Priming Short Children with Sex Steroids prior to Growth Hormone Testing Decreases the Frequency of Divergent Results Open Access

Diagnosing growth hormone deficiency (GHD) in children is challenging due to the complexities involved in GH testing and interpretation of results. Various stimuli, including clonidine, insulin, glucagon, and arginine, are used for GH stimulation tests [1]. However, no single stimulus is considered superior to others [2]. Current guidelines recommend that at least two different stimuli are used, as relying on a single test often yields low GH levels even in healthy children, increasing the risk of misdiagnosis [3]. In addition to stimulation tests, spontaneous GH secretion is commonly used to assess GHD [4], with somewhat better reproducibility but lower sensitivity than stimulation tests [3, 5, 6]. We and others have suggested that combining spontaneous nocturnal GH measurements with stimulation tests can reduce the risk of overdiagnosing GHD in children [7, 8]. An exact cutoff level for peak GH concentration when diagnosing GHD in children has not been established, and peak levels differ between stimuli given [9, 10]; however, a suggested cutoff level of approximately 7 µg/L is widely used and accepted [2, 3, 11]. Severe GHD is often defined as a stimulated peak GH concentration <3 µg/L [12, 13].

Another challenge in diagnosing GHD in children and adolescents arises from the influence of puberty on circulating GH levels [14]. Prepubertal children have physiologically low activity in the hypothalamic-pituitary-gonadal axis, leading to low concentrations of both sex steroids and GH [15]. Simultaneously, the physiological decrease in growth velocity prior to pubertal onset may trigger GHD investigations [16]. Previous studies have shown that at least half of prepubertal children diagnosed with GHD later prove to be GH sufficient when retested during puberty or adulthood [17‒19], indicating that the GHD diagnosis may have been incorrect.

Short-term sex steroid treatment prior to GH testing, a procedure called priming, increases stimulated GH secretion in healthy children with no or early puberty [20, 21]. This increase could lower the risk of false-positive results from GH testing and improve the diagnostic accuracy in prepubertal children and children in early puberty [22‒24]. Adolescent boys with growth failure who responded to GH stimulation tests only after priming reached normal final heights even though they did not receive GH treatment [25]. In an audit involving eight European countries and the USA, priming was recommended in five of nine countries [4]. However, no overall consensus regarding priming has been reached [2], different protocols are used, and the short-term effects of priming in a clinical setting using both spontaneous nocturnal GH secretion and stimulation tests are not fully understood. We have previously shown a high frequency of divergent results when combining a spontaneous nocturnal GH test and the arginine-insulin tolerance test (AITT), i.e., only one of the two tests showed pathological values [7], and we hypothesized that priming reduces the frequency of divergent results. Thus, this study investigated whether priming was associated with a reduced frequency of divergent results. We also studied the effects of priming on spontaneous GH secretion, potential short-term adverse events linked to priming, and the feasibility of priming in a clinical setting.

This was a single-center, retrospective, population-based, observational study.

Eligible patients were children (0–18 years of age) who had been evaluated for possible GHD at the Department of Pediatrics at Örebro University Hospital, Örebro, Sweden, from the 1st of January 1993 until the 28th of February 2023. During the investigation, the children were routinely evaluated using a nocturnal spontaneous GH test, followed by a stimulation test (AITT). The clinical decision to perform GH testing was based on a comprehensive evaluation of each patient’s medical history, parental heights, auxological parameters including height velocity and sitting height index, as well as clinical and biochemical findings, indicating that the short child might have GHD. In some cases, the intravenous insulin injection or another part of the testing were considered unsuitable. In such instances, alternative testing was performed, for example nocturnal sampling only. Children who underwent only one test or with an incomplete stimulation test, e.g., only one stimulus was given, or those who exhibited insufficient hypoglycemia (defined as plasma glucose >2.7 mmol/L), were excluded from the study. The study was approved by the Regional Board of Ethics, Uppsala, Sweden (registration number: 2017/358, with an amendment approved by the Swedish Ethical Review Authority, registration number: 2023-01820-02).

To identify patients who underwent an investigation for GHD during the study period, we searched hospital registries for appropriate diagnostic and procedure codes (online suppl. material; for all online suppl. material, see https://doi.org/10.1159/000546884). Some data from the first part of the study period (1993–2017) have been presented previously [7]; however, those data did not focus on priming. In this study, additional data (2017–2023) were merged with the preexisting data (Fig. 1). All patients who fulfilled the inclusion criteria for this study but not the exclusion criteria were included, and their data were extracted in a structured format from the medical records. All medical records were available for new patients.

