Abstract
Despite the role of branched-chain amino acids (BCAAs) in physiological processes such as nutrient signaling and protein synthesis, there is ongoing debate about the link between circulating BCAAs and insulin resistance (IR) in various populations. In healthy women, IR mildly increases during pregnancy, whereas both BCAAs and markers of BCAA catabolism decrease, indicating that fetal growth is being prioritized. Exercise reduces IR in nonpregnant adults, but less is known about the effect of exercise during pregnancy in women with obesity on IR and BCAA breakdown. The aim of this study was to determine the effect of a moderate-intensity exercise intervention during pregnancy on maternal circulating BCAAs and markers of BCAA catabolism [short-chain acylcarnitines (ACs)], and their associations with IR. Healthy obese [n = 80, means ± SD; body mass index (BMI): 36.9 ± 5.7 kg/m2] pregnant women were randomized into an exercise (n = 40, aerobic/resistance 3×/wk, ∼13th gestation week until birth) or a nonexercise control (n = 40) group. Blood was collected at 12.2 ± 0.5 and 36.0 ± 0.4 gestation weeks and analyzed for BCAA-derived acylcarnitine concentrations as markers of BCAA breakdown toward oxidative pathways, and glucose and insulin concentrations [updated homeostatic model assessment of IR (HOMA2-IR)]. After adjusting for HOMA2-IR, there were no interaction effects of group by time. In addition, there was a main positive effect of time on HOMA2-IR (12 wk: 2.3 ± 0.2, 36 wk: 3.0 ± 0.2, P = 0.003). A moderate-intensity exercise intervention during pregnancy in women with obesity was not associated with changes in BCAA-derived ACs versus standard of care. The decrease in BCAA-derived ACs throughout gestation could not be explained by IR.
NEW & NOTEWORTHY This research showed an increase in insulin resistance (IR) and decrease in branched-chain amino acid catabolism throughout gestation in women with obesity, and addition of a moderate exercise intervention (known to attenuate IR in nonpregnant populations) did not alter these shifts. Findings provide support for metabolic safety of exercise during pregnancy.
Keywords: BCAA, gestation, insulin resistance, obesity, physical activity
INTRODUCTION
Normal metabolic shifts in metabolism occur throughout gestation, facilitating the shunting of nutrients toward the developing fetus, and thus optimizing fetal tissue growth. For example, insulin resistance (IR) mildly increases within normal limits throughout pregnancy in healthy women, promoting the redirection of substrates from maternal tissues toward growing fetal tissues. Furthermore, cross-sectional analysis of IR in late pregnancy has been shown to positively relate to circulating concentrations of branched-chain amino acids (BCAAs; leucine, isoleucine, and valine) and markers of BCAA breakdown toward oxidation [e.g., short-chain acylcarnitines (ACs)] (1). Furthermore, maternal BMI is positively related to circulating concentrations of BCAA and related short-chain ACs (1).
Despite there being clear cross-sectional relationships between IR, BCAA, and BCAA-related metabolites, when considering the longitudinal changes in these metabolites (e.g., early to late pregnancy), the data are sparse. Upon comparison of the changes in IR and the changes in circulating BCAA concentrations throughout gestation, our group recently determined that the changes in IR and circulating BCAAs are significantly negatively associated, wherein IR either increases or stays the same, whereas circulating BCAAs decrease throughout gestation (2). Thus, although these findings do not support a cause-and-effect relationship between circulating BCAA and IR in pregnancy, a relationship is nonetheless present and important to understand, considering the imperative role of BCAAs on tissue growth during pregnancy.
Quantification of circulating short-chain BCAA-derived ACs provide a less invasive (e.g., no biopsy needed) way of measuring BCAA catabolism throughout the metabolic chain, which ultimately leads to BCAA oxidation. BCAA-derived ACs are produced through a series of metabolic reactions, beginning with the breakdown of circulating BCAAs that ultimately produce short-chain coenzyme A metabolites (CoAs) within the mitochondria. In situations of IR, BCAA flux through catabolic pathways increases (3) (e.g., more BCAAs enter into the mitochondria to be oxidized because of increased supply). However, when the metabolic machinery responsible for further breakdown cannot adequately keep up with the increased supply, ACs are formed from the CoAs by mitochondrial enzymes (e.g., carnitine acyltransferases) (4), which can then be trafficked to the blood. C3 ACs are by-products of valine and isoleucine catabolism [e.g., propionylcarnitine (C3) (5)], and C5 ACs are by-products of leucine and isoleucine catabolism [e.g., isovalerylcarnitine (C5), 3-hydroxyisovalerylcarnitine (C5-OH), 2-methylbutyrylcarnitine (MBC) (5)].
Exercise in nonpregnant (6–8) and pregnant (9) populations reduces IR. With the notion that IR is positively related to BCAA flux through catabolic pathways (3), this should mean that exercise has the capacity to also reduce the elevated BCAA flux (above normal rates) through the system and therefore reduce the production of BCAA-derived ACs. Nonetheless, the impact of exercise on the relationship between BCAA-derived ACs and IR throughout gestation is unknown. This information is imperative to determine because the increase in IR and resultant decrease in markers of BCAA breakdown throughout gestation allow for the shifting of the flux of nutrients away from maternal tissues and toward growing fetal tissues.
Therefore, the purpose of this study was to determine the effect of an exercise intervention during pregnancy in healthy women with obesity on circulating BCAA-derived ACs, and if IR is related to these changes. We hypothesized that the exercise intervention would result in a greater decrease in BCAA-derived ACs (indicating less breakdown toward oxidative pathways) throughout gestation compared with a standard of care (nonexercise) group. Furthermore, although it could be expected that exercise would not completely attenuate the increase in IR throughout gestation, we expected that exercise would attenuate IR to a greater extent than the nonexercise group and would be related to the change in BCAA-derived ACs.
