Aquaculture 233 (2004) 423 – 437

www.elsevier.com/locate/aqua-online

Replacement of dietary fish oil with palm

fatty acid distillate elevates tocopherol and
tocotrienol concentrations and increases
oxidative stability in the muscle of
African catfish, Clarias gariepinus
Wing-Keong Ng a,*, Yan Wang a,
Preyah Ketchimenin a, Kah-Hay Yuen b
a
Fish Nutrition Laboratory, School of Biological Sciences, Universiti Sains Malaysia,
Penang 11800, Malaysia
b
School of Pharmaceutical Sciences, Universiti Sains Malaysia, Penang 11800, Malaysia

Received 25 August 2003; received in revised form 14 October 2003; accepted 14 October 2003

Abstract

An 8-week feeding trial was conducted to determine the effect of substituting fish oil with palm
fatty acid distillate (PFAD), a by-product from crude palm oil refining, on growth, feed utilization,
vitamin E deposition, and oxidative stability in the muscles of African catfish. Five isonitrogenous,
isoenergetic, and isolipidic (10%) practical diets were formulated to contain 0%, 25%, 50%, 75%, or
100% substitution of fish oil with PFAD. Growth of African catfish was significantly ( P < 0.05)
better in fish fed the 25% PFAD diet compared to fish fed the control diet, with fish oil as the sole
lipid source. Higher levels of dietary PFAD supplementation did not cause further improvement in
growth performance of catfish. Muscle concentrations of a-tocopherol, a-tocotrienol, and g-
tocotrienol increased linearly in response to increasing dietary concentrations originating from the
added PFAD. a-Tocopherol constituted 68.5 – 80.2% of the vitamin E composition of catfish muscle,
with total tocotrienols maintaining an equilibrium of 13.4 – 26.7% irrespective of dietary
composition. Thiobarbituric acid-reactive substance (TBARS) concentrations in the muscles of
catfish decreased significantly with increasing levels of dietary PFAD. The present study showed that
the use of PFAD in catfish diets could offer protection against lipid peroxidation in muscles that
would translate to longer shelf-life for seafood products. In fish meal-based diets for African catfish,

* Corresponding author. Tel.: +60-4-6533888x4005; fax: +60-4-6565125.

E-mail address: wkng@usm.my (W.-K. Ng).

0044-8486/$ - see front matter D 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.aquaculture.2003.10.013
424 W.-K. Ng et al. / Aquaculture 233 (2004) 423–437

PFAD can totally replace added fish oil. The use of PFAD is a practical and cost-effective way to
produce high-energy diets without the damaging side effects of increased lipid radicals.
D 2004 Elsevier B.V. All rights reserved.

Keywords: Palm fatty acid distillate; Tocopherol; Tocotrienol; Lipid peroxidation; African catfish

1. Introduction

Fish oil is a major dietary lipid source used in commercial aquafeeds. In recent years,
technological advances in the aquafeed manufacturing industry have made possible the
incorporation of high levels of dietary oils in fish feeds to produce energy-dense diets.
Improvements in growth and feed utilization efficiency have been reported in fish due to
the protein-sparing effect of dietary lipid (DeSilva et al., 1991; Hemre and Sandnes, 1999;
Chaiyapechara et al., 2003). However, feeding high levels of dietary fish oils, which
contain a high proportion of polyunsaturated fatty acids (PUFAs) that are highly
susceptible to oxidation, can lead to increased oxidative stress for the fish that can result
in pathological conditions (Murai and Andrews, 1974; Sakai et al., 1998) and deterioration
of fillet quality (Scaife et al., 2000; Chaiyapechara et al., 2003). Increases in the lipid
content of commercial fish feeds are usually not followed by appropriate antioxidant
supplementation in order to maintain normal antioxidant status, which further exacerbates
the deleterious effects of lipid peroxidation especially in cellular biomembranes that
contain high amounts of PUFA.
One practical and cost-effective way to produce high-energy diets without the
damaging side effects of increased lipid radicals is to use a more saturated lipid source
that contains high levels of natural antioxidants such as vitamin E. Palm fatty acid distillate
(PFAD), a by-product from the physical refining of crude palm oil (Young, 1987; Ng et al.,
2003), is a potential substitute for fish oil in aquafeeds. In Malaysia alone, it was estimated
that about 402,500 metric tons of PFAD was generated for the year 2002, and with the
rapid increase in global palm oil production, the amount of PFAD produced will rise
substantially (Sundram et al., 2002). PFAD contains about 80% free fatty acids and 14.5%
acylglycerols (Ong and Choo, 1997), with a fatty acid composition of about 64% saturates,
30% monoenes, and 6% PUFA (Ng et al., 2003). Vitamin E is known to accumulate in the
PFAD fraction during palm oil processing and can be as high as 0.8%, making it one of the
richest natural sources of this antioxidant (Ong and Choo, 1997).
Vitamin E is the generic name given to a group of lipid-soluble compounds, which
include four tocopherols—a-tocopherol (a-T), h-tocopherol (h-T), g-tocopherol (g-T),
and y-tocopherol (y-T)—and four tocotrienols—a-tocotrienol, h-tocotrienol, g-tocotrie-
nol, and y-T3 isoforms. While most vegetable oils contain almost exclusively tocopherols,
palm oil is unique because tocotrienols represent about 80% of the vitamin E content
(Slover, 1971). For dietary purposes, each of these vitamin E isoforms is assigned a
biopotency factor based on the amount of vitamin E necessary to prevent fetal resorption
in pregnant, vitamin E-deficient rats (Sheppard and Pennington, 1993). When expressed as
D-a-tocopherol equivalents, the following values were assigned to the various isoforms: a-
T, 1.0; h-T, 0.5; g-T, 0.1; y-T, 0.03; a-T3, 0.3; h-T3, 0.05; g-T3 and y-T3, unknown. It is
 W.-K. Ng et al. / Aquaculture 233 (2004) 423–437 425

