Journal Logo


An Ethanolic Extract of Artemisia dracunculus L. Enhances the Metabolic Benefits of Exercise in Diet-induced Obese Mice


Author Information
Medicine & Science in Sports & Exercise: April 2021 - Volume 53 - Issue 4 - p 712-723
doi: 10.1249/MSS.0000000000002516



The authors of “Ethanolic Extract of Artemisia dracunculus L. Enhances the Metabolic Benefits of Exercise in Diet-induced Obese Mice” () report that in supplemental methods two of the units of measure are not correct. Specifically, the units for oxygen concentration should be μM (or nmols/mL) not nM. In addition, the units for the antimycin A concentration should also be μM not mM. (See Supplemental Digital Content, supplemental methods—amended, Note that these corrections do not affect the results presented.

Medicine & Science in Sports & Exercise. 53(12):2703, December 2021.

The worldwide prevalence of type 2 diabetes mellitus (T2DM) has grown from 4.7% in 1980 to a reported 8.5% as of 2018 (1). Furthermore, prediabetes accounts for 352.1 million (7.3%) of the adult population worldwide (2). The prevention and treatment of prediabetes and T2DM include pharmacotherapies (i.e., metformin) and recommended increased physical activity and exercise. Notably, the effect of these two interventions can prevent development from prediabetes to T2DM (2). Patients seeking to improve metabolic health often consume dietary or botanical supplements for their reported health benefits. An ethanolic extract of Artemisia dracunculus L. (Russian tarragon, termed 5011) has been studied extensively in preclinical models of obesity and prediabetes/T2DM. A 5011 extract improves skeletal muscle insulin signaling in vitro despite accumulated lipid species and the presence of inflammatory cytokines (3). Recently, 5011 was shown to enhance the switch from fatty acid oxidation to carbohydrate oxidation in skeletal muscle homogenates when exposed to palmitate indicating improved tissue-specific metabolic flexibility (4).

Acute exercise and exercise training have well-known beneficial effects on glucose (i.e., glucose disposal and insulin sensitivity) and fatty acid metabolism (i.e., increased fat oxidation) in those with prediabetes and T2DM (5–7). However, there is an emerging body of literature to suggest that when pharmacotherapies and or dietary supplements are combined with exercise, there is an attenuation in the cardiometabolic benefits of exercise (8,9). Studying the concurrent effects of exercise and botanical supplementation, with known independent beneficial effects, respectively, is important to understand the overall potential risks and benefits of additional dietary supplementation to exercise training. The purpose of this study was to assess the effect of 5011 combined with voluntary wheel running exercise on in vivo glucose and fat metabolism in a diet-induced obese animal model.


Animal housing and study design

Four-week-old male C57BL/6J mice (n = 52) were purchased (The Jackson Laboratory, Bar Harbor, ME) and group-housed (four mice per cage). All mice were immediately placed on a 45% high-fat diet (HFD; no. 12451, Research Diets Inc., New Brunswick, NJ) for 8 wk. After the 8-wk HFD only period, body composition was measured via nuclear magnetic resonance, and a stable isotope–labeled oral glucose tolerance test (SI-OGTT) was performed to provide an index of endogenous glucose production and glucose disposal. Mice were then randomly allocated to one of four groups: 1) HFD sedentary (HFD Sed), 2) HFD exercise (HFD Ex), 3) HFD + 5011 extract sedentary (5011 Sed), and HFD + 5011 extract exercise (5011 Ex) and individually housed. The HFD groups correspond to the same 45% HFD consumed during the initial 8 wk of the experiment. Animals were housed at room temperature (22°C) for the entire experimental period. The 5011 groups correspond to the 45% HFD combined with 1% w/w of 5011 that was incorporated into the food by Research Diets. Weekly food intake was assessed by weighing the content of the food hopper place in and at the end of the week accounting for any remaining food pellets and crumbs to determine the amount consumed (10). Exercise groups had free access to an in-cage running wheel, whereas sedentary groups remained in a standard cage without a running wheel. All animals were individually housed for the remaining 4 wk of the experiment. In total, the study lasted 12 wk (Fig. 1A).

Voluntary wheel running (exercise) reduces fat mass. A, The experimental design. C57BL/6J mice (4 wk old, n = 52) were started on a 45% HFD without access to a running wheel (Sed) for 8 wk. At the 8-wk (Pre) time point, NMR was conducted to measure body composition and mice underwent an SI-OGTT. Mice were then randomly allocated to one of four groups; HFD with no running wheel (HFD Sed, n = 13), HFD with running wheel access (HFD Ex, n = 13), HFD with 1% 5011 with no running wheel (5011 Sed, n = 13), or the 5011 HFD with running wheel access (5011 Ex, n = 13). At week 9 of the experiment, mice were placed in metabolic cages for 6 d. At week 12 (Post), NMR and SI-OGTT were repeated. Bodyweight over the 12-wk experiment (B). NMR was used to measure body fat percentage (C) and lean mass (D). NMR, nuclear magnetic resonance; 5011, HFD + 5011; Ex, exercise; Sed, sedentary. All data are expressed as mean ± SD. *P < 0.01, **P < 0.001, ***P < 0.001, ****P < 0.0001.

At the end of the 12-wk experiment (Post), mice were euthanized by isoflurane administration with cardiac puncture. A subgroup of mice (HFD Sed, n = 2; HFD Ex, n = 2; 5011 Sed, n = 2; 5011 Ex, n = 2) from all groups was IP injected with Human Insulin (Humulin; Eli Lilly, Indianapolis, IN) at a dose of 1.5 U·kg−1 10 min before euthanasia to examine the modulation of insulin signaling proteins. Pennington Biomedical Research Center Animal Care and Use Committee approved all study experiments.

Stable isotope glucose tolerance test

We used an SI-OGTT consisting of two complimentary deuterium glucose labels (2-[2H] and 6-6-[2H]; Cambridge Isotope, Tewksbury, MA) at a one-to-one ratio to provide an index of the dynamic changes in glucose disposal and patterns in endogenous glucose production (11,12). Mice were fasted with no access to running wheels for 4 h and were then gavaged with a 1:1 ratio of the glucose tracer at a total dose of 2 g·kg−1 lean body mass. Tail blood glucose was measured at 0, 15, 30, 45, 60, 90, 120 min with a glucometer (Bayer, Leverkusen, Germany). Fractional amounts of [2-2H]-glucose and [6,6-2H]-glucose tracer concentration was measured at 0, 15, 30, 60, 120 min via gas chromatography–mass spectrometry and matrix method to account for natural background isotopic skew. The glucose tracer concentration over time was calculated by using the measured plasma glucose (unlabeled) concentrations. The use of stable isotope–labeled glucose permits the partitioning of the blood glucose concentration into the glucose derived from endogenous (unlabeled) and exogenous gavage-derived (labeled [2H] and 6-6-[2H]) sources (12). Hepatic futile glucose cycling occurs frequently in T2DM, and rodents were fed a western diet (13). The isotopic position of [2-2H]-glucose can undergo recycling in the tricarboxylic acid cycle and be positioned on glucose leaving the hepatocyte. Hepatic futile glucose cycling can also be inspected by measuring the difference between [2-2H]-glucose area under the curve (AUC) and [6,6-2H]-glucose AUC (11,14). Insulin enzyme-linked immunoabsorbance assay (Crystal Chem, Downers Grove, IL) and nonesterified fatty acids (NEFA) enzymatic colorimetric assays (Wako Chemical, Richmond, VA) were performed on tail blood samples for 0- and 15-min time points during Pre and Post SI-OGTT.