MATERIALS AND METHODS
Participants
This study made use of a subset of participants within an ongoing randomized controlled trial called Expecting (ClinicalTrials.gov Identifier: NCT02125149), conducted at the Arkansas Children’s Nutrition Center (ACNC). Expecting was established to investigate the impact of an exercise intervention during pregnancy in sedentary women with obesity on pregnancy and postnatal outcomes in the mother and the offspring. Participants were recruited in the first 12 wk of pregnancy and were included if they met the following criteria: prepregnancy body mass index (BMI) ≥30.0 kg/m2, singleton pregnancy, ≥18 yr of age, sedentary (defined as individuals who do not engage in purposeful physical activity and who have a sedentary work activity level; this information was attained during phone screening), and conception without assisted fertility treatments. Participants were excluded if they had preexisting medical conditions (e.g., diabetes, hypertension), if they used recreational drugs, tobacco or alcohol during pregnancy (self-reported), and if they had any contraindications to exercise during pregnancy as determined by their physician. Beyond these criteria, specific medications were not listed as part of the exclusion criteria, and women were able to continue in the study if they were diagnosed with gestational diabetes mellitus (GDM) during the intervention.
Upon enrollment, participants were randomized to one of two groups: an exercise intervention group and a standard of care control group. The standard of care group received their standard of care through their OB/GYN clinic. Optional pregnancy and child development education classes were offered monthly at the ACNC to provide engagement and attention. During the first visit, physical activity recommendations for pregnancy were provided. Participants in both groups were asked to maintain their normal dietary habits and patterns. Participants were required to provide a signed release from their OB/GYN or primary care physician/nurse practitioner before participation. All experimental procedures were conducted according to the Declaration of Helsinki principles and were approved by the Institutional Review Board at the University of Arkansas for Medical Sciences. Written informed consent was provided by each participant.
Study Design
The current analyses assessed blood samples obtained from two study visits within the parent study: first trimester before the intervention (visit 1, 12.2 ± 0.5 wk gestation) and third trimester of pregnancy (visit 2, 36.0 ± 0.4 wk gestation). Visit 2 was the last assessment of the mother within the parent study; however, participants continued to exercise even after this visit (ideally until birth). The current study did not add any visits to the parent study.
Exercise Training
The exercise prescription adhered to the current recommendations from the US Department of Health and Human Services Physical Activity Guidelines for Americans, which recommend at least 150 min of moderate-intensity aerobic activity per week during pregnancy (10). The exercise program started at ∼13 wk of pregnancy, and the periodization was tailored such that there was an increase in both exercise duration (15 min/bout to 30 min/bout) and intensity during the first 6 wk of training to allow sedentary women to progressively become more active, and then maintenance of exercise duration and intensity until birth, if possible. If needed (e.g., doctor recommendation or discomfort), exercise session duration, intensity, and weekly frequency could be varied. All exercise sessions were supervised by a qualified exercise trainer.
Participants visited the ACNC Fitness Center three times per week and performed 30–45 min of aerobic and resistance exercise, as well as flexibility exercises. The aerobic exercise was performed on any combination of a treadmill (walking), a recumbent stationary cycle, and/or an elliptical machine (all: TechnoGym, Cesena, Italy). The resistance exercise consisted of whole body circuit training using hydraulic resistance equipment (Easy Line Circuit, TechnoGym, Cesena, Italy) that provided resistance throughout the full range of motion and emphasized the following exercises at 15–20 repetitions per set, for two to three sets: chest press and low row, latissimus dorsi pull-down and shoulder press, chest fly and reverse fly, arm extension and arm flexion, hip abduction and hip adduction, and leg extension and leg flexion. Flexibility consisted of stretching of the major muscle groups (e.g., neck, shoulders, back, triceps, calves, and hamstrings) for three sets of 20 s, each, taking care to not perform stretching exercises that would compromise balance.
Monitoring of proper intensity throughout aerobic and resistance exercise was achieved using the Borg Rating of Perceived Exertion Scale (scale of 6–20 with score 7 = very, very light and score of 19 very, very hard; aerobic exercise was kept at RPE ≤ 15, targeting an RPE of 12–14), and the talk test (able to hold a conversation) throughout the respective bout. In addition, participants wore a heart rate monitor (Polar Electro, Kempele, Finland) to track heart rate continuously during each workout.
In addition to the exercise regimen, participants were encouraged to increase their average daily steps throughout pregnancy. Step count was monitored weekly using a Garmin (Garmin Ltd., Olathe, KS) or Fitbit (Fitbit, San Francisco, CA) device and was recorded in a log. Motivational interviewing by a trained staff member was conducted weekly to address barriers to achieving weekly session and average daily step goals.
Anthropometric Measurements
Anthropometric measurements were obtained using standardized techniques in duplicate or triplicate if not within acceptable agreement. Weight was measured to the nearest 0.1 kg using a tared standing digital scale (Perspective Enterprises, Portage, MI) with the participant wearing a hospital gown. Height was measured to the nearest 0.1 cm standing against a wall-mounted stadiometer (Tanita Corp., Tokyo, Japan). Body mass index (BMI) was determined using the standard equation [BMI = body mass (kg)/height (m)2].
Dietary Assessment
Dietary intake was assessed at 12 wk and 36 wk of gestation using 3-day food records (two weekdays and one weekend day) analyzed with the Nutrition Data System for Research (NDSR, Nutrition Coordinating Center, University of Minnesota, MN) software. Total energy intake, total absolute and relative (to body mass) macronutrient, and percent macronutrient intake, as well as BCAA (leucine, isoleucine, and valine), was assessed.
Physical Activity Assessment
Physical activity was assessed on at least 3 days (one weekend day and two weekdays) with an Actical accelerometer (Philips Respironics Co. Inc., Bend, OR) before visit 1 and visit 2, respectively. The monitor was placed on the participant’s nondominant ankle, was programmed to begin recording at 11:59 PM, and recorded continuously throughout the day and night. Total activity counts (AC) per day over the course of time with which the monitor was worn were calculated. Then, to derive average total AC per day, this number was divided by the total number of valid days worn. Ultimately, total average daily steps, as well as the time that was spent in sedentary, light, moderate, and strenuous activity were provided. At the 12-wk time point, participants had not yet begun the exercise protocol, and the accelerometer thus captured daily activity data. At the 36-wk time point, the accelerometer captured both daily activity and exercise activity if scheduled workouts were overlapping with accelerometer days.
