Indian J Med Res 129, March 2009, pp 285-292

A case-control study on insulin resistance, metabolic co-variates &

prediction score in non-alcoholic fatty liver disease
S. Bajaj, P. Nigam*, A. Luthra**, R.M. Pandey+, D. Kondal+, S.P. Bhatt++, J.S. Wasir** & A. Misra*,**

Department of Medicine, Motilal Nehru Medical College, Allahabad, *Center for Diabetes, Obesity &
Cholesterol Disorders (C-DOC), Diabetes Foundation (India), **Department of Diabetes &
Metabolic Diseases, Fortis Hospital & Departments of +Biostatistics & ++Medicine
All India Institute of Medical Sciences, New Delhi, India

Received July 31, 2007

Background & objectives: Asian Indians have a high prevalence of insulin resistance and the metabolic
syndrome. Currently, non-alcoholic fatty liver disease (NAFLD) is considered to be an integral part of the
metabolic syndrome with insulin resistance as a central pathogenic factor. We studied anthropometric
parameters, insulin resistance and metabolic co-variates in subjects with NAFLD as compared to those
without NAFLD, and also developed a prediction score for NAFLD.
Methods: Thirty nine subjects with NAFLD and 82 controls were selected for the study after
ultrasonography of 121 consecutive apparently healthy subjects. Anthropometric profile [body mass
index (BMI), waist circumference (WC) etc,], lipid profile, hepatic aminotransferases, fasting blood
glucose (FBG), insulin were recorded and value of homeostasis model assessment of insulin resistance
(HOMA-IR) was analysed. Step-wise logistic regression analysis and area under the receiver operator
curve (aROC) were analysed to arrive at a prediction score.
Results: Overall, prevalence of NAFLD was 32.2 per cent and prevalence of metabolic syndrome was
seen in 41 per cent among cases and 19.5 per cent in controls (P<0.01). Subjects with NAFLD had
significantly higher values of BMI, WC, hip circumference, FBG, fasting insulin, total cholesterol and
serum triglycerides. Step-wise logistic regression analysis showed odds ratio (OR) and 95 per cent
confidence interval (CI) for BMI [ 4.3 (1.6, 11.3)], FBG [5.5 (1.5, 19.8)] and fasting insulin [ 2.4 (1.0, 5.8)]
as independent predictors of NAFLD. The prediction score for NAFLD was; 1 (fasting insulin) +1.6
(BMI) + 1.9 (FBG) (sensitivity of 84.6%, specificity of 51.2% and aROC 76%).
Interpretation & conclusion: In this study, presence of NAFLD indicated close relationship with multiple
features of metabolic syndrome. The prediction score developed could be used as a screening tool to
predict NAFLD among Asian Indians in north India.
Key words Insulin resistance - metabolic syndrome - non-alcoholic fatty liver disease - north Asian Indians

Non-alcoholic fatty liver disease (NAFLD) is now

believed to be an integral part of the metabolic syndrome,
which comprises a cluster of abnormalities (dysglycaemia,
dyslipidaemia, hypertension, procoagulant tendency, etc.)

with insulin resistance as a central pathogenic factor1,2.

Several investigators have reported that NAFLD is
significantly associated with reduced biological effects of
insulin3,4. In this scenario, it is suspected that lipotoxicity
285

286

INDIAN J MED RES, MARCH 2009

could be more important than organ-specific insulin

resistance as the predominant abnormality. It has been
opined that there is some reduction in hepatic insulin
sensitivity due to triglyceride accumulation in liver
as lipids overflow from adipocytes into liver5. This
could also mean that suppression of insulin sensitivity
is a secondary phenomenon in NAFLD, as opposed to
a primary pathophysiological event. More importantly,
NAFLD has been shown to be independently related
to insulin resistance (as assessed by hyperinsulinaemiceuglycaemic clamp technique) independent of obesity
and abdominal adiposity6,7.
The data on the prevalence of fatty liver in Asian
populations are limited. Chitturi et al8 highlighted the
potential burden of the disease in the Asian-Pacific area;
1.8 million Asians and at least 400,000 Australians had
fatty liver, thus eclipsing the disease burden of hepatitis
B and C. An autopsy study from western India and data
from coastal eastern India showed prevalence of fatty
liver of 15.8 and 24 per cent, respectively9,10 .
The prevalence of the metabolic syndrome is
rapidly increasing globally11,12. The data from National
Health and Nutrition Examination Survey (1988-1994)
indicated the prevalence of metabolic syndrome to be
in the range of 24 to 44 per cent in US population11.
Several studies in India indicated the prevalence of
the insulin resistance and the metabolic syndrome
ranging from about 11 to 41 per cent13 depending
on the region and urbanization. A recent study
showed that almost 1/3rd of the urban population in
major metropolitan cities in India had the metabolic
syndrome14. The central pathogenic feature of the
metabolic syndrome, insulin resistance is also widely
prevalent in Asian Indians and is of higher magnitude
than white Caucasians7. However, inter-relationships
of NAFLD, insulin resistance and the metabolic
syndrome have been sparsely studied in Asian Indians.
In the only study from north India, 38/39 patients
showed insulin resistance as assessed by homeostasis
model assessment of insulin resistance (HOMA-IR),
and 100 per cent patients (n=54) had the metabolic
syndrome. However, these subjects were not assessed
for detailed anthropometry and were not compared to
controls without NAFLD15.
We hypothesized that subjects with NAFLD would
be more obese, would have fasting hyperinsulinaemia,
higher value of fasting blood glucose and greater
degree of insulin resistance. To test this hypothesis,
we studied anthropometric and metabolic profiles

