Clinical, anthropometric, biochemical and histological character of nonalcoholic fatty liver disease without Insulin Resistance

Abstract

Objectives

Nonalcoholic Fatty Liver Disease (NAFLD) is thought to be a hepatic manifestation of Metabolic Syndrome (MS) or Insulin Resistance (IR). The aim of the study was to explore the clinical, anthropometric, metabolic, biochemical and histological profile of NAFLD patients without IR by comparing it with NAFLD with IR.

Methods

Total 851 patients with sonographic evidence of fatty liver were included. These patients underwent clinical, anthropometric, biochemical and histological evaluation. IR was calculated using the homeostatic model assessment. Liver biopsy done in 285 patients who consented for the procedure and who had MS or raised ALT.

Results

Among 851 NAFLD patients, 561(65.9%) patients were without IR and 290 (34.1%) patients were with IR. The proportion of male sex [230 (41.0%) vs. 89 (30.7%); P = 0.046] were higher but diabetes [19.10% vs. 39.0%; P = 0.000] and MS were [58.80%vs. 78.10%; P = 0.014] significantly lower in non IR group. Body Mass Index (BMI) kg/m² and Waist Circumference (WC) in cm were also lower in non IR group: [26.6 ± 3.5 vs. 27.9 ± 4.3; P = 0.002] and [93.3 ± 8.4 vs. 95.9 ± 8.4; P = .003]. Lipid profile, ALT, AST and ALP were not differed between the groups. Histopathology reports revealed that lobular inflammation, ballooning and fibrosis were similar in two groups, only steatosis score was higher in IR group [2.0 ± 0.7 vs. 1.8 ± 0.8; P = 0.007].

Conclusion

There are significant proportion of NAFLD patients without IR in Bangladesh. NAFLD patients without IR predominantly male, had lower BMI, WC, MS and diabetes. Histologically NAFLD without IR equally severe with ballooning, lobular inflammation and fibrosis except steatosis. Insulin resistance is the principal but not the sole factor for NAFLD in our population.

Introduction

Nonalcoholic Fatty Liver Disease (NAFLD) includes a wide spectrum of liver pathologies, ranging from pure steatosis, which is usually benign, to Nonalcoholic Steatohepatitis (NASH), which can progress to cirrhosis, liver failure and hepatocellular carcinoma.^{1, 2, 3} NAFLD is now recognized as the commonest cause of hepatic dysfunction in the general population.⁴ Prevalence of NAFLD in western countries ranges from 24% to 42% and it is the most prevalent chronic liver disease.^{5, 6, 7} Data from USA suggest that up to 30% & 3-5% of the general population has NAFLD and NASH, respectively and prevalence of NASH may attain 25–70% among morbidly obese patients.⁸ The prevalence of fatty liver in the general population of India has been shown to be 16.6–24%, which is similar to that reported from the developed countries.^7,9 Prevalence of sonologically detected nonalcoholic fatty liver disease in Bangladesh is 18.5%–34.34%.^10,11 Another study revealed that prevalence of NASH among the NAFLD patients in Bangladesh is 42.40%.¹² The burden of NAFLD is expected to rise due to increasing obesity and type 2 diabetes mellitus as a result of economic prosperity, sociodemographics and life style changes. It is thought that NAFLD is a hepatic manifestation of metabolic syndrome or Insulin Resistance (IR).¹³ It is also considered an early expression of the metabolic syndrome.¹⁴ Although NAFLD is more common in subject with obesity and diabetes mellitus (DM), it also occurs in lean and non-diabetic subject.^{15, 16, 17} In developed countries, 15 % of NAFLD patients are nonobese.¹⁸ Developing country scenario is quite different, a recent study in rural India found that 75 % of NAFLD patients had a Body Mass Index (BMI) of less than 25 kg/m² and 54 % had neither overweight nor abdominal obesity.⁹ Also, among the Bangladeshi NAFLD patients, 25.6% of them are nonobese and 53.1% of nonobese NAFLD cases were NASH.¹⁹ Though IR is a major pathogenic mechanism of developing NAFLD,²⁰ some studies in this subcontinent failed to confirm any relation of IR to NAFLD.^15,21 The two recent studies Das et al. and Singh et al. from India had found prevalence of IR in only 46% and 54.46% respectively of NAFLD patients; almost half of the NAFLD population was devoid of IR.^15,21 Therefore, existing data showed that IR is not the sole predictor of the pathogenesis of NAFLD rather NAFLD is a polygenic disease caused by both genetic and environmental factors. This created the suspicion that NAFLD sans IR could be a different entity from NAFLD with IR. There are scarcity of studies in this regard. So to characterize NAFLD without IR we have compared the clinical, anthropometric, biochemical and histological profile of NAFLD patients with and without IR in this study.

