ENVIRONMENT AND HEALTH

Intermittent Lighting Reduces the Incidence of Ascites in Broilers:

An Interaction with Protein Content of Feed on Performance
and the Endocrine System
N. BUYS,1 J. BUYSE, M. HASSANZADEH-LADMAKHI, and E. DECUYPERE
Laboratory for Physiology and Immunology of Domestic Animals,
K.U. Leuven, Kardinaal Mercierlaan 92, B-3001 Heverlee, Belgium

ABSTRACT An experiment with 840 day-old male
broiler chicks (Ross) was initiated to investigate the
effect of intermittent lighting schedules, combined with
two isocaloric feeds differing in protein content, on
ascites mortality. At 9 d of age, chicks were randomly
distributed over two rooms: one room with a 23 h
light(L):1 h dark (D) lighting schedule (CL), and another
room with an intermittent lighting schedule (IL, 1L:3D).
In each room, two isocaloric feeds differing in CP
content, supplemented or not supplemented with 1.5
ppm triiodothyronine (T3), were provided. The experi-
ment was repeated under identical conditions except
that lighting schedules were changed between rooms.
From Day 14 onwards, biweekly growth and feed intake
were determined and feed conversion (FCR) was
calculated. Daily mortality was recorded and necropsies
were performed. Weekly blood samples were taken
(Key words: ascites, broiler, intermittent light, hormones, mortality)
from 10 randomly chosen birds per experimental group
for measurements of growth hormone (GH), thyroid
hormones [thyroxine (T4) and triiodothyronine (T3)],
and hematocrit values.
Birds reared in IL manifested a more concave growth
trajectory than those in CL, which was associated with a
lower FCR and higher plasma GH levels during the
finisher period. Mortality due to ascites was markedly
increased by dietary T3 supplementation. In hyper-
thyroid chickens, ascites mortality was lower in IL than
in CL and lower in birds fed normal protein than in
those fed subnormal protein levels. Dietary T3 decreased
plasma GH and T4 levels, whereas T3 levels were
increased. It is concluded that IL or a higher protein
content of the feed reduces the incidence of T3-induced
ascites mortality. Possible causal mechanisms are dis-
cussed.
1998 Poultry Science 77:54–61

INTRODUCTION 3D) improve FCR and reduce abdominal fat content

Modern broilers are intensively selected for fast
growth rate combined with a low feed conversion ratio
(FCR). However, adverse selection responses, such as
augmented fat deposition, and an increased incidence of
metabolic diseases, such as ascites or sudden death
syndrome, have also occurred. An imbalance between
oxygen supply and oxygen need can be the cause of
hypoxemia, cardiopulmonary dysfunction, and ascites
(Scheele et al., 1992; Julian, 1993; Maxwell et al., 1995).
The long-term solution for ascites in broilers should be
found in selection programs, but for a short-term
solution, much research has focused on the possible
beneficial effect of different management strategies in
reducing the incidence of ascites.
Intermittent lighting programs (IL) consisting of
repeated cycles of, e.g., 1 h light and 3 h darkness (1L:
relative to these parameters in broiler chickens raised
under nearly continuous illumination (Buyse et al.,
1994a, 1996). Additionally, during the dark period of
each dark-light cycle, heat production is markedly
decreased (Buyse et al., 1994b). Because it is claimed that
all factors that increase metabolic rate will predispose
birds to develop ascites (Albers and Frankenhuis, 1990),
the use of intermittent light programs might offset this
tendency.
The purpose of this study was to investigate the effect
of the interaction of intermittent lighting schedules with
a diet containing suboptimal levels of protein on the
incidence of ascites, performance, and hormonal
parameters. Dietary triiodothyronine (T3) was used to
increase ascites mortality (Buys et al., 1993; Decuypere et
al., 1994) and, hence, the discrimination power of the
two other experimental variables.

Received for publication February 24, 1997. Abbreviation Key: C = chicken; CL = continuous light; D = dark;
Accepted for publication August 14, 1997. FCR = feed conversion ratio; GH = growth hormone; IL = intermittent
1 To whom correspondence should be addressed: nadine. light; L = light; T3 = triiodothyronine; T4 = thyroxine; TRH = thyroid
buys@agr.kuleuven.ac.be releasing hormone.

