Content uploaded by Andrea Ermolao
Author content
All content in this area was uploaded by Andrea Ermolao
Content may be subject to copyright.
ORIGINAL ARTICLE
Marco Zaccaria Æ Andrea Ermolao Æ Giulio Sergio Roi
Piera Englaro Æ Giovanna Tegon Æ Maurizio Varnier
Leptin reduction after endurance races differing in duration
and energy expenditure
Accepted: 7 February 2002 / Published online: 26 April 2002
Ó Springer-Verlag 2002
Abstract Serum leptin concentrations are reduced in the
presence of a negative energy balance. It has been
demonstrated, however, that strenuous and prolonged
exercise, which induces a marked negative energy bal-
ance, is not always followed by a reduction in serum
leptin levels. We therefore analysed serum leptin con-
centrations before and after three endurance races,
which differed in duration and energy expenditure (EE),
with the aim of clarifying the relationship between the
level of EE and the reduction in leptin levels. Forty-five
males participated in one of three competitive endurance
races, a half-marathon run [21.097 km, estimated EE
1,400 kcal (5,852 kJ)], a ski-alpinism race [about 45 km,
estimated EE 5,000 kcal (20,900 kJ)], and an ultramar-
athon race [100 km, estimated EE 7,000 kcal
(29,269 kJ)]. Blood samples for analysis of serum leptin,
and plasma free fatty acids (FFA) were collected before
and after the races. Pre-race leptin values were signifi-
cantly correlated with both body mass index and body
fat mass (r=0.672 and r=0.699, respectively;
P<0.0001). After exercise, serum leptin levels decreased
significantly in the ultramarathon [from 4.15 (0.63) lg/l
to 1.01 (0.15) lg/l; P<0.001] and in the ski-alpinism
race [from 1.10 (0.28) lg/l to 0.62 (0.15) lg/l; P<0.01],
but not in the half-marathon [from 1.38 (0.40) lg/l to
1.20 (0.36) lg/l]. Plasma FFA were found to have sig-
nificantly increased in all three of the races, showing a
negative correlation with the percent reduction in leptin
(r=0.369, P<0.02). Our data indicate that only a pro-
longed endurance exercise involving a high EE can in-
duce a marked reduction in circulating serum leptin
levels.
Keywords Endurance exercise Æ Leptin Æ Energy
expenditure Æ Free fatty acids
Introduction
Leptin, a protein that is expressed and secreted by adi-
pose tissue, regulates food intake by acting as a hypo-
thalamic signal of satiety. Under steady-state conditions
of energy balance, leptin is considered a good index of
the amount of adipose stores, while under non-steady-
state conditions, leptin is no longer a reliable marker of
body fat stores. In fact, a negative energy balance state,
such as during prolonged fasting, decreases leptin con-
centrations, whereas a positive energy balance, such as
during overfeeding, increases leptin levels (Kolaczynski
et al. 1996).
Exercise is a physiological condition that effects a
rapid increase of energy expenditure that is positively
related to the intensity and duration of the exercise itself.
Thus, a strenuous and prolonged exercise bout that in-
duces a marked negative energy balance state would be
expected to decrease leptin levels. However, reports on
the effects of prolonged exercises on serum leptin con-
centrations have yielded contradictory findings. Serum
leptin has been reported to be reduced or unchanged
both after exercises performed in the laboratory, and
after strenuous endurance races, such as marathons or
ultramarathons (Hickey et al. 1996; Koistinen et al.
1998; Landt et al. 1997; Tuominen et al. 1997). These
contradictory results may be due to differences in the
experimental procedures used in the different studies
(exercise protocols, diet before and during the exercise,
circadian rhythm of leptin). We therefore analysed se-
rum leptin concentrations before and after three com-
petition endurance races that differed in duration and
Eur J Appl Physiol (2002) 87: 108–111
DOI 10.1007/s00421-002-0606-4
M. Zaccaria (&) Æ A. Ermolao Æ P. Englaro Æ G. Tegon
M. Varnier
U.O.A. di Medicina dello Sport,
Dipartimento di Scienze Mediche e Chirurgiche,
Universita
`
di Padova,
Via Ospedale Civile, 105, 35128 Padova, Italy
E-mail: marco.zaccaria@unipd.it
Tel.: +39-49-8215650
Fax: +39-49-8215810
G.S. Roi
Casa di cura ‘‘S. Maria’’,
Viale Piemonte, 70 Castellanza, Varese, Italy
energy expenditure, with the aim of verifying the rela-
tionship between the level of energy expenditure and the
reduction in serum levels of leptin.