including fasting insulin levels and analysed the

HOMA-IR value in subjects with NAFLD and in
healthy controls without NAFLD. Using these data, we
also aimed to develop a prediction score for NAFLD.
Material & Methods
Subjects were selected from those attending
the medical outpatient department of Motilal Nehru
Medical College, Allahabad, in response to local
advertisement during the period February 2005 to
June 2005. A total of 160 apparently healthy subjects
were randomly enrolled which were voluntarily
participated in response to local advertisement, out of
which 39 patients were excluded based on significant
alcohol consumption (>20 g/day). Subjects with
known diabetes, cardiovascular disease (CVD) and
any chronic illness like renal failure, tuberculosis,
and advanced liver diseases were excluded from the
study. All subjects were assessed for demographic and
socio-economic profiles, smoking and physical activity
pattern. A family history of type 2 diabetes mellitus
(T2DM), overweight, hypertension and CVD was
obtained. A written informed consent was obtained
from the patients. Approval for conducting the study
was obtained from the institutional ethics committee of
Motilal Nehru Medical College, Allahabad.
For measuring weight, subject was instructed to
stand still in the platform, with the body weight evenly
distributed between both the feet. After removing heavy
clothing weight (Seca 803, digital scale, Germany) was
measured to the nearest of 0.1 kg. Height was measured
using stadiometer (Seca 206, Germany) with head held
in Frankfort plane to the nearest of 0.1 cm. Body
mass index (BMI) was calculated by the following
formula; weight (kg)/height (m2). Waist circumference
(WC) was measured mid-way between iliac crest and
lowermost margin of the ribs, in quiet breathing. Hip
circumference (HC) was measured at the maximum
protruding part of buttocks at the level of the greater
trochanter with the patient wearing minimal clothing
and feet together. Mid-thigh circumference was taken
at the point in anterior midline of the thigh, midway
between the inguinal ligament and base of patella to
the nearest of 0.1mm.
Pulse rate was recorded after 5 min of rest. Blood
pressure was also recorded after at least 5 min of rest
in a chair, with feet on the floor, and arm supported at
heart level, using a mercury sphygmomanometer. An
appropriate-sized cuff (cuff bladder encircling at least
80% of the arm) was used to ensure accuracy. Systolic

BAJAJ et al: NAFLD & INSULIN RESISTANCE IN NORTH INDIANS

blood pressure was measured at the point where the

first of two or more sounds was heard (phase 1), and
diastolic blood pressure before the disappearance of
sounds (phase 5).
Biochemical analysis: A fasting venous blood sample
(15 ml) was obtained for metabolic profile. Estimations
for fasting blood glucose (FBG), total cholesterol (TC),
serum triglycerides (TG) and high-density lipoprotein
cholesterol (HDL-C) were performed using commercial
kits (Randox Laboratory, San Francisco, CA, USA) on
a automated analyzer (Micro Semi-Autoanalyser 2000,
C.L. Micromed Italy). Value of low-density lipoprotein
cholesterol (LDL-C) was calculated using Friedwalds
equation16.
Serum insulin assay: Serum fasting insulin was
determined using a commercially available
radioimmunoassay kit (Immunotech, A Beckman
Coulter Company, Insulin (e) IRMA kit, Canada).
The principle of this assay was based on competitive
binding of labelled and unlabelled insulin to the
binding sites of anti-insulin antibodies immobilized on
inner wall of the tube. The radioactivity of iodinated
insulin bound to the anti-insulin antibodies on solid
phase was measured using a gamma counter. The intraassay and inter-assay percentage coefficient variables
were 1.95 and 2.23 per cent, respectively. The lower
limit of detection is 0.01 U/ml. Reference range of
the assay was 2.1 to 22 U/ml. Insulin resistance was
measured by two surrogate measures: fasting insulin
and homeostasis model assessment (HOMA). The
value of HOMA was calculated by the following
equation17: (fasting insulin (U/ml) X fasting glucose
(mmol/l))/22.5 and depicted as HOMA-IR value.
Ultrasound imaging: Abdominal ultrasound was
carried out using 3.5 MHz curvilinear probe (SiemensG 60 S 2004, Germany) by a trained operator who was
blinded to all clinical and laboratory characteristics of
participants. Both subcostal and intercostal scanning
was done. Normal liver parenchyma was seen as
solid homogenous echo texture, which was midway
between the renal cortex and pancreatic echogenicity.
The findings of hepatic steatosis at sonography include
increased echogenicity and sound attenuation. The
grading of NAFLD (Table I) was based on the criteria
of Brunt et al 18.
Definitions: Overweight and obesity were defined as
BMI 23-24.9 kg/m2 and BMI 25 kg/m2 respectively19.
Waist circumference >90 cm for males and >80 cm for
females was considered an indicator of abdominal