54
 INTERMITTENT LIGHT REDUCES ASCITES INCIDENCE 55
TABLE 1. Composition of experimental feeds experiment was repeated under identical conditions,
except that lighting schedules were exchanged between
Ingredients and analyses B1 diet B2 diet
rooms. Starting dates were 17 February and 13 May 1994.
Ingredients (%) The design thus was a 2 · 2 · 2 factorial experiment with
Wheat 31.00 32.00 lighting schedule (IL or CL), feed (B1 or B2), and treatment
Peas 20.00 0
Soybean meal 10.40 23.50
(control or T3 supplemented) as factors (seven repetitions
Tapioca 8.60 10.20 per group per experiment).
Sorghum 7.00 7.00
Rape seed 6.60 9.00
Animal fat 6.00 6.00 Measurements
Animal meal 6.00 6.00
Fish meal 3.00 3.00 From Day 9 onwards and repeated every 2 wk (Days 9,
Sunflower meal 0 2.30 14, 28, and 42), chickens were weighed on a pen basis and
Vitamin and mineral premix 0.60 0.60
L-lysine 0.50 0 biweekly growth and feed intake was determined. Feed
DL-methionine 0.35 0.21 conversion was calculated on a pen basis. Daily mortality
Limestone 0.30 0.30 was recorded and necropsies were performed. Scores
Calculated analysis were given for hydropericardia, heart dilation, liver
CP 20.50 23.50
CF 11.00 12.00
cirrhosis, and accumulation of fluid in the abdomen
Crude fiber 3.60 4.00 (ascites).
L-Lysine 1.11 1.11 On Days 14, 21, 28, 35, and 42 of age, blood samples
Methionine + Cystine 0.81 0.81 were taken into heparinized tubes from a wing vein from
ME, kcal/kg 3,300 3,300
10 randomly chosen birds per lighting schedule, feed, and
T3 treatment combination. From the seven pens per
experimental group, two chickens were taken from three
randomly assigned pens and one chicken from the
MATERIALS AND METHODS remaining four pens. Chickens were sampled during the
second half of a 1-h light period (0930 h to 1000 h). Blood
samples were centrifuged immediately (3,000 · g, 10 min)
Experimental Design and plasma was stored at –20 C until assayed for growth
Eight hundred and forty day-old male broiler chicks hormone (GH) and thyroid hormones (T4 and T3). From
(Ross) were obtained from a commercial hatchery2 and the same birds, a venous blood sample was taken in
placed in 28 floor pens (1.0 · 0.8 m; 30 chickens per pen). capillaries for measurements of hematocrit.
All chicks were vaccinated against infectious bronchitis, Plasma T3 and T4 levels were measured by RIA using
Gumboro, and Newcastle diseases. Each pen contained antisera and standard solutions from Byk-Sangtec,4
125I-T3 and 125I-T4 from ICN.5 Intra-assay variabilities
two drinking nipples with cups and a 0.8-m feeder trough.
Wood shavings were used as litter. During the first 8 d, the were 4.5 and 5.4% for T3 and T4 respectively. Chicken GH
lighting schedule provided 23 h light/d (23L:1D, CL) and (cGH) was measured with a homologous RIA as deve-
an experimental feed containing 3,300 kcal ME/kg and loped and validated by Berghman et al. (1988). The intra-
23.5% CP (B2) was provided for ad libitum consumption. assay coefficient of variation was 4.0%. All measurements
Temperature was set initially at 32 C and gradually of a given hormone were performed in a single assay.
reduced by 1 C/2 d until 22 C was reached.
At 9 d of age, chicks were randomly distributed into Statistical Analysis
two light-proof, temperature-controlled rooms, each con-
Combined response variable data from both experi-
taining 28 floor pens (15 chickens per pen). In one room,
ments were analyzed by the General Linear Models
the 23L:1D lighting schedule was maintained, whereas in
procedure (SAS Institute, 1986). Lighting schedule, feed,
the other an intermittent lighting schedule (IL) consisting
and T3 supplementation treatment were classification
of 1L:3D cycles, repeated six times daily, was imposed
variables and interactions were calculated. Pen means
(lights on: 0900 to 1000 h; 1300 to 1400 h; 1700 h to 1800 h;
were used as experimental units for calculation of
2100 to 2200 h; 0100 to 0200 h; 0500 to 0600 h). All chickens
mortality, growth, and FCR. If a significant model (P <
were subjected to a light intensity of 20 lx at bird height. In
0.05) was discerned, treatment means were compared by
each room, two isocaloric feeds (3,300 kcal ME/kg)
the LSMeans test. Mortality data were analyzed by
differing in CP content (B1: 20.5% CP; and B2: 23.5% CP)
weighted ANOVA after arc sine square root percentage
were provided (Table 1). Both feeds contained the same transformation (Snedecor and Cochran, 1967).
amounts of essential amino acids and were or were not
supplemented with 1.5 ppm triiodothyronine (T33). The
RESULTS

2Avibel, Halle-Zoersel, Belgium. Mortality

3T2627, Sigma Chemical Co., St. Louis, MO 63178-9916.
4Byk-Sangtec Diagnostica GmbH, Dietzenbach, Germany. The total mortality and the results of the necropsy are
5ICN Biomedicals Inc., Costa Mesa, CA 92626. given in Table 2. Total mortality as well as mortality due to
56 BUYS ET AL.
TABLE 2. Effect of dietary triiodothyronine (T3) supplementation, protein content of feed, and
lighting schedule on mortality of broilers due to ascites, hydropericardium, or other causes

