Effects of 18 Months of Growth Hormone Replacement Therapy on Bone Mineral Density in Patients with Adult Growth Hormone Deficiency: A Retrospective Study

Objective The effect of physiological dose growth hormone (GH) replacement therapy on bone mineral density (BMD) in adults with growth hormone deficiency (GHD) is not well defined. We aimed to investigate the effects of 18 months of treatment with recombinant human growth hormone (rhGH) at physiological doses on BMD, body composition (BC), and quality of life (QoL). Methods Sixty-eight patients diagnosed with adult growth hormone deficiency (AGHD) in our hospital were included in this retrospective study. All patients received individualized rhGH replacement to maintain normal serum insulin-like growth factor-1 (IGF-1) levels. BMD and BC measurements were performed by dual energy X-ray absorptiometry (DXA). Excluding those with incomplete follow-up data, we analyzed BMD in 68 patients, as well as BC and QoL in 36 of them. Results Compared with the baseline, lumbar spine BMD decreased by 0.008 g/cm2 (P=0.006) and increased by 0.011 g/cm2 (P=0.045) at month 18, and total hip BMD decreased by 0.005 g/cm2 (P=0.008) and did not change significantly from the baseline at month 18. The changes in BMD did not differ by sex, and the increase in BMD was more pronounced in patients with low Z-scores at the baseline (lumbar spine: P=0.005 and total hip: P=0.018). The percentage change from the baseline in BMD was greater for the lumbar spine than for the total hip (P=0.003). Lean body mass (LBM) increased significantly (P=0.012), total body fat ratio (TBF%) decreased significantly (P=0.011), visceral adipose tissue (VAT) decreased significantly (P=0.016), and QoL improved significantly (P < 0.001). Conclusions Within 18 months of treatment, bone resorption manifested first, BMD decreased to a nadir at month 6, and then it increased. The increase in BMD was greater in the lumbar spine than in the hip, and the increase was more pronounced in patients with low BMD. Eighteen months of rhGH replacement therapy significantly improved lumbar spine BMD and improved BC and QoL.


Introduction
Adult growth hormone defciency (AGHD) is a multifactorial group of endocrine disorders characterized by reduced bone mineral density (BMD) and lean body mass (LBM), increased visceral fat, insulin resistance, dyslipidemia, reduced muscle strength, and reduced quality of life (QoL) [1].
Conclusive evidence indicates that GH defciency in adulthood results in decreased BMD, decreased bone turnover, and higher fracture risk rates [2]. Growth hormone (GH) exerts its efects mainly indirectly through insulin-like growth factor-1 (IGF-1), while GH itself can also act directly on diferent cells, leading to the growth of bones and diferent organs. GH and IGF-1 have important roles in maintaining bone metabolic homeostasis in humans [3]. Te GH/IGF-1 axis promotes longitudinal bone growth, skeletal maturation, and bone mass acquisition during adolescence and acts to maintain bone mass in adults. Te GH/IGF-1 axis exerts its anabolic efects on bone through multiple mechanisms, and it can stimulate the proliferation of bone marrow mesenchymal stem cells and promote their diferentiation into osteoblasts. Te GH/IGF-1 axis also maintains bone mass by maintaining a balance between osteoblasts and osteoclasts [4][5][6].
Several studies have shown that BMD is positively affected by rhGH replacement therapy in patients with AGHD. Several long-term studies [7][8][9][10][11], the longest of which, up to 15 years, have shown that rhGH replacement signifcantly improved BMD in patients with AGHD, with a peak bone mass in the seventh year followed by a downward trend [7]. However, in some short-term studies within 24 months, these fndings have been inconsistent or even contradictory due to diferences in the dose of GH, duration of treatment, study population, study methods, and equipment used to detect BMD [12][13][14][15][16][17][18]. Several fndings suggest that during the frst 12 months of rhGH replacement therapy, in which bone resorption predominates, no improvement in BMD is observed and even a downward trend [17,19,20]. Te response of bone to GH is not the same in various parts of the body, and the changes observed in the long bones of the extremities, the lumbar spine, and the hip joints are inconsistent [16,19]. Te efect of GH on bone mass is diferent between patients with adult-onset GH defciency (AO-GHD) and those with childhood-onset GH defciency (CO-GHD) [21] because patients with CO-GHD may not have reached peak bone mass, and some may have used rhGH before, in some studies, the two groups were not distinguished. All of these factors have infuenced the outcome of response to GH replacement. Tese studies were all conducted in Caucasians, and there were no relevant studies in Asians, so these conclusions do not serve as an absolute reference for the efects of rhGH treatment on BMD in Asian patients with AGHD.
Multiple studies have shown that replacement therapy with rhGH in patients with AGHD also produces some improvement in body composition and can reduce body fat mass (BFM), increase lean body mass [22,23], and enhance patient quality of life [18,24].
We conducted a retrospective study aimed at exploring the efects of short-term rhGH replacement on bone mineral density, body composition, and quality of life.