Exercise Testing
A submaximal graded walking treadmill fitness test was conducted during the first trimester before the intervention (visit 1, ∼12 wk gestation) and during the second trimester in the middle of the intervention (∼24 wk gestation). A fitness test was not conducted during the third trimester to ensure that extra precautions were being taken. The test followed exercise recommendations and guidelines from the American College of Sports Medicine guidelines during pregnancy (11). Breath composition and ventilation rate were measured at rest and during the test using a metabolic cart (Medgraphics Ultima PFX system, MGC Diagnostics Corporation, St. Paul, MN).
The participant began with a 3–5-min warm-up, and then the incline on the treadmill was increased in intervals until the participant reached 15 on the Borg Rating of Perceived Exertion Scale (RPE-15). The percent grade achieved (% grade at RPE-15), respiratory exchange ratio (RER at RPE-15, ratio of the volume of carbon dioxide produced to the volume of oxygen consumed), heart rate achieved (HR at RPE-15, beats per minute), VO2 at RPE-15 (mL·min−1), VO2 relative to body mass at RPE-15 (mL·kg−1·min−1), VO2 pulse at RPE-15; calculated as VO2 divided by heart rate (mL·kg−1·beat−1), and work (watts) were recorded at both time points. Delta values were calculated for each variable (visit 2 − visit 1).
Blood Collection and Analysis
Prior to blood collection, participants were fasted overnight. Overnight-fasted blood samples were collected (from the antecubital vein) in EDTA-coated plasma vacutainers (Becton, Dickinson & Company, Franklin Lakes, NJ) for quantification of insulin concentrations, and serum vacutainers (Becton, Dickinson & Company, Franklin Lakes, NJ) for quantification of BCAA-derived ACs (C3, C5, C5-OH, MBC), BCAAs (leucine, isoleucine, and valine), and glucose concentrations. All concentrations were measured in duplicate.
BCAA-derived acylcarnitine quantification.
C3, C5, C5-OH, and MBC were quantified using liquid chromatography-mass spectrometry [LC-MS/MS, UltiMate 3000 UHPLC system fitted with an XSelect CSH C18 XP reverse phases column, and SCIEX QTRAP 4000 (Framingham, MA) mass spectrometer]. C3, C5, C5-OH, MBC, propionylcarnitine-N-methyl-d3, 3-hydroxyisovaleryl-l-carnitine-N-methyl-d3, and isovaleryl-l-carnitine-N,N,N-trimethyl-d9 were obtained from Sigma Aldrich (St. Louis, MO), and 2-methylbutyrl-l-carnitine-d3 chloride was obtained from Cayman Chemical (Ann Arbor, MI). All solvents used were of optima grade from Fisher Scientific (Pittsburg, PA).
Serum samples (100 µL) plus 100 µL of an internal standard mix (50 mg/mL: C3-D3, C5-D9, C5-OH -D3, and MBC-D3) were extracted in methanol (2:1). Extracts were evaporated to dryness under a nitrogen stream and reconstituted in 100 µL 25% aqueous methanol. Chromatographic separation was performed on an UltiMate 3000 UHPLC system fitted with an XSelect CSH C18 XP reverse phases column (100× 2.1 mm, 2.5 µm, Watters, Milford, MA) kept at 30°C, while samples were kept at 4°C. A flow rate of 200 µL/min and injection volume of 5 µL was used. Mobile phases consisted of 0.1% trifluoroacetic acid in water (A) and 0.1% trifluoroacetic acid in methanol (B) with a 12-min elution gradient as follows: 0–3 min, 10% B; 3–5 min, 30%; 5–8 min, 99% B; and 8–12 min 10% B. Identification was carried out on a SCIEX QTRAP 4000 (Framingham, MA) mass spectrometer. Data were acquired by multiple reaction monitoring (MRM) in positive Turbo Spray ionization mode. Nitrogen as curtain, CAD, GS1, and GS2 gas was set at 50, high, 60, 50 units, respectively. Ion spray voltage and source temperature were at 4,500 V and 350°C. Results were quantitated using SCIEX Analyst 1.7 software; calibration curves were allowed a 1/w weighting and had a linear regression of >0.99.
Amino acid quantification.
An amino acid quantifying kit (EZ:faast, Phenomenex, Torrance, CA) was used to prepare samples, and gas chromatography-mass spectrometry (GC-MS, Shimadzu QP-2010 Ultra GC-MS, Shimadzu Scientific Instruments, Columbia, MD) was used to quantify amino acid concentrations (the BCAAs; leucine, isoleucine, and valine), as previously described (12). Samples were concentrated by solid-phase extraction, derivatized, and then separated by liquid–liquid extraction. Samples were analyzed in duplicate on a GC-MS outfitted with an EZ:faast amino acid analysis column (10 m × 0.25 mm × 0.25 μm). Amino acid concentration corrected for injection differences [norvaline internal standard (200 nmol/mL)] was determined using a three-point standard curve for individual amino acids (0–200 nmol/mL) (12).
IR calculation.
Insulin (Mesoscale Discovery Platform Multi-Array Assay System, Gaithersburg, MD) and glucose (Randox Daytona Clinical Analyzer, Randox Daytona, Crumlin, United Kingdom) concentrations were quantified to determine the homeostatic model assessment of IR (HOMA2-IR) using the Oxford Centre for Diabetes, Endocrinology and Metabolism homeostasis model assessment (HOMA)-2 calculator (13).
Statistical Analysis
A sample size calculation was performed for a two-factor ANCOVA using the GLMPOWER procedure in SAS version 9.4 (SAS Institute, Carey, NC). Based on a two-tailed alpha of 0.05 and a medium effect size (i.e., 0.25), the total N was determined to be 80 (i.e., 40 in each treatment group). This yielded a power of 0.88 to detect a difference of 0.06 µmol/L in mean BCAA concentrations for the main effects of time and treatment and the time × treatment interaction and the covariate, HOMA2-IR.
All results are presented as least-squares means and standard errors. Statistical tests were considered statistically significant at the 0.05 level. Analyses were performed in SAS software, version 9.4 (SAS Institute, Carey, NC).
Linear model for analysis of differences in dietary intake.
A linear model with week, group and week × group interaction terms was used to determine whether there were any differences in dietary intake. We also used Pearson’s correlations to compare dietary intake (protein, animal protein, plant protein, BCAAs) to HOMA2-IR, as well as dietary intake of these variables and circulating BCAAs, to determine whether there were any relationships at both the 12- and 36-wk time points. Furthermore, after comparing the relationship between the delta dietary values and the delta ACs/HOMA2-IR, there were no significant relationships.