0 ppm T3 1.5 ppm T3

B1: 20.5% CP B2: 23.5% CP B1: 20.5% CP B2: 23.5% CP
Experiment Mortality1 CL2 IL2 CL IL CL IL CL IL
1 Ascites . . . . . . . . . . . . 1 2 . . . . . .
Hydropericardium . . . . . . . . . . . . 9 1 3 2
Other causes . . . 2 . . . 1 3 3 . . . . . .
2 Ascites . . . . . . 1 . . . 8 1 2 3
Hydropericardium . . . . . . 1 . . . 11 9 6 2
Other 1 . . . 1 1 21 15 19 8
Sum of 1 and 2 Ascites . . . . . . 1 . . . 9 3 2 3
Hydropericardium . . . . . . 1 . . . 20 10 9 4
Other 1 2 1 2 24 18 19 8
Total 1 2 3 2 53 31 30 15
1Absolute numbers, initial number of chickens per group is 105.
2CL = 23 h light (L):1 h dark (D), IL = IL:3D.

ascites or to hydropericardium was increased in T3-
supplemented birds compared to control chickens (P <
0.0001), and in the second experiment compared to the
first one (P < 0.005). Moreover, mortality due to ascites or
to hydropericardium was decreased in birds reared under
IL compared to those reared under CL (P < 0.05); this
result was especially obvious in T3-fed chickens. Also in
T3-fed chickens, ascites mortality was much higher in
birds fed a subnormal protein diet (20.5%) than in those
fed a diet containing normal protein (23.5%), regardless of
the lighting schedule. However, in control birds, no effect
of dietary protein content was found (interaction T3
treatment by feed: P < 0.05).
Performance Data
Performance data and the results of the statistical
analyses are given in Tables 3 and 4, respectively. The
change from CL to IL on Day 9 significantly reduced body
weight on Day 14 (P < 0.0001), but this effect was less
obvious in birds fed a low protein diet (interaction light by
feed: P < 0.01). At 28 d of age, birds reared at IL continued
to have lower BW than their CL counterparts, and those
fed B1 were heavier than those fed B2 feed. Final body
weights at 42 d of age, however, did not differ between
lighting schedules or feed. Dietary T3 supplementation
significantly reduced growth from Day 14 on and
throughout the whole experimental period in all groups.
Final BW at 42 d of age was reduced in T3-supplemented
birds compared to control birds, and this reduction was
more pronounced in birds fed B1 than in those fed the B2
diet (interaction feed by T3 supplementation treatment: P
< 0.05).
The data concerning the absolute body weight gain,
calculated per 2 wk, clearly showed an initial growth-
depressive effect of IL, especially in birds fed B2
(interaction light by feed: P < 0.0001). From Day 14 until
28, no effect of lighting schedule on growth was found, but
there was an effect of the feed: birds fed B1 showed
slightly faster growth than those fed B2. During the
finisher period (Day 28 until 42), the chickens reared in IL
showed a compensatory growth (P < 0.005) in that they
reached the same final body weight as those reared in CL.
During this period, the growth of birds fed B2 was higher
than the growth of chickens fed B1, especially if the feed
was supplemented with T3 (interaction feed by T3
treatment: P < 0.05).
Feed intake between Days 9 and 14 of birds reared in IL
was significantly reduced compared to that of their CL
counterparts (P < 0.0001). A slight interaction of light by
feed on feed intake between Days 9 and 14 was due to the
lower FI of birds that consumed the B1 diet and were
reared under the CL treatment (P < 0.05). Dietary T3
supplementation significantly reduced feed intake (P <
0.0001). From Day 14 until 28, chickens reared under IL
consumed less feed than those reared under CL (P <
0.0001). During the finisher period, the lighting schedule
did not continue to influence feed intake, but a significant
interaction between lighting schedule and feed was found
(P < 0.005), as feed intake was higher in IL chicks than in
CL chicks, when fed B2, but feed intake was lower in the IL
chicks than in the CL chicks when the B1 diet was fed,
independent of the dietary T3 supplementation. Cumula-
tive feed intake, calculated over the whole experimental
period, was lower in birds reared under IL than in those
reared under CL (P < 0.001), irrespective of the dietary
protein content or the T3 supplementation. Because the
reduction in growth rate in IL chicks between Days 9 and
14 was more pronounced than the reduction in feed
intake, a less favorable FCR was found in chicks reared
under IL than in chicks reared under CL in this period (P <
0.005), especially in birds fed B2 (interaction light by feed:
P < 0.005). Dietary T3 supplementation did not alter FCR,
as growth and feed intake were reduced to the same
extent. However, from Day 14 to 42, a lower FCR was
found in IL chickens than in CL chickens (P < 0.0001), in
chickens fed B1 than in those fed B2 (P < 0.01), and in
control birds than in T3-supplemented chickens (P <
0.0005). The cumulative FCR was lower in chickens reared
under IL than in those reared under CL, regardless the
protein content of the feed or the T supplementation (P <
0.0001).
 INTERMITTENT LIGHT REDUCES ASCITES INCIDENCE 57
TABLE 3. Effect of dietary triiodothyronine (T3) supplementation, protein content of feed, and lighting
schedule on body weight, body weight gain, feed intake, and feed conversion ratio of broilers1