287

Table I. Grading of NAFLD

Grading of macrovesicular steatosis:
Grade 0: None
Grade 1: Up to 33%
Grade 2: 33-66%
Grade 3: >66%
Grading of nonalcoholic steatohepatitis:
Mild (Grade I) Steatosis- up to 66%
Balloned hepatocytes-occasional (zone 3)
Scattered intra-acinar neutrophils lymphocytes
Portal inflammation none or mild
Moderate
Steatosis- any degree
(Grade II)
Balloned hepatocytes-obvious (zone 3)
Intra- acinar neutrophils ++ lymphocytes +
Portal inflammation- mild to moderate
Perisinusoidal zone III fibrosis
Severe
Steatosis- panacinar
(Grade III) Balloned hepatocytes- widespread
Intra-acinar neutrophils ++(associated with
ballooned hepatocytes) chronic inflammation
Portal inflammation- mild to moderate
Source: Ref. 17

obesity20. T2DM and impaired fasting glucose were

diagnosed according to the diagnostic criteria of
the American Diabetes Association21. The modified
criteria of National Cholesterol Education Program,
Adult Treatment Panel III (NCEP, ATP III) were used
to define the metabolic syndrome; waist circumference,
males >90 cm, females >80 cm, fasting blood glucose
>100 mg/dl, serum TG >150 mg/dl, blood pressure
>130/85 mmHg and HDL-C; males <40 mg/dl, and
females <50 mg/dl22. Fasting hyperinsulinaemia was
defined as fasting insulin 2.7 U/ml23 and physical
activity as >20 min brisk walk per day. Any degree of
current smoking was taken for defining smoking.
Statistical analysis: Data were presented as either mean
SD or median (range) as appropriate. The differences
in mean values of the variables between cases and
controls were tested using student t-test and t-test
with adjustment for variables with unequal variances
(Welchs test). Difference between proportions was
tested using Chi-square test. Wilcoxon rank-sum test
was used to detect the differences between cases and
controls for the fasting insulin and HOMA values.
A bivariate analysis followed by multiple logistic
regression was carried out to identify the independent
predictors of NAFLD considering age, BMI, WC, blood
pressure, serum TG, HDL-C, fasting blood glucose
(FBG) and fasting insulin as risk factors and to estimate
odds ratio (OR) and 95 per cent confidence interval
(95% CI).To develop the prediction score, weights

288

INDIAN J MED RES, MARCH 2009

for the predictor variables were obtained by dividing

each of -coefficient by the lowest -coefficient. The
prediction score was obtained as a sum of weight of
each variables multiplied by the variables in the model.
The receiver operator characteristics (ROC) curve was
drawn between sum of these weighted coefficients and
the NAFLD to determine the appropriate cut off value
of the prediction score. The area under ROC curve
(AUC) with its 95 per cent confidence interval (CCI)
was also estimated. STATA 9.0 (STATA Corporation,
College Station, TX, USA) was used for the statistical
analysis. P<0.05 was considered as significant.
Results
Complete clinical profile, anthropometric and
ultrasonography data and serum samples were
available for 121 subjects. Among these, 39 subjects
who were diagnosed to have fatty liver classified as
cases while 82 subjects who had normal abdominal
ultrasonography were taken as controls. The monthly
income of the majority of cases and controls was in the
Table II. Demographic and lifestyle profiles of cases and controls
Variables
Sex (Male), n (%)
Per capita income
(INR*/mo.), n (%)
<1000
1001-5000
5001-10,000
>10,001
Education, n (%)
Illiterate
Primary (class 1st to 9th)
Intermediate
(class 10th-12th )
Graduate
Post graduate
Living standard habitat, n (%)
Rural
Semi-urban
Urban
Family history, n (%)
Type 2 diabetes mellitus
Hypertension
Overweight
Cardiovascular disease
Physical activity, n (%)
Never
Once a week
Two-three times in a week
Daily
Smoking, n (%)
*Indian National Rupees

