ARTICLE IN PRESS

G Model
ATH-11615; No. of Pages 6

Atherosclerosis xxx (2010) xxx–xxx

Contents lists available at ScienceDirect

Atherosclerosis
journal homepage: www.elsevier.com/locate/atherosclerosis

Resistance exercise but not aerobic exercise lowers remnant-like lipoprotein

particle cholesterol in type 2 diabetes: A randomized controlled trial夽
Claire Gavin a,b , Ronald J. Sigal b,c,d,e,f , Marion Cousins a,b , Michelle L. Menard a,b ,
Michelle Atkinson a,b , Farah Khandwala f , Glen P. Kenny c,d ,
Spencer Proctor g , Teik Chye Ooi a,b,∗ , on behalf of the Diabetes Aerobic and Resistance Exercise
(DARE) trial investigators
a
Clinical Research Laboratory, Division of Endocrinology and Metabolism, Faculty of Medicine, University of Ottawa, Ottawa, Canada
b
Department of Medicine, Faculty of Medicine, University of Ottawa, Ottawa, Canada
c
School of Human Kinetics, Faculty of Health Sciences University of Ottawa, Ottawa, Canada
d
Clinical Epidemiology Program, Ottawa Hospital Research Institute, Ottawa, Canada
e
Department of Medicine and Cardiac Sciences, Faculties of Medicine and Kinesiology, University of Calgary, Calgary, Alberta, Canada
f
Department of Community Health Sciences, Faculties of Medicine and Kinesiology, University of Calgary, Calgary, Alberta, Canada
g
Alberta Institute for Human Nutrition, University of Alberta, Edmonton, Alberta, Canada

a r t i c l e i n f o a b s t r a c t

Article history: The comparative effects of aerobic and resistance exercise on triglyceride-rich lipoproteins including
Received 12 April 2010 remnant lipoproteins are controversial. This study examined exercise effect on remnant-like lipoprotein
Received in revised form 12 August 2010 particle cholesterol (RLP-C) in type 2 diabetes. Participants were randomized to control (Control), aerobic
Accepted 19 August 2010
(Aerobic), resistance (Resistance), or both (Combined) exercise groups. Baseline and 6-month fasting
Available online xxx
RLP-C and apolipoprotein B48 concentrations were measured. Data analysis was on an intention-to-treat
basis.
Keywords:
At 6 months, RLP-C was lower in all groups; RLP-C mg/dl, (95% confidence interval), Control −3.91,
Resistance exercise
Aerobic exercise
(−6.21 to −1.6), p = 0.001; Aerobic −3.89, (−6.41 to −1.36), p = 0.003, Resistance −7.52, (−9.89 to −5.15),
Remnant lipoproteins p = 0.0001, Combined −7.50, (−9.87 to −5.13), p = 0.0001. Total triglycerides were significantly lower
Diabetes mellitus in Resistance and Combined groups only; −17.7 mg/dl (−32.8 to −2.7), p = 0.02 and −27.5 (−42.5 to
−11.5), p = 0.001, respectively. Inter-group comparisons showed no difference in RLP-C change between
Aerobic and Control and a significant difference in RLP-C change only where groups incorporating resis-
tance exercise were compared with those without. There was no significant difference in RLP-C change
between Resistance and Combined. Inter-group comparisons of total triglycerides change were significant
only between Combined and Control. Changes in apolipoprotein B48 were not significant in inter-group
comparisons.
In conclusion, our data indicate that resistance exercise training, not aerobic, lowers RLP-C in type
2 diabetes. This effect was not revealed by changes in total triglycerides and apolipoprotein B48. The
discordance between changes in RLP-C and apolipoprotein B48 in response to resistance exercise may
indicate (a) a decrease in VLDL remnant and not chylomicron remnant particle number and/or (b) a
depletion of cholesterol in chylomicron and/or VLDL remnants.
© 2010 Elsevier Ireland Ltd. All rights reserved.

1. Introduction
Higher levels of physical fitness are associated with reduced
all-cause mortality, primarily due to reduced rates of cardiovas-
夽 Trial registration: ClinicalTrials.gov study ID NCT00195884, http://www.
clinicaltrials.gov.
∗ Corresponding author at: The Ottawa Hospital Riverside Campus, 1967 Riverside
Drive, Ottawa, ON K1H 7W9 Canada. Tel.: +1 613 738 8400x81937;
fax: +1 613 738 8327.
E-mail address: tcooi@toh.on.ca (T.C. Ooi).
cular disease and cancer [1]. In diabetes, where risk of death from
coronary artery disease (CAD) is 2 to 4 times that of the general
population [2], high aerobic fitness is associated with a 45–70%
reduction in overall mortality [3,4]. We also know that in men, mus-
cular strength adds to the survival advantage of aerobic fitness [5]
although the mechanism is unknown.
There is a mismatch between the magnitude of survival benefit
seen with exercise and the magnitude of change in cardiovascular
risk indices. Improvements in traditional cardioprotective indices
such as glycemia, blood pressure and concentrations of low den-
sity lipoprotein cholesterol (LDL-C), triglycerides and high-density

0021-9150/$ – see front matter © 2010 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.atherosclerosis.2010.08.071

