ORIGINAL
E n d o c r i n e
ARTICLE
R e s e a r c h
Sugar-Sweetened Beverages With Moderate Amounts
of Fructose, but Not Sucrose, Induce Fatty Acid
Synthesis in Healthy Young Men: A Randomized
Crossover Study
Michel Hochuli,* Isabelle Aeberli,* Adrienne Weiss, Martin Hersberger,
Heinz Troxler, Philipp A. Gerber, Giatgen A. Spinas, and Kaspar Berneis‡
Division of Endocrinology, Diabetes, and Clinical Nutrition (M.Ho., I.A., P.A.G., G.A.S., K.B.), University
Hospital Zurich, 8091 Zurich, Switzerland; Division of Clinical Chemistry and Biochemistry (A.W., M.He.,
H.T.), University Children’s Hospital Zurich, 8032 Zurich, Switzerland; Human Nutrition Laboratory (I.A.),
Institute of Food, Nutrition, and Health and Competence Center for Systems Physiology and Metabolic
Diseases (P.A.G., G.A.S.), ETH Zurich, 8092 Zurich, Switzerland; Zurich Center for Integrative Human
Physiology (K.B.), 8057 Zurich, Switzerland
Context: The impact of sugar-sweetened beverages (SSB) on lipid metabolism when consumed in
moderate amounts by normal weight subjects is debated.
Objective: The objective of the study was to investigate the effect of different types of sugars in
SSB on fatty acid metabolism (ie, fatty acid synthesis and oxidation) in healthy young men.
Design: Thirty-four normal-weight men were studied in a randomized crossover study. Four isocaloric 3-week interventions with SSB were performed in random order: medium fructose (MF; 40
g/d); high fructose (HF; 80 g/d), high sucrose (HS; 80 g/d), and high glucose (HG; 80g/d). Fasting total
plasma fatty acid composition was measured after each intervention. Acylcarnitines were measured in the fasting state and after a euglycemic hyperinsulinemic clamp in nine subjects.
Results: The relative abundance of palmitate (16:0) and the molar fatty acid ratio of palmitate to
linoleic acid (16:0 to18:2) as markers of fatty acid synthesis were increased after HF [relative abundance of palmitate: 22.97% ⫾ 5.51% (percentage of total fatty acids by weight ⫾SD)] and MF
(26.1% ⫾ 1.7%) compared with HS (19.40% ⫾ 2.91%, P ⬍ .001), HG (19.43% ⫾3.12 %, P ⬍ .001),
or baseline (19.40% ⫾ 2.79%, P ⬍ .001). After HS and HG, the relative abundance of palmitate was
equal to baseline. Fasting palmitoylcarnitine was significantly increased after HF and HS (HF and
HS vs. HG: P ⫽ .005), decreasing after inhibition of lipolysis by insulin in the clamp.
Conclusions: When consumed in moderate amounts, fructose but not sucrose or glucose in SSB
increases fatty acid synthesis (palmitate), whereas fasting long-chain acylcarnitines are increased
after both fructose and sucrose, indicating an impaired ␤-oxidation flux. (J Clin Endocrinol Metab
99: 2164 –2172, 2014)
D
ietary intake of free fructose in the United States increased during the last decades, in parallel with an
increased use of high fructose corn syrup instead of sucrose. Currently a daily amount of approximately 70 g of
added fructose (or 10% of total calories) is an average
intake in young males, with a high proportion of added
fructose stemming from sugar-sweetened beverages (1).
According to the Swiss Nutritional Board, added sugar in
ISSN Print 0021-972X ISSN Online 1945-7197
Printed in U.S.A.
Copyright © 2014 by the Endocrine Society
Received October 22, 2013. Accepted February 14, 2014.
First Published Online March 6, 2014
‡
, Deceased, March 28, 2014.
* M.Ho. and I.A. contributed equally to this work.
Abbreviations: BMI, body mass index; HF, high fructose; HG, high glucose; HS, high sucrose; LDL, low-density lipoprotein; MF, medium fructose; MUFA, monounsaturated fatty
acid; PUFA, polyunsaturated fatty acid; SCD, stearoyl-CoA desaturase; SSB, sugar sweetened beverage; VLDL, very low-density lipoprotein.
2164
jcem.endojournals.org
J Clin Endocrinol Metab, June 2014, 99(6):2164 –2172
doi: 10.1210/jc.2013-3856
doi: 10.1210/jc.2013-3856
food and beverages in Switzerland is mostly in the form of
sucrose and is consumed at approximately 18% of total
daily calories (corresponding to 9% from fructose) in
2002. Fructose and other caloric sweeteners have negative
metabolic effects, such as dyslipidemia (2) with decreased
low-density lipoprotein (LDL)-particle size (3, 4), a rise of
inflammatory markers (5, 6), and an increased risk for
type 2 diabetes (7). Mechanistic studies revealed an increased de novo lipogenesis, blood triglycerides, and hepatic insulin resistance (3, 8, 9) as key findings after high
to very high doses of fructose. We found previously that
consumption of even moderate amounts of fructose result
in smaller LDL particle size and increased hepatic insulin
resistance in healthy young men (10, 11). Fructose appears
to be a stronger inducer of lipogenesis than glucose (12–
14). Hepatic fructose metabolism bypasses steps of glycolysis (ie, glucose metabolism) regulated by the energy
status of the cell, leading to enhanced synthesis of precursors for fatty acid synthesis. Fructose stimulates hepatic de
novo lipogenesis (3, 15–17) and very low-density lipoprotein (VLDL)-triglyceride secretion, and may decrease
VLDL-triglyceride clearance (18, 19). Furthermore, there
is also an enhanced and extended activity of regulator
proteins involved in de novo lipogenesis (eg, sterol regulatory element-binding protein-1 [SREBP-1], carbohydrate-response element-binding protein [ChREBP]) with
fructose, which is independent from the action of insulin
(17).