0 ppm T3 1.5 ppm T3

B1: 20.5% CP B2: 23.5% CP B1: 20.5% CP B2: 23.5% CP
Variable Age CL2 IL2 CL IL CL IL CL IL
(d)
Body weight,
g/chicken 9 162 – 4 161 – 2 168 – 3 163 – 3 166 – 3 162 – 3 167 – 4 162 – 4
14 311 – 5ab 289 – 2bcd 326 – 5a 276 – 8de 298 – 5bc 264 – 4e 295 – 5bc 246 – 8f
28 1,143 – 25a 1,089 – 14b 1,092 – 24b 1,056 – 8b 904 – 18c 865 – 15cd 904 – 14c 840 – 11d
42 2,042 – 40a 2,008 – 27a 1,981 – 30a 2,023 – 22a 1,418 – 33c 1,453 – 30bc 1,524 – 41b 1,505 – 27bc
Body weight
gain, g/chicken/
period 9 to 14 148 – 3a 129 – 2a 158 – 3a 109 – 6c 132 – 3b 102 – 2c 129 – 3b 85 – 5d
14 to 28 832 – 21a 800 – 12ab 766 – 22b 780 – 12b 606 – 16c 601 – 12c 609 – 13c 592 – 13c
28 to 42 899 – 20a 919 – 19ab 889 – 22b 966 – 21a 514 – 21e 588 – 25d 620 – 32cd 665 – 27c
9 to 42 1,880 – 37a 1,840 – 26a 1,813 – 29a 1,855 – 20a 1,253 – 33c 1,291 – 29bc 1,358 – 40b 1,343 – 28b
Feed intake,
g/chicken/period 9 to 14 224 – 4b 188 – 2de 237 – 6a 181 – 2e 202 – 2c 160 – 2f 193 – 3cd 153 – 2f
14 to 28 1,223 – 47ab 1,186 – 21ab 1,243 – 24a 1,147 – 13b 981 – 18cd 912 – 22d 1,035 – 20c 923 – 26d
28 to 42 1,864 – 37a 1,790 – 24ab 1,761 – 68ab 1,838 – 29a 1,652 – 49b 1,474 – 30d 1,506 – 62d 1,576 – 69cd
9 to 42 3,311 – 74a 3,164 – 39a 3,241 – 83a 3,166 – 34a 2,835 – 54b 2,547 – 35d 2,734 – 70bc 2,654 – 63cd
Feed conversion
ratio, kg feed
intake:kg growth 9 to 14 1.52 – 0.05c 1.46 – 0.02c 1.51 – 0.04c 1.72 – 0.09ab 1.54 – 0.04c 1.58 – 0.04bc 1.52 – 0.06c 1.88 – 0.10a
14 to 28 1.46 – 0.03c 1.48 – 0.03c 1.63 – 0.03ab 1.47 – 0.02c 1.63 – 0.04ab 1.52 – 0.04c 1.71 – 0.04a 1.56 – 0.04bc
28 to 42 2.07 – 0.03c 1.96 – 0.04c 1.99 – 0.09c 1.91 – 0.03c 3.25 – 0.12a 2.56 – 0.11b 2.50 – 0.17b 2.39 – 0.09b
9 to 42 1.76 – 0.01c 1.71 – 0.02c 1.79 – 0.04c 1.71 – 0.02c 2.27 – 0.05a 1.98 – 0.04b 2.03 – 0.08b 1.98 – 0.04b
a–fMeansin a row with no common superscript differ significantly (P < 0.05).
1Valuesare means – SEM of 14 pens of 15 birds initially.
2CL = 23 h light (L):1 h dark (D), IL = 3L:1D.

Plasma Hormone Levels
Mean plasma hormone levels are given in Table 5 and
the results of the statistical analysis are summarized in
Table 6. Dietary T3 supplementation significantly in-
creased plasma T3 levels at all ages (P < 0.0001). At 14 d of
age, this increase was more pronounced in birds fed B1,
especially when reared under IL. At 28 and 42 d of age,
respectively, lower and higher T3 levels were found in IL
chickens than in CL ones (P < 0.01).
Plasma T4 levels were higher in 14-d-old chicks reared
under IL than in those reared under CL (P < 0.0005),
although this was only the case for B2 chickens (interac-
tion light by feed: P < 0.05). Dietary T3 supplementation

TABLE 4. Probability values resulting from the factorial analysis of variance with as factors lighting schedule,
feed, and dietary triiodothyronine (T3) supplementation and their interactions as calculated for
body weight, body weight gain, feed intake, and feed conversion ratio

Factor
Lighting Feed T3 suppl. L · F
Variable Age schedule (L) (F) (T) L · F L · T F · T · T
(d)
Body weight 9 NS NS NS NS NS NS NS
14 <0.0001 NS <0.0001 0.0084 NS NS NS
28 0.0002 0.0281 <0.0001 NS NS NS NS
42 NS NS <0.0001 NS 0.0240 NS
Body weight gain 9 to 14 <0.0001 0.0019 <0.0001 <0.0001 NS NS NS
14 to 28 NS 0.0409 <0.0001 NS NS NS NS
28 to 42 0.0015 0.0012 <0.0001 NS NS 0.0284 NS
9 to 42 NS NS <0.0001 NS NS 0.0103 NS
Feed intake 9 to 14 <0.0001 NS <0.0001 0.0416 NS 0.0244 0.0231
14 to 28 <0.0001 NS <0.0001 NS NS NS NS
28 to 42 NS NS <0.0001 0.0045 NS NS NS
9 to 42 <0.0001 NS <0.0001 NS NS NS NS
Feed conversion 9 to 14 0.0018 0.0034 NS 0.0012 NS NS NS
14 to 28 <0.0001 0.0061 0.0003 0.0341 NS NS NS
28 to 42 0.0002 <0.0001 <0.0001 0.0218 0.0228 0.0034 0.0413
9 to 42 <0.0001 (0.053) <0.0001 NS NS 0.0224 0.0177
58 BUYS ET AL.
TABLE 5. Effect of dietary triiodothyronine (T3) supplementation, protein content of feed, and lighting schedule
on plasma T3, thyroxine (T4), and growth hormone (GH) concentrations and hematocrit in broilers1

0 ppm T3 1.5 ppm T3

B1: 20.5% CP B2: 23.5% CP B1: 20.5% CP B2: 23.5% CP
Variable Age CL2 IL2 CL IL CL IL CL IL
(d)
Plasma T3
concentration,
ng/mL 14 2.59 – 0.14d 2.16 – 0.17d 3.47 – 0.22d 2.14 – 0.28d 9.47 – 0.75b 11.68 – 0.83a 5.72 – 0.70c 6.31 – 0.88c
21 2.63 – 0.15c 2.27 – 0.23c 2.21 – 0.21c 2.00 – 0.19c 9.47 – 0.73a 9.28 – 1.48a 6.36 – 0.74b 9.92 – 0.65a
28 1.82 – 0.18c 1.30 – 0.10c 1.87 – 0.13c 0.94 – 0.11c 4.85 – 0.88ab 4.19 – 0.53b 6.07 – 0.55a 4.70 – 0.54b
35 0.97 – 0.07c 0.84 – 0.10c 0.95 – 0.09c 1.23 – 0.08c 3.95 – 0.54ab 2.77 – 0.53b 5.02 – 0.46a 4.55 – 0.62a
42 0.67 – 0.05d 1.19 – 0.10d 0.88 – 0.08d 1.47 – 0.10d 4.53 – 0.51c 7.91 – 0.42b 4.54 – 0.38c 9.71 – 1.08a
Plasma T4
concentration,
ng/mL 14 5.5 – 0.6b 6.9 – 0.4b 5.9 – 0.6bc 11.6 – 1.24a 2.4 – 0.1d 2.8 – 0.3cd 2.3 – 0.1d 5.4 – 2.1bc
21 8.6 – 0.7c 11.6 – 1.1ab 10.8 – 0.6b 12.9 – 0.9a 4.5 – 0.3d 5.4 – 0.4d 3.9 – 0.2d 5.1 – 0.2d
28 12.2 – 0.8b 13.0 – 0.6b 12.0 – 0.9b 14.8 – 0.7a 4.0 – 0.4c 3.6 – 0.2c 4.6 – 0.2c 3.6 – 0.2c
35 9.4 – 1.1c 9.7 – 0.8b 11.3 – 1.0ab 12.2 – 0.6a 3.4 – 0.2c 2.7 – 0.3c 3.6 – 0.2c 3.5 – 0.2c
42 10.8 – 0.8ab 10.3 – 0.9b 12.3 – 0.8a 12.4 – 0.5a 3.3 – 0.1c 4.1 – 0.1c 3.6 – 0.2c 4.8 – 0.2c
Plasma GH
concentration,
ng/mL 14 102.6 – 13.0a 107.9 – 16.5a 69.9 – 11bc 83.6 – 10ab 35.4 – 2.8d 42.4 – 2.4cd 55.8 – 5.5bcd 65.0 – 7bcd
21 67.2 – 9.2b 67.5 – 14.3b 107.9 – 15.2a 71.9 – 16.4b 32.2 – 1.5c 43.9 – 2.3bc 46.4 – 4.7bc 46.9 – 3.8bc
28 83.9 – 7.2a 53.0 – 5.5b 76.1 – 6.1a 88.8 – 11.4a 40.3 – 4.7bc 38.3 – 1.6bc 28.9 – 2.2c 42.1 – 2.9bc
35 46.2 – 2.8bc 62.2 – 6.0a 55.4 – 5.4ab 61.4 – 7.2a 33.1 – 2.7cd 42.0 – 4bcd 29.0 – 2.6d 39.3 – 3.6cd
42 40.2 – 2.8ab 51.6 – 4.7a 40.5 – 4.1ab 48.1 – 4.3ab 26.5 – 1.9c 35.4 – 2.4bc 24.7 – 1.7c 45.5 – 8.6ab
Hematocrit 14 28.4 – 0.5abc 26.3 – 0.5c 27.0 – 0.6bc 28.7 – 0.8abc 26.7 – 0.7c 27.8 – 0.4abc 29.9 – 1.8a 29.2 – 0.5ab
21 30.6 – 0.8ab 27.9 – 1.0b 31.4 – 0.8a 29.6 – 0.8ab 30.4 – 0.9ab 28.0 – 1.0b 30.6 – 0.7ab 30.3 – 0.8ab
28 28.7 – 0.3b 26.5 – 0.3c 28.0 – 0.6bc 29.5 – 0.7ab 29.4 – 1.2ab 27.6 – 0.5bc 29.0 – 0.5b 31.1 – 0.8a
35 30.0 – 0.6 29.0 – 0.8 28.9 – 0.9 28.9 – 0.8 31.4 – 0.8 30.5 – 0.9 28.9 – 0.8 31.0 – 0.9
42 30.8 – 0.9 28.9 – 1.3 29.2 – 0.6 28.3 – 0.7 29.4 – 0.8 31.0 – 1.3 28.6 – 0.5 28.2 – 0.7
a–dMeans in a row with no common superscript differ significantly (P < 0.05).
1Valuesare means – SEM of 20 chickens per treatment.
2CL = 23 h light (L):1 h dark (D). IL = 3L:1D.

reduced plasma T4 levels at all ages (P < 0.0001) in control Hematocrit

birds reared under IL. In both feed groups, plasma T4
levels were higher in birds reared under IL than in birds
reared under CL (P < 0.005) at 14 and 21 d of age. Dietary
T3 supplementation, however, abolished the effect of
lighting schedule (interaction light by T3 treatment: P <
0.05) and of feed (Feed by T3 treatment interaction: P <
0.005). At the end of the experiment, higher plasma T4
concentrations were found in euthyroid B2 chickens,
regardless of the lighting schedule, and influences of
lighting schedule disappeared.
Plasma GH levels were influenced mainly by dietary T3
supplementation. By Day 14, GH concentrations were
lower in hyperthyroid chickens, especially when fed B1
(feed by T3 treatment interaction: P < 0.001). At 21 d of age,
birds fed B2 had higher plasma GH concentrations than
those fed B1. At the age of 4 wk, these higher concentra-
tions were only obvious in birds reared in IL (light by feed
interaction: P < 0.001) and not supplemented with T3 (feed
by T3 treatment interaction: P < 0.05). At Days 35 and 42,
plasma GH levels were significantly higher in chickens
reared under IL than in chicks reared under CL (P < 0.001).
In chickens fed an unsupplemented diet, an age-related
decrease was observed for T3 (P < 0.001) and GH (P <
0.005), whereas T4 levels increased with age (P < 0.005).
Only at the age of 21 d was a significant effect of the
imposed lighting schedule on hematocrit values observed
(P < 0.01); whereas birds reared under IL had lower
hematocrit values than those reared under CL. During the
first 4 wk of age, hematocrit levels were higher in birds fed
B2 than in those fed B1, but this difference disappeared
afterwards. Dietary T3 supplementation did increase
hematocrit values at 28 d of age (P < 0.05), whereas at
other ages no effect was found.
DISCUSSION
As described earlier (Buyse et al., 1994a,b, 1996), the
change of CL to IL at an early age is followed by initial
growth depression. This depression is, however, fol-
lowed by a period of compensatory growth, in a way
that the birds reared in IL reach the same final body
weight by 6 wk of age, as those reared in CL. Moreover,
in this experiment, an interaction between the lighting
schedule and the protein content of the feed was
observed. The initial growth-depressive effect of IL was
more pronounced in birds fed a normal protein ration
(23.5% CP, B2) than in those receiving feed with a
 INTERMITTENT LIGHT REDUCES ASCITES INCIDENCE 59
TABLE 6. Probability values resulting from the factorial analysis of variance with the factors lighting schedule,
feed, and dietary triiodothyronine (T3) supplementation and their interactions as calculated for the
plasma T3, thyroxine (T4), and growth hormone (GH) concentrations and hematocrit

Factor
Lighting Feed T3 supplemen-
Variable Age schedule (L) (F) tation (T) L · F L · T F · T L · F · T
(d)
T3 concentration 14 NS <0.0001 <0.0001 NS 0.0078 <0.0001 NS
21 NS NS <0.0001 0.0418 0.0393 NS NS
28 0.0075 NS <0.0001 NS NS NS NS
35 NS 0.0087 <0.0001 NS NS 0.0417 NS
42 <0.0001 NS <0.0001 NS <0.0001 NS NS
T4 concentration 14 0.0002 0.0066 <0.0001 0.0105 NS NS NS
21 0.0017 NS <0.0001 NS 0.0312 0.0028 NS
28 NS NS <0.0001 NS 0.0025 NS NS
35 NS 0.0041 <0.0001 NS NS NS NS
42 NS 0.0087 <0.0001 NS NS NS NS
GH concentration 14 NS NS <0.0001 NS NS 0.0006 NS
21 NS 0.0344 <0.0001 NS NS NS NS
28 NS NS <0.0001 0.0008 NS 0.0395 NS
35 0.0021 NS <0.0001 NS NS NS NS
42 0.0002 NS <0.0001 NS NS NS NS
Hematocrit 14 NS 0.0123 NS NS NS NS 0.0130
21 0.0059 0.0494 NS NS NS NS NS
28 NS 0.0078 0.0266 0.0002 NS NS NS
35 NS NS NS NS NS NS NS
42 NS NS NS NS NS NS NS

subnormal protein content (20.5% CP, B1). The same
phenomenon was found in T3-fed chickens. The initial
slower growth rate in B2 chicks is linked to a less
favorable FCR. Nevertheless, IL resulted in an overall
decreased FCR, irrespective of the protein content of the
feed or the dietary T3 supplementation. No overall feed
effect between 9 and 42 d of age was observed, nor was
there a light by feed interaction. Therefore, it can be
concluded that a reduction of the CP content of feed
from 23.5 to 20.5% does not negatively influence the
performance of the birds, regardless of lighting pro-
gram, if the required levels of essential amino acids are
met. This result will have a beneficial effect on the
nitrogen emission with both lighting schedules, particu-
larly with the IL treatment, in view of the improvement
of dietary N retention under the latter lighting schedule
(Buyse et al., 1996).
It was clearly demonstrated that IL reduces the
incidence of ascites. This result can be explained by
different mechanisms. First, it has been shown that with
IL, heat production and oxygen consumption are
significantly lower during the dark periods of each L:D
cycle (Buyse et al., 1994b). Because a lack of oxygen is
known to be the primary cause of the development of
ascites, lower oxygen consumption in IL chicks could
reduce the incidence of ascites. A second mechanism
could be the altered growth pattern between IL and CL
chickens. Decuypere et al. (1994) showed that the
incidence of ascites is much higher in fast-growing
broiler lines than in slower growing ones. Although IL
chickens reached the same final body weight as CL
chickens by 42 d of age, IL chicks have a reduced
growth in the 2nd wk, followed by a compensatory
growth in the period thereafter. This difference in
growth rate results in a more concave growth pattern
compared to that of CL chickens. As shown by Buyse et
al. (1994b), heat production, and hence oxygen require-
ments per kilogram of metabolic BW, of chickens
following a normal growth trajectory (with CL) are the
highest at about 2 wk of age, which is indicative of an
amplified metabolic demand. This metabolic demand
predisposes chickens for ascites development. The
accumulation of fluid in the pericardium and the
abdomen is only the final step in a cascade of events
leading to ascites. The predisposition for the develop-
ment of the syndrome already occurs during the first
weeks of the growing period. It is exactly during this
initial period that the growth rate, and, thus, the oxygen
requirements of the IL chickens, are reduced, which
alleviates the metabolic load and, hence, the develop-
ment of ascites. The lower hematocrit values found in IL
chicks at 3 wk of age support this hypothesis. The
beneficial effect of IL on ascites incidence could be
compared with the effect of early feed restriction
(Shlossberg et al., 1991). Temporary growth reduction
has no negative effects on proportional weights of lungs
and heart (as is the case for muscle); but appear to have
beneficial effects of these variables (Buyse, personal
communication). Acar et al. (1995) also found that a
temporary feed reduction of 25% during the 2nd wk of
the growing period resulted in a growth reduction, but
did not affect proportional heart and lung weights. Also,
ascites mortality was significantly reduced in feed-
restricted broilers compared to their counterparts that
60 BUYS ET AL.
consumed feed ad libitum. The beneficial effect of a more
concave growth pattern in avoiding ascites is also found
in chicks fed B2 compared to those fed B1. Surprisingly,
B2-fed birds had relatively high hematocrit values. The
rather low hematocrit values of T3-supplemented chick-
ens that showed a high ascites incidence, however,
indicates that the relationship between hematocrit and
ascites mortality is not always obvious.
The management factors investigated in this experi-
ment not only influenced the performance and ascites
incidence of the birds, but also their endocrine functions.
Dietary T3 supplementation increased plasma T3 levels,
as found earlier (Buys et al., 1993; Decuypere et al., 1994).
The reason for the larger increase in 14-d-old birds fed a
high protein diet, especially when reared under IL, is
not clear. It can not be due to a higher feed intake and,
thus T3 intake. Nor does this appear to be the case per
unit BW. It is possible that there are differences in T3
absorption, entero-hepatic cycles, or breakdown be-
tween B1 and B2 chickens, although these differences
cannot be concluded from this experiment. Plasma T3
levels are higher when feed intake is higher: chickens
fed B2 show higher plasma T3 than those fed B1; and in
the first weeks, birds under CL have higher plasma T3
concentrations than their IL counterparts, whereas at
later stages, the opposite relationship is found. A direct
effect of dietary protein content on T4 secretion was
found earlier and is mediated by an effect on
hypothalamo-pituitary activity (Buys, unpublished