Please cite this article in press as: Gavin C, et al. Resistance exercise but not aerobic exercise lowers remnant-like lipoprotein particle cholesterol
in type 2 diabetes: A randomized controlled trial. Atherosclerosis (2010), doi:10.1016/j.atherosclerosis.2010.08.071
 ARTICLE IN PRESS
G Model
ATH-11615; No. of Pages 6
2 C. Gavin et al. / Atherosclerosis xxx (2010) xxx–xxx
lipoprotein cholesterol (HDL-C) have been reported, but they are
often minor [6]. One explanation is that these classic cardiovascu-
lar disease risk factors do not always reflect sub-clinical etiology
accurately, and that risk assessment limited to these markers fails
to identify a significant proportion of patients at risk for coronary
artery disease [7]. Emerging evidence suggests a role for the mea-
surement of remnant-like lipoprotein particle cholesterol (RLP-C)
concentration, identifying cholesterol associated with very low
density lipoprotein (VLDL) and chylomicron remnants, in coronary
artery risk assessment [8–10].
Despite the well characterized health benefits of exercise,
there remains some controversy surrounding the optimal exercise
modality, duration, frequency and/or intensity for cardioprotec-
tion. Although many reports, including a recent meta-analysis [6]
and a large Italian study [11], have examined the effects of both
aerobic and resistance exercise on the lipoprotein profile, none has
examined the effect of exercise modality on remnant lipoproteins
specifically.
Recently, the Diabetes Aerobic and Resistance Exercise (DARE)
trial investigated the effect of aerobic and resistance exercise train-
ing, alone or in combination, on levels of HbA1c and several other
coronary artery risk factors in patients with type 2 diabetes (T2D)
[12]. The DARE trial was the largest exercise trial (n = 251) of its kind
and revealed that aerobic and resistance exercise each improved
glycemic control, reflected in HbA1c, but their combination reduced
HbA1c at least twice as much as either type of exercise alone. The
objective of this current study, was to assess the effects of aerobic
and resistance exercise alone and in combination, on levels of RLP-C
in the DARE trial participants. We hypothesized that in DARE par-
ticipants, RLP-C would decrease to a greater extent in Aerobic and
in Resistance training groups than in Control. Further, we hypoth-
esized that the decrease in RLP-C would be greater in Combined
than either Aerobic or Resistance groups.
2. Methods
2.1. Study design
The design of the DARE trial, including details of the exer-
cise intervention programs, is described in detail elsewhere [12].
Briefly, the DARE study was a single-center, randomized, controlled
trial with a parallel group design. After a 4-week run-in period
to assess compliance, previously inactive people with T2D were
randomized to one of 4 groups: aerobic training group (“Aero-
bic”), resistance exercise group (“Resistance”), combined aerobic
training and resistance exercise group (“Combined”), and waiting-
list control (“Control”). The participants followed their designated
program for 22 weeks after randomization. Previously sedentary
individuals, 39–70 years of age were recruited. Inclusion criteria
included T2D for at least 6 months and baseline hemoglobin A1C
between 6.6% and 9.9%. Participants exercised at community-based
facilities, supervised by personal trainers. Prior to randomization,
subjects entered a 4-week run-in period to assess compliance. Sub-
jects performed 15–20 min of aerobic exercise at a target intensity
• lyzer. The RLP-C intra and inter assay CVs were 6.0% and 6.33%,
of 60% of maximum heart rate (∼50% of VO2peak ) and 1–2 sets of
7 exercises. All resistance training exercises were performed on
weight machines. Subjects attending at least 10 of the scheduled
12 exercise sessions were eligible for randomization. During the
intervention phase (weeks 5–26), participants exercised 3 times
per week, and training duration and intensity increased on a weekly
basis to a maximum of 45 min per session at 75% of maximum
heart rate. Heart rate monitors (Polar Electro Oy, Kempele, Finland)
displaying the participant’s heart rate were used to standardize
exercise intensity. Target heart rates were based on maximum
heart rate achieved during the maximal treadmill exercise test
Please cite this article in press as: Gavin C, et al. Resistance exercise but not aerobic exercise lowers remnant-like lipoprotein particle cholesterol
in type 2 diabetes: A randomized controlled trial. Atherosclerosis (2010), doi:10.1016/j.atherosclerosis.2010.08.071
performed at baseline. All aerobic activities were performed three
times per week on a treadmill or cycle ergometer. For the resis-
tance training group, the frequency of training increased from 2
to 3 days per week, the number of sets performed for each exer-
cise increased from 2 to 3 sets, the amount of weight lifted for
a given exercise also increased whereas the number of repeti-
tions decreased to a maximum of 8 repetitions. The Combined
group did the full Aerobic program plus the full Resistance pro-
gram. Subsequent to the run-in phase, participants assigned to
the Control group were asked to revert to their level of activity
at baseline and to maintain this level for the remainder of the
study.
This study was approved by the Ottawa Hospital Research Ethics
Board and all subjects gave informed consent. The authors were
solely responsible for designing the DARE study, collection and
analysis of the data and writing the manuscript.
RLP-C was not included as an outcome in the original trial reg-
istration, but was added as a planned secondary study outcome
shortly after the trial began in 2000. Apolipoprotein B-48 (ApoB48)
measurement was added as a secondary outcome in 2008.
For the present analyses, we examined samples from each of the
DARE trial participants and measured RLP-C and ApoB48 at baseline
and at 6 months.
All samples for RLP-C and ApoB48 measurement were stored
at −80 ◦ C until the completion of the study and samples for each
subject were analysed in the same assay. Differences in RLP-C levels
are unlikely to be due to freezing and storage as it has been shown
that samples stored for 6 months were not different from those
frozen for 1 week [13].
Blood was sampled at least 48 h and sometimes up to a full week
after the most recent exercise session. It is known that exercise can
acutely affect TRL metabolism but it is not clear what the duration
of this acute effect is. In an endurance exercise study, fasting TG and
postprandial lipemia increased rapidly with detraining some time
between 15 h and 60 h after the last bout of exercise [14]. There is
no equivalent information for resistance training. It is possible but
unlikely that any change in RLP-C and TG in this study would be
due to the acute effect of exercise.
2.2. Sample handling and assay procedures
Fasting blood was collected in evacuated tubes containing EDTA
(Becton Dickinson Vacutainer Systems, Franklin Lakes, NJ). All sam-
ples were immediately centrifuged at 3200 rpm for 10 min at 4 ◦ C,
then separated and plasma samples were frozen at −80 ◦ C until
completion of the study.
RLP-C was measured in plasma by the method of Nakajima
et al. [15]. As described in our previous study [16], remnant-
like lipoprotein particles were separated by mixing 5 ␮l plasma
with 300 ␮l immunoseparation gel consisting of monoclonal
antibodies to apolipoprotein-B100 and apolipoprotein-A1. Fol-
lowing 2 h incubation at room temperature, cholesterol in the
supernatant (unbound fraction) was measured using a sensitive
enzymatic cholesterol assay adapted for a Roche Cobas Mira ana-
respectively.
Total cholesterol, total triglyceride and HDL-C concentrations
were measured by enzymatic methods, all on a Beckman-Coulter
LX20 analyzer (Beckman Instruments, Brea, CA). LDL-C concen-
tration was calculated using the Friedewald equation. The total
cholesterol, triglycerides and HDL-C inter assay CVs were 1.3%, 1.7%
and 2.8%, respectively.
ApoB48 in plasma was measured using a high sensitiv-
ity sandwich ELISA (Enzyme Linked ImmunoSorbent Assay) kit
from Shibayagi Company, Gunma, Japan. ApoB48 was recog-
nized using a monoclonal antibody as previously described [17].
 ARTICLE IN PRESS
G Model
ATH-11615; No. of Pages 6