It has long been standard practice at our clinic to assess both spontaneous nocturnal GH secretion and conduct an AITT the following morning. A pediatric endocrinologist assessed the pubertal development according to Tanner shortly before the GH investigation [26]. Before 2017, priming with sex steroids was rarely performed and when it was given, only children in Tanner stages I–II aged >10 years were primed. Starting in 2017, sex steroid priming has been more frequently used and consistently at our clinic in accordance with changes in local guidelines. Since 2017, the criteria for administering priming have been Tanner stages I–III and age ≥7 years initially [27], later revised to 8 years for girls and 9 years for boys [3]. In most cases, ethinyl estradiol was given to both boys and girls at a daily dosage of 40 µg/m² body surface, divided into two doses on the 2 days prior to GH testing [20, 28]. When ethinyl estradiol was not available, estradiol valerate was given instead as a single daily dose of 1 mg (body weight <20 kg) or 2 mg (>20 kg) for 3 days prior to GH testing [29]. Testosterone enanthate was occasionally administered to boys as an intramuscular injection (100 mg as a single dose 5–10 days prior to GH testing) [30, 31]. Blood samples were taken every 30 min for 12 h to measure spontaneous nocturnal GH secretion. The AITT started immediately after nocturnal sampling with an arginine infusion (30 min, 0.5 mg/kg body weight), followed by an intravenous injection of insulin (0.1 IU/kg body weight). The insulin dose was reduced to 0.05–0.08 IU/kg body weight for younger children [32]. Blood samples were collected every 15 min during the AITT to determine the GH concentration. During both the nocturnal and stimulation tests, all children fasted. A divergent result was defined as only one of the tests, spontaneous or stimulation test, yielding a peak GH value of ≥7 μg/L (the chosen cutoff value in this study). The target height was calculated as the mean parental height minus 6.5 cm for girls or plus 6.5 cm for boys. Standard deviation (SD) scores for weight and height were calculated using Swedish reference data [33]. Bone age was assessed manually by a radiologist according to the Greulich-Pyle standard at the Department of Radiology, Örebro University Hospital, until mid-2017. Thereafter, assessments were performed using the automated BoneXpert software [34].

GH concentrations were analyzed at the Department of Clinical Chemistry at Örebro University Hospital. During the study period, the analysis method changed (see online suppl. material). The detection range for GH was 0.05–40 µg/L. Values below and above these limits were set to 0.03 µg/L and 40.1 µg/L, respectively, in this study, as in our previous study [7], to enable the inclusion of all values in the statistical analyses.

Data are presented as medians (min–max), means ± SD, or absolute numbers (percent). The Kolmogorov-Smirnov and the Shapiro-Wilk tests were used for normality testing, alongside ocular assessment of Q-Q plots and boxplots. Student’s t test and the Mann-Whitney U test were used when comparing continuous variables between groups, as appropriate. Proportions were compared using the chi-square test or Fisher’s exact test, as appropriate. IBM SPSS Statistics version 29 (IBM Corporation^®, Armonk, NY, USA) was used for the statistical analyses. A p value <0.05 was considered statistically significant.

The search identified 33 new patients, three (9%) of whom had incomplete tests and were excluded. The remaining 30 patients were included and added to our former cohort of 102 patients with complete tests (Fig. 1). The new patients included 19 boys (63%) and 11 girls (37%) aged 7.6 years (2.6–14.0 years) and 10.1 years (4.8–13.0) years, respectively, at the time of GH testing. Sixteen (53%) had been primed with sex steroids. In the total study population, 25 patients (19%) received priming. The primed children were older than the non-primed children but did not differ in other clinical characteristics. A bone age assessment was performed within 1 year from the GH testing in 19 of the primed (76%) and 59 of the non-primed children (55%), respectively. The bone ages did not differ significantly between the groups (Table 1).

Two of the 25 primed children (8%) failed both GH tests, whereas 19 of the 107 non-primed children did (18%, p = 0.363). In total, 26 children (20%) had divergent results between the two tests. Among the 25 primed children, only one child (4%) had a divergent result (<7 μg/L in the AITT only) compared with 25 of the 107 non-primed children (23%, p = 0.027, Fig. 2). Among the non-primed children with divergent results, 16 had GH peak values <7 µg/L in the AITT only, and 9 had peak values <7 µg/L in the nocturnal test only.