Patients and methods

We have included 851 patients in whom fatty liver was diagnosed on ultrasonography during the period March’2013 to December’2016 at the Department of Hepatology of the University Hospital. Patient with significant alcohol intake (>30 g m/day in case of male and >20 g m/day in case of female), evidence of acute/chronic viral hepatitis, drug induced hepatitis, autoimmune hepatitis, other metabolic liver diseases, pregnancy and patients of ongoing drugs that may cause fatty liver were excluded.

The procedures followed were in accordance with the ethical standards of the Helsinki Declaration of 1975. All individuals provided written informed consent prior to enrollment in the study. These patients underwent clinical evaluation, anthropometric measurements, and blood tests. Blood pressure, basal metabolic index (BMI), and waist circumference (WC) were recorded for every patients. The Regional Office for the Western Pacific Region of WHO, the International Association for the Study of Obesity and the International Obesity Task Force proposed new BMI criteria for Asians.²² According to this criteria, patients with BMI of ≥25 kg/m² were labeled as obese, those with a BMI ≤ 25 kg/m² were nonobese and patients with BMI 23–24.9 kg/m² were overweight.²³

Liver function tests were performed prior to the liver biopsy. Blood samples were obtained under fasting conditions, and the following tests were performed using standard laboratory methods: ALT, AST, alkaline phosphatase, Gamma-Glutamyl Transpeptidase (GGT). FBG, lipid profile and Serum fasting insulin were assessed using the method of indirect chemiluminescence (MEIA).

IR was calculated using the Homeostatic Model Assessment (HOMA) method using a mathematical model derived from FBG and plasma insulin. The value of HOMA was calculated by the following equation: (fasting insulin (μU/ml) × FBG (mg/dl))/405, and depicted as HOMA-IR value.²⁴ Study done in Indian population had shown that the normal value of IR as assessed by HOMA-IR is less than 2.^25,26 For our study, HOMA-IR value above 2 was considered to indicate IR.²⁰

The modified criteria of National Cholesterol Education Program, Adult Treatment Panel III (NCEP, ATP III) for Asians were considered for diagnosis of metabolic syndrome (MS).^{27, 28, 29} Reaching the defining level for any three out of five of the parameters satisfies the clinically applicable definition of metabolic syndrome, three of the five listed criteria were considered: waist circumference (WC) ≥80 cm for women and ≥90 cm for men; serum triglyceride ≥150 mg/dL (1.7 mmol/L); serum HDL cholesterol <50 mg/dL (1.3 mmol/L) for women and <40 mg/dL (1 mmol/L) for men; elevated blood pressure (systolic blood pressure ≥130 and or diastolic blood pressure ≥85 mmHg or drug treatment for hypertension); and plasma glucose concentration ≥100 mg/dL (5.6 mmol/L) or drug treatment for diabetes.

Hepatic ultrasonography had been performed following 8 hours of fasting using a 3.5 MHz probe. NAFLD was diagnosed by the sonographic findings of the echogenicity of the liver is greater than that of the renal cortex and were graded as per Gore et al.:

• •Grade 1 (mild): normal visualization of diaphragm/ intrahepatic vessels;
• •Grade 2 (moderate): impaired visualization of diaphragm/ intrahepatic vessels;
• •Grade 3 (severe): poor visualization of diaphragm/ intrahepatic vessels.³⁰

Liver biopsy had been performed in 285 patients who consented for the procedure and who had metabolic syndrome or raised ALT (in case of male >30 U/L and in case of female >19 U/L), using Trucut biopsy needle (14 F) through the intercostal approach. Biopsy sample of <1.5 cm and or containing less than six portal tracts was considered as inadequate tissue. Five biopsy samples were inadequate to comment for histopathology, and five patients were withdrawn from the study. So, at the end, 275 histopathology reports were available for the study, which were analyzed by a histopathologist blinded to the patients clinical and biochemical results. The diagnosis of NASH was based on the criteria of Brunt et al., modified by Kleiner et al.^31,32. In this scoring system, the degree of disease activity in NAFLD was evaluated using the NAFLD Activity Score (NAS), which was calculated as the unweighted sum of the scores for steatosis (0–3), lobular inflammation (0–3), and hepatocyte ballooning (0–2) and thus ranged from 0 to 8. A NAS of 5 or more was diagnosed as “NASH,” and other than NASH, a non-NASH fatty liver (NNFL) was considered. Hepatic fibrosis staging was as follows: 0 = no fibrosis; 1 = zone 3 fibrosis only; 2 = zone 3 and portal/periportal fibrosis; 3 = bridging fibrosis; and 4 = cirrhosis.

