non-Hispanic white (18%)
	3 Asian (11%)	3 Asian (11%)
	1 African-American (3%)	1 African-American (3%)
Weight, kg	67.5 ± 17.7	74 ± 18.6
Height, cm	156.5 ± 10.5	164.8 ± 8.3b
Height z-score	−0.96 ± 1.29	−0.24 ± 0.80
BMI, kg/m2a	27.8 ± 8.2	27.2 ± 6.7
BMI z-scorea	1.31 ± 1.25 (n = 7 with z-score > 2)	1.06 ± 1.12 (n = 5 with z-score > 2)
WC, cm	89.68 ± 17.9	90.68 ± 15.32
WHtR	0.55 ± 0.11	0.55 ± 0.10
Glucocorticoid dose, mg/m2 · d	19.5 ± 5.4	N/A
Fludrocortisone dose, mg/d	0.1 ± 0.05 (n = 24)	N/A

N/A, not available.

^aMatching criteria (mean ± SD).

^bP < .01.

The mean glucocorticoid dose of hydrocortisone equivalents was 19.5 ± 5.4 mg/m² · d with nine patients on a longer-acting glucocorticoid (eight on dexamethasone). The average fludrocortisone dose was 0.1 ± 0.05 mg/d, with four simple-virilizing patients not on this treatment. There was no difference in mineralocorticoid doses between hypertensive and nonhypertensive subjects (P = .35). CAH males had a mean 17-OHP level of 9725.2 ± 9844 ng/dL (SI unit conversion × 0.0302 = nanomoles per liter), androstenedione 284.2 ± 261.1 ng/dL (SI unit conversion × 0.0349 = nanomoles per liter), T 405.9 ± 211.5 ng/dL (SI unit conversion × 0.0347 = nanomoles per liter), free T 63.6 ± 34.8 pg/mL (SI unit conversion × 3.47 = picomoles per liter), and SHBG 35.7 ± 13 nmol/L. CAH females had a mean 17-OHP level of 2169.1 ± 2447.5 ng/dL; androstenedione, 124.6 ± 79.6 ng/dL; T, 21.7 ± 14.5 ng/dL; free T, 4.1 ± 3.7 pg/mL; and SHBG, 28.3 ± 16.2 nmol/L. As assessed by cross-sectional 17-OHP levels, there were four suppressed patients, seven patients in acceptable control, eight patients in intermediate control, and nine patients in poor control. CAH patients collectively had a mean plasma renin activity 5.42 ± 4.74 ng/mL · h (SI unit conversion × 1.0 = micrograms per liter per hour).

Abdominal adiposity

The main finding of this study was significantly increased abdominal adiposity in adolescents and young adults with CAH compared with controls, for both VAT (Figure 1A; CAH, 43.8 ± 45.5 vs controls, 26.4 ± 29.6 cm², P < .001) and SAT (Figure 1B; CAH, 288.1 ± 206.5 vs controls, 226.3 ± 157.5 cm², P < .01). The VAT to SAT ratio was also significantly higher in CAH than controls (Figure 1C; CAH, 0.154 ± 0.067 vs controls, 0.118 ± 0.064, P < .05).

Figure 1.

Abdominal adiposity in adolescents and young adults with classical CAH and matched controls. A, VAT is higher in CAH compared with controls (P < .001). B, SAT is higher in CAH compared with controls (P < .01). C, VAT to SAT ratio is higher in CAH compared with controls (P < .05).

Within the CAH group, measures of obesity and body composition correlated positively with VAT and SAT, including BMI z-score, WHtR, and trunk and total body fat mass (Table 2). In terms of circulating metabolic markers, leptin correlated positively with VAT and SAT (Table 2). Of note, 50% of the CAH cohort had elevated leptin levels, which were significantly higher in obese (29.2 ± 18.9) compared with nonobese subjects (9.4 ± 7.3; P < .05) and correlated strongly with measures of obesity (BMI z-score, P < .05; WHtR, P < .001; and total fat mass, P < .001). The HOMA-IR index was elevated in 18% of patients and correlated positively with VAT and SAT (Table 2). Fasting lipids (total cholesterol, LDL, VLDL, and triglycerides) correlated positively with both VAT and SAT as well (Table 2). There were no significant mean differences between hypertension categories, when comparing categories with abdominal adiposity. Family history for CVD or CVD risk factors such as smoking, hypertension, and type 2 diabetes did not show any correlations with abdominal adiposity.

Table 2.