Study Population.
We selected 68 patients (22 males and 46 females; mean age 42.90 ± 11.52 years) with AGHD who visited in the Endocrinology Department of the First Affliated Hospital of Chongqing Medical University from September 2018 to September 2022. AGHD diagnosis was based on the American College of Endocrinology recommendation of an insulin tolerance test (ITT) as the gold standard [25] and a peak growth hormone level <5.0 mg/L. Tey were all AO-GHD patients and had not received prior GH therapy. All were assessed for thyroid, gonadal, and adrenal function, and all patients with defciencies of other hormones except for GH were on stable hormone replacement therapy for more than 6 months. Te exclusion criteria were as follows: (1) CO-GHD; (2) known or suspected secondary osteoporosis; (3) antiosteoporotic drugs were used or have been clearly established to have an efect on BMD; (4) BMD was not reviewed at regular follow-up visits, and data were incomplete; (5) systolic blood pres-sure>140 mmHg or diastolic blood pressure >90 mmHg under treatment; (6) diabetes or poor fasting glucose control; (7) combined severe liver and kidney disease; (8) severe heart disease; and (9) present, past, or family history of malignancy. Finally, we selected 68 patients, all of whom had been on GH therapy for 18 months and had BMD and other indexes assessed at the baseline before GH treatment and at 6, 12, and 18 months, respectively. Only 36 subjects had body composition data available at the baseline and 18 months, so only 36 subjects were included in the comparison statistics of body composition and physical data (7 males and 29 females; mean age: 38.75 ± 8.75 years).
An individualized rhGH dosage regimen was used in all patients to maintain IGF-1 at normal levels. Patients were subcutaneously injected daily with rhGH (Jintrolong ® , China) by using an automatic pen device. Te initial dose was 0.5 IU/day, and in the frst month after treatment initiation, doses were adjusted based on subject changes in efcacy and safety indicators and IGF-1 levels. Patients were followed up every 3 months with questionnaires, physical measurements, and blood biochemical measurements. Bone mineral density was measured every 6 months. Te rhGH dose was adjusted according to the actual situation with each increase or decrease of 0.17 mg/day (0.5 IU/day), and the highest dose did not exceed 0.67 mg/day (2 IU/day).
All protocols were approved by the ethics committee of our hospital and conformed to the Declaration of Helsinki, and all subjects gave written informed consent.

General Clinical Information
Collection and Anthropometric Measurements. All patients completed a uniformly formulated questionnaire, and clinical history taking and physical examinations were performed on all subjects by the same investigator. Te anthropometric measurements of height, weight, waist circumference (WC), hip circumference (HC), grip strength, and blood pressure were measured as described in our previous study [22]. Te body mass index (BMI) was calculated by weight (kg)/height 2 (m 2 ), and waisthip ratio (WHR) was calculated by waist circumference (cm)/hip circumference (cm).

Blood Biochemical Measurements.
All subjects had a light diet the previous night and fasted for 8-12 hours, and then blood was collected from the elbow vein the next morning and centrifuged to obtain fresh serum. Fasting blood glucose (FPG), fasting insulin (FINS), liver and kidney function, and blood lipid profles were measured. GH, IGF-1, and thyroid function were also measured. FPG was measured using the glucose oxidase method (Biosen5030 rapid glucose detector, Necar, Germany). Lipid profles were measured by a biochemical autoanalyzer (Olympus AU5400, Japan). FINS, GH, IGF-1, and thyroid function were measured by chemiluminescence (Chemi immuno luminescence kit, Roche).