Cases (n=39)
20 (51)

Controls (n=82)
54 (65.9)

14 (35.9)
18 (46.2)
3 (7.7)
4 (10.3)

33 (40.2)
43 (52.4)
3 (3.7)
3 (3.7)

2 (5.1)
3 (7.6)
12 (30.7)

5 (6.0)
6 (7.3)
32 (39.0)

20 (51.3)
2 (5.1)

35 (42.7)
4 (4.9)

2 (5.1)
3 (7.7)
34 (87.2)

14 (17.3)
5 (6.2)
62 (76.5)

14 (36.8)
9 (23.7)
9 (23.1)
3 (7.7)

22 (27.2)
13 (16.3)
14 (17.1)
3 (3.7)

19 (48.8)
13 (7.7)
4 (10.3)
13 (33.3)
8 (20.5)

46 (56.8)
6 (7.4)
9 (11.1)
20 (24.7)
14 (17.0)

range of 1001-5000 INR. Most of the cases (87.2%)

and controls (76.5%) belonged to urban areas. Definite
family history of T2DM, hypertension, overweight and
CVD was more prevalent in cases then controls but not
statistically significant. A total of 48.8 per cent of cases
and 56.8 per cent controls were defined as physically
inactive (Table II).
Significantly higher values of BMI (P<0.001), WC
(P<0.01), HC (P<0.001), FBG (P<0.01), TC (P<0.001)
and TG (P<0.001) were recorded in cases than in
controls. Although, value of HOMA-IR and fasting
insulin levels were also high in cases as compared to
controls, the difference was not statistically significant.
Among the 39 patients with NAFLD 15.4 per cent
were overweight and 66.7 per cent were obese while
in 82 controls, 23.2 per cent were overweight and 24.4
Table III. Clinical and biochemical profiles of cases and controls
Variable

Cases (n=39)
(Mean SD)

Age (yr)
40.9 11.1*
Systolic blood pressure (mmHg)
128.2 17.4
Diastolic blood pressure (mmHg)
83.2 12.3
Pulse rate ( per min)
81.5 10.1
Body mass index (kg/m2)
26.7 4.4***
Waist circumference (cm)
89.2 13.9**
Hip circumference (cm)
98.3 1.9***
Mid-thigh circumference (cm)
48.0 10.7
Fasting blood glucose (mg/dl)
96.6 24.0**
Fasting insulin (U/ml) median
2.6 (0.01,19.6)
(range)
Total cholesterol (mg/dl)
176.4 40.9***
Serum triglycerides (mg/dl)
136.4 68.9**
HDL-C (mg/dl)
42.6 8.7
Aspartate aminotransferase (U/l)
33.3 17.3
Alanine aminotransferase (U/l)
42.3 4.7
HOMA-IR median (range)
0.6 (0, 6.1)
Overweight (BMI 23-24.9 kg/m2)
6 (15.4)
No. (%)
Obesity (BMI 25 kg/m2)
26 (66.7)***
No. (%)
The metabolic syndrome No.,
16 (41)**
n (%)
Impaired fasting glucose No.,
9 (23.1)**
n (%)

Controls
(n=82)
(Mean SD)
35.0 13.4
126.2 14.9
81.6 9.8
79.2 12.5
22.7 3.9
80.6 18.8
89.4 1.5
45.9 6.6
84.9 8.2
1.3 (0.01,9.5)
153.2 32.7
104.2 45.9
41.9 8.0
30.7 19.9
34.8 36.8
0. 9 (0, 2.2)
19 (23.2)
20 (24.4)
16 (19.5)
4 (4.9)

Values are mean SD

HDL-C, High-density lipoprotein cholesterol; HOMA-IR, Insulin
resistance value calculated by Homeostasis model of assessment;
Numbers (percentage); According to the modified criteria of
National Cholesterol Education Program, Adult Treatment Panel III
(NCEP, ATP III); Two cases (5.1%) and none of the controls had
type 2 diabetes mellitus (T2DM)
P *<0.05, **<0.01, ***<0.001 compared to controls

