A single bout of moderate-intensity exercise increases whole-body insulin sensitivity for 12–48 h post-exercise; however, the relationship between exercise energy expenditure and the improvement in insulin sensitivity is not known. We hypothesized that the exercise-induced increase in whole-body insulin sensitivity, assessed with HOMAIR (homoeostasis model assessment of insulin resistance), is directly related to the energy expended during exercise. We studied 30 recreationally active non-obese men (age, 27±5 years; body mass index, 24±2 kg/m2) in the post-absorptive state on two separate occasions: once after exercising at 60% of V̇O22peak (peak oxygen consumption) for 30–120 min on the preceding afternoon (expending a total of 1.28–5.76 MJ) and once after an equivalent period of rest. Blood samples were obtained the following morning. Exercise-induced changes in HOMAIR were curvilinearly related to exercise energy expenditure (r=−0.666, P=0.001) with a threshold of approx. 3.77 MJ (900 kcal) for improvements in HOMAIR to be manifested. In particular, HOMAIR was reduced by 32±24% (P=0.003) in subjects who expended more than 3.77 MJ during exercise, but did not change for those who expended fewer than 3.77 MJ (−2±21%; P=0.301). Furthermore, the magnitude of change in HOMAIR after exercise was directly associated with baseline (i.e. resting) HOMAIR (r=−0.508, P=0.004); this relationship persisted in multivariate analysis. We conclude that improved whole-body insulin resistance after a single bout of exercise is curvilinearly related to exercise energy expenditure, and requires unfeasible amounts of exercise for most sedentary individuals.

INTRODUCTION

Regular, moderate-intensity, endurance-type physical activity is associated with significantly reduced risk for Type 2 diabetes and cardiovascular disease, in part due to enhanced insulin action [1]. Whole-body insulin sensitivity is higher in trained athletes than in untrained subjects, and improves considerably with exercise training in previously sedentary individuals [2]. Most of the enhancement in insulin action associated with exercise training is attributed to the last bout of exercise, and is lost after 3–6 days of inactivity [25]. In fact, insulin-mediated whole-body glucose uptake is increased 12–48 h after a single exercise session in both healthy [69] and insulin-resistant subjects [8,10,11], but returns to baseline values thereafter [25].

The amount of exercise required to elicit an enhancement in insulin sensitivity remains uncertain. Furthermore, there is considerable inter-individual variability in the metabolic response to acute exercise and the concomitant changes in glucose and insulin dynamics [12], some of which may be related to baseline insulin resistance [10,11,13]. Understanding the dose–response nature of these relationships will have important physiological and public health implications. Current public guidelines advocate 30–60 min of moderate-intensity exercise on most days of the week [1]; however, approx. 70% of the adult population fails to meet the recommended 30-min goal of regular exercise and approx. 40% does not engage in any kind of physical activity [14]. It was therefore our objective to examine the relationship between the energy expended during exercise and basal whole-body insulin sensitivity, assessed by using the HOMAIR (homoeostasis model assessment of insulin resistance), in healthy untrained men.

MATERIALS AND METHODS

Subjects and preliminary testing

A total of 30 non-obese recreationally active but untrained men participated in the study (Table 1). All subjects engaged in moderate-intensity physical activities ≤twice/week, and were considered to be in good health after completing a medical evaluation, which included a history and physical examination and standard blood tests. None were smokers or taking medication. Body composition (fat mass and fat-free mass) was determined by dual-energy X-ray absorptiometry [Delphi-W (Hologic) and DPX-MD+ (Lunar)], and V̇O2peak [peak V̇O2 (oxygen consumption)] was measured with an incremental exercise test on a treadmill (Vmax229; SensorMedics) or a cycloergometer (TrueOne 2400; ParvoMedics), depending on whether subjects were assigned to perform running or cycling exercise (see below). Written informed consent was obtained from all subjects before participation in the study, which was approved by the Human Studies Committee and the General Clinical Research Center Advisory Committee at Washington University School of Medicine in St. Louis, MO, U.S.A. and the Bioethics Committee of Harokopio University, Athens, Greece.

Table 1
Subject characteristics

Values are means±S.D. *P<0.001 compared with the corresponding value in subjects who expended <3.77 MJ (900 kcal) during exercise. †Values in parentheses are expressed as a percentage of the V̇O2peak; ‡values in parentheses are in kcal.