Indirect calorimetry

At week 9, 1-wk postrandomization, mice were housed in metabolic cages (Promethion Metabolic Analyzer; Sable Systems International) at room temperature (20°C–22°C). Mice had ad libitum access to their respective diets (HFD or HFD + 5011) and water, which were continuously monitored. Sedentary (HFD Sed and 5011 Sed) mice had normal cages, and exercise groups (HFD Ex and 5011 Ex) had access to a running wheel. Continuous assessment of oxygen consumption and carbon dioxide production was monitored to calculate estimates of energy expenditure (EE) and RER. Mouse activity was monitored by beam breaks in the x, y, and z axes.

Wheel running EE and substrate utilization

Wheel running EE was calculated as previously reported (15). Briefly, net joules for each running bout was calculated by first subtracting the resting EE from the EE data trace and then integrating below the net EE peak corresponding to each wheel running bout. Only running bouts greater than 60 s in duration or longer were accepted for analysis. Running distance (m) and running speed (m·s−1) were organized into data bins with a minimum of 300 bouts per bin. Running wheel activity was also continuously monitored and analyzed specifically for running time, running speed, running EE, and running RER via a custom macro provided by Sable Systems.

Preparation of tissue extracts

At the time of necropsy, adipose tissues, gastrocnemius muscle, and liver were collected, weighed, and frozen. Mice were removed from wheel running cages at least 6 h from the time of necropsy. Triacylglycerol was measured in serum, muscle, and liver using a colorimetric assay (Eagle Biosciences, Amherst, NH). Liver and muscle glycogen were processed by enzymatic hydrolysis and measured via a colorimetric assay (Abcam, Cambridge, MA). Serum NEFA were quantified by the Acyl-CoA oxidase method (Wako Chemical). Serum insulin was measured by ELISA (Biovision, Milpitas, CA).

Protein and mRNA expression

Gastrocnemius and liver tissues were powdered using a frozen mortar and pestle and homogenized in a protein extraction buffer containing 150 mM NaCl, 10 mM Tris (pH 7.4), 1 mM EDTA, 1 mM EGTA, 0.5% IGEPAL CA-630, 1% Triton X-100, 2 mM PMSF, 2 μM pepstatin, 100 trypsin inhibitory mU of aprotinin, 20 μM leupeptin, 2 mM 1, 10-phenanthroline, 0.8 mM sodium orthovanadate, and 400 μM sodium fluoride. Tissue homogenates were then clarified using centrifugation at 13,000g for 10 min at 4°C and stored at −80°C until analysis. For RNA extraction, powdered tissues were homogenized in TRIzol (Life Technologies, Carlsbad, CA) followed by isolation using standard phenol–chloroform extractions. RNA was then isolated using the RNeasy Mini Kit (Qiagen, Hilden, Germany) and stored at −80°C until further analysis.


Protein concentrations from tissue homogenates were quantified using a BCA Kit for Protein Determination (Sigma-Aldrich, St. Louis, MO). A total of 50–100 μg of protein were loaded on 10% sodium dodecyl sulfate–polyacrylamide gels (National Diagnostics, Atlanta, GA) and transferred onto 0.45-μm nitrocellulose membranes (Bio-Rad, Hercules, CA). Membranes were blocked with 4% milk for 1 h at room temperature and probed with antibodies (see Table Supplemental Digital Content 1, List of antibodies and primers, overnight at 4°C. Proteins were detected using horseradish peroxidase-conjugated secondary antibodies (Jackson ImmunoResearch, West Grove, PA) and SuperSignal West Pico PLUS reagents (Thermo Fisher, Rockford, IL). Blots were imaged using x-ray film or the ChemiDoc XRS+ Imaging System (Bio-Rad). Protein signals were quantified in either Image Studio 5.2 (LI-COR Biosciences, Lincoln, NE) or Image Lab 4.0 (Bio-Rad).

Gene expression analysis

RNA concentrations were quantified using the NanoDrop ND-1000 UV-Vis Spectrophotometer. Reverse transcription was performed using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA) with 2 μg of RNA. Quantitative PCR was performed using SYBR Premix with ROX Plus (Takara Bio, Mountain City, CA) with 4 ng of cDNA and run on the Applied Biosystems 7900HT system with SDS 2.4 software (Applied Biosystems). Thermal cycling conditions were as follows: 2 min at 50°C, 10 min at 95°C, 40 cycles of 15 s at 95°C, and 1 min at 60°C; dissociation stage: 15 s at 95°C, 15 s at 60°C, and 15 s at 95°C. Relative expression of genes was interpolated using standard curves, and all target genes were normalized to the numerical average of peptidyl prolyl isomerase A and ubiquitin B.

Statistical analysis

Independent t-test and repeated-measures two-way and three-way ANOVA were used to analyze data where appropriate (Prism, Graphpad version 8.3). Multiple comparisons were made using the Holm–Sidak test. Diet effects indicated differences between normal HFD and 5011 + HFD mice. Exercise effects indicated differences between voluntary wheel running exercises (Ex) mice compared with nonrunning sedentary mice (Sed). ANCOVA was used to determine differences between groups in EE and substrate utilization (RER) for complete data set according to consensus statements regarding analysis of EE (16). Fat-free mass, fat mass, and total body mass were used as covariates. Time effects indicated differences between Pre (8 wk, prerandomization time point) and Post (12 wk, 4 wk postrandomization). Statistical significance was declared at P < 0.05. All data were reported as mean ± SD. Calculations of the incremental AUC were made by the trapezoidal method.


Body composition

We fed 52 4-wk-old C57BL/6J mice with HFD (45%) for 8 wk to generate a model of diet-induced obesity (DIO). At the end of the 8-wk DIO period, body composition was measured before randomization. Figure 1A depicts study design and experimental groups (HFD Sed, HFD Ex, 5011 Sed, and 5011 Ex) and the experimental schedule. Bodyweight (Fig. 1B), percent body fat (Fig. 1C), and percent fat-free mass (Fig. 1D) did not differ before randomization (week 8). After 4 wk of voluntary wheel running (Ex), both the HFD Ex and the 5011 HFD Ex mice had significant reductions in body weight by the end of the study (week 12) (Fig. 1B). The significant reduction in body mass from Pre to Post coincided with a reduced percent body fat (Fig. 1C) and increased percent fat-free mass (Fig. 1D). There were no differences between groups for weekly food intake at Pre and Post (data not shown).