Human and animal studies suggest that ingestion of
fructose may also affect lipid degradation via ␤-oxidation,
although the evidence appears less clear compared with
the effects of fructose on lipogenesis (20 –22). The differential effect of various sugars (sucrose, fructose, or glucose) contained in sugar-sweetened beverages (SSBs) on
fatty acid metabolism in men has so far been compared
only after feeding high amounts (at 25% of total daily
energy intake) of sugar. Although both fructose and sucrose (but not glucose) resulted in similarly increased postprandial triglyceride synthesis (23, 24), to our knowledge
it is not known whether sucrose or fructose consumed in
lower amounts (representing current average added sugar
consumption in the United States or Switzerland) behave
differently regarding lipid metabolism (ie, fatty acid synthesis and oxidation).
The aim of the present study, therefore, was to assess
the differential effect of moderate amounts of various sugars (including fructose, sucrose, and glucose) in SSBs on
fatty acid synthesis and degradation in healthy young men
by analysis of plasma fatty acid profiles and acylcarnitine
concentrations.
jcem.endojournals.org
2165
Subjects and Methods
Study design
The use of plasma fatty acid analysis to quantify de novo
lipogenesis has been validated previously (25). Newly synthesized fatty acids increase the fraction of palmitate (16:0) (the
primary product of fatty acid synthase) and dilute the amount of
linoleic acid (18:2), an essential fatty acid. In addition, concentrations of monounsaturated fatty acid (MUFA) and polyunsaturated fatty acid (PUFA) species were quantified to estimate the
process of fatty acid desaturation. ␤-Oxidation was assessed by
measuring concentrations of blood acylcarnitines in the fasting
state. In addition, acylcarnitines were measured during a euglycemic hyperinsulinemic clamp to assess the dynamics of acylcarnitine concentrations after inhibition of lipolysis by insulin.
Total plasma fatty acid and acylcarnitine profiles were analyzed with plasma and blood samples collected from a randomized controlled crossover trial, designed to study glucose and
lipid metabolism after dietary interventions with different sugars
in SSBs (10, 11). The study is registered and was approved by the
Ethics Committee of the University Hospital Zurich. The study
consisted of two parts, as described previously (10, 11). Four
different dietary interventions were performed in random sequence in both parts of the study. During each dietary intervention, subjects were supplied with SSBs containing various sugars
in different concentrations in random order during 3 weeks: 40 g
fructose per day [medium fructose (MF)], 80 g fructose per day
[high fructose (HF)], 80 g glucose per day [high glucose (HG)],
and 80 g sucrose per day [high sucrose (HS)], as described previously (10, 11). In the second part of the study, in addition,
hyperinsulinemic euglycemic clamps were performed with nine
participants after each intervention to assess glucose metabolism
and the dynamics of acylcarnitines, as described in reference (10)
and in the Supplemental Material.
Subjects
For the analysis of fasting plasma fatty acid profiles, 34
healthy, normal-weight male volunteers from both parts of the
study who completed all four dietary interventions were included. Acylcarnitine profiles were analyzed in the nine participants undergoing euglycemic hyperinsulinemic clamps.
Volunteers were eligible for the study if they were male, had
a normal body mass index (BMI; 19 –25 kg/m2), were healthy,
and were 20 –50 years old. Volunteers taking regular medication
or consuming SSBs with a total content of more than 60 g of
carbohydrates per day were not included in the study. Written
informed consent was obtained from all subjects before entering
the study.
Dietary assessment
Dietary assessment was performed as described previously
(10, 11). In the week prior to each examination as well as before
the start of the first intervention, all subjects filled in a 3-day (2
weekdays and 1 weekend day) weighed food record (26). A nutrition software system [EBISpro for Windows 8.0 (Swiss version), Dr J. Erhardt, University of Hohenheim, Hohenheim, Germany] was used to convert the amount of food eaten into
individual nutrients. Three-day energy and nutrient intakes were
averaged to obtain a mean daily energy and nutrient intake for
each subject.
2166
Hochuli et al
Sugar-Sweetened Beverages and Fatty Acid Metabolism
J Clin Endocrinol Metab, June 2014, 99(6):2164 –2172
Analysis of total plasma fatty acid profiles
Statistical analysis
Fasting blood samples (EDTA tubes) were collected at baseline and after each dietary intervention, between 7:00 and 9:30
AM after a 12-hour overnight fast. Samples were centrifuged at
3000 rpm for 10 minutes, and the plasma was stored at ⫺80°C
prior to sample preparation. For one participant, plasma fatty
acid profiles after the MF intervention were not analyzed because
the corresponding plasma sample could not be retrieved.
Statistical analysis was performed using the statistical package SPSS 21 (IBM SPSS Statistics). All variables were checked for
normal distribution prior to data analysis. Data are presented as
arithmetic mean ⫾ SD. Nonnormally distributed data were log
transformed, and statistical analysis was carried out with the
transformed data. The effect of the interventions on anthropometric and metabolic parameters, individual plasma fatty acids
and acylcarnitines was examined using multiple linear regression
(28), always controlling for between-patient differences. Post
hoc sequential Bonferroni correction was applied to account for
multiple comparisons. For comparison of fasting acylcarnitine
concentrations to concentrations after the euglycemic hyperinsulinemic clamp, a paired samples t test was applied. A value of
P ⬍ .05 after sequential Bonferroni correction was considered
significant.
Production of fatty acid methyl esters
One hundred twenty-five microliters of plasma were diluted
with ultrapure water to a total volume of 250 ␮L. One milliliter
of methanol-methylene chloride [3:1 (vol/vol)] was added together with 20 ␮g deuterated C17:0 (Cambridge Isotope Laboratories, Inc) and 0.8 ␮g deuterated C22:0 (Dr Ehrenstorfer
GmbH, Augsburg, Germany) as internal standards. For the preparation of fatty acid methyl esters, 200 ␮L acetyl chloride was
added to each sample and incubated at 75°C for 1 hour. The
reaction was neutralized by the addition of 4 mL 7% K2CO3, and
fatty acid methyl esters were extracted with 2 mL hexane (27).
Subsequent to centrifugation at 2500 rpm for 20 minutes, 1.6 mL
of the hexane layer was dried under nitrogen and redissolved in
280 ␮L heptane.