therefore not surprising that almost all vitamin E research in fish nutrition has concentrated
on a-T, usually supplied as the synthetic all-rac-a-tocopheryl acetate, as it is deemed the
most potent of all the isoforms (Frigg et al., 1990; Roem et al., 1990; Baker and Davies,
1996, 1997). There are a few reports on the deposition of other tocopherols in salmonids
(Sigurgisladottir et al., 1994; Hamre and Lie, 1997; Parazo et al., 1998), but there is
currently no previous studies on the deposition of palm tocotrienols in fish and its potential
role in enhancing fish fillet quality.
The essentiality of vitamin E in human and animal diets lies in its role as a potent
antioxidant that inhibits lipid peroxidation in biological membranes. Recent research
seems to indicate that the assigned vitamin E activity of the various isoforms may not
predict their role as biological antioxidants (Packer et al., 2001). In vitro studies of rat liver
microsomal membranes showed that a-T3 possesses 40 –60 times higher antioxidant
activity against lipid peroxidation and 6.5 times better protection of cytochrome P450
(a membrane protein) against oxidative damage than a-T (Serbinova et al., 1991). Ikeda et
al. (2003) reported that in some tissues in rats fed equivalent dietary levels of a-T or a-T3,
both isoforms provided equal protection against lipid peroxidation. The protective ability
of tocotrienols from palm oil was reported to be significantly higher compared to a-T as
effective inhibitors of protein oxidation and lipid peroxidation in rat liver microsomes
(Kamat et al., 1997), with g-T3 being the most effective.
The purpose of the present study was to evaluate the nutritive value of PFAD as a
potential dietary lipid source, at the expense of marine fish oil, using the African catfish as
an experimental model. The need to evaluate alternatives to fish oils is crucial for the
expanding global aquaculture industry considering the escalating costs and stagnation in
the world supply of fish oils. The effects of graded dietary levels of PFAD on growth, feed
utilization efficiency, and body composition of African catfish were investigated. Of major
emphasis in this study was the potential deposition of the tocotrienols and tocopherols
present in PFAD onto the fish muscle, thereby adding value to the resultant fish fillet
product.

2. Materials and methods

2.1. Diet preparation

Five isonitrogenous (36% crude protein) and isoenergetic (14.64 kJ/g diet) practical
diets were formulated (Table 1). Danish fish meal and solvent-extracted soybean meal
were used as protein sources in all diets. Diet lipid levels were fixed at 10%, with 4%
coming from residual oil in fish meal. The control diet was supplemented with 6%
menhaden fish oil and no PFAD. Subsequent diets were supplemented with 1.5%, 3.0%,
4.5%, or 6.0% of PFAD, at the expense of fish oil, which constituted 25%, 50%, 75%, and
100%, respectively, of dietary fish oil substitution. All diets contained 80 mg all-rac-a-
tocopheryl acetate/kg diet, which was added as part of the vitamin premix (Ng et al.,
2003). This more than meets the vitamin E requirement of African catfish, which was
estimated to be 30 –40 mg all-rac-a-tocopheryl acetate/kg diet (Baker and Davies, 1997).
All experimental diets were prepared as previously described by Ng et al. (2003) and
426 W.-K. Ng et al. / Aquaculture 233 (2004) 423–437

Table 1
Ingredient and proximate composition of the experimental diets (g/100 g dry diet)
% Substitution of fish oil with PFAD
0 25 50 75 100
Ingredient
Menhaden fish oil 6.00 4.50 3.00 1.50 0.00
PFADa 0.00 1.50 3.00 4.50 6.00
Danish fish meal 34.69 34.69 34.69 34.69 34.69
Soybean meal 20.91 20.91 20.91 20.91 20.91
Othersb 38.40 38.40 38.40 38.40 38.40

Proximate composition
Moisture 14.0 15.2 15.0 15.9 15.0
Crude protein 36.2 36.2 36.0 36.3 36.2
Crude lipid 10.5 10.5 10.4 9.8 10.2
Ash 9.2 9.2 9.3 9.3 9.3
Crude fiber 1.8 2.1 2.0 2.0 2.4
NFEc 42.3 42.0 42.3 42.6 41.9
a
Palm fatty acid distillate was obtained fresh from a local palm oil distillery and stored frozen at 20 jC
until used.
b
Includes corn flour (25.36%), mineral premix (3.00%; Ng et al., 2003), vitamin premix (2.00%; Ng et al.,
2003), dicalcium phosphate (1.00%), carboxymethyl cellulose (1.50%), and cellulose (5.54%).
c
Nitrogen-free extract = 100 (protein + lipid + ash + fiber).

stored in airtight freezer bags at 20 jC until needed. PFAD is in a semisolid state at

room temperature and was warmed slightly above its melting point before it was mixed
with the dry ingredients.

2.2. Experimental procedure

African catfish (Clarias gariepinus) fingerlings were obtained from a local fish farm
and held in a 1000-l fiberglass tank upon arrival at our laboratory. All fish were fed a
commercial catfish pellet for 2 weeks during this acclimatization period. At the com-
mencement of the feeding trial, groups of 10 catfish fingerlings (mean weight F S.E.,
5.31 F 0.02 g) were stocked into 95-l glass aquaria, and assigned one of the five
experimental diets. Each diet was fed close to apparent satiation at 4% of body weight
per day in two equal feedings to triplicate groups of catfish. Fish were batch-weighed
weekly by aquarium, and the daily feeding ration was adjusted accordingly. The aquaria
system setup, operation of the fish culture system, and water quality were as previously
described (Lim et al., 2001). The feeding trial was conducted for 8 weeks.