289

BAJAJ et al: NAFLD & INSULIN RESISTANCE IN NORTH INDIANS

Table IV. Characteristics of metabolic syndrome in cases and controls
Variables
BMI (kg/m2)
23
< 23
Waist circumference (cm)
Action Level-II
>80 for females, 90 for males
80 for females, 90 for males
Serum triglycerides (mg/dl)
>150
150
Fasting blood glucose (mg/dl)
>100
100
HDL-C (mg/dl)
>40 for males, >50 for females
40 for males, 50 for females
Blood pressure (mmHg)
>130 SBP, 85 DBP
130 SBP, 85 DBP
Fasting insulin (U/ml)
2.7
<2.7

Cases
n (%)

Controls
n (%)

P value

Unadjusted
OR (95% CI)

32 (82.1)
7 (17.9)

39 (47.6)
43 (52.4)

0.000

5.0 (1.9,12.7)

23 (58.9)
16(41)

22 (26.8)
60(73.2)

0.001

3.9 (1.76, 8.76)

9(23.1)
30(76.9)

14(17.1)
68 (82.9)

0.431

1.4 (0.57,3.73)

11(28.2)
28(71.8)

4 (4.9)
78(95.1)

0.001

7.7 (2.25, 26.02)

26(66.7)
13(33.3)

53(64.6)
29(35.8)

0.826

1.1 (0.49, 2.45)

19 (48.72)
20 (51.28)

32 (39.02)
50(60.98)

0.313

1.5 (0.67,3.20)

18(46.15)
21(53.85)

23(28.05)
59(71.95)

0.05

2.2 (0.99,4.86)

SBP, Systolic blood pressure; DBP, diastolic blood pressure

Source: Ref. 22

per cent were obese. The metabolic syndrome was

more prevalent in cases as compare to controls (41 vs.
19.5%, P=0.01). Impaired fasting glucose was present
in 9 (23.1%) cases and 4 (4.9%) controls (Table III).
Two cases (5.1%) and none of the controls had T2DM.
Fasting plasma glucose was repeated twice to confirm
the status of T2DM in these two cases.
On bivariate analysis, BMI was found to be
more in the cases (82.1%) than the controls (47.6%,
P=0.001). The odds ratio and 95 per cent confidence
interval [(OR) 95% CI] for BMI [5.0 (1.9, 12.7)], WC
[3.9 (1.76, 8.76)] and FBG [7.7 (2.25, 26.02)] showed
significant association with NAFLD (Table IV). On
multiple logistic regression analysis, the OR (95%
CI) for BMI, FBG and fasting insulin were [4.3 (1.6,
11.3)], [5.4 (1.5, 19.8)] and [2.4(1.0-5.8)] respectively
(Table V). These variables were independent predictors
of NAFLD.
The median (range) of FBG (mg/dl) for cases [96.6
(74-221)] was comparatively higher than controls [84.9
(71-116)] (P=0.0001). Similarly, range and median of
fasting insulin (U/ml) were comparatively higher in
cases [3.5 (0.01-19.6)] than controls [2.3 (0.01-9.5)]
(P=0.03). The range and median for HOMA-IR values

in cases [0.9 (0.01-6.1)] was higher than controls but not

statistically significant [0.5 (0.01-2.2)] (P=0.06) (Fig.).
Weighted coefficients of BMI, FBG and fasting
insulin were used to generate a prediction score.
To determine the appropriate cut-off value of the
prediction score, area under the receiver operator curve
(aROC) was drawn between sums of these weighted
coefficients and the NAFLD (Table V).
Prediction score for NAFLD = 1 (fasting insulin)
+1.6 (body mass index) + 1.9 (fasting blood glucose)
(maximum score = 4.5, minimum score = 0). Fasting
Table V. Stepwise logistic regression model for analysis of
predictors of NAFLD
Variables

Body mass index

Fasting blood glucose
Fasting insulin

Logistic regression model

OR (95% CI)

Coefficient

4.3 (1.6,11.3)
5.4 (1.5,19.8)
2.4 (1.0,5.8)

1.45
1.69
0.88

Weighted
coefficient
1.6
1.9
1

Prediction score for NAFLD= 1 (fasting insulin) +1.6 (body mass

index) +1.9 (fasting blood glucose)
Maximum score= 4.5, minimum score= 0

290

INDIAN J MED RES, MARCH 2009

(a)

Insulin: 1: (>2.7 U/ml); 0: otherwise; body mass

Index: 1: ( 23 kg/m2); 0, otherwise; fasting blood
glucose: 1 (>100 mg/dl); 0, otherwise.
Using aROC analysis, a score cut-off of 1.6
ensured the best balance between sensitivity and
specificity. The sensitivity (95% CI); specificity
(95%CI) and aROC (95% CI) were 84.6 (69.5, 94.1),
51.2 (39.9, 62.4) and 0.76 (0.7, 0.9). The probability of
subject being NAFLD was more if score was 1.6.
Discussion
In this study, the metabolic syndrome was found to
be associated with NAFLD in apparently healthy north
Indians. Further, a prediction score was developed to
predict NAFLD for adult north Indians.