Exercise energy expenditure
ParameterAll subjects<3.77 MJ (900 kcal)>3.77 MJ (900 kcal)
n 30 18 12 
Age (years) 27±5 27±5 27±4 
Body mass index (kg/m224±2 24±2 24±3 
Weight (kg) 79±11 77±7 82±15 
Body fat (% of body weight) 17±4 17±4 17±5 
Fat-free mass (kg) 63±8 61±6 66±9 
V̇O2peak (l/min) 3.3±0.7 3.2±0.7 3.4±0.6 
V̇O2 during exercise (l/min)† 2.0±0.3 (62±2) 1.9±0.3 (61±2) 2.0±0.3 (63±2) 
Total energy expenditure during exercise (MJ)‡ 3.23±1.40 (771±336) 2.25±0.80 (537±190) 4.69±0.56 (1121±134)* 
Exercise energy expenditure
ParameterAll subjects<3.77 MJ (900 kcal)>3.77 MJ (900 kcal)
n 30 18 12 
Age (years) 27±5 27±5 27±4 
Body mass index (kg/m224±2 24±2 24±3 
Weight (kg) 79±11 77±7 82±15 
Body fat (% of body weight) 17±4 17±4 17±5 
Fat-free mass (kg) 63±8 61±6 66±9 
V̇O2peak (l/min) 3.3±0.7 3.2±0.7 3.4±0.6 
V̇O2 during exercise (l/min)† 2.0±0.3 (62±2) 1.9±0.3 (61±2) 2.0±0.3 (63±2) 
Total energy expenditure during exercise (MJ)‡ 3.23±1.40 (771±336) 2.25±0.80 (537±190) 4.69±0.56 (1121±134)* 

Experimental protocol

Each subject completed two studies within 3 weeks in a randomized order: once after resting and once after exercising on the preceding afternoon. Subjects were instructed to adhere to their regular diet and to refrain from exercise for a minimum of 3 days before the start of each study (rest and exercise). For the exercise study, they came to the laboratory and cycled on a semi-recumbent cycloergometer (Cateye Fitness) or ran on a motor-driven treadmill (Technogym Runrace) at 60% of their V̇O2peak for 30–120 min between 17.30 and 19.30 hours (seven subjects exercised for 30 min; seven for 60 min; nine for 90 min, and seven for 120 min). There were no significant differences in V̇O2peak between individuals who exercised for different amounts of time (3.3±0.8, 3.2±0.6, 3.3±0.7 and 3.2±0.5 l/min, respectively; P=0.883 as determined by one-way ANOVA). V̇O2 was measured at regular intervals during exercise (every 5, 15 or 30 min, depending on the duration of exercise) to calculate total energy expenditure of the bout (range, 1.28–5.76 MJ; i.e. 306–1376 kcal) [15] and adjust the workload, if necessary, to maintain the desired V̇O2 within ±5%. The different modes and durations of exercise were chosen in order to cover the range from the minimum recommended amount of exercise (i.e. 30 min of moderate-intensity) [1] to an exercise session that most untrained individuals would find it difficult to complete (i.e. 2 h in duration) [16], and ensured that the experimental conditions would closely reflect a realistic situation, in which unaccustomed individuals engage in various activities for variable amounts of time. Subjects were not allowed access to any kind of food or beverage during exercise, other than water. For the resting study, subjects lay in bed or sat in a chair. After completion of the exercise or the equivalent period of rest, they consumed the same standardized meal [50 kJ (12 kcal)/kg of body weight; 50–55% carbohydrate, 30–35% fat and 15–20% protein] at approx. 20.00 hours, in order to eliminate between-trial differences in energy balance other than the energy deficit incurred by exercise, and then fasted (except for water) until completion of the study the next day.

At 08.00 hours the following morning, an arterialized blood sample was obtained from a heated forearm vein for the determination of basal plasma glucose and insulin concentrations. Blood was collected in chilled tubes containing heparin (for glucose) or EDTA plus aprotinin (for insulin) and placed immediately on ice. Plasma was separated by centrifugation within 30 min of collection, and samples were stored at −80 °C until analysis. Plasma glucose concentration was determined by the glucose oxidase method on an automated glucose analyser (YSI 2300 STAT+; Yellow Spring Instruments). Plasma insulin concentration was measured by RIA (Linco Research). The HOMAIR score, which reflects whole-body insulin resistance, was calculated as the product of fasting plasma insulin [in mU/l (milli-international units/l)] and glucose (in mmol/l) concentrations divided by 22.5 [17].

