Lipolytic suppression following carbohydrate ingestion
limits fat oxidation during exercise
JEFFREY F. HOROWITZ, RICARDO MORA-RODRIGUEZ,
LAURI O. BYERLEY, AND EDWARD F. COYLE
The Human Performance Laboratory, Department of Kinesiology and Health Education
and Division of Nutritional Sciences, The University of Texas at Austin, Austin, Texas 78712
insulin; preexercise meal; glycemic index; stable isotopes
IT HAS LONG BEEN recognized that carbohydrate ingestion before exercise reduces fat oxidation during a
subsequent exercise bout (3). More recently, it has been
reported that this suppression persists for at least 4 h
after a meal (20). Thus fat oxidation in active people is
often under the influence of the insulin response from
normal dietary carbohydrate. However, little is known
about the mechanisms by which carbohydrate ingestion reduces fat oxidation during physical activity.
Fat oxidation during exercise involves several steps,
beginning with the hydrolysis of triglycerides (i.e.,
lipolysis) to liberate fatty acids that are transported,
either via plasma and/or through myoplasm, to the
muscle mitochondria for oxidation (30). Triglycerides
must first be hydrolyzed before the resultant fatty acids
can be oxidized. Therefore, equivalent rates of lipolysis
and fat oxidation could imply that lipolysis limits fatty
acid oxidation. After an overnight fast, lipolysis exceeds
fat oxidation by as much as 50% both at rest and during
exercise (25, 32), and thus lipolysis clearly does not
limit fat oxidation when a person is fasted. However,
administration of the antilipolytic agent nicotinic acid
E768
reduces plasma free fatty acid (FFA) concentration and
fat oxidation during exercise (13), but it has not been
determined if fat oxidation equals lipolysis under these
conditions. Like nicotinic acid, small elevations in
plasma insulin concentration (i.e., 10–30 µU/ml) dramatically reduce lipolysis at rest (4, 5), yet lipolysis still
appears to be in excess of fat oxidation at rest (5).
During exercise, however, fat oxidation increases severalfold above resting levels, making it more likely for a
suppression of lipolysis to limit fat oxidation.
No study, to our knowledge, has directly quantified
lipolysis during exercise after an increase in plasma
insulin concentration to determine the relationship
between lipolysis and fat oxidation during exercise.
This was the primary purpose of the present study. In
that lipolysis appears sensitive to even small increases
in plasma insulin, we fed subjects small amounts of
both a low-glycemic (i.e., fructose) and high-glycemic
carbohydrate (i.e., glucose) before exercise to raise
plasma insulin concentration only 10–30 µU/ml above
fasting levels. To more directly determine if an insulininduced suppression of lipolysis limits fat oxidation, we
also raised lipolysis via intravenous triglyceride and
heparin infusion to determine the extent to which this
raised fat oxidation.
A reduction in fat oxidation during exercise after the
ingestion of carbohydrate requires a compensatory
increase in carbohydrate oxidation to maintain energy
production. The source of the increase in carbohydrate
oxidation (i.e., blood glucose vs. muscle glycogen) is
debated (1, 8, 14). Whereas leg glucose uptake and
oxidation increase during exercise after a preexercise
carbohydrate meal compared with during fasting (1, 2),
conflicting evidence indicates that muscle glycogen
utilization either increases (8) or remains the same
(1, 14).
The principal aims of this study were to determine 1)
how the magnitude of the plasma insulin concentration
after ingestion of a low- vs. a high-glycemic carbohydrate meal before exercise influenced lipolysis during a
subsequent exercise bout, 2) if the suppression in
lipolysis after a relatively small preexercise carbohydrate meal reduced fat oxidation during exercise, and
3) how the relative contributions of blood glucose and
muscle glycogen were influenced by a reduction in
lipolysis and fat oxidation during exercise after carbohydrate ingestion.
METHODS
Subjects. Six healthy, active males participated in this
experiment. Their peak oxygen consumption (V̇O2 peak) while
cycling, body weight, and age were 3.91 6 0.24 l/min (52.8 6
0193-1849/97 $5.00 Copyright r 1997 the American Physiological Society
Downloaded from http://ajpendo.physiology.org/ by 10.220.33.6 on June 17, 2017
Horowitz, Jeffrey F., Ricardo Mora-Rodriguez, Lauri
O. Byerley, and Edward F. Coyle. Lipolytic suppression
following carbohydrate ingestion limits fat oxidation during
exercise. Am. J. Physiol. 273 (Endocrinol. Metab. 36): E768–
E775, 1997.—This study determined if the suppression of
lipolysis after preexercise carbohydrate ingestion reduces fat
oxidation during exercise. Six healthy, active men cycled 60
min at 44 6 2% peak oxygen consumption, exactly 1 h after
ingesting 0.8 g/kg of glucose (Glc) or fructose (Fru) or after an
overnight fast (Fast). The mean plasma insulin concentration
during the 50 min before exercise was different among Fast,
Fru, and Glc (8 6 1, 17 6 1, and 38 6 5 µU/ml, respectively;
P , 0.05). After 25 min of exercise, whole body lipolysis was
6.9 6 0.2, 4.3 6 0.3, and 3.2 6 0.5 µmol · kg21 · min21 and fat
oxidation was 6.1 6 0.2, 4.2 6 0.5, and 3.1 6 0.3
µmol · kg21 · min21 during Fast, Fru, and Glc, respectively (all
P , 0.05). During Fast, fat oxidation was less than lipolysis
(P , 0.05), whereas fat oxidation approximately equaled
lipolysis during Fru and Glc. In an additional trial, the same
subjects ingested glucose (0.8 g/kg) 1 h before exercise and
lipolysis was simultaneously increased by infusing Intralipid
and heparin throughout the resting and exercise periods
(Glc1Lipid). This elevation of lipolysis during Glc1Lipid
increased fat oxidation 30% above Glc (4.0 6 0.4 vs. 3.1 6 0.3
µmol · kg21 · min21; P , 0.05), confirming that lipolysis limited
fat oxidation. In summary, small elevations in plasma insulin
before exercise suppressed lipolysis during exercise to the
point at which it equaled and appeared to limit fat oxidation.