C. Gavin et al. / Atherosclerosis xxx (2010) xxx–xxx 3

Table 1
Baseline characteristics.

Characteristic Control (n = 63) Aerobic (n = 60) Resistance (n = 64) Combined (n = 64)

Men/women, n/n 41/22 39/21 40/24 40/24

Mean age, y 54.8 (7.2) 53.9 (6.6) 54.7 (7.5) 53.5 (7.3)
Non-Hispanic white race/other race, n/n 61/2 59/1 55/9 55/9
Mean duration of diabetes, y 5.0 (4.5) 5.1 (3.5) 6.1 (4.7) 5.2 (4.8)
Mean hemoglobin A1c , % 7.66 (0.89) 7.68 (0.85) 7.71 (0.86) 7.67 (0.91)
Lipid values, mg/dla
LDL cholesterol, mg/dlb 115.5 (52.8) 124.9 (56.8) 117.0 (55.6) 119.2 (56.0)
HDL cholesterol, mg/dlc 41.0 (13.5) 42.1 (15.5) 42.7 (15.2) 42.7 (15.2)
Non-HDL cholesterol, mg/dl 153.7 (58.7) 157.1 (64.3) 153.1 (63.2) 151.2 (63.2)
Triglycerides, mg/dlc 166.5 (134.1) 157.2 (137.9) 161.5 (139.2) 142.4 (124.0)
Total cholesterol-HDL cholesterol ratio 4.82 (1.90) 4.78 (2.09) 4.73 (2.08) 4.67 (2.08)
Lipid-lowering agents, n (%)
Total 27 (43) 24 (40) 26 (41) 25 (39)
Statin 24 (38) 17 (28) 26 (41) 22 (34)
Fibrate 6 (10) 9 (15) 1 (2) 7 (11)
Other 0 (0) 2 (3) 0 (0) 1 (2)
Oral hypoglycemic agents, n (%)
Total 50 (83) 49 (82) 48 (75) 43 (67)
Metformin 43 (68) 42 (70) 41 (64) 36 (56)
Sulfonylurea 32 (51) 33 (55) 28 (44) 23 (36)
Meglitinide 4 (6) 2 (3) 4 (6) 1 (2)
Alpha-glucosidase inhibitor 1 (2) 1 (2) 2 (3) 1 (2)
Thiazolidinedione 7 (11) 13 (22) 15 (23) 14 (22)
a
Results are means (SD). Lipid values were entered into linear mixed-effects models, adjusted for age, sex, exercise training site, body mass index, and use or non-use of
oral hypoglycemic medication.
Unless otherwise indicated the sample for analysis was 64 combined exercise training participants, 60 aerobic training participants, 64 resistance training participants and
63 control participants.
b
The sample for analysis was 64 combined exercise training participants, 59 aerobic training participants, 63 resistance training participants and 62 control participants.
Plasma triglyceride levels were too high in 3 participants to use the Friedewald equation to calculate the LDL cholesterol level.
c
Values were transformed to the logarithm for analysis and then exponentiated.