Statistical analysis

Data were analysed with SPSS 13 for Windows (SPSS Inc). All datasets were normally distributed according to the Kolmogorov–Smirnov procedure. Results are presented as means±S.D. Results after rest and exercise were compared with a two-tailed paired Student's t test. Relationships between variables of interest were examined with correlation and regression analyses. P<0.05 was considered statistically significant.

RESULTS

Fasting plasma glucose and insulin concentrations, and therefore the HOMAIR score, were significantly lower in the morning after exercise than at rest (Table 2). These responses varied considerably between individuals (Table 2), but were not affected by the mode of exercise (running or cycling) in any consistent manner (results not shown).

Table 2
Fasting plasma glucose and insulin concentrations and HOMAIR score after a single evening exercise bout or an equivalent period of rest

Values are means±S.D.

ParameterRestExerciseExercise-induced change (%)P value
Glucose (mmol/l) 5.3±0.4 5.1±0.4 −3±6 0.005 
Insulin (mU/l) 6.5±2.6 5.4±2.1 −12±26 0.004 
HOMAIR score 1.56±0.67 1.25±0.52 −14±26 0.003 
ParameterRestExerciseExercise-induced change (%)P value
Glucose (mmol/l) 5.3±0.4 5.1±0.4 −3±6 0.005 
Insulin (mU/l) 6.5±2.6 5.4±2.1 −12±26 0.004 
HOMAIR score 1.56±0.67 1.25±0.52 −14±26 0.003 

Total energy expenditure during exercise correlated negatively with exercise-induced changes in fasting plasma glucose (r=−0.482, P=0.007) and insulin (r=−0.521, P=0.003) concentrations.

There was a negative curvilinear relationship between total energy expenditure during exercise and changes in HOMAIR score, with no apparent effect of exercise below a caloric expenditure of approx. 3.77 MJ (i.e. 900 kcal) (r=−0.666, P=0.001; Figure 1). Hence HOMAIR was reduced by approx. 30% in subjects who expended more than 3.77 MJ during exercise (n=12; P=0.003), but did not change for those who expended fewer than 3.77 MJ (n=18; P=0.301; Figure 1). There were no significant differences in subject characteristics between these two groups (Table 1).

Exercise-induced changes in HOMAIR as a function of total energy expenditure during exercise (a), and for subjects who expended less than or more than 3.77 MJ (900 kcal) during the exercise bout (b)

Figure 1
Exercise-induced changes in HOMAIR as a function of total energy expenditure during exercise (a), and for subjects who expended less than or more than 3.77 MJ (900 kcal) during the exercise bout (b)

In (b), values are means±S.D. *P<0.05 compared with rest.

Figure 1
Exercise-induced changes in HOMAIR as a function of total energy expenditure during exercise (a), and for subjects who expended less than or more than 3.77 MJ (900 kcal) during the exercise bout (b)

In (b), values are means±S.D. *P<0.05 compared with rest.

The relationship between changes in HOMAIR score and exercise energy expenditure was linear when the latter was expressed relative to body weight (r=−0.577, P=0.001). However, the percentage variance explained by this linear relationship (R2=33.3%) was less than that explained by the curvilinear relationship with total energy expenditure of exercise (R2=44.4%).

The magnitude of change in HOMAIR after exercise was inversely correlated with the baseline (i.e. resting) HOMAIR score (r=−0.508, P=0.004; Figure 2).

Exercise-induced changes in HOMAIR as a function of baseline (resting) HOMAIR score

Figure 2
Exercise-induced changes in HOMAIR as a function of baseline (resting) HOMAIR score
Figure 2
Exercise-induced changes in HOMAIR as a function of baseline (resting) HOMAIR score

In multivariate linear regression analysis, including exercise energy expenditure, age, body mass index, body weight, fat mass and fat-free mass, V̇O2peak, and baseline (i.e. resting) plasma glucose and insulin concentrations and HOMAIR score, total energy expenditure during exercise (standardized β=−0.494, P=0.001) and baseline HOMAIR score (β=−0.420, P=0.005) were the only significant independent predictors of the exercise-induced change in HOMAIR, accounting for 49.4% of the total variance (F=13.2, P<0.001).