EE and substrate oxidation

One week after randomization (week 9; 13 wk of age), mice were placed in metabolic cages with (HFD Ex and 5011 Ex) or without (HFD Sed and 5011 Sed) running wheels. We conducted these metabolic cage experiments 1 wk after randomization to minimize the weight alerting effects of wheel running. By doing so, we were able to evaluate early phase adaptions to Ex, the combined 5011 extract and HFD diet, and the combination of 5011 and Ex. Figure 2A demonstrates that both HFD Ex and 5011 Ex mice had significantly (P < 0.0001) increased dark cycle EE (kcal·h−1 per mouse) when compared with their Sed control (HFD Ex and 5011 Sed). There were the expected differences for RER when comparing the light versus dark cycle (Fig. 2B) with no difference between groups. Analysis of the 30-min period with the lowest EE (Fig. 2C, Rest) and corresponding RER (Fig. 2D) did not reveal any differences between groups. However, an analysis of the 15-min period with the greatest EE (Fig. 2E, active) did reveal a significant (P < 0.0001, Sed vs Ex) increase in EE (kcal·h−1 per mouse) for HFD Ex and 5011 Ex mice when compared with HFD Sed and 5011 Sed mice, respectively. When analyzing the same time running periods (Fig. 2F, active) for RER, both HFD Ex mice and 5011 Ex mice had lower RER than their sedentary controls. However, 5011 Sed (0.083 ± 0.02) mice experienced a reduction (P = 0.007) in RER during that active period when compared with HFD Sed mice (0.86 ± 0.02).

EE and fat oxidation are increased with exercise. Mice were placed in metabolic cages for 7 d 1 wk after randomization (weeks 9–10). Sedentary groups had normal cages, and wheel running groups had access to electronically monitored running wheels. Diets and water were provided ad libitum. EE (A) and RER (B) for total light and dark cycle. Data for the period of EE (rest period) were measured for EE (C) and RER (D). Conversely, periods of the highest measured activity (active period) are reported for corresponding EE (E) and RER (F). 5011, HFD + 5011; Ex, exercise; Sed, sedentary. All data are expressed as mean ± SD. n = 13 mice per group. *P < 0.01, **P < 0.001, ***P < 0.001, ****P < 0.0001. #Difference between groups at Post time point (HFD Sed vs HFD Ex, P < 0.001) (5011 Sed vs 5011 Ex, P < 0.05).

To assess potential differences between HFD Ex and 5011 Ex mice for the effects of wheel running, statistical analysis of EE and RER was performed on periods of continuous wheel running (no resting periods greater than 5 s) corresponding to distances between 15 and 100 m (Fig. 3A). Data were sorted into bins of distance and were matched between groups. Figure 3A shows a significant increase in the net energy (joules) expended over the distance traveled and as a function of running speed (Fig. 3B). As expected, there was a significant increase in net joules expended with increased distance (P < 0.0001) or speed increase (P < 0.0001). However, there were no differences in net joules expended at the binned distances or speeds between the two wheel running groups. A significant increase in net joules expended across all distances was detected by ANCOVA (Fig. 3C; HFD Ex vs 5011 Ex, P < 0.001). There was, however, no linear relationship between RER and distance (Fig. 3D; HFD Ex, P = 0.86; 5011 Ex, P = 0.34) or speed (Fig. 3E; HFD Ex, P = 0.11; 5011 Ex, P = 0.26). RER was significantly lower at speeds between 0.46 m·s−1 (P < 0.001) and 0.50 m·s−1 (P < 0.0001). The running speeds were recorded over a period of at least 300 bouts. Because of the lack of linearity between RER and most covariates, we plotted the RER during all wheel running bouts as a percent cumulative relative frequency. This allows for the visualization of the RER ordered from lowest to highest with the ability to identify the EC50 (average). ANCOVA analysis revealed a lower average RER (P < 0.0001) during all running periods for 5011 Ex mice when compared with HFD Ex mice (Fig. 3F).

5011 HFD shifts fuel preference to fat oxidation during extended bouts of wheel running. An analysis of wheel running mice (HFD Ex and 5011 Ex) was conducted by removing the resting EE for the data and then integrating below the net EE peak that corresponded to each wheel running bout. The final units are net joules expended. Net joules were expressed as a function distance (A) or speed (B). To compare HFD Ex and 5011, EE during wheel running ANCOVA was performed (C). The relationship between RER and distance (D) and speed (E) was also plotted. To visualize the ANCOVA results, the percent relative cumulative frequency was performed on RER during the wheel running period (F). 5011, HFD + 5011; Ex, exercise; Sed, sedentary. All data are expressed as mean ± SD. n = 11–12 per group. **P < 0.001, ****P < 0.0001.

Wheel running, food/water intake behavior, and energy balance

During the metabolic cage experiment, we analyzed total wheel running behavior as well as food intake to determine total energy and fat balance between groups. The 5011 Ex mice spent a greater percentage of time during the metabolic cage experiment wheel running (Fig. 4A; HFD Ex 18.0% ± 4.1% vs 5011 Ex 22.85% ± 1.9%, P = 0.02) as compared with HFD Ex mice. The total average distance ran each day was not different between groups (Fig. 4B; HFD Ex, 5923 ± 2704 m·d−1, vs 5011 Ex, 6145 ± 1359 m·d−1; P = 0.80). Non–wheel running ambulation (pedometers) was reduced in the HFD Ex mice (Fig. 4C; 124.7 ± 31.0 m·d−1, P = 0.002) when compared with HFD Sed mice (90.4 ± 16.8 m·d−1). Analysis of individual axis beam breaks revealed significant reduction (P = 0.02) for Ex mice (HFD Ex and 5011 Ex mice) when compared with sedentary controls in only the y-axis (Fig. 4D). Food intake (Fig. 4E, P = 0.028) and water intake (Fig. 4F, P = 0.0008) were significantly increased in 5011 Ex mice (0.47 ± 0.14 kcal·g−1 of body weight) when compared with 5011 Sed mice (0.355 ± 0.063 kcal·g−1 of body weight). There were no significant differences detected in the number of bouts of feeding (Fig. 4G; P = 0.24) or drinking (Fig. 4H; P = 0.25) measured between groups during the experiment. Energy balance (Fig. 4I; energy intake − EE) was significantly (P = 0.0008) lower in HFD Ex mice (−1.56 ± 1.62 kcal) versus HFD Sed mice (1.31 ± 1.49 kcal) with no differences in energy balance between 5011 Ex and 5011 Sed mice. Likewise, there was a significant reduction in fat balance (fat intake − fat oxidation) only in HFD Ex mice (Fig. 4J; 0.5 ± 0.1 g, P = 0.02), but with no significant reduction in 5011 Ex mice (Fig. 4J; 0.65 ± 0.1 g).