Gas chromatography-mass spectrometry of fatty
acid methyl esters
One-microliter sample was injected into a Finnigan PolarisQ
ion trap gas chromatography-mass spectrometry system
(Thermo Quest) with an injector temperature of 280°C. Fatty
acid methyl esters were separated on a 30-m BGB-1701 column.
The initial oven temperature of 50°C was held for 8 minutes and
then increased to a final temperature of 280°C with a temperature gradient of 5°C/min. The ion source temperature and transfer line temperature was 230°C and 300°C, respectively. Analytes were detected as positive ions in full-scan mode. The specific
mass traces were extracted for the quantification of each analyte.
The fatty acids were identified by comparison of the retention
time and the mass spectrum with authentic standards.
Analysis of acylcarnitine profiles in whole-blood
spots
Dried blood spots on filter paper were collected at baseline
and after each dietary intervention in the fasting state and after
the hyperinsulinemic euglycemic clamp (including a 3 h tracer
equilibration period using [6, 6]2H2 glucose and a 2 h clamp
period with insulin) (see Supplemental Material).
Dry filter paper cards were sealed into plastic bags and stored
at ⫺20°C until analysis. Dried blood spots were extracted with
methanol, the extract dried under nitrogen, and redissolved in
acetonitrile/water (50:50) for analysis according to the diagnostic standard procedure of the Division of Clinical Chemistry and
Biochemistry of the University Children’s Hospital Zurich. Acylcarnitines were quantified using eight isotopically labeled internal standards (Cambridge Isotopes Laboratories). Precursor ions
of a mass to charge ratio of 85 in the mass range mass to charge
ratio 150 – 450 were acquired on a SCIEX QTrap 4000 LC-ESIMS/MS instrument (AB Sciex Switzerland GmbH). Routine
chemistry analysis of blood samples were performed as described
(10, 11).
Results
Baseline characteristics of the participants and important
anthropometric and biochemical data after each of the
interventions are summarized in Table 1. All participants
were normal weight (mean BMI 22.4 ⫾ 1.79 kg/m2).
There were no significant changes of cholesterol (total,
high-density lipoprotein, and LDL), triglyceride, and total
free fatty acid concentrations after the dietary interventions compared to baseline (Table 1). Fasting glucose was
slightly but significantly increased after all interventions
compared with baseline (Table 1), as discussed in an earlier publication (11). Mean total energy intake was
2446 ⫾ 554 kcal and was not significantly different between the four interventions and baseline (Table 2). The
proportion of energy intake from carbohydrates was
higher after each intervention (51%–56%) compared
with baseline (48%), and the proportion of energy intake
from fat (31%–32%) was lower after each intervention
(except for MF) compared with baseline (37%) (P ⬍ .05)
(Table 2). The consumption of the individual sugars varied
according to the intervention. Total fructose intake was
highest in the HF intervention, and the amount of fructose
consumed in HS intervention corresponded to the amount
in MF. The amount of fructose consumed in HG corresponded to baseline. The energy (kilocalories) provided by
added sugars in the SSBs was 13% of the mean total energy
intake for HG, HF, and HS and 6.7% for MF. Intake of
saturated fatty acids, monounsaturated (MUFAs), and
PUFAs was not significantly different between baseline
and the different dietary interventions (data not shown).
Fasting total plasma fatty acid profiles
The relative abundance of palmitic acid (16:0), the primary fatty acid synthesized in de novo lipogenesis, was
increased after HF [22.97% ⫾ 5.51% (percentage of total
fatty acids by weight)] and MF (26.1% ⫾ 1.7%) but equal
doi: 10.1210/jc.2013-3856
Table 1.
jcem.endojournals.org
2167
Baseline Characteristics of Participants
n
Anthropometrics
Weight, kga
Height, ma
BMI, kg/m2a
Waist circumference, cma
Body fat, %a
Body fat, kga
Laboratory parameters
Fasting glucose, mmol/La
Total plasma cholesterol, mmol/La
HDL-C, mmol/Lc
LDL-C, mmol/Lc
Triglycerides, mmol/Lc
Free fatty acids, ␮mol/La
Baseline
HG
MF
HF
HS
34
34
33
34
34
73.3 ⫾ 8.6
1.81 ⫾ 0.07
22.3 ⫾ 1.82
82.6 ⫾ 5.94
15.4 ⫾ 3.32
11.5 ⫾ 3.45
73.9 ⫾ 8.8b
1.81 ⫾ 0.07
22.5 ⫾ 1.85b
82.7 ⫾ 5.76
15.4 ⫾ 3.32
11.6 ⫾ 3.58
73.4 ⫾ 9.0
1.81 ⫾ 0.07
22.4 ⫾ 1.91
82.7 ⫾ 5.72
15.5 ⫾ 3.06
11.6 ⫾ 3.34
73.6 ⫾ 8.5
1.81 ⫾ 0.07
22.4 ⫾ 1.78
82.6 ⫾ 5.83
15.7 ⫾ 3.00
11.7 ⫾ 3.19
73.8 ⫾ 8.4
1.81 ⫾ 0.07
22.5 ⫾ 1.71
83.0 ⫾ 5.67
15.5 ⫾ 3.05
11.6 ⫾ 3.22
4.39 ⫾ 0.41
3.98 ⫾ 0.85
1.30 ⫾ 0.30
2.17 ⫾ 0.68
0.76 ⫾ 0.31
380.4 ⫾ 163.1
4.56 ⫾ 0.38b
4.07 ⫾ 0.80
1.29 ⫾ 0.34
2.23 ⫾ 0.70
0.82 ⫾ 0.36
377.0 ⫾ 195.7
4.60 ⫾ 0.38b
4.18 ⫾ 0.70
1.36 ⫾ 0.32
2.27 ⫾ 0.63
0.85 ⫾ 0.49
407.1 ⫾ 211.0
4.55 ⫾ 0.48b
4.16 ⫾ 0.90
1.33 ⫾ 0.37
2.23 ⫾ 0.77
0.85 ⫾ 0.43
370.5 ⫾ 180.1
4.63 ⫾ 0.38b
4.19 ⫾ 0.88
1.33 ⫾ 0.31
2.33 ⫾ 0.69
0.83 ⫾ 0.56
396.7 ⫾ 207.4
Abbreviation: HDL-C, high-density lipoprotein cholesterol. HG was 80 g of glucose per day; MF was 40 g of fructose per day; HF was 80 g of
fructose per day; HS was 80 g of sucrose per day.