LIPOLYSIS AND FAT OXIDATION
To ensure that in vitro lipolysis of the infused triglycerides
during Glc1Lipid did not occur after sampling, we compared
plasma FFA concentrations using our blood storage method
with two methods reported to prevent in vitro lipolysis during
hypertriglyceridemia. These methods were 1) the same
method as presently used, with the exception of substituting
ethylene glycol-bis(b-aminoethyl ether)-N,N,N8,N8-tetraacetic acid (EGTA) for EDTA (11), and 2) blood samples immediately centrifuged and plasma placed in a test tube containing
0.2 ml of a 5 M NaCl solution and incubated for 30 min at
56°C (23). If the plasma FFA concentration were greater
using our present storage method compared with the others,
this would suggest that in vitro lipolysis may have occurred
after sampling during Glc1Lipid. However, no differences
existed in the mean plasma fatty acid concentration among
the different methods (i.e., 0.28 6 0.03, 0.30 6 0.02, 0.30 6
0.04 mM for EDTA-, EGTA-, and NaCl-treated samples,
respectively), indicating that significant in vitro lipolysis did
not presently occur.
Isotope enrichment sample preparation. Plasma samples (1
ml) were deproteinized by adding 1 ml of 0.3 N Ba(OH)2 and 1
ml of 0.3 N Zn(SO)4. Each tube was then vortexed and
incubated in an ice bath for 20 min. After centrifugation at
3,000 rpm for 15 min at 4°C, the supernatant was placed in
separate tubes for glucose (0.5 ml) and glycerol analysis (1.5
ml) and the water was removed from the tubes via vacuum
centrifugation (Savant Instruments, Farmingdale, NY). The
aldonitrile acetate derivative of glucose was prepared by
adding 100 µl of hydroxolamine-hydrochloride solution (20
mg/ml in pyridine) to the dried sample. After a 30-min
incubation at 100°C, 75 µl of acetic anhydride (Supelco,
Bellefonte, PA) were added. The samples remained incubating for an additional hour and then were evaporated under
N2. Before injection into the GC-MS, the samples were
reconstituted with ethyl acetate. The Tris-trimethylsilyl derivative of glycerol was prepared by reconstituting the dried
sample with 30 µl of a trimethylsilyl solution (Tri-Sil, Pierce,
Rockford, IL).
Preliminary testing. V̇O2 peak was determined while subjects
cycled an ergometer (Monark, model 819, Varberg, Sweden)
by using a continuous protocol that lasted 7–10 min. In addition,
2 days before each experimental trial, the subjects performed the
experimental exercise protocol (60 min at 45% V̇O2 peak) to
ensure homogeneity of the last exercise bout. The subjects ate
the exact same meals at the same time of day before each of
the four trials.
Measurement of gas exchange. As subjects inhaled through
a two-way Daniel’s valve, inspired air volume was measured
with a Parkinson-Cowan CD4 dry gas meter (Rayfield Equipment, Waitsfield, VT). The expired gases were continuously
sampled from a mixing chamber and analyzed for oxygen
(Applied Electrochemistry, model SA3, Ametek, Pittsburgh,
PA) and carbon dioxide (Beckman, model LB-2, Schiller Park,
IL). These instruments were interfaced with a computer
for calculations of the rate of oxygen consumption (V̇O2) and
respiratory exchange ratio (RER).
Calculations. Whole body lipolysis was quantified by measuring the rate of glycerol appearance in plasma (Ra glycerol ).
This calculation assumes that one mole of glycerol is liberated
for every mole of triglyceride hydrolyzed. It is theoretically
possible that Ra glycerol may underestimate whole body lipolysis due to phosphorylation of glycerol within the cell and/or by
incomplete triglyceride hydrolysis (16). Both of these possibilities would reduce glycerol appearance in the systemic circulation after lipolysis. Conversely, others indicate that neither of
these events occur in skeletal muscle or in adipose tissue (21,
31) and that nearly all glycerol liberated via lipolysis mixes
Downloaded from http://ajpendo.physiology.org/ by 10.220.33.6 on June 17, 2017
1.9 ml · kg21 · min21 ), 75.0 6 3.3 kg, and 26.5 6 3.8 yr,
respectively. Subjects were informed of the possible risks, and
each signed a consent form approved by the Internal Review
Board of The University of Texas at Austin.
Experimental protocol. On four separate occasions, the
subjects arrived at the laboratory in the morning after a 12-h
fast. Exactly 1 h before exercise, the subjects ingested 0.8 g of
carbohydrate/kg body weight (i.e., ,60 g) from 1) glucose
(Glc) to elicit a relatively high plasma insulin concentration,
2) fructose (Fru) to elicit a small elevation in plasma insulin
concentration, or 3) glucose ingestion followed by an intravenous infusion of a 20% triglyceride emulsion (Intralipid) and
heparin (Glc1Lipid) to increase lipolysis and prevent a
decline in plasma FFA concentration after glucose ingestion;
or they remained fasted (Fast). Both Glc and Fru were
provided as a 20% solution in water. During Glc1Lipid, as the
subjects ingested the 20% glucose solution 1 h before exercise,
we initiated a constant rate infusion of a 20% triglyceride
emulsion (i.e., Intralipid) and sodium heparin (after a bolus of
7.1 U/kg) to promote lipolysis of the infused triglycerides.
Infusion rates in the first two subjects (0.7 ml · kg21 · h21 of
triglyceride and 7.1 U heparin · kg21 · h21 ) produced plasma
FFA concentrations higher than Fast. A lower infusion rate in
the remaining four subjects (0.4 ml · kg21 · h21 of triglyceride
and 4.8 U heparin · kg21 · h21 ) elicited plasma FFA that closely
matched levels during Fast. Exactly 1 h after ingestion, the
subjects cycled for 60 min at 44 6 2% V̇O2 peak. The order of the
trials was counterbalanced, and they were separated by a
minimum of 48 h.