This method correlates significantly with previous methods of 3. Results

densitometric gel scanning following gel electrophoresis [18].
The ApoB48 intra and inter assay CVs were 4.2% and 5.0%,
respectively.
2.3. Statistical analysis
Analyses were intention-to-treat, and all randomized subjects
(including those who later withdrew) were included. SAS version 9
(SAS Institute, Cary, NC, USA) was used for all analyses of continuous
variables. Differences were considered significant if p-value was
≤0.05; we did not adjust for multiple comparisons.
For all continuous variables (RLP-C concentration, ApoB48 con-
centration, other lipids), linear mixed-effects models were used for
repeated measures over time with the outcome as the dependent
variable and effects for time, group (Combined, Aerobic, Resistance
or Control), and time-by-group interaction with an unstructured
covariance matrix. Within the mixed model, 95% confidence inter-
vals and p-values for inter-group contrasts for change in the
outcome variable between baseline and 6 months, and for changes
over time within each group were estimated. For all linear mixed
model analyses, distribution of residuals was examined, and trans-
formations were applied to achieve normality when necessary.
To determine whether effects of the exercise programs differed
according to sex, analyses were repeated with terms for sex, sex-
by-time interaction and sex-by-group-by-time interaction added
to the models.
Several sensitivity analyses were also employed. The above
models were re-run, with age, sex, body mass index, and specific
exercise facility as additional covariates. Additional models were
applied to incorporate all of the above plus use of statins and use
of fibrates and subjects with changes in lipid-altering medication
were excluded. The distributions of residuals were skewed for total
triglycerides, RLP-C and ApoB48; therefore analyses were done
on log-transformed values and results exponentiated to express
results as geometric means.
The participants in all groups were similar in age, sex, dura-
tion of diabetes and medication use (see Table 1). Details of
screening, enrollment, adherence to the intervention protocol and
follow-up assessments have been published previously [12]. Partic-
ipants had dietary counseling with the aim of maintaining baseline
weight. However all groups had similar slight decreases in overall
caloric intake over time. No significant inter-group differences in
macronutrient composition were observed. Changes in BMI were
greater for aerobic training vs control (difference = −0.75 kg/m2 ,
p = 0.009), and almost significantly greater in combined exercise
vs resistance exercise (difference = −0.50 kg/m2 , p = 0.075), but dif-
ferences between resistance training and control, and between
combined training and aerobic training, were not significant.
Aerobic training reduced waist circumference vs control (dif-
ference = −2.1 cm, p = 0.030), and resistance training was close
to having a significant impact vs control, (difference = −1.8 cm,
p = 0.054), but there were no significant differences between com-
bined training and either resistance or aerobic training alone. There
were no statistically significant inter-group differences in changes
of HDL-C, LDL-C, TG, non-HDL-C or total cholesterol to HDL-C ratio
[12].
There were no significant differences in baseline RLP-C, TG and
ApoB48 among the 4 groups (Kruskal Wallis p values = 0.239 for
RLP-C, 0.393 for TG and 0.380 for ApoB48). Within all four groups,
there was a significant decrease in RLP-C at the end of the 6-
month intervention period compared to baseline (Control p = 0.001;
Aerobic p = 0.003; Resistance p = 0.0001; Combined p = 0.0001) (see
Table 2). Inter-group comparisons showed a significant difference
in the change of RLP-C in groups with a resistance exercise com-
ponent only (see Table 3), i.e. Resistance vs Control (p = 0.03),
Combined vs Control (p = 0.03), Aerobic vs Resistance (p = 0.04) and
Combined vs Aerobic (p = 0.04). There was no difference between
groups when both included a resistance exercise component (Com-
bined vs Resistance p = 0.99), and no difference between Aerobic

Please cite this article in press as: Gavin C, et al. Resistance exercise but not aerobic exercise lowers remnant-like lipoprotein particle cholesterol
in type 2 diabetes: A randomized controlled trial. Atherosclerosis (2010), doi:10.1016/j.atherosclerosis.2010.08.071
 ARTICLE IN PRESS
G Model
ATH-11615; No. of Pages 6

4 C. Gavin et al. / Atherosclerosis xxx (2010) xxx–xxx

and Control. There was no material change in all the p values shown

ApoB-48
in Tables 2 and 3 when further adjustment was made for use/non-

1.6
1.4
−0.15* (−0.30 to −0.01) −15.5
−13.3
use of oral diabetes medication, or for each individual class of oral

(%)
hypoglycemic medication. Because of a slight discrepancy in eth-
nic distribution among the groups, we re-analysed the data after

0.01 (−0.13 to 0.16)

0.01 (−0.15 to 0.17)

−0.10 (−0.25 to 0.05)

further adjustment for ethnicity (whites vs non-whites) and the
results remained materially unchanged. Finally, we have performed
a simple linear regression model on the differences, with baseline
ApoB-48

RLP-C as a covariate, and the significance did not change.