2.3. Sample collection and analysis

At the start of the experiment, an initial sample of 10 fish was taken and kept frozen at
20 jC for subsequent whole-body proximate analysis. Muscle tissue was excised from
another 10 fish for subsequent vitamin E analysis. At the end of the feeding trial, all fish
were weighed and five fish were randomly sampled from each aquarium. Blood was
 W.-K. Ng et al. / Aquaculture 233 (2004) 423–437 427

collected from each fish into heparinized capillary tubes by severing the caudal peduncle.
The capillary tubes were microcentrifuged and the relative volume of the packed red blood
cells was measured to determine the percent hematocrit value. The fish were then dissected
and their livers and viscera weighed for estimation of hepatosomatic index (HSI) and
viscerosomatic index (VSI). These indices were calculated as a percentage of organ or
tissue to the whole-body weight of individual fish. Viscera comprised liver, gastrointes-
tinal tract, and intraperitoneal fat. The fish was then skinned and muscle tissues removed,
pooled, sealed in parafilm, and stored frozen for subsequent vitamin E and thiobarbituric
acid-reactive substance (TBARS) analysis. The remaining fish were pooled and stored
frozen for whole-body composition analysis. Fish carcasses were blended, dried in an
oven, and ground into powder before proximate analysis.
The moisture, crude protein, lipid, fiber, and ash contents of diet ingredients,
experimental diets, and fish whole body were determined by standard Association of
Official Analytical Chemists (AOAC) (1997) methods.

2.4. Determination of tocopherol and tocotrienol concentrations

The tocopherol and tocotrienol concentrations of PFAD, diets, and muscles were
determined by high-performance liquid chromatography (HPLC) with fluorescence
detection. The HPLC system consisted of a 5-Am Kromasil silica column (250  4.6
mm; Metaphase, Kuala Lumpur, Malaysia) equipped with a guard column (Upchurch
Scientific, Oak Harbor, WA), a Rheodyne 7125 sample injector fitted with a 50-Al sample
loop, a Jasco PU-980 pump, and a Jasco FP-1520 Fluorescence Detector (Jasco, Tokyo,
Japan) with the excitation and emission wavelength set at 296 and 330 nm, respectively.
The chromatograms were recorded by a Hitachi D-2500 chromatointegrator (Hitachi,
Tokyo, Japan) at a chart speed of 2.5 mm/min. The isocratic mobile phase used was
hexane/tetrahydrofuran (100:4, vol/vol) at a flow rate of 2 ml/min. Tocotrienol peaks were
identified and quantified with the help of an in-house reference material extracted from
palm oil (TocominR 50%: Carotech, Malaysia), which was initially calibrated using a
tocotrienol standard kit purchased from Merck (Darmstadt, Germany). Tocopherol stand-
ards were purchased from Sigma-Aldrich (St. Louis, MO, USA).
For the determination of vitamin E concentrations in PFAD, the oil sample was
prepared according to the cold saponification procedure recommended for vegetable oils
as described by Pocklington and Dieffenbacher (1988), with slight modifications. Samples
of diet and muscle were prepared for HPLC analyses using a modification of the method
of Ikeda et al. (2003). The experimental diets were finely ground, weighed (0.2 g), and
transferred into a centrifuge tube, and 0.2 ml of 2% NaCl, 2 ml of ethanol containing
0.2% butylated hydroxytoluene (BHT), and 0.4 ml of 60% KOH were added and the tube
flushed with nitrogen and capped. The contents of the tube were thoroughly mixed and
saponified at 70 jC for 20 min. After cooling on ice, 2 ml of 2% NaCl was added and the
tocopherols and tocotrienols were extracted with 5 ml of hexane containing 10% (vol/vol)
ethyl acetate. For the muscle, tissue samples were homogenized in 4 ml of absolute
ethanol containing 0.2% BHT in an Ultra-Turrax tissue disrupter (IKA, Malaysia). The
tissue homogenate (2 ml) was pipetted into a centrifuge tube, 0.2 ml of 60% KOH was
added, the tube was flushed with nitrogen, then capped, vortexed, and saponified at 70 jC
428 W.-K. Ng et al. / Aquaculture 233 (2004) 423–437

for 20 min. After cooling on ice, 2 ml of 2% NaCl was added and the tocopherols and
tocotrienols were extracted with 3 ml of hexane containing 10% (vol/vol) ethyl acetate.
All operations were conducted in subdued lighting conditions and all solvents were kept
on ice until used.

2.5. Determination of TBARS concentrations in muscle

The oxidative stability of the fish muscle was determined by the TBARS test according
to the procedure outlined in Baker and Davies (1996). Malondialdehyde (MDA), which is
formed during the breakdown of PUFA, was used as an index for determining the extent of
lipid peroxidation. Iron and ascorbic acid were used to induce lipid peroxidation in
homogenized muscle samples and the resultant MDA reacted with 2-thiobarbituric acid.
The color intensity of the MDA – TBA adduct was measured spectrophotometrically at 535
nm, and the MDA concentration of the sample was calculated from a calibration curve
using MDA standards.

2.6. Statistical analysis

Final fish weight, percentage weight gain, feed efficiency ratio, protein efficiency ratio,
net protein utilization (NPU), whole-body composition, HSI, VSI, hematocrit, TBARS, as
well as vitamin E concentrations and composition were all subjected to one-way analysis
of variance (SAS Institute, NC, USA) to determine if significant differences occurred
among the dietary treatments. Differences between means were assessed by Duncan’s
multiple range test (Duncan, 1955). Effects with a probability of P < 0.05 were considered
significant. Regression analysis between dietary and muscle tocopherol and tocotrienol
concentrations was performed using Lotus 123 statistical software.

3. Results

3.1. Vitamin E concentration and composition of experimental diets

The analyzed proximate composition of the experimental diets (Table 1) was consistent
with formulated values and did not vary among the five diets. The tocopherol and
tocotrienol concentrations of the diets generally increased with increasing levels of dietary
PFAD (Table 2). The PFAD used in the present study contained 3889.2 F 165.3 Ag
vitamin E/g oil, with 79.5% as tocotrienols. Dietary g-T3, the major tocotrienol in PFAD,
increased from 1.4 Ag/g in the control diet with no PFAD to 104.5 Ag/g in the diet with
100% substitution of fish oil with PFAD (6% PFAD). y-T3 and a-T3 were not detected in
the 0% PFAD diet, but increased 30- and 40-fold, respectively, in the 100% PFAD diet.
Low levels of h- and g-T were detected in all diets, but a-T, the major tocopherol in
PFAD, increased at graded levels from 74.8 to 118.7 Ag/g diet, with increasing dietary
levels of PFAD. y-T and h-T3 were not detected in all diets. Total vitamin E content in the
diets increased from 78.6 to 304.8 Ag/g when dietary PFAD was increased from 0% to
100% of the supplemented oil.
 W.-K. Ng et al. / Aquaculture 233 (2004) 423–437 429

Table 2
Total tocopherols and tocotrienols of the experimental diets
Vitamin Ea % Substitution of fish oil with PFADb
0 25 50 75 100
Concentration (lg/g dry diet)
a-T 74.88 86.95 91.74 98.86 118.72
h-T 1.07 1.32 1.48 0.98 1.09
g-T 1.37 2.77 4.06 5.32 7.52
a-T3 0.00 11.62 17.03 28.98 40.53
g-T3 1.36 26.96 45.39 72.31 104.55
y-T3 0.00 6.53 12.40 20.31 32.40
Total T + T3 78.68 136.16 172.10 226.76 304.81

Composition (%)
a-T 95.1 63.8 53.3 43.6 38.9
h-T 1.4 1.0 0.9 0.4 0.4
g-T 1.7 2.0 2.3 2.4 2.5
a-T3 0.0 8.6 9.9 12.8 13.3
g-T3 1.8 19.8 26.4 31.9 34.3
y-T3 0.0 4.8 7.2 8.9 10.6
Total T 98.2 66.8 56.5 46.4 41.8
Total T3 1.8 33.2 43.5 53.6 58.2
a
a = Alpha, h = beta, g = gamma, and y = delta tocopherols (T) and tocotrienols (T3). y-T and h-T3 were not
detected in the diet samples.
b
Palm fatty acid distillate. Contains 3889.2 F 165.3 Ag total vitamin E/g oil, with 79.5% consisting of
tocotrienols and 20.4% tocopherols.

When expressed as a percentage, tocotrienols constituted 1.8%, 33.2%, 43.5%, 53.6%,

and 58.2% of the total vitamin E content of the 0%, 25%, 50%, 75%, and 100% PFAD
diet, respectively (Table 2). The major vitamin E isoform present in all diets was a-T,
which constituted 95.1% of the total vitamin E in the control diet down to 38.9% in the
100% PFAD diet.

3.2. Growth performance and body composition of African catfish

African catfish fingerlings fed the control diet with fish oil as the sole lipid source
had significantly ( P < 0.05) lower final weight and percentage weight gain compared to
fish fed the 25% PFAD diet (Table 3). The growth response of catfish fed diets with up
to 100% substitution of fish oil with PFAD was not significantly different. However,
there appeared to be a slight reduction in growth performance with increasing dietary
PFAD, and the NPU of catfish fed the 100% PFAD diet was significantly lower
compared to the NPU of catfish fed the 25% PFAD diet. Feed efficiency ratio and
protein efficiency ratio did not differ significantly among fish fed the various diets. All
fish appeared healthy at the end of the experiment and there was no mortality during the
experimental period.
Whole-body composition of catfish fed the various diets was not significantly different,
with the exception of body lipid content, which was significantly lower in fish fed the
100% PFAD diet (Table 4). HSI, VSI, and hematocrit of catfish fed increasing dietary
430 W.-K. Ng et al. / Aquaculture 233 (2004) 423–437

Table 3
Growth performance and feed utilization efficiency of African catfish fingerlings fed increasing levels of PFAD
for 8 weeksa
% Fish oil substitution Final weight (g) Weight gainb(%) FERc PERd NPUe(%)
b b
0% PFAD 28.1 426.8 1.03 2.77 44.0a,b
25% PFAD 36.2a 581.2a 1.11 3.07 49.4a
50% PFAD 32.4a,b 512.8a,b 1.06 2.96 45.6a,b
75% PFAD 29.6a,b 458.2a,b 0.92 2.54 41.1a,b
100% PFAD 29.5a,b 458.1a,b 0.89 2.47 38.5b
Pooled S.E.M. 1.4 27.2 0.04 0.12 1.9
a
Values are the mean of triplicate groups of 10 fish. Average initial body weight of individual fish was
5.31 + 0.02 g. Mean values in columns with different superscripts are significantly different ( P < 0.05).
b
Expressed as the percent of initial body weight at the end of 8 weeks.
c
Feed efficiency ratio = wet weight gain (g)/total dry weight of diet fed (g).
d
Protein efficiency ratio = wet weight gain (g)/total protein intake (g).
e
Net protein utilization (%) = 100  (final initial fish body protein)/total protein intake.

levels of PFAD were not significantly different compared to fish fed the control diet
(Table 4).

3.3. Muscle tocopherol and tocotrienol concentrations

The major vitamin E isoform deposited in the muscle of African catfish was a-T (Table
5). The concentrations of a-T were significantly different among the muscle of catfish fed
the various diets with a 2.3 times higher concentration in fish fed the 100% PFAD diet
(13.9 Ag/g tissue) compared to fish fed the 0% PFAD diet (6.0 Ag/g tissue). It was
observed that after 8 weeks on the 0%, 25%, or 50% PFAD diet, the concentration of a-T
was slightly lower in the muscle compared to the initial value at the commencement of the
feeding trial. Muscle concentration of a-T had a direct and linear relationship with dietary
a-T concentrations ( y = 0.20x 10.07, R2 = 0.86). The concentrations of all non-a-T in the
muscle were low and did not rise above initial values found at the start of the feeding trial.

Table 4
Whole-body composition (% wet weight), hepatosomatic index, viscerosomatic index, and hematocrit of African
catfish fed increasing levels of PFAD for 8 weeksa
% Fish oil substitution Whole body HSIb VSIc Hematocrit (%)
Moisture Ash Protein Lipid
0% PFAD 71.7 3.6 15.6 7.2a 0.91 2.27 38.0
25% PFAD 71.9 3.4 15.9 6.8a,b 0.89 2.33 40.8
50% PFAD 71.8 3.6 15.3 7.2a 0.93 2.55 39.0
75% PFAD 72.2 3.4 15.6 7.2a 0.93 2.70 42.1
100% PFAD 73.6 3.4 15.4 6.1b 1.00 2.61 38.8
Pooled S.E.M. 0.4 0.04 0.1 0.22 0.02 0.08 0.7
a
Values are the mean of triplicate groups of five fish. Mean values in columns with different superscripts are
significantly different ( P < 0.05). Initial whole-body composition was 75.2% moisture, 3.4% ash, 14.8% protein,
and 5.4% lipid.
b
HSI=[100  liver weight (g)]/body weight (g).
c
VSI=[100  visceral weight (g)]/body weight (g).
 W.-K. Ng et al. / Aquaculture 233 (2004) 423–437 431

Table 5
Muscle concentration (Ag/g tissue) of tocopherols and tocotrienols and their relative composition (% of total
vitamin E) in African catfish fingerlings fed increasing levels of PFAD for 8 weeksa
Vitamin Eb Initial % Substitution of fish oil with PFAD S.E.M.
0 25 50 75 100
Concentration (lg/g tissue)
a-T 10.00 6.00d 6.41c,d 6.81c 11.25b 13.87a 1.56
h-T 0.48 0.32 0.32 0.31 0.41 0.39 0.02
g-T 0.16 0.16a 0.16a 0.16a 0.03b 0.00b 0.04
a-T3 0.36 0.24c 0.64c 0.74b,c 1.25b 2.01a 0.30
g-T3 0.00 0.76c 1.34b,c 1.67b 1.90b 2.76a 0.32
y-T3 0.00 0.00b 0.13a 0.24a 0.14a 0.23a 0.04
Total T + T3 11.00 7.48d 9.01c 9.94c 14.98b 19.25a 2.18

Composition (%)
a-T 90.9 80.2a 71.2b,c 68.5c 75.4a,b 72.1b,c 2.0
h-T 4.4 4.3a 3.6a,b 3.2b 2.8bc 2.0c 0.4
g-T 1.4 2.2a 1.8b 1.6b 0.2c 0.0c 0.4
a-T3 3.3 3.1b 7.1a 7.5a 8.2a 10.4a 1.2
g-T3 0.0 10.3c 14.8a,b 16.8a 12.5b,c 14.3a,b 1.1
y-T3 0.0 0.0c 1.4a,b 2.4a 1.0b,c 1.2a,b,c 0.4
Total T 96.7 86.6a 76.6b 73.3b 78.4b 74.1b 2.4
Total T3 3.3 13.4b 23.4a 26.7a 21.7a 25.9a 2.4
a
Values are the mean of triplicate groups of four to five fish. Mean values in rows with different superscripts
are significantly different ( P < 0.05).
b
a = Alpha, h = beta, g = gamma, and y = delta tocopherols (T) and tocotrienols (T3). y-T and h-T3 could not
be detected in the muscle samples.

The g-T3 and a-T3 concentrations in catfish muscle were linearly related to dietary
concentrations ( y = 0.02x + 0.77 and y = 0.04x + 0.14, respectively) with a good fit
(R2 = 0.97 for both) (Table 5). Similar levels of g-T3 and a-T3 were deposited in the
muscle of fish fed the 75% or 100% PFAD diet despite dietary g-T3 concentrations being
more than twice the amounts of a-T3. Low levels of y-T3 were detected in all samples
except in the muscle of fish fed the control diet. Total tocotrienols increased linearly from
1.0 to 5.0 Ag/g tissue when dietary PFAD substituted 0– 100% of added fish oil.
Total vitamin E concentrations in catfish muscle were reflective of the concentration
found in the respective diet ( y = 0.06x + 2.02, R2 = 0.95), but the same was not true of the
vitamin E composition (Table 5). Even though the percentage of tocotrienols increased
from 33.2% to 58.2% of the total dietary vitamin E with increasing supplementation of
PFAD (Table 4), muscle tocotrienol composition remained somewhat constant at 21.7 –
26.7% in the muscle of catfish. The only significant increase in percent tocotrienols (and
corresponding decrease in tocopherols) in muscles was observed in fish fed the 25% PFAD
diet compared to fish fed the control diet without any added PFAD. a-T constituted 68.5 –
80.2% of the vitamin E composition of catfish muscles, with g-T3 a distant second at
10.3 –16.8%. All other vitamin E isoforms detected contributed less than 10% each to the
vitamin E concentrations found in catfish muscle. The percentage contribution of h-T and
g-T in muscle vitamin E content significantly decreased with increasing supplementation
of PFAD.
432 W.-K. Ng et al. / Aquaculture 233 (2004) 423–437

Fig. 1. Thiobarbituric acid-reactive substances in African catfish muscles from fish fed diets with increasing levels
of PFAD expressed as nanomoles of MDA per gram of tissue. Values are mean F S.E., n = 3. Means with different
letters are significantly different ( P < 0.05).

3.4. Oxidative stability of fish muscle

The TBARS concentration (expressed as nmol MDA/g tissue) in the muscles of catfish
fed PFAD at all levels of dietary inclusion were significantly lower ( P < 0.05) than those
of fish fed the control diet with fish oil as the sole dietary lipid source (Fig. 1). The
oxidative stability of muscle from catfish fed the 100% PFAD diet was significantly better
compared to fish fed the 25% or 50% PFAD diet.

4. Discussion

Traditionally, high-quality marine fish oils have been used almost exclusively as
sources of dietary lipids in the formulation of commercial fish feeds. However, recent
research has shown that substantial quantities of vegetable oils, including palm oils, can be
used to substitute fish oil in fish diets without negatively affecting growth (Tortensen et al.,
2000; Lim et al., 2001; Rosenlund et al., 2001; Bell et al., 2002; Ng, 2002; Ng et al.,
2003).
African catfish fed diets containing only fish oil as the sole lipid source showed
significantly lower weight gain compared to fish fed diets where the added fish oil was
blended with PFAD in a 3:1 ratio. The results of the present study are in agreement with
previous studies (Hoffman and Prinsloo, 1995; Ng et al., 2003) that also reported growth
depression in African catfish fed high levels of cod liver oil. The present study lends
 W.-K. Ng et al. / Aquaculture 233 (2004) 423–437 433

further credence to our earlier suggestion that the high n 3 PUFA present in fish oils is
the probable cause of the growth depression and that there exist an optimum ratio between
n 3 and n 6 fatty acids for maximum growth of the African catfish (Ng et al., 2003).
Increasing dietary PFAD in the diets of African catfish did not significantly affect growth
performance, but there was a general trend of decreased feed efficiency leading to a slight
depression in growth. Being a by-product, PFAD has a wide range of quality and is known
to contain trace amounts of impurities such as high-molecular-weight ketones that might
decrease palatability to animals (Wong, 1983). Although there were no marked changes in
body composition, body indices such as HSI and VSI, or in the hematocrit of catfish fed
diets with high levels of PFAD, further research would be necessary to determine the
antinutritive and toxicological effects, if any, of this by-product.
As far as we know, the present study is the first reported data on the deposition of
dietary palm tocotrienols in fish tissues. Muscle tocotrienol concentrations of African
catfish were observed to increase significantly concomitant with increasing dietary PFAD.
However, when tocotrienol concentrations were expressed as a percentage of total vitamin
E, it was interesting to note that despite an increasing percentage of tocotrienols in the diet
(1.8 –58.2%), tocotrienols constituted only 13.4 –26.7% of the total vitamin E deposited in
catfish muscle. In catfish fed PFAD-supplemented diets, an equilibrium in T:T3 ratio of
about 7.5:2.5 in the muscle was reached in 8 weeks irrespective of dietary vitamin E
composition. In the catfish muscle, 68.5 – 80.2% of the total vitamin E deposited was
present as a-T. Similar high ratios of a-T to the sum of other vitamin E isoforms in tissues
also have been reported for laboratory land animals (Goh et al., 1992; Ikeda et al., 2003).
The reasons for the preferential deposition of a-T in animal tissues are currently the
subject of active investigation. Based on current knowledge, there is consensus among
scientists that biodiscrimination takes place largely in the liver where a a-tocopherol
transfer protein (a-TTP) has been identified in rats (for review, see Packer et al., 2001). a-
TTP has high affinity for a-T and preferentially incorporates this vitamin E isoform into
lipoproteins that transport a-T to the various tissues. Our data seems to suggest that a
similar a-T binding protein may exist in catfish liver, although the isolation of such a
protein has not been reported in fish. Studies with dietary g-tocopherols and y-tocopherols
in Atlantic salmon have shown the preferential deposition of a-T in fish muscles over the
non-a-T isoforms (Hamre and Lie, 1997; Parazo et al., 1998). The existence of a
mechanism similar to that present in mammals for biodiscrimination between the
tocopherols mediated through competition for a a-TTP in salmon liver was similarly
presumed by the authors. Goh et al. (1992) postulated that apart from the high affinity of
a-TTP for a-T, the lower retention of tocotrienols in animal tissues might also be due to
them being more reactive antioxidants in vivo, which accounts for their faster disappear-
ance after absorption, and presented indirect evidence that in vivo conversion of
tocotrienols to tocopherols may be likely in the rabbit. Other mechanisms of biodiscrimi-
nation at the cellular membranes of specific tissues and other routes of vitamin E delivery
also have been proposed to account for the varying deposition of vitamin E isoforms in
various tissues (Parazo et al., 1998; Ikeda et al., 2003).
Assuming the existence of a a-TTP in catfish liver, the results of the present study
indicated that there exists a different relative affinity of the three tocotrienols for a-TTP.
For example, in catfish fed the 100% PFAD diet, dietary concentration of g-T3 was 2.5
434 W.-K. Ng et al. / Aquaculture 233 (2004) 423–437

times greater than a-T3, but similar concentrations of g-T3 and a-T3 were detected in the
muscles, indicating that the relative affinity of g-T3 for a-TTP is lower than that of a-T3.
A similar conclusion was made by Ikeda et al. (2003) in rats fed with either g-T3 or a-T3.
Based on urinary excretion of tocotrienol metabolites, they further suggested that g-T3
might be preferentially metabolized in rats. Dietary concentrations of y-T3 were only
slightly lower than a-T3 in the present study, but very small quantities of y-T3 were
deposited in catfish muscle. In catfish fed the 75% or 100% PFAD diet, muscle
concentrations of y-T3 compared to a-T3 were 10 times lower. Indirect evidence from
the present study seems to indicate that the relative affinity of a-T3 for a-TTP is the
highest, followed by g-T3 and then y-T3.
Low concentrations of h-T and g-T were present in the experimental diets and similar
low levels were detected in catfish muscles. Muscle concentrations of h-T and g-T in fish
fed the various diets did not exceed initial muscle values after 8 weeks on the diets,
indicating that the two isoforms could not compete for the a-TTP especially since the
dietary concentrations of a-T were overwhelmingly higher. With the exception of
Sigurgisladottir et al. (1994), who found g-T and a-T to be deposited to the same extent
in Atlantic salmon muscle, other later authors (Hamre and Lie, 1997; Parazo et al., 1998)
have reported significantly lower deposition of g-T, h-T, and y-T compared to a-T in
salmon tissues.
The suggested relative affinities of the various vitamin E isoforms for a-TTP as
indicated by the data from the present study are consistent with their assigned
biopontency values (Sheppard and Pennington, 1993). There is currently consensus
among vitamin E researchers that the low biological activities of tocotrienols and non-a-
tocopherols may be due to their low affinity for a-TTP, rather than the actual antioxidant
potency of the various vitamin isoforms (Packer et al., 2001). Research has clearly
shown that the antioxidant efficacy of tocotrienols in cell membranes is higher than that
of tocopherols in vitro (Serbinova et al., 1991; Kamat et al., 1997; Packer et al., 2001).
If higher levels of tocotrienols can be deposited in fish tissues, tocotrienols would be
more effective antioxidants than tocopherols in vivo. The presence of a-T has been
reported to decrease tocotrienol retention in rat tissues (Ikeda et al., 2003). In the present
study, all diets were supplemented with equal amounts of all-rac-a-tocopheryl acetate on
top of the a-T naturally occurring in PFAD and other dietary ingredients. Further
research to investigate the deposition of tocotrienols in fish fed a tocotrienol-rich fraction
extracted from PFAD as the sole vitamin E source is currently being carried out in our
laboratory.
Elevated concentrations of a-T in African catfish (Baker and Davies, 1996, 1997) and
other fish species (Frigg et al., 1990; Scaife et al., 2000; Chaiyapechara et al., 2003)
have been reported to improve oxidative stability of fish fillets. To date, all studies on
the role of vitamin E in decreasing lipid peroxidation in fish tissues had depended on the
use of synthetic all-rac-a-tocopheryl acetate as the sole dietary source of vitamin E. We
have previously shown that a-T concentrations in African catfish muscle increased
linearly in response to increasing dietary a-T originating from crude palm oil (Lim et al.,
2001). A similar linear increase in total vitamin E concentrations in the muscles of
African catfish fed increasing dietary levels of PFAD was also observed in the present
study. However, the contribution of the palm oil-based vitamin E deposited in catfish
 W.-K. Ng et al. / Aquaculture 233 (2004) 423–437 435

muscles in enhancing the oxidative stability of the tissues could not be conclusively
shown in the present study, although we believe it does play a significant role. With
increasing dietary PFAD, it is anticipated that the fatty acid composition of the muscle
would be less unsaturated compared to the muscles of catfish fed higher levels of fish
oil. It is generally known that the fatty acid composition of fish muscle lipids reflects the
fatty acid composition of the dietary oils used (Tortensen et al., 2000; Rosenlund et al.,
2001; Bell et al., 2002) and a similar correlation also had been previously reported for
African catfish fed various palm oil products (Ng et al., 2003). The lower lipid
peroxidation observed in the muscle of catfish fed increasing dietary levels of PFAD
could be due to both the decreasing unsaturation of muscle lipids as well as the elevated
concentrations of vitamin E. Further research is currently being conducted in our
laboratory to elucidate the antioxidant potency of palm-based vitamin E compared to
synthetic a-tocopheryl acetate.
PFAD is being suggested as a practical and cost-effective substitute for fish oils. Since
oils with high PUFA content are highly susceptible to oxidation and rancidity, the oxidative
stability of both the feed pellets and fish fillets can be substantially improved with the use
of PFAD. PFAD is also a rich source of vitamin E, having both tocotrienols and
tocopherols, which are highly potent natural antioxidants, and its bioaccumulation in fish
fillets can impart antioxidative properties that would translate to longer shelf-life for
seafood products. The deposition of tocotrienols in fish fillets also adds value to the product
since the potential health benefits of tocotrienols in the human diet may include beneficial
effects on the prevention of cardiovascular diseases and cancer (Watkins et al., 1999).

Acknowledgements

We would like to thank the Malaysian Ministry of Science, Technology, and

Environment (MOSTE) for the research grant (IRPA Project no. 01-02-05-2216 EA008)
that was provided for this study. We would also like to thank David Ho (Carotech) for
providing the tocotrienol internal reference standard. The technical input of Irene Yap for
the tocotrienol analysis is greatly appreciated.

References

Association of Official Analytical Chemists (AOAC), 1997. In: Cunniff, P.A. (Ed.), Official Methods of Analysis
of AOAC International, 16th ed. OAC International, Arlington, VA.
Baker, R.T.M., Davies, S.J., 1996. Changes in tissue a-tocopherol status and degree of lipid peroxidation with
varying a-tocopherol acetate inclusion in diets for the African catfish. Aquacult. Nutr. 2, 71 – 79.
Baker, R.T.M., Davies, S.J., 1997. The quantitative requirement for a-tocopherol by juvenile African catfish,
Clarias gariepinus Burchell. Anim. Sci. 65, 135 – 142.
Bell, J.G., Henderson, R.J., Tocher, D.R., McGhee, F., Dick, J.R., Porter, A., Smullen, R.P., Sargent, J.R., 2002.
Substituting fish oil with crude palm oil in the diet of Atlantic salmon (Salmo salar) affects muscle fatty acid
composition and hepatic fatty acid metabolism. J. Nutr. 132, 222 – 230.
Chaiyapechara, S., Casten, M.T., Hardy, R.W., Dong, F.M., 2003. Fish performance, fillet characteristics, and
health assessment index of rainbow trout (Oncorhynchus mykiss) fed diets containing adequate and high
concentrations of lipid and vitamin E. Aquaculture 219, 715 – 738.
436 W.-K. Ng et al. / Aquaculture 233 (2004) 423–437

