Clinical characteristics of non-alcoholic fatty liver disease in Chinese adult hypopituitary patients

Xian-Xian Yuan, Hui-Juan Zhu, Hui Pan, Shi Chen, Ze-Yu Liu, Yue Li, Lin-Jie Wang, Lin Lu, Hong-Bo Yang,Feng-Ying Gong

Abstrac t BACKGROUND Patients with hypothalamic-pituitary disease have the feature of central obesity,insulin resistance, and dyslipidemia, and there is increased prevalence of liver dysfunction consistent with non-alcoholic fatty liver disease (NAFLD) in this population. The causes of hypopituitarism in the reported studies varied and combined pituitary hormone deficiency including central diabetes insipidus is much common in this population. This retrospective cross-sectional study was performed to analyze the clinical characteristics and related factors with NAFLD and cirrhosis in Chinese adult hypopituitary/panhypopituitary patients.AIM To analyze the clinical characteristics of and related risk factors for NAFLD in Chinese adult hypopituitary patients.METHODS Adult Chinese patients with hypopituitarism and/or panhypopituitarism were enrolled at the Pituitary Center of Peking Union Medical College Hospital because the analysis used anonymous clinical data that w ere obtained from the Electronic Medical Record of Peking Union Medical College Hospital.

Key words: Hypopituitarism; Non-alcoholic fatty liver disease; Cirrhosis; Diabetes insipidus; Plasma osmolality

INTRODUCTION

Primary non-alcoholic fatty liver disease (NAFLD) is caused by an excess of fat in the liver (steatosis) that is not the result of excessive alcohol consumption or other second ary causes[1,2]. The d isease can be presented in the form of hepatic steatosis,inflammatory non-alcoholic steatohep atitis (NASH), fibrosis and/or cirrhosis[1,2].Ap p roximately 2%-3% of NAFLD p atients d evelop NASH, w hich carries a 20%potential risk of evolving further and causing cirrhosis[1,3]. Furthermore, NAFLD patients are five times more likely to develop hepatocellular carcinoma in the absence of cirrhosis compared w ith hepatitis C patients[4,5]. How ever, the pathogenesis of NAFLD and its progression are a complex process, and the “multiple-hit” hypothesis proposed by Buzzetti et al[6]in 2016 suggested that simple steatosis and NASH not only exhibited d ifferent risks of progression but might also reflect different d isease entities in terms of p athogenesis. Multip le insults, includ ing insulin resistance,obesity, and gut microbiota, contribute to the d evelop ment of steatosis and liver inflammation[6].

Patients w ith hypothalamic-pituitary disease present w ith central obesity, insulin resistance, and dyslipid emia, w hereas liver d ysfunction and NAFLD are common symptoms in this subpopulation[7-10]. Hong et al[8]reported that the frequency of fatty liver on abd ominal ultrasonograp hy in male subjects w ith hyp op ituitarism w as significantly higher compared with that of control subjects (70.6% vs 32.5%). Grow th hormone (GH) d eficiency (GHD) has been consid ered an important contributing factor to these metabolic changes in hypopituitary patients[11-13]. How ever, the causes of hypopituitarism in the reported studies varied, and combined pituitary hormone d eficiency, includ ing central d iabetes insip id us, w as more common in this subpop ulation. To the best of our know led ge, the imbalance of central d iabetes insipidus and concomitant serum electrolyte concentration has not been investigated with regard to the pathophysiology of NAFLD in patients w ith hypopituitarism. The aim of this retrosp ective cross-sectional stud y w as to comp are Chinese ad ult hyp op ituitary p atients w ith NAFLD and those w ithout NAFLD in terms of their clinical manifestations, end ocrine function, metabolic p arameters, and serum electrolyte p arameters and to analyze the associated factors w ith NAFLD and cirrhosis.

MATERIALS AND METHODS

The present study is a retrospective cross-sectional observational study. The study protocol was approved by the Ethics Committee of the Peking Union Medical College Hospital (PUMCH; No. JS1233).