Both the maximum peak and mean GH values during the nocturnal test were significantly greater in primed children than in non-primed children. However, the stimulated peak values did not differ between the two groups (Table 2). The interindividual variation, expressed as variance and SD, of stimulated and spontaneous GH peak values and mean spontaneous values was greater in the non-primed group (Table 2). However, the intraindividual variation in GH peak values between the two tests did not differ between primed and non-primed children. The three primed children diagnosed with GHD after the testing (see Table 2) consisted of two children with low peaks in both tests (see Fig. 2), one with panhypopituitarism and one with isolated GHD. The third child was diagnosed clinically with isolated GHD even though he had a peak stimulated GH value of 8.7 μg/L and a peak spontaneous value of 7.3 μg/L, reflecting the previous clinical use of a higher cutoff level than the one applied in this study.

Since priming was used more frequently and consistently after a particular year (2017) and only applicable to certain ages, we performed sensitivity analyses. Compared with the children investigated in 2017 or later, the children investigated before 2017 had, as expected, a greater frequency of divergent results (24% vs. 6%, respectively, p = 0.028). However, after excluding the primed children, this difference was no longer statistically significant, indicating that the year of testing had no profound impact on our results. Furthermore, when the year of the GH test was analyzed as a continuous variable and age at the time of testing and still excluding the primed children, no significant differences were found for these variables between children with and without divergent results.

Three primed children (12%) experienced potential adverse events during the GH test, and they were all given ethinyl estradiol. All three experienced vomiting. One child also reported abdominal pain, and another reported dizziness. No cases of gynecomastia or development of pubertal signs were reported. With respect to feasibility, there were no problems with access to or handling of the medication used for priming. There were neither any reports suggesting that GH testing was postponed or aborted due to adverse events or a shortage of priming medication.

We report the short-term effects of sex steroid priming in children undergoing GH testing in a clinical setting. Compared with non-primed children, primed children had a lower frequency of divergent results between a nocturnal spontaneous GH test and an AITT. The spontaneous GH values, but not the stimulated peaks, were higher in primed children. Few adverse events were linked to priming, and the feasibility appeared good.

The frequency of divergent results (20%) found in the study population aligned with previously reported data [5, 6, 35, 36]. However, in the primed children in this study, the frequency was considerably lower (4%). The only difference in clinical characteristics between primed and non-primed children in this study was the age at testing (see Table 1), and sensitivity analyses indicated that age did not influence the frequency of divergent results. Neither did the year of testing impact our results, and the pubertal stage did not differ between primed and non-primed patients, indicating the robustness of the results. Thus, the low frequency of divergent results in primed children may be a direct consequence of priming and, more specifically, its effect on spontaneous nocturnal GH peak values, which were found to increase after priming. Furthermore, priming appeared to be negatively associated with the tendency to diagnose a child with GHD in the present study. Compared with 21% of the non-primed children, only 12% of the primed children were diagnosed with GHD; however, owing to the small number of participating children, this difference did not reach statistical significance. In addition, not all primed children in this study reached GH peaks ≥7 µg/L, indicating that priming may not conceal true or severe GHD. Indeed, a child with panhypopituitarism in the present study had low GH peaks in both tests, even though being primed. Furthermore, none of the primed children with GH peaks ≥7 µg/L exhibited clinical or radiological evidence indicative of severe GHD. Overall, we suggest that the use of sex steroid priming in short children evaluated with both spontaneous and stimulation tests may yield a more accurate GHD diagnosis. This suggestion is in accordance with the findings of previous studies, where the effects of priming and its consequences for the use of GH replacement have been followed to final height [24, 25]. The latest American guidelines recommend priming for certain children [22], with similar suggestions from the UK [3] and Germany [37]. However, observational studies such as our study can provide only limited evidence. Large, randomized, controlled studies are needed for a more comprehensive investigation of the effects of priming, but such studies are almost impossible to perform now when GH testing and therapy have been practiced for decades. Nevertheless, future studies need to confirm our findings, and they should include long-term follow-up to assess final height outcomes.

Priming has predominantly been used prior to stimulation tests [38, 39] and has previously been shown to increase stimulated GH peak levels when various stimuli are used [20, 21, 40], but to the best of our knowledge, this has not been previously reported for spontaneous GH secretion. Although this finding is not surprising, it is essential to understand the short-term effect of priming on spontaneously secreted GH concentrations since a spontaneous nocturnal GH secretion test can be the only GH investigation performed in some pediatric departments. Reluctance to use priming and reliance solely on spontaneous or stimulated GH secretion may increase the risk of overdiagnosing GHD in children, which we reported in our previous publication [7], where children investigated with one GH test only were more often diagnosed with GHD than children investigated with both stimulation and spontaneous tests. The reason why we could not confirm the findings of others concerning increased stimulated GH peaks after priming is unclear but may be linked to the considerable variation in GH response to stimulation in our study relative to the size of our study population. Similar findings, i.e., no difference in stimulated GH peaks between primed and non-primed children, have been reported in two previous studies [39, 41] with similar or smaller study populations than the one used in our study.

