Hydrocortisone dosage at 3 years of age is positively correlated with body mass index at 10 years in individuals with 21-hydroxylase deficiency

Abstract

Obesity is a major complication in individuals with 21-hydroxylase deficiency (21-OHD) and excessive hydrocortisone (HC) dosing is a likely contributing factor. However, the association between the HC dosage and body mass index (BMI) during childhood remains unclear. We previously reported a positive correlation between HC dosage during late infancy and a subsequent increase in BMI. In this study, we expanded our research to assess the impact of HC dosage during childhood. We retrospectively collected data on HC dosages and BMI in 28 patients with 21-OHD (9 boys and 19 girls) at both 3 and 10 yr of age and conducted correlation analyses between HC dosage and BMI. No significant cross-sectional associations were observed between the body-surface-area adjusted HC dosage (HC/BSA) and BMI standard deviation score (BMI-SDS) at either 3 or 10 yr of age. However, the HC/BSA at 3 yr positively correlated with the BMI-SDS at 10 yr, indicating a longitudinal relationship. Furthermore, children whose HC/BSA at 3 yr fell almost within the guideline-recommended range exhibited significantly lower BMI-SDS at 10 yr compared to those receiving higher doses. The longitudinal association between HC dosage and BMI suggests the importance of careful HC dose monitoring, although causality could not be established.

Highlights

● The HC dose did not cross-sectionally correlate with BMI at either 3 or 10 yr.

● HC dosage at 3 yr was longitudinally associated with BMI at 10 yr.

● Adjusting HC to remain within the guideline-recommended range is important.

Introduction

21-hydroxylase deficiency (21-OHD), the most common form of congenital adrenal hyperplasia, is a genetic disorder caused by loss-of-function variants of the CYP21A2 gene, leading to adrenal insufficiency and androgen excess (1, 2). The mainstay treatment for 21-OHD is the replacement of glucocorticoids, most commonly hydrocortisone (HC) and mineralocorticoids, to prevent adrenal crisis and manage excess androgen (3, 4). However, in clinical practice, supraphysiological doses of HC are often required to achieve adequate control of androgen excess (3, 4), which may result in adverse effects such as obesity and obesity-related complications (5,6,7,8,9,10,11,12).

Since childhood obesity is strongly associated with the future development of obesity (13), optimizing HC dosage during childhood is likely critical for improving long-term outcomes in patients with 21-OHD; however, there is insufficient evidence evaluating the impact of HC dosage on BMI during childhood. Accordingly, we previously assessed the association between HC dosage and BMI during the infancy and toddler periods and reported that higher HC dosages during late infancy were positively correlated with increases in BMI between 1 and 3 yr of age (14), highlighting the potential importance of appropriate HC dose adjustment during infancy in the management of body weight in the toddler period. In the present study, we expanded our research to investigate the association between HC dosage at 3 yr of age and BMI at 10 yr and to clarify the association between HC management during childhood and subsequent BMI status.

Materials and Methods

Ethical considerations

This study was approved by the Ethics Committee of Osaka Women’s and Children’s Hospital (Approval No. 1529). The opt-out recruitment method for participating in the study was applied with permission from the Ethical Review Board.

Participants

We retrospectively reviewed the medical records of patients with 21-OHD who received medical management at Osaka Women’s and Children’s Hospital between 1991 and 2020. Patients with nonclassical 21-OHD were excluded from the study. 21-OHD was diagnosed based on genetic analysis, urinary steroid profiling, and/or clinical findings. Individuals born with a birthweight < 2,500 g and/or classified as small for gestational age (SGA) were excluded to minimize the influence of birth size on anthropometric outcomes at evaluation. SGA was defined as either birthweight or length below −2.0 standard deviations (SD) based on sex-specific reference data for birthweight and length according to gestational age in the Japanese population (15). Of these 55 patients, 20 were excluded because they had not yet reached 10 yr of age. Among the remaining 35 individuals, data at either 3 or 10 yr of age were not available for 7, as many of them were referred to our hospital specifically for genital reconstructive surgery and, therefore, did not have regular follow-up visits. Accordingly, complete data on anthropometric parameters and medication dosages at both 3 and 10 yr of age were available for 28 patients (9 male and 19 female) who were included in the study. None of the included participants had disorders other than 21-OHD, which is known to cause growth retardation or obesity.