Abdominal Fat Correlations With Body Composition and Metabolic and Inflammatory Markers in Adolescents and Young Adults With CAH due to 21-Hydroxylase Deficiency

	VAT	SAT
Body composition
BMI z-score	0.44a	0.92b
WC, cm	0.82c	0.95b
WHtR	0.84c	0.96b
Trunk fat mass, g	0.75b	0.94b
Total fat mass, g	0.75b	0.96b
Metabolic
Leptin, ng/mL	0.43a	0.73b
HOMA-IR	0.54a	0.52a
Total cholesterol, mg/dL	0.61c	0.60c
HDL, mg/dL	0.39	0.16
LDL, mg/dL	0.50a	0.58c
VLDL, mg/dL	0.53a	0.47a
TG, mg/dL	0.52a	0.47a
Inflammatory
PAI-1, ng/mL	0.56c	0.48a
hs-CRP, mg/L	0.42a	0.63c

Abbreviation: TG, triglycerides.

^aP < .05.

^bP < .01.

^cP < .001.

We found positive correlations between abdominal adipose tissue amount and inflammatory markers in CAH subjects (Table 2). Both VAT and SAT showed positive correlations with the adipokine, PAI-1, and with hs-CRP, which is known to correlate with CVD risk in adults (24). When examining obese vs nonobese CAH subjects, PAI-1 differed between the groups (higher in obese subjects, 27.2 ± 26.4 vs nonobese, 11.4 ± 5.0 ng/mL, P = .04), although hs-CRP did not differ between the groups (obese, 3.08 ± 3.46 vs nonobese, 1.59 ± 1.48 mg/L, P = .14).

Within the CAH group, we found no sex differences in VAT (males, 47.2 ± 60.1 vs females, 40.8 ± 29.5 cm²) or SAT (males, 246 ± 198.1 vs females, 324.6 ± 213.4 cm²). We also did not find a sex difference in our controls for VAT (males, 25.4 ± 33.8 vs females, 27.4 ± 26.5 cm²) or SAT (males, 180.5 ± 134.7 vs females, 265.9 ± 169.3 cm²). Although serum androgen levels did not show significant correlations with abdominal adiposity in CAH, there was a negative correlation between SHBG and both VAT and SAT (Table 3). There were no correlations between other markers of hormonal control (17-OHP) or the glucocorticoid dose with VAT or SAT in the CAH group. Because the total catecholamines were measured in only six subjects, statistical associations between catecholamines and adiposity were not examined.

Table 3.

Abdominal Fat Correlations With Androgens and Hormonal Control in Adolescents and Young Adults With CAH due to 21-Hydroxylase Deficiency

	VAT	SAT
SHBG	−0.45a	−0.56b
T	−0.07	−0.17
Androstenedione	−0.20	−0.29
17-OHP	−0.14	−0.17
Glucocorticoid dose	0.07	0.004

^aP < .05.

^bP < .01.

Discussion

This study demonstrates that adolescents and young adults with CAH have significantly increased amounts of both visceral and sc abdominal adipose tissue compared with matched controls. We already see an increased prevalence of obesity in adolescents with CAH, above the epidemic rates seen in normal children, and now we see that these adolescents and young adults have a more unfavorable abdominal fat distribution for the same degree of obesity. This places CAH patients at even greater risk for harmful metabolic sequelae from obesity, with the distinct possibility that increased abdominal adiposity could be actively promoting pathogenic processes linked to CVD risk in CAH.

Over the past 2 decades, there has been an increasing focus in obese individuals on specific regions of fat accumulation, in particular VAT (10), and adipose tissue is now regarded as an active endocrine organ. Rich in macrophages and producing adipokines such as PAI-1, VAT is well known to play a harmful role in obesity and actively promotes inflammation in the body (24). The strong correlations between VAT and several adipokines/inflammatory markers (leptin, PAI-1, and hs-CRP) in our CAH youth support this potentially pathological role of VAT. The implications of increased abdominal adipose tissue in CAH include the following: 1) an association with insulin resistance and metabolic syndrome exists in youth (25); 2) adipokines can act centrally and stimulate regions of the brain regulating appetite; and 3) inflammation is a CVD risk factor itself (24) and it is possible that adolescents and young adults with CAH have similar systemic low-grade inflammation as obese individuals without CAH (25).

