Full text of DOI:10.1007/s11745-001-0822-5

Postprandial Triglyceride-Rich Lipoprotein
Metabolism and Insulin Sensitivity
in Nonalcoholic Steatohepatitis Patients
Maurizio Cassadera,*, Roberto Gambinoa, Giovanni Mussoa, Nadia Depetrisa,
Fabio Meccaa, Paolo Cavallo-Perina, Giovanni Pacinib, Mario Rizzettoc,
and Gianfranco Paganoa
aDepartment of Internal Medicine, University of Turin, Turin, Italy, bInstitute of Systems Science and Biomedical Engineering
(LADSEB-CNR), Padua, Italy, and cDepartment of Gastroenterology, University of Turin, Corso Dogliotti 14, Turin, Italy
ABSTRACT: Nonalcoholic steatohepatitis (NASH) is a syndrome lipid metabolism, in accordance with the high prevalence of the
frequently associated with obesity, diabetes mellitus, and dys- metabolic syndrome in NASH patients. This paper suggests that
lipidemia. Increased fasting insulinemia and blood glucose lev- a fat load may be useful in early detection of atherogenic risk in
els may trigger a reduced catabolism of lipoproteins rich in the presence of otherwise normal fasting plasma lipids.
triglycerides by lipoprotein lipase (LPL) and an increase in their
fasting and postprandial levels. An association between post-
prandial lipemia and coronary heart disease has been observed,
and many studies now support this concept. The most impor-
tant result of our study is the increase in triglyceride-rich
lipoproteins response after a fat load in NASH patients, the in-
crease of incremental area under the postprandial curve, and
the duration of the hypertriglyceridemic peaks. The persisting
postprandial plasma triglyceride elevation in NASH patients
was mostly due to the elevated plasma level of large triglyc-
eride-rich particles. These data are coupled with lower plasma
HDL2-cholesterol levels. As for lipoprotein analyses, the num-
Paper no. L8603 in Lipids 36, 1117–1124 (October 2001).
Hyperlipidemia (hypertriglyceridemia, hypercholesterolemia,
or both) is a common abnormality in 20 to 81% of patients
with nonalcoholic steatohepatitis (NASH) (1), an increasingly
recognized chronic liver disease characterized by the histo-
logical pictures of alcoholic disease in the absence of signifi-
cant ethanol consumption (2). Other conditions commonly as-
sociated with NASH are obesity and non-insulin-dependent
diabetes mellitus (1,3). An increase in fasting blood glucose
ber of apolipoprotein B100 (ApoB100) particles is not signifi- with reduced glucose tolerance is coupled with depressed me-
cantly different between the two groups, and the higher content
of triglycerides in NASH very low density lipoproteins (VLDL)
increases the triglyceride-to-ApoB ratio and the particle size. A
decreased enzymatic activity of LPL or a defective assembly and
secretion of VLDL from hepatocytes due to a moderate reduc-
tion in microsomal triglyceride transfer protein could be in-
volved in the overloading of VLDL. Moreover, the undetectable
levels of ApoB48 in triglyceride-rich lipoproteins fraction A
could be related to the synthesis of smaller and denser chylomi-
crons. NASH patients not only are insulin resistant but also tend
to present alterations in fatty meal delivery, suggesting that an
tabolism of triglyceride-rich lipoproteins (TRL) and an in-
crease in their fasting and postprandial levels (4,5). The links
between glucose and lipoprotein metabolism have received
particular attention in the context of the metabolic syndrome
(6). Plasma insulin concentration, an indirect index of insulin
sensitivity, was shown to be related to the production of large
very low density lipoproteins (VLDL) particles (Svedberg
flotation rate, Sf, of 60 to 400) in experimental studies with
humans (7).
The postprandial phase is regarded as potentially athero-
increase in fasting plasma insulin and glucose, with insulin re- genic (6–8). Investigation of dynamic metabolism after a
sistance, joins with depressed metabolism of triglyceride-rich
lipoproteins. An increase in postprandial triglyceride levels with
production of large VLDL suggests an atherogenic behavior of
mixed or a fatty meal provides more complete data about the
behavior of other lipoprotein classes. TRL are a heteroge-
neous population of lipoprotein particles. Those containing
apolipoprotein (Apo) B100 are secreted from the liver
(VLDL), whereas those with ApoB48 are secreted from the
intestine as chylomicrons after fat intake. Several reports
have concluded that postprandial alterations in healthy men
induce transient compositional alterations of VLDL that link
these lipoprotein species to the formation of lipid deposits on
arterial walls (4,6,7,9,10). TRL can also trigger an impaired
fibrinolysis and endothelial dysfunction (11), which may be
ultimately responsible for cardiovascular disease (12). More-
over, the fasting lipid profile may be in the normal range and
fail to reflect the atherogenic risk of subjects with impaired
postprandial triglycerides (Tg) removal (12,13).
*To whom correspondence should be addressed at Dipartimento di Medicina
Interna, Corso A.M. Dogliotti 14, 10126 Torino, Italy.
E-mail: maurizio.cassader@unito.it
Abbreviations: Apo, apolipoprotein; AUC, area under the postprandial
triglyceride curve; BMI, body mass index; Chol, cholesterol; FFA, free fatty
acid; FPG, fasting plasma glucose; FSIGT, frequently sampled intravenous
glucose tolerance test; HDL, high density lipoproteins; LDL, low density
lipoproteins; LPL, lipoprotein lipase; MTP, microsomal triglyceride transfer
protein; NASH, nonalcoholic steatohepatitis; OFL, oral fat load; PCR, poly-
merase chain reaction; SDS, sodium dodecyl sulfate; Sf, Svedberg flotation
rate; SI, sensitivity index; Tg, triglyceride; TRL, triglyceride-rich lipopro-
teins; VLDL, very low density lipoproteins.