(b)

(c)

Fig. Box plot representation of fasting insulin levels (a), fasting

blood glucose (b), HOMA-IR (c) in subjects with non-alcoholic
fatty liver disease and in controls. Each box comprises the values
between the 25th and the 75th percentiles, and the bold horizontal
line is the median value; the whiskers stretch from the 10th and
to the 90th percentile. Circles represent individual outliers value.
Stars represent extremes value of individual.

Association between the metabolic syndrome and

NAFLD has been shown24,25, and many researchers
now consider NAFLD to be a manifestation of the
metabolic syndrome26. Earlier, we had emphasized that
the prevalence of NAFLD in India would approximate
prevalence of the metabolic syndrome since most of the
metabolic co-variates of NAFLD are highly prevalent
in Asian Indians27. As evidenced by other investigators
in India, profile of various chronic diseases remains
largely similar13,14,28-31. Characteristic phenotype
makes Asian Indians more prone to diabetes and other
metabolic diseases.
Three clear independent risk factors for NAFLD
emerged in our study; BMI, fasting blood glucose
levels and fasting insulin levels. These findings are in
line with the observations in other ethnic groups32,33.
By using proton magnetic resonance spectroscopy
(MRS-P), Shulman et al34 measured hepatic
triglyceride (HTG) content and plasma interleukin-6
(IL-6) concentrations in different ethnic groups in
USA. Interestingly, these authors reported that the
HTG content and plasma IL-6 concentrations were
nearly 2-fold higher in Asian Indians as compared
to white Caucasians. The increased HTG content
and IL-6 concentrations were associated with higher
prevalence of insulin resistance in Asian Indians (59%)
as compared to white Caucasians (20%). It is likely that
such ethnic differences in HTG content exist in other
ethnic groups as well, but remain to be proved. The
reason(s) of such ethnic differences are not clear, but
may relate to genetic and metabolic factors or dietary
intake.
Interestingly, abdominal adiposity was not shown
to be an independent predictor of NAFLD in our study,

BAJAJ et al: NAFLD & INSULIN RESISTANCE IN NORTH INDIANS

291

though it was more prevalent in those with NAFLD.

These data are in contrast to that of Fan et al35, who
found the strongest association of central (WC) and
overall obesity (BMI) with the metabolic syndrome in
Asia Pacific region. However, the inter-relationship of
abdominal obesity and NAFLD in adult Asian Indians
need to be further researched. Further, it is important to
find out correlation of various regional adipose tissue
compartments (truncal subcutaneous, intra-abdominal),
which are important determinants of insulin sensitivity
in Asian Indians, with NAFLD13,36,37.

predictive of NAFLD. The prediction score developed

can be used as screening tool for NAFLD. Further
studies are needed to identify simple prediction tools
for NAFLD, and whether NAFLD could predict
metabolic perturbations (e.g., diabetes) in Asian
Indians.
1.

Meigs JB. Insulin resistance syndrome? Syndrome X?

Multiple metabolic syndrome? A syndrome at all? Factor
analysis reveals patterns in the fabric of correlated metabolic
risk factors. Am J Epidemiol 2000; 152 : 908-11.

Marchesini et al3 have reported that insulin

resistance as assessed by HOMA-IR to be higher in
NAFLD cases as compared to controls largely due to
increased insulin concentration with normal or nearnormal glucose levels. However, our data indicated
that insulin resistance was higher primarily due to
higher fasting glucose levels. Further, this is in line
with our finding that fasting blood glucose level was
an independent predictor of NAFLD. This discrepancy
between our data and that of Marchesini et al 3 indicates
that the clinicopathological profile of Indian NAFLD
patients may be somewhat different from that seen in
other ethnic groups15.

2.

Marchesini G, Bugianesi E, Forlani G, Cerrelli F, Lenzi M,

Manini R, et al. Nonalcoholic fatty liver, steatohepatitis, and
the metabolic syndrome. Hepatology 2003; 37 : 917-23.

3.

Marchesini G, Brizi M, Morselli-Labate AM, Bianchi

G, Bugianesi E, McCullough AJ, et al. Association of
nonalcoholic fatty liver disease with insulin resistance. Am J
Med 1999; 107 : 450-5.

4.

Angelico F, Del Ben M, Conti R, Francioso S, Feole K,

Fiorello S, et al. Insulin resistance, the metabolic syndrome,
and nonalcoholic fatty liver disease. J Clin Endocrinol Metab
2005; 90 : 1578-82.

5.

DeFronzo RA. Dysfunctional fat cells, lipotoxicity and type 2

diabetes. Int J Clin Pract 2004; 58 (Suppl 143): 9-21.