Isotope infusion. When the subject arrived at the laboratory, Teflon catheters were inserted into a forearm vein of
both arms of the subject (one for infusion, the other for blood
sampling), and a heating pad was affixed to the forearm and
hand of the sampling arm. A blood sample was then withdrawn for determination of background isotopic enrichment,
followed by a primed, constant-rate infusion of 6,6-d2-glucose
(i.e., 0.41 µmol · kg21 · min21; prime of 35 µmol/kg) and d5glycerol (i.e., 0.24 µmol · kg21 · min21; prime of 3.6 µmol/kg),
using calibrated syringe pumps (Harvard Apparatus, South
Natick, MA). Subjects were infused for at least 2 h before the
start of exercise to allow for attainment of isotopic equilibrium. These rates of isotope infusion were maintained
throughout the entire experiment.
Blood sampling and analysis. Blood samples were withdrawn immediately before and every 10 min after ingestion
during both rest and exercise. Each blood sample was divided
into three different tubes for subsequent analysis and immediately placed in an ice bath until the end of the trial. Three
milliliters of each blood sample were placed in evacuated
tubes containing 143 USP units of sodium heparin (Vacutainer, Becton Dickinson, Rutherford, NJ). These samples
were later analyzed for isotopic enrichment of the aldonitrile
acetate derivative of 6,6-d2-glucose (29) and the tris(hydroxymethyl)aminomethane (Tris)-trimethylsilyl derivative of d5glycerol (31), via gas chromotography-mass spectrophotometry (GC-MS). An additional 2 ml of each blood sample were
placed in a test tube containing 0.2 ml of an Aprotinin (0.5
TiU/ml) and EDTA (82 mM) solution and later analyzed for
plasma insulin concentration (radioimmunoassay; ICN Biomedicals, Costa Mesa, CA). The final 3 ml of each blood
sample were placed in a test tube containing 0.15 ml of EDTA
(82 mM) for later determination of plasma glycerol [fluorometric assay (12)], glucose (glucose oxidase autoanalyzer, Yellow
Springs Instruments, Yellow Springs, OH), and FFA [colorimetric assay (22)]. In each tube, plasma was separated by
centrifugation (3,000 rpm for 20 min at 4°C), immediately
frozen, and stored at 270° C until analysis.
E769
E770
LIPOLYSIS AND FAT OXIDATION
with the circulation after triglyceride hydrolysis, thus suggesting that Ra glycerol provides a reasonably accurate index of
lipolysis. Ra glycerol as well as the rate of appearance and
disappearance of glucose in plasma (i.e., Ra Glc and Rd Glc ) were
calculated using the non-steady-state equation of Steele (28),
modified for use with stable isotopes
Ra 5
F 2 Vd[C/(1 1 E)(dE/dt)]
Rd 5 Ra 2
E
Vd[(dC/dt)(1 1 E) 2 C(dE/dt)]
(1 1 E)2
RESULTS
Plasma glucose concentration. The average plasma
glucose concentration during the 50-min period before
exercise was high during both Glc and Glc1Lipid (7.3 6
0.4 and 6.6 6 0.6 mM, respectively; P , 0.05 vs. Fru
and Fast), moderate during Fru (5.4 6 0.2 mM; P ,
0.05 vs. Fast), and remained at basal levels during Fast
(4.8 6 0.2 mM). During exercise, no differences in
plasma glucose concentration existed among the trials
(Fig. 1).
Plasma insulin concentration. As designed, the
plasma insulin concentration was elevated to different
levels after the different carbohydrate meals. The mean
plasma insulin concentration during the 50-min period
before exercise was 38.5 6 5.4, 17.0 6 1.3, and 8.2 6 0.8
µU/ml during Glc, Fru, and Fast, respectively (all P ,
0.05; Fig. 2). Insulin concentration decreased during
exercise in both Glc and Fru (Fig. 2), and no differences
existed among trials by 40 min of exercise.
Ra glycerol. At rest 5 min before exercise, Ra glycerol during
Glc was ,60% lower than Fast (1.64 6 0.11 vs. 3.81 6
0.7 µmol · kg21 · min21; P , 0.05; Fig. 3). Fructose ingestion reduced resting Ra glycerol ,30% below Fast (2.75 6
0.51 µmol · kg21 · min21 ). However, this difference did
not reach statistical significance. During exercise,
Ra glycerol during Fast was significantly greater than both
Glc and Fru (P , 0.05). Ra glycerol during Fru was
intermediate to Fast (P , 0.05 at all times) and Glc
(P , 0.05 at 10 and 25 min; Fig. 3).
Plasma FFA concentration. The reduction in lipolysis
during Glc and Fru lowered (P , 0.05) plasma FFA
Fig. 1. Plasma glucose concentration during rest and exercise.
* Mean plasma glucose concentration at rest during Glucose,
Glucose1Lipid, or Fructose significantly greater than Fast experiments (see Experimental protocol for description of experimental
groups), P , 0.05. † Mean plasma glucose concentration at rest
during Glucose and Glucose1Lipid significantly greater than Fructose, P , 0.05.
concentration immediately before exercise compared
with Fast (0.16 6 0.01, 0.18 6 0.01, and 0.36 6 0.06
mM, respectively), and it remained low throughout
exercise (Fig. 4). Lipolysis of the infused triglycerides
during Glc1Lipid prevented the reduction in plasma
FFA concentration after glucose ingestion (Fig. 4). In a
subgroup of four subjects, a relatively low Intralipid
infusion rate was used, and in these subjects, the
plasma FFA concentration during exercise was identical to that observed while subjects fasted (Fig. 4B).