A combination of 5011 and exercise increase food and water intake during a metabolic cage experiment. Wheel running and food/water intake behaviors were measured while in the metabolic cages. The time spent (%) running in wheels (A) as well as the average daily running (m·d−1) (B) were recorded. All mice had normal ambulatory walking recorded (C) as well as spatial activity (D). The amount (F and H) and frequency of water and food intake (G and I) were monitored respectively. Energy balance (I) (EE − energy intake) and fat balance (J) (fat burned − fat consumed) were calculated for each mouse. 5011, HFD + 5011; Ex, exercise; Sed, sedentary. All data are expressed as mean ± SD. n = 11–12 per group. *P < 0.01, **P < 0.001, ***P < 0.001.

Stable isotope glucose tolerance test

We used an SI-OGTT to resolve to measure changes in endogenous glucose patterns and glucose disposal after the ingestion of glucose. The SI-OGTT was performed after the 8-wk DIO period (Pre) and after the 4-wk wheel running/5011 diet period (Post). Mice were fasted for 4 h and gavaged with a 1:1 ratio of 2-[2H] glucose and 6-6-[2H] glucose (2 g·kg−1). Blood was collected from the tail before (0 min, fasted) and after gavage 15, 30, 60, and 120 min. Blood glucose concentration was measured by glucometer, and serum was preserved for later stable isotope analysis by GC-MS. We first evaluated total blood glucose AUC (Fig. 5A) without consideration of tracer glucose. There were no differences in the Pre glucose AUC between groups. Both exercise groups (HFD Ex and 5011 Ex) had reduced glucose AUC when compared with their respective Pre measurements (time effect P < 0.0001). Likewise, at the Post time point, HFD Ex and 5011 Ex mice had lower glucose AUC than HFD Sed and 5011 Sed, respectively (activity effect, <0.0001, Sed vs Ex). A two-way repeated-measures ANOVA did not reveal any differences between groups for the glycemic excursion over the 120-min period for the Pre measurement (data not shown). However, there were significant activity effects (P = 0.001) for the Post measurement (Fig. 5B). At 0-min (fasted/preglucose gavage) HFD Ex (178 ± 12.1 mg·dL−1), blood glucose was lower than HFD Sed (218 ± 28.0 mg·dL−1; P = 0.02). By contrast, 5011 Ex mice (187.6 ± 30.5 mg·dL−1) demonstrated a positive trend for improved 0-min glucose (P = 0.07) when compared with 5011 Sed mice (222.2 ± 29.7 mg·dL−1). Peak glucose occurred at 15-min after glucose gavage for all groups. The 5011 Ex group demonstrated significantly lower peak glucose response (301.8 ± 36.7 mg·dL−1; P = 0.01) when compared with 5011 Sed mice (358.7 ± 30.5 mg·dL−1). The peak glucose response was not significantly (P = 0.12) different between HFD Ex (334.9 ± 30.5 mg·dL−1) and 5011 Ex mice according to the repeated-measures ANOVA. There were no differences between groups for fasted insulin measured during the glucose tolerance test at the Pre SI-OGTT measurement (data not shown). At the Post SI-OGTT, insulin concentration (Fig. 5C) measured at 0 min (fasted) and 15 min after glucose gavage showed an increased fasted insulin in HFD Sed mice (2.2 ± 1.2 ng·mL−1, P < 0.001) when compared HFD Ex mice (1.5 ± 1.0 ng·mL−1) and was nearly significant (P = 0.06) when compared with 5011 Sed mice (1.47 ± 0.8 ng·mL−1). There were no significant differences in the NEFA response (Fig. 5D) at 0- and 15-min time points take during the Post SI-OGTT. HOMA-IR (Fig. 5E) measured from fasted blood before necropsy revealed both HFD Ex and 5011 Ex had reduced (activity effect, P = 0.02) insulin resistance when compared with Sed groups.

Combined 5011 HFD and exercise reduces peak glucose during glucose tolerance test. Mice were fasted for 4 h before the stable isotope glucose tolerance test. Wheel running mice had the running wheel removed during the 4-h fast. Mice were gavage with a 1:1 mixture of 2-[2H] glucose and 6-6-[2H] glucose. Blood glucose was checked at 0 (fasted), 15, 30, 60, and 120 min after glucose gavage with a glucometer. Whole blood glucose (unlabeled) AUC (A) and glycemic excursion (unlabeled) (B, Post measurement) were plotted. Tail blood was collected to measure serum insulin (C) and NEFA (D). Before necropsy, fasted blood was used to measure HOMA-IR (E). 5011, HFD + 5011; Ex, exercise; Sed, sedentary; HOMA-IR, homeostatic model of insulin resistance. All data are expressed as mean ± SD. n = 13 per group. *P < 0.01, **P < 0.001, ****P < 0.0001.

There were no differences (other than time effects) for the Pre SI-OGTT between groups for endogenous glucose excursion and AUC (Fig. 6A and C), 2-[2H] glucose (Fig. 6D and F), and 6-6-[2H] glucose (Fig. 6G and I). We detected significant time effects (Pre vs Post) for HFD Ex (P < 0.0001) and 5011 Ex (P < 0.0001) to show reduced endogenous glucose AUC in response exercise, whereas endogenous glucose AUC was increased in 5011 Sed mice (Pre vs Post, P < 0.001). When comparing endogenous AUC at the Post SI-OGTT, HFD Ex and 5011 Ex mice had significant reductions when compared with HFD Sed and 5011 Sed mice, respectively (HFD Ex vs HFD Sed, P < 0.0001, and 5011 Ex vs 5011 Sed, P < 0.0001). There was a significant reduction in 2-[2H] glucose AUC (time effect, P < 0.001) for 5011 Sed and 5011 Ex mice only. Likewise, 6-6-[2H] glucose AUC was reduced only in the 5011 Sed and 5011 Ex mice. For both tracer AUC, there were no differences between 5011 Sed and 5011 Ex at Post SI-OGTT measurement.

Exercise with and without 5011 reduced endogenous glucose AUC, but 5011 is required to improve glucose disposal. Analysis of endogenous glucose (all tracer glucose − no tracer glucose), 2-[2H] glucose, and 6-6-[2H] glucose concentration was measured from the tail blood collected during the glucose tolerance test measured at Pre and Post. The endogenous glucose (A and B), 2-[2H] glucose (D and E) and 6-6-[2H] (G and H), and glucose glycemic excursion at Pre and Post P < 0.0001 were used to calculate endogenous and tracer AUC (C, F, and I), respectively. 5011, HFD + 5011; Ex, exercise; Sed, sedentary. All data are expressed as mean ± SD. n = 13 per group. *P < 0.01, **P < 0.001, ***P < 0.00.

Glycogen and triglycerides

Our previous reports indicate that male mice fed a diet supplemented with 5011 had a robust improvement in skeletal muscle metabolic flexibility and reduced ectopic lipid accumulation (4). In the current study, serum triglycerides (Fig. 7A) trended lower in HFD Ex mice when compared with HFD Sed mice (P = 0.07). Muscle triglycerides (Fig. 7B) were elevated in both HFD Ex and 5011 Ex mice (P = 0.04). Liver triglycerides (Fig. 7C) were significantly reduced in HFD Ex (P < 0.0001), 5011 Sed (P = 0.001), and 5011 Ex (P = 0.002) mice when compared with HFD Sed mice. Muscle glycogen trended (P = 0.10) to be increased as in HFD Ex and 5011 Ex mice (Fig. 7D). Liver glycogen (Fig. 7E) was reduced in wheel running mice when compared with sedentary mice (P = 0.04).