Delayed skeletal maturation may be found in children with GHD and is typically seen in children with delayed puberty [4, 42, 43]. Previous studies have reported on an increased GH response following sex steroid priming in individuals with constitutional delay of growth and puberty [44] and priming in such situations has been recommended in clinical guidelines [22, 45]. In the present study, the bone ages were similarly delayed in primed and non-primed children.

The adverse events of priming reported in the present study were few (12%), transient, and mild. Previously reported adverse events include transient, moderate breast enlargement in girls after the administration of estradiol valerate [29] and priapism and testicular pain after administration of testosterone enanthate to boys [46]; however, these adverse events were not observed in this study. This difference may be related to the type of priming agent, as we most often used ethinyl estradiol. Furthermore, we found that the priming procedure was uncomplicated, as the medication used for priming was easily accessible and manageable.

In this study, the cutoff value used to define GHD was set at 7 µg/L for both spontaneous nocturnal and stimulated GH peaks, consistent with Swedish national guidelines [47]. While the evidence for a correct cutoff level for GHD diagnosis in children is weak, this is especially true for spontaneous nocturnal peak values. However, efforts to disentangle the optimal cutoff level suggest that 7.3 µg/L may be a suitable cutoff for spontaneous nocturnal GH peaks [48]. This finding aligns with our prior research, where a stimulated peak cutoff of 7 µg/L corresponded best with a spontaneous nocturnal cutoff of 7.8 µg/L [7].

The strengths of this study are the standardized protocol used (nocturnal test and AITT) throughout the entire study period, the fact that the stimulation test always followed immediately after the spontaneous test, meaning that the divergent results were not due to differences in patient characteristics or laboratory methods, and the fact that the study was performed in a clinical setting, increasing the generalizability of the results. Furthermore, throughout the study period, only two pediatric endocrinologists were predominantly involved in patient care, likely resulting in a consistent clinical examination and interpretation of the GH test results. The limitations of this study include those intrinsic to retrospective, observational studies: the risk of missing data in medical records and the potential of missing measurements of known or unknown confounding factors. Furthermore, the subgroups were relatively small, which might have hindered the identification of some statistically significant results, and we did not follow the patients to their near adult height. Lastly, we lack data on sleep quality during the spontaneous GH test, the assay methods for measuring GH concentrations changed during the study period, and three different priming agents were used.

In conclusion, our data show that priming with sex steroids significantly decreases the occurrence of divergent results between spontaneous and stimulated GH secretion in children with short stature. Priming is associated with significantly higher peak and mean spontaneous nocturnal GH values, is feasible in the clinical setting, and is well tolerated by children. We suggest that priming should be considered when investigating children for potential GHD. Future research should validate our findings in a larger population and include long-term follow-up to assess outcomes through adult height.

This study protocol was reviewed and approved by the Regional Ethics Board at Uppsala, Sweden, approval No. 2017/358. An amendment application was approved by the Swedish Ethical Review Authority, registration No. 2023-01820-02). Written informed consent was not obtained or required from participants or participants’ legal guardian/next of kin in this study in accordance with national legislation and institutional requirements.

The authors have no conflicts of interest to declare.

The authors declare that this study received funding from the Futurum Academy for Health and Care, Jönköping County Council, Region Jönköping County, the Swedish Research Council (project 2021-01807), the Swedish Society of Medicine, Promobilia, Sällsynta Fonden, Nyckelfonden Research Foundation, Sällskapet Barnavård, Stiftelsen Frimurare Barnhuset i Stockholm, the Stockholm County Council, Karolinska Institutet, Stockholm, Sweden, and Örebro University, Örebro, Sweden. No funder had any role in the design of this study, data collection, data analysis, decision to publish, or preparation of the manuscript.

O.L. retrieved all the data from the medical records and merged the new data with preexisting data, generated the figures and tables, performed the statistical analyses, and wrote the first draft of and revised the manuscript. O.N. and M.L. were responsible for the study design and interpretation of the analysis results, supervised the statistical analysis, and participated in the writing of the manuscript. All authors approved the submitted version.

The raw data used in the analyses performed in this study to support our conclusions are not publicly available due to their containing information that could compromise the privacy of research participants but are available from O.L. upon reasonable request.