Subjects

A total of 36 male and 14 female adult Chinese patients w ith hypopituitarism and/or panhypopituitarism w ere consecutively enrolled at the Pituitary Center of PUMCH between August 2012 and April 2018. According to abdominal ultrasonography, these patients w ere d ivided into an NAFLD (-) group and an NAFLD (+) group, and the latter w as further d ivided into an NAFLD group and a cirrhotic group according to the histological and/or clinical diagnostic criteria of cirrhosis.

Hypopituitarism/p anhypopituitarism w as d iagnosed clinically based on clinical manifestations, basal evaluation of pituitary function, and provocative tests. Hypopituitarism/panhypopituitarism was confirmed as follows: GHD was diagnosed as a peak serum GH level low er than 3 μg/L in both insulin tolerance tests (ITTs) and Larginine tests[14]. Hypogonadotropic hypogonad ism (HH) was defined by a reduction in the levels of luteinizing hormone (LH) and follicle stimulating hormone (FSH) for premenopausal w omen with a low serum estradiol concentration, and for adult men with a morning serum total testosterone concentration low er than 10.4 nmol/L[15]. HH was diagnosed via gonadotropin releasing hormone analogue (triptorelin) stimulation tests as follow s: LH60minw as low er than 4 IU/L (indisp ensable tests for delayed or absent puberty)[16]. Secondary hypocortisolism w as diagnosed based on a peak serum cortisol resp onse low er than 18 μg/d L (497.5 nmol/L) in ITTs[14]. Second ary hypothyroid ism w as diagnosed from a serum free thyroxine (FT4) level below the laboratory reference range in conjunction w ith a low, normal, or mild ly elevated thyroid stimulating hormone (TSH) level in the setting of pituitary disease[17]. Central d iabetes insip id us w as consid ered w hen serum osmolality w as higher than 295 mOsm/kg, w hile the corresponding urine osmolality w as lower than 300 mOsm/kg in fluid d eprivation tests, and a subsequent response to arginine vasopressin (AVP)was observed[18].

Diagnosis of NAFLD was confirmed by fatty infiltration of the liver on abdominal ultrasonography in association w ith or without abnormal liver enzymes. The patients were diagnosed as cirrhotic by the development of clinically evident complications of liver diseases (such as esophageal varices, ascites, jaund ice, hepatic encephalopathy,portal hypertension, hypersplenism, and hepatocellular carcinoma) or by liver biopsy.

The exclusion criteria w ere as follows: (1) patients w ith isolated pituitary hormone d eficiencies; (2) p atients w ith primary and second ary autoimmune hyp ophysitis treated w ith glucocorticoid s; (3) patients w ith hypop ituitarism second ary to functional p ituitary ad enomas, includ ing somatotrophin, lactotrop hin, thyrotrophin,corticotrophin, gonadotrophin, and plurihormonal and double adenomas; (4) patients w ith positive hepatitis B or C serology or w ith evid ence of inherited, autoimmune,cholestatic, d rug-ind uced, or metabolic liver d isease using stand ard clinical,laboratory, imaging, and histologic criteria; (5) patients with weekly alcohol intake of more than 21 units for men or 14 units for w omen; (6) patients with liver d ysfunction of a specific etiology other than hypopituitarism and NAFLD; and (7) children and adolescents.

Data extraction and management

The data were extracted from medical records, which included patient characteristics,such as age, gender, race, anthropometric parameters including height, weight, body mass index (BMI), waist circumference (WC), and blood pressure, diagnosis of the cause of hyp op ituitarism/p anhyp op ituitarism, laboratory results, abd ominal ultrasonography results, complications and comorbidities, treatments, and hormone rep lacements. The time of diagnosis of hypopituitarism/panhypopituitarism w as obtained from the date of symptoms and signs if ap plicable, or the time of clinical d iagnosis. The time of d iagnosis of NAFLD w as consid ered to be the d ate of abd ominal ultrasonograp hy that w as ind icative of p rimary NAFLD. Hormone replacement therapy refers to regular and dose titration rep lacement for at least 3 months. Biochemical and hormone levels w ere measured w ithin three days after admission. Homeostasis mod el assessment of insulin resistance (HOMA-IR) w as obtained according to the formula: (fasting plasma glucose × fasting serum insulin)/22.5.