Basing on IR, the patients were categorized into 2 groups [A (Insulin Resistant): IR > 2; B (non IR): IR ≤ 2], and comparison was made between the two groups.

Statistical analysis

It was done by using SPSS 17.0 software. Results are presented as mean ± standard deviation (SD) for quantitative data and as numbers or percentages. Quantitative data was compared by Student t test or Man Whitney U test to compare the variables between IR and non-IR groups. Categorical data were compared using the χ² test. Logistic regression analysis was done to measure the odds ratio of different variable. In the present study ‘P’ value <0.05 was considered statistically significant.

Results

The total number of patients with NAFLD in this study was 851, of whom 561(65.9%) patients were without IR, and 290 (34.1%) patients were having IR. The patients were in age range 18–70 years. The mean BMI was (27.2 ± 3.9) kg/m² and Waist Circumference (WC) was (89.9 ± 1.7) cm, which were predominantly in the range of obesity and metabolic syndrome, respectively. The mean FBG and HOMA-IR were (5.8 ± 1.9) mmol/l and (2.0 ± 1.5) respectively, with range of HOMA-IR 0.4–9.7.

The females outnumbered males in both the groups with sex [male: female] ratio of almost 2:3 in patients without IR and 1:2 in IR group. Among the male patients between IR and non-IR group, males had significantly lower in IR group (30.7% vs 41.0%; P = 0.046).

On analysis of metabolic parameters, 2/3^rd patients 574(67.50%) were obese, one fourth 220 (25.80%) patients were diabetics, two third 557(65.50%) patients had MS and one forth 238 (27.90%) patients were hypertensive. The proportion of patients with diabetes [39.0% vs. 19.10%; P = 0.000] and MS [78.10% vs. 58.80%; P = 0.014] were significantly higher in IR group. Though not significant, IR group had greater number of hypertensive patients (36.60% vs 23.50%; P = 0.085). Waist circumference (95.9 ± 8.4 vs. 93.3 ± 8.4) cm P = 0.003 and BMI (27.9 ± 4.3 vs 26.6 ± 3.5) kg/m² P = 0.002 were also significantly higher in IR group (Table 1).

Table 1.

Clinical, Anthropometric and Biochemical Character of NAFLD Patients With or Without Insulin Resistance.

Variable	Total	Non IR	IR	P
Number (%)	851(100)	561(65.9)	290 (34.1)
Age in years	39.9 ± 9.6	39.5 ± 9.4	40.5 ± 10.0	0.297
Sex male n (%)	319 (37.5)	230 (41.0)	89 (30.7)	0.046
BMI kg/m2	27.2 ± 3.9	26.6 ± 3.5	27.9 ± 4.3	0.002
Diabetes n (%)	220 (25.8)	107 (19.1)	113 (39.0)	0.000
Hypertension n (%)	238 (27.9)	132 (23.5)	106 (36.6)	0.085
Metabolic syndrome n (%)	557 (65.5)	330 (58.8)	227 (78.1)	0.014
Waist circumference in cm	89.9 ± 1.7	93.3 ± 8.4	95.9 ± 8.4	0.003
Fasting blood sugar mmol/Lt	5.8 ± 1.9	5.6 ± 1.7	6.3 ± 2.5	0.003
S. cholesterol mg/dl	205.7 ± 45.8	207.2 ± 47.2	208.7 ± 42.9	0.741
S. LDL mg/dl	126.6 ± 41.3	126.2 ± 38.8	127.6 ± 49.6	0.756
S. HDL mg/dl	37.4 ± 9.1	36.8 ± 9.4	38.3 ± 8.9	0.116
S. triglyceride mg/dl	224.6 ± 131.3	224.4 ± 124.0	238.3 ± 142.6	0.290
ALT u/l	56.1 ± 45.8	56.3 ± 59.2	55.1 ± 35.2	0.813
AST u/l	42.4 ± 38.8	41.0 ± 41.3	41.4 ± 26.4	0.920
GGT U/l	49.5 ± 37.4	46.5 ± 32.8	56.2 ± 48.1	0.030
ALP U/l	93.5 ± 49.1	89.3 ± 33.9	93.5 ± 37.5	0.366
HOMA-IR	2.0 ± 1.5	1.3 ± 0.4	3.5 ± 1.7	0.000

On comparison of the biochemical parameters, FBG (6.3 ± 2.5) mmol/l vs. (5.6 ± 1.7) mmol/l; P = 0.003 and GGT (56.2 ± 48.1) U/L vs (46.5 ± 32.8) U/l; P = 0.03 were significantly lower in Non-IR group. Other biochemical markers including lipid profile, ALT, AST and ALP differences were not statistically significant between the two groups.