Copyright © 2001 by AOCS Press
1117
Lipids, Vol. 36, no. 10 (2001)

1118
M. CASSADER ET AL.
The aim of the study was to examine the dynamic post- and 30% from fat, with a polyunsaturated to saturated fat ratio
prandial metabolism of TRL and assess its relationship to in- of 1.50). Mean daily cholesterol (Chol) intake was 200–250 mg.
sulin sensitivity in 11 nonobese, nondiabetic NASH patients
and 16 healthy subjects with a normal fasting lipid profile medical examination was carried out at 9
Determination of lipid parameters at baselines. A hospital
. after a fast of
A.M
after a standardized oral fat-loading test.
12 h or more. Blood samples drawn from the antecubital vein
were used to determine the Tg, total Chol, low density
lipoprotein (LDL) Chol, high density lipoprotein (HDL)
Chol, HDL2-Chol, and HDL3-Chol, Apo AI, Apo B, glucose,
MATERIALS AND METHODS
Study subjects. Lipoproteins from 11 NASH patients and 16 creatinine, albumin, and thyroid-stimulating hormone base-
healthy subjects were studied. None of the participants was lines. The subjects were then asked to return on different
obese [(body mass index (BMI) 25.9 2.0 kg/m2 for patients mornings for the oral fat load (OFL) test after a fast of not
and 25.0 2.5 kg/m2 for controls] or diabetic. Fasting plasma less than 12 h. Blood (1 mg/mL EDTA-Na2) was centrifuged
lipoprotein lipid concentrations are shown in Table 1. There was for 30 min at 2500 rpm and 4°C in a J6B centrifuge (Beck-
no clinical or laboratory evidence of thyroid dysfunction or man Instruments, Palo Alto, CA) and stored at –20°C until
other conditions leading to secondary hyperlipoproteinemia in processed. Total Chol and Tg were measured enzymatically
either group. Inclusion criteria for the 11 NASH patients (35
(Poli Diagnostici, Milan, Italy). Plasma HDL-Chol was de-
9 yr, 10 male/1 female) were as follows: histological feature of termined after precipitation of ApoB-containing lipoproteins
alcoholic steatohepatitis with chronically elevated aminotrans- with heparin and manganese chloride (14). Plasma HDL2-
ferases and/or γ-glutamyl transpeptidase; negative history of al- and HDL3-Chol levels were determined according to Gidez
cohol consumption (>40 g/wk); negative markers for hepatitis et al. (15): HDL2 and HDL3 were separated after precipita-
B virus/hepatitis C virus; no recent use of known steatogenic tion of ApoB-containing lipoproteins with heparin and man-
drugs; absence of autoantibodies related to autoimmune liver ganese chloride, and HDL2 were further precipitated with
disease; normal serum iron and copper metabolism profile; ab- dextran sulfate. The formula of Friedewald et al.(16) was
sence of serum immunoglobulin M antibodies, anti-Epstein- used to calculate LDL-Chol. Plasma ApoAI and ApoB were
Barr virus, herpes virus, and cytomegalovirus. The controls determined by an automated immunoturbidimetric method
were 16 healthy subjects matched for age, sex (33 9 yr, 15 (Poli Diagnostici) with a Shimadzu CL-7000 (Shimadzu In-
male/1 female), and the main anthropometric parameters. All struments, Kyoto, Japan).
subjects provided their informed consent. Participants received
DNA analysis. APOE (apolipoprotein E) genotypes were
a questionnaire about fat intake, alcohol intake, and physical ac- determined by polymerase chain reaction (PCR) amplifica-
tivity. Dietary habits were similar between young and elderly tion of genomic DNA using specific oligonucleotide primers
subjects (20% of calories of protein, 50% from carbohydrates, (17). Genomic DNA was isolated from frozen EDTA whole
blood through the QIAamp Blood Kit (Qiagen-Genenco M-
Medical s.r.l., Florence, Italy). Genomic DNA (0.5 µg) was
amplified in 25 µL reaction mixture containing 10 mmol/L
TABLE 1
Tris-HCl (pH 8.3); 50 mmol/L KCl; 200 µmol each of dATP,
dCTP, dGTP, and dTTP; 1.5 mmol/L MgCl2; 200 pmol of
each primer, and 1.8 U Taq DNA polymerase (Roche Diag-
nostics S.p.A., Milan, Italy). The amplification cycle was per-
formed in a Gene Cycler (Bio-Rad, Milan, Italy). Five min-
utes of denaturation at 94°C was followed by 30 cycles of
1 min at 90°C, 1 min at 60°C, and 2 min at 70°C. PCR prod-
ucts were cleaved with 4 U of HhaI restriction endonuclease
(New England Biolabs, Ipswich, MA), as recommended by
the manufacturer. Subsequently, the samples were electro-
phoresed through a 10% nondenaturing polyacrylamide gel
and visualized by silver staining (Bio-Rad, Milan, Italy). The
sizes of HhaI fragments were estimated by comparison with
known size markers (MspI-digested pUC18 DNA).
Frequently sampled intravenous glucose tolerance test
(FSIGT). Insulin sensitivity was assessed by performing an
FSIGT. The data were analyzed with the minimal modeling
technique to estimate the insulin sensitivity index (SI) as de-
scribed in detail elsewhere (18).