a
Arithmetic mean ⫾ SD.
b
Significantly different compared with baseline (P ⬍ .05, multiple linear regression with Bonferroni correction for four comparisons).
c
Geometric mean ⫾ SD.
to baseline (19.40% ⫾ 2.79%) after HS (19.40% ⫾
2.91%) and HG (19.43% ⫾ 3.12%) (P ⬍ .001 HF or MF
compared with baseline, HS, or HG) (Figure 1A and Supplemental Table 1), indicating a higher proportion of
newly synthesized fatty acids after HF and MF. The relative abundance of linoleic acid (18:2), an essential fatty
acid, was lower after HF and MF (P ⬍ .001 compared with
baseline) but was not significantly different from the baseline after HS and HG (Figure 1B), most probably due to the
dilution of the plasma fatty acid pool by de novo synthesized fatty acids (eg, palmitate) after HF and MF (25). The
molar ratio of palmitic acid to linoleic acid (16:0 to 18:2),
a marker for de novo lipogenesis, was increased after MF
(1.13 ⫾ 0.14) and HF (1.00 ⫾ 0.75) when compared with
baseline (0.71 ⫾ 0.17, P ⬍ .001) but not after HS and HG
(Figure 1C).
Levels of monounsaturated fatty acids were lower after
diets containing fructose. Both the relative abundances of
palmitoleic acid (16:1) and oleic acid (18:1) (primarily
synthesized via ⌬9-desaturase) were decreased after HF
and MF compared with baseline, HS, or HG (P ⬍ .001).
Similarly, the molar ratios palmitoleic acid to palmitic acid
(16:1 to 16:0) and oleic acid to stearic acid (18:1 to 18:0)
were decreased after the interventions with SSBs containing fructose (HF, MF) (P ⬍ .001 compared with baseline,
HS, or HG) but remained unchanged from baseline after
the interventions with SSBs supplemented with sucrose or
glucose (HS and HG) (Figure 2) (Supplemental Table 1).
Table 2. Dietary Intake Per Day (Mean ⫾ SD) of All Subjects at Baseline and After Each of the Four 3-Week
Interventions
N
Energy, kcal/d
Carbohydrates, %
Protein, %
Fat, %
Free fructose, g/d
Total fructose, g/d
Free glucose, g/d
Total glucose, g/d
Sucrose, g/d
Baseline
HG
MF
HF
HS
34
2340 ⫾ 508
48 ⫾ 7
16 ⫾ 3
37 ⫾ 6
16.3 ⫾ 8.6
47.3 ⫾ 21.7
14.5 ⫾ 6.5
45.5 ⫾ 19.7
61.9 ⫾ 32.1
34
2457 ⫾ 536
56 ⫾ 5a
13 ⫾ 2a
31 ⫾ 4a
14.9 ⫾ 9.4
42.6 ⫾ 19.9
92.2 ⫾ 5.7a
119.9 ⫾ 15.2a
55.3 ⫾ 27.2
33
2374 ⫾ 631
51 ⫾ 6a
14 ⫾ 3a
35 ⫾ 6
54.4 ⫾ 9.5a
82.8 ⫾ 23.9a
11.6 ⫾ 6.3
40.0 ⫾ 22.1
56.8 ⫾ 36.0
34
2444 ⫾ 506
55 ⫾ 5a
13 ⫾ 2a
31 ⫾ 5a
90.9 ⫾ 8.1a
115.2 ⫾ 17.2a
10.36 ⫾ 6.0
34.9 ⫾ 15.2a
48.6 ⫾ 23.1a
34
2507 ⫾ 559
54 ⫾ 6a
14 ⫾ 3a
32 ⫾ 6a
13.0 ⫾ 8.8
77.0 ⫾ 20.3a
13.1 ⫾ 12.4
76.7 ⫾ 20.3a
126.2 ⫾ 27.1a
Data are arithmetic mean ⫾ SD (all parameters). HG was 80 g of glucose per day; MF was 40 g of fructose per day; HF was 80 g of fructose per
day; HS was 80 g of sucrose per day. Free fructose and free glucose refer to fructose and glucose that is contained in the food as monosaccharide,
whereas total fructose and total glucose refer to both the monosaccharides and the part derived from the disaccharide sucrose (50% fructose and
50% glucose).
a
Significantly different compared with baseline (P ⬍ .05, multiple linear regression with Bonferroni correction for four comparisons).
Hochuli et al
A
20.00
10.00
0.00
B
HG
HS
Linoleic acid (18:2) [%]
B
HF
MF
‡
‡
30.00
20.00
10.00
0.00
Palmitic acid (16:0) / Linoleic acid (18:2)
*
Palmitoleic acid (16:1) [%]
*
2.00
B
HG
HS
C
HF
MF
*
*
1.50
1.00
0.50
0.00
J Clin Endocrinol Metab, June 2014, 99(6):2164 –2172
B
HG
HS
HF
MF
Diet
Figure 1. Relative abundance of palmitic acid (16:0) and linoleic acid
(18:2) in plasma at baseline and after the dietary interventions. Palmitic
acid (16:0) (A) and linoleic acid (18:2) (B) in percentage of total fatty
acids (by weight) are shown. C, Molar ratio of palmitic to linoleic acid
(16:0 to 18:2), calculated from the concentrations of (16:0) and (18:2)
in micromoles per liter. The relative abundance of palmitate is
increased after SSBs containing fructose (HF and MF) but not sucrose
(HS) or glucose (HG). The horizontal line indicates the baseline level.
Dietary interventions: HG, glucose, 80 g/d; HS, sucrose, 80 g/d; HF,
fructose, 80 g/d; MF, fructose, 40 g/d. B denotes the measurement at
baseline before the interventions. *, Significant differences compared
with baseline, HS, or HG, with P ⱕ .002; ‡, significant differences
compared with baseline (P ⱕ .007).