Data collection and study design

BMI was calculated by dividing the body weight (BW) (kg) by the square of standing height (HT) (m²). The body surface area (BSA) was calculated based on the formula provided by Du Bois as follows: BSA (m²) = WT (kg)^0.425 ×HT (cm)^0.725 × 0.007184. HT-SDS, BW-SDS, and BMI-SDS were determined based on the normal growth standards for Japanese children from a national survey in 2020 (16, 17). Anthropometric parameters, and HC and fludrocortisone (FC) dosages were measured at 3 and 10 yr of age. The HC dose was divided by BSA (HC/BSA) to adjust for body size.

Statistical analysis

Values were expressed as medians with interquartile ranges (IQRs). Comparisons of HT-SDS, BW-SDS, BMI-SDS, HC dosage, and FC dosage between 3 and 10 yr of age were conducted using Wilcoxon signed-rank tests. To assess the association between HC/FC dosage and obesity, correlations between HC/FC dosage (HC [mg/d], HC/BSA[mg/m²/d], FC [mg/d], and FC/BSA [mg/m²/d]) and BMI-SDS were evaluated using Spearman’s rank correlation coefficient. Additionally, the participants were divided into three groups based on HC/BSA at 3 yr of age: < 25th percentile (n = 6), 25th–75th percentile, and ≥ 75th percentile. The Kruskal-Wallis test was used to compare the BMI-SDS scores at 3 and 10 yr of age among the three groups. P was set than 0.05 were considered significant. Statistical analyses were performed using EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan) with a graphical user interface in R (R Foundation for Statistical Computing, Vienna, Austria).

Results

Characteristics of participants

The clinical characteristics at birth are shown in Table 1. The diagnosis of 21-OHD was based on genetic testing in 2 patients, biochemical analysis of urine steroid profiling in 6 patients, and both in 2 patients. The remaining 18 patients were clinically diagnosed based on clinical features, elevated blood 17-hydroxyprogesterone levels, and the presence of adrenal hypertrophy on imaging analysis. All the patients underwent HC and FC. Sodium chloride was not prescribed to the children at either 3 or 10 yr of age.

Table 1. Characteristics of participants at birth.

As shown in Table 2, the median ages at 3 and 10 yr of age were 2.99 yr (IQR: 2.95 to 3.06) and 10.21 yr (IQR: 10.13 to 10.30), respectively. The median of HC/BSA at 3 and 10 yr of age was 18.75 (IQR: 16.83 to 23.84) and 17.74 (14.97 to 20.18), and the dosage of HC was higher than that recommended in the current Japanese guideline (18). Spearman’s rank correlation coefficient revealed that HC/BSA at 3 yr was positively associated with current age (rS = 0.56, p = 0.002), suggesting that older individuals received higher dosages of HC. The dose of HC/BSA remained unchanged between the two time points. Although the FC was higher at 10 yr of age than at 3 yr, the BSA-adjusted value (FC/BSA) was significantly lower at 10 yr. Both HT-SDS and WT-SDS scores at 10 yr were significantly higher than those at 3 yr, whereas BMI-SDS remained unchanged. BMI SDS remained below +2.0 at both time points, except for 1 patient at 10 yr of age with a BMI-SDS of 2.24, suggesting that most patients were not obese in the study population.

Table 2. Clinical characteristics of participants at 3-yr and 10-yr examinations.

The HC dosage at 3 yr of age was positively correlated with BMI-SDS at 10 yr

As shown in Table 3, we investigated the association between HC dosage and BMI-SDS. Neither HC nor HC/BSA were correlated with BMI-SDS when cross-sectionally assessed at 3 or 10 yr of age; however, BMI-SDS at 10 yr of age was positively correlated with HC (rS = 0.48, p = 0.01) and HC/BSA (rS = 0.52, p < 0.01) at 3 yr. No association was detected between the FC dosage and BMI-SDS. These results indicate the potential longitudinal effect of HC exposure on BMI during childhood.