Oral fat load. Five minutes after drawing a blood sample
(t0), the subjects consumed a liquid OFL consisting of a mix-
ture of dairy cream (35% fat) and egg yolk. The total amount
Baseline Characteristics, Frequently Sampled Intravenous Glucose
Tolerance Test (FSIGT), and Minimal Model Parameters for NASH
Patients and Controlsa
NASH patients
Controls
(n = 11)
(n = 16)
P
Age (yr)
35
9
33
9
NS
Sex (M/F)
10/1
25.9 2.0
91
97 12
13.9 4.5
3.66 2.86
15/1
25 2.5
83
89 14
6.9 2.8
9.15 2.94
NS
BMI (kg/m2)
0.330
0.002
0.135
0.0001
0.000
Waist (cm)
FPG (mg/dL)
IRI (µU/mL)
6
5
Insulin sensitivity
index (min–1/µU/mL)
Triglycerides (mg/dL)
Total cholesterol (mg/dL)
LDL-cholesterol (mg/dL)
HDL-cholesterol (mg/dL)
HDL2-cholesterol (mg/dL)
HDL3-cholesterol (mg/dL)
Apo AI (mg/dL)
97 67
183 40
108 34
75 27
185 22
106 25
59 16
18.4 7.5
40.5 7.1
121 26
85 21
0.246
0.868
0.861
0.187
0.047
0.648
0.841
0.840
52
7
13.0 5.0
39.4 4.1
119 24
83 30
Apo B (mg/DL)
aData are presented as mean standard deviation. NASH, nonalcoholic
steatohepatitis; BMI, body mass index; FPG, fasting plasma glucose; IRI, im-
muno-reactive insulin; LDL, low density lipoprotein; HDL, high density
lipoprotein; Apo, apolipoprotein; NS, not significant.
Lipids, Vol. 36, no. 10 (2001)

POSTPRANDIAL LIPID METABOLISM IN NASH
1119
of fat was based on the subject’s body surface area (77.5 g fat, tial ultracentrifugation and used to quantify ApoB100 and
0.5 g Chol per m2). The OFL was consumed during a period ApoB48. The bands were quantified by densitometry using
of 15 min, and the subjects then remained without food for Gel Doc equipment (Bio-Rad). Density values were assigned
10 h. A catheter (Venflon Viggo AB, Helsingborg, Sweden) to the ApoB100 bands of the standard LDL and a standard
inserted in the antecubital vein and kept patent during the test curve was constructed. The values were recalculated by lin-
was used to draw samples after 2, 4, 6, 8, and 10 h (t2, t4, t6, ear regression.
t8, and t10) for the determination of plasma TRL. Strenuous
activity was forbidden during the test, because exercise can culated for the lipid baseline parameters, and differences be-
reduce postprandial lipemia. tween the patients and controls were determined with the t
Statistics. Means and standard deviations (SD) were cal-
Laboratory analyses. Plasma total Chol and Tg were mea- test, or the Mann-Whitney test when comparing postprandial
sured by means of automated enzymatic methods. The area area. Tg levels were log-transformed before testing. Differ-
under the postprandial triglyceride curve (AUC) was evalu- ences in the postprandial responses (plasma Tg and VLDL
ated using the trapezoid rule to estimate the overall response fraction Tg values) were assessed by connecting Tg plasma
of plasma Tg during the entire 10-h postprandial period.
levels vs. time with segments. The AUC was computed by the
Separation of VLDL and TRL subfractions. VLDL were trapezoid method on the area exceeding baseline (incremen-
isolated through preparative ultracentrifugation and subse- tal area). Spearman’s coefficient was used to assess the corre-
quently assayed for their Tg and total Chol content from all lation between parameters concerning insulin sensitivity and
the subjects’ blood samples. Briefly, one aliquot of plasma basal and postprandial lipid profile.
was brought to densities of 1006 g/L by addition of a KBr so-
lution (d = 1330 g/L) and centrifuged at 105,000 × g in a type
50 rotor for 21 h at 10°C in a Beckman L8-70M ultracen-
RESULTS
trifuge. Concentrations of Tg and total Chol were then deter- Patients’ clinical characteristics and fasting plasma lipids
mined in lipoprotein fractions enzymatically. The first amount and lipoproteins. Mean SD of clinical anthropometric and
of blood (10 mL) was drawn for subfractionating TRL by ul- basal laboratory determinations of glucose and lipid metabo-
tracentrifugation on a discontinuous density gradient (19). lism in NASH patients and controls are reported in Table 1.
Plasma separated as already described was brought to a den- Although cases and controls showed similar BMI, there was
sity of 1.10 g/mL by the addition of solid KBr. The density a tendency to visceral fat accumulation in NASH patients
gradient was prepared manually by adding to 4 mL of this (waist: 91 6 vs. 83 5 cm, P = 0.002). Fasting total plasma
plasma 3 mL of a 1.065 g/mL solution containing 0.05% Chol (183 40 vs. 185 22 mg/dL, P = 0.618) and Tg (97
KBr/NaCl plus EDTA (pH 7.4); 3 mL of a similar solution at 67 vs. 75 27 mg/dL, P = 0.246) were not significantly dif-
1.020 g/mL; and 3 mL of physiological saline at 1.006 g/mL. ferent. HDL2-Chol levels were significantly lower in the pa-
The sample was ultracentrifuged in a SW4O rotor in a Beck- tients (13.0 5.0 vs. 18.4 7.5 mg/dL, P = 0.047). The other
man L8-70M centrifuge at 20°C in stages allowing the sepa- Chol lipoprotein fractions had similar levels in the two groups
ration of four VLDL fractions with decreasing Sf values: A > as well as ApoAI and ApoB (Table 1). Fasting plasma insulin
400; B = 175–400; C = 100–175; D = 20–100. The first cen- (13.9 4.5 vs. 6.9 2.8 µU/mL, P = 0.0001) and fasting
trifugation (28,300 rpm for 43 min) isolated fraction A in a plasma glucose (FPG) levels (97 12 vs. 89 14 mg/dL, P =
floating volume of 0.5 mL. The volume removed was re- 0.135) were increased in the group of NASH patients. There
placed with physiological saline, and the sample was cen- was no difference in ApoE allelic frequency: 12 NASH pa-
trifuged at 40,000 rpm for 67 min to isolate fraction B. This tients were E3/E3, 2 were 4/3, and 2 were 3/2; 8 controls were
procedure was then repeated at 40,000 rpm for 71 min and at E3/E3, 2 were 4/3, and 1 was 3/2. There was no difference in
37,000 rpm for 18 h to isolate fractions C and D, respectively. percentage of smokers: 3 patients and 2 controls.