Furthermore, our subjects with CAH exhibit a higher VAT to SAT ratio than controls, which is concerning, given that an increased proportion of VAT to SAT constitutes a particularly adverse metabolic phenotype in obese adolescents (25). Although only one CAH subject in our study had metabolic syndrome, we found strong correlations between VAT and the main components of the metabolic syndrome in our CAH cohort, including obesity, insulin resistance, and LDL levels.

Patients with CAH have several reasons to develop elevated levels of the adipokine, leptin, including epinephrine deficiency (26) and an altered leptin axis due to decreased soluble leptin receptor (27). Our classical CAH cohort exhibited elevated leptin levels, with a predominance in obese subjects, similar to other CAH studies (1, 26, 27). However, it may be the dysregulation of leptin suppression seen in CAH (28) that is important in the development of obesity in CAH. Leptin has already been linked to insulin resistance in CAH (26) and acts centrally as a key regulator of food intake and energy balance. We confirmed that leptin levels correlate with abdominal adipose tissue, more strongly in SAT, similar to studies of adiposity in individuals without CAH (29). More study is needed to understand the exact role of leptin and related neuroendocrine pathways between abdominal adipose tissue and the brain in the development of metabolic sequelae in individuals with CAH.

The inherent hormonal imbalances in CAH of hyperandrogenism and glucocorticoid deficiency are important to consider in the pathophysiology of obesity in this cohort, with evidence of cross talk between androgens and glucocorticoids at the neuroendocrine and peripheral levels and functional hyperandrogenism associated with increased abdominal adiposity in adult women (30). An interesting finding in this study was the lack of sex differences in CAH adolescents with regard to abdominal adipose tissue, which could suggest a propensity toward an android phenotype of abdominal obesity in females with CAH secondary to exposure to elevated androgens over a lifetime. Although our cohort of controls did not show a difference in VAT between males and females, regardless of BMI, normative data show increased VAT in males compared with females as of late adolescence (14), with typical sex differences in body composition primarily attributable to sex steroids (31). It is well known that adult men have twice the visceral fat of adult women (16) and that there could either be a pathological role for excess androgens or a lack of protective effect from estrogen (5). We did not see significant correlations between either androstenedione or T with abdominal adipose tissue in our subjects with CAH. However, a strong negative correlation between abdominal adiposity and SHBG was noted, with known associations between low SHBG levels and higher free T as well as insulin resistance, obesity, and dyslipidemia in adults (32, 33).

Understanding the mechanisms by which androgens could potentially equalize sex differences between females and males with CAH could help understand typical sex differences in CVD in non-CAH subjects. However, untreated CAH patients would need to be studied to better assess the relationship between hyperandrogenism and adiposity, which would not be ethically possible in classical CAH. Hormonal control or treatment was not a modulating factor in these patients, with no correlations between 17-OHP or glucocorticoid doses and abdominal adipose tissue, despite our patients being on relatively high mean glucocorticoid doses overall. There was also no difference in glucocorticoid dose between obese and nonobese subjects. However, it is known that patients with much higher glucocorticoid levels as seen in Cushing syndrome have increased visceral adipose tissue and obesity (34), and thus, the contribution of chronic, supraphysiological glucocorticoid replacement over a lifetime to the development of CVD risk in CAH needs to be further evaluated.

A limitation of our study was the lack of dietary intake information at the time of the study visit. In addition, we performed a single assessment as part of a cross-sectional design. It would be ideal to follow up patients longitudinally to perform an association analysis of glucocorticoid and androgen exposure and visceral fat. Although most CAH subjects were not in acceptable hormonal control as assessed by their early morning 17-OHP level, this is similar to a large CAH cohort reported to have approximately 30% of patients in acceptable control (1). Additionally, we do not have plasma catecholamine levels for all CAH cases studied. The association between total catecholamines and abdominal adiposity merits further study in CAH youth, with implications in the regulation of leptin and lipolysis in the development of increased abdominal adiposity in CAH youth.

We conclude that adolescents and young adults with CAH exhibit a concerning and likely pathological tendency toward visceral adiposity. The combination of proinflammatory VAT, hormonal imbalances inherent in CAH, and increased propensity toward obesity is concerning in these individuals. More studies are necessary to elucidate underlying mechanisms of obesity in CAH, including neuroendocrine pathways, and the role of catecholamine deficiency. In the clinical setting, adolescents with classical CAH should be evaluated for central abdominal obesity, with a simple obesity measure such as WHtR that highly correlates with VAT. We can then target these at-risk individuals for preventive weight loss and therapeutics to decrease long-term adverse outcomes in CAH.