Statistical analysis

The statistical package for the social science (SPSS, version 22.0; SPSS Inc., Chicago, IL,United States) was used for data and statistical analyses. Continuous data conforming to a normal distribution are expressed as the mean ± standard deviation, w hile continuous d ata of partial distribution are represented by the med ian values[interquartile range (IQR)]. Frequency data are expressed as the number (proportion)of patients with a condition. The Student's t-test and Wilcoxon Signed-Rank test were used for continuous variables, and the chi-square or Fisher's exact test was used for categorical variables. A P-value less than 0.05 was considered significant.

RESULTS

Clinical characteristics of studied subjects

A total of 36 male and 14 female p atients w ith hyp op ituitarism and/or panhypopituitarism were included in the present study. Their mean age was 26.9 years (range from 18.0 to 41.5 years). The frequency of fatty liver on abdominal ultrasonography was 54.0%. A total of two patients were diagnosed as cirrhotic by liver biopsy, while five patients were diagnosed as cirrhotic by the development of clinically evident complications of liver disease as indicated in the Materials and Methods section. Three cirrhotic patients were decompensated, one of whom had upper gastrointestinal bleeding, another two had hypersplenism, ascites, esophageal and gastric varices, and hepatopulmonary syndrome. The clinical characteristics of hypopituitary patients with NAFLD and without NAFLD are shown in Table 1, and a comparative analysis was completed. No significant differences were noted in the gender ratio among patients with cirrhosis, those with NAFLD, and those without NAFLD. Cirrhotic patients were younger than NAFLD patients (P = 0.009), with no significant difference in the course of hypopituitarism. The course of hypopituitarism of cirrhotic patients was significantly longer than that of patients without NAFLD(median course 13.0 years and 3.0 years, respectively, P = 0.023), although there was no significant difference in the age between the two groups.

Etiology of hypopituitarism

The underlying etiology of hypopituitarism is illustrated in Table 2. The profile of the causes of hypopituitarism in the cirrhotic group was similar to those of the NAFLD (-)and the NAFLD groups (P = 0.431 and 0.466, respectively). Intracranial germ cell tumors were the most common cause of hypopituitarism in patients without NAFLD(43.5%), while congenital hypopituitarism was the most common cause in NAFLD and cirrhotic patients (45.0% and 42.9%, respectively), follow ed by craniopharyngioma and other causes.

Table 1 Clinical characteristics of hypopituitary patients with or without non-alcoholic fatty liver disease

Distribution of anterior pituitary deficiencies and hormone replacement

The distribution of deficiency in anterior pituitary hormones is shown in Table 2, and all patients exhibited gonadotropin deficiency. The subjects who were on hormone replacement w ere evaluated clinically and biochemically at regular intervals,according to standard clinical practice, to ensure the physiological replacement of the deficient hormones. Although four (57.1%) cirrhotic patients were panhypopituitary,no significant differences were noted with regard to the distribution of deficiency in anterior pituitary hormones and to the p roportion of patients w ho received physiological hormone replacement. No patient received GH therapy in the present study. No significant difference was noted in serum GH or insulin-like growth factor 1 (IGF1). Serum thyroid hormone levels were similar in cirrhotic, NAFLD, and NAFLD (-) patients, as shown in Table 1. In addition, there was no difference in the proportion of patients w ith hypothalamic dysfunction in cirrhotic, NAFLD, and NAFLD (-) patients.