Bone Mineral Density and Body
Composition. BMD and BC measurements were performed by a dual energy X-ray absorptiometry (DXA) scanner (Hologic Discovery QDR Series, USA). Te subjects removed metals and jewelry, lying fat on the scanner, which was manipulated by the same professionally trained staf. Quality control assays for DXA were performed daily with a coefcient of variation (CV) <1.0%. BMD of the lumbar spine L2-L4 and total hip were measured, and T-scores and Z-scores were automatically generated using standard procedures and DXA reference database software as previously described [22]. To eliminate the infuence of age on BMD, the reduction in BMD of the subjects was assessed with Z-scores. In this study, subjects with Z < −1 were classifed into the low bone mass group, and subjects with Z ≥ −1 were classifed into the high bone mass group. Body fat mass, lean body mass, total body fat ratio a (TBF%), and visceral adipose tissue (VAT) were also estimated by DXA. Total body lean/fat ratio was calculated by LBM (g)/BFM (g).

Quality of Life. Quality of Life-Assessment of Growth
Hormone Defciency in Adults (QoL-AGHDA) was used to assess the quality of life, a questionnaire specifcally designed to assess the QoL of AGHD patients, an indicator used in daily monitoring or clinical trials of patients, and each language version has been shown to have good reliability and internal consistency [26,27]. Tis questionnaire consists of a total of 25 individual questions, which are scored as 0 if the patient indicates the absence of this problem, 1 if the patient indicates the presence of a problem, and a high total score indicates poor quality of life.
2.6. Statistics. Statistical analysis was performed using SPSS 25.0 software. We performed the Kolmogorov-Smirnov test or the Shapiro-Wilk test a priori on the data of all subjects to determine whether they belonged to a normal distribution. Data are expressed as the mean ± standard deviation (SD) if they conformed to a normal distribution and as M (P25 and P75) if they did not conform. Te comparisons of means between the groups that conformed to a normal distribution and homogeneity of variance were performed using the independent samples t-test, and those that were not consistent were performed by the Mann-Whitney U test. Diferences in data at each time point of rhGH treatment were compared using the paired sample t-test or Wilcoxon test. Diferences in change between BMD of the hip and lumbar spine were compared using the Wilcoxon test. Te comparisons of diferences in BMD change between the groups for sex and Z-scores were performed using the Mann-Whitney U test. Te correlation analysis was performed using Pearson, Spearman's correlation analysis. P values <0.05 were considered signifcant.

Baseline Characteristics.
According to the inclusion criteria, a total of 68 subjects (22 males, 46 females; mean age 43.22 ± 11.32 years) were included in the study. Te baseline characteristics of all patients are shown in Table 1. Te males in this study had signifcantly greater WHR than females (P < 0.001), as well as the male subjects possessing lower lumbar spine BMD T-scores (P < 0.05) and Z-scores (P < 0.01). Te peak GH after ITT correlated signifcantly with IGF-1 values (r � 0.531; P < 0.001). We categorized the baseline BMD and Z-score of all subjects into peak GH ≥ 3 and peak GH < 3 groups according to the GH peak, and we did not observe diferences in BMD or Z-scores between the two groups. We also did not observe correlations between peak GH with baseline BMD or Z-scores.

Serum IGF-1 Concentration.
Serum IGF-1 signifcantly increased after 18 months of rhGH replacement therapy compared to the baseline (P < 0.001). Serum IGF-1 was signifcantly elevated from the baseline at each follow-up visit, with the increase being most pronounced within the frst month of treatment ( Figure 1).

BMD.
BMD changes in the lumbar spine and total hip are shown in Table 2 and Figure 2. Compared with the baseline, BMD of the lumbar spine increased signifcantly after 18 months of rhGH therapy by 1.4% in percentage and by 0.011 g/cm 2 in actual value (P � 0.045). BMD decreased signifcantly at month 6, with a percentage decrease of 0.89% and an actual value of 0.008 g/cm 2 (P � 0.006). BMD was lower than the baseline at month 12, but the diference was not statistically signifcant (P � 0.412).
Compared with the baseline, BMD of the total hip did not change signifcantly after 18 months (P � 0.701) and decreased signifcantly at 6 months, with a percentage decrease of 0.62% and an actual value decrease of 0.005 g/cm 2 (P � 0.008). BMD began to increase after 6 months, by 0.46% at month 18 relative to that at month 6, with an actual increase of 0.0041 g/cm 2 ; however, it was not signifcant (P � 0.201), and the BMD of the hips remained lower than that at the baseline after 18 months of rhGH therapy, although the diference was not signifcant (Table 2).
We performed post hoc stratifcation of BMD data, grouping BMD of the lumbar spine and total hip separately by sex ( Figure 3) and Z-scores ( Figure 4). We found that the percentage changes from the baseline in BMD at the lumbar spine and total hip did not difer by sex (Figure 3), but subjects with Z < −1 had a greater increase, and the changes were signifcantly greater than those with Z ≥ −1 (lumbar spine: P � 0.005 and total hip: P � 0.018) (Figure 4). We also found that the percentage change from the baseline in BMD at the lumbar spine was greater after 18 months of rhGH therapy than that at the total hip (P � 0.003) ( Figure 5).
We found no correlation between serum IGF-1 levels and BMD or Z-scores at any time point (data not shown in the article).