In contrast, unsaturated fatty acids synthesized via ⌬6desaturase were increased after the HF intervention. The
relative abundance of ␥-linolenic acid (18:3) and the molar
ratio of ␥-linolenic (18:3) to linoleic acid (18:2) were increased after HF (but not MF) when compared with baseline, HS, or HG (P ⬍ .001), whereas the levels after HG or
HS remained unchanged compared with baseline levels
(Figure 3A and Supplemental Table 1).
A
3.00
*
*
HF
MF
2.00
1.00
0.00
Palmitoleic (16:1) / palmitic acid (16:0)
Palmitic acid (16:0) [%]
30.00
Sugar-Sweetened Beverages and Fatty Acid Metabolism
Oleic (18:1) / stearic acid (18:0)
2168
B
HG
HS
0.20
B
0.15
*
*
HF
MF
0.10
0.05
0.00
B
HG
HS
C
3.00
*
*
HF
MF
2.00
1.00
0.00
B
HG
HS
Diet
Figure 2. Relative abundance of monounsaturated fatty acids
synthesized via ⌬9-desaturase in plasma. A, Palmitoleic acid (16:1) in
percentage of total fatty acids (by weight). B, Ratio of palmitoleic to
palmitic acid (16:1 to 16:0). C, Ratio of oleic to stearic acid (18:1 to
18:0). Molar ratios are calculated from the concentrations of the
respective fatty acids in micromoles per liter. The horizontal line
indicates the baseline level. The relative abundance of
monounsaturated fatty acids synthesized via ⌬9-desaturase is
decreased after diets containing fructose (HF and MF). For
abbreviations of dietary interventions, see Figure 1. *, Significant
differences compared with baseline, HS, or HG, with P ⬍ .001.
A different pattern was observed for eicosapentaenoic
acid (20:5), a fatty acid produced via ⌬5-desaturase. Levels of 20:5 depended on the amount but not the type of
sugar in the SSBs: the ratio 20:5 to 20:4 was increased after
all SSBs containing 80 g of added sugar (HG, HS, HF)
compared with baseline (P ⬍ .016) (Figure 3B and Supplemental Table 1).
jcem.endojournals.org
*
0.03
‡
2.0
‡
A
1.6
0.02
0.01
1.2
0.8
0.4
0.0
0.00
HF
B
HG
HS
HF
‡
‡
‡
MF
HG
HS
MF
Baseline
Long chain acylcarnitines (C16)
Fasting
Clamp
B
[μmol/l]
0.30
20:5 / 20:4
2169
Palmitoylcarnitine (C16)
A
[μmol/l]
γ-Linolenic (18:3) / Linoleic acid (18:2)
doi: 10.1210/jc.2013-3856
0.20
B
*
1.6
*
1.2
0.8
0.4
0.0
HF
0.10
HG
HS
MF
Baseline
Short and medium chain acylcarnitines (C6 - C14)
Fasting
Clamp
1.2
B
HG
HS
HF
MF
Diet
Figure 3. Fatty acids synthesized via ⌬6-desaturase (␥-linolenic acid)
and ⌬5-desaturase (eicosapentaenoic acid) in plasma. A, Ratio ␥linolenic to linoleic acid (18:3 to 18:2). B, Ratio eicosapentaenoic to
eicosatetraenoic (arachidonic) acid (20:5 to 20:4). Molar ratios are
calculated from the concentrations of the respective fatty acids in
micromoles per liter. The horizontal line indicates the baseline level.
The relative abundance of ␥-linolenic acid (18:3) is increased only after
HF. Eicosapentaenoic acid (20:5) is increased in a dose-dependent
manner after all interventions with 80 g added sugar but not after MF
or at baseline. For abbreviations of dietary interventions, see Figure 1.
*, Significant differences compared with baseline, HS, or HG with P ⬍
.001; ‡, significant differences compared with baseline (P ⱕ .016).
Acylcarnitine profiles
Fasting long-chain acylcarnitines (ie, intermediates
prior to ␤-oxidation) were increased after HF and HS,
compared with HG (eg, C16 palmitoylcarnitine: HF 1.36
and HS 1.34 vs HG 0.96 ␮mol/L, P ⫽ .005 for comparison
between these interventions), whereas C16 acylcarnitine
concentrations were not significantly different from baseline after HG (Figure 4A). During the euglycemic hyperinsulinemic clamp, long-chain acylcarnitine concentrations (eg, C16 palmitoylcarnitine) decreased compared
with fasting concentrations (HS, MF, P ⬍ .05, HF, P ⫽
NS), but virtually no change in palmitoylcarnitine levels
could be observed for HG and baseline, in which the fasting concentrations of palmitoylcarnitine were also lower
(Figure 4B and Supplemental Figure 1A).
Levels of short- and medium-chain acylcarnitines decreased during the clamp after all interventions (P ⬍ .03)
(Figure 4C and Supplemental Figure 1B).
[μmol/l]
0.00
C
0.8
*
*
*
*
*
0.4
0.0
HF
HG
HS
MF
Baseline
Figure 4. Acylcarnitine concentrations, under fasting conditions and
during a euglycemic hyperinsulinemic clamp. A, Fasting levels of
palmitoylcarnitine (long chain; C16) are significantly higher after the
HF and HS interventions compared with HG. B, During the clamp,
palmitoylcarnitine levels decrease with insulin (HS and MF, P ⬍ .05; HF,
P ⫽ NS), but no change is seen for HG and baseline. C, In contrast to
long-chain acylcarnitines, levels of short- and medium-even-chain
acylcarnitines (C6 to C14) decrease to the same level during the clamp
after all interventions and at baseline. ‡, Significant differences of HS
and HF compared with HG (P ⫽ .005); *, significant differences of
acylcarnitine concentrations after the euglycemic clamp compared with
the respective fasting concentrations (P ⬍ .05).
