Target Audience: Researchers interested in obesity and diabetes.

Introduction: Increased accumulation of intrahepatic lipids (IHL) and intramyocellular lipids (IMCL) in obesity is strongly associated with insulin resistance [1]. Long-term exercise training has been reported to increase the IMCL levels, while improving insulin sensitivity. The goal of this study was to investigate how dietary fat intake modulates the effect of long term exercise training on body weight, IHL and IMCL/Cr, and the resultant effect on insulin sensitivity in an animal model of long-term exercise.

Animals and Methods: All in-vivo experiments were in compliance and approved by institutional animal care and use committee. 20 male Wistar rats were equally divided into the following four groups: ‘Sedentary + Chow diet’, ‘Exercise + Chow diet’, ‘Sedentary + High fat diet’ and ‘Exercise + High fat diet’. Exercised rats were made to run on an animal treadmill (Columbus-1055SRM-E54 Exer-3/6-Dual). Rats were allowed to get habituated to exercise activity using a low intensity exercise protocol (10 to 14 m/min for 10 min per day) for one week. The long-term exercise protocol involved treadmill running at the rate of 18m/min for 30 min/day for 12 weeks. Rats in the sedentary cohorts did not undergo any exercise intervention. Post-measurements of body weight, IHL, IMCL/Cr and insulin sensitivity were made after 12 weeks of intervention. In vivo imaging experiments were performed using a 7 T Bruker ClinScan MRI/MRS scanner with motion compensation, using a 72mm volume resonator for RF transmits and 20mm receive-only coil. Volume localized PRESS experiments were performed on a 64 mm3 and 27 mm3 voxel with TR=4 s, TE=13ms, 64/128 averages and 2048 complex points acquired with a spectral width of 3500 Hz for liver and muscle respectively. Spectra were analyzed using LC Model [2]. Animals were also subjected to an oral glucose tolerance test (OGTT) after 12 hours of fasting. The Matsuda index [3] of insulin sensitivity was estimated using serum glucose and insulin values at 0, 30, 60, 90 and 120 mins of the OGTT. The effects of diet and exercise on body weight, IHL, IMCL/Cr and insulin sensitivity in the pooled cohort were analyzed using two-way analysis of variance (two-way ANOVA). Practical and statistical significance in pairwise group comparisons were performed using Cohen’s d effect size measure [4] and the Student’s t-test.

Results: The statistical significance of the ‘diet’, ‘exercise’, and ‘diet × exercise’ interaction term, when body weight , IHL, IMCL/Cr and insulin sensitivity were entered as dependent variables for the two-way ANOVA analysis are shown in Table 1. Fig. 1(a)-1(d) show the boxplots of these dependent variables. The main term for diet was significant for all the dependent variables, while the main term for exercise was significant only for IHL and IMCL/Cr. The ‘diet × exercise’ interaction term was significant for IHL and IMCL/Cr but not for body weight and insulin sensitivity. With low dietary fat (chow diet), long-term exercise resulted in non-significant (P>0.05) increases in IHL (Cohen’s d = 0.70), IMCL/Cr (Cohen’s d = 0.43) and Matsuda’s index (Cohen’s d = 0.35), and a non-significant decrease in body weight (Cohen’s d = -0.17). Under the high-fat diet regimen, long-term exercise resulted in significant (P<0.05) reduction in IHL (Cohen’s d = -6.22), IMCL (Cohen’s d = -2.56) and body weight (Cohen’s d = -1.67), and significant increases in Matsuda’s index (Cohen’s d = 1.61).

Discussion and Conclusions: Our results suggest that the dietary fat content modulates the effect of long-term exercise on IHL and IMCL/Cr. The so called ‘athlete’s paradox’ of high insulin sensitivity despite elevated IMCL has been well reported [5]. We found that there are practically significant elevations in both IHL and IMCL/Cr under a low-fat background diet in an animal model of long-term exercise, despite a practically significant increase in insulin sensitivity. Under a high-fat background diet, long-term exercise resulted in a sharp reduction in IHL and IMCL compared to diet-matched sedentary control, which might be explained by the reduced mismatch between lipid oversupply and mitochondrial fat oxidative capacity [6, 7]. The effect of long-term exercise on body weight and insulin sensitivity does not seem to be dependent on the dietary fat content. There were practically significant reductions in body weight and elevation in insulin sensitivity under both low-fat and high-fat background diets. Our results indicate the amount of dietary fat content that needs to be factored in, while assessing the effect of exercise-induced mobilization of ectopic fat on insulin sensitivity.