The automated methods mentioned above were used to deter-
mine Chol and Tg on the four fractions.
Insulin sensitivity. NASH patients had significantly lower
insulin sensitivity (SI 3.66 2.86 vs. 9.15 2.94 min–1/
ApoB48 and ApoB100 analysis. ApoB48 and ApoB100 µU/mL, P = 0.000)
analysis were performed on five subjects with NASH and on
Postprandial response of plasma Tg. The total OFL ingested
five controls matched for age, sex, and the main anthropomet- by the NASH group was 390 5.2 g (mean SD), and by the
ric parameters.
control group was 385 4.8 g. It was well tolerated, and there
VLDL ApoB48 and ApoB100 were separated by sodium were no instances of gastrointestinal disturbance or steatorrhea
dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis during or after the test. The postprandial plasma Tg and VLDL-
using 3.9% gel according to Battula et al. (20). Nondelipi- Tg curves are reported in Figure 1 (right-hand panel). Plasma
dated lipoprotein samples were reduced in SDS sample buffer Tg peaked 4 h after the meal and returned to fasting levels after
for 4 min at 96°C. Samples were applied to the gel and run at 8 h in both groups (Fig. 1). Total Chol and LDL-Chol displayed
40 mA in 0.025 M Tris, 0.192 M glycine, and 0.1% SDS. Gel a flat pattern in both groups, and the HDL-Chol levels were un-
was stained with Silver Stain (Bio-Rad). Since the chro- changed. Plasma Tg were higher in the patients at all time points
mogenicity of ApoB48 is similar to that of ApoB100 (21), a with a significant difference at 4 h (p = 0.006). The response of
protein standard was prepared from LDL isolated by sequen- VLDL-Tg is illustrated in Figure 1. VLDL-Tg are higher in the
Lipids, Vol. 36, no. 10 (2001)

1120
M. CASSADER ET AL.
A
B
Time (h)
Time (h)
FIG. 1. Change in (A) plasma triglyceride (Tg) and in (B) very low density lipoprotein (VLDL) Tg concentrations after the fat
load in 11 nonalcoholic steatohepatitis (NASH) subjects (■) patients and 16 controls (●). Results are expressed as mean
standard error of the mean. *P < 0.05. **P < 0.01.
NASH group at all time points with significant differences at similar way in both control and NASH individuals. The
4 and 6 h (P = 0.043 and P = 0.019). There was no significant ApoB100 increased and peaked at 4 h in VLDL fraction D in
difference in the incremental AUC for Tg: patients, 1329
596.2 mg/dL × 10 h, P = 0.063; controls, 984.6 322 mg/dL
× 10 h. The NASH group SI was significantly correlated with
the fasting total Tg concentration (rs = –0.67, P = 0.042), fast-
the control group, but this difference was not significant.
DISCUSSION
ing VLDL-Tg (r = –0.61, P = 0.048), VLDL-Tg (fraction B) The present study was designed to highlight some mecha-
(rs = –0.67, P = 0.033), and VLDL-Tg (fraction C) (rs = nisms that potentially link atherosclerosis to an abnormal fat
–0.64, P = 0.044), but not with any other basal or postpran- tolerance in NASH patients. NASH is a syndrome frequently
dial parameter. A correlation between fasting plasma Tg and associated with obesity, diabetes mellitus, and dyslipidemia
the AUC was observed in both groups: NASH, r = 0.835, P = (1,3). Elevated levels of free fatty acids (FFA) in the liver are
0.0013 (rs = 0.763, P = 0.0062); controls, r = 0.487, P = supposed to be responsible for the development of steatohep-
0.0551 (rs = 0.633, P = 0.0083). The slope for the regression atitis as a part of the metabolic syndrome (6). Increased fast-
line between fasting Tg and AUC was less steep in the con- ing blood glucose levels may trigger a depressed metabolism
trol group (AUC = 574.67 + 5.46 Tg) than that in the NASH of lipoproteins rich in Tg and an increase in their fasting and
group (AUC = 584.26 + 7.66 Tg).
postprandial levels (4,5).
Postprandial Tg levels in TRL subfractions. The lipid con-
tent of large and small TRL particles isolated by ultracentrifu-
gation after the OFL is shown in Figure 2. The lipid composi-
tion of fraction A showed a transient Tg increase after 4 and 6 h
in the NASH group (P = 0.03 and 0.04, respectively). The dif-
ference was no longer significant during the last postprandial
hours. The Tg content of VLDL subfractions B, C, and D dis-
played similar curves in both groups throughout the postpran-
dial load period. There were no significant differences between
the two groups in the total Chol variations at any time.
FRACTION A
FRACTION C
Time (h)
Time (h)
FRACTION D
FRACTION B
ApoB48 and ApoB100 in TRL subfractions. Absolute post-
prandial changes of ApoB48 and ApoB100 concentrations in
VLDL fractions are shown in Figures 3 and 4, respectively.