Discussion
To the best of our knowledge, this is the first study demonstrating that even amounts of free fructose that are
lower than the average daily consumption in the United
States, consumed daily during 3 weeks added to SSBs, have
distinct effects on fatty acid synthesis in healthy young
men, when compared with sweetened beverages containing sucrose, or glucose.
The relative abundance of palmitate (16:0) is significantly increased, and the fraction of the essential fatty acid
linoleate (18:2) is decreased after interventions with SSBs
containing added fructose (HF and MF), in contrast to the
interventions with SSBs containing sucrose (HS) or glucose (HG), indicating enhanced fatty acid synthesis (25)
2170
Hochuli et al
Sugar-Sweetened Beverages and Fatty Acid Metabolism
only after diets containing added fructose. It is well known
that ingestion of fructose in high amounts enhances de
novo lipogenesis (3, 15, 16). The present data expand this
knowledge and reveal that increased fatty acid synthesis,
even occurs after consumption of moderate amounts of
fructose (MF) in SSB, corresponding to 6.7% of total energy intake. In addition, the present data show that this
effect is completely abolished when sucrose instead of
fructose is added to the SSBs (HS), providing the same
amount of fructose (40 g) from SSBs as in the MF intervention. This demonstrates that sucrose, if consumed in
moderate amounts, has diverging effects on lipogenesis
when compared with SSBs containing only fructose. Most
likely this is explained by a preferential lipogenic effect of
fructose; the precise mechanism behind this finding, however, remains to be further elucidated.
The effect of sucrose on fatty acid synthesis is probably
dose related, and similar effects to fructose may possibly
be observed when higher amounts of sucrose are ingested.
Differences in the intestinal absorption kinetics of fructose
vs sucrose may also play a role (29, 30). In a study using
higher amounts of sugar (25% of energy) than in the present investigation, postprandial triglyceride profiles were
comparably elevated after providing both fructose or sucrose (or high fructose corn syrup) in sweetened beverages
(31). Postprandial hepatic de novo lipogenesis specifically
was increased during 10 weeks of consumption of fructose
in humans but not during consumption of glucose (3).
However, the quantitative effect of fructose consumption
on fatty acid synthesis compared with other pathways of
fructose disposal is debated (32). We speculate that the
effects of fructose on plasma palmitate and linoleate in the
present study will primarily result from enhanced postprandial de novo lipogenesis, as supported by previous
data with isotope tracers (3). It has been shown previously
that distinct fatty acid patterns reflecting enhanced de
novo lipogenesis persist over at least 24 hours (25); therefore, the measured plasma fatty acid composition will represent an integrated result over postprandial and fasting
states.
Although SSBs containing only fructose increased
palmitate as a marker of de novo fatty acid synthesis, fructose did not stimulate the production of monounsaturated
fatty acids derived from ⌬9-desaturase, ie, palmitoleic acid
(16:1) and oleic acid (18:1). Stearoyl-CoA desaturase
(SCD) is the rate-limiting enzyme in the biosynthesis of
monounsaturated fatty acids (16:1, 18:1). Numerous previous studies in humans and animals have shown that ⌬9desaturase activity is up-regulated after high-carbohydrate diets, eg, via the lipogenic transcription factors sterol
regulatory element-binding protein-1 (SREBP-1) and carbohydrate-response element-binding protein (ChREBP)
J Clin Endocrinol Metab, June 2014, 99(6):2164 –2172
and also by insulin (33–36). Contrary to our expectations,
the relative abundances of palmitoleic acid (16:1) and
oleic acid (18:1) were decreased after MF and HF. It can
be speculated that de novo synthesis of palmitic acid (16:0)
or stearic acid (18:0) is up-regulated to a higher extent as
compared with SCD activity after diets containing smaller
amounts of fructose. This may be due to the fact that the
metabolism of fructose with production of precursors for
de novo lipogenesis (ie, 16:0) is not primarily regulated by
insulin, whereas the regulation of SCD activity also depends on insulin and may be more strongly up-regulated
when SSB also contain glucose. These results also indicate
that the use of palmitoleic acid (16:1) fatty acid ratios as
a surrogate marker of de novo lipogenesis may underestimate the true rate under these conditions.
A different pattern was observed for fatty acids synthesized via ⌬6- and ⌬5-desaturase. The regulation of ⌬6and ⌬5-desaturase activity depends on multiple factors
(22, 37). HF clearly increases fatty acids synthesized via
⌬6-desaturase (eg, ␥-linolenic acid), whereas this is not
observed for MF, HG, and HS (Figure 3). In contrast,
␦5-desaturase appears to be induced by all types of sugars
studied (HF, HS, and HG). The exact mechanism behind
this finding, however, remains to be elucidated further.
Our data show that fructose and sucrose added to SSBs
may also affect fatty acid oxidation. Long-chain fatty acids are esterified to acylcarnitines for transport across the
inner mitochondrial membrane prior to ␤-oxidation.
Thus, acylcarnitine levels will depend on fluxes of free
fatty acids from lipolysis (intrahepatic and from adipose
tissue) and the rate of ␤-oxidation. Whereas increased
palmitate synthesis is observed after ingestion of fructose
but not sucrose, long-chain acylcarnitine levels (eg, palmitoylcarnitine) are higher after both fructose and sucrose,
suggesting a limited clearance of long chain acylcarnitines
by fatty acid oxidation. During the euglycemic hyperinsulinemic clamp, palmitoylcarnitine decreased after the
HF (not significant) and significantly after HS and MF
diets, most probably by inhibition of lipolysis and decreased flux toward ␤-oxidation, whereas palmitoylcarnitine remained unaltered at baseline and after HG. Thus,
the glucose load with HG may not be sufficient to change
␤-oxidation rate enough to result in a visible effect on
long-chain acylcarnitines during the clamp. Short- and
medium-chain acylcarnitines, arising from intermediates
within ␤-oxidation, decrease to the same level during the
clamp after all interventions and at baseline, demonstrating a net global reduction of fluxes through ␤-oxidation
during the hyperinsulinemic clamp as expected.
So far, only limited information is available about the
effect of added fructose or sucrose on fatty acid oxidation.