DISCUSSION

In the present study, we investigated the relationship between the energy expended during a single bout of moderate-intensity endurance exercise, performed in the evening, and fasting plasma glucose and insulin concentrations and insulin resistance (HOMAIR) the following morning in healthy, non-obese, untrained men. Our findings suggest that improvements in whole-body insulin sensitivity, reflected by a decreased HOMAIR score, are curvilinearly related to exercise energy expenditure with a threshold of approx. 3.77 MJ (900 kcal) for a beneficial effect to be manifested. However, exercise-mediated changes in whole-body insulin resistance are inversely related to baseline HOMAIR score, even within this group of normoglycaemic and normoinsulinaemic subjects with normal HOMAIR [18], implying that insulin-resistant individuals may benefit more from exercise than insulin-sensitive ones. Nonetheless, single sessions of typical recreational activities are unlikely to have a significant impact on whole-body insulin sensitivity in healthy sedentary individuals.

It is well established that a single session of strenuous exercise leads to increased insulin-mediated whole-body glucose uptake for some 12–48 h post-exercise in healthy and insulin-resistant subjects [68,10,11]. In our present study, whole-body insulin sensitivity improved proportionally with the energy expended during exercise when this exceeded approx. 3.77 MJ, whereas no changes in HOMAIR score occurred after a total energy cost of exercise less than 3.77 MJ (900 kcal). This energy expenditure threshold is equivalent to ≥60–90 min of exercise at 60% of V̇O2peak in our recreationally active men, which by far exceeds current public recommendations for physical activity [1]. In fact, most sedentary individuals will probably not be able to exercise for more than 1 h at this intensity [16]. The existence of an energy expenditure threshold in the insulin-sensitizing effect of exercise is in line with findings on other physiological outcomes of acute exercise, for example the lowering of plasma triacylglycerol (triglyceride) concentrations [19,20]. Our present findings also help to explain the lack of a consistent effect of exercise on insulin sensitivity in some previous studies because too little exercise was performed (corresponding to total caloric expenditures ≤2.1 MJ or 500 kcal) [13,21]. These results collectively suggest that improvements in basal whole-body insulin sensitivity after a single bout of exercise require unfeasible amounts of exercise for most untrained individuals.

It is possible that regular exercise training lowers the threshold of exercise required to improve whole-body insulin sensitivity; the effects of acute exercise on insulin action appear to be more pronounced in the trained state [2] and everyday leisure-time physical activity is associated with dose-dependent reductions in diabetes risk in epidemiological surveys [2225]. However, whether exercise training or the intensity of exercise modify the dose–response relationship between energy expenditure and insulin sensitivity in any way remains to be determined. Furthermore, the inverse association between exercise-induced changes in HOMAIR and baseline HOMAIR implies that less exercise may be required to improve insulin sensitivity in insulin-resistant subjects than those with good insulin sensitivity at baseline, i.e. before engaging in exercise. For instance, some of our subjects with higher baseline HOMAIR scores, albeit within the normal range [18], exhibited large improvements in insulin resistance (i.e. approx. 20–40% reduction) after expending only 1.26–2.51 MJ (i.e. 300–600 kcal) during exercise. Likewise, insulin-mediated whole-body glucose disposal increased to a greater extent in insulin-resistant obese subjects compared with insulin-sensitive lean subjects approx. 12 h after a strenuous session of exercise [10], and HOMAIR decreased in the morning after a single 90-min exercise bout at 60% of V̇O2peak in men, but not in women, whose baseline HOMAIR was half that of men [26]. Regular exercise, even of low energy cost, should therefore not be rejected as a means of improving insulin sensitivity in insulin-resistant subjects.