(95% CI)

It is noteworthy that changes in RLP-C (reflective of smaller,

denser triglyceride-rich lipoproteins (TRL)) were not mirrored by
changes in total circulating triglycerides, which reflect all TRL,
Results (mg/dl) are means (SE) adjusted for age, BMI, centre, sex, use of statins and use of fibrate. RLP-C and triglycerides n = 251, ApoB-48 n = 250 (control n = 62) unless otherwise stated.
0.85 (0.16)
0.81 (0.18)
0.83 (0.17)
0.68 (0.14)

including the triglyceride-rich very low density lipoproteins (VLDL)

6 months
ApoB-48

and to a much lesser extent, chylomicrons, which constitute a very

small fraction of TRL in the fasting state. Plasma triglycerides were
lower at 6 months in the Resistance and Combined groups only
(Resistance p = 0.02; Combined p = 0.001). Change in triglycerides
0.83 (0.16)
0.79 (0.17)
0.99 (0.21)
0.77 (0.16)
ApoB-48
Baseline

was significantly different between Combined and Control only

(p = 0.01). In clinical practice, changes in triglycerides are often
interpreted to reflect an overall change in RLP-C [16,19] however
our data suggest that this may be an over simplification.
0.5
−4.8
−11.2
−19.3
TG

To ascertain which TRL particles were predominantly affected

(%)

by resistance exercise (chylomicron remnants or apo E-rich VLDL-

−27.5** (−42.5 to −11.5)
0.9 (−14.14 to 15.94)

remnants), we measured ApoB48, which is present in chylomicrons

−7.1 (−23.91 to 8.86)
−17.7* (−32.8 to −2.7)

and chylomicron remnants only. One sample was accidentally

excluded when the ApoB48 assay was run giving n = 250 (not 251).
Interestingly, ApoB48 decreased in the Resistance group only at the
end of the 6-month supervised exercise period (p = 0.04) (Table 2).
(95% CI)

Changes in ApoB48 were not significant in any inter-group com-

TG

parisons (Table 3).

Statin and fibrate medications lower RLP-C [20–22]. Our results
Values were transformed to the logarithm for analysis and results exponentiated to express results as geometric means.

were unchanged when adjusted for lipid-lowering therapy and

166.5 (17.7)
148.8 (17.7)
142.6 (15.9)
118.7 (13.3)
6 months

also when a separate analysis was carried out for patients with no
changes in lipid-lowering therapy during the study (n = 166) (data
TG

not shown).
165.6 (16.8)
155.9 (17.7)
159.4 (17.7)
140.8 (15.9)

4. Discussion
Baseline
TG

Cholesterol in remnant-like lipoprotein particles (RLP-C) corre-

lates with coronary artery disease and is an independent risk factor
RLP-C

for this in the presence of normal total cholesterol levels [23]. The
−22.9
−23.4
−39.1
−51.6
(%)

main finding in this study is that resistance exercise lowers RLP-

C in subjects with T2D. Our study does not support a beneficial
−7.52*** (−9.89 to −5.15)
−7.50*** (−9.87 to −5.13)
−3.91** (−6.21 to −1.60)

contribution of aerobic exercise in T2D on RLP-C. To the best of

−3.89* (−6.41 to −1.36)

our knowledge, this is the first time a sustained beneficial effect of

resistance exercise on remnant lipoproteins has been shown in any
population.
Our findings are important given the association between RLP-
(95% CI)
RLP-C

C and atherosclerosis shown in populations with and without T2D

in epidemiological studies. In a study of 120 subjects (mean age
65.6 years) with T2D and angiographically proven coronary artery
disease, RLP-C was confirmed to be a risk factor for subsequent
Changes in RLP-C, triglycerides and ApoB-48.

13.52 (1.84)

9.32 (1.38)
13.21 (2.04)
12.30 (1.81)

coronary events [8]. Furthermore, RLP-C was a better predictor of

6 months

Significance expressed as p < 0.0001.

coronary events than high HbA1c levels. RLP-C has also been shown
Significance expressed as p < 0.001.
RLP-C

Significance expressed as p < 0.05.

to predict carotid artery intima-media thickness independently of

plasma triglycerides and LDL-C [9]. Cellular mechanisms for the
atherogenicity of remnant-like particles may include upregulation
16.64 (2.58)
19.21 (2.86)
14.54 (2.19)
17.05 (2.36)

