Hypertriglyceridemia and younger age are associated with effectiveness of growth hormone therapy on hepatic steatosis

Abstract

Adult growth hormone deficiency (AGHD) is often accompanied with metabolic dysfunction-associated steatotic liver disease (MASLD). Although some studies reported that MASLD is ameliorated by growth hormone replacement therapy (GHRT), the characteristics of AGHD that are associated with an improvement of hepatic steatosis by GHRT remain unknown. We aimed to investigate whether GHRT affects hepatic lipid accumulation as well as biochemical parameters, and investigated the association between these parameters (UMIN000044989). Thirty people with AGHD were recruited, and assigned to either the GHRT group or the non-GHRT group. Serum laboratory data were analyzed before and after GHRT. Hepatic lipid content was evaluated using magnetic resonance imaging-proton density fat fraction (MRI-PDFF). Correlations between MRI-PDFF and other clinical parameters were investigated. Twenty-nine people completed this study (19 in the GHRT group and 10 in the non-GHRT group). In the GHRT group, significant decreases in MRI-PDFF and serum levels of aspartate aminotransferase, alanine aminotransferase, and γ-glutamyl transpeptidase were observed after the treatment. The decrease in MRI-PDFF levels after GHRT significantly correlated with initial MRI-PDFF, triglyceride (TG), lactate dehydrogenase, and ALT levels, and age. Multiple regression analysis demonstrated that younger age and high serum TG levels were independent predictors of a decrease in MRI-PDFF levels. GHRT in people with AGHD significantly reduced lipid accumulation in the liver on MRI, and improved serum liver parameters. Age and serum TG levels were found to be associated with the effectiveness of GHRT.

Introduction

Adult growth hormone deficiency (AGHD) is a rare endocrine disorder caused by hypothalamic-pituitary lesions. People with AGHD often have complications of metabolic dysfunction-associated steatotic liver disease (MASLD), which is characterized by excessive accumulation of lipids in the liver [1].

Although it has been reported that MASLD caused by AGHD is ameliorated by growth hormone replacement therapy (GHRT) [2, 3], other studies have shown no effects of GHRT on MASLD [4, 5], suggesting that GHRT may improve certain types of hepatic steatosis accompanied with AGHD.

Magnetic resonance spectroscopy (MRS) has been used to evaluate lipid accumulation in the liver, as a less invasive method than liver biopsy. Several clinical studies have demonstrated the effects of GHRT on hepatic lipid content using MRS [4, 5]. Whereas MRS requires specific software and stringent conditions to ensure that the sequence acquisitions are reliable, magnetic resonance imaging (MRI) enables simple quantification of hepatic lipid content and repeated imaging in daily clinical practice.

In this study, to thoroughly investigate whether GHRT affects hepatic steatosis in people with AGHD, hepatic lipid content was evaluated using MRI-proton density fat fraction (MRI-PDFF) before and after GHRT.

Materials and Methods

Study design and people

An open-label nonrandomized study was performed on thirty people with AGHD who were recruited from the Department of Diabetes, Endocrinology and Metabolism of Kitasato University Hospital between 2016 and 2022. Patients were diagnosed as having AGHD using an insulin tolerance test or a growth hormone (GH)-releasing peptide-2 test [6]. Inclusion criteria were an age of 20 to 79 years, naive to GH or not treated with GH for 180 days or more before the recruitment, and being diagnosed as having AGHD according to the guidelines of the Japan Endocrine Society. Exclusion criteria were an alcohol consumption of more than 30 g per day in men and more than 20 g per day in women, known hepatic disease or new diagnosis of hepatic disease, known diabetes mellitus, known malignant tumors, pregnant or likely to get pregnant, mental incapacity, unwillingness or language barriers precluding adequate understanding or cooperation. Subjects with viral hepatitis were identified by their clinical records and serological markers of hepatitis viruses, and excluded from the study. The possibility of drug-induced adverse events was carefully reviewed at each visit.

Thirty people with AGHD were recruited, and they received an explanation regarding the possible benefits of GHRT from an endocrinologist. As a result, 20 subjects agreed to undergo GHRT, and 10 GHD patients declined to undergo the therapy, owing to financial issues or injection burden.

The dose of recombinant GH was titrated to achieve an insulin-like growth factor 1 (IGF-1) level within the normal range. The GHRT group was reassessed after more than 6 months of GHRT, and the non-GHRT group was also reassessed after more than 6-months of follow-up at our hospital.

The study protocol was approved by the Kitasato University Medical Ethics Organization (study approval no.: KMEO B21-050), and the trial registration number is UMIN000044989. People data were obtained using an opt-out methodology from a retrospective subgroup, and written informed consent was obtained from participants in the prospective subgroups. All methods were performed in accordance with the relevant guidelines and regulations of Kitasato University Hospital as well as the Ethical Guidelines for Medical and Health Research Involving Human Subjects in Japan.

Clinical and Laboratory tests

Body composition parameters, including body weight, muscle mass, lean body mass and fat mass were measured with the eight-electrode multifrequency bioelectrical impedance technology (body composition analyzer MC-180; Tanita, Tokyo, Japan). Lipid metabolic parameters included triglycerides, high-density lipoprotein (HDL) cholesterol, and low-density lipoprotein (LDL) cholesterol. Liver functions were assessed by measuring aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase (ALP), gamma-glutamyl transpeptidase (γ-GTP) and lactate dehydrogenase (LDH). The indirect non-invasive fibrosis scores of Fibrosis-4 (FIB-4) index were calculated using the following formula: FIB-4 index = age (years) × AST (U/L)/platelet count (10⁹/L) × √ALT (U/L) [7]. Glucose metabolism parameter included hemoglobin A1c (HbA1c). Blood tests were performed with the participants in a fasting state.

Assessment of hepatic steatosis using MRI-PDFF

Hepatic lipid content was assessed using MRI-PDFF, in which lipid content is accurately measured by MRI analysis of the iterative decomposition of water and lipid with echo asymmetry and least-squares estimation (IDEAL IQ, GE Healthcare, IL, USA) [8]. IDEAL-IQ images were acquired during a single breath hold using a Discovery MR750w Expert 3.0 Tesla. MASLD was evaluated based on MRI-PDFF [9]. MRI and blood tests were performed either on the same day or within one month of each other.

Statistical analysis

Statistical analysis was performed using JMP Pro17 (SAS Institute, Inc., Cary, NC, USA) and Prism 10.05 (GraphPad Software Inc., San Diego, CA, USA) software. Data are presented as the mean and standard deviation (SD), unless otherwise indicated. Two independent samples were assessed using the paired t-test. In addition, the correlation between the change in MRI-PDFF values and changes in clinical parameters or clinical parameters at the beginning of the GHRT were analyzed using a simple linear regression model. Stepwise multiple linear regression analysis was performed to identify the independent predictors of the reduction in MRI-PDFF. A p-value of less than 0.05 (two-tailed) was considered to indicate a statistically significant difference between groups.

Results

Characteristics of the subjects

Thirty people with AGHD were recruited for this study; 20 subjects were assigned to the GHRT group, and 10 people were followed up as the non-GHRT group. A subject assigned to the GHRT group was excluded from the study owing to drug-induced liver injury, which was presumably not associated with GHRT (Fig. 1). The clinical characteristics of the 29 patients are shown in Table 1. Pituitary hormone deficiencies ranged from single axis to panhypopituitarism. Benign tumors or cystic diseases were the most common etiology of AGHD among the participants. Twenty-three of the people underwent surgery. The baseline characteristics of the participants are shown in Table 2. The median time from the diagnosis of pituitary disorders to participation in the study was 1.58 years (range: 0.25–42 years). The duration of GHRT in this study was 10.8 months (range: 6–40 months). None of the participants reported an apparent family history of familial hypercholesterolemia. Some participants were taking medications for dyslipidemia (Table 1), and these medications were not changed during the study period.

Fig. 1  CONSORT diagram for subjects recruitment, allocation and follow-up

GHRT, Growth hormone replacement therapy

Table 1 Clinical characteristics of the subjects

Patient no.	Age (years)	Sex	Diagnosis	Treatment	rGH (mg/day)	Hydrocortisone (mg/day)	Estrogen replacement	No. of hormone deficiencies in addition to GHD	Medication for dyslipidemia
1	40	Male	Craniopharingioma	S	0	15	None	3	None
2	33	Female	NFPitNET	S	0	0	Oral	2	None
3	66	Female	NFPitNET	S	0	15	None	2	Statins
4	72	Male	NFPitNET	S	0	15	None	3	None
5	77	Male	NFPitNET	S	0	15	None	2	Statins
6	78	Male	NFPitNET	S	0	15	None	2	Statins
7	46	Female	Rathke’s cleft cyst	S	0	15	Oral	4	None
8	26	Male	Germinoma	R	0	15	None	4	None
9	48	Female	Meningioma	S	0	15	None	2	Statins
10	76	Female	Lymphocytic hypophysitis	—	0	15	None	2	Statins
11	22	Male	Craniopharingioma	S	0.45	15	None	4	Statins
12	41	Female	Craniopharingioma	S	0.2	15	Transdermal	4	None
13	43	Female	Craniopharingioma	S	0.35	15	Transdermal	4	None
14	46	Female	Craniopharingioma	S	0.15	15	None	2	None
15	50	Male	Craniopharingioma	S	0.15	15	None	4	Fibrates, ezetimibe
16	60	Male	Craniopharingioma	S	0.15	15	None	2	Statins, ezetimibe
17	44	Male	NFPitNET	S	0.2	15	None	2	Statins
18	59	Male	NFPitNET	S	0.1	15	None	2	None
19	47	Female	Rathke’s cleft cyst	S	0.1	15	None	1	None
20	70	Male	Rathke’s cleft cyst	S	0.1	15	None	1	None
21	71	Female	Rathke’s cleft cyst	S	0.225	0	None	2	Statins, omega-3 fatty acid ethyl esters
22	40	Female	Empty sella	—	0.3	15	Transdermal	3	None
23	50	Female	Empty sella	—	0.2	15	Transdermal	2	None
24	47	Male	Acromegaly	S	0.1	0	None	0	None
25	55	Female	Acromegaly	S	0.25	15	None	2	Statins
26	57	Female	Germ cell tumor	S, R	0.15	15	None	3	None
27	21	Female	Idiopathy	—	0.4	15	Transdermal	3	None
28	42	Male	Prolactinoma	S	0.2	15	None	3	None
29	48	Male	PSIS	—	0.3	15	None	2	Statins

NFPitNET, non-functioning pituitary neuroendocrine tumor; PSIS, pituitary stalk interruption syndrome; rGH, recombinant growth hormone; S, surgery; R, radiation; GHD, growth hormone deficiency

Table 2 Clinical characteristics of the subjects

	Overall population n = 29	non-GHRT n = 10	GHRT n = 19
Age (years)	51 ± 16	56 ± 19	48 ± 13
Male sex	14 (48%)	5 (50%)	9 (47%)
Height (cm)	165 ± 8	165 ± 7	164 ± 9
Body weight (kg)	66 ± 14	62 ± 12	69 ± 15
BMI (kg/m2)	24.3 ± 4.2	22.4 ± 2.8	25.3 ± 4.5
Body fat (%)	29.0 ± 8.8	26.2 ± 7.0	30.4 ± 9.4
Fat mass (kg)	20.2 ± 9.5	16.6 ± 5.1	21.8 ± 10.7
Muscle mass (kg)	44.6 ± 8.9	44.3 ± 9.1	44.7 ± 9.0
Lean body mass (kg)	47.2 ± 9.3	46.8 ± 9.5	47.4 ± 9.4
Time from diagnosis (years)	1.58 (0.25–42)	4 (0.25–25)	1.5 (0.20–42)
Adrenal insufficiency (n)	25 (86%)	8 (80%)	17 (89%)
Hypogonadism (n)	22 (76%)	9 (90%)	13 (68%)
Hypothyroidism (n)	23 (79%)	8 (80%)	15 (79%)
Central DI (n)	15 (52%)	5 (50%)	10 (53%)
No. of hormone deficiencies in addition to GHD	2.45 ± 1.0	2.60 ± 0.8	2.37 ± 1.1
Serum IGF-1 (ng/dL)	61.0 ± 28.8	56.1 ± 38.4	63.6 ± 23.1
Serum IGF-1 SD score	–3.2 ± 1.4	–3.1 ± 1.6	–3.2 ± 1.3
AST (U/L)	31 ± 16	24 ± 6	36 ± 19
ALT (U/L)	36 ± 31	18 ± 6	46 ± 35
γ-GTP (U/L)	37 ± 29	18 ± 5.8	46 ± 31
ALP (U/L)	63 ± 27	51 ± 26	70 ± 26
LDH (U/L)	189 ± 63	183 ± 41	192 ± 72
HbA1c (%)	5.7 ± 0.4	5.6 ± 0.4	5.7 ± 0.4
MRI-PDFF (%)	10.9 ± 12.0	4.0 ± 3.5	14.6 ± 13.3
MASLD (n)	16 (55%)	2 (20%)	14 (74%)
FIB-4 index	1.23 ± 0.64	1.53 ± 0.82	1.07 ± 0.47

Data are presented as the mean ± standard deviation (SD), except for time from diagnosis, which is the median (min-max).

BMI, Body mass index; DI, diabetes insipidus; IGF-1, insulin like growth factor-1; AST, aspartate aminotransferase; ALT, alanine aminotransferase; γ-GTP, gamma-glutamyl transpeptidase; LDH, lactate dehydrogenase; HbA1c, hemoglobin A1c; MRI-PDFF, MRI-derived proton density fat fraction; FIB-4, Fibrosis-4; GHRT, growth hormone replacement therapy

Effects of GHRT on clinical laboratory data and body composition

Analysis of serum hepatic enzyme levels demonstrated that in the GHRT group, AST, ALT, and γ-GTP levels significantly decreased from 36 ± 19 U/L, 46 ± 35 U/L, and 46 ± 31 U/L, respectively, at baseline, to 25 ± 10 U/L (p = 0.0010), 28 ± 15 U/L (p = 0.010), and 32 ± 24 U/L (p = 0.011), respectively, at the last visit. (Table 3). On the other hand, there were no significant differences in changes in ALP, LDH, HbA1c, and LDL-cholesterol levels, and the FIB-4 index in both groups.

Table 3 Baseline and final clinical parameters of the subjects

	non-GHRT		p-valuea	GHRT		p-valuea
	At baseline	At the last visit		At baseline	At the last visit
Weight (kg)	61.7 ± 11.6	61.1 ± 8.9	0.20	69.0 ± 15.4	68.8 ± 15.2	0.84
BMI (kg/m2)	22.4 ± 2.8	22.1 ± 2.2	0.17	25.3 ± 4.5	25.3 ± 4.4	0.83
Body fat (%)	26.2 ± 7.0	25.9 ± 6.2	0.79	30.4 ± 9.4	28.9 ± 9.9	0.10
Fat mass (kg)	16.6 ± 5.1	15.9 ± 4.0	0.49	21.8 ± 10.7	21.0 ± 11.1	0.16
Muscle mass (kg)	44.3 ± 9.1	44.6 ± 8.5	0.46	44.7 ± 9.0	45.2 ± 8.5	0.10
Lean body mass (kg)	46.8 ± 9.5	45.9 ± 8.9	0.37	47.4 ± 9.4	48.2 ± 8.7	0.17
Serum IGF-1 (ng/dL)	56.1 ± 38.4	59.6 ± 35.8	0.59	63.6 ± 23.1	155 ± 87.3	<0.001
Serum IGF-1 SD score	–3.1 ± 1.6	–2.9 ± 1.7	0.55	–3.2 ± 1.3	–0.27 ± 1.2	<0.001
AST (U/L)	24 ± 6.4	23 ± 3.6	0.82	36 ± 19	25 ± 10	0.0010
ALT (U/L)	18 ± 6.4	19 ± 7.6	0.64	46 ± 35	28 ± 15	0.0010
γ-GTP (U/L)	18 ± 5.8	17 ± 7.3	0.50	46 ± 31	32 ± 24	0.011
ALP (U/L)	51 ± 26	49 ± 20	0.62	70 ± 26	72 ± 21	0.77
LDH (U/L)	183 ± 41	180 ± 36	0.93	192 ± 72	172 ± 39	0.15
Triglycerides (mg/dL)	103 ± 47	112 ± 46	0.52	164 ± 100	155 ± 61	0.51
HDL-cholesterol (mg/dL)	61 ± 14	66 ± 16	0.13	59 ± 26	56 ± 16	0.24
LDL-cholesterol (mg/dL)	103 ± 33	98 ± 14	0.58	126 ± 39	116 ± 27	0.66
HbA1c (%)	5.6 ± 0.4	5.7 ± 0.5	0.065	5.7 ± 0.4	5.8 ± 0.6	0.17
MRI-PDFF (%)	4.0 ± 3.5	4.4 ± 4.2	0.33	14.6 ± 13.3	11.1 ± 10.5	0.041
FIB-4 index	1.53 ± 0.82	1.64 ± 0.91	0.20	1.07 ± 0.47	1.05 ± 0.50	0.80

Data are presented as the mean ± standard deviation (SD).

^aStatistical testing was performed using the paired t-test between at baseline and at the last visit.

BMI, Body mass index; IGF-1, insulin like growth factor-1; AST, aspartate aminotransferase; ALT, alanine aminotransferase; γ-GTP, gamma-glutamyl transpeptidase; LDH, lactate dehydrogenase; HbA1c, hemoglobin A1c; MRI-PDFF, magnetic resonance imaging-derived proton density fat fraction; FIB-4, Fibrosis-4; GHRT, growth hormone replacement therapy

Effect of GH replacement therapy on hepatic lipid content

Hepatic lipid content in the GHRT group, as measured by MRI-PDFF, was significantly reduced at the last visit compared with at baseline (from 14.6% ± 13.3% to 11.1% ± 10.5%; p = 0.041), but not in the non-GHRT group (from 4.0 ± 3.5 to 4.4 ± 4.2; p = 0.33) (Table 3).

When the correlation between the MRI-PDFF reduction and clinical parameters in the GHRT group were analyzed, the reduction in MRI-PDFF was found to be significantly correlated with changes in body weight (r = –0.58, p = 0.0097), BMI (r = –0.55, p = 0.015), AST (r = –0.57, p = 0.012), ALT (r = –0.73, p = 0.00040), and triglycerides (TGs) (r = –0.73, p = 0.00040) (Table 4, Fig. 2). In addition, the reduction in MRI-PDFF significantly correlated with the baseline data of LDH (r = 0.47, p = 0.042), ALT (r = 0.52, p = 0.024), TGs (r = 0.70, p = 0.00080), MRI-PDFF (r = 0.62, p = 0.0043), and age (r = –0.56, p = 0.012) (Table 4, Fig. 3). Notably, there was no significant correlation between MRI-PDFF reduction and changes in IGF-1 (r = 0.44, p = 0.056) or IGF SD score (r = 0.24, p = 0.33) in the GHRT group (Table 4). To investigate the correlation of MRI-PDFF reduction with other clinical parameters, a multiple linear regression analysis was performed for the significant parameters from the univariate analysis, which demonstrated that age (β = 0.22, p = 0.016), and TG levels (β = –0.041, p = 0.0012) at baseline were independent predictors of MRI-PDFF reduction; young age and high serum TG levels correlated with a significant decrease in MRI-PDFF upon GHRT (Table 5). Notably, initial MRI-PDFF values correlated with initial serum TG levels (r = 0.7582, p = 0.0002; Fig. 4).

Table 4 Correlations between reduction in proton density fat fraction and clinical parameters

	Clinical parameter at baseline		Change in clinical parameter
	r	p-value	r	p-value
Age (years)	–0.56	0.012	—	—
Weight (kg)	0.18	0.45	0.58	0.0097
BMI (kg/m2)	0.23	0.33	0.55	0.015
Body fat (%)	0.25	0.34	0.067	0.80
Fat mass (kg)	0.16	0.53	0.32	0.21
Muscle mass (kg)	0.14	0.59	0.35	0.19
Lean body mass (kg)	0.14	0.58	0.30	0.24
Serum IGF-1 (ng/dL)	0.25	0.30	–0.44	0.056
Serum IGF-1 SD score	–0.16	0.51	–0.24	0.33
AST (U/L)	0.30	0.22	0.57	0.012
ALT (U/L)	0.52	0.024	0.73	0.00040
γ-GTP (U/L)	–0.027	0.91	0.42	0.076
ALP (U/L)	–0.14	0.60	0.056	0.84
LDH (U/L)	0.47	0.042	0.30	0.21
Triglycerides (mg/dL)	0.70	0.00080	0.69	0.0016
HDL-cholesterol (mg/dL)	–0.33	0.17	–0.053	0.84
LDL-cholesterol (mg/dL)	0.14	0.57	–0.14	0.58
HbA1c (%)	–0.35	0.17	0.46	0.070
MRI-PDFF (%)	0.62	0.0043	—	—
FIB-4 index	–0.24	0.33	0.050	0.84

BMI, Body mass index; IGF-1, insulin like growth factor-1; AST, aspartate aminotransferase; ALT, alanine aminotransferase; γ-GTP, gamma-glutamyl transpeptidase; LDH, lactate dehydrogenase; HbA1c, hemoglobin A1c; MRI-PDFF, magnetic resonance imaging-derived proton density fat fraction; FIB-4, Fibrosis-4

Fig. 2  Correlations between MRI-PDFF reduction and changes in body weight, body mass index, and liver enzyme and TG levels

A scatterplot demonstrating the correlation of MRI-PDFF reduction (–ΔMRI-PDFF) with changes in (A) body weight (–ΔBW), (B) body mass index (–ΔBMI), (C) AST (–ΔAST), (D) ALT (–ΔALT), and (E) TG (–ΔTG) after GHRT. MRI-PDFF, magnetic resonance imaging-proton density fat fraction; BW, body weight; BMI, body mass index; AST, aspartate aminotransferase; ALT, alanine aminotransferase; TG, triglyceride

Fig. 3  Correlations between MRI-PDFF reduction and baseline levels of MRI-PDFF, age, TGs, liver enzymes, and LDH

Scatterplots demonstrating the correlation of MRI-PDFF reduction (–ΔMRI-PDFF) with (A) MRI-PDFF at baseline, (B) age, (C) serum TG level at baseline, (D) ALT at baseline, and LDH at baseline. MRI-PDFF, magnetic resonance imaging-proton density fat fraction; TG, Triglyceride; ALT, alanine aminotransferase; LDH, lactate dehydrogenase

Table 5 Stepwise multiple linear regression analysis (Model R²: 0.61)

Variable	β-value	t-value	SE	95% CI	p-value
Age (years)	0.22	2.71	0.081	0.047/0.39	0.016
TG (mg/dL)	–0.041	–3.91	0.010	–0.063/–0.019	0.0012

SE, standard error; CI, confidence interval

Fig. 4  Correlation between initial MRI-PDFF values and initial TG levels

Scatterplot demonstrating the correlation between initial MRI-PDFF values and initial serum TG levels. MRI-PDFF, magnetic resonance imaging-proton density fat fraction; TG, triglyceride

Discussion

In the present study, we demonstrated using MRI that hepatic lipid content is reduced after GHRT in people with AGHD. GHRT has been shown to prevent or reverse the progression of MASLD in AGHD people [2, 3, 10], which is consistent with the present data. In contrast, some other studies found no significant improvement in the hepatic steatosis of AGHD people after GHRT [4, 5]. As the present study suggested that higher initial MRI-PDFF levels resulted in an improvement of lipid contents in the liver (Fig. 3), the severity of hepatic lipid accumulation might affect the effectiveness of GHRT on hepatic steatosis. To our knowledge, this is the first study to demonstrate that initial serum TG levels can predict the effect of GHRT on hepatic lipid content, as demonstrated by the reliable quantification of intrahepatic fat content using MRI.

It has been reported that GH directly suppresses liver steatosis through the regulation of both lipid uptake and de novo lipogenesis in the liver, independently of IGF-1 [11]. In contrast, IGF-1 has been demonstrated to improve liver fibrosis in vivo by inducing the senescence of activated stellate cells [12-14]. Thus, GHRT is likely to reduce lipid contents directly through IGF-1 or through IGF-1-mediated signaling.

Regarding liver-associated parameters, serum AST and ALT levels in addition to MRI-PDFF were significantly improved by GHRT (Table 3). In our study, three subjects had increased liver enzyme levels after GHRT, and all of them experienced body weight gain, suggesting that weight control may affect the effectiveness of GHRT. This appears to be consistent with a previous study reporting that a higher proportion of patients showing an increase in liver enzyme levels after GHRT experienced weight gain than those who did not [15]. Our study also showed a possible association between body weight changes and MRI-PDFF, indicating that weight gain may exacerbate hepatic steatosis during GHRT (Table 4).

In our study, neither initial IGF-1 levels nor changes in serum IGF-1 levels during the observation period were correlated with the reduction in MRI-PDFF. As a previous study reported that IGF-1 mRNA levels in the liver, but not serum IGF-1 levels, were correlated with the parameters of liver steatosis [16, 17], serum IGF-1 levels may not be a sensitive marker of the severity of liver steatosis.

Whereas histological analysis of liver biopsies can directly evaluate hepatic steatosis lesions, a histological method for the quantitative comparison of hepatic lipids between subjects has not been established to date. On the other hand, quantification of MRI-PDFF levels can easily be standardized between individuals, so that we were able to repeatedly analyze 29 people with AGHD under the same conditions. In addition, this method is less invasive than histological analyses of liver biopsies. A long-term study would provide highly informative insights into the effects of GHRT on hepatic steatosis.

Multiple regression analysis demonstrated that a high serum TG level at baseline and young age predicts a decrease in MRI-PDFF levels (Fig. 3, Table 5, Graphical Abstract). Hepatic lipid contents have been shown to closely correlate with serum TG levels and very-low-density lipoprotein-TG levels [18, 19], which is consistent with our data that initial serum TG levels correlated with initial MRI-PDFF levels (Fig. 4). This positive correlation between the two parameters may result in more pronounced effects of GHRT on hepatic steatosis in patients with high TG levels at baseline. GH was shown to be directly responsible for the regulation of both lipid uptake and de novo lipogenesis in the liver [20]. Furthermore, a liver from an elderly individual was shown to have compromised cellular maintenance, which may be associated with GHRT resistance [21]. Taking these previous findings into account, GHRT may have an additional effect on hepatic steatosis in people with AGHD, particularly in patients with a younger age and/or hypertriglyceridemia.

Graphical Abstract

There are several limitations to this study. First, this study was launched as a nonrandomized trial. The control subjects were those who declined to undergo GHRT, owing to financial issues or injection burden, although all subjects received an explanation regarding the possible benefits of GHRT from an endocrinologist. As a result of this nonrandomization bias, MRI-PDFF values at baseline in the GHRT group were higher than those in the non-GHRT group, which may have resulted in the different prognoses of liver steatosis. Second, the small sample size and the short follow-up duration make it difficult to generalize the conclusions. The positive association between reduction in hepatic lipid content and basal TG levels might have been influenced by the limited sample size. In addition, lipid accumulation in the liver at baseline may lead to a substantial improvement in hepatic steatosis. To address these limitations, a study with a larger patient cohort and a longer follow-up period measuring MRI-PDFF values is necessary in the future. Third, as most participants demonstrated multiple hormone deficiencies in addition to AGHD in this study, it is possible that these additional hormone deficiencies may have influenced the effects of GHRT on hepatic lipid content. A previous study reported that hydrocortisone (HC) replacement doses of 15 to 25 mg did not increase the risk of nonalcoholic fatty liver disease in people with primary or secondary adrenal insufficiency [22]. As all participants with adrenal insufficiency in this study received either 15 mg or no HC, the risk of NAFLD caused by HC replacement therapy was likely minimal. On the other hand, as gonadal hormones have been implicated in the regulation of hepatic steatosis [23, 24], impaired gonadal function may affect the development of hepatic steatosis. Thus, these confounding factors should be taken into account for evaluating the effects of GHRT.

In summary, we found that GHRT for people with AGHD is likely to improve hepatic steatosis, and its effectiveness correlates with hypertriglyceridemia and young age. Predicting the effectiveness of GHRT by serum TG levels and age is a simple and noninvasive approach for optimizing therapeutic strategies for hepatic steatosis. A future study with a longer observation period and a larger number of people with AGHD than the present study is expected to clarify the clinical significance of GHRT on hepatic steatosis, which may provide a better quality of life for these patients.

Author Contribution Statement

T.T., K.T., and T.M. designed the whole project and wrote the manuscript. All authors revised the manuscript critically for important intellectual content, and approved the final version of the manuscript. T.M. is the guarantor of this work.

Data Availability

The datasets analyzed in this study are available from the corresponding author upon reasonable request.

Acknowledgements

We thank Kazuko Sagara and Satomi Takebe for their secretarial assistance with this study.

Disclosure

This study was supported by a research grant from Novo Nordisk (T.T. and T.M.). Novo Nordisk did not influence the content of the publication, data collection, or analysis of this study. The other authors have no competing interests to disclose. T.M. is a member of Endocrine Journal’s Editorial Board.

References

• 1   Melmed  S (2019) Pathogenesis and diagnosis of growth hormone deficiency in adults. N Engl J Med 380: 2551–2562.
• 2   Takahashi  Y,  Iida  K,  Takahashi  K,  Yoshioka  S,  Fukuoka  H, et al. (2007) Growth hormone reverses nonalcoholic steatohepatitis in a patient with adult growth hormone deficiency. Gastroenterology 132: 938–943.
• 3   Nishizawa  H,  Iguchi  G,  Murawaki  A,  Fukuoka  H,  Hayashi  Y, et al. (2012) Nonalcoholic fatty liver disease in adult hypopituitary patients with GH deficiency and the impact of GH replacement therapy. Eur J Endocrinol 167: 67–74.
• 4   Gardner  CJ,  Irwin  AJ,  Daousi  C,  McFarlane  IA,  Joseph  F, et al. (2012) Hepatic steatosis, GH deficiency and the effects of GH replacement: a Liverpool magnetic resonance spectroscopy study. Eur J Endocrinol 166: 993–1002.
• 5   Meienberg  F,  Yee  M,  Johnston  D,  Cox  J,  Robinson  S, et al. (2016) Liver fat in adults with GH deficiency: comparison to matched controls and the effect of GH replacement. Clin Endocrinol (Oxf) 85: 76–84.
• 6   Chihara  K,  Shimatsu  A,  Hizuka  N,  Tanaka  T,  Seino  Y, et al. (2007) A simple diagnostic test using GH-releasing peptide-2 in adult GH deficiency. Eur J Endocrinol 157: 19–27.
• 7   Sterling  RK,  Lissen  E,  Clumeck  N,  Sola  R,  Correa  MC, et al. (2006) Development of a simple noninvasive index to predict significant fibrosis in patients with HIV/HCV coinfection. Hepatology 43: 1317–1325.
• 8   Hope  TA,  Ohliger  MA,  Qayyum  A (2014) MR imaging of diffuse liver disease: from technique to diagnosis. Radiol Clin North Am 52: 709–724.
• 9   Imajo  K,  Kessoku  T,  Honda  Y,  Tomeno  W,  Ogawa  Y, et al. (2016) Magnetic resonance imaging more accurately classifies steatosis and fibrosis in patients with nonalcoholic fatty liver disease than transient elastography. Gastroenterology 150: 626–637.e627.
• 10   Takahashi  Y (2017) The role of growth hormone and insulin-like growth factor-I in the liver. Int J Mol Sci 18: 1447.
• 11   Doycheva  I,  Erickson  D,  Watt  KD (2022) Growth hormone deficiency and NAFLD: An overlooked and underrecognized link. Hepatol Commun 6: 2227–2237.
• 12   Sanz  S,  Pucilowska  JB,  Liu  S,  Rodríguez-Ortigosa  CM,  Lund  PK, et al. (2005) Expression of insulin-like growth factor I by activated hepatic stellate cells reduces fibrogenesis and enhances regeneration after liver injury. Gut 54: 134–141.
• 13   Nishizawa  H,  Iguchi  G,  Fukuoka  H,  Takahashi  M,  Suda  K, et al. (2016) IGF-I induces senescence of hepatic stellate cells and limits fibrosis in a p53-dependent manner. Sci Rep 6: 34605.
• 14   Luo  X,  Jiang  X,  Li  J,  Bai  Y,  Li  Z, et al. (2019) Insulin-like growth factor-1 attenuates oxidative stress-induced hepatocyte premature senescence in liver fibrogenesis via regulating nuclear p53-progerin interaction. Cell Death Dis 10: 451.
• 15   Matsumoto  R,  Fukuoka  H,  Iguchi  G,  Nishizawa  H,  Bando  H, et al. (2014) Long-term effects of growth hormone replacement therapy on liver function in adult patients with growth hormone deficiency. Growth Horm IGF Res 24: 174–179.
• 16   Stanley  TL,  Fourman  LT,  Zheng  I,  McClure  CM,  Feldpausch  MN, et al. (2021) Relationship of IGF-1 and IGF-binding proteins to disease severity and glycemia in nonalcoholic fatty liver disease. J Clin Endocrinol Metab 106: e520–e533.
• 17   Hribal  ML,  Procopio  T,  Petta  S,  Sciacqua  A,  Grimaudo  S, et al. (2013) Insulin-like growth factor-I, inflammatory proteins, and fibrosis in subjects with nonalcoholic fatty liver disease. J Clin Endocrinol Metab 98: E304–E308.
• 18   Westerbacka  J,  Cornér  A,  Tiikkainen  M,  Tamminen  M,  Vehkavaara  S, et al. (2004) Women and men have similar amounts of liver and intra-abdominal fat, despite more subcutaneous fat in women: implications for sex differences in markers of cardiovascular risk. Diabetologia 47: 1360–1369.
• 19   Fabbrini  E,  Mohammed  BS,  Magkos  F,  Korenblat  KM,  Patterson  BW, et al. (2008) Alterations in adipose tissue and hepatic lipid kinetics in obese men and women with nonalcoholic fatty liver disease. Gastroenterology 134: 424–431.
• 20   Liu  Z,  Cordoba-Chacon  J,  Kineman  RD,  Cronstein  BN,  Muzumdar  R, et al. (2016) Growth hormone control of hepatic lipid metabolism. Diabetes 65: 3598–3609.
• 21   Zhang  C,  Cuervo  AM (2008) Restoration of chaperone-mediated autophagy in aging liver improves cellular maintenance and hepatic function. Nat Med 14: 959–965.
• 22   Meyer  G,  Gruendl  M,  Chifu  I,  Hahner  S,  Werner  J, et al. (2023) Glucocorticoid replacement for adrenal insufficiency and the development of non-alcoholic fatty liver disease. J Clin Med 12: 6392.
• 23   Apostolov  R,  Gianatti  E,  Wong  D,  Kutaiba  N,  Gow  P, et al. (2022) Testosterone therapy reduces hepatic steatosis in men with type 2 diabetes and low serum testosterone concentrations. World J Hepatol 14: 754–765.
• 24   McKenzie  J,  Fisher  BM,  Jaap  AJ,  Stanley  A,  Paterson  K, et al. (2006) Effects of HRT on liver enzyme levels in women with type 2 diabetes: a randomized placebo-controlled trial. Clin Endocrinol (Oxf) 65: 40–44.