Fat oxidation and lipolysis. During Fast, lipolysis
was 15–25% in excess of fat oxidation throughout
exercise (6.9 6 0.6 vs. 6.1 6 0.2 µmol · kg21 · min21 at
20–30 min and 8.4 6 0.6 vs. 6.8 6 1.2 µmol · kg21 · min21
at 50–60 min; both P , 0.05; Fig. 5). During both Glc
and Fru, however, lipolysis and fat oxidation were
suppressed to the point at which fat oxidation was
closely matched to lipolysis (Fig. 5). Increasing lipolysis
after glucose ingestion via lipid-heparin infusion during Glc1Lipid increased fat oxidation 30% above Glc
Fig. 2. Plasma insulin concentration during rest and exercise. * Mean
plasma insulin concentration at rest and after 20 min of exercise
during Glucose, Glucose1Lipid, and Fructose significantly greater
than Fast, P , 0.05. † Mean plasma insulin concentration at rest
during Glucose and Glucose1Lipid significantly greater than Fructose, P , 0.05.
Downloaded from http://ajpendo.physiology.org/ by 10.220.33.6 on June 17, 2017
where F is isotope infusion rate, Vd is volume of distribution
(i.e., estimated to be 230 ml/kg for glycerol and 100 ml/kg for
glucose), C is plasma concentration of the tracee, E is tracer
isotopic enrichment, and dE/dt and dC/dt are maximum rates
of change of enrichment and concentration, respectively, with
respect to time. Fat (i.e., triglyceride) and carbohydrate
oxidation were calculated from V̇O2 and RER (nonprotein
respiratory quotient), measured from expired air during the
20- to 30- and 50- to 60-min periods of exercise (17). Muscle
glycogen oxidation was calculated as the difference between
total carbohydrate oxidation and Rd Glc. Coggan et al. (7)
reported that ,90% of Rd Glc is oxidized, and thus it provides a
reasonable representation of blood glucose oxidation. These
calculations have been used in previous studies (25, 26).
Statistical analysis. Significant differences among trials
were identified using a two-way analysis of variance (treatment by time) for repeated measures with Tukey’s post hoc
analysis. Planned comparisons for mean values were evaluated using a paired Student’s t-test with a Bonferroni correction factor, P , 0.05.
E771
LIPOLYSIS AND FAT OXIDATION
(4.0 6 0.4 vs. 3.1 6 0.3 µmol · kg21 · min21, respectively)
at 20–30 min of exercise. However, fat oxidation during
Glc1Lipid was not restored to levels observed during
Fast (Table 1). In addition, fat oxidation was similar
when the lipid-heparin infusion increased plasma FFA
concentrations to either 0.4 mM (i.e., fat oxidation 5
Fig. 5. Relationship between lipolysis (Ra glycerol ) and fat oxidation.
* Lipolysis significantly greater than fat oxidation, P , 0.05. A: 20- to
30-min period of exercise. B: 50- to 60-min period of exercise.
5.0 6 0.6 µmol · kg21 · min21; n 5 4) or 0.9 mM (i.e., fat
oxidation 5 4.1 6 0.4 µmol · kg21 · min21; n 5 2).
Plasma glucose kinetics. At rest, the mean Ra Glc and
Rd Glc during both Glc and Glc1Lipid were two- to
fourfold greater than Fast (P , 0.05), whereas Fru was
20–30% greater than Fast (P , 0.05; Fig. 6). Throughout exercise, Ra Glc and Rd Glc during Glc and Glc1Lipid
remained more than twofold greater than Fast and
,50% greater than Fru (P , 0.05). Ra Glc and Rd Glc
during Fru was 50–60% greater than Fast (P , 0.05).
No difference in Ra Glc or Rd Glc was observed between Glc
and Glc1Lipid at any time (Fig. 6).
Muscle glycogen oxidation. During the 20- to 30-min
period of exercise, muscle glycogen oxidation during
Table 1. Fat oxidation rate
Exercise Time
Fig. 4. Plasma free fatty acid (FFA) concentration during rest and
exercise. * Mean plasma FFA concentration during Glucose and
Fructose significantly less than Fast and Glucose1Lipid immediately
before (0 min) and throughout exercise, P , 0.05. A: response of all 6
subjects. B: response of 4 subjects in which FFA concentration was
increased during Glucose1Lipid to exactly match the Fast response.
Trial
20–30 min
50–60 min
Fast
Fructose
Glucose
Glucose 1 Lipid
6.1 6 0.2
4.2 6 0.5*
3.1 6 0.3*†‡
4.0 6 0.4*
6.8 6 0.4
5.3 6 0.6*
4.9 6 0.7*
5.3 6 0.4*
Values are means 6 SE and are in µmol · kg21 · min21. * Significantly different from Fast, P , 0.05. † Significantly different from
Fructose, P , 0.05. ‡ Significantly different from Glucose 1 Lipid, P ,
0.05. See Experimental protocol for description of experimental trials.
Downloaded from http://ajpendo.physiology.org/ by 10.220.33.6 on June 17, 2017
Fig. 3. Rate of appearance of glycerol (Ra glycerol ) during rest and
exercise. * Significantly less than Fast, P , 0.05. † Significantly less
than Fructose, P , 0.05.
E772
LIPOLYSIS AND FAT OXIDATION
Glc and Fru was significantly greater (P , 0.05) than
Fast (87 6 8, 84 6 9, and 70 6 8 µmol · kg21 · min21,
respectively; Table 2). During this same period, the
increase in lipolysis and the resultant elevation in
plasma FFA concentration during Glc1Lipid lowered
muscle glycogen oxidation compared with Glc (74 6 8
vs. 87 6 8 µmol · kg21 · min21, respectively; P , 0.05)
and there was no difference in muscle glycogen oxidation in Fast compared with Glc1Lipid (70 6 8 vs. 74 6
8 µmol · kg21 · min21, respectively). During the 50- to
60-min period of exercise, no differences in muscle
Table 2. Total carbohydrate oxidation, rate of glucose
disappearance, and muscle glycogen oxidation
20- to 30-Min Period of Exercise
Trial
Total
carbohydrate oxidation
Rd glucose
Muscle
glycogen oxidation
Fast
Fructose
Glucose
Glucose 1 Lipid
85 6 9
105 6 8*
117 6 6*†‡
107 6 7*
15 6 1
21 6 2*
30 6 3*†
33 6 2*†
70 6 8
84 6 9*
87 6 8*‡
74 6 8
Values are means 6 SE and are in µmol · kg21 · min21. Rd glucose , rate
of glucose disappearance. * Significantly different from Fast, P ,
0.05. † Significantly different from Fructose, P , 0.05. ‡ Significantly
different from Glucose 1 Lipid, P , 0.05.