of intercellular adhesion molecule-1 (ICAM-1), vascular cell adhe-

Baseline

sion molecule-1 (VCAM-1) and tissue factor on endothelial cells

RLP-C

[24]. These molecules aid monocyte recruitment into arterial wall

and thrombus formation and thus are “proatherothrombogenic”
Resistance

[24]. Remnant lipoproteins are also able to enter the subendothe-

Combined
Aerobic
Control

lial layer and be taken up by macrophages, thus contributing

Table 2

to foam cell formation and atherogenesis. In addition to promo-

*

**

***

tion of atherosclerosis, RLP-C may have a role in coronary artery

Please cite this article in press as: Gavin C, et al. Resistance exercise but not aerobic exercise lowers remnant-like lipoprotein particle cholesterol
in type 2 diabetes: A randomized controlled trial. Atherosclerosis (2010), doi:10.1016/j.atherosclerosis.2010.08.071
 ARTICLE IN PRESS
G Model
ATH-11615; No. of Pages 6
C. Gavin et al. / Atherosclerosis xxx (2010) xxx–xxx
Table 3
Difference in change in RLP-C, triglycerides and ApoB-48 between treatment groups.
RLP-C
Difference p value
Aerobic vs Control 0.02 (−3.40 to 3.44) 0.99
Resistance vs Control −3.62 (−6.92 to −0.31) 0.03
Combined vs Control −3.60 (−6.89 to −0.30) 0.03
Aerobic vs Resistance 3.64 (0.18 to 7.10) 0.04
Combined vs Aerobic −3.62 (−7.07 to −0.17) 0.04
Combined vs Resistance 0.02 (−3.32 to 3.36) 0.99
Results are means (95% CI) adjusted for age, BMI, centre, sex, use of statins and use of fibrate. RLP-C and triglycerides (mg/dl) n = 251, ApoB-48 (mg/dl) n = 250 (control n = 62).
Values were transformed to the logarithm for analysis and results exponentiated to express results as geometric means.
spasm. In human studies, RLP-C correlates independently with
abnormal vasomotor reactivity in coronary arteries. Higher lev-
els of RLP-C are associated with decreased coronary nitric oxide
bioactivity and impairment of endothelium-dependent dilatation
[10].
RLP-C levels have been studied to a limited extent in previously-
published randomized trials in people with dysglycemia. Ai et
al. [25] looked at plasma RLP-C in patients following an energy-
restricted diet and aerobic exercise and found that levels fell in
patients who had a fall in insulin resistance only. This leads to
the question of whether changes in RLP-C in our study correlated
with changes in body composition. In DARE [12], body composi-
tion changed significantly and in an almost identical manner in the
Aerobic and Resistance groups compared to Control, however the
RLP-C responses were quite different. Neither aerobic training nor
resistance training resulted in visceral fat changes differing from
the control group. A small study (n = 20), examining the effect of
a single session of aerobic exercise on postprandial lipid profiles
measured on the following day in a non-diabetic population, found
a significant reduction in RLP-C and total triglycerides, in addition
to chylomicron and VLDL number [26]. A study from the same group
carried out in 10 men with T2D [27] did not show a significant
change in any of the lipid parameters measured, although RLP-C
was not measured.
The mechanisms responsible for lipid/lipoprotein changes are
better described for aerobic than resistance exercise. The decrease
in TG and increase in HDL-C with aerobic exercise are usually
attributed in other studies to increased lipolytic activity of lipopro-
tein lipase, increasing TG removal rates [28]. Our data show that
while changes in ApoB48 relative to baseline were significant for
the Resistance group, there was no difference in change among
groups, including the Control group. This suggests that increased
clearance of RLP may not be a major mechanism responsible for the
decrease in RLP-C with resistance training. The difference in RLP-C
response seen between resistance and aerobic training is unlikely
to be due to different effects on plasma insulin levels and insulin
resistance since effects on plasma insulin and HOMA did not differ
among the exercising groups. Each group experienced a 10–20%
decline in both plasma insulin and HOMA, and inter-group com-
parisons showed no significant difference among the groups (data
not shown).
Within all four groups, there was a significant decrease in RLP-C
at the end of the 6-month intervention period compared to base-
line. A fall in RLP-C within all groups including Control, suggests
that a dietary mechanism is at least partly responsible for the find-
ing.
In our study, changes in RLP-C with resistance exercise were
not mirrored by changes in total triglycerides. As total triglycerides
is the sum of triglycerides from all lipoprotein fractions, espe-
cially TRL, and RLP-C is the cholesterol associated with smaller,
denser remnant lipoproteins only, it is perhaps not surprising that
changes in total triglycerides do not always reveal changes in RLP-C.
Please cite this article in press as: Gavin C, et al. Resistance exercise but not aerobic exercise lowers remnant-like lipoprotein particle cholesterol
in type 2 diabetes: A randomized controlled trial. Atherosclerosis (2010), doi:10.1016/j.atherosclerosis.2010.08.071
5
TG ApoB-48
Difference p value Difference p value
−7.97 (−30.12 to 14.17) 0.46 −0.002 (−0.22 to 0.21) 0.98
−18.60 (−39.86 to 2.66) 0.09 −0.17 (−0.38 to 0.04) 0.11
−28.34 (−49.60 to −7.09) 0.01 −0.12 (−0.32 to 0.09) 0.27
10.63 (−11.51 to 32.77) 0.36 0.17 (−0.05 to 0.38) 0.13
−19.49 (−42.52 to 2.66) 0.08 −0.11 (−0.33 to 0.10) 0.30
−9.74 (−31.00 to 12.40) 0.39 0.05 (−0.16 to 0.26) 0.62
However multiple studies have demonstrated a strong correlation
between total triglyceride concentrations and RLP-C in blood, sug-
gesting that there might not be a need to measure RLP-C when
triglyceride levels are available [16,19]. Our data provide evidence
for a beneficial change in an important atherogenic sub-set of TRL
from resistance exercise, and this is not reflected in total triglyc-
erides which are a broader marker of all TRL. This underlines the
importance of measuring sub-sets of TRL, especially those that have
a known higher atherogenic potential such as the smaller remnants
of chylomicrons and VLDL. Parameters in this study were mea-
sured at one point in time in the fasting but not postprandial state.
Improvements to the lipolytic cascade and liberation of plasma TG
were not measured. Our study does not lend itself to derive firm
conclusions on why there was greater reduction in RLP-C than TG
with resistance exercise. However, it is possible that the reduction
in RLP may have been associated with an increase in the num-
ber of other larger TRL fractions, resulting in little or no change
in serum total TG levels. The exact underlying mechanism(s) for
this is unclear.
To ascertain which triglyceride-rich lipoprotein particles were
predominantly affected by resistance exercise (chylomicron-
remnants or apo E-rich VLDL-remnants), we measured ApoB48,
which is present in chylomicrons and their remnants only. The
fact that there was no difference in ApoB48 change among all the
groups tells us that the effect of resistance exercise was not on
the number of chylomicron remnant particles but, more likely on
the number or composition of VLDL remnants. However, as we
did not measure apolipoprotein B-100 in our remnant-like particle
fraction, we cannot draw definite conclusions about VLDL rem-
nant number from our data. Alternatively, the effect of resistance
exercise on RLP may have been qualitative and not quantitative.
Thus, resistance exercise may have depleted remnant particles of
cholesterol so that the amount of cholesterol associated with them
was reduced (reflected in a decrease in RLP-C) without a change
in the number of particles. This effect may have been on both
VLDL remnants and chylomicron remnants or on one subgroup
only.
In a meta-analysis of 27 exercise studies in type 2 diabetes [6],
aerobic exercise had a small benefit over resistance exercise in low-
ering total cholesterol (−0.23 ± 0.33), whereas combined exercise
had a small benefit for HDL-C (0.41 ± 0.27). Studies of resistance
exercise have sometimes shown a reduction in total cholesterol
and LDL-C [29], correlating with changes in body composition in
some studies and not in others [30]. An earlier Cochrane systematic
review of the same subject [30] looking at all randomized controlled
trials comparing any type of well-documented aerobic, fitness or
progressive resistance training with no exercise in people with T2D,
concluded only that exercise reduces plasma triglycerides but not
total cholesterol. The survival benefits of aerobic and resistance
exercise cannot be easily attributed to these findings. In both the
meta-analysis and the Cochrane review the effect on RLP-C was not
examined.
 ARTICLE IN PRESS
G Model
ATH-11615; No. of Pages 6
6 C. Gavin et al. / Atherosclerosis xxx (2010) xxx–xxx
5. Conclusion
This report has provided evidence of the effects of exercise train-
ing on RLP-C in subjects with T2D. In particular resistance exercise
but not aerobic exercise induced substantial reduction in concen-
tration of RLP-C which was not reflected in triglyceride levels.
Acknowledgements
The measurements of remnant-like particle cholesterol and
apolipoprotein B-48 for the present paper were supported by inter-
nal funds (TCO and SP). SP is supported by a New Investigator Award
from the Heart and Stroke Foundation of Canada. The main DARE
trial was supported by grants from the Canadian Institutes of Health
Research (Grant #MCT-44155) and the Canadian Diabetes Associ-
ation (The Lillian Hollefriend Grant). Dr. Sigal was supported by
a New Investigator Award from the Canadian Institutes of Health
Research and a Research Chair from the Ottawa Health Research
Institute; he is currently a Health Senior Scholar of the Alberta Her-
itage Foundation for Medical Research. Dr. Kenny was supported by
a Career Scientist Award from the Ontario Ministry of Health and
Long Term Care. We are indebted to the DARE study coordinators,
research supporters and participants and all those involved in the
DARE study. Thanks to Annals of Internal Medicine for permission
to include baseline characteristics data from the DARE paper [12].
References
[1] Blair SN, Kohl III HW, Paffenbarger Jr RS, et al. Physical fitness and all-
cause mortality. A prospective study of healthy men and women. JAMA
1989;262:2395–401.
[2] Kannel WB, McGee DL. Diabetes and glucose tolerance as risk factors for car-
diovascular disease: the Framingham study. Diabetes Care 1979;2:120–6.
[3] Hu FB, Stampfer MJ, Solomon C, et al. Physical activity and risk for cardiovas-