The exact cellular mechanisms responsible for the increase in insulin sensitivity late into recovery from a single bout of exercise are not well understood. Exercise-induced changes in insulin sensitivity have been linked to the depletion of skeletal muscle glycogen and/or triacylglycerol stores [27]. Depletion of muscle glycogen [68,10,11] leads to enhanced post-exercise uptake of glucose to facilitate glycogen replenishment [2,3,28]. The major cellular event underlying this phenomenon is increased translocation of the GLUT4 isoform of the glucose transporter from its intracellular storage sites to the cell surface [29]; however, the mechanism(s) responsible for mediating this event, the signals involved and the amount of exercise required to elicit these signals remain poorly defined [28]. Glycogen repletion in the recovery from exercise occurs in two distinct phases: an early insulin-independent period of rapid glycogen resynthesis (lasting approx. 1 h after cessation of exercise) and a subsequent period (up to 1–2 days post-exercise) of slow glycogen resynthesis which is insulin-dependent [30]. This mechanism fits well with the results from our present study, in which muscle glycogen stores were probably not depleted to any significant extent, except for those subjects who exercised for ≥60–90 min and expended more than approx. 3.77 MJ (900 kcal), as skeletal muscle glycogen is reduced dose-dependently with the total energy cost and, hence, the duration of exercise [30,31]. We are not aware of any studies examining the possible dose-dependency of exercise-induced changes in relevant signalling pathways. Furthermore, glycogen resynthesis rates are substantially slower in insulin-resistant subjects compared with insulin-sensitive subjects in the late post-exercise period [32], which could help explain our present observation that relatively insulin-resistant subjects (as indicated by higher baseline HOMAIR scores) enjoyed greater decreases in HOMAIR after exercise than relatively insulin-sensitive ones, because persistence of low glycogen stores augments the exercise-induced enhancement in muscle insulin action [28]. Intramuscular triacylglycerol content is also closely associated with insulin sensitivity [33]. Diminution of skeletal muscle lipid stores and/or enhanced lipid oxidation after exercise could therefore also facilitate muscle insulin action [34]. The relative contribution of changes in skeletal muscle glycogen and triacylglycerol metabolism to the exercise-induced improvements in insulin sensitivity and the role, if any, of the energy expenditure of exercise are currently not known.

In summary, in the present study, we investigated the relationship between the total energy expended during a single bout of moderate-intensity evening exercise and changes in fasting plasma glucose and insulin concentrations and HOMAIR in healthy, non-obese, untrained men. Our results indicate that more than 1 h of moderate-intensity exercise is required to improve basal whole-body insulin sensitivity, assessed with the HOMAIR score. Whether the same applies for repeated exercise sessions (i.e. training) remains to be studied. Although the interpretation of our findings is limited by the use of HOMAIR as a surrogate index of whole-body insulin resistance, there is a good correlation (r=0.7–0.9) between HOMAIR and estimates of insulin sensitivity derived from hyperinsulinaemic–euglycaemic clamp and minimal model analysis [35]. These findings may therefore be useful for the development of appropriate exercise protocols targeted at ameliorating insulin resistance.

Abbreviations

     
  • HOMAIR

    homoeostasis model assessment of insulin resistance

  •  
  • mU

    milli-international unit

  •  
  • V̇O2

    oxygen consumption

  •  
  • V̇O2peak

    peak V̇O2

We thank Amalia Yanni and Maria Maraki for technical support, and the study subjects for their participation. This study was supported by grants from the American Heart Association (0365436Z and 0510015Z), the Hellenic Heart Foundation, and National Institutes of Health grants AR 49869, DK 56341 (Clinical Nutrition Research Unit) and RR 00036 (General Clinical Research Center).

References

References
1
Bassuk
 
S. S.
Manson
 
J. E.
 
Epidemiological evidence for the role of physical activity in reducing risk of type 2 diabetes and cardiovascular disease
J. Appl. Physiol.
2005
, vol. 
99
 (pg. 
1193
-
1204
)
2
Borghouts
 
L. B.
Keizer
 
H. A.
 
Exercise and insulin sensitivity: a review
Int. J. Sports Med.
2000
, vol. 
21
 (pg. 
1
-
12
)
3
Henriksen
 
E. J.
 
Effects of acute exercise and exercise training on insulin resistance
J. Appl. Physiol.
2002
, vol. 
93
 (pg. 
788
-
796
)
4
Henriksson
 
J.
 
Influence of exercise on insulin sensitivity
J. Cardiovasc. Risk
1995
, vol. 
2
 (pg. 
303
-
309
)
5
van Baak
 
M. A.
Borghouts
 
L. B.
 
Relationships with physical activity
Nutr. Rev.
2000
, vol. 
58
 (pg. 
S16
-
S18
)
6
Mikines
 
K. J.
Sonne
 
B.
Farrell
 
P. A.
Tronier
 
B.
Galbo
 
H.
 