DeSilva, S.S., Gunasekera, R.M., Shim, K.F., 1991. Interaction of varying protein and lipid levels in young red
tilapia: evidence of protein sparing. Aquaculture 95, 305 – 318.
Duncan, D., 1955. Multiple range tests and multiple F tests. Biometrics 11, 1 – 42.
Frigg, M., Prabucki, A.L., Buhdel, E.U., 1990. Effect of dietary vitamin E levels on oxidative stability of trout
fillets. Aquaculture 84, 145 – 158.
Goh, S.H., Hew, N.F., Khor, H.T., 1992. An apparent in vivo conversion of tocotrienols to tocopherols in the
rabbit. Malaysian Oil Sci. Technol. 1, 56 – 59.
Hamre, K., Lie, O., 1997. Retained levels of dietary a-, g- and y-tocopherol in tissues and body fluids of Atlantic
salmon, Salmo salar, L. Aquacult. Nutr. 3, 99 – 107.
Hemre, G.I., Sandnes, K., 1999. Effect of dietary lipid level on muscle composition in Atlantic salmon Salmo
salar. Aquacult. Nutr. 5, 9 – 16.
Hoffman, L.C., Prinsloo, J.F., 1995. The influence of different dietary lipids on the growth and body composition
of the African sharptooth catfish, Clarias gariepinus (Burchell). S. Afr. J. Sci. 91, 315 – 320.
Ikeda, S., Tohyama, T., Yoshimura, H., Hamamura, K., Abe, K., Yamashita, K., 2003. Dietary a-tocopherol
decreases a-tocotrienol but not g-tocotrienol concentration in rats. J. Nutr. 133, 428 – 434.
Kamat, J.P., Sarma, H.D., Devasagayam, T.P.A., Nesaretnam, K., Basiron, Y., 1997. Tocotrienols from palm oil
as effective inhibitors of protein oxidation and lipid peroxidation in rat liver microsomes. Mol. Cell. Biochem.
170, 131 – 138.
Lim, P.K., Boey, P.L., Ng, W.K., 2001. Dietary palm oil level affects growth performance, protein retention and
tissue vitamin E concentration of African catfish, Clarias gariepinus. Aquaculture 202, 101 – 112.
Murai, T., Andrews, J.W., 1974. Interaction of dietary a-tocopherol, oxidized menhaden oil and ethoxyquin on
channel catfish (Ictalurus punctatus). J. Nutr. 104, 1416 – 1431.
Ng, W.K., 2002. Potential of palm oil utilisation in aquaculture feeds. Asia Pacific J. Clin. Nutr. 11, 473 – 476.
Ng, W.K., Lim, P.K., Boey, P.L., 2003. Dietary lipid and palm oil source affects growth, fatty acid compo-
sition and muscle a-tocopherol concentration of African catfish, Clarias gariepinus. Aquaculture 215,
229 – 243.
Ong, A.S.H., Choo, Y.M., 1997. Carotenoids and tocols from palm oil. In: Shahidi, F. (Ed.), Natural Antioxidants:
Chemistry, Health Effects, and Application. American Oil Chemists Society, Champaign, IL, pp. 133 – 149.
Packer, L., Weber, S.U., Rimbach, G., 2001. Molecular aspects of a-tocotrienol antioxidant action and cell
signalling. J. Nutr. 131, 369 – 373.
Parazo, M.R.M., Lall, S.P., Castell, J.D., Ackman, R.G., 1998. Distribution of a- and g-tocopherols in Atlantic
salmon (Salmo salar) tissues. Lipids 33, 697 – 704.
Pocklington, W.D., Dieffenbacher, A., 1988. Determination of tocopherols and tocotrienols in vegetable oils and
fats by high performance liquid chromatography. Pure Appl. Chem. 60, 877 – 892.
Roem, A.J., Kohler, C.C., Stickney, R.R., 1990. Vitamin E requirements of the blue tilapia, Oreochromis aureus
(Steindachner), in relation to dietary lipid level. Aquaculture 87, 155 – 164.
Rosenlund, G., Obach, A., Sandberg, M.G., Standal, H., Tveit, K., 2001. Effect of alternative lipid sources on
long-term growth performance and quality of Atlantic salmon (Salmo salar L.). Aquacult. Res. 32, 323 – 328.
Sakai, T., Murata, H., Endo, M., Shimomura, T., Yamauchi, K., Ito, T., Yamaguchi, T., Nakajima, H., Fuku-
dome, M., 1998. Severe oxidative stress is thought to be a principle cause of jaundice of yellowtail Seriola
quinqueradiata. Aquaculture 160, 205 – 214.
Scaife, J.R., Onibi, G.E., Murray, I., Fletcher, T.C., Houlihan, D.F., 2000. Influence of a-tocopherol acetate on
the short- and long-term storage properties of fillets from Atlantic salmon Salmo salar fed a high lipid diet.
Aquacult. Nutr. 6, 65 – 71.
Serbinova, E., Kagan, V., Han, D., Packer, L., 1991. Free radical recycling and intramembrane mobility in the
antioxidant properties of alpha-tocopherol and alpha-tocotrienol. Free Radic. Biol. Med. 10, 263 – 275.
Sheppard, A.J., Pennington, J.A.T., 1993. Analyses and distribution of vitamin E in vegetable oils and foods. In:
Packer, L., Fuchs, J. (Eds.), Vitamin E in Health and Disease. Marcel Dekker, New York, pp. 9 – 31.
Sigurgisladottir, S., Parrish, C.C., Ackman, R.G., Lall, S.P., 1994. Tocopherol deposition in the muscle of
Atlantic salmon (Salmo salar). J. Food Sci. 59, 256 – 259.
Slover, H.T., 1971. Tocopherols in foods and fats. Lipids 6, 291 – 296.
Sundram, K., Thiagarajan, T., Gapor, A., Basiron, Y., 2002. Palm tocotrienols: new antioxidants for the new
millenium. Inform 13, 634 – 640.
 W.-K. Ng et al. / Aquaculture 233 (2004) 423–437 437

Tortensen, B.E., Lie, O., Froyland, L., 2000. Lipid metabolism and tissue composition in Atlantic salmon (Salmo
salar L.)—effects of capelin oil, palm oil and oleic-enriched sunflower oil as dietary lipid sources. Lipids 35,
653 – 664.
Watkins, T.R., Bierenbaum, M.L., Giampaolo, A., 1999. Tocotrienols: biological and health effects. In: Papas,
A.M. (Ed.), Antioxidant Status, Diet, Nutrition and Health. CRC Press LLC, Boca Raton, FL, pp. 479 – 496.
Wong, M.H., 1983. Quality of byproducts from chemical and physical refining of palm oil and other oils. J. Am.
Oil Chem. Soc. 60, 316 – 321.
Young, F.V.K., 1987. Refining and fractionation of palm oil. Gunstone, F.D. (Ed.), Palm Oil: Critical Reports on
Applied Chemistry, vol. 15. Wiley, New York, pp. 39 – 69.