Table 3. Correlation analysis between BMI-SDS and dosage of hydrocortisone.

Patients treated with recommended dosage of hydrocortisone at 3 yr old showed a lower BMI at 10 yr old than those treated with higher dosages

We further examined the association between HC dosage and BMI by subdividing the participants into three groups based on HC/BSA at 3 yr of age: < 25th percentile, 25th–75th percentile, and ≥ 75th percentile. The median HC/BSA in each group was 12.83 mg/m²/d (IQR: 11.11–15.45; n = 6), 18.59 mg/m²/d (IQR: 17.37–21.80; n = 15), and 25.65 mg/m²/d (IQR: 24.90–35.45; n = 7), respectively (Fig. 1A). Notably, only the < 25th percentile group had HC/BSA values that fell predominantly within the guideline-recommended dosage range (individual values: 8.0, 12.1, 12.4, 13.3, 15.4, and 15.6 mg/m²/d). No significant difference in BMI-SDS was observed among the three groups at 3 yr of age, whereas there was a significant difference in BMI-SDS at 10 yr of age (Fig. 1B).

Fig. 1.

Association of Hydrocortisone dosage at 3 yr of age with BMI-SDS at 10 yr. A: The participants were divided into three groups based on HC/BSA at 3 yr of age: < 25th percentile, 25th–75th percentile, and ≥ 75th percentile, and the corresponding hydrocortisone dosage per body surface area was shown. B. BMI-SDS at 3 and 10 yr of age was shown. Box-and-whisker plots show that the box represents the interquartile range, with the line inside the box indicating the median. The whiskers extend to the minimum and maximum values within 1.5 × IQR. Cross marks denote the mean, while circles represent outliers.

Discussion

Obesity is a serious complication in individuals with classic 21-OHD, with a prevalence of 10.3% to 70% (19). A cohort study in the UK indicated an increased risk of obesity, metabolic syndrome, and cardiovascular comorbidities in adults with 21-OHD (20). A similar finding was reported in a retrospective analysis in the US (21). In line with these findings in adult populations, several cross-sectional studies have shown a higher rate of obesity in children with 21-OHD (22). For instance, Völkl et al. conducted a cross-sectional analysis of 89 children and adolescents with 21-OHD, aged between 0.2 and 17.9 yr, and showed that 16.8% had a BMI-SDS above 2.0 (23). Finkielstain et al. cross-sectionally assessed the frequency of obesity in 170 children aged 0.6 to 17 yr and found that approximately 35% were obese (24). Sarafoglou et al. showed that approximately 50% of the children had at least one BMI measurement ≥ 95 percentile in 194 children with 21-OHD (7). A cross-sectional study from Vietnam involving 201 children aged 1.1–16.5 yr showed that 28.9% of the participants were obese and 24.4% were overweight (25). Similar to findings reported overseas, evidence from Japan indicates elevated average BMI values, reaching up to +1.23 SD in male patients and up to +1.75 SD in female patients (26). Since pediatric obesity is strongly correlated with the development of adult obesity (27), these lines of evidence demonstrate the potential importance of managing obesity during childhood to improve long-term metabolic outcomes in individuals with 21-OHD.

To understand the pathogenesis of childhood obesity in 21-OHD, we previously examined the effect of HC dosage on BMI, as HC overdose is generally associated with an increased risk of obesity. In our previous study, we demonstrated that HC dosage during late infancy was positively associated with a subsequent increase in BMI between 1 and 3 yr of age in patients with 21-OHD (14), providing evidence of the potential importance of appropriate HC dosing during infancy for managing body weight during the toddler period. In the present study, we expanded our understanding of this association by investigating whether the relationship between HC dosage and BMI persists into childhood. We found a positive correlation between HC/BSA at 3 yr and BMI-SDS at 10 yr, suggesting that appropriate HC dosing patterns in toddlers are also important in the management of body weight during childhood.