Exercise and 5011 reduce hepatic triglyceride accumulation. Serum triglycerides (A) were measured from the fasted mice at the time necropsy. Muscle samples (mixed gastrocnemius) and liver sections were snap frozen and powered to analyze triglycerides (B and C, respectively) and glycogen (D and E, respectively). 5011, HFD + 5011; Ex, exercise; Sed, sedentary. All data are expressed as mean ± SD. n = 8–9 per group. *P < 0.01, **P < 0.001.

Skeletal muscle and liver mRNA expression and protein content

A hallmark adaptation of exercise endurance training increased mitochondrial capacity and content (17). We measured gene expression of some early exercise response and mitochondrial genes to interrogate if 5011, alone or in combination with exercise, could alter skeletal muscle and liver adaptations. Skeletal muscle (gastrocnemius) gene expression of endothelial nitric oxide synthase was significantly increased (P = 0.03) by exercise with and without 5011 (Fig. 8A). Analysis of mitochondrial-associated genes (Fig. 8B) shows increased expression of malate dehydrogenase 1 (Mdh1) and citrate synthase (Cs) in HFD Ex (Mdh1, P < 0.0001; Cs, P < 0.001) and 5011 Ex (Mdh1, P < 0.001; Cs, P < 0.001) gastrocnemius muscle when compared with sedentary controls. Only the combination of 5011 and exercise increased the expression of Pcg1α (P < 0.05) and Tfam (P < 0.001) when compared with sedentary control (5011 Sed). Early exercise response gene analysis for the liver (Fig. 8C) revealed a significant increase in Foxo1 (HFD Ex vs HFD Sed, P = 0.02) and Pepck (HFD Ex vs HFD Sed, P < 0.05; 5011 Ex vs 5011 Sed, P < 0.05). However, Pdk4 expression was significantly reduced in gastrocnemius of exercising mice (HFD Ex and 5011 Ex) when compared with sedentary mice (HFD Sed and 5011 Sed). We did not detect any significant difference in mitochondrial-associated genes in the liver (Fig. 8D) between groups. Both in vivo and in vitro work has demonstrated enhanced insulin signaling in the skeletal muscle of insulin-resistant male mice supplemented with 5011 (3,18). In addition, exercise can improve skeletal muscle Akt signaling in some, but not all, cases (19). Therefore, we measured Akt phosphorylation with and without insulin stimulation in the mixed gastrocnemius muscle and liver. In addition, we examined phosphorylated AMPK and PDK4 levels in the muscle and liver as these proteins have also been shown to increase with exercise (20,21). The ratio of phosphorylated Akt to total Akt in the muscle and liver was increased with insulin, but not further elevated in HFD Ex, 5011 Sed, or 5011 Ex mice. The ratio of phosphorylated AMPK to total AMPK was not different between groups Sed or Ex for muscle or liver (see Figure, Supplemental Digital Content 2, Effect of exercise, 5011, and 5011 with exercise on expression of insulin signaling proteins in skeletal muscle and liver, Likewise, there were no significant differences in PDK4 protein abundance between groups for muscle or liver.

A combination of 5011 and exercise increases the expression of specific mitochondrial genes. Total RNA was isolated from mixed gastrocnemius and liver tissues and subjected to RT-qPCR to examine a panel of genes involved in early metabolic responses to exercise (A and C) and mitochondrial function (B and D). 5011, HFD + 5011; Ex, exercise; Sed, sedentary. All data are expressed as fold change vs sedentary controls (Sed CTLs) ± SEM. n = 8 per group. *P < 0.01, **P < 0.001, ***P < 0.001, ****P < 0.0001.


There is a significant body of literature demonstrating the attenuation of the benefits of exercise with antihyperglycemic medications or natural products (22–24). However, some agents have been shown to improve (or not interfere) with the benefits of exercise (25,26). In the current study, we designed an experiment to determine whether a supplemental ethanolic extract of A. dracunculus L. (5011) could enhance or attenuate exercise-mediated benefits on fat and glucose metabolism in a preclinical model of obesity and prediabetes. We observed that 5011 increased fat oxidation mice during voluntary wheel running when compared with a control HFD diet (Fig. 3). We used an SI-OGTT containing two complimentary deuterium glucose tracers (2-[2H] glucose and 6-6-[2H] glucose) to differentiate between endogenous glucose patterns and whole-body glucose disposal. We provide evidence that when combined with exercise, 5011 improves glucose disposal to a greater degree than exercise alone (Fig. 6). Despite improved glucose disposal in the 5011 Ex group, there was an equivalent reduction in total blood glucose AUC between both exercise groups (Fig. 5A; HFD Ex and 5011 Ex). The lack of difference in total blood glucose difference between these groups underlies the importance of an integrative metabolic response between the liver and the skeletal muscle to balance glucose output and disposal to achieve glycemic control.

Previous work from our group showed no effect of 5011 to reduce body fat while consuming an HFD (45%) (4). Similarly, we did not observe any reduction in body fat in mice consuming the 5011 supplemented HFD (5011 Sed) compared with nonsupplemented diet. However, exercise reduced both body weight and adiposity that was not attenuated in the 5011 supplemented diet (5011 Ex, Fig. 1). Mice were housed at room temperature in metabolic cages 1 wk after randomization (week 9), and we performed our metabolic cage experiments early in the exercise intervention and supplemental 5011 phase (1 wk after randomization) to avoid the confounding influence of changes in body weight and adiposity shown at the end of the study. By doing so, we were able to detect early adaptions to voluntary wheel running and 5011 to increase fat oxidation. We used methods similar to Lark et al. (15) to isolate the metabolic effects of the wheel running. From that data, we used rigorous statistical approaches (ANCOVA and percent relative cumulative frequency) that compared EE and RER between HFD Ex and 5011 Ex mice during distance- and speed-matched wheel running (Fig. 3). We only analyzed running bouts that exceeded 60 s to acquire stable metabolic measurements. Furthermore, we only included bins that included at least 300 bouts. Bouts of running that exceeded approximately 100 m were not included in our binned analysis because of low bout attempts. We also observed a similar phenomenon for running speed, whereby the total bouts exceeding 0.50 m·s−1 were below our accepted threshold. The net EE during wheel running is relatively equivalent between HFD Ex and 5011 Ex mice across all binned distances (Fig. 3A and B). However, there was a trend for higher EE at lower absolute work rates (0.42 to 0.46 m·s−1) in the 5011 Ex mice. Conversely, we noticed the opposite trend (lower EE) at higher work rates (0.48 to 0.50 m·s−1). The ANCOVA of the entire data set revealed that wheel running EE was increased in 5011 Ex mice when compared with HFD Ex mice (Fig. 3C). This could be explained by the majority of wheel running bouts (~75%) occurring between speeds of 0.42 and 0.46 m·s−1. We used identical methods to determine differences in substrate utilization (RER) during wheel running (Figs. 3D and 3E). As the work rate increased (0.46 to 0.50 m·s−1), we detected significantly reduced RER in 5011 Ex mice compared with HFD Ex mice, suggesting a shift to greater fat oxidation. We visualized our ANCOVA RER data by percent relative cumulative frequency (Fig. 3F) (27). This analysis demonstrates a significant shift toward fat oxidation in 5011 Ex mice when accounting for distance and running speed. To our knowledge, this the first report to demonstrate differences in fat oxidation during voluntary wheel running with any dietary supplement, botanical or otherwise.

Other wheel running, activity, feeding, and drinking behaviors were monitored during the metabolic cage experiment period (Fig. 4). We noticed a general trend for reduced non–wheel running activity in HFD Ex and 5011 Ex mice, which is typical when mice are provided a running wheel (15). Interestingly, 5011 Ex mice consumed more food (Fig. 4F) and water (Fig. 4H) during this phase of the experiment. This was directly reflected by the negative energy balance displayed by the HFD Ex mice, but not the 5011 Ex mice (Fig. 4I). This was also mirrored by the reduction in fat balance only observed in HFD Ex mice and not the 5011 Ex mice. These data suggest that in response to wheel running, 5011 Ex mice had a heightened motivation to consume food and return to energy balance. The lower RER observed in 5011 Ex mice during wheel running (Fig. 3D–F) may potentially trigger the intrinsic drive to replace fat calories lost during running.

We used a stable isotope glucose tolerance test to resolve the differences in endogenous glucose versus glucose disposal. We have previously demonstrated that endogenous glucose changes after a 4-wk period of wheel running were responsible for the improvement in glucose tolerance (11). In general, wheel running reduced (Pre vs Post) endogenous glucose (Fig. 6A–C). Endogenous glucose was increased in 5011 Sed mice when comparing the Pre versus Post time points (4 wk). Interestingly, only the 5011 Sed and 5011 Ex mice experienced improved glucose disposal (tracer glucose AUC; Figs. 6D–I). We did not detect significantly greater serum insulin levels at 15 min after glucose gavage (peak glucose) in the 5011 Sed mice when compared with HFD Sed mice, ruling out the possibility of hyperinsulinemia driving enhanced glucose disposal. Only the 5011 Ex mice had enhanced glucose disposal (reduced 2-[2H] glucose and 6-6-[2H] glucose AUC), reduced endogenous glucose (Fig. 6C), and lower peak (15-min) total glucose concentration when comparing Pre versus Post time points (Fig. 5B). These effects were not dependent on increased insulin concentration measured at 0 or 15 min (Fig. 5C). Therefore, 5011 appears to improve glucose disposal and endogenous glucose, whereas exercise alone (HFD Ex) only improved endogenous glucose AUC. The use of 2-[2H]-glucose, in combination with 6-6-[2H]-glucose, provides an index of hepatic futile cycling (12). However, we did not detect any overt hepatic futile cycling in our mice (data not shown) in agreement with our previous work using the same diet and conditions (11). Mice fed a western diet (high fat/high sucrose) were more susceptible to hepatic futile cycling, suggesting that this may be a feature of more severe metabolic dysfunction than we have shown in the current study (12,14).

A major obesity-related risk factor is the accumulation of ectopic lipids in the skeletal muscle and liver (28). Dysregulated fatty acid metabolism is thought to promote the partitioning of lipids toward storage rather than oxidation (29). However, a consistent adaptation for aerobic exercise is increased skeletal muscle lipid accretion known as the “athlete’s paradox” (30). This exercise-mediated adaptation is positively associated with improved insulin sensitivity after aerobic exercise training. We detected a significant increase in muscle triglyceride (Fig. 7B) content in both HFD Ex and 5011 Ex mice while demonstrating improved metabolic health. These data would suggest that the skeletal muscle adaptation from exercise to store lipid is not abolished when consuming 5011. Liver triglycerides (Fig. 7C) were markedly reduced in HFD Ex, 5011 Ex, and 5011 Sed mice when compared with HFD Sed mice. We have previously shown that 5011 reduces liver triglyceride content after 3 months of feeding (4). Recent evidence demonstrates that glycerol (from the breakdown of triglycerides) represents a major carbon source for gluconeogenesis, which may explain the reduction in liver triglycerides and increased hepatic glucose production in 5011 Sed mice (31–33).

Both exercise groups showed slight improvements in skeletal muscle glycogen levels, but no differences were detected between HFD Ex and 5011 Ex mice. Liver glycogen levels were substantially reduced in both exercise groups. Liver glycogen is an important source of glucose during exercise in mice, whereas muscle glycogen is less crucial (34). Exercise trials in humans have shown improved metabolic flexibility, resulting from exercise, which was mediated primarily by a greater capacity to store glucose as muscle glycogen (35). The degree of muscle glycogen depletion during exercise drives muscle carbohydrate storage into the skeletal muscle (36). Whether liver glycogen depletion during exercise dictates this response in mice is unknown. However, our data suggest that liver glycogen reduction is associated with a positive metabolic response.

Exercise training notably promotes mitochondrial biogenesis and the expression of key tricarboxylic acid cycle enzymes (17). However, the combination of metformin or resveratrol with aerobic exercise training has been frequently reported to attenuate exercise-induced mitochondrial adaptations (37,38). We assayed skeletal muscle expression of genes associated with mitochondrial function, biogenesis, and respiratory capacity (see Figure, Supplemental Digital Content 3, OXPHOS permeabilized skeletal muscle fibers, We did not detect any attenuation in the expression of mitochondrial genes or respiratory function that was elevated in HFD Ex mice. Rather, we observed significant increases in Pgc1-α, citrate synthase, and Tfam when 5011 was combined with exercise. Previous work from our group showed that 5011 diet alone increased malate dehydrogenase 1 enzyme expression (Mdh1) while decreasing muscle triglyceride content in male mice after 3 months (4). Although we did not detect any 5011 diet effects on Mdh1 in the current study (4 wk of supplementation), we did observe that Mdh1 and Mdh2 levels were robustly increased in HFD Ex and 5011 Ex mice.

A major limitation of this study is the lack of inclusion of female mice. Our previous work demonstrates increased adiposity and hepatic glucose output (via pyruvate tolerance test) in female mice consuming the same 5011 HFD (45%) for 12 wk (39). Furthermore, female mice also demonstrate differences in responses to exercise when compared with male mice (40). Developing an understanding of sexual dimorphism in the response to combined exercise training with dietary supplementation is a necessary endeavor considering that women represent the largest consumer group of botanical dietary supplements (41). Despite these limitations, our study used two high-resolution, state-of-the-art techniques to determine the influence of supplemental 5011 on the beneficial metabolic effects of exercise. First, we provided evidence that 5011 can increase fat oxidation during voluntary wheel running. The data are further supported by the increase in mitochondrial gene expression in 5011 Ex mice, in addition to our previously reported enhanced ex vivo skeletal muscle fat oxidation (4). We also demonstrated improved glucose disposal in 5011 Ex mice after 4 wk of voluntary wheel running measured by the reduction in 6-6-[2H]-glucose AUC. Although 5011 Sed mice did not show any improvement in total blood glucose AUC after 4 wk of supplementation, we did detect improved glucose disposal. However, at the same time, there was an increase in endogenous glucose levels during the SI-OGTT that counteracted the changes in total glucose levels. An explanation for the inability to suppress endogenous glucose during the SI-OGTT was not revealed in our analysis but deserves future exploration.

A potential consideration that must be considered in our interpretation of the data is animal housing temperature and timing experiments. Our experiments were conducted at room temperature. Recent reports demonstrate that the metabolic benefits of exercise (i.e., glucose tolerance) are diminished when mice exercise in their thermo neutral zone (29°C–30°C) (42,43). The activity of rodent brown adipose tissue is speculated to contribute these findings (44). However, there are some discrepancies regarding the activity of brown adipose tissue and overall metabolic adaptations to mice exercising at room temperature or thermoneutrality that are likely related to sex differences between studies. In the current study, a thorough examination of adipose tissue was not conducted. In agreement with previous work conducted at room temperature, we did see a reduction in the brown adipose tissue mass (data not shown) in the exercising mice (HFD Ex and 5011 Ex). Interestingly, mice housed at thermoneutrality have increased brown adipose tissue mass with altered thermogenic and metabolic capacity that is considered to more accurately reflect human physiology (42,44). In addition, the duration of time between the final bout of wheel running and the measurement of glucose tolerance should be considered. Although previous studies using intense, forced-exercise models have shown that insulin action and glucose metabolites (45,46) are returned to resting levels within 4 h postexercise, we cannot exclude the possibility that some residual effects of previous voluntary wheel running persisted. Therefore, housing temperature and residual effects of acute wheel running should be considered when interpreting the translational effect of our findings.

In summary, we have shown that 5011 enhances the beneficial effects after 4 wk of voluntary wheel running to promote fat oxidation and glucose disposal in DIO mice. Cardiorespiratory fitness is a very strong and independent predictor of longevity (47) and, therefore, should be a focus of lifestyle modification. Future efforts should be made to identify dietary supplements or medications that enhance, or at least do not interfere with, the health-promoting effects of exercise. Studying the combination of exercise and dietary supplementation may also provide an opportunity to elucidate the mechanisms of action of dietary supplements or natural compounds according to their interactions with exercise training.

This publication was supported by the NCCIH and the Office of Dietary Supplements of the National Institutes of Health under Award Number P50AT002776. T. A. is supported by NCCIH T32 AT004094. This project/work used facilities within the Animal Metabolism and Behavior Core, Genomics Core, and Cell Biology and Bioimaging Core at PBRC that are supported in part by NIH center COBRE (P30GM118430) and NORC (P30DK072476) centers as well as an NIH equipment award S10OD023703. The authors also thank Tamra Mendoza for assisting in animal experiments.

The authors declare that they have no conflict of interest. The results of this study are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation. The authors report that result of the present study do not constitute endorsement by the American College of Sports Medicine.

T. D. A. and J. M. S. designed the experiments. T. D. A., J. S., B. A. I., and J. R. B. L, conducted the experiments and analyzed the data. T. A. wrote the manuscript, and G. M. K., Z. E. F., J. R. B. L., and J. M. S. edited the final version.