Table 2 Etiology of hypopituitarism and distribution of pituitary deficiencies in our patients n (%)

Diabetes insipidus and electrolytic imbalance

No patients were taking medications, such as diuretics, that could affect plasma osmolality and serum sodium for the treatment of cirrhotic conditions. A total of 36(72%) patients also had diabetes insipidus; however, 21 of them (58.3%) did not receive treatment with desmopressin. No significant differences were noted in the proportion of patients with diabetes insipidus or in the proportion of patients who received desmopressin treatment, between patients with and w ithout NAFLD. In addition, no significant differences were observed between the NAFLD and the cirrhotic groups, as shown in Table 2. However, plasma osmolality of patients with cirrhosis w as significantly higher than that of NAFLD patients (median plasma osmolality = 314.9 mOsm/kgH2O and 299.3 mOsm/kgH2O, respectively, P = 0.036).Similar differences w ere observed in the concentrations of serum sodium and chlorine, as show n in Table 1, but no significant differences were noted between patients with cirrhosis and those without NAFLD.

Association with metabolic syndrome

The comparison of the clinical characteristics demonstrated that no significant differences were noted in systolic blood pressure (SBP) or diastolic blood pressure(DBP) between patients with and without NAFLD. SBP and DBP of cirrhotic patients were lower than those of the NAFLD patients, although no statistically significant differences were observed (P adjusted for age = 0.116 and 0.050, respectively). The median BMI and WC in cirrhotic patients were 27.7 kg/m2and 110.5 cm, respectively.These values were significantly higher than those of patients without NAFLD (22.2 kg/m2and 73.0 cm, P = 0.011 and 0.040, respectively) but similar to those of the NAFLD patients (BMI = 27.9 kg/m2and WC = 102.0 cm, P = 0.912 and 0.396,respectively). The HOMA-IR of cirrhotic patients was significantly higher than that of patients without NAFLD (P = 0.045), although no significant difference was noted betw een the cirrhotic patients and NAFLD patients. A correlation analysis was performed, and it showed that fasting insulin concentration was positively associated with plasma osmolality in patients with NAFLD after adjusting for gender, age, and BMI (r = 0.540, P = 0.046), but no correlation was noted in total hypopituitary patients or in patients without NAFLD. There was no correlation between HOMA-IR and plasma osmolality. In contrast to previous studies[7-9], no significant difference was noted in the levels of fasting plasma glucose, glycated hemoglobin A1C(HbA1C), total cholesterol (TC), low density lipoprotein cholesterol (LDL-C), high density lipoprotein cholesterol (HDL-C), or triglycerides (TG), as shown in Table 1.

Abnormal liver function

Serum aminotransferase and bilirubin levels were considered elevated if they were above the normal laboratory reference range. No significant differences were noted among cirrhotic, NAFLD, and NAFLD (-) patients with regard to tserum aminotransferase and bilirubin levels, as show n in Table 3. The median levels of γ-glutamyltransferase (GGT) in cirrhotic patients w ere higher than those in patients w ithout NAFLD (52.0 U/L and 23.0 U/L, resp ectively, P = 0.046), although no significant difference was noted between NAFLD patients and cirrhotic patients (P = 0.862). The comp arison of the tests for hepatic synthetic function, includ ing serum albumin,prothrombin time (PT), and international normalized ratio (INR), d emonstrated that the average PT and INR in cirrhotic patients w ere significantly higher than those in the NAFLD p atients and NAFLD (-) patients (as show n in Table 3), although no significant d ifference w as noted betw een patients w ith and w ithout NAFLD. The cirrhotic p atients frequently exhibited a number of hematological abnormalities,including varying d egrees of cytop enia. Thrombocytopenia w as the most common type of hematological abnormality, w hile leukopenia and anemia developed later in the disease course. As show n in Table 3, w hite blood counts and blood platelet counts were low er in cirrhotic patients than in NAFLD and NAFLD (-) patients, although no significant differences w ere noted betw een patients w ith and without NAFLD.

Changes in biochemical indicators following hormone replacement in cirrhotic patients

The biochemical indicators of the seven cirrhotic patients were retested following regular hormone replacement, as shown in Figure 1. Liver function was generally stable, which was reflected by serum aminotransferase levels, white blood count, and blood p latelet count. TG w as significantly low er than that before hormone rep lacement (P = 0.043). The fasting serum insulin and HOMA-IR levels were d ecreased follow ing regular hormone rep lacement, although no significant differences were obtained (P = 0.593 and 0.593, respectively). Serum sodium w as decreased from 152.9 mmol/L to 143.6 mmol/L after hormone replacement (P =0.046); however, no significant difference was noted in plasma osmolality (319.6 mOsm/kgH2O and 300.1 mOsm/kgH2O, P = 0.080).

DISCUSSION

NAFLD is the most common cause of chronic liver dysfunction and a major global health problem over the past decades, whereas hypopituitarism/panhypopituitarism is a rare cause of NAFLD, w hich is easily misdiagnosed by endocrinologists and gastroenterologists. The data reported in the present study showed that the frequency of fatty liver in hypopituitary patients based on abdominal ultrasonography was 54.0%, which is significantly higher than that of the general adult population in China(approximately 15%), reported in 2013[19]. The incidence of cirrhosis in the present study population (patients with hypopituitarism and NAFLD) was 28%, which was significantly higher than that in the general NAFLD population but similar to that reported in a longitudinal cohort study (29%)[10,20].

Disturbances in endocrine hypothalamic-pituitary axes have been associated with NAFLD, including GHD, hypothyroidism, and hypogonadism[21]. GHD has been considered an important contributing factor to the metabolic change encountered in hypopituitary patients. The majority of the patients in the present study had GHD,and the average IGF1 level was considerably low in this patient group. Comparative analysis demonstrated no significant differences in serum growth hormone or IGF1 levels between hypopituitary patients with NAFLD and those without. Moreover, no significant differences were noted in the distribution of anterior pituitary deficiencies or hormone replacement in current hypopituitary patients with and without NAFLD.The causes of hypopituitarism varied, and combined pituitary hormone deficiency,including central diabetes insipidus, was more common in this population. Therefore,we speculated that this condition would be the result of multihormonal deficiency and that additional aggravating factors would be responsible for the pathophysiology of NAFLD in hypopituitary patients.

Congenital hypopituitarism, intracranial germ cell tumor, and craniopharyngioma w ere common causes of hypopituitarism in the present stud y. The latter tw o disorders usually affect the secretion of AVP. Combined deficits in AVP secretion and thirst sensation can result in life-threatening hyperosmolality and hypernatremia. In the current study, 72% of patients with hypopituitarism presented with central diabetes insipidus. Thus, hypernatremia manifested in these hypopituitary patients.Plasma osmolality levels of cirrhotic patients were significantly higher than those of NAFLD patients, and similar differences were noted with regard to serum sodium and serum chlorine levels. Homeostasis of electrolytes and the normal concentration of plasma osmolality are essential for the maintenance of normal cellular function.Cell volume changes are triggered by osmotic substances, thereby activating signalsthat contribute to the control and regulation of metabolism and gene exp ression.Follow ing the ind uction of hyp erosmolality, cells shrink and may und ergo apoptosis[22-24]. Moreover, some components of insulin signaling, such as the enzyme protein kinase B, are sensitive to hyperosmolarity[22-24]. Hyperosmolality can inhibit insulin-ind uced glucose up take, glycogen synthesis, and lip ogenesis in 3T3-L1 adip ocytes[25,26]. Therefore, the d ehyd ration of hep atocytes, w hich is triggered by hyperosmolarity, can also ind uce insulin and cytokine resistance and promote the progression of NAFLD[27,28].