ApoB48 in VLDL fraction A was detected only in the control
group. In VLDL fractions B, C, and D, ApoB48 was at lower
levels in NASH patients. In the control group, the ApoB48
concentrations increased significantly in the VLDL fraction
D at the baseline and at 2 and 10 h. The ApoB100 concentra-
tion in VLDL fraction A, B, and C responded to the OFL in a
Time (h)
Time (h)
FIG. 2. Tg concentrations during the fat load test in the Sf A > 400, Sf B
= 175–400, Sf C = 100–175, Sf D = 20–100 lipoprotein fractions (for
method see Materials and Methods section) in 11 NASH patients (■)
and 16 controls (●). Results are expressed as means. *P < 0.05. For ab-
breviations see Figure 1.
Lipids, Vol. 36, no. 10 (2001)

POSTPRANDIAL LIPID METABOLISM IN NASH
1121
ApoB 48-VLDL: A
ApoB100-VLDL: A
Hours
Hours
ApoB 48-VLDL: B
ApoB100-VLDL: B
Hours
Hours
ApoB 48-VLDL: C
ApoB100-VLDL: C
Hours
Hours
ApoB 48-VLDL: D
ApoB100-VLDL: D
Hours
Hours
FIG. 3. Line plots of changes in plasma apolipoprotein B48 (ApoB48)
concentrations during the fat load test in Sf A > 400, Sf B = 175–400, Sf
C = 100–175, Sf D = 20–100 lipoprotein fractions (for method see Ma-
terials and Methods section) in 5 NASH patients (■) and 5 controls (●).
Results are expressed as means. *P < 0.05. VLDL, very low density
lipoprotein; for other abbreviations see Figure 1.
FIG. 4. Line plots of changes in plasma ApoB100 concentrations during
the fat load test in Sf A > 400, Sf B = 175–400, Sf C = 100–175, Sf D =
20–100 lipoprotein fractions (for method see Materials and Methods
section) in 5 NASH patients (■) and 5 controls (●). Results are expressed
as means. For abbreviations see Figures 1 and 3.
tion in the postprandial phase. The curves of both groups for
the other VLDL classes were superimposed. The persistent
postprandial plasma Tg elevation in NASH patients was mostly
due to the elevated level of large TRL particles (fraction A).
The production of large VLDL particles (Sf rate of 60 to 400)
seems to be regulated by plasma insulin concentrations in ex-
perimental studies in humans (7). Some authors have shown
that this persistent elevation is ascribable to reduced clearance
of postprandial large VLDL, which could also be related to in-
creased competition with chylomicrons for the same lipolytic
pathway (22). NASH patients present an impaired metabolism
in controlling plasma Tg levels after fat intake, and this may
depend on their insulin-resistant state. The role of postprandial
insulin in the regulation of postprandial lipid metabolism is still
poorly understood. The roles of hyperinsulinemia and insulin
resistance in the alteration of lipid metabolism are not clear ei-
ther (23). Nevertheless, some authors have reported that post-
prandial lipemia is related to the degree of insulin resistance
In our patients, fasting plasma insulin was significantly in-
creased and HDL2-Chol levels were significantly decreased.
SI was negatively correlated to fasting total Tg concentration,
fasting VLDL-Tg, VLDL-Tg (fraction B), and VLDL-Tg
(fraction C) but not to any other basal or postprandial param-
eter. The most important results of our study are the increase
of TRL response after a fat load in the NASH patients, the in-
crease of the incremental area under the postprandial curve,
and the duration of the hypertriglyceridemic peaks. Further-
more, we observed a significant correlation between plasma
Tg and the Tg AUC in the NASH group. The slope of the re-
gression line between fasting and the Tg AUC was steeper in
the NASH group. If the fasting Tg level mirrors the postpran-
dial Tg level, then NASH patients have a higher postprandial
Tg level than the control for the same level of fasting Tg.
NASH patients seem to have a disturbance in the metabo-
lism of TRL particles, since they displayed a rapid increase
of Tg fraction A levels of lipoprotein and a sustained eleva-
Lipids, Vol. 36, no. 10 (2001)

1122
M. CASSADER ET AL.
even in nondiabetic subjects (24). In the plurimetabolic syn- ing to atherogenesis (8); other investigators maintain that the
drome the simple fasting hyperinsulinemia is associated with postprandial phase is a critical moment, marked by an abnor-
the increased postprandial lipemia even if hypertriglyc- mal metabolism of Tg transportation, its carriers (38), and
eridemia and diabetes mellitus are absent (25).
other related enzymes or lipoprotein classes (39,40). It has
Our data may show that there is no significant difference in also been suggested that a prolonged postprandial hyper-
the content of ApoB100 in VLDL fraction A between NASH triglyceridemic response alters the coagulation-fibrinolytic
and control subjects. These results mean that the number of balance and favors thrombosis (13).
ApoB100 particles is not significantly different between the two
High plasma Tg is also closely correlated with low HDL-
groups. Because postprandial Tg are higher in NASH subjects, Chol (41–43) and may influence the composition of the HDL;
the net effect is to increase the Tg-to-ApoB ratio and to increase an inverse relationship between HDL2-Chol levels and the
particle size. In our study, VLDL fraction A of NASH subjects postprandial Tg levels is also evident (44). This shows that
carry more Tg than VLDL isolated from control subjects, sug- the substantial alterations in TRL metabolism upstream from
gesting a decreased enzymatic activity of lipoprotein lipase the lipolytic and maturative cascade of the lipoproteins have
(LPL), a common abnormality in insulin resistance, or a defec- a profound effect on the variations in the other fractions.
tive assembly and secretion of VLDL from hepatocytes due to a These observations lead to the conclusion that HDL levels
moderate reduction in microsomal Tg transfer protein (MTP) as and composition reflect the efficiency of TRL catabolism
a consequence of different gene polymorphism (26).
(38,42,43), and that high HDL-Chol indicates a low produc-
The most striking result of our study is that ApoB48 in tion of atherogenic remnants according to a mechanism that
TRL fraction A in NASH subjects is below the sensitivity of also involves the cholesteryl ester transfer protein (45).
the silver stain. ApoB48 levels are extremely low in fraction
It seems appropriate to stress that NASH patients not only
A, and this fact could be related to an altered functioning of are insulin-resistant but also tend to present alterations in fatty
the MTP. Authors have reported that MTP play a key role in meal delivery, suggesting that an increase in fasting plasma
the first steps of chylomicron assembly (27). Even if MTP has insulin and FPG, with insulin resistance, is coupled with
no effect on ApoB48 synthesis, in the intestine MTP appears lower plasma HDL2-Chol and depressed metabolism of TRL.
to be important in regulating the Tg content of the chylomi- An increase in fasting and postprandial Tg levels with pro-
cron particle (28). Thus, an impaired functioning of MTP in duction of large VLDL suggests an atherogenic behavior of
NASH subjects could give rise to smaller and denser chy- lipid metabolism, in accordance with the high prevalence of
lomicrons, and it could explain why ApoB48 is absent in frac- the metabolic syndrome in these patients (6,46). Longitudinal
tion A and present in the other fractions.
studies are needed to confirm this finding.
The reduction in ApoB48 and ApoB100 in VLDL fraction
These findings indicate that the presence of NASH should
D in NASH subjects could be explained with a flaw of lipoly- be systematically sought both in patients with unexplained in-
sis of larger and less dense lipoproteins. crease of liver enzymes in the absence of other known mech-
Since LDL are produced during VLDL catabolism, aug- anisms of liver injury and in subjects with two or more ele-
mented secretion and availability of the Tg bound to the ments of the metabolic syndrome, irrespective of the presence
VLDL and the events subsequent to these metabolic changes of obesity or visceral adiposity. Owing to the frequent associ-
may account for the production of altered particles (29). Ex- ation of the two syndromes, an oral glucose tolerance test and
aggerated postprandial lipemia has consequences for the par- plasma insulin assay together with an OFL should be system-
ticle composition of plasma lipoprotein: Tg-enriched LDL are atically performed in NASH patients for the early detection
good substrates for hepatic lipase and the result is smaller and of atherogenic risk in the presence of quite normal fasting
denser LDL (1.040 < d <1.063 g/mL) constituting an athero- plasma lipids.
genic lipoprotein phenotype (19,29,30). Because the smaller,
denser LDL are more susceptible to oxidation (31), the de-
gree of oxidation may also vary in function of the main
ACKNOWLEDGMENTS
changes in the metabolism of the Tg bound to the VLDL. This study was funded in part from a grant from Ministero dell’Uni-
versità e della Ricerca Scientifica e Tecnologica (MURST) 60%,
Italy.
Moreover, the greater the degree of LDL oxidation, the
greater the atherogenic potential of the particle (32).
These results may suggest that impaired postprandial Tg
metabolism in NASH patients is associated with their in-
creased visceral obesity. Reduced SI, FFA elevation (33), and
reduced TRL catabolism by LPL are common in the meta-
bolic syndrome (34,35).
An association between postprandial lipemia and coronary
heart disease had been observed since the early 1950s (36),
and many studies now support this concept (8–10,37).
According to some workers, chylomicrons and their rem-
nants deliver dietetic Chol to the arterial wall, thus contribut-
REFERENCES
1. Sheath, S.G., Gordon, F.D., and Chopra, S. (1997) Nonalcoholic
Steatohepatitis, Ann. Intern. Med. 126, 137–145.
2. Ludwig, J., Viggiano, T.R., McGill, D.B., and Ott, B.J. (1980)
Nonalcoholic Steatohepatitis. Mayo Clinic Experiences with a
Hitherto Unnamed Disease, Mayo Clin. Proc. 55, 434–438.
3. Powell, E.E., Cooksley, W.G., Hanson, R., Searle, J., Halliday,
J.W., and Powell, L.W. (1990) The Natural History of Nonalco-
holic Steatohepatitis: A Follow-Up Study of Forty-Two Patients
for up to 21 Years, Hepatology 11, 74–80.
Lipids, Vol. 36, no. 10 (2001)

POSTPRANDIAL LIPID METABOLISM IN NASH
1123
4. Cohn, J.S., McNamara, J.R., Cohn, S.D., Ordovas, J.M., and 23. Harbis, A., Defoort, C., Narbonne, H., Juhel, C., Senft, M.,
Schaefer, E.J. (1988) Postprandial Plasma Lipoprotein Changes
in Human Subjects of Different Ages, J. Lipid Res. 29, 469–479.
5. Krasinski, S.D., Cohn, J.S., Schaefer, E.J., and Russel, R.M.
(1990) Postprandial Plasma Retinyl Ester Response Is Greater
in Older Subjects Compared with Younger Subjects. Evidence
Latge, C., Delenne, B., Portugal, H., Atlan-Gepner, C.,
Vialettes, B., and Lairon, D. (2001) Acute Hyperinsulinism
Modulates Plasma Apolipoprotein B-48 Triglyceride-Rich
Lipoproteins in Healthy Subjects During the Postprandial Pe-
riod, Diabetes 50, 462–469.
of Delayed Plasma Clearance of Intestinal Lipoproteins, J. Clin. 24. Jeppesen, J., Hollenbeck, C.B., Zhou, M.-Y., Coulston, A.M.,
Invest. 85, 883–892.
Jones, C., Ida Chen, Y.-D., and Reaven, G.M. (1995) Relation
Between Insulin Resistance, Hyperinsulinemia, Postheparin
Plasma Lipoprotein Lipase Activity, and Postprandial Lipemia,
Arterioscerl. Thromb. Vasc. Biol. 15, 320–324.
6. Grundy, S.M. (1998) Hypertriglyceridemia, Atherogenic Dys-
lipidemia, and the Metabolic Syndrome, Am. J. Cardiol. 81
(4A), 18B–25B.
7. Malmström, R., Packard, C.J., Watson, T.D., Rannikko, S., 25. Schrezenmeir, J., Keppler, I., Fenselau, S., Weber, P., Biesalski,
Caslake, M., Bedford, D., Stewart, P., Yki-Jarvinen, H., Shep-
herd, J., and Taskinen, M.R. (1997) Metabolic Basis of Hy-
potriglyceridemic Effects of Insulin in Normal Men, Arte-
rioscler. Thromb. Vasc. Biol. 17, 1454–1464.
H.K., Probst, R., Laue, C., Zuchhold, H.D., Prellwitz, W., and
Beyer, J. (1993) The Phenomenon of a High Triglyceride Re-
sponse to an Oral Lipid Load in Healthy Subjects and Its Link
to the Metabolic Syndrome, Ann. NY Acad. Sci. 683, 302–314.
8. Zilversmit, D.B. (1979) Atherosclerosis: A Postprandial Phe- 26. Bernard, S., Touzet, S., Personne, I., Lapras, V., Bondon, P.J.,
nomenon, Circulation 60, 473–485.
9. Havel, R.J. (1994) Postprandial Hyperlipidemia and Remnants
Lipoproteins, Curr. Opin. Lipidol. 5, 102–109.
10. Cohn, J.S. (1994) Postprandial Lipid Metabolism, Curr. Opin.
Lipidol. 5, 185–190.
11. Björkegren, J., Hamsten, A., Milne, R.W., and Karpe, F. (1997)
Berthezene, F., and Moulin, P. (2000) Association Between Mi-
crosomal Triglyceride Transfer Protein Gene Polymorphism and
the Biological Features of Liver Steatosis in Patients with Type
II Diabetes, Diabetologia 43, 995–999.
27. Hussain, M.H. (2000) A Proposed Model for Assembly of Chy-
lomicrons, Atherosclerosis 148, 1–15.
Alterations of VLDL Composition During Alimentary Lipemia, 28. Gleeson, A., Anderton, K., Owens, D., Bennett, A., Collins, P.,
J. Lipid Res. 38, 301–314.
Johnson, A., White, D., and Tomkin, G.H. (1999) The Role of
Microsomal Triglyceride Transfer Protein and Dietary Choles-
terol in Chylomicron Production in Diabetes, Diabetologia 42,
944–948.
12. Karpe, F., Hellénius, M.L., and Hamsten, A. (1999) Differences
in Postprandial Concentrations of Very-Low-Density Lipopro-
tein and Chylomicron Remnants Between Normotriglyceri-
demic and Hypertriglyceridemic Men With and Without Coro- 29. Patsch, J.R., Misenböck, G., Hopferwieser, T., Muhlberger, V.,
nary Heart Disease, Metabolism 48, 301–307.
Knapp, E., Dunn, J.K., Gotto, A.M., Jr., and Patsch, W. (1992)
Relation of Triglyceride Metabolism and Coronary Artery Dis-
ease. Study in the Postprandial State, Arterioscl. Thromb. 12,
1336–1345.
13. Hamsten, A., and Karpe, F. (1996) Triglycerides and Coronary
Heart Disease—Has Epidemiology Given Us the Right An-
swer?, in Lipids: Current Perspective (Betteridge, D.J., ed.), pp.
43–68, Martin Dunitz, London.
14. Warnick, G.R., and Albers, J.J. (1978) A Comprehensive Evolu-
tion of the Heparin Manganese Precipitation Procedure for Esti-
30. Misenböck, G., and Patsch, J.R. (1992) Postprandial Hyper-
lipemia: The Search for the Atherogenic Lipoprotein, Curr.
Opin. Lipidol. 3, 186–201.
mating High Density Lipoprotein Cholesterol, J. Lipid Res. 29, 31. Mann, C.J., Yen, F.T., Grant, A.M., and Bihain, B.E. (1991)
65–76.
Mechanism of Plasma Cholesterol Ester Transfer in Hyper-
triglyceridemia, J. Clin. Invest. 88, 2059–2066.
15. Gidez, L.I., Miller, G.J., Burnstein, M., Slagle, S., and Eder,
H.A. (1982) Separation and Quantitation of Subclasses of 32. Patsch, J.R., Prasad, S., Gotto, A.M., Jr., and Patsch, W. (1987)
Human Plasma HDL by a Single Precipitation Procedure, J.
Lipid Res. 23, 1206–1216.
16. Friedewald, W.T., Levy, R.I., and Frederickson, D.S. (1972) Es-
timation of the Concentration of Low Density Lipoprotein Cho-
High Density Lipoprotein2: Relationship of the Plasma Levels
of this Lipoprotein Species to its Composition, to the Magnitude
of Postprandial Lipemia and to the Activities of Lipoprotein Li-
pase and Hepatic Lipase, J. Clin. Invest. 80, 341–347.
lesterol in Plasma Without Use of the Preparative Ultracen- 33. Davis, C.E., Gordon, D., LaRosa, J., Wood, P.D., and Halperin,
trifuge, Clin. Chem. 18, 499–502.
M. (1980) Correlation of Plasma High-Density Lipoprotein
Cholesterol Levels with Other Plasma Lipid and Lipoprotein
Concentrations. The Lipid Research Clinics Program Prevalence
Study, Circulation 62, 24–30.