results). Dietary T3 supplementation decreased plasma
T4 and GH levels as was observed earlier (Decuypere et
al., 1994). This result can be explained by a negative
feedback of T3 on the hypothalamus, resulting in a
decreased thyroid-releasing hormone (TRH) secretion.
As TRH is also a potent stimulator of GH release in
avian species (Harvey et al., 1991), reduction in TRH can
cause the reduced GH levels in T3-treated birds. The
decrease in IGF-I after T3 treatment could be related to
the decreased GH levels but perhaps also to a direct
effect of T3 on decreasing GH receptor numbers
independently from circulating GH concentrations (Buys
et al., 1993).
The higher plasma GH levels in birds reared under IL
than in birds reared under CL confirm earlier results
that IL broilers manifesting compensatory growth have
higher mean plasma GH levels than their age-matched
counterparts (Kühn et al., 1996). These higher mean GH
levels are the mechanistic result of the enhanced
pulsatile GH secretion from the pituitary somatotrophs,
as reflected by the higher GH mass per burst, amplitude,
and, hence, production rate that is a predominant causal
mechanism for the improved efficiency of dietary
nitrogen retention during compensatory growth (Buyse
et al., 1997).
It is concluded that IL or a higher protein content of
the feed reduces the incidence of T3-induced ascites
mortality and that this could be due to different causal
mechanisms.
ACKNOWLEDGMENTS
This research was funded by the “Instituut ter
bevordering van het Wetenschappelijk Onderzoek in
Nijverheid en Landbouw (IWONL)”, project nr. D1/
4-5260/5507A. G. Nackaerts and I. Geerts were grate-
fully appreciated for technical assistance.
REFERENCES
Acar, N., F. G. Sizemore, G. R. Leach, R. F. Wideman, Jr., R. L.
Owen, and G. F. Barbato, 1995. Growth of broiler chickens
in response to feed restriction regimens to reduce ascites.
Poultry Sci. 74:833–843.
Albers, G., and M. T. Frankenhuis, 1990. Ascites, a high
altitude disease in the lowlands. Poultry Misset (Feb.–
March 1990):24–25.
Berghman, L., J. Van Beemen, E. Decuypere, E. R. Kühn and F.
Vandesande, 1988. One step purification of chicken
growth hormone from a crude pituitary extract by use of a
monoclonal immunoadsorbant. J. Endocrinol. 112:381–387.
Buys, N., T. Bartha, J. Buyse, G.A.A. Albers, and E. Decuypere,
1993. Effect of dietary T3 supplementation on ascites
incidence in an ascites-sensitive and -resistant broiler line:
influence on liver deiodination activity and on hormonal
profiles in plasma, pericardial and abdominal ascitic fluid.
Pages 300–304 in: Proceedings of the 9th European
symposium on Poultry nutrition, September 5–11, 1993
Jelenia Gora, Poland.
Buyse, J., E. Decuypere, and H. Michels, 1994a. Intermittent
lighting and broiler production. 1. Effect on female broiler
performance. Arch. Geflügelkd. 58:69–74.
Buyse, J., E. Decuypere, and H. Michels, 1994b. Intermittent
lighting and broiler production. 2. Effect on energy and on
nitrogen metabolism. Arch. Geflügelkd. 58:78–83.
Buyse, J., E. R. Kühn, and E. Decuypere, 1996. The use of
intermittent lighting in broiler raising: 1. Effect on broiler
performance and efficiency of nitrogen retention. Poultry
Sci. 75:589–594.
Buyse, J., E. Decuypere, and J. D. Veldhuis, 1997. Compensa-
tory growth of broiler chickens is associated with an
enhanced pulsatile growth hormone (GH) secretion:
preferential amplification of GH secretory burst mass. Br.
Poult. Sci. 38:291–296.
Decuypere, E., C. Vega, T. Bartha, J. Buyse, J. Zoons, and
G.A.A. Albers, 1994. Increased sensitivity to triiodothyro-
nine (T3) of broiler lines with a high susceptibility for
ascites. Br. Poult. Sci. 35:287–297.
Harvey, S., R. A. Fraser, and R. W. Lea, 1991. Growth hormone
secretion in poultry. CRC Rev. Poult. Biol. 3:239–282.
Kühn, E. R., V. M. Darras, C. Gysemans, E. Decuypere, L. R.
Berghman, and J. Buyse, 1996. The use of intermittent
lighting in broiler raising. 2. Effects on the somatotrophic
and thyroid axes and on plasma testosterone levels.
Poultry Sci. 75:595–600.
Julian, R. J., 1993. Ascites in poultry (review article). Avian
Pathol. 22:419–454.
Maxwell, H. M., I. A. Alexander, G. W. Robertson, M. A.
Mitchell and C. McCorquodale, 1995. Identification of
tissue hypoxia in the livers of ascitic and hypoxia induced
broilers using trypan blue. Br. Poult. Sci. 36:791–798.
 INTERMITTENT LIGHT REDUCES ASCITES INCIDENCE 61
SAS Institute, 1986. SAS User’s Guide: Statistics. SAS Institute Schlosberg, A., E. Berman, U. Bendheim, and I. Plavnik, 1991.
Inc., Cary, NC. Controlled early feed restriction as a potential means of
Scheele, C. W., E. Decuypere, P.F.G. Vereijken, and F.J.G. reducing the incidence of ascites in broilers. Avian Dis. 35:
Schreurs, 1992. Ascites in Broilers: 2. Disturbances in the 681–684.
hormonal regulation of metabolic rate and fat metabolism. Snedecor, G. W., and W. G. Cochran, 1967. Statistical Methods.
Poultry Sci. 71:1971–1984. 6th ed. The Iowa State University Press, Ames, IA.