Table 3 Liver function including liver enzymes, routine blood tests, and coagulation function tests in our patients

The pathogenesis of NAFLD has not been fully elucidated, and multiple etiologies have been proposed. The most prevalent hypothesis includes insulin resistance as the key mechanism leading to hepatic steatosis and possibly to steatohepatitis and liver fibrosis[6,29]. The d ata rep orted in the present stud y ind icated that the metabolic changes accompanied by hyp opituitarism/panhypopituitarism notably caused an increase in the weight and obesity parameters, insulin resistance, and hyperlipidemia,which is consistent w ith the findings reported previously[7-10]. HOMA-IR in patients w ith hyp op ituitarism and cirrhosis w as significantly higher than that in p atients w ithout NAFLD. Moreover, fasting insulin concentration w as positively associated with plasma osmolality in patients with NAFLD, although no correlation w as noted betw een HOMA-IR and plasma osmolality. However, peripheral IR, pancreatic β-cell dysfunction, and even diabetes mellitus may occur in cirrhotic patients[30,31]. Therefore,the effect of the hyperosmolar state on the progression of NAFLD in hypopituitary patients requires further studies.

Figure 1 Measurement of biochemical indicators after hormone replacement in cirrhotic patients. The P-values were compared before and after hormone replacement. A-C: Changes of total cholesterol (A), triglyceride (B), and homeostasis model assessment of insulin resistance (C). D-F: Changes of serum aminotransferase alanine aminotransferase (D), aspartate aminotransferas (E), and γ-glutamyltransferase (F). G-I: Changes of white blood count (G), neutrophil count(H), and blood platelet count (I). Significant differences were defined by a P-value being lower than 0.05. ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; GGT: γ-glutamyltransferase; WBC: White blood cells; NEUT: Neutrophil granulocyte; PLT: Blood platelet; TC: Total cholesterol; TG: Triglyceride;HOMA-IR: Homeostasis model assessment of insulin resistance.

The sensitivity of ultrasound has been reported to range from 53% to 100% and its sp ecificity from 77% to 98%[32]. Higher d iagnostic sensitivities and specificities are achieved during the evaluation of moderate to severe hepatic steatosis cases, whereas low er values are noted during all grades of hepatic steatosis[32]. The accurate diagnosis of NASH and cirrhosis can be conducted by liver biopsy. Adams et al[10]reported that in ten p atients w ith hyp othalamic and p ituitary d ysfunction w hose NAFLD symptoms w ere confirmed by liver biopsy, six were cirrhotic (29% of the total cohort),two exhibited NASH with fibrosis (10.5% of the total cohort), and two presented with simple steatosis. In the present stud y, tw o of our patients underw ent liver biopsy,which indicated the lack of differential diagnosis between simple steatosis and NASH from NAFLD. Consequently, the incid ence of NASH and fibrosis in our cohort w as not clear. Only tw o of the seven cirrhotic p atients received regular hormone rep lacement, w hile three of them received this therap y until they d evelop ed symptoms of decompensated cirrhosis, such as upper gastrointestinal bleeding and hypersplenism. Comparative analysis of our patients indicated that the variables of hepatic synthetic function, w hite blood count, and blood platelet count of cirrhotic patients w ere significantly low er than those of patients w ith hypopituitarism without NAFLD. How ever, no significant d ifferences w ere noted in the level of serum aminotransferase or bilirubin. Follow ing a period of regular hormone replacement,the biochemical ind icators of the seven cirrhotic p atients exhibited no further deterioration and were reflected by serum aminotransferase levels, white blood count,and blood p latelet count, although it w as unknow n w hether the p athological ind icators of liver cirrhosis had improved. Therefore, routine liver blood tests could not be used to assess the progression of liver fibrosis and cirrhosis in NAFLD subjects,and liver biopsy w as performed if necessary[1,2].

In conclusion, our data demonstrated a high prevalence of NAFLD and cirrhosis in p atients w ith hyp op ituitarism. Notably, hyp op ituitary p atients w ith cirrhosis exhibited significantly higher BMI and HOMA-IR comp ared w ith those w ithout NAFLD. In ad d ition, fasting insulin concentration w as positively associated w ith plasma osmolality in patients with NAFLD, following adjustment for gender, age, and BMI. We reported for the first time that plasma osmolality and serum sodium levels of hyp op ituitary p atients w ith cirrhosis w ere significantly higher than those of hypopituitary patients w ith NAFLD. Ad ditional studies are required to confirm that hyperosmolality may be a significant contributor to the deterioration of NAFLD in hypopituitary p atients. Routine liver blood tests could not be used to assess the progression of liver fibrosis and cirrhosis in subjects with NAFLD.