Several studies have evaluated the relationship between HC dosage and BMI in children with 21-OHD. Völkl et al. conducted a cross-sectional analysis of 89 children and adolescents aged 0.2 to 17.9 yr and reported a positive association between HC dosage and BMI-SDS, although Pearson’s correlation coefficient was modest (23). Sarafoglou et al. found no significant association between HC dosage and height and age-adjusted BMI in a cohort of 194 children with 21-OHD (7). Similarly, a cross-sectional multicenter study involving 101 patients aged 8–18 yr showed no association between glucocorticoid dosage and BMI (12). These lines of evidence indicate that the association between the HC dosage and BMI is weak, even when assessed cross-sectionally. Consistent with these findings, our cross-sectional analysis did not reveal a significant association between the HC dosage and BMI. However, a key finding of our study was the longitudinal association between HC dosage at 3 yr of age and BMI at 10 yr. This longitudinal relationship suggests that the HC dosage in early childhood may have an impact on subsequent weight status during childhood.

Both global and Japanese clinical guidelines recommend an HC dose of 10–15 mg/m²/d during childhood (4, 18). In the present study, we showed that individuals treated with HC dosages almost within this recommended range at 3 yr of age exhibited normal BMI-SDS at 10 yr and individuals with higher HC dosages at 3 yr showed greater BMI-SDS at 10 yr. These findings underscore the importance of maintaining HC dosage within the recommended range during the toddler period to ensure appropriate weight control during childhood. However, achieving and maintaining this recommended dosage range in clinical practice may be challenging, partly because of the necessity of administering higher doses of HC to suppress excess adrenal androgen production. Indeed, data from the CAH registry in Germany demonstrated high rates of overtreatment, defined as an HC dose exceeding 15 mg/m²/d (28). Simultaneously, reducing the HC dose may increase the risk of adrenal insufficiency. Therefore, clinicians must carefully balance the risk of overtreatment with that of insufficient replacement, and HC dosing should be individualized from a comprehensive perspective on disease control.

Our study has several limitations that should be acknowledged. First, this study had a single-center design, which may have introduced selection bias, as the current population might not be representative of the broader population with 21-OHD. Additionally, the sample size was relatively small; therefore, multivariate analysis was not conducted. Large-scale studies in collaboration with multiple nationwide centers are necessary to better understand the association between HC dosage and obesity in individuals with 21-OHD. Second, the diagnosis of 21-OHD in many patients was based on clinical, laboratory, and imaging findings, partly because of the limited availability of genetic testing and/or urinary steroid profiling, especially in older patients. Although these patients were treated as 21-OHD and no clinical features inconsistent with 21-OHD were observed during follow-up, the possibility of other forms of congenital adrenal hyperplasia, such as 11β-hydroxylase deficiency, 3β-hydroxysteroid dehydrogenase deficiency, or cytochrome P450 oxidoreductase deficiency, cannot be completely excluded. Nevertheless, given that 21-OHD is the predominant form of congenital adrenal hyperplasia, we believe that the current results are minimally affected by the absence of definitive diagnostic confirmation. Third, owing to the observational nature of this study, causality could not be established form the present study. Fourth, unmeasured confounders, including disease severity, genetic factors, family history of obesity, physical activity, and dietary patterns, were not controlled in this analysis. These limitations indicate that the findings should not be interpreted as evidence that reducing HC dosage prevents obesity, as such changes could compromise disease control.

In conclusion, we demonstrated that the HC dosage at 3 yr of age was associated with elevated BMI at 10 yr. Children whose HC dosage was maintained within the guideline-recommended range at 3 yr exhibited normal BMI-SDS at 10 yr. These results suggest the importance of appropriate adjustment of HC dosage in weight management; however, causality has not been established, and verification through prospective interventional studies is required.

These findings provide preliminary evidence of an association between early HC dosing and later weight outcomes. However, clinical decision-making should continue to prioritize the established treatment guidelines that balance disease control and minimize side effects. Further interventional studies are needed to determine whether modifications to HC dosing protocols can safely influence long-term weight outcomes without compromising disease management.

Conflict of interests

The authors have no conflicts of interest to disclose.