Body Composition, Physical Measurements, and QoL.
As previously mentioned, only 36 subjects were included in the comparison statistics of body composition and physical data (Table 3). Compared with the baseline, no signifcant changes were observed in height, weight, BMI, WC, HC, and WHR at month 18, and BMI means before and after treatment were within the normal range (BMI < 24 kg/m 2 ).        BFM was reduced but not signifcantly (P � 0.094), LBM was signifcantly increased (P � 0.012), total body lean/fat ratio was signifcantly increased (P � 0.003), TBF% was signifcantly decreased (P � 0.011), and VAT was signifcantly decreased (P � 0.016). It illustrates the increase in the proportion of lean mass, the decrease in the proportion of fat mass, and the unchanged body weight. Grip strength increased signifcantly (P � 0.003). Diastolic blood pressure decreased signifcantly (P � 0.036), and blood pressure was within normal limits before and after treatment. Qol-AGHDA scores decreased signifcantly (P < 0.001).

Discussion
In this retrospective study, we investigated the efects of rhGH replacement therapy at physiological doses on BMD in patients with AGHD, and we adjusted the dosage of rhGH by using the levels of serum IGF-1 as a reference index to ensure that the IGF-1 levels of the patients were below the upper limit of the normal range for the same age. Some previous studies may have employed low doses of rhGH replacement for observation, which may interfere with the results because of insufcient elevation of serum IGF-1 levels [18], and some studies used higher doses, resulting in a high incidence of side efects in patients [28]. Our individualized rhGH dose replacement is more clinically valuable in the treatment of patients with AGHD. Some of the previous studies included patients with CO-GHD, but their response to rhGH is not consistent with AO-GHD patients, thereby causing interference with the study results, and the two populations cannot be combined in such studies [20]. In order to exclude the confounding efect of CO-GHD on the study results, only patients with AO-GHD were included in this study.
Studies have shown that BMD in adults with GHD is reduced [2], there is no diference in BMD reduction between isolated GHD patients and multiple pituitary hormone defciency patients [29], and there is no signifcant diference in BMD changes between the two populations after receiving rhGH replacement. Tis shows that the decrease in bone mass in patients with AGHD is mainly due to the lack of GH but is not related to other pituitary hormones. In our study, the mean T-score of BMD of all subjects at the baseline was within the range of osteopenia, indicating that the BMD of AGHD patients was lower than that of normal healthy people. Terefore, theoretically speaking, the supplementation of physiological dose of GH in AGHD patients has a positive impact on BMD, and in some long-term research reports, the replacement of rhGH has signifcantly improved the BMD of AGHD patients [7][8][9][10][11]24].
Tis is an 18-month short-term study. We observed that, compared with the baseline, the BMD of both the lumbar spine and total hip decreased signifcantly at the 6th month of rhGH replacement therapy, and BMD did not increase or even decreased during the frst 12 months of treatment. Te frst 6 months showed the most signifcant decline, and no signifcant change in BMD was observed between the 6th month and the 12th month. At the 18th month, the BMD of the lumbar spine was signifcantly improved compared with the 6th month and was signifcantly higher than the baseline at the 18th month. Te BMD of the total hip also increased in the 18th month compared with the 6th month, although it was not signifcant, and after 18 months of treatment, it was still lower than the baseline, but it was not signifcant. Tis is consistent with some previous research results. Within 12 months of starting rhGH replacement therapy, BMD cannot be observed to increase or even signifcantly decrease, and only after 18-24 months can BMD increase be observed. Te study of Sartorio [17] showed that BMD during rhGH replacement therapy declined in 1-12 months, with the largest decline in the frst 6 months, and bone absorption reached its peak in the 3rd month; Välimäki et al. [30] and Holmes et al. [20] showed that BMD decreased to the lowest point in the sixth month of rhGH replacement therapy and then increased, but no net increase in BMD was observed in 12 months; however, several studies have shown that BMD has a consistent rise in the process of rhGH replacement therapy even in the frst 12 months [12,13,19]. Data are expressed as the mean ± SD or M (P25 and P75). * P < 0.05; * * P < 0.01; * * * P < 0.001.