Methods
Forty-five Caucasian males, all runners usually engaged in com-
petitive races, voluntarily agreed to be included in the study and
gave their written informed consent to participate. Each subject
participated in one of three endurance competitive races. The en-
ergy expenditure for each race was estimated following methods
described elsewhere (di Prampero 1986).
The first race was a sea-level, flat half-marathon run
(21.097 km), starting at 9.00 a.m., with a mean exercise time of 1 h
31 min and an estimated mean energy expenditure of 1,400 kcal
(5.852 kJ).
The second race was a 45-km ski-alpinism race, alternating up-
and-downhill running and cross-country skiing. Athletes started at
8.00 a.m. from 2,080 m, reached a maximum altitude of 4,226 m,
and then went down to the finish at 1,637 m. The mean time of the
race was 7 h 21 min, and the mean estimated energy expenditure
was 5,000 kcal (20,900 kJ).
The third race was a 100-km ultramarathon race starting at
3.00 p.m., with a mean time of 15 h 27 min, and a mean estimated
energy expenditure of 7.00 kcal (29,269 kJ).
All subjects were in a post-absorptive condition, having eaten a
carbohydrate-rich meal about 2 h before the start of the race.
During the race they drank water, and energy drinks (sweet tea,
isotonic drinks), and ate energy bars ad libitum.
Before each race, in all subjects the body mass index (BMI) and
body composition were assessed by bioelectrical impedance (BIA
101 RJL/Akern, Florence, Italy). Blood samples for measurement
of serum leptin and plasma free fatty acids (FFA) were collected
immediately before and within 20 min after the race, except in the
ski-alpinism race, in which blood samples were collected from
subjects in a post-absorptive condition the day before the race, at
about the same time of day as the beginning of the race.
Serum leptin was measured using a specific double-antibody
radioimmunoassay method (Sensitivity 0.03lg/l; inter- and intra-
assay coefficient of variability lower than 7.6% and 5%, respec-
tively). Plasma FFA was determined by an enzymatic method
(NEFA Quick ‘‘BMY’’, Boehringer Mannheim Yamanouchi,
Tokyo, Japan).
Statistical analysis
Data are expressed as mean (SEM). The statistical analysis of pre-
and post-exercise values was made using the Wilcoxon test for
paired data. Pearson’s correlation coefficient was used to test the
relationship between leptin and the fat mass percentage and
between the percentage of variation of leptin and FFA. The level of
statistical significance was set at P<0.05.
Results
The characteristics of the subjects under basal condi-
tions are reported in Table 1: ski-alpinism subjects were
the youngest, while the ultramarathon subjects were the
oldest and had the highest BMI, body fat mass and
resting leptin concentrations. The overall pre-race leptin
values of subjects, were significantly correlated with
BMI and with body fat mass (r=0.672 and r=0.699,
respectively; P<0.0001).
After exercise, serum leptin levels significantly de-
creased in subjects who participated in the two ult-
raendurance races [ultramarathon: from 4.15 (0.63) lg/l
to 1.01 (0.15) lg/l, P<0.0005; ski-alpinism race: from
1.10 (0.28) lg/l to 0.62 (0.15) lg/l, P<0.01], while in the
half-marathon, no significant reduction was found [from
1.38 (0.40) lg/l to 1.20 (0.36) lg/l; P=0.065, Fig. 1].
The mean percent decrease in leptin was –71 (3)% in the
ultramarathon, –40 (7)% in the ski-alpinism, and –2.5
(10)% in the half-marathon.