cular events in diabetic women. Ann Intern Med 2001;134:96–105.
[4] Lee CD, Blair SN, Jackson AS. Cardiorespiratory fitness, body composition
and all-cause and cardiovascular disease mortality in men. Am J Clin Nutr
1999;69:373–80.
[5] Ruiz JR, Sui X, Lobelo F, et al. Association between muscular strength and mor-
tality in men: prospective cohort study. BMJ 2008;337:a439.
[6] Snowling NJ, Hopkins WG. Effects of different modes of exercise training on
glucose control and risk factors for complications in type 2 diabetic patients: a
meta-analysis. Diabetes Care 2006;29:2518–27.
[7] Stamler J, Vaccaro O, Neaton JD. Diabetes, other risk factors and 12-yr cardio-
vascular mortality for men screened in the Multiple Risk Factor Intervention
Trial. Diabetes Care 1993;16:434–44.
[8] Fukushima H, Sugiyama S, Honda O, et al. Prognostic value of remnant-like
lipoprotein particle levels in patients with coronary artery disease and type II
diabetes mellitus. J Am Coll Cardiol 2004;43:2219–24.
[9] Karpe F, Boquist S, Tang R, et al. Remnant lipoproteins are related to intima-
media thickness of the carotid artery independently of LDL cholesterol and
plasma triglycerides. J Lipid Res 2001;42:17–21.
Please cite this article in press as: Gavin C, et al. Resistance exercise but not aerobic exercise lowers remnant-like lipoprotein particle cholesterol
in type 2 diabetes: A randomized controlled trial. Atherosclerosis (2010), doi:10.1016/j.atherosclerosis.2010.08.071
[10] Kugiyama K, Doi H, Motoyama T, et al. Association of remnant lipoprotein lev-
els with impairment of endothelium-dependent vasomotor function in human
coronary arteries. Circulation 1998;97:2519–26.
[11] Balducci S, Zanuso S. The Italian diabetes and exercise study (IDES): design
and methods for a prospective Italian multicentre trial of intensive lifestyle
intervention in people with type 2 diabetes and the metabolic syndrome. Nutr
Metab Cardiovasc Dis 2008;18:585–95.
[12] Sigal RJ, Kenny GP, Boulé NG, et al. Effects of aerobic training, resistance
training, or both on glycemic control in type 2 diabetes. Ann Intern Med
2007;147:357–69.
[13] McNamara JR, Shah PK, Nakajima K, et al. Remnant lipoprotein cholesterol and
triglyceride reference ranges from the Framingham Heart Study. Clin Chem
1998;44:1224–32.
[14] Hardman AE, Lawrence JEM, Herd SL. Postprandial lipemia in endurance-
trained people during a short interruption to training. J Appl Physiol
1998;84:1895–901.
[15] Nakajima K, Okazaki M, Tanaka A, et al. Separation and determination of
remnant-like particles in human serum using monoclonal antibodies to Apo
B-100 and Apo A-1. J Clin Lig Assay 1996;19:177–83.
[16] Ooi TC, Cousins M, Ooi DS, et al. Postprandial remnant-like lipoproteins in
hypertriglyceridemia. J Clin Endocrinol Metab 2001;86:3134–42.
[17] Kinoshita M, Kojima M, Matsushima T, Teramoto T. Determination of
apolipoprotein B-48 in serum by a sandwich ELISA. Clin Chim Acta
2005;351:115–20.
[18] Su JW, Nzekwu MM, Cabezas MC, Redgrave T, Proctor SD. Methods to assess
impaired post-prandial metabolism and the impact for early detection of car-
diovascular disease risk. Eur J Clin Invest 2009;39:741–54.
[19] Sacks F. Triglycerides are just a marker for specific atherogenic remnant
lipoprotein particles. Atherosclerosis Suppl 2003;4:9.
[20] Stein DT, Devaraj S, Balis D, Adams-Huet B, Jialal I. Effect of statin therapy on
remnant lipoprotein cholesterol levels in patients with combined hyperlipi-
demia. Arterioscler Thromb Vasc Biol 2001;21:2026–31.
[21] Ooi TC, Cousins M, Ooi DS, Nakajima K, Edwards AL. Effect of fibrates on post-
prandial remnant-like particles in patients with combined hyperlipidemia.
Atherosclerosis 2004;172:375–82.
[22] Kolovou GD, Anagnostopoulou KK, Daskalopoulou SS, Mikhailidis DP, Cokki-
nos DV. Clinical relevance of postprandial lipaemia. Curr Med Chem
2005;12:1931–45.
[23] Nakamura T, Kugiyama K. Triglycerides and remnant particles as risk factors
for coronary artery disease. Curr Atheroscler Rep 2006;8:107–10.
[24] Doi H, Kugiyama K, Oka H, et al. Remnant lipoproteins induce proatherothrom-
bogenic molecules in endothelial cells through a redox-sensitive mechanism.
Circulation 2000;102:670–6.
[25] Ai M, Tanaka A, Ogita K, et al. Relationship between hyperinsulinemia and rem-
nant lipoprotein concentrations in patients with impaired glucose tolerance. J
Clin Endocrinol Metab 2000;85:3557–60.
[26] Gill JM, Al-Mamari A, Ferrell WR, et al. Effects of a moderate exercise ses-
sion on postprandial lipoproteins, apolipoproteins and lipoprotein remnants
in middle-aged men. Atherosclerosis 2006;185:87–96.
[27] Gill JM, Al-Mamari A, Ferrell WR, et al. Effect of prior moderate exercise on
postprandial metabolism in men with type 2 diabetes mellitus, heterogeneity
of responses. Atherosclerosis 2007;194:134–43.
[28] Cohen JC, Noakes TD, Spinnler Benade AJ. Postprandial lipemia and chylomi-
cron clearance in athletes and in sedentary men. Am J Clin Nutr 1998;49:
443–7.
[29] Boyden T, Pamenter R. Resistance exercise training is associated with decreases
in serum low-density lipoprotein cholesterol levels in pre-menopausal women.
Arch Intern Med 1993;153:97–100.
[30] Thomas D, Elliott EJ, Naughton GA. Exercise for type 2 diabetes mellitus.
Cochrane Database Syst Rev 2006;3:002968.