17. Hixson, J.E., and Vernier, D.T. (1990) Restriction Isotyping of
Human Apolipoprotein E by Gene Amplification and Cleavage
with HhaI, J. Lipid Res. 31, 545–548.
18. Bergman, R.N., Phillips, L.S., and Cobelli, C. (1981) Physio- 34. Karpe, F., Bard, J.M., Steiner, G., Carlson, L.A., Fruchart, J.C.,
logic Evaluation of Factors Controlling Glucose Tolerance in
Man, J. Clin. Invest. 68, 1456–1467.
19. Redgrave, T.G., and Carlson, L.A. (1979) Changes in Plasma
and Hamsten, A. (1993) HDLs and Alimentary Lipemia. Stud-
ies in Men with Previous Myocardial Infarction at a Young Age,
Arterioscler. Thromb. 13, 11–22.
Very Low Density and Low Density Lipoprotein Content, Com- 35. Deckelbaum, R.J., Granot, E., Oschry, Y., Rose, L., and Eisem-
position, and Size After a Fatty Meal in Normo- and Hyper-
triglyceridemic Man, J. Lipid Res. 20, 217–229.
20. Battula, S.B., Fitzsimons, O., Moreno, S., Owens, D., Collins,
berg, S. (1984) Plasma Triglyceride Determines Structure-Com-
position in Low and High Density Lipoproteins, Arteriosclero-
sis 4, 225–231.
P., Johnson, A., and Tomkin, G.H. (2000) Postprandial 36. Patsch, J.R., Prasad, S., Gotto, A.M., Jr., and Patsch, W. (1987)
Apolipoprotein B48- and B100-Containing Lipoproteins in
Type 2 Diabetes: Do Statins Have a Specific Effect on Triglyc-
eride Metabolism? Metabolism 49, 1049–1054.
High Density Lipoprotein2. Relationship of the Plasma Levels
of This Lipoprotein Species to its Composition, to the Magni-
tude of Postprandial Lipemia, and to the Activities of Lipopro-
tein Lipase and Hepatic Lipase, J. Clin. Invest. 80, 341–347.
21. Karpe, F., and Hamsten, A. (1994) Determination of
Apolipoproteins B48 and B100 in Triglyceride-Rich Lipopro- 37. Karpe, F., Tornvall, P., Olivecrona, T., Steiner, G., Carlson,
teins by Analytical SDS-PAGE, J. Lipid Res. 35, 1311–1317.
22. Syvänne, M., and Taskinen, M.R. (1997) Lipids and Lipopro-
teins as Coronary Risk Factor in Non-Insulin-Dependent Dia-
betes Mellitus, Lancet 350 (suppl. 1), 20–23.
L.A., and Hamsten, A. (1993) Composition of Human Low Den-
sity Lipoprotein: Effects of Postprandial Triglyceride-Rich
Lipoproteins, Lipoprotein Lipase, Hepatic Lipase and Choles-
teryl Ester Transfer Protein, Atherosclerosis 98, 33–49.
Lipids, Vol. 36, no. 10 (2001)

1124
M. CASSADER ET AL.
38. Björkegren, J., Packard, C.J., Hamsten, A., Bedford, D., 42. Steinberg, D., Parthasarathy, S., Carew, T.E., Khoo, J.C., and
Caslake, M., Foster, L., Shepherd, J., and Karpe, F. (1996) Ac-
cumulation of Large Very Low Density Lipoprotein in Plasma
During Intravenous Infusion of a Chylomicron-Like Triglyc-
Witztum, J.L. (1989) Modifications of Low-Density Lipopro-
tein That Increase Its Atherogenicity, N. Engl. J. Med. 320,
915–924.
eride Emulsion Reflects Competition for a Common Lipolytic 43. Couillard, C., Bergeron, N., Prud’homme, D., Bergeron, J.,
Pathway, J. Lipid Res. 37, 76–86.
39. Boquist, S., Hamsten, A., and Ruotolo, G. (2000) Insulin and
Nonesterified Fatty Acid Relations to Alimentary Lipaemia and
Tremblay, A., Bouchard, C., Mauriège, P., and Després, J.P.
(1998) Postprandial Triglyceride Response in Visceral Obesity
in Men, Diabetes 47, 953–960.
Plasma Concentrations of Postprandial Triglyceride-Rich 44. Taskinen, M.R. (1995) Insulin Resistance and Lipoprotein Me-
Lipoproteins in Healthy Middle-Aged Men, Diabetologia 43,
185–193.
40. Coresh, J., Kwiterovich, P.O., Smith, H.H., and Bachorik, P.S.
tabolism, Curr. Opin. Lipidol. 6, 153–160.
45. Björntorp, P. (1994) Fatty Acids, Hyperinsulinemia, and Insulin
Resistance: Which Comes First? Curr. Opin. Lipidol. 5, 166–174.
(1993) Association of Plasma Triglyceride Concentration and 46. Marchesini, G., Brizi, M., Morselli-Labate, A.M., Bianchi, G.,
LDL Particle Diameter, Density, and Chemical Composition
with Premature Coronary Artery Disease in Men and Women,
J. Lipid Res. 34, 1687–1697.
Bugianesi, E., McCullough, A.J., Forlani, G., and Melchionda,
N. (1999) Association of Nonalcoholic Fatty Liver Disease with
Insulin Resistance, Am. J. Med. 107, 450–455.
41. Dejager, S., Bruckert, E., and Chapman, M.J. (1993) Dense Low
Density Lipoprotein Subspecies with Diminished Oxidative Re-
sistance Predominate in Combined Hyperlipidemia, J. Lipid [Received August 14, 2000, and in revised form July 24, 2001; revi-
Res. 34, 295–308.
sion accepted August 13, 2001]
Lipids, Vol. 36, no. 10 (2001)