Student’s t test for analysis of performance outcomes.
A Student’s t test was used to compare the delta values of all performance variables between exercise and nonexercise groups.
ANCOVA analysis of circulating BCAA-derived AC and BCAA concentrations.
Data on BCAA and AC concentrations were analyzed by an ANCOVA, incorporating the main effects of group (exercise vs. control), time (12 vs. 36 wk gestation) and a time × group interaction. IR (HOMA2-IR) was included in the statistical model as a covariate. Initially, gestational weight gain was considered as a covariate but was removed from the statistical model when it was found to be nonsignificant (P > 0.05) for each metabolite. BCAAs and ACs were examined for normality and homogeneity of variance. Additionally, BCAAs and ACs were assessed for homogeneity of slopes with the covariate, HOMA2-IR. All assumptions of the ANCOVA were met. However, one observation yielded an outlier for IR (1.5 times greater than the interquartile range) and therefore was removed from the analyses.
Analysis after collapsing groups.
A one-factor analysis of covariance (factors: week; covariate: HOMA2-IR) was run after combining groups.
RESULTS
Table 1 details baseline characteristics of all participants, as well as the exercise group and nonexercise control group independently. No baseline characteristics were different between groups. The average compliance to the exercise program (which was set to be 3 days per week) was 76%. We found that there was no association between compliance and BCAA-derived ACs in the EX group. Compliance accounted for 8.9% of the variation in MBC, 0.2% of the variation in C5-OH, 10.9% of the variation in C5, and 0.06% of the variation in C3. Although not shown in Table 1, four women were diagnosed with GDM throughout the study. One woman was diagnosed during her second trimester (no treatment), and three women were diagnosed during their third trimester (one treated with diet, the two others treated with Glyburide medication). All four of these women were in the nonexercise control group. Other medications were used to treat common problems such as heartburn, allergies, anemia or low iron levels, seasonal colds, pregnancy-related nausea, and urinary tract infections.
Table 1.
Baseline characteristics by exercise and control groups in sedentary women with obesity
| Variable | Overall n = 80 | Control n = 40 | Exercise n = 40 |
|---|---|---|---|
| Age at enrollment, yr | 28.9 ± 4.5 | 28.7 ± 4.4 | 29.2 ± 4.6 |
| Parity | 2.2 ± 0.2¥ | 2.1 ± 0.2§ | 2.2 ± 0.2§ |
| Race | |||
| Caucasian | 58 (73%) | 29 (72%) | 29 (72%) |
| African American | 20 (25%) | 11 (28%) | 9 (22%) |
| More than one/missing | 2 (2%) | 0 (0%) | 2 (4%) |
| Ethnicity | |||
| Non-Hispanic | 73 (91%) | 36 (90%) | 37 (92%) |
| Hispanic | 7 (9%) | 4 (10%) | 3 (8%) |
| Weight, 12 wk, kg | 99.4 ± 17.4 | 99.5 ± 21.2 | 99.2 ± 12.4 |
| BMI, 12 wk, kg/m2 | 36.9 ± 5.7 | 37.6 ± 6.8 | 36.2 ± 4.2 |
| Fat free mass, 12 wk, kg | 50.6 ± 7.1* | 50.6 ± 8.4 | 50.5 ± 5.4 |
| Fat mass, 12 wk, kg | 48.7 ± 12.5* | 48.9 ± 14.4 | 48.6 ± 10.1 |
| Body fat percentage, 12 wk, % | 48.6 ± 5.4* | 48.5 ± 5.5 | 48.7 ± 5.2 |
| Gestational weight gain, kg | 9.3 ± 5.3 | 8.7 ± 5.1 | 10.0 ± 5.5 |
Means ± SD; *n = 79 because of missing data; ¥n = 68 because of missing data; §n = 34 because of missing data. Student’s t-tests were used to compare control and exercise groups. No statistical differences between exercise and control for any of the presented variables (P ≥ 0.05).
Table 2 details the physical activity at baseline before the intervention (12 wk of gestation) and at 36 wk of gestation by group and overall. Of these participants, 32 of 40 exercised during accelerometer wear, but on average, they only exercised for ∼25% of the days that they wore the accelerometer. At 36 wk, average daily steps, moderate activity time, and vigorous activity time were greater, whereas sedentary time was lower in the exercise group compared with the nonexercise control group. Further, the delta values (24 vs. 12 wk) of total daily steps and moderate activity time were significantly greater in the exercise compared with the nonexercise control group.
Table 2.
Physical activity recorded by accelerometry at 12 and 36 wk of gestation by group and overall
| Variable | Overall n = 80 | Control n = 40 | Exercise n = 40 |
|---|---|---|---|
| 12 wk (n = 80) | |||
| Steps, steps/day | 6,631 ± 2,152 | 6,417 ± 2,239 | 6,845 ± 2,039 |
| Sedentary time, min/day | 1,065 ± 88 | 1,078 ± 81 | 1,052 ± 92 |
| Light activity time, min/day | 198 ± 52 | 190 ± 34 | 206 ± 64 |
| Moderate activity time, min/day | 174 ± 54 | 170 ± 56 | 179 ± 53 |
| Vigorous activity time, min/day | 3 ± 3 | 2 ± 3 | 3 ± 3 |
| 36 wk (n = 73) | |||
| Steps, steps/day | 6,378 ± 2205 | 5,522 ± 2,082 | 7,145 ± 2,023‡,* |
| Sedentary time, min/day | 1,072 ± 94 | 1,097 ± 102 | 1,049 ± 79¥,† |
| Light activity time, min/day | 198 ± 56 | 194 ± 69 | 202 ± 42 |
| Moderate activity time, min/day | 167 ± 53 | 147 ± 54 | 185 ± 45‡,* |
| Vigorous activity time, min/day | 3 ± 3 | 2 ± 2 | 3 ± 3¥,† |
| Change over time (n = 73) | |||
| Steps, steps/day | −320.8 ± 365.4 | −1,044 ± 379 | 345 ± 321¥ |
| Sedentary time, min/day | 9 ± 15 | 19 ± 17 | −0.2 ± 13 |
| Light activity time, min/day | −1 ± 10 | 4 ± 11 | −6 ± 8 |
| Moderate activity time, min/day | −8 ± 9 | −23 ± 10 | 5 ± 8¥ |
| Vigorous activity time, min/day | 0.2 ± 0.6 | −0.2 ± 0.4 | 1 ± 1 |
Means ± SD. Student’s t-tests were used to compare groups and time points.‡P < 0.001 vs. control; ¥P < 0.05 vs. control; *P < 0.001 vs. 12 wk; †P < 0.05 vs. 12 wk.