Fig. 7. Percent energy expenditure derived from muscle glycogen,
blood glucose, and fat during 20- to 30-min period of exercise.
* Significantly different from Fast, P , 0.05. † Significantly different
from Glucose, P , 0.05.
Downloaded from http://ajpendo.physiology.org/ by 10.220.33.6 on June 17, 2017
Fig. 6. A: rate of appearance of glucose (Ra glucose ) during rest and
exercise. B: rate of disappearance of glucose (Rd glucose ) during rest and
exercise. * Mean Ra and Rd of glucose during Glucose, Glucose1Lipid,
and Fructose significantly greater than Fast, P , 0.05. † Mean Ra and
Rd of glucose during Glucose and Glucose1Lipid significantly greater
than Fructose, P , 0.05.
glycogen oxidation were observed among trials (62 6 7,
64 6 6, 68 6 5, and 70 6 8 µmol · kg21 · min21 for Fast,
Fru, Glc, and Glc1Lipid, respectively).
Energy expenditure. Figure 7 summarizes the estimated relative caloric contribution from fat, blood
glucose, and muscle glycogen during the 20- to 30-min
period of exercise. The total rate of energy expenditure
was similar during all trials (108 6 11 cal · kg21 · min21 ).
During Fast, the relative energy expenditure from fat,
blood glucose, and muscle glycogen was 44 6 2, 9 6 2,
and 47 6 6, respectively (Fig. 7). Compared with Fast,
Glc reduced fat oxidation by one-half, representing only
22 6 2% of the total energy, whereas the contributions
of both blood glucose and muscle glycogen increased to
20 6 2 and 58 6 6%, respectively (all P , 0.05 vs. Fast).
Lipolysis of the infused triglycerides during Glc1Lipid
increased the relative contribution of fat to 30% (P ,
0.05 vs. Glc) and returned the relative contribution of
muscle glycogen back to fasting levels (50 6 6%; P ,
0.05 vs. Glc). During Fru, the relative contributions
from fat (30 6 4%) and blood glucose (14 6 2%) were
intermediate to those during Fast and Glc (P , 0.05 vs.
Fast and Glc). However, the relative contribution of
muscle glycogen was the same as Glc (56 6 6%). During
the 50- to 60-min period of exercise, the relative contribution of fat during Fast (48 6 3%) was significantly
greater than Glc, Glc1Lipid, and Fru (34 6 4, 37 6 3,
and 37 6 5, respectively; P , 0.05 vs. Fast), whereas
the contribution from blood glucose during Fast (11 6
3%) was lower (22 6 4, 19 6 4, and 22 6 4%,
respectively; P , 0.05 vs. Fast). However, no differences
in fat oxidation or Rd Glc existed among Glc, Glc1Lipid,
and Fru. During this same period, the contribution
from muscle glycogen was not different among any of
the trials (41 6 5%, 44 6 4%, 45 6 5%, and 41 6 4% for
Fast, Glc, Glc1Lipid, and Fru, respectively).
LIPOLYSIS AND FAT OXIDATION
DISCUSSION
has additional effects on muscle that concomitantly
reduced fat oxidation.
We have attributed the suppression in lipolysis to the
elevation in plasma insulin concentration despite an
elevation in plasma glucose concentration and a likely
fall in plasma glucagon concentration after carbohydrate ingestion. Previous reports indicate that alterations in glucose (6) and glucagon concentration (19) do
not affect lipolysis when plasma insulin concentration
does not change even slightly. In addition, we recognize
the possibility that the measurement of Ra glycerol may
underestimate whole body lipolysis. Potential causes
for this underestimation include some activity of glycerol kinase within adipose and/or muscle tissue and
incomplete hydrolysis of a triglyceride into di- or
monoglycerides (16). Both of these events would reduce
glycerol appearance in the systemic circulation after
lipolysis. To the contrary, reports indicate that neither
of these phenomena occur within skeletal muscle or in
adipose tissue (21, 31). However, despite the possible
underestimation of whole body lipolysis in the present
study, we have clearly demonstrated that lipolysis was
suppressed during exercise after carbohydrate ingestion, compared with an overnight fast. Additionally, by
demonstrating that an increase in lipolysis during
Glc1Lipid increased fat oxidation 30% above Glc, we
established that this suppression in lipolysis during
exercise after glucose ingestion was sufficient to limit
fat oxidation.
Increasing lipolysis during Glc1Lipid significantly
raised fat oxidation from 3.1 6 0.3 to 4.0 6 0.4
µmol · kg21 · min21, but it did not restore fat oxidation to
levels observed during Fast (6.1 6 0.2 µmol · kg21 ·
min21 ). This occurred despite the fact that plasma FFA
concentration was elevated to a level as great (0.4 mM
for n 5 4; Fig. 3B) or greater than when fasted (0.9 mM
for n 5 2). Thus the remaining reduction in fat oxidation below fasted levels was not likely a result of an
inadequate increase in plasma FFA concentration. This
indicates that the lower fat oxidation during Glc1Lipid
vs. Fast was due to an inhibition within the exercising
muscle. Unfortunately, we were unable to differentiate
between adipose and intramuscular triglyceride lipolysis in the present study. Because infusion of Intralipid
and heparin only increases lipolysis in the vasculature,
it is possible that a reduction in intramuscular triglyceride lipolysis after the ingestion of glucose may have
reduced fat oxidation during Glc1Lipid compared with
Fast (9). We have also recently found that acute hyperglycemia, hyperinsulinemia, and the resultant increased glycolytic flux reduce long-chain fatty acid
transport and oxidation within muscle mitochondria
(9). Therefore, it is possible that carbohydrate ingestion
regulates fat oxidation through the combined effects of
a reduction in whole body lipolysis (i.e., both adipose
and intramuscular) superimposed on an inhibition of
long-chain fatty acid oxidation in muscle.