Histopathology reports after liver biopsy were available in 275 patients. Of these, 181 were without IR while 94 had IR. On comparison of histology between patients with IR and without IR, lobular inflammation, ballooning and fibrosis were having no significant differences between the two groups. Only steatosis score was significantly higher in IR group (2.0 ± 0.7 vs 1.8 ± 0.8; P = 0.007). There was also significantly higher the mean NAFLD Activity Score (NAS) (4.9 ± 1.2 vs. 4.5 ± 1.4; P = 0.03) and NASH (NNFL/NASH) [28/66 vs. 87/94; P = 0.004] in IR group. Regarding fibrosis score, there was no significant difference between non IR and IR group (1.3 ± 0.6 vs 1.4 ± 0.8; P = 0.369) (Table 2).

Table 2.

Histological Difference of NAFLD Patients With or Without Insulin Resistance.

Variable	Total N = 275	Non IR N = 181	IR N = 94	P
Steatosis	1.9 ± 0.8	1.8 ± 0.8	2.0 ± 0.7	0.007
Ballooning	1.3 ± 0.5	1.3 ± 0.5	1.4 ± 0.6	0.145
Lobular inflammation	1.5 ± 0.6	1.5 ± 0.6	1.5 ± 0.6	0.636
NAS	4.6 ± 1.3	4.5 ± 1.4	4.9 ± 1.2	0.030
NNFL/NASH	115/160	87/94	28/66	0.004
Fibrosis	1.3 ± 0.7	1.3 ± 0.6	1.4 ± 0.8	0.369

In comparison of NNFL and NASH group, patients with higher Age (37.4 ± 8.8 vs 40.3 ± 9.8; P = 0.014), presence of DM (11/104 vs 108/52; P = 0.000), higher FBG (5.4 ± 1.7 vs 5.8 ± 1.9; P = 0.027) and presence of IR (28/87 vs 66/94; P = 0.004) had significantly higher chance of having NASH in univariate analysis. BMI, Waist Circumference and Metabolic syndrome had no significant differences between the groups in univariate analysis. In multivariate logistic regression analysis, presence of DM and IR [P = 0.003 and P = 0.041, respectively] were significantly associated with NASH. Among the 275 patients only 16 patients had advance fibrosis (F ≥ 3). Age, DM, MS, BMI, WC, IR and biochemical profile had no significant association of having advance fibrosis (F ≥ 3) or not (Table 3).

Table 3.

NASH, Fibrosis, Insulin Resistance and Associated Factors.

Variable	Univariate analysis for NASH			Multivariate analysis for NASH		Univariate analysis for fibrosis
	NNFL N = 115	NASH N = 160	P	Odds ratio	P	Fibrosis <3 N = 259	Fibrosis ≥3 N = 16	P
Sex male/female	44/71	52/108	0.323			91/168	5/11	0.752
Age in years	37.4 ± 8.8	40.3 ± 9.8	0.014	1.02	.173	38.9 ± 9.3	41.6 ± 12.3	0.408
Diabetes present/absent	11/104	108/52	0.000	.224	.003	57/202	5/11	0.396
Metabolic syndrome present/absent	77/38	118/42	0.221			185/76	10/6	0.445
BMI kg/m2	26.6 ± 4.3	27.3 ± 4.0	0.228			27.1 ± 4.1	26.7 ± 4.4	0.733
Waist circumference in cm	93.4 ± 9.0	95.4 ± 8.6	0.091			94.7 ± 8.9	94.1 ± 6.1	0.817
Fasting blood sugar mmol/Lt	5.4 ± 1.7	5.8 ± 1.9	0.027	.879	.190	5.6 ± 1.8	5.6 ± 1.9	0.859
Serum Triglyceride mg/dl	216.4 ± 110.7	236.4 ± 137.3	0.198			228.8 ± 128.1	215.4 ± 110.8	0.684
Insulin resistance present/absent	28/87	66/94	0.004	.540	.041	173/86	8/8	0.169