Table 3 outlines habitual dietary intake at 12 and 36 wk of gestation, with P values indicating group differences, accounting for differences due to weeks. There were no significant week by group interactions or main effect of time for any measured variable. Groups differed in relative (to body mass) intake of protein (P = 0.031) and animal protein (P = 0.033), as well as absolute intake of leucine (P = 0.024), isoleucine (P = 0.027), and valine (P = 0.017). There were no group differences in the macronutrient percentages of total energy intake.
Table 3.
Dietary intake in exercise and nonexercise control groups at baseline preintervention (12 wk) and 36 wk of gestation
| Control |
Exercise |
P Value | |||
|---|---|---|---|---|---|
| 12 Wk n = 31 | 36 Wk n = 31 | 12 Wk n = 31 | 36 Wk n = 36 | ||
| Total energy, kcal/day | 1810 (115) | 2,082 (91.7) | 2,048 (121) | 2,160 (87.1) | 0.113 |
| Fat, g·kg−1·day−1 | 0.8 (0.1) | 0.8 (0.1) | 0.9 (0.1) | 0.8 (0.04) | 0.283 |
| Fat, % | 36.8 (1.2) | 36.2 (1.2) | 36.4 (0.9) | 36.5 (0.9) | 0.931 |
| Carbohydrate, g·kg−1·day−1 | 2.2 (0.1) | 2.5 (0.1) | 2.5 (0.1) | 2.4 (0.1) | 0.441 |
| Carbohydrate, % | 48.2 (1.4) | 48.9 (1.5) | 48.1 (1.1) | 46.9 (1.1) | 0.405 |
| Protein, g·kg−1·day−1 | 0.7 (0.05) | 0.7 (0.04) | 0.8 (0.05) | 0.8 (0.04) | 0.031 |
| Protein, % | 14.9 (0.6) | 14.8 (0.7) | 15.3 (0.6) | 16.5 (0.6) | 0.097 |
| Animal protein, g·kg−1·day−1 | 0.5 (0.03) | 0.5 (0.03) | 0.5 (0.03) | 0.6 (0.03) | 0.033 |
| Plant protein, g·kg−1·day−1 | 0.2 (0.01) | 0.2 (0.01) | 0.3 (0.02) | 0.2 (0.01) | 0.269 |
| Leucine, g | 5.2 (0.4) | 5.9 (0.3) | 6.0 (0.3) | 6.7 (0.3) | 0.024 |
| Isoleucine, g | 3.0 (0.2) | 3.3 (0.2) | 3.4 (0.2) | 3.9 (0.2) | 0.027 |
| Valine, g | 3.4 (0.2) | 3.7 (0.2) | 3.9 (0.2) | 4.4 (0.2) | 0.017 |
Values in the table are least squares means and (SE). A linear model with week, group, and week-by-group interaction terms was used to determine whether there were any differences in dietary intake. n does not equal 40 in either group at either time point because several participants did not return diet logs. P values indicate between group differences adjusted for variation due to week. There were no significant week × group interactions. Bolded values are statistically significant.
After conducting Pearson’s correlations, at 12 wk, when dietary intake was expressed relative to body mass (g/kg), there was no relationship between HOMA2-IR and dietary protein, animal or plant protein, leucine, isoleucine, or valine. At 36 wk, only animal protein expressed relative to body mass was significantly related to HOMA2-IR (correlation coefficient: 0.26; P = 0.04). There were no relationships between dietary intake of protein, animal or plant protein, leucine, isoleucine, or valine and circulating BCAAs at either time point.
Results of Student’s t test for Analysis of Performance Outcomes during the First Trimester before the Intervention and during the Second Trimester in the Middle of the Intervention
Table 4 presents the performance data at both 12 and 36 wk of gestation and the delta values between the two tests. During the test, workload was increased to an intensity corresponding to an RPE of 15. Thus, there were no difference in RPE from pre- to postintervention. All delta performance variables at RPE-15 (% grade, VO2, VO2 relative to body mass, RER, heart rate, VO2 pulse, and work) were significantly and positively greater in the exercise than control group, showing a higher level of fitness in the exercise group compared with the control group as a result of the intervention.
Table 4.
Performance variables at a workload corresponding to the Borg scale rate of perceived exertion (RPE) of 15 at 12 wk (pre-intervention) and ∼24 wk (middle of the intervention) of gestation in sedentary women with obesity, and the delta values between the ∼24 versus 12 wk variables
| Control n = 35 |
Exercise n = 38 |
||||||
|---|---|---|---|---|---|---|---|
| 12 Wk | 24 Wk | Delta | 12 Wk | 24 Wk | Delta | P Value | |
| % Grade at RPE-15 | 8 ± 3 | 7 ± 2 | −1 ± 0.3 | 8 ± 2 | 10 ± 2 | 2 ± 0.4 | <0.0001 |
| VO2 at RPE-15, mL·kg−1·min−1 | 17.4 ± 3.0 | 16.3 ± 2.5 | −1.1 ± 0.4 | 17.3 ± 3.1 | 18.3 ± 2.4 | 1.0 ± 0.4 | 0.0002 |
| VO2 at RPE-15, mL·min−1 | 1700 ± 386 | 1640 ± 316 | −58 ± 36 | 1720 ± 349 | 1890 ± 300 | 171 ± 39 | <0.0001 |
| RER at RPE-15 | 1.1 ± 0.1 | 1.0 ± 0.1 | −0.02 ± 0.02 | 1.0 ± 0.1 | 1.1 ± 0.1 | 0.04 ± 0.02 | 0.008 |
| HR at RPE-15, beats/min | 163 ± 13 | 162 ± 12 | −1 ± 2 | 159 ± 18 | 161 ± 22 | 2 ± 3 | 0.222 |
| VO2 Pulse at RPE-15, mL·kg−1·bpm−1 | 0.11 ± 0.02 | 0.10 ± 0.01 | −0.006 ± 0.002 | 0.11 ± 0.02 | 0.12 ± 0.04 | 0.009 ± 0.007 | 0.03 |
| Work at RPE-15, W | 100 ± 47 | 94 ± 35 | −6 ± 5 | 109 ± 45 | 139 ± 9 | 30 ± 5 | <0.0001 |
| RPE-15 | 15 ± 1 | 15 ± 1 | 0.00 ± 0.13 | 15 ± 1 | 15 ± 1 | −0.11 ± 0.11 | 0.27 |
Values in the table are means ± SE. A Student’s t-test was used to compare the delta values of all performance variables between exercise and nonexercise groups. n does not equal 40 in either group at either time point because some participants did not complete an exercise test. P values: Delta values of control versus exercise groups. % grade, percent grade on treadmill; HR, heart rate; RER, respiratory exchange ratio (VCO2/VO2); RPE, rate of perceived exertion; VO2 pulse, VO2/HR; VO2, oxygen uptake.
Results of ANCOVA Analysis of Circulating BCAA-Derived AC and BCAA Concentrations
The coefficient of variation for each measured variable was as follows: C3 (8.8%), C5 (3.3%), C5-OH (13.2%), MBC (3.5%), leucine (8.7%), isoleucine (9.4%), valine (10.5%), insulin (3.5%), and glucose (2.8%). The two-factor ANCOVA revealed no significant group × time interactions for the BCAA-derived ACs (P values: C3 = 0.59, C5 = 0.34, C5-OH = 0.65, and MBC = 0.30), BCAAs (P values: leucine = 0.48, isoleucine = 0.43, and valine = 0.60), insulin (P value = 0.60), or glucose (P value = 0.77) after adjusting for HOMA2-IR. Table 5 presents the measured metabolite concentrations and IR measures in exercise and nonexercise control groups at baseline preintervention (12 wk) and 36 wk of gestation.
Table 5.
Metabolite concentrations and insulin resistance measures in exercise and nonexercise control groups at baseline pre-intervention (12 wk) and 36 wk of gestation
| Control n = 40 |
Exercise n = 40 |
|||||||
|---|---|---|---|---|---|---|---|---|
| 12 Wk | 36 Wk | Delta | P | 12 Wk | 36 Wk | Delta | P | |
| Leu, µmol/L | 100.5 (3.2) | 91.9 (2.7) | −8.8 (2.5) | 0.063 | 99.6 (3.1) | 84.5 (2.9) | −12.5 (2.5) | 0.001 |
| Ile, µmol/L | 59.0 (2.5) | 56.6 (2.6) | −2.9 (1.8) | 0.527 | 58.1 (2.4) | 51.0 (2.3) | −6.5 (1.8) | 0.027 |
| Val, µmol/L | 225.5 (9.5) | 206.3 (9.6) | −27.5 (6.4) | 0.164 | 223.8 (9.2) | 184.1 (8.7) | −37.1 (6.4) | 0.002 |
| C3, µmol/L | 1.00 (0.04) | 0.64 (0.04) | −0.31 (0.03) | 0.001 | 1.00 (0.04) | 0.66 (0.04) | −0.29 (0.03) | 0.001 |
| C5, µmol/L | 0.34 (0.02) | 0.24 (0.02) | −0.08 (0.02) | 0.003 | 0.30 (0.02) | 0.25 (0.02) | −0.04 (.02) | 0.072 |
| C5-OH, µmol/L | 0.18 (0.01) | 0.15 (0.01) | −0.02 (0.01) | 0.006 | 0.19 (0.01) | 0.16 (0.01) | −0.02 (0.01) | 0.019 |
| MBC, µmol/L | 0.32 (0.02) | 0.22 (0.02) | −0.08 (0.02) | 0.003 | 0.28 (0.02) | 0.23 (0.02) | −0.04 (0.02) | 0.094 |
| HOMA2-IR† | 1.9 (0.2) | 2.9 (0.2) | 0.81 (0.31) | 0.003 | 2.3 (0.2) | 2.8 (0.2) | 0.91 (0.31) | 0.097 |
| Glucose, mmol/L | 5.0 (0.1) | 5.0 (0.1) | 0.03 (0.12) | 0.870 | 4.9 (0.1) | 4.8 (0.1) | 0.05 (0.12 | 0.727 |
| Insulin, µIU/mL | 15.2 (1.9) | 23.5 (1.9) | 6.6 (2.6) | 0.003 | 18.0 (1.9) | 22.3 (1.8) | 7.7 (2.6) | 0.093 |
Values in the table are least squares means and (SE) with means adjusted for HOMA2-IR, total protein, and body mass index (BMI) and VO2 at ∼12 wk of gestation. Data were analyzed by an ANCOVA, incorporating the main effects of group (exercise vs. control), time (12 vs. 36 wk gestation), and a time-by-group interaction. IR (HOMA2-IR) was included in the statistical model as a covariate. Initially, gestational weight gain was considered as a covariate but was removed from the statistical model when it was found to be nonsignificant (P > 0.05) for each metabolite. P values indicate within-group time effects. No Delta values differed significantly between groups. Plasma samples were used to quantify insulin concentration. Serum samples were used to quantify glucose and all BCAAs and ACs. †No HOMA2-IR adjustment; BCAAs, branched-chain amino acids; C3, propionylcarnitine; C5, isovalerylcarnitine; C5-OH, 3-hydroxyisovalerylcarnitine; HOMA-IR, homeostatic model assessment of insulin resistance quantified using fasting plasma insulin and serum glucose concentrations; ile, isoleucine; leu, leucine; MBC, 2-methylbutyrylcarnitine; val, valine. ANCOVA yielded no significant group × time interaction (NS). Bolded values are statistically significant.
Further, there was no main effect of group in any of the BCAA-derived ACs (P values: C3 = 0.51, C5 = 0.99, C5-OH = 0.16, and MBC = 0.96), BCAAs (P values: leucine = 0.47, isoleucine = 0.24, and valine = 0.32), glucose (P value = 0.84), or insulin (P value = 0.45) after adjusting for HOMA2-IR.
There was a main effect of time on several metabolites within the control group (P values: C3 = 0.013, C5 = 0.006, MBC = 0.008, and insulin = 0.013), and some metabolites approached significance in the exercise group (P values: C5 = 0.059, MBC = 0.050, and insulin = 0.077; Table 5). However, there were no other main effects of time in any other variable within the respective groups.
Results of Analyses after Collapsing Groups
After collapsing the groups together because of no group × time interactions, all BCAA-derived ACs and BCAAs decreased across time (i.e., 12 vs. 36 wk; Table 6). Importantly, even after adjustment for HOMA2-IR, the effect of time remained for each metabolite (Table 6). There was a significant effect of time for insulin to increase from week 12 to week 36, but there was no main effect of time on glucose (Table 6).
Table 6.
Metabolite concentrations and insulin resistance measures after combining the exercise and control groups at 12 wk and 36 wk gestation in sedentary women with obesity (n = 40)
| 12 Wk | 36 Wk | P | |
|---|---|---|---|
| Leu, µmol/L | 101.8 (1.9) | 88.6 (1.9) | <0.0001 |
| Ile, µmol/L | 61.7 (1.6) | 54.6 (1.6) | 0.002 |
| Val, µmol/L | 236.4 (6.0) | 197.3 (6.0) | <0.0001 |
| C3, µmol/L | 1.00 (0.02) | 0.67 (0.02) | <0.0001 |
| C5, µmol/L | 0.32 (0.01) | 0.25 (0.01) | 0.0002 |
| C5-OH, µmol/L | 0.19 (0.01) | 0.16 (0.01) | <0.0001 |
| MBC, µmol/L | 0.29 (0.01) | 0.22 (0.01) | 0.0003 |
| HOMA2-IR† | 2.3 (0.2) | 3.0 (0.2) | 0.003 |
| Glucose, Mmol/L† | 4.94 (0.06) | 4.94 (0.06) | 0.9887 |
| Insulin, µIU/mL† | 18.4 (1.3) | 24.1 (1.3) | 0.0028 |
Values in the table are least squares means and (SE) adjusted for HOMA2-IR. A one-factor analysis of covariance (factors: week; covariate: HOMA2-IR) was run after combining groups. †No HOMA2-IR adjustment; leu, leucine; ile, isoleucine; val, valine; C3, propionylcarnitine; C5, isovalerylcarnitine; C5-OH, 3-hydroxyisovalerylcarnitine; MBC, 2-methylbutyrylcarnitine; HOMA2-IR, homeostatic model assessment of insulin resistance quantified using fasting plasma insulin and serum glucose concentrations.
DISCUSSION
This project aimed to determine the effect of an exercise intervention during pregnancy in healthy women with obesity on circulating BCAA-derived ACs, and if IR is related to these changes. Overall, we found that although exercise improved performance-related variables in the exercise group compared with the nonexercise control group, exercise exerted no effect on circulating BCAA-derived ACs, and adjusting for IR did not alter these results. Furthermore, when focusing exclusively on the change over the course of time, circulating BCAA-derived ACs and BCAAs significantly decreased even after adjusting for IR. IR estimated by HOMA2-IR increased similarly in both groups (ΔHOMA2-IR, exercise group: +0.90 ± 2.30; control group: +0.81 ± 1.44; P = 0.415). These findings are surprising considering that the exercise intervention elicited improvements in performance outcomes, and there is overwhelming evidence suggesting the beneficial effects of exercise during pregnancy on glucose regulation. For example, the time spent in moderate-to-vigorous physical activity is negatively associated with IR (Matsuda Index) and insulin response (first- and second-phase insulin response) measured after an oral glucose tolerance test (OGTT) in overweight and obese pregnant women (14). The absence of an impact of exercise on IR in our cohort may be the result of the healthy metabolic status of our cohort (despite having obesity), with the understanding that many reports of alterations in glucose regulation with an exercise intervention are in pregnant women afflicted with GDM (15–18).
Furthermore, the increase in IR throughout the gestational period in otherwise healthy women is a naturally occurring, nonpathological phenomenon (19). Therefore, this phenotype may be defended in pregnancy such that any exercise effects on this outcome would be limited. This idea is important because individuals with insulin-resistant obesity have higher transcription of enzymes related to impaired BCAA breakdown (20), which would reflect the generation of high concentrations of BCAA-derived ACs. However, in the case of our cohort, as IR increased modestly (within normal healthy limits), BCAA-derived ACs (and BCAAs) actually decreased. Thus, these findings confirm that: 1) There is a disconnect between BCAA-derived ACs/BCAAs and IR during pregnancy (2), which refutes studies in other populations (21, 22); 2) Our group of women was “metabolically healthy,” that is, no other significant comorbidities despite having obesity; and 3) A subtle increase in IR throughout pregnancy may be a protected phenotype to promote optimal growth and health of the baby and mother.
Regardless of IR, our finding of no effect of exercise on circulating BCAA-derived AC concentrations has been confirmed in obese, sedentary, insulin-resistant, nonpregnant women after a long-term (∼14 wk) aerobic training and weight loss (500–600 kcal/day reduction) program (23). Thus, it may be that there is a particularly tight regulation of these metabolites and that circulating concentrations are not affected by exercise perturbations. Nonetheless, although the effect of exercise on resting BCAA-derived AC concentrations are underwhelming, an exercise intervention may affect the acute BCAA catabolic responsiveness of an individual to an exercise bout. This is supported by a significant increase in the circulating concentration of one of the AC derivatives of valine (isobutyrylcarnitine) during an acute exercise bout after ∼14 wk of aerobic training and weight loss (23). These results suggest varying catabolic flux of valine compared with leucine and isoleucine (confirmed by no difference in C5 or MBC ACs).
BCAAs are known to decrease throughout pregnancy (∼13 to 31 wk of gestation) (24, 25) and to a greater extent in women that are more insulin resistant (2, 26). The decrease in BCAAs throughout gestation is associated with protein accretion and tissue growth via enhanced placental amino acid uptake (27). In the present study, we were able to confirm the decrease in circulating BCAAs and BCAA-derived ACs, reflecting an increase in tissue growth (e.g., gestational weight gain, exercise: 10.0 ± 5.5 kg; nonexercise control: 8.7 ± 5.1 kg). Therefore, independent of exercise, it seems that there is a “pull” of BCAAs toward growing fetal tissue reflected by a reduction of markers in BCAA catabolism. Nonetheless, the lack of differences in markers of BCAA catabolism between exercise and nonexercise groups in the current study may be explained by an improvement in muscle quality (e.g., mitochondrial efficiency) in the exercise group and therefore increased protein turnover and amino acid transport (28). However, to confirm this theory, future studies should measure amino acid turnover (e.g., using stable isotope technologies).
It was interesting to note that the exercise group consumed higher dietary protein, animal protein, leucine, isoleucine, and valine. However, because dietary protein relative to total energy intake (e.g., percent protein intake) was not different compared with the control group and because total energy intake was higher (although nonsignificantly) compared to the control group, it is likely that the driving factor in these differences was total energy intake. Regardless, we have previously reported that total protein and plant protein intake during late pregnancy is negatively related to insulin resistance, but this relationship disappears after adjustment for BMI (29). Furthermore, there is no relationship between animal protein and insulin sensitivity before or after BMI-adjustment. In addition, other reports have determined that dietary intake of proteins, and BCAAs are not related to circulating concentrations of BCAAs (30). Furthermore, it has been suggested that rather than dietary BCAA intake, the dietary pattern is a more influential contributing factor to high-circulating concentrations of BCAA (31). Thus, future work should focus on assessing these relationships in the lens of dietary patterns rather than individual nutrients.
One of the limitations of the current study may have been the selected exercise intensity, frequency, and duration over time. It is well known that these variables affect an individual’s physiological response to an exercise program (32). Nonetheless, there is considerable difficulty in practically increasing exercise intensity, duration, and frequency during pregnancy in previously sedentary women, considering the need to prescribe exercise that is deemed “safe” for clinical recommendations and also acceptable for current cultural and personal beliefs. Even so, moderate- to high-intensity exercise during pregnancy is consistently shown to be safe for both the mother and offspring, considering preconception fitness (e.g., if the pregnant woman was already active) (33, 34). To our knowledge, there are no other studies in pregnancy that have measured similar outcomes, and therefore it is difficult to determine the optimal exercise frequency, duration, and intensity for these specific metabolic changes.
In conclusion, there was no effect of a moderate-intensity exercise regimen from ∼13 to 36 wk gestation in sedentary women with obesity on circulating concentrations of BCAA-derived ACs or BCAAs themselves. Further, the decrease in BCAA-derived ACs and BCAAs throughout gestation cannot be explained by changes in IR. However, it does not seem like exercise has a negative effect on BCAA metabolism during pregnancy (i.e., exercise did not increase concentrations of markers of BCAA catabolism and did not “drain” BCAA plasma levels). Therefore, in this regard, moderate exercise may be deemed metabolically safe and not disruptive to the increased fuel partitioning to the growing fetal tissues in the latter stages of pregnancy. This point is particularly important when considering the implementation of an exercise intervention during pregnancy in previously sedentary women. However, there is a need for more high-quality randomized controlled trials of exercise in pregnancy measuring longitudinal changes in metabolites related to amino acid metabolism, and more importantly, changes in protein metabolic fluxes.
GRANTS
This work was supported by Arkansas Children’s Research Institute and Arkansas Biosciences Institute Postgraduate Grant (to B.R.A. and B.J.S.), US Department of Agriculture—Agricultural Research Service Projects 6026-51000-010-05S and 6026-51000-012-06S (to B.R.A., A.A., and E.B.), National Institutes of Health (NIH)/National Institute of Diabetes and Digestive and Kidney Diseases R01 DK107516 (to B.R.A, A.A., and E.B.). Authors were also partly supported by NIH/National Institute of General Medical Science 1P20GM109096-01A1 (to A.A. and E.B.).
DISCLOSURES
No conflicts of interest, financial or otherwise, are declared by the authors.
AUTHOR CONTRIBUTIONS
B.R.A., A.A., and E.B. conceived and designed research; B.R.A., K.E.M., and A.A. performed experiments; B.R.A. and B.J.S. analyzed data; B.R.A. and E.B. interpreted results of experiments; B.R.A. prepared figures; B.R.A. and E.B. drafted manuscript; B.R.A., B.J.S., K.E.M., A.A., and E.B. edited and revised manuscript; B.R.A., B.J.S., K.E.M., A.A., and E.B. approved final version of manuscript.
ACKNOWLEDGMENTS
The authors thank the participants for time and dedication to the study, and the Clinical Research Core at the ACNC for participation in data collection. We thank Alvin Dupens III in the ACNC physical activity core for exercise training of the pregnant participants, and the physical activity core for exercise testing the participants. We thank Matthew Cotter and Lindsey Pack for blood sample analyses. The authors also thank ACNC data managers Lindsey Fullen and Jennifer Ford for data support.
REFERENCES
- 1.Sandler V, Reisetter AC, Bain JR, Muehlbauer MJ, Nodzenski M, Stevens RD, Ilkayeva O, Lowe LP, Metzger BE, Newgard CB, Scholtens DM, Lowe WL, HAPO Study Cooperative Research Group. Associations of maternal BMI and insulin resistance with the maternal metabolome and newborn outcomes. Diabetologia 60: 518–530, 2017. doi: 10.1007/s00125-016-4182-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Allman B, Diaz E, Andres A, Borsheim E. Divergent changes in serum branched-chain amino acid concentrations and estimates of insulin resistance throughout gestation in healthy women. J Nutr 150: 1757–1764, 2020. doi: 10.1093/jn/nxaa096. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Newgard CB. Interplay between lipids and branched-chain amino acids in development of insulin resistance. Cell Metab 15: 606–614, 2012. doi: 10.1016/j.cmet.2012.01.024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Lynch CJ, Adams SH. Branched-chain amino acids in metabolic signalling and insulin resistance. Nat Rev Endocrinol 10: 723–736, 2014. doi: 10.1038/nrendo.2014.171. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Schooneman MG, Vaz FM, Houten SM, Soeters MR. Acylcarnitines: reflecting or inflicting insulin resistance? Diabetes 62: 1–8, 2013. doi: 10.2337/db12-0466. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.AminiLari Z, Fararouei M, Amanat S, Sinaei E, Dianatinasab S, AminiLari M, Daneshi N, Dianatinