While essential for linear growth during childhood, GH promotes lipolysis, especially in visceral adipose tissue and protein synthesis and decreases peripheral insulin sensitivity and glucose uptake in adults [10]. Adult GH deficiency is a clinical syndrome that mainly results from pituitary tumors or the treatment of such tumors, namely, with surgery or radiation. Regardless of etiology, GH deficiency is generally associated with several metabolic changes, including increased visceral adipose tissue, decreased lean body mass, and dyslipidemia, with higher total and low-density lipoprotein (LDL) cholesterol, triglycerides, and hypertension [11,12]. Although GH has antagonistic effects on insulin action, its deficiency leads to insulin resistance and glucose intolerance, probably due to specific changes in fat distribution, such as increased visceral adipose tissue and ectopic fat accumulation [10,13]. These frequently lead to metabolic syndrome (MS) in patients with untreated GH deficiency [13]. Given the intricate association between MS and NAFLD, a focus has been given to the role of GH in the pathogenesis of NAFLD in the last decade (Figure 1).

Several cross-sectional studies reported an increased prevalence of liver dysfunction and NAFLD in patients with hypopituitarism, particularly those with GH deficiency [14,15,16]. Moreover, patients with GH deficiency and NAFLD have an accelerated progression of the hepatic disease [15]. Additionally, a study by Koehler et al. showed that obese patients with NASH and advanced fibrosis have low serum GH levels and that normal GH levels essentially excluded advanced fibrosis [17,18]. Patients with Laron syndrome, which is characterized by GH resistance due to inactivating mutations of the GH receptor, were also found to have a higher incidence of NAFLD [19]. Interestingly, acromegalic patients treated with the GHR antagonist pegvisomant showed increases in hepatic triglyceride content [20]. Despite the aforementioned results, evidence on this matter remains conflicting, with some studies not finding differences in the prevalence of NAFLD or in the intrahepatic lipid content assessed by magnetic resonance spectroscopy between GH deficient patients and healthy adults [21]. Heterogeneity in sample sizes and the clinical characteristics of included patients, such as sex and ethnicity, may help to explain these differences in the studies’ results.

Animal and cell culture studies further support a role for the GH/ insulin-like growth factor-1 (IGF-1) axis in the pathophysiology of NAFLD and its progression to NASH and fibrosis [11]. Animal models with liver-specific mutations in the GH receptor or downstream signaling pathways (JAK2/STAT5) develop metabolic syndrome, hepatic steatosis, steatohepatitis, and fibrosis [22,23,24]. Moreover, in animal models of adult-onset hepatic GH resistance, steatosis and NASH evolve rapidly after the loss of hepatic GH signaling, regardless of other signs of metabolic dysfunction [25].

The restoration of GH levels in adults with GH deficiency reduces body fat, increases lean mass, and ameliorates the lipid profile, decreasing total and LDL cholesterol and increasing high-density lipoprotein (HDL) cholesterol [26,27]. A few small studies have shown that GH replacement improves hepatic injury, as observed by a rapid decrease in serum liver transaminases and gamma-glutamyl transferase levels, steatosis, lobular inflammation, hepatocyte ballooning, and the severity of fibrosis [15,28]. Additionally, GH supplementation in pediatric patients with GH deficiency has been associated with an improvement in NAFLD, as it decreases visceral fat accumulation and lipid deposition on the liver and enhances mitochondrial function [29,30,31]. Animal studies have reinforced these results [24]. It is well known that the actions of GH are mediated both directly and indirectly through the stimulation of IGF-1 production [10]., and the majority of circulating IGF-1 (>90%) is produced by hepatocytes in response to growth hormone receptor stimulation [32]. Increasing evidence suggests that both GH and IGF-1 have direct and indirect effects on hepatic structure and function [33]. Moreover, decreased GH and, consequently, IGF-1 might be responsible for the muscle mass changes, particularly sarcopenia, which is seen in NAFLD [34]. GH was recently proposed to directly inhibit de novo lipogenesis and the expression of peroxisome proliferator-activated receptor-gamma (PPAR-γ) and CD36, key regulators of free fatty acid uptake [23,25]. IGF-1 is an anti-inflammatory molecule that contributes to mitochondrial function and reduces oxidative stress in the liver [24]. IGF-1 prevents cholesterol accumulation through the stimulation of the expression of ATP-binding cassette transporter A1 (ABCA1), a pivotal regulator of lipid efflux from cells to apolipoproteins [35]. Furthermore, IGF-1 limits the activity of hepatic stellate cells and induces their senescence, therefore attenuating hepatic fibrosis [36]. Lower levels of IGF-1 might result in lower protection against liver inflammation and fibrosis [37]. By impairing the adipose tissue phenotype, GH deficiency increases the expression of proinflammatory cytokines and adipokines [38], which compromises insulin sensitivity and impairs the ability of adipose tissue to store fat, increasing lipid influx into ectopic organs, such as the liver [22]. Chronic liver diseases, including NASH, are associated with a reduction of GH receptor (GHR) expression and, therefore, reduced IGF-1 levels [39]. As described above, lower levels of IGF-1 impair liver homeostasis, with a higher risk of fibrosis, leading to a vicious cycle of both worsening hepatic homeostasis and increasing growth hormone function.

Given the crucial role of GH in hepatic lipid metabolism, there are some clinical trials examining the impact of low-dose GH supplementation in patients with hepatic steatosis and NASH without known hypothalamic/pituitary disease. A new clinical trial had its results recently published, showing that treatment with recombinant human GH may have the potential to reduce liver fat content in adolescents with NAFLD and obesity [40]. Other clinical trials studying the impact of GH supplementation on NAFLD are underway, such as the clinical trial named Growth Hormone and Intrahepatic Lipid Content in Patients With Nonalcoholic Fatty Liver Disease (NCT02217345). Lastly, IGF-1 replacement is also being considered as an option to treat patients with liver diseases [41]. Experimental studies show that treatment with IGF-1 is particularly beneficial in the reduction of liver fibrosis, although positive effects in hepatic steatosis and inflammation can also be seen [24,42].

Acromegaly is characterized by excessive GH and, consequently, IGF-1 and is the most frequent cause of GH-secreting pituitary adenoma. Increased levels of GH are associated with increased lipolysis and favorable body composition, with increased lean body mass and decreased visceral and subcutaneous adipose tissue [43]. However, acromegaly also promotes insulin resistance, with consequent hyperglycemia, hyperinsulinemia, hypertriglyceridemia, and an increased risk of overt diabetes [44]. These paradoxical effects may justify contradictory evidence on this topic. While some studies including patients with active acromegaly found that intrahepatic lipid, measured by magnetic resonance spectroscopy, is relatively low in comparison to healthy subjects [20,45], others showed that hepatic steatosis is a common comorbidity in acromegaly, hypothesizing that lipotoxicity and insulin resistance may outweigh the direct hepatic effects of GH [46].

Acromegaly treatment with surgery or medical therapy improves metabolic risk by increasing insulin sensitivity [45]. However, GH, IGF-1, insulin-like growth factor binding proteins (IGFBPs), and medical treatment have a complex relationship with insulin sensitivity and hepatic steatosis. GHR antagonists (GHRA) induce an improvement in acromegaly glycemic control through the decrease of glucose and the normalization of insulin secretion [47]. This effect enables one to understand the important effect of GH on hepatic and peripheral IGF-1 action. Hepatic GH-induced IGF-1 production is regulated by portal insulin levels, as insulin promotes the translocation of hepatic GHR to the surface. When portal insulin levels are high, the liver becomes GH sensitive, regardless of the cause of the increased production of insulin. In addition, portal insulin also inhibits hepatic IGFBP1 production, which may increase the bioavailability of circulating IGF-1. Insulin suppression by somatostatin analogs (SSA) also selectively results in hepatic GH resistance, which itself decreases hepatic IGF-1 production [48]. Therefore, the consequent reduction in circulating IGF-1 does not necessarily reflect GH activity in peripheral tissues. It, thus, makes sense that the normalization of serum IGF-1 levels during SSA does not necessarily imply the control of disease’s activity in peripheral tissues, which is a condition that Neggers coined as being “extra-hepatic acromegaly” [49]. This concept received support in a study that evaluated surgically and SSA-treated acromegalics. Despite the normalization of IGF-1, SSA-treated patients had less suppressed GH levels and less symptom relief [50]. On the other hand, GHRA do not block all tissues with equal effectiveness for the GH actions. Adipose tissue seems to require less GHRA to reduce GH actions when compared to the liver, where more GHRA are required to reduce IGF-1 production [51]. This could be a reason for local GHRA-induced lipomatosis. In further support of this hypothesis, it was recently reported that short-term GHRA administration in healthy subjects can suppress lipolysis without affecting either circulating or local IGF-1 [52]. Accordingly, it is possible that peripheral suppression of GH activity is obtained prior to the normalization of hepatic IGF-1 production. Therefore, GHRA-treated patients with acromegaly and normal peripheral IGF-1 can have peripheral GH deficiency [49]. If all this is true, then patients with acromegaly and diabetes should need higher GHRA doses to normalize IGF-1 compared to patients without diabetes. Recently, this was demonstrated by Droste et al. [53].

In a cross-sectional study including patients previously treated for acromegaly, hepatic steatosis, measured by magnetic resonance spectroscopy, was found to be increased compared to healthy controls, even several years after successful treatment [54]. In a recent prospective study, Ciresi et al. found no differences in the prevalence of steatosis after 12 months of treatment with somatostatin analogs, measured by abdominal ultrasonography [55]. The heterogeneity in the chosen imaging method to assess hepatic steatosis and in the treatments for acromegaly may explain these differences.