1. Kaiser AB, Zhang N, DER PLUIJM WVAN. Global prevalence of type 2 diabetes over the next ten years (2018–2028). Diabetes. 2018;67(1 Suppl):202-LB.
2. Hostalek U. Global epidemiology of prediabetes—present and future perspectives. Clin Diabetes Endocrinol. 2019;5(1):5.
3. Vandanmagsar B, Haynie KR, Wicks SE, et al. Artemisia dracunculus L. extract ameliorates insulin sensitivity by attenuating inflammatory signalling in human skeletal muscle culture. Diabetes, Obes Metab. 2014;16:728–38.
4. Yu Y, Mendoza TM, Ribnicky DM, et al. An extract of Russian tarragon prevents obesity-related ectopic lipid accumulation. Mol Nutr Food Res. 2018;62:e1700856.
5. Van Tienen FH, Praet SF, De Feyter HM, et al. Physical activity is the key determinant of skeletal muscle mitochondrial function in type 2 diabetes. J Clin Endocrinol Metab. 2012;97:3261–9.
6. Nyholm B, Mengel A, Nielsen S, et al. Insulin resistance in relatives of NIDDM patients: the role of physical fitness and muscle metabolism. Diabetologia. 1996;39:813–22.
7. Meex RC, Schrauwen-Hinderling VB, Moonen-Kornips E, et al. Restoration of muscle mitochondrial function and metabolic flexibility in type 2 diabetes by exercise training is paralleled by increased myocellular fat storage and improved insulin sensitivity. Diabetes. 2010;59(3):572–9.
8. Ristow M, Zarse K, Oberbach A, et al. Antioxidants prevent health-promoting effects of physical exercise in humans. Proc Natl Acad Sci U S A. 2009;106(21):8665–70.
9. Malin SK, Gerber R, Chipkin SR, Braun B. Independent and combined effects of exercise training and metformin on insulin sensitivity in individuals with prediabetes. Diabetes Care. 2012;35:131–6.
10. Yang Y, Smith DL Jr, Keating KD, Allison DB, Nagy TR. Variations in body weight, food intake and body composition after long-term high-fat diet feeding in C57BL/6J mice. Obesity (Silver Spring). 2014;22(10):2147–55.
11. Allerton TD, Kowalski G, Hang H, Stephens J. Dynamic glucose disposal is driven by reduced endogenous glucose production in response to voluntary wheel running: a stable isotope approach. Am J Physiol Endocrinol Metab. 2020;ajpendo.00450.2019.
12. Kowalski GM, Hamley S, Selathurai A, et al. Reversing diet-induced metabolic dysregulation by diet switching leads to altered hepatic de novo lipogenesis and glycerolipid synthesis. Sci Rep. 2016;6:27541.
13. Efendic S, Wajngot A, Vranic M. Increased activity of the glucose cycle in the liver: early characteristic of type 2 diabetes. Proc Natl Acad Sci U S A. 1985;82(9):2965–9.
14. Kowalski GM, Kraakman MJ, Mason SA, Murphy AJ, Bruce CR. Resolution of glucose intolerance in long-term high-fat, high-sucrose-fed mice. J Endocrinol. 2017;233(3):269–79.
15. Lark DS, Kwan JR, McClatchey PM, et al. Reduced nonexercise activity attenuates negative energy balance in mice engaged in voluntary exercise. Diabetes. 2018;67:831–40;db171293.
16. Tschöp MH, Speakman JR, Arch JR, et al. A guide to analysis of mouse energy metabolism. Nat Methods. 2012;9(1):57–63.
17. Holloszy JO, Oscai LB, Don IJ, Molé PA. Mitochondrial citric acid cycle and related enzymes: adaptive response to exercise. Biochem Biophys Res Commun. 1970;40(6):1368–73.
18. Wang ZQ, Ribnicky D, Zhang XH, Raskin I, Yu Y, Cefalu WT. Bioactives of Artemisia dracunculus L enhance cellular insulin signaling in primary human skeletal muscle culture. Metabolism. 2008;57:S58–64.
19. Sakamoto K, Aschenbach WG, Hirshman MF, Goodyear LJ. Akt signaling in skeletal muscle: regulation by exercise and passive stretch. Am J Physiol Endocrinol Metab. 2003;285(5):E1081–8.
20. Chen ZP, Stephens TJ, Murthy S, et al. Effect of exercise intensity on skeletal muscle AMPK signaling in humans. Diabetes. 2003;52:2205–12.
21. Kurth-Kraczek EJ, Hirshman MF, Goodyear LJ, Winder WW. 5’ AMP-activated protein kinase activation causes GLUT4 translocation in skeletal muscle. Diabetes. 1999;48:1667–71.
22. Gliemann L, Schmidt JF, Olesen J, et al. Resveratrol blunts the positive effects of exercise training on cardiovascular health in aged men. J Physiol. 2013;591(20):5047–59.
23. Nathan DM, Barret-Connor E, Crandall JP, et al. Long-term effects of lifestyle intervention or metformin on diabetes development and microvascular complications: the DPP outcomes study. Lancet Diabetes Endocrinol. 2016;3(11):866–75.
24. Malin SK, Nightingale J, Choi S-E, Chipkin SR, Braun B. Metformin modifies the exercise training effects on risk factors for cardiovascular disease in impaired glucose tolerant adults. Obesity (Silver Spring). 2013;21(1):93–100.
25. Hällsten K, Virtanen KA, Lönnqvist F, et al. Rosiglitazone but not metformin enhances insulin- and exercise-stimulated skeletal muscle glucose uptake in patients with newly diagnosed type 2 diabetes. Diabetes. 2002;51(12):3479–85.
26. Regensteiner JG, Bauer TA, Reusch JE. Rosiglitazone improves exercise capacity in individuals with type 2 diabetes. Diabetes Care. 2005;28(12):2877–83.
27. Riachi M, Himms-Hagen J, Harper M-E. Percent relative cumulative frequency analysis in indirect calorimetry: application to studies of transgenic mice. Can J Physiol Pharmacol. 2004;82(12):1075–83.
28. Greco AV, Mingrone G, Giancaterini A, et al. Insulin resistance in morbid obesity: reversal with intramyocellular fat depletion. Diabetes. 2002;51(1):144–51.
29. Koves TR, Ussher JR, Noland RC, et al. Mitochondrial overload and incomplete fatty acid oxidation contribute to skeletal muscle insulin resistance. Cell Metab. 2008;7(1):45–56.
30. Goodpaster BH, He J, Watkins S, Kelley DE. Skeletal muscle lipid content and insulin resistance: evidence for a paradox in endurance-trained athletes. J Clin Endocrinol Metab. 2001;86:5755–61.
31. Perry RJ, Zhang D, Guerra MT, et al. Glucagon stimulates gluconeogenesis by INSP3R1-mediated hepatic lipolysis. Nature. 2020;579(7798):279–83.
32. Wang Y, Kwon H, Su X, Wondisford FE. Glycerol not lactate is the major net carbon source for gluconeogenesis in mice during both short and prolonged fasting. Mol Metab. 2020;31:36–44.
33. Hasenour CM, Rahim M, Young JD. In vivo estimates of liver metabolic flux assessed by 13C-propionate and 13C-lactate are impacted by tracer recycling and equilibrium assumptions. Cell Rep. 2020;32(5):107986.
34. Pederson BA, Cope CR, Schroeder JM, et al. Exercise capacity of mice genetically lacking muscle glycogen synthase: in mice, muscle glycogen is not essential for exercise. J Biol Chem. 2005;280(17):17260–5.
35. Malin SK, Haus JM, Solomon TP, Blaszczak A, Kashyap SR, Kirwan JP. Insulin sensitivity and metabolic flexibility following exercise training among different obese insulin-resistant phenotypes. Am J Physiol Endocrinol Metab. 2013;305(10):E1292–8.
36. Richter EA, Garetto LP, Goodman MN, Ruderman NB. Muscle glucose metabolism following exercise in the rat: increased sensitivity to insulin. J Clin Invest. 1982;69(4):785–93.
37. Boulé NG, Robert C, Bell GJ, et al. Metformin and exercise in type 2 diabetes: examining treatment modality interactions. Diabetes Care. 2011;34:1469–74.
38. Polley KR, Jenkins N, O’Connor P, McCully K. Influence of exercise training with resveratrol supplementation on skeletal muscle mitochondrial capacity. Appl Physiol Nutr Metab. 2016;41:26–32.
39. Fuller S, Yu Y, Mendoza T, Ribnicky DM, Cefalu WT, Floyd ZE. Potential adverse effects of botanical supplementation in high-fat-fed female mice. Biol Sex Differ. 2018;9(1):41.
40. Lightfoot JT. Sex hormones’ regulation of rodent physical activity: a review. Int J Biol Sci. 2008;4(3):126–32. [cited 2020 Jun 24 ] Available from: /pmc/articles/PMC2359866/?report=abstract.
41. Mahady GB, Parrot J, Lee C, Yun GS, Dan A. Botanical dietary supplement use in peri- and postmenopausal women. Menopause. 2003;10(1):65–72.
42. Raun SH, Henriquez-Olguín C, Karavaeva I, et al. Housing temperature influences exercise training adaptations in mice. Nat Commun. 2020;11(1):1560.
43. McKie GL, Medak KD, Knuth CM, et al. Housing temperature affects the acute and chronic metabolic adaptations to exercise in mice. J Physiol. 2019;597(17):4581–600.
44. Aldiss P, Lewis JE, Lupini I, et al. Exercise training in obese rats does not induce browning at thermoneutrality and induces a muscle-like signature in brown adipose tissue. Front Endocrinol (Lausanne). 2020;11:97.
45. Cartee GD, Young DA, Sleeper MD, Zierath J, Wallberg-Henriksson H, Holloszy JO. Prolonged increase in insulin-stimulated glucose transport in muscle after exercise. Am J Physiol. 1989;256(4 Pt 1):E494–9.
46. Knudsen JG, Biensø RS, Hassing HA, Jakobsen AH, Pilegaard H. Exercise-induced regulation of key factors in substrate choice and gluconeogenesis in mouse liver. Mol Cell Biochem. 2015;403:209–17.
47. Kodama S, Saito K, Tanaka S, et al. Cardiorespiratory fitness as a quantitative predictor of all-cause mortality and cardiovascular events in healthy men and women: a meta-analysis. JAMA. 2009;301(19):2024–35.


Supplemental Digital Content

Copyright © 2020 by the American College of Sports Medicine