Plasma FFA levels significantly increased at the end
of the races, with the absolute highest level being ob-
served in the longest race [ultramarathon: from 471.3
(92.4) lmol/l to 1,801.4 (188.6) lmol/l, P<0.001; ski-
alpinism: from 486.9 (55.0) lmol/l to 1,484.0
(221.9) lmol/l, P<0.001; half-marathon: from 733.8
(44.1) lmol/l to 1,234.8 (97.5) lmol/l, P<0.001]. The
percent variation in FFA showed a significant inverse
relationship with the percent variation in leptin
(r=0.369, P<0.02).
Discussion
The aim of our study was to investigate the effect on
serum leptin concentration of three endurance races of
different duration and energy expenditure. Two of the
races, the ski-alpinism race and the ultramarathon, were
very prolonged, medium- to high-intensity endurance
races (7 h 15 min and 15 h 27 min, respectively),
Table 1. Characteristics of
races and subjects. Please note
that energy expenditure is given
in kcal, where 1 kcal=4.19 kJ.
(BMI Body mass index)
Parameter Half-marathon Ski-alpinism Ultramarathon
Distance (km) 21.097 45 100
Mean [range] duration
(h:min)
1:30 (1:08–1:50) 7:21 min (5:16–8:08) 15:27 min (13:00–17:00)
Energy expenditure
(estimated kcal)
1,400 5,000 7,000
Number of subjects 23 11 11
Age (years) 44.3 (2.7) 34.6 (2.5)* 46.1 (3.2)
BMI (kg/m
2
) 23.2 (0.4) 22.0 (0.4) 26.4 (0.9)**
,
***
Body fat mass (%) 14.5 (0.7) 12.8 (0.9) 17.7 (0.8)**
,
***
*P<0.05 vs ultramarathon and half-marathon;
**P<0.05 vs half-marathon;
***P<0.05 vs ski-alpinism
109
involving a very high energy expenditure (5,000 kcal and
7,000 kcal, respectively). The third race, the half-mara-
thon, was a shorter, high-intensity race (1 h 31 min)
involving a lower energy expenditure (1,400 kcal). Se-
rum leptin concentration decreased significantly in the
ultramarathon (–71%) and in the ski-alpinism race
(–40%), but not in the half-marathon (–2.5%). Thus, the
major finding of our study was that only prolonged
strenuous endurance exercises involving elevated energy
expenditure reduced serum leptin concentrations.
Previous reports on the effect of exercise on serum
leptin concentrations are contradictory: acute exercise
lasting 30 min, performed at different intensities and
caloric expenditure (Weltman et al. 2000), and a 20-mile
(32.2 km) treadmill run (Hickey et al. 1996) did not af-
fect serum leptin concentrations, while a 3-h cycle erg-
ometer exercise bout induced a 42% decrease in serum
leptin (Koistinen et al. 1998). Studies on marathon
runners, on the other hand, showed either no changes
(Koistinen et al. 1998) or a slight (11%), but significant
reduction (Tuominen et al. 1997) in leptin levels. In the
only report available on ultraendurance exercise
[101 miles (162.5 km) at a medium altitude], a significant
reduction (32%) in leptin levels was found (Landt et al.
1997). These contradictory or non-homogeneous find-
ings on the leptin response to different kinds of exercise
could be explained, at least partially, by differences in
the experimental procedures (i.e. exercise protocols, diet
before and during the exercise, circadian rhythm of
leptin).
Fasting causes a fall in circulating leptin levels, be-
ginning after 12 h (Kolaczynski et al. 1996). Since most
of the available studies on the effect of exercise on leptin
levels were performed in overnight-fasted subjects, the
effect of fasting could have influenced basal leptin levels
and, consequently, their response to exercise (Hickey
et al. 1996; Landt et al. 1997). To avoid any interference
from fasting, our subjects were studied in a post-ab-
sorptive condition, having eaten a carbohydrate-rich
meal about 2 h before the start of the race. This eating
pattern, which is usual for people participating in
endurance races, could have counteracted the fasting-
induced fall in leptin; in fact, it has been shown that
small amounts of glucose prevents the reduction in se-
rum leptin that usually occurs during prolonged fasting
(Kolaczynski et al. 1996). However, if a small amount of
glucose has been reported to be sufficient to prevent the
fall in leptin induced by fasting, this does not appear to
happen during prolonged exercise, because we found a
marked decrease in leptin concentrations in spite of the
consumption of energy drinks and energy bars ad libi-
tum during the races. So, from our data, it seems that
the effect of the negative energy balance induced by
exercise prevails over the effect of glucose on leptin
levels.