But the evidence suggests that ingestion of fructose may
doi: 10.1210/jc.2013-3856
reduce lipid oxidation (8, 9, 18, 20 –22, 38), which is in
line with our findings. Fructose-containing SSBs increase
de novo lipogenesis but may also interfere with ␤-oxidation through feedback mechanisms of product overflow
and possibly down-regulation of ␤-oxidation enzymes
(eg, via peroxisome proliferator-activated receptor ␣
[PPAR␣]) (22, 39).
This study has limitations. First, we measured total
plasma fatty acids. Analysis of fatty acids in different lipoproteins, eg, VLDL particles, would probably have provided an even better qualitative estimate of fatty acid synthesis by the liver. However, the analysis of total plasma
fatty acid profiles demonstrated consistent and biologically plausible results, which can be explained within the
established framework of lipid biochemistry. Furthermore, given the fact that plasma lipoprotein concentrations were similar after all interventions, we expect negligible effects on plasma fatty acid composition due to
changes of plasma lipoproteins (40). Second, we cannot
completely exclude that changes in dietary fatty acid intake may have influenced plasma fatty acid composition
(40). However, the intake of fats (in percentage of total
energy intake) was similar across all interventions, and
there were no significant differences in the calculated intake of fatty acids, MUFAs, and PUFAs between the interventions. The observed changes of the fatty acid composition were almost exclusively confined to diets with
SSBs containing free fructose, which makes it unlikely that
these effects would be explained by changes in dietary
habits beyond intake of SSBs. There was a trend toward a
slightly increased body weight after all interventions compared with baseline but only statistically significant for the
HG intervention. However, the absolute differences are
very small (100 – 600 g) and are barely relevant to this
study.
Taken together, the present study clearly demonstrates
that even when consumed in moderate amounts, SSBs containing fructose but not sucrose stimulate fatty acid synthesis (palmitate), whereas both fructose and sucrose increase long-chain acylcarnitine levels. Further studies
using specific isotopic tracers are needed to quantify the
effects of SSBs on fatty acid metabolism. The effects of
varying amounts of fructose, sucrose, and glucose in diet
on insulin resistance, obesity, and cardiovascular disease
will have to be evaluated in long-term prospective studies.
Acknowledgments
We thank Valeria Meyer and Cornelia Zwimpfer (University
Hospital Zurich, Zurich, Switzerland) for the expert laboratory
and technical work. We also thank all the subjects participating
jcem.endojournals.org
2171
in this study and the Nestlé Product and Technology Centre
(Konolfingen, Switzerland) for providing the study drinks.
Contributions of the authors include the following: M.Ho.
and K.B. designed the research; M.Ho., I.A., M.He., A.W.,
P.A.G., H.T., and K.B. conducted the research; M.Ho., I.A.,
A.W., P.A.G., and K.B. analyzed the data; M.Ho., I.A., M.He.,
A.W., G.A.S., and K.B. wrote the paper; and K.B. had primary
responsibility for the final content.
The Nestlé Product and Technology Centre (Konolfingen,
Switzerland) provided the sugar-sweetened beverages for the
study.
The study had a trial registration number of NCT01021969
(www.clinicaltrials.gov).
Address all correspondence and requests for reprints to: Michel Hochuli, MD, PhD, University Hospital Zurich, Division of
Endocrinology, Diabetes, and Clinical Nutrition, 8091 Zurich,
Switzerland. E-mail: [email protected].
The study was supported by Swiss National Science Foundation Grant 32003B-119706 (to K.B.) and the Vontobel Foundation (to G.A.S.).
Disclosure Summary: The authors have nothing to disclose.
References
1. Marriott BP, Cole N, Lee E. National estimates of dietary fructose
intake increased from 1977 to 2004 in the United States. J Nutr.
2009;139:1228S–1235S.
2. Welsh JA, Sharma A, Abramson JL, Vaccarino V, Gillespie C, Vos
MB. Caloric sweetener consumption and dyslipidemia among US
adults. JAMA. 2010;303:1490 –1497.
3. Stanhope KL, Schwarz JM, Dyachenko A, et al. Consuming fructose-sweetened, not glucose-sweetened, beverages increases visceral
adiposity and lipids and decreases insulin sensitivity in overweight/
obese humans. J Clin Invest. 2009;119:1322–1334.
4. Aeberli I, Zimmermann MB, Molinari L, et al. Fructose intake is a
predictor of LDL particle size in overweight schoolchildren. Am J
Clin Nutr. 2007;86:1174 –1178.
5. Schulze MB, Hoffmann K, Manson JE, et al. Dietary pattern, inflammation, and incidence of type 2 diabetes in women. Am J Clin
Nutr. 2005;82:675– 684; quiz 714 –715.
6. Sorensen LB, Raben A, Stender S, Astrup A. Effect of sucrose on
inflammatory markers in overweight humans. Am J Clin Nutr.
2005;82:421– 427.
7. de Koning L, Malik VS, Rimm EB, Willett WC, Hu FB. Sugarsweetened and artificially sweetened beverage consumption and risk
of type 2 diabetes in men. Am J Clin Nutr. 2011;93:1321–1327.
8. Le KA, Ith M, Kreis R, et al. Fructose overconsumption causes dyslipidemia and ectopic lipid deposition in healthy subjects with and
without a family history of type 2 diabetes. Am J Clin Nutr. 2009;
89:1760 –1765.
9. Le KA, Faeh D, Stettler R, et al. A 4-wk high-fructose diet alters lipid
metabolism without affecting insulin sensitivity or ectopic lipids in
healthy humans. Am J Clin Nutr. 2006;84:1374 –1379.
10. Aeberli I, Hochuli M, Gerber PA, et al. Moderate amounts of fructose consumption impair insulin sensitivity in healthy young men: a
randomized controlled trial. Diabetes Care. 2013;36:150 –156.
11. Aeberli I, Gerber PA, Hochuli M, et al. Low to moderate sugarsweetened beverage consumption impairs glucose and lipid metabolism and promotes inflammation in healthy young men: a randomized controlled trial. Am J Clin Nutr. 2011;94:479 – 485.