International Journal of Endocrinology Te serum IGF-1 level of the subjects increased significantly compared with the baseline, indicating that the initial decrease in BMD was not caused by an insufcient increase in IGF-1 levels. Te most likely explanation is that GH stimulates bone remodeling [6,31], a complete bone remodeling cycle is approximately 4 months, and bone resorption occurs before new bone is formed. Terefore, in the initial stage of rhGH replacement therapy, the level of bone resorption is greater than that of bone formation, which is manifested by a decrease in bone density, and it takes 6-12 months for newly formed bone to be completely mineralized. Several studies have shown that rhGH increased the markers of bone turnover between 12 months, reached the peak of bone turnover between the 6th and 10th months, and then decreased, but BMD still showed an upward trend [14,15,31]. Tis provides a good explanation for our research results; rhGH replacement therapy stimulates bone remodeling and increases bone turnover. In the frst 6-12 months, it is mainly manifested as bone resorption, and then BMD begins to increase. Te deposition and mineralization of new bone takes a certain time, so it lags behind the changes in bone turnover markers. It is also possible that rhGH increases the bone area, which will also afect the results of BMD changes [12,13,16,19]. In general, GH has a positive efect on BMD in patients with AGHD. Although results seen in short-term studies are relatively limited, we can speculate that long-term rhGH replacement therapy should lead to a net increase in BMD, which has also been demonstrated in some long-term studies.
Although the changes we observed in BMD in the lumbar spine and total hip were similar, there were also diferences. Te change in BMD of the hip was not as signifcant as that of the lumbar spine, and BMD of the total hip did not change signifcantly or even decreased slightly from the baseline after 18 months, but BMD of the lumbar spine was signifcantly higher than the baseline after 18 months of treatment. Our data also show that the percentage change from the baseline in BMD of the lumbar spine was signifcantly higher than that of the hip after completing 18 months of rhGH replacement therapy, and the BMD of the lumbar spine appeared to be more responsive to GH than that of the hip. Tis is possibly because the lumbar spine contains more trabecular bone, whereas the femoral neck contains more cortical bone, and trabecular bone is more metabolically active and more susceptible to hormonal infuences than cortical bone [20,32].
We did not observe diferences in BMD changes by sex, but we found that patients in the low Z-scores' group at the baseline had a greater percentage increase in BMD at both the lumbar spine and total hip than those in the high Zscores' group. Te response to GH seems to be more pronounced in patients with low bone mass. Several reports have shown a more pronounced increase in BMD in patients with lower BMD on rhGH replacement therapy, but the diferences in BMD changes by gender are controversial. Similar to our results, an 18-month randomized, doubleblind, placebo-controlled study of GH replacement in Copenhagen [14] showed no diference in BMD changes between males and females after 18 months of treatment with physiological doses of rhGH replacement in patients with AGHD, while BMD increased more signifcantly in patients with low Z-scores. In a Finnish multicenter study of rhGH replacement therapy for 42 months [30], the increase in BMD was signifcantly greater in males than in females, and the percentage increase in BMD was signifcantly greater in patients with osteopenia at the baseline than in patients with normal bone mass. A 2-year prospective study from Sweden [13] showed that the percentage increase in total body BMD was greater in females than males at the end of rhGH replacement therapy, but no gender diferences were observed in BMD changes at other sites, while the percentage increase in BMD was greater in patients with Z < −1 at the baseline. A prospective study in the Netherlands [12] of rhGH treatment for two years, in which patients were divided into men, estrogen-replete women, and estrogendepleted women, showed that changes in Z-scores of the lumbar spine did not difer among the three groups, but the study measured only the lumbar spine.
Unfortunately, we did not observe a correlation between IGF-1 levels and BMD at any time point. It may be due to the following reasons: (1) changes in serum IGF-1 are more sensitive to changes in BMD, with a signifcant increase in IGF-1 during the frst month of rhGH treatment, but the changes in BMD are relatively slow and complex, and the formation of new bone takes 6-12 months, so changes in BMD are not concordant with changes in serum IGF-1 levels [3,6]; (2) the duration of the disease varied in each subject, as did the duration of GH defciency, IGF-1, which can decrease markedly within a short time after onset, but the decrease in BMD lagged relatively behind; (3) changes in BMD are associated with various factors, such as serum 25-(OH)VD₃ levels, dietary habits, sunlight exposure, and exercise, and failure to use sex hormone replacement therapy in postmenopausal women can also interfere with BMD changes, as can previous hyperpituitarism in some patients with pituitary tumors [33]; (4) although GH acts mainly indirectly through IGF-1, in addition to that, GH can also act directly on bone [23]; (5) the analysis using only the values of IGF-1 did not exclude the interference of age on IGF-1, and the IGF-1 SD score for age should be widely used in the future clinical or scientifc research. We also found no correlation between peak GH and BMD or Z-scores, and when we grouped BMD and Z-scores according to peak GH, no diference was found between the groups. Probably because of the variable course of the disease in each patient, the peak GH could only indicate the degree of GH defciency present in the patient; however, it was not clear how long the defcient period was, and it was not possible to predict the patients' BMD simply by the peak GH. However, the results of Annamaria, who contradicted our study, divided 101 adult patients with hypopituitarism into very severe GHD, severe GHD, partial GHD, and non-GHD groups according to the peak GH on a GH provocation test and found that in the lumbar spine, the T-scores for BMD correlated with the degree of GHD defciency and that both the peak GH and serum IGF-1 correlated signifcantly with the T-scores [29].
Our data also show that 18 months of rhGH replacement therapy improved body composition. Weight, WC, HC, 8 International Journal of Endocrinology WHR, and BMI did not change signifcantly, which is consistent with most long-term studies. However, the efect of rhGH replacement therapy on BMI is controversial, with reports showing an increase in BMI and reports showing no change in BMI [22,24]. A possible explanation is that the increase in BMI does not necessarily represent an increase in BFM, possibly an increase in LBM and total bone mineral content, but also an increase in body water retention. Patients had signifcant increases in LBM and total body lean/ fat ratio and signifcant decreases in both TBF% and VAT. However, adipose tissue is not well distinguished from water by DXA, and intracellular water content increases after rhGH replacement therapy. Te grip strength of patients was also signifcantly elevated, which seems to indirectly refect increased LBM. However, subjects may not have consistent posture when performing grip strength tests and may be afected by developing force in positions such as the shoulder or large arm, and we should standardize their posture in the future grip strength tests. Te signifcant decrease in QoL scores indicates that the patients' quality of life was improved. After patients had passed through rhGH replacement therapy, there was an increase in life well-being and exercise performance, which all helped elevate patients' LBM and BMD, as well as decrease BFM. We also observed a signifcant decrease in diastolic blood pressure possibly because IGF-1 stimulates the vascular endothelium to produce nitric oxide (NO) to dilate the blood vessels or by increasing the activity of Na/K-ATPase in vascular smooth muscle cells, and it is also possible that rhGH replacement reduced lipid levels [34].
Tere are several limitations to our study. First, this is a retrospective study, and we cannot have better control of the patients' test indicators: for example, the patients were not tested for serum 25-(OH)VD₃, and we are not clear about specifcs such as patients' lifestyle habits and exercise patterns, which are associated with BMD. In addition, our study also lacked a placebo or a control group not treated with rhGH because aging in humans over time counteracts a portion of the efects of rhGH, so the conclusions drawn by relying solely on the pre-and post-treatment controls in the rhGH-treated group are of relatively limited value. Additionally, we only studied the efect of rhGH on BMD in AGHD patients, but it is unclear how GH treatment afects BMD in non-AGHD patients. Prospective controlled studies with large samples are needed in the future to explore the efect of GH on BMD.

Conclusions
In conclusion, short-term treatment with rhGH replacement for 18 months resulted in benefcial efects on BMD, but these benefcial efects were shown only for the lumbar spine. Although BMD initially decreased, it tended to increase after the 6th month. Meanwhile, both body composition and QoL were improved, which may not only indirectly increase BMD by increasing exercise capacity but also reduce cardiovascular risk. Our study provides some theoretical support for whether rhGH can be used clinically to increase BMD in patients.

Data Availability
Te data that support the fndings of this study are not publicly available for ethical reasons but are available on request.

Conflicts of Interest
All authors declare that they have no conficts of interest.