6.

Seppl-Lindroos A, Vehkavaara S, Hkkinen AM, Goto T,

Westerbacka J, Sovijrvi A, et al. Fat accumulation in the
liver is associated with defects in insulin suppression of
glucose production and serum free fatty acids independent of
obesity in normal men. J Clin Endocrinol Metab 2002; 87 :
3023-8.

7.

Garg A, Misra A. Hepatic steatosis, insulin resistance and

adipose tissue disorders. J Clin Endocrinol Metab 2002; 87 :
3019-22.

8.

Chitturi S, Farrell GC, George J. Non-alcoholic steatohepatitis

in the Asia-Pacific region: future shock? J Gastroenterol
Hepatol 2004; 19 : 368-74.

9.

Amarapurkar A, Ghansar T. Fatty liver: experience from

western India. Ann Hepatol 2007; 6 : 37-40

How should we identify patients with NAFLD?

A simple clinical dictum should be that obese patient
would have high chances of having NAFLD. We report
here a refinement in prediction of NAFLD in north
Indians by including, besides BMI, fasting insulin and
fasting blood glucose levels. It is possible that these
variables could be replaced by more simple variables,
however, this issue needs further data and analysis in
future studies, so that sensitivity and specificity profile
for prediction of NAFLD increases and it becomes
cost-effective to apply this prediction score in general
population in a developing country like India. Further,
before applying it to the general population in a random
fashion, it should be studied in a large cohort.

References

10. Singh SP, Navak S, Swain M, Rout N, Mallik RN, Agrawal O,

et al. Prevalence of nonalcoholic fatty liver disease in coastal
eastern India: a preliminary ultrasonographic survey. Trop
Gasteroenterol 2004; 25 : 76-9.

A limitation of our study was that the diagnosis

of NAFLD was based on liver ultrasonography. It has
been argued that other methods; magnetic resonance
spectroscopy, and liver biopsy are better tools for
defining NAFLD, and could be considered as gold
standard.

11. Ford ES, Gile WH, Dietz WH. Prevalence of the metabolic
syndrome among US adults: findings from the third National
Health and Nutrition Examination Survey. JAMA 2002; 287 :
356-9.

In summary, our findings showed important

clinical (BMI) and metabolic (fasting insulin, and
fasting blood glucose levels) predictors of NAFLD
in north Indians. It is important to note that variability
in the anthropometric and biochemical parameters is

13. Misra A, Vikram NK. Insulin resistance syndrome (metabolic

syndrome) and Asian Indians. Curr Sci 2002; 83 : 1483-96.

12. Cameron AJ, Shaw JE, Zimmet PZ. The metabolic syndrome:
prevalence in worldwide populations. Endocrinol Metab Clin
North Am 2004; 33 : 351-75.

14. Vikram NK, Misra A, Pandey RM, Dwivedi M, Luthra

K, Dhingra V, et al. Association between subclinical
inflammation & fasting insulin in urban young adult north
Indian males. Indian J Med Res 2006; 124 : 677-82.

292

INDIAN J MED RES, MARCH 2009

15. Duseja A, Das A, Das R, Dhiman RK,Chawla Y, Bhansali A,

et al. The clinicopathological profile of Indian patients with
nonalcoholic fatty liver disease (NAFLD) is different from
that in the West. Dig Dis Sci 2007; 52 : 2368-74.
16. Friedewald WT, Levy RI, Fredrikson DS. Estimation of the
concentration of low-density lipoprotein cholesterol in plasma
without use of the preparative ultracentrifuge. Clin Chem
1972; 18 : 499-502.
17. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher
DF, Turner RC. Homeostasis model assessment: insulin
resistance and -cell function from fasting plasma glucose
and insulin concentrations in man. Diabetologia 1985; 28 :
412-9.

26. Angulo P. GI epidemiology: nonalcoholic fatty liver disease.

Aliment Pharmacol Ther 2007; 25 : 883-9.
27. Misra A, Misra R, Wijesuriya M, Banerjee D. The metabolic
syndrome in South Asians: continuing escalation & possible
solutions. Indian J Med Res 2007; 125 : 345-54.
28. Mohan V, Sandeep S, Deepa R, Shah B, Varghese C.
Epidemiology of type 2 diabetes: Indian scenario. Indian J
Med Res 2007; 125 : 217-30.
29. Gupta A, Gupta R, Sarna M, Rastogi S, Gupta VP, Kothari K.
Prevalence of diabetes, impaired fasting glucose and insulin
resistance syndrome in an urban Indian population. Diabetes
Res Clin Pract 2003; 61 : 69-76.