The factors that regulate the relative contribution of
fat and carbohydrate to total energy production are not
well understood. More than 30 years ago, Randle and
co-workers (24) first proposed the classic hypothesis
Downloaded from http://ajpendo.physiology.org/ by 10.220.33.6 on June 17, 2017
The present study is the first to our knowledge to
quantify lipolysis during exercise after an elevation in
plasma insulin concentration and to demonstrate that
fat oxidation was suppressed to levels closely matching
the rate of lipolysis. The notion that fat oxidation can be
‘‘limited’’ by a low lipolytic rate stems from the fact that
a triglyceride must first be hydrolyzed before the
resultant fatty acids can be oxidized. Because there is a
very small pool of unesterified fatty acids within the
human body (,10–40 µmol/kg), the rate of fat oxidation cannot exceed the rate of lipolysis for more than a
few minutes during exercise. It follows that when
lipolysis is suppressed to low levels, the rate of fat
oxidation may be limited by the reduced availability of
unesterified fatty acids due to the low lipolytic rate.
Lipolysis exceeds fat oxidation both at rest and
during exercise when subject is fasted (25, 32). In the
present study, lipolysis exceeded fat oxidation by 15–
25% when subjected was fasted. Thus, because more
fatty acids are liberated via lipolysis than are oxidized,
fat oxidation is not limited by lipolysis when subjects
are fasted (25, 32). It has been reported that lipolysis
was suppressed .60% at rest after an elevation in
plasma insulin concentration to only ,30 µU/ml (5).
This suppression, together with the high rate of triglyceride reesterification at rest (5, 32), appears to account
for the observation of Campbell et al. (5) that small
elevations in plasma insulin concentration (e.g., to only
,30 µU/ml) in resting subjects reduced the rate of FFA
appearance in plasma, plasma FFA concentration, and
the rate of FFA disappearance from plasma (Rd FFA ) to
the point that Rd FFA equaled fat oxidation. Therefore, at
rest, a reduction in mobilization of FFA from adipose
tissue after a relatively small increase in plasma
insulin concentration may limit fat oxidation, whereas
whole body lipolysis remains in excess of fat oxidation.
During exercise, both lipolysis and fat oxidation
increase above resting levels (25, 32). However, we have
presently reported that a relatively small elevation in
plasma insulin concentration before exercise (9 µU/ml
during Fru and 30 µU/ml during Glc) attenuated the
rise in lipolysis during exercise to the point that the
rate of lipolysis was similar to the rate of fat oxidation
(Fig. 5). Thus, because fat oxidation cannot exceed the
rate of lipolysis, a close matching of lipolysis and fat
oxidation suggests that the lipolytic suppression was
sufficient to limit fat oxidation. However, lipolysis and
fat oxidation were measured using different methods
(i.e., isotope dilution and indirect calorimetry, respectively). As such, the close matching of these measurements provides only indirect evidence that lipolysis
limited fat oxidation during exercise after carbohydrate
ingestion. Direct support for the idea that lipolysis did
indeed limit fat oxidation was obtained from our observation that fat oxidation increased 30% with an increase in lipolysis (i.e., via Intralipid and heparin
infusion) after glucose ingestion. However, this increase in lipolysis did not restore fat oxidation to
fasting levels, suggesting that carbohydrate ingestion
E773
E774
LIPOLYSIS AND FAT OXIDATION
0.05 Glc vs. Fast). Therefore, muscle glycogen oxidation
increased in response to a reduced supply of energy
from the combination of fat (i.e., from both plasma FFA
and intramuscular stores) and blood glucose. Although
AMP and Pi concentrations were not measured in the
present study, it has recently been reported that low
plasma FFA concentrations during exercise, similar to
those reported during Glc in the present study (i.e.,
,0.2 mM), increased muscle AMP and Pi concentrations (10). Because AMP and Pi increase glycogen
phosphorylase activity, this may explain the greater
rate of muscle glycogen oxidation during Glc compared
with Fast.
The transient decline in plasma glucose concentration during the first 20 min of exercise is a phenomenon
characteristic of preexercise carbohydrate ingestion (1,
8). Obviously this reflects an imbalance between Rd Glc
and Ra Glc (i.e., Rd Glc . Ra Glc ). However, limited information is available describing this imbalance (18). Presently, the reduction in plasma glucose concentration
during the first 20 min of exercise after carbohydrate
ingestion was due to a large increase in Rd Glc without a
compensatory increase in Ra Glc. Muscle contraction and
exercise both increase glucose uptake (15), resulting in
the relatively large increase in Rd Glc at the onset of
exercise when subjects are fed carbohydrate. However,
Ra Glc did not increase markedly with the onset of
exercise. After 10 min of exercise Rd Glc declined, likely
due to the rapid fall in plasma glucose and insulin
concentration during exercise (18). After 25 min of
exercise, Ra Glc slightly (not significant) exceeded Rd Glc,
and thereafter plasma glucose concentration returned
to fasted levels.
In summary, when fasted and plasma insulin concentrations are at basal levels (,10 µU/ml), lipolysis is in
excess of fat oxidation during exercise. However, only a
relatively small increase in plasma insulin concentration before exercise (,10 µU/ml after fructose ingestion
and ,30 µU/ml after glucose ingestion) reduced lipolysis to the point that it equaled fat oxidation. The idea
that lipolysis did indeed limit fat oxidation was demonstrated by the fact that fat oxidation increased 30%
with an increase in lipolysis (i.e., via Intralipid and
heparin infusion) after glucose ingestion. However, this
increase in lipolysis did not restore fat oxidation to
fasting levels, suggesting that carbohydrate ingestion
has additional effects within skeletal muscle that concomitantly reduced fat oxidation. The increase in lipolysis and plasma FFA concentration during Glc1Lipid,
however, did increase fat oxidation sufficiently to reverse the increase in muscle glycogen oxidation observed during exercise after glucose ingestion alone.
The present findings demonstrate that carbohydrate
ingestion before exercise resulting in a 10–30 µU/ml
elevation in plasma insulin concentration reduced fat
oxidation primarily by suppressing lipolysis during
exercise.
We greatly appreciate the technical support of Dr. Andrew Coggan
and Michael Sullivan. We additionally appreciate the assistance from
Paul Below, Melissa Domenick, Pete Flatten, Ricardo Fritzsche, Wes
Downloaded from http://ajpendo.physiology.org/ by 10.220.33.6 on June 17, 2017
that fat oxidation regulates carbohydrate oxidation
(i.e., ‘‘glucose-fatty acid cycle’’). This theory states that
increases in mitochondrial acetyl-CoA and cytosolic
citrate concentrations, secondary to an increase in fat
oxidation, inhibit pyruvate dehydrogenase and
phosphofructokinase and subsequently reduce glycolysis and glucose uptake (24). However, little evidence is
available to support the functional significance of this
mechanism in human skeletal muscle during exercise
(10). Investigations have failed to detect elevations in
muscle acetyl-CoA, citrate, and glucose 6-phosphate
concentrations coincident to an elevation in fat oxidation during exercise after an increase in plasma FFA
concentration (10, 11). Additional evidence for the
absence of this regulatory mechanism in exercising
humans was the present observation that a 30% increase in fat oxidation during Glc1Lipid compared
with Glc did not reduce Rd Glc (i.e., glucose uptake),
which agrees with the recent finding of Romijn et al.
(26). Carbohydrate ingestion and the resultant increase in plasma insulin concentration suppressed
lipolysis while additionally increasing Rd Glc. These are
two powerful mechanisms by which carbohydrate metabolism regulates fat oxidation.
Exercise and fasting are effective means to increase
fat oxidation. Thus exercise is often prescribed to
individuals attempting to reduce body fat. Our present
findings indicate that when exercise is preceded by a
carbohydrate meal, lipolysis and fat oxidation are
reduced. How this affects body fat and weight regulation over periods of days to months is not clear. It has
recently been reported that total fat oxidation was
reduced ,30% over an 8-h period when carbohydrate
was ingested before exercise, compared with ingestion
after exercise (27). Reducing lipolysis and impairing
daily fat oxidation by ingesting carbohydrate before
exercise increases the likelihood of being in a positive
lipid balance and thus may increase fat deposition.
Presently, we have demonstrated that ingestion of
either a low-glycemic carbohydrate (i.e., fructose) or a
high-glycemic carbohydrate (i.e., glucose) suppresses
lipolysis and fat oxidation, at least during the 2-h
period of this study. Due to the sensitivity of lipolysis to
even small elevations of insulin, it appears that to
maintain high rates of fat oxidation at rest and during
subsequent exercise, people should not eat even small
amounts of carbohydrate before exercise.
The reduction in lipolysis and fat oxidation after
carbohydrate ingestion necessitates a compensatory
increase in carbohydrate oxidation to maintain energy
production during exercise. Carbohydrate ingestion
and the resultant insulin response increased blood
glucose uptake and presumably oxidation during exercise, as previously reported (1, 2). However, in the
present study, the increase in blood glucose uptake
during the 20- to 30-min period of exercise during Glc
compared with Fast (12 cal · kg21 · min21; P , 0.05) was
only about one-half as great as the reduction in fat
oxidation (23 cal · kg21 · min21; Fig. 7), and, as a result,
the calculated rate of muscle glycogen oxidation appeared to increase 11 cal · kg21 · min21 (i.e., 22%; P ,
LIPOLYSIS AND FAT OXIDATION
Earwood, Matt Oseto, Tom Switzer, Trace Adams, and the participants of this study.
This study was supported by Grant DK-46017 from National
Institute of Diabetes and Digestive and Kidney Diseases (to R. R.
Wolfe and E. F. Coyle), Mars, and the Gatorade Sports Science
Institute.
Address for reprint requests: E. F. Coyle, The Univ. of Texas at
Austin, Rm. 222 Bellmont Hall, Austin, TX 78712.
Received 13 February 1997; accepted in final form 25 June 1997.
REFERENCES
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
endurance cycling performance. Med. Sci. Sports Exercise 19:
33–36, 1987.
Ivy, J. L., and J. O. Holloszy. Persistent increase in glucose
uptake by rat skeletal muscle following exercise. Am. J. Physiol.
241 (Cell Physiol. 10): C200–C203, 1981.
Kurpad, A., K. Khan, A. G. Calder, S. Coppack, K. Frayn,
and I. MacDonald. Effect of noradrenaline on glycerol turnover
and lipolysis in the whole body and subcutaneous adipose tissue
in humans in vivo. Clin. Sci. (Lond.) 86: 177–184, 1994.
Lusk, G. The Elements of the Science of Nutrition. Philadelphia,
PA: Saunders, 1928.
Marmy-Conus, N., S. Fabris, J. Proietto, and M. Hargreaves. Preexercise glucose ingestion and glucose kinetics
during exercise. J. Appl. Physiol. 81: 853–857, 1996.
Miles, J. M., and M. D. Jensen. Does glucagon regulate adipose
tissue lipolysis? J. Clin. Endocrinol. Metab. 77: 5A–5B, 1993.
Montain, S. J., M. K. Hopper, A. R. Coggan, and E. F. Coyle.
Exercise metabolism at different time intervals after a meal. J.
Appl. Physiol. 70: 882–888, 1991.
Newsholme, E. A., and K. Taylor. Glycerol kinase activity in
muscles from vertebrates and invertebrates. Biochem. J. 112:
465–474, 1969.
Novak, M. Colorimetric ultramicro method for determination of
free fatty acids. J. Lipid Res. 6: 431–433, 1965.
Peterson, J., B. E. Bihain, G. Bengtsson-Olivecrona, R. J.
Deckelbaum, Y. A. Carpentier, and T. Olivecrona. Fatty acid
control of lipoprotein lipase: a link between energy metabolism
and lipid transport. Proc. Natl. Acad. Sci. USA 87: 909–913,
1990.
Randle, P. J., P. B. Garland, C. N. Hales, and E. A. Newsholme. The glucose fatty-acid cycle: its role in insulin sensitivity
and the metabolic disturbances of diabetes mellitus. Lancet 1:
785–789, 1963.
Romijn, J. A., E. F. Coyle, L. Sidossis, A. Gastaldelli, J. F.
Horowitz, E. Endert, and R. R. Wolfe. Regulation of endogenous fat and carbohydrate metabolism in relation to exercise
intensity and duration. Am. J. Physiol. 265 (Endocrinol. Metab.
28): E380–E391, 1993.
Romijn, J. A., E. F. Coyle, X.-J. Zhang, L. S. Sidossis, and
R. R. Wolfe. Relationship between fatty acid delivery and fatty
acid oxidation during strenuous exercise. J. Appl. Physiol. 79:
1939–1945, 1995.
Schneiter, P., V. Di Vetta, E. Jéquier, and L. Tappy. Effect of
physical exercise on glycogen turnover and net substrate utilization according to the nutritional state. Am. J. Physiol. 269
(Endocrinol. Metab. 32): E1031–E1036, 1995.
Steele, R. Influences of glucose loading and of injected insulin on
hepatic glucose output. Ann. NY Acad. Sci. 82: 420–430, 1959.
Tserng, K.-Y., and S. C. Kalhan. Estimation of glucose carbon
recycling and glucose turnover with [U-13C]glucose. Am. J.
Physiol. 245 (Endocrinol. Metab. 8): E476–E482, 1983.
Van der Vusse, G. J., and R. S. Reneman. Lipid metabolism in
muscle. In: Handbook of Physiology. Exercise: Regulation and
Integration of Multiple Systems. Bethesda, MD: Am. Physiol.
Soc., 1996, sect. 12, chapt. 21, p. 952–994.
Wolfe, R. R. Radioactive and Stable Isotope Tracers in Biomedicine: Principles and Practice of Kinetic Analysis (1st ed.). New
York: Wiley-Liss, 1992.
Wolfe, R. R., S. Klein, F. Carraro, and J.-M. Weber. Role of
triglyceride-fatty acid cycle in controlling fat metabolism in
humans during and after exercise. Am. J. Physiol. 258 (Endocrinol. Metab. 21): E382–E389, 1990.
Downloaded from http://ajpendo.physiology.org/ by 10.220.33.6 on June 17, 2017
1. Ahlborg, G., and O. Bjorkman. Carbohydrate utilization by
exercising muscle following pre-exercise glucose ingestion. Clin.
Physiol. 7: 181–195, 1987.
2. Ahlborg, G., and P. Felig. Substrate utilization during prolonged exercise preceded by ingestion of glucose. Am. J. Physiol.
233 (Endocrinol. Metab. Gastrointest. Physiol. 2): E188–E194,
1977.
3. Bøje, O. Arbeithypoglykamie nach glukoseeingabe. Skand. Arch.
Physiol. 83: 308–312, 1940.
4. Bonadonna, R. C., L. C. Groop, K. Zych, M. Shank, and R. A.
DeFronzo. Dose-dependent effect of insulin on plasma free fatty
acid turnover and oxidation in humans. Am. J. Physiol. 259
(Endocrinol. Metab. 22): E736–E750, 1990.
5. Campbell, P. J., M. G. Carlson, J. O. Hill, and N. Nurjhan.
Regulation of free fatty acid metabolism by insulin in humans:
role of lipolysis and reesterification. Am. J. Physiol. 263 (Endocrinol. Metab. 26): E1063–E1069, 1992.
6. Cersosimo, E., S. Coppack, and M. Jensen. Lack of effect of
hyperglycemia on lipolysis in humans. Am. J. Physiol. 265
(Endocrinol. Metab. 28): E821–E824, 1993.
7. Coggan, A. R., W. M. Kohrt, R. J. Spina, D. M. Bier, and J. O.
Holloszy. Endurance training decreases plasma glucose turnover and oxidation during moderate-intensity exercise in men. J.
Appl. Physiol. 68: 990–996, 1990.
8. Costill, D. L., E. F. Coyle, G. Dalsky, W. Evans, W. Fink, and
D. Hoopes. Effects of plasma FFA and insulin on muscle
glycogen usage during exercise. J. Appl. Physiol. 43: 693–699,
1977.
9. Coyle, E. F., A. E. Jeukendrup, A. J. M. Wagenmakers, and
W. H. M. Saris. Fatty acid oxidation is directly regulated by
carbohydrate metabolism during exercise. Am. J. Physiol. 273
(Endocrinol. Metab. 36): E268–E275, 1997.
10. Dyck, D. J., S. J. Peters, P. Wendling, A. Chesley, E.
Hultman, and L. L. Spriet. Regulation of muscle glycogen
phosphorylase activity during intense aerobic cycling with elevated FFA. Am. J. Physiol. 270 (Endocrinol. Metab. 33): E116–
E125, 1996.
11. Dyck, D. J., C. T. Putman, G. J. F. Heigenhauser, E.
Hultman, and L. L. Spriet. Regulation of fat-carbohydrate
interaction in skeletal muscle during intense aerobic cycling.
Am. J. Physiol. 265 (Endocrinol. Metab. 28): E852–E859, 1993.
12. Eggstein, M., and E. Kuhlmann. Triglycerides and glycerol.
Determination after alkaline hydrolysis. In: Methods of Enzymatic Analysis (2nd ed.), edited by H. U. Bergmeyer. New York:
Academic, 1974, p. 1825–1831.
13. Gollnick, P. D., B. Pernow, B. Essen, E. Jansson, and B.
Saltin. Availability of glycogen and plasma FFA for substrate
utilization in leg muscle of man. Clin. Physiol. 1: 27–42, 1981.
14. Hargreaves, M., D. L. Costill, W. J. Fink, D. S. King, and
R. A. Fielding. Effect of pre-exercise carbohydrate feedings on
15.
E775