Variations in the circadian rhythm of leptin may also
influence leptin responses to exercise, since it has been
demonstrated that the highest concentrations of leptin
levels occur between midnight and early morning, fol-
lowed by a progressive decrease, with the lowest con-
centrations occurring at noon and in the early-afternoon
(Licinio et al. 1997). Our findings allow us to rule out
any influence of the circadian rhythm on the two ult-
raendurance races, because pre- and post-exercise ski-
alpinism race samples were collected at the same time
(3.00 p.m.) on two consecutive days, and the post-exer-
cise samples for the ultramarathon were obtained in the
early morning (6.00 a.m.), when leptin is declining but is
still at higher levels than at the time of the start of the
race (3.00 p.m.).
In agreement with other authors (Landt et al. 1997),
we found that FFA were affected markedly by pro-
longed exercise, the mean percentage increase ranging
from 50% to 500%, and showing a significant inverse
relationship with the percent variation of serum leptin.
On the other hand, in the half-marathon we found a
significant increase of FFA not related to leptin varia-
tions, showing that in man, acute variations of FFA
concentrations do not change leptin levels. Thus, the
observed inverse relationship of leptin versus FFA
during exercise seems not consequent to a cause-and-
effect relationship.
In conclusion, our findings indicate that a significant
reduction of serum leptin concentration can be observed
Fig. 1. Mean (SEM) leptin and
free fatty acid (FFA) levels
before (white bars) and after
(shaded bars) the three endur-
ance races.*P<0.01;
**P<0.001
110
only when endurance exercise is very prolonged, with an
energy expenditure by far higher than the energy ex-
pended in 24 h of normal activity. However, the physi-
ological mechanisms regulating the reduction in leptin
that occurs during exercise remain unclear, and are de-
serving of further investigation.
Acknowledgement The authors acknowledge the Federation of
Sport at Altitude for its kind cooperation.
References
di Prampero PE (1986) The energy cost of human locomotion on
land and water. Int J Sports Med 7:55–72
Hickey MS, Considine RV, Israel RG, Mahar TL, McCammon
MR, Tyndall GL, Houmard JA, Caro JF (1996) Leptin is re-
lated to body fat content in male distance runners. Am J Physiol
271:E938–E940
Kolaczynski JW, Considine RV, Ohannesian J, Marco CC,
Opentanova I, Nyce MR, Myint M, Caro JF (1996) Responses
of leptin to short-term fasting and refeeding in humans: a link
with ketogenesis but not ketones themselves. Diabetes 45:
1511–1515
Koistinen HA, Tuominen JA, Ebeling P, Heiman ML, Stephens
TW, Koivisto VA (1998) The effect of exercise on leptin con-
centration in healthy men and in type 1 diabetic patients. Med
Sci Sports Exerc 30:805–810
Landt M, Lawson GM, Helgeson JM, Davila-Roman VG,
Ladenson JH, Jaffe AS, Hickner RC (1997) Prolonged exercise
decreases serum leptin concentrations. Metabolism 46:
1109–1112
Licinio J, Mantzoros C, Negra
˜
o AB, Cizza G, Wong M, Buongi-
orno PB, Chrousos GP, Karp B, Allen C, Flier JS, Gold PW
(1997) Human leptin levels are pulsatile and inversely related to
pituitary-adrenal function. Nature Med 3:575–579
Tuominen JA, Ebeling P, Laquier FW, Heiman ML, Stephens T,
Koivisto VA (1997) Serum leptin concentration and fuel
homeostasis in healthy man. Eur J Clin Invest 27:206–211
Weltman A, Pritzlaff CJ, Wideman L, Considine RV, Fryburg DA,
Gutgesell ME, Hartman ML, Veldhuis JD (2000) Intensity of
acute exercise does not affect serum leptin concentrations in
young men. Med Sci Sports Exerc 32:1556–1561
111