**2.3 DHEA-S levels in NAFLD**

Recently, Charlton *et al*. observed that levels of DHEA are significantly lower in patients with histologically advanced NASH, as compared with patients with mild NASH or simple fatty liver. (Charlton M, 2009). DHEA levels exert a good sensitivity and specificity in discriminating patients with more advanced histological disease, as shown by the receiver operating characteristic (ROC) analysis. To validate their results, we also determined circulating DHEA levels in Japanese patients with 133 biopsy-proven NAFLD. Of 133 patients, 90 patients were diagnosed as NASH: 73 patients had stage 0–2, and 17 had stage 3 or 4. In addition, 399 sex- and age-matched healthy people participating in health checkups who had normal levels of alanine aminotransferase (ALT) levels (≤ 30 IU/L) were also enrolled as the control group. Body mass index (BMI), aspartate aminotransferase (AST), ALT, γGT, triglyceride, and HOMA-IR were significantly higher in NAFLD patients than those in the control group, whereas serum DHEA-S levels were similar between both groups. Consistent with our result, in patients with polycystic ovary syndrome (PCOS), DHEA-S levels were

Whereas DHEA levels naturally reach their peak in the early morning hours, DHEAS levels show no diurnal variation. From a practical point of view, measurement of DHEAS is preferable to DHEA, as levels are more stable. The Baltimore Longitudinal Study of Aging (BLSA) is a multidisciplinary observational study of the physiological and psychological aspects of human aging and diseases and conditions that increase with age. In BLSA, men who had higher DHEAS levels had significantly greater longevity than men with lower levels. (Roth et al., 2002) In Japan, a 27-year study in a community-based cohort (Tanushimaru study) indicated that DHEAS level may be a predictor of longevity in men, independent of age, blood pressure, and plasma glucose (Enomoto et al, 2008). Low serum levels of DHEA(-S) predict death from all causes, cardiovascular disease, and ischemic heart disease in elderly Swedish men. (Ohlsson et al., 2010) On the basis of these results, serum DHEA level is known to be an indicator of longevity at least in men and is often determined in anti-aging checkups (Nishizaki et al., 2009) . Elevated levels of DHEA are found in patients with Cushing syndrome or congenital adrenal hyperplasia, while DHEA levels are reported to be low in some people with anorexia, end-stage kidney disease, type 2 diabetes, AIDS, adrenal insufficiency, and in the critically ill. Some studies suggested that low serum DHEA-S levels were associated with the metabolic syndrome (Muller et al., 2005, Chen et al., 2010). In contrast, several studies found that DHEA levels are not different between subjects with metabolic syndrome and without. (Fukui et al., 2007, Haring et al., 2009, Akishita et al., 2010) It is suggested that age per se is an important correlate of the associations between DHEA-S and metabolic variables. In this way, the previous studies regarding the association between endogenous DHEA-S level and metabolic syndrome are inconsistent. Previous studies have shown that diabetic patients with high serum levels of insulin have lower serum levels of DHEA and DHEA-S. (Yamaguchi et al., 1998). A negative correlation between DHEA and hyperinsulinemia has been repeatedly demonstrated. (Kauffman et al., 2006, Saygili et al., 2005, Vasarhelyi et al., 2003). Fukui and colleagues reported that low levels of DHEA are associated with atherosclerosis and deterioration of urinary albumin excretion in male patients with type 2 diabetes (Fukui et al., 2004, 2005, 2006). Similarly, Serum DHEA-S level seem to be associated with atherosclerosis in diabetic postmenopausal women independent of age, body

stature, diabetic status, and other atherosclerotic risk factors (Kanazawa et al., 2008).

Recently, Charlton *et al*. observed that levels of DHEA are significantly lower in patients with histologically advanced NASH, as compared with patients with mild NASH or simple fatty liver. (Charlton M, 2009). DHEA levels exert a good sensitivity and specificity in discriminating patients with more advanced histological disease, as shown by the receiver operating characteristic (ROC) analysis. To validate their results, we also determined circulating DHEA levels in Japanese patients with 133 biopsy-proven NAFLD. Of 133 patients, 90 patients were diagnosed as NASH: 73 patients had stage 0–2, and 17 had stage 3 or 4. In addition, 399 sex- and age-matched healthy people participating in health checkups who had normal levels of alanine aminotransferase (ALT) levels (≤ 30 IU/L) were also enrolled as the control group. Body mass index (BMI), aspartate aminotransferase (AST), ALT, γGT, triglyceride, and HOMA-IR were significantly higher in NAFLD patients than those in the control group, whereas serum DHEA-S levels were similar between both groups. Consistent with our result, in patients with polycystic ovary syndrome (PCOS), DHEA-S levels were

**2.3 DHEA-S levels in NAFLD** 

**2.2 The significance of serum DHEA-S levels** 

similar between those with NAFLD and without. (Kauffman et al., 2010). According to a crosssectional population-based study derived from data of 1912 men, however, the highest risk of hepatic steatosis was found in subjects with the highest serum DHEA-S levels (Völzke H et al., 2010). DHEA and DHEAS levels of post menopausal women with fatty liver were greater than those of post menopausal women with normal histology. (Saruç et al., 2003) These results are contrast to our study. Discrepancies between these studies and ours might be explained by differences in the selection of subjects, sex, size of the study populations and ethnicity.

Only in our NAFLD patients, NASH patients had lower levels of serum DHEA-S levels compared to non-NASH patients (Fig 4). Serum DHEA levels were negative correlated with age in males and females (Fig 5). A " dose effect " of lower DHEA-S and advanced fibrosis was observed, with a mean DHEA-S of 170.4±129.2, 137.6±110.5, 96.2±79.3, 61.2±46.3, and

Fig. 4. Serum DHEA levels in biopsy-proven NAFLD.

Fig. 5. The relationship between serum DHEA-S levels and age in NAFLD patients

Dehydroepiandrosterone in Nonalcoholic Fatty Liver Disease 103

Variables Odds ratio 95% confidence interval *P-value* DHEA-S ≦66 μg/dl 4.9549 1.1691-20.9996 0.0229 age≧65 yr 2.8962 0.7843-10.6948 0.1106 sex (female) 1.9494 0.3765-10.0935 0.4264 HOMA-IR≧5 2.3671 0.6276-8.9273 0.2033 BMI ≧28 kg/m2 1.0446 0.2619-4.1658 0.9508 Diabetes 1.6007 0.3904-6.5023 0.5107 Dyslipidemia 0.2500 0.0682-0.9162 0.0364 Hypertension 0.4184 0.1022-1.7126 0.2256

HOMA-IR, homeostasis assessment model for insulin resistance; BMI, body mass index

Table 1. Logistic regression models of the association of NAFLD (advanced versus mild) with dehydroepiandrosterone sulfate (DHEA-S) levels and other clinical variables

Fig. 7. Laparoscopic findings and liver histology of a case of NASH-LC who was referred to Center for Digestive and Liver Diseases, Nara City Hospital. A: laparoscopy (lt lobe), B: laparoscopy (rt lobe), C: microscopy (HE stain), D: microscopy (Masson-trichrome stain).

(Franz et al., 1979). It was also important to consider whether low levels of DHEA-S might occur as a result of CLD in general versus a specific phenomenon of histologically more

Nakajima T et al revealed that telomere shortening, a marker of senescence, could be associated with hepatic steatosis, insulin resistance, oxidative stress in the liver, and impaired regenerative response in NAFLD patients (Nakajima et al., 2006). The hepatic

advanced NAFLD.

30.0±32.0, for fibrosis stages 0, 1, 2, 3 and 4, respectively. The area under the ROC curve for DHEA in separating patients with and without advanced brosis was 0.788. The sensitivity of a DHEA-S-value of 66 mg/dL or less for the presence of more advanced NAFLD was 76.5% and specicity was 73.3% (85/116) (Fig 6)(Sumida et al., 2010a). Our data suggest that patients with DHEA-S levels greater than 66 μg/dL are highly unlikely to have advanced NAFLD (4/89 patients, sensitivity 76% and specificity 73%). Multivariate logistic regression analyses found that serum level of DHEA-S below 66μg/ml was selected as an independent predictor for advanced fibrosis even after adjusting for age, gender and insulin resistance (Table 1).We intended to support the concept that the association between low levels of DHEA and worsening histology is independent of age, sex and insulin resistance. Decreased levels of DHEA can have important roles in the progression hepatic fibrosis in NAFLD. It is expected that determinant of serum DHEA become a predictor of hepatic fibrosis in NAFLD. A 53-year female who had been pointed out her fatty liver without any medications was referred to our hospital because of thrombocytopenia (platelet count 4.6×104/μl). Her BMI was 31.6kg/m2 and she had mildly elevated transaminase activities (AST 61IU/l, ALT 59IU/l) and prolonged prothrombin time (66%). Laparoscopic findings revealed nodular liver and her liver histology showed NASH (Brunt grade 3, stage 4) (Fig 7). Her DHEA-S levels was the lowest (5μg/dl) among our NAFLD patients.

Fig. 6. ROC analysis for predicting severe fibrosis (stage 3-4).

Free fatty acids (FFAs), which lead to oxidative stress in NASH, are the major source of DHEA (Fig 3). The inability to produce appropriate amounts of DHEA in response to FFAs may translate into a more rapid and worsening progression toward NASH (Manco et al., 2008). Serum DHEA-S levels depend on adrenal DHEA production and its hepatic metabolism mediated by DHEA sulfotransferase (DHEA-ST) which catalyzes sulfonation of DHEA to form DHEA-S. It is hypothesized that a low level of DHEA-S was due to a defect in sulfurylation in patients with hepatic cirrhosis, since DHEA-ST is synthesized in the liver

30.0±32.0, for fibrosis stages 0, 1, 2, 3 and 4, respectively. The area under the ROC curve for DHEA in separating patients with and without advanced brosis was 0.788. The sensitivity of a DHEA-S-value of 66 mg/dL or less for the presence of more advanced NAFLD was 76.5% and specicity was 73.3% (85/116) (Fig 6)(Sumida et al., 2010a). Our data suggest that patients with DHEA-S levels greater than 66 μg/dL are highly unlikely to have advanced NAFLD (4/89 patients, sensitivity 76% and specificity 73%). Multivariate logistic regression analyses found that serum level of DHEA-S below 66μg/ml was selected as an independent predictor for advanced fibrosis even after adjusting for age, gender and insulin resistance (Table 1).We intended to support the concept that the association between low levels of DHEA and worsening histology is independent of age, sex and insulin resistance. Decreased levels of DHEA can have important roles in the progression hepatic fibrosis in NAFLD. It is expected that determinant of serum DHEA become a predictor of hepatic fibrosis in NAFLD. A 53-year female who had been pointed out her fatty liver without any medications was referred to our hospital because of thrombocytopenia (platelet count 4.6×104/μl). Her BMI was 31.6kg/m2 and she had mildly elevated transaminase activities (AST 61IU/l, ALT 59IU/l) and prolonged prothrombin time (66%). Laparoscopic findings revealed nodular liver and her liver histology showed NASH (Brunt grade 3, stage 4) (Fig 7).

Her DHEA-S levels was the lowest (5μg/dl) among our NAFLD patients.

Fig. 6. ROC analysis for predicting severe fibrosis (stage 3-4).

Free fatty acids (FFAs), which lead to oxidative stress in NASH, are the major source of DHEA (Fig 3). The inability to produce appropriate amounts of DHEA in response to FFAs may translate into a more rapid and worsening progression toward NASH (Manco et al., 2008). Serum DHEA-S levels depend on adrenal DHEA production and its hepatic metabolism mediated by DHEA sulfotransferase (DHEA-ST) which catalyzes sulfonation of DHEA to form DHEA-S. It is hypothesized that a low level of DHEA-S was due to a defect in sulfurylation in patients with hepatic cirrhosis, since DHEA-ST is synthesized in the liver


HOMA-IR, homeostasis assessment model for insulin resistance; BMI, body mass index

Table 1. Logistic regression models of the association of NAFLD (advanced versus mild) with dehydroepiandrosterone sulfate (DHEA-S) levels and other clinical variables

Fig. 7. Laparoscopic findings and liver histology of a case of NASH-LC who was referred to Center for Digestive and Liver Diseases, Nara City Hospital. A: laparoscopy (lt lobe), B: laparoscopy (rt lobe), C: microscopy (HE stain), D: microscopy (Masson-trichrome stain).

(Franz et al., 1979). It was also important to consider whether low levels of DHEA-S might occur as a result of CLD in general versus a specific phenomenon of histologically more advanced NAFLD.

Nakajima T et al revealed that telomere shortening, a marker of senescence, could be associated with hepatic steatosis, insulin resistance, oxidative stress in the liver, and impaired regenerative response in NAFLD patients (Nakajima et al., 2006). The hepatic

Dehydroepiandrosterone in Nonalcoholic Fatty Liver Disease 105

Arlt W. (2004). Dehydroepiandrosterone and ageing. *Best Pract Resn Clin Endocrinol Metab*

Baranova, A. & Tran TP. & Birerdinc A. & Younossi ZM. (2011). Systematic review:

Bednarek-Tupikowska, G. & Gosk, I. & Szuba, A. &Bohdanowicz-Pawlak, A. & Kosowska, B.

men. *Hypertens Res* Vol.33, No.6, pp. 587-591

Vo.18, No.3, pp. 363-380.

*S A* Vol. 97, No. 8, pp. 4279-4284.

No.4, pp. 842-845.

54, No.3, pp. 765-769.

No.5, pp. 563–573.

testosterone with metabolic syndrome and its components in middle-aged Japanese

association of polycystic ovary syndrome with metabolic syndrome and nonalcoholic fatty liver disease. Aliment Pharmacol Ther Vol. 33, No.7, pp. 801-814. Baulieu, EE. &Thomas, G. & Legrain, S . & Lahlou, N. & Roger, M. & Debuire, B. &

Faucounau, V. & Girard, L. & Hervy, MP. & Latour, F. & Leaud, MC. & Mokrane, A. & Pitti-Ferrandi, H. & Trivalle, C. & de Lacharrière, O. & Nouveau, S. & Rakoto-Arison, B. & Souberbielle, JC. & Raison, J. & Le Bouc, Y. & Raynaud, A. & Girerd, X. & Forette F. (2000). Dehydroepiandrosterone (DHEA), DHEA sulfate, and aging: contribution of the DHEAge Study to a sociobiomedical issue. *Proc Natl Acad Sci U* 

& Bidzinska B, et al. (2006). Inuence of dehydroepiandrosterone on platelet aggregation, superoxide dismutase activity and serum lipid peroxide concentrations in rabbits with induced hypercholesterolemia. *Med SciMonit* Vol. 6, No.1, pp. 40-45. Charlton M. & Angulo, P. & Chalasani, N. & Merriman, R. & Viker, K. &

Charatcharoenwitthaya, P. & Sanderson, S. & Gawrieh, S. & Krishnan, A. & Lindor K (2008). Low circulating levels of dehydroepiandrosterone in histologically advanced nonalcoholic fatty liver disease. *Hepatology* Vo. 47, No.2, pp. 484-492. Chen, YC. & Chang, HH. & Wen, CJ. & Lin, WY. & Chen, CY. & Hong, BS. & Huang KC. (2010).

Elevated serum dehydroepiandrosterone sulphate level correlates with increased risk for metabolic syndrome in the elderly men. *Eur J Clin Invest* Vo. 40, No.3, 220-225. Day, CP. & James, OF. (1998). Steatohepatitis: a tale of two "hits"? *Gastroenterology* Vol. 114,

replacement on insulin sensitivity and lipids in hypoadrenal women. *Diabetes* Vol.

& Nanjo, Y. & Shigetoh, Y. & Imaizumi T. (2008). Serum dehydroepiandrosterone sulfate levels predict longevity in men: 27-year follow-up study in a community-based

sulfate concentrations in normal subjects and men with cirrhosis. *Steroids* Vol, 34,

(2004). Association between urinary albumin excretion and serum dehydroepiandrosterone sulfate concentration in male patients with type 2 diabetes: a possible link between urinary albumin excretion and cardiovascular

Wada, K. & Hasegawa, G. & Yoshikawa T. (2005). Serum dehydroepiandrosterone sulfate concentration and carotid atherosclerosis in men with type 2 diabetes.

Dhatariya, K. & Bigelow, ML. & Nair, KS. (2005). Effect of dehydroepiandrosterone

Enomoto, M. & Adachi, H. & Fukami, A. & Furuki, K. & Satoh, A. & Otsuka, M. & Kumagae, S.

Fukui, M. & Kitagawa, Y. & Nakamura, N. & Kadono, M. & Hasegawa, G. & Yoshikawa T.

Fukui, M. & Kitagawa, Y. & Nakamura, N. & Kadono, M. & Yoshida, M. & Hirata, C. &

cohort (Tanushimaru study). *J Am Geriatr Soc* Vol. 56, No.6, pp. 994-998. Franz, C. & Watson, D. & Longcope, C. (1979). Estrone sulfate and dehydroepiandrosterone

disease. *Diabetes Care* Vol. 27, No.12, pp. 2893-2897.

*Atherosclerosis* Vol. 181, No.2, pp. 339-344

expression of senescence marker protein-30 (SMP30), which was identified as an antioxidant and anti-apoptotic protein, decreased in the proportion of the hepatic fibrosis in NAFLD patients (Park et al., 2010). These results suggest that the association of aging with NASH pathogenesis is noteworthy.