Effect of physical exercise on sensitivity and responsiveness to insulin in humans
Am. J. Physiol.
1988
, vol. 
254
 (pg. 
E248
-
E259
)
7
Bogardus
 
C.
Thuillez
 
P.
Ravussin
 
E.
Vasquez
 
B.
Narimiga
 
M.
Azhar
 
S.
 
Effect of muscle glycogen depletion on in vivo insulin action in man
J. Clin. Invest.
1983
, vol. 
72
 (pg. 
1605
-
1610
)
8
Perseghin
 
G.
Price
 
T. B.
Petersen
 
K. F.
, et al 
Increased glucose transport-phosphorylation and muscle glycogen synthesis after exercise training in insulin-resistant subjects
N. Engl. J. Med.
1996
, vol. 
335
 (pg. 
1357
-
1362
)
9
Annuzzi
 
G.
Riccardi
 
G.
Capaldo
 
B.
Kaijser
 
L.
 
Increased insulin-stimulated glucose uptake by exercised human muscles one day after prolonged physical exercise
Eur. J. Clin. Invest.
1991
, vol. 
21
 (pg. 
6
-
12
)
10
Devlin
 
J. T.
Horton
 
E. S.
 
Effects of prior high-intensity exercise on glucose metabolism in normal and insulin-resistant men
Diabetes
1985
, vol. 
34
 (pg. 
973
-
979
)
11
Devlin
 
J. T.
Hirshman
 
M.
Horton
 
E. D.
Horton
 
E. S.
 
Enhanced peripheral and splanchnic insulin sensitivity in NIDDM men after single bout of exercise
Diabetes
1987
, vol. 
36
 (pg. 
434
-
439
)
12
Wallberg-Henriksson
 
H.
 
Acute exercise: fuel homeostasis and glucose transport in insulin-dependent diabetes mellitus
Med. Sci. Sports Exercise
1989
, vol. 
21
 (pg. 
356
-
361
)
13
Kang
 
J.
Robertson
 
R. J.
Hagberg
 
J. M.
, et al 
Effect of exercise intensity on glucose and insulin metabolism in obese individuals and obese NIDDM patients
Diabetes Care
1996
, vol. 
19
 (pg. 
341
-
349
)
14
Schoenborn
 
C. A.
Barnes
 
P. M.
 
Leisure-time physical activity among adults: United States, 1997–1998, Advance data from vital and health statistics; no. 325.
2002
Hyattsville, MD
National Center for Health Statistics
15
Frayn
 
K. N.
 
Calculation of substrate oxidation rates in vivo from gaseous exchange
J. Appl. Physiol.
1983
, vol. 
55
 (pg. 
628
-
634
)
16
Cullinane
 
E.
Siconolfi
 
S.
Saritelli
 
A.
Thompson
 
P. D.
 
Acute decrease in serum triglycerides with exercise: is there a threshold for an exercise effect?
Metab. Clin. Exp.
1982
, vol. 
31
 (pg. 
844
-
847
)
17
Matthews
 
D. R.
Hosker
 
J. P.
Rudenski
 
A. S.
Naylor
 
B. A.
Treacher
 
D. F.
Turner
 
R. C.
 
Homeostasis model assessment: insulin resistance and β-cell function from fasting plasma glucose and insulin concentrations in man
Diabetologia
1985
, vol. 
28
 (pg. 
412
-
419
)
18
Ascaso
 
J. F.
Pardo
 
S.
Real
 
J. T.
Lorente
 
R. I.
Priego
 
A.
Carmena
 
R.
 
Diagnosing insulin resistance by simple quantitative methods in subjects with normal glucose metabolism
Diabetes Care
2003
, vol. 
26
 (pg. 
3320
-
3325
)
19
Magkos
 
F.
Patterson
 
B. W.
Mohammed
 
B. S.
Mittendorfer
 
B.
 