Discussion

In this study, the majority of the patients were female, which were similar with other studies conducted among the Bangladeshi population.^12,19 This female preponderance in our population might be due to less physical activity as a result of social conservative attitude which bounded most of the females to stay at home for household work leading to sedentary lifestyle and also due to more carbohydrate predominant food intake. Male NAFLD patients were predominantly in the non-IR. Similarly, another study from Bangladesh revealed that males were the predominant NAFLD patients in the nonobese.¹⁹ The mean BMI of the patients was 27.20 ± 3.9 kg/m² which is consistent with other studies from different parts of India and the West.^{33, 34, 35, 36} However, prevalence of IR in our study subjects was 34.1%, which is much lower as compared to other previous study.^15,21 Surprisingly, majority of the NAFLD patients (65.90%) were without IR, which might be due to genetic association (PNPLA3 polymorphism), food habit like consumption of high carbohydrate and food adulteration. Various genetic factors including PNPLA3 Polymorphism are known to confer susceptibility to NAFLD in individuals without increasing the level of IR.^37,38 An earlier study had shown that patients taking high carbohydrate had dyslipidemia especially hypertriglyceredemia³⁹; this may lead to fatty liver before development of IR.⁴⁰ Though HOMA is a commonly used method to assess insulin resistance and β-cell function and easy to estimate but it has got its own limitations. The HOMA has been correlated with the hyperinsulinemia and β-cell function, primarily in younger and middle-aged people. Previous studies have explored that HOMA-IR may not be a good predictor of insulin resistance in older people with diabetes.²⁴

There was significantly higher MS (78.1%), obesity (74.5%) and DM (39.0%) in patients with IR compared to patients without IR, supporting the central role of IR and MS in the pathogenesis of NAFLD in IR group.⁴¹ The findings were similar with other study.²¹ According to Kotronen et al., in non-diabetic individuals, liver fat content is 4 folds higher in subjects with MS compared to those without MS irrespective of age, gender and BMI.⁴² In patients with IR, this might be due to hyperglycemia, hyperinsulinemia, increased free fatty acid, high oxidative stress and altered profile of adipocytokines in hepatocytes which lead to development of steatosis and progression to steatohepatitis.⁴³ Waist circumference is markedly higher in the patients with IR, this indicates that central obesity is the major indicator of IR. A Korean study by Deok et al., explained that waist circumference contributes to insulin resistance (IR) and development of NASH.⁴⁴ In contrast with our study, another study had found a poor association of abdominal adiposity with NAFLD patients in Asians.⁴⁵

GGT level significantly higher in NAFLD patients with IR. In this study, differences among the lipid profile between the IR and non-IR group was not statistically significant, which is not consistent with another study Singh et al. from India.²¹

Among the biopsy findings regarding NAS, only steatosis was significantly more marked in the patients with IR. Similarly, the study Singh et al., showed that moderate fatty change/steatosis on ultrasonography was seen more commonly in patients with IR.²¹ In our study, it was revealed that the NAFLD patients with IR had significantly increased risk of having NASH. This scenario of a significant proportion of NAFLD and NASH patients with IR is in similar to the situation in developed countries, where a very high association of NASH with IR has been reported.⁴¹ Despite the absence of IR, proportion of patients with high grades of fibrosis were almost similar with patients with presence of IR. T The study Singh et al., found that a proportion of NAFLD had high grades of fibrosis, which is nearly similar with our study.²¹ Multifactorial pathogenesis of this disease, genetic factors, and/or specific dietary habits might be responsible for the development of NAFLD in non-IR individuals, even in the absence of metabolic disorders.^46,47

The limitation of our study is that this was a single-tertiary center study which may reflect the presence of advanced cases only. The higher cutoff for IR (HOMA-IR of 2) in our study could have increased the prevalence of non-IR NAFLD population. Most of our patients are of middle age group. Further to this we have not address issues like genetics and gut microbiota in the patients sans IR.

In conclusion, this is the first attempt to characterize patients of NAFLD without IR and establish this group as a distinct entity in Bangladesh. Nearly two-third of our NAFLD populations were without IR. NAFLD patients without IR predominantly male, had lower BMI, waist circumference, FBS and MS, were less diabetic and obese. Histologically NAFLD without IR equally severe with ballooning, lobular inflammation and fibrosis except steatosis which is much more in patients of NAFLD with IR.

So, insulin resistance is not the sole factor for NAFLD in our population, since only about one-third of the NAFLD patients had IR. NAFLD in our region is possibly a heterogeneous disorder, with a distinct group of NAFLD patients without IR. Further studies are necessary in this group of NAFLD patients to explore the role of other possible etiological factors: dietary habit, gut flora and genetic association in the development of NAFLD in the absence of IR.

Conflicts of interest

There is no conflict of interest.