The present study w as limited by insufficient sample size, as hypopituitarism is a rare cond ition. We d id not use corrections for multip le comparisons because the find ings from this analysis were general associations rather than affirmative findings.Thus, prospective, multicenter, cohort studies and animal experiments are required in the future to facilitate further und erstand ing of the NAFLD pathop hysiology in hypopituitary patients.

ARTICLE HIGHLIGHTS

Research background

Non-alcoholic fatty liver disease (NAFLD) is a major global health problem with a substantial rise in prevalence over the last decades. Patients with hypothalamic-pituitary disease have the features of central obesity, insulin resistance, and dyslipidemia, and there is an increased prevalence of liver dysfunction consistent with NAFLD in this population. Growth hormone deficiency (GHD) has been considered as an important contributing factor to these metabolic changes in hypopituitary patients. However, the causes of hypopituitarism in the reported studies varied, and combined pituitary hormone deficiency, including central diabetes insipidus,was more common in this population. This retrospective cross-sectional study was performed to analyze clinical characteristics of NAFLD in Chinese adult hypopituitary/panhypopituitary patients, and to explore the risk factors that lead to rapid progression to cirrhosis.

Research motivation

The research motivation of the present study was to identify possible risk factors related to NAFLD in patients with hypopituitarism by summarizing the characteristics of NAFLD in this patient population, in order to prevent and delay the occurrence and progression of NAFLD in patients with hypopituitarism in future.

Research objectives

The main objectives of the present study were to analyze clinical characteristics of NAFLD in Chinese adult hypopituitary/panhypopituitary patients, and to explore the risk factors that lead to rapid progression to cirrhosis. Additional studies are required to research the mechanism of rapid progression of NAFLD to cirrhosis in hypopituitary patients.

Research methods

The present study is a retrospective cross-sectional observational study. A total of 50 adult Chinese patients with hypopituitarism and/or panhypopituitarism were enrolled. The data were extracted from the medical records, including patients' characteristics, diagnosis and treatment,biochemical and hormonal tests, and abdominal ultrasound. And then statistical analysis was performed.

Research results

Fifty-four percent of hypopituitary patients in this study were diagnosed with NAFLD, and seven patients were cirrhotic, which was significantly higher than that of the general population.Body mass index (BMI) and homeostasis model assessment of insulin resistance (HOMA-IR) of the cirrhotic patients were significantly higher than those of the patients without NAFLD.Moreover, plasma osmolality and serum sodium concentration of the cirrhotic patients were significantly higher than those of the NAFLD patients, and fasting insulin concentration was positively associated with plasma osmolality in patients with NAFLD, following adjustment for gender, age, and BMI.

Research conclusions

The present study demonstrated a high prevalence of NAFLD and cirrhosis in patients with hypopituitarism. Hypopituitary patients with cirrhosis exhibited significantly higher BMI and HOMA-IR compared with those without NAFLD. In addition, fasting insulin concentration was positively associated with plasma osmolality in patients with NAFLD, following adjustment for gender, age, and BMI. Moreover, we report for the first time that plasma osmolality and serum sodium levels of hypopituitary patients with cirrhosis were significantly higher than those of hypopituitary patients with NAFLD. Additional studies are required to confirm that hyperosmolality may be a significant contributor to the rapid progression of NAFLD in hypopituitary patients.

Research perspectives

There is a high prevalence of NAFLD and cirrhosis in patients with hypopituitarism, but the factors that lead to rapid progression to cirrhosis are not clear. Further studies are needed to determine whether hyperosmolality contributes to the deterioration of NAFLD in hypopituitary patients.

ACKNOWLEDGEMENTS

We are grateful to Yue-Lun Zhang for his recommendations on the statistical analysis.