12. Dentin R, Girard J, Postic C. Carbohydrate responsive element bind-
2172
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
Hochuli et al
Sugar-Sweetened Beverages and Fatty Acid Metabolism
ing protein (ChREBP) and sterol regulatory element binding protein-1c (SREBP-1c): two key regulators of glucose metabolism and
lipid synthesis in liver. Biochimie. 2005;87:81– 86.
Koo HY, Miyashita M, Cho BH, Nakamura MT. Replacing dietary
glucose with fructose increases ChREBP activity and SREBP-1 protein in rat liver nucleus. Biochem Biophys Res Commun. 2009;390:
285–289.
Koo HY, Wallig MA, Chung BH, Nara TY, Cho BH, Nakamura
MT. Dietary fructose induces a wide range of genes with distinct
shift in carbohydrate and lipid metabolism in fed and fasted rat liver.
Biochim Biophys Acta. 2008;1782:341–348.
Elliott SS, Keim NL, Stern JS, Teff K, Havel PJ. Fructose, weight
gain, and the insulin resistance syndrome. Am J Clin Nutr. 2002;
76:911–922.
Mayes PA. Intermediary metabolism of fructose. Am J Clin Nutr.
1993;58:754S–765S.
Dekker MJ, Su Q, Baker C, Rutledge AC, Adeli K. Fructose: a highly
lipogenic nutrient implicated in insulin resistance, hepatic steatosis,
and the metabolic syndrome. Am J Physiol Endocrinol Metab. 2010;
299:E685–E694.
Chong MF, Fielding BA, Frayn KN. Mechanisms for the acute effect
of fructose on postprandial lipemia. Am J Clin Nutr. 2007;85:1511–
1520.
Parks EJ. Dietary carbohydrate’s effects on lipogenesis and the relationship of lipogenesis to blood insulin and glucose concentrations. Br J Nutr. 2002;87(suppl 2):S247–S253.
Abdel-Sayed A, Binnert C, Le KA, Bortolotti M, Schneiter P, Tappy
L. A high-fructose diet impairs basal and stress-mediated lipid metabolism in healthy male subjects. Br J Nutr. 2008;100:393–399.
Cox CL, Stanhope KL, Schwarz JM, et al. Consumption of fructosesweetened beverages for 10 weeks reduces net fat oxidation and
energy expenditure in overweight/obese men and women. Eur J Clin
Nutr. 2012;66:201–208.
Roglans N, Vila L, Farre M, et al. Impairment of hepatic Stat-3
activation and reduction of PPAR␣ activity in fructose-fed rats.
Hepatology. 2007;45:778 –788.
Stanhope KL, Havel PJ. Endocrine and metabolic effects of consuming beverages sweetened with fructose, glucose, sucrose, or
high-fructose corn syrup. Am J Clin Nutr. 2008;88:1733S–1737S.
Schaefer EJ, Gleason JA, Dansinger ML. Dietary fructose and glucose differentially affect lipid and glucose homeostasis. J Nutr. 2009;
139:1257S–1262S.
Hudgins LC, Hellerstein M, Seidman C, Neese R, Diakun J, Hirsch
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
J Clin Endocrinol Metab, June 2014, 99(6):2164 –2172
J. Human fatty acid synthesis is stimulated by a eucaloric low fat,
high carbohydrate diet. J Clin Invest. 1996;97:2081–2091.
Gibson RS. Principles of Nutritional Assessment. 2nd ed. New
York: Oxford University Press; 2005.
Moser AB, Jones DS, Raymond GV, Moser HW. Plasma and red
blood cell fatty acids in peroxisomal disorders. Neurochem Res.
1999;24:187–197.
Senn S. Crossover Trials in Clinical Research. 2nd ed. Chichester,
UK: Wiley Publishing; 2002.
Le MT, Frye RF, Rivard CJ, et al. Effects of high-fructose corn syrup
and sucrose on the pharmacokinetics of fructose and acute metabolic and hemodynamic responses in healthy subjects. Metabolism.
2012;61:641– 651.
Rumessen JJ, Gudmand-Hoyer E. Absorption capacity of fructose in
healthy adults. Comparison with sucrose and its constituent monosaccharides. Gut. 1986;27:1161–1168.
Stanhope KL, Griffen SC, Bair BR, Swarbrick MM, Keim NL, Havel
PJ. Twenty-four-hour endocrine and metabolic profiles following
consumption of high-fructose corn syrup-, sucrose-, fructose-, and
glucose-sweetened beverages with meals. Am J Clin Nutr. 2008;87:
1194 –1203.
Sun SZ, Empie MW. Fructose metabolism in humans—what isotopic tracer studies tell us. Nutr Metab (Lond). 2012;9:89.
Attie AD. Relationship between stearoyl-CoA desaturase activity
and plasma triglycerides in human and mouse hypertriglyceridemia.
J Lipid Res. 2002;43:1899 –1907.
Chong MF, Hodson L, Bickerton AS, et al. Parallel activation of de
novo lipogenesis and stearoyl-CoA desaturase activity after 3 d of
high-carbohydrate feeding. Am J Clin Nutr. 2008;87:817– 823.
Ntambi J. Regulation of stearoyl-CoA desaturases and role in metabolism. Prog Lipid Res. 2004;43:91–104.
Flowers MT, Ntambi JM. Stearoyl-CoA desaturase and its relation
to high-carbohydrate diets and obesity. Biochim Biophys Acta.
2009;1791:85–91.
Li Y, Nara TY, Nakamura MT. Peroxisome proliferator-activated
receptor ␣ is required for feedback regulation of highly unsaturated
fatty acid synthesis. J Lipid Res. 2005;46:2432–2440.
Roberts R, Bickerton AS, Fielding BA, et al. Reduced oxidation of
dietary fat after a short term high-carbohydrate diet. Am J Clin Nutr.
2008;87:824 – 831.
Bartlett K, Eaton S. Mitochondrial ␤-oxidation. Eur J Biochem.
2004;271:462– 469.
Hodson L, Skeaff CM, Fielding BA. Fatty acid composition of adipose tissue and blood in humans and its use as a biomarker of
dietary intake. Prog Lipid Res. 2008;47:348 –380.