18. Brunt EM, Janney CG, Di Bisceglie AM, NeuschwanderTetri BA, Bacon BR. Nonalcoholic steatohepatitis: a proposal
for grading and staging the histological lesions. Am J
Gastroenterol 1999; 94 : 2467-74.

30. Prabhakaran D, Chaturvedi V, Ramakrishnan L, Jeemon P,

Shah P, Snehi U, et al. Risk factors related to the development
of diabetes in men working in a north Indian industry. Natl
Med J India 2007; 20 : 4-10.

19. World Health Organization Western Pacific Region/

International Assoc for the Study of Obesity/Int Obesity Task
Force. The Asia-Pacific perspective: redefining obesity and
its treatment. Health Communications Australia: Melbourne,
2000.
Available
from:
http://www.wpro.who.int/NR/
rdonlyres/0A35147B-B1D5-45A6-9FF2-F7D86608A4DE/0/
Redefiningobesity.pdf.

31. Yajnik CS, Lubree HG, Rege SS, Naik SS, Deshpande JA,
Deshpande SS, et al. Adiposity and hyperinsulinemia in
Indians are present at birth. J Clin Endocrinol Metab 2002;
87 : 5575-80.

20. Misra A, Vikram NK, Gupta R, Pandey RM, Wasir JS, Gupta
VP. Waist circumference cut-off points and action levels for
Asian Indians for identification of abdominal obesity. Int J
Obes (Lond) 2006; 30 : 106-11.
21. Expert Committee on the Diagnosis and Classification of
Diabetes Mellitus; Report of the expert committee on the
diagnosis and classification of diabetes mellitus. Diabetes
Care 2003; 26 (Suppl 1): S5-20.
22. Grundy SM, Cleeman JI, Daniels SR, Donato KA, Eckel RH,
Franklin BA, et al. Diagnosis and management of the metabolic
syndrome: An American Heart Association/National Heart,
Lung, and Blood Institute Scientific Statement. Circulation
2005; 112 : 2735-52.
23. Ramachandran A, Snehalatha C, Yamuna A, Murugesan
N, Narayan KM. Insulin resistance and clustering of
cardiometabolic risk factors in urban teenagers in southern
India. Diabetes Care 2007; 30 : 1828-33.
24. Bellentani S, Saccoccio G, Masutti F, Croce S, Brandi G,Sasso
F, et al. Prevalence of and risk factors for hepatic steatosis in
Northern Italy. Ann Intern Med 2000; 132 : 112-7.
25. Gupte P, Amarapurkar D, Agal S, Baijal R, Kulshrestha P,
Pramanik S, et al. Non-alcoholic steatohepatitis in type 2
diabetes mellitus. J Gastroenterol Hepatol 2004; 19 : 854-8.

32. Gambarin-Gelwan M, Kinkhabwala SV, Schiano TD, Bodian

C, Yeh HC, Futterweit W. Prevalence of nonalcoholic fatty
liver disease in women with polycystic ovary syndrome. Clin
Gastroenterol Hepatol 2007; 5 : 496-501.
33. Oh SY, Cho YK, Kang MS, Yoo TW, Park JH, Kim HJ, et al.
The association between increased alanine aminotransferase
activity and metabolic factors in nonalcoholic fatty liver
disease. Metabolism 2006; 55 : 1604-9.
34. Petersen KF, Dufour S, Feng J, Befroy D, Dziura J, Dalla
Man C, et al. Increased prevalence of insulin resistance and
nonalcoholic fatty liver disease in Asian-Indian men. Proc
Natl Acad Sci USA 2006; 103 : 18273-7.
35. Fan JG, Saibara T, Chitturi S, Kim BI, Sung JJ, Chutaputti
A, Asia Pacific Working Party at NAFLD. What are the risk
factors and settings for non-alcoholic fatty liver disease in
Asia-Pacific?.J gasteroenterol Hepatol 2007; 22 : 794-800.
36. Misra A, Vikram NK. Clinical and pathophysiological
consequences of abdominal adiposity and abdominal adipose
tissue depots. Nutrition 2003; 19 : 457-66.
37. Misra A, Vikram NK, Arya S, Pandey RM, Dhingra V,
Chatterjee A, et al. High prevalence of insulin resistance
in postpubertal Asian Indian children is associated with
adverse truncal body fat patterning, abdominal adiposity and
excess body fat. Int J Obes Relat Metab Disord 2004; 28 :
1217-26.

Reprint requests: Dr Anoop Misra, Director & Head, Department of Diabetes & Metabolic Diseases Fortis Flt. Lt. Rajan Dhall Hospital
Vasant Kunj, New Delhi 110070, India
e-mail: anoopmisra@metabolicresearchindia.com