A single 1-h bout of evening exercise increases basal FFA flux without affecting VLDL-triglyceride and VLDL-apolipoprotein B-100 kinetics in untrained lean men
Am. J. Physiol. Endocrinol. Metab.
2007
, vol. 
292
 (pg. 
E1568
-
E1574
)
20
Superko
 
H. R.
 
Exercise training, serum lipids, and lipoprotein particles: is there a change threshold?
Med. Sci. Sports Exercise
1991
, vol. 
23
 (pg. 
677
-
685
)
21
Cusi
 
K.
Maezono
 
K.
Osman
 
A.
, et al 
Insulin resistance differentially affects the PI 3-kinase- and MAP kinase-mediated signaling in human muscle
J. Clin. Invest.
2000
, vol. 
105
 (pg. 
311
-
320
)
22
Helmrich
 
S. P.
Ragland
 
D. R.
Leung
 
R. W.
Paffenbarger
 
R. S.
 
Physical activity and reduced occurrence of non-insulin-dependent diabetes mellitus
N. Engl. J. Med.
1991
, vol. 
325
 (pg. 
147
-
152
)
23
Hu
 
F. B.
Leitzmann
 
M. F.
Stampfer
 
M. J.
Colditz
 
G. A.
Willett
 
W. C.
Rimm
 
E. B.
 
Physical activity and television watching in relation to risk for type 2 diabetes mellitus in men
Arch. Intern. Med.
2001
, vol. 
161
 (pg. 
1542
-
1548
)
24
Folsom
 
A. R.
Kushi
 
L. H.
Hong
 
C. P.
 
Physical activity and incident diabetes mellitus in postmenopausal women
Am. J. Public Health
2000
, vol. 
90
 (pg. 
134
-
138
)
25
Hu
 
F. B.
Sigal
 
R. J.
Rich-Edwards
 
J. W.
, et al 
Walking compared with vigorous physical activity and risk of type 2 diabetes in women: a prospective study
JAMA, J. Am. Med. Assoc.
1999
, vol. 
282
 (pg. 
1433
-
1439
)
26
Gill
 
J. M.
Herd
 
S. L.
Tsetsonis
 
N. V.
Hardman
 
A. E.
 
Are the reductions in triacylglycerol and insulin levels after exercise related?
Clin. Sci.
2002
, vol. 
102
 (pg. 
223
-
231
)
27
Thompson
 
P. D.
Crouse
 
S. F.
Goodpaster
 
B.
Kelley
 
D.
Moyna
 
N.
Pescatello
 
L.
 
The acute versus the chronic response to exercise
Med. Sci. Sports Exercise
2001
, vol. 
33
 (pg. 
S438
-
S445
)
28
Holloszy
 
J. O.
 
Exercise-induced increase in muscle insulin sensitivity
J. Appl. Physiol.
2005
, vol. 
99
 (pg. 
338
-
343
)
29
Hansen
 
P. A.
Nolte
 
L. A.
Chen
 
M. M.
Holloszy
 
J. O.
 
Increased GLUT-4 translocation mediates enhanced insulin sensitivity of muscle glucose transport after exercise
J. Appl. Physiol.
1998
, vol. 
85
 (pg. 
1218
-
1222
)
30
Price
 
T. B.
Rothman
 
D. L.
Shulman
 
R. G.
 
NMR of glycogen in exercise
Proc. Nutr. Soc.
1999
, vol. 
58
 (pg. 
851
-
859
)
31
Hultman
 
E.
 
Physiological role of muscle glycogen in man, with special reference to exercise
Circ. Res.
1967
, vol. 
20–21
 (pg. 
I99
-
I114
)
32
Price
 
T. B.
Perseghin
 
G.
Duleba
 
A.
, et al 
NMR studies of muscle glycogen synthesis in insulin-resistant offspring of parents with non-insulin-dependent diabetes mellitus immediately after glycogen-depleting exercise
Proc. Natl. Acad. Sci. U.S.A.
1996
, vol. 
93
 (pg. 
5329
-
5334
)
33
Goodpaster
 
B. H.
Brown
 
N. F.
 
Skeletal muscle lipid and its association with insulin resistance: what is the role for exercise?
Exercise Sport Sci. Rev.
2005
, vol. 
33
 (pg. 
150
-
154
)
34
Bruce
 
C. R.
Hawley
 
J. A.
 
Improvements in insulin resistance with aerobic exercise training: a lipocentric approach
Med. Sci. Sports Exercise
2004
, vol. 
36
 (pg. 
1196
-
1201
)
35
Wallace
 
T. M.
Levy
 
J. C.
Matthews
 
D. R.
 
Use and abuse of HOMA modeling
Diabetes Care
2004
, vol. 
27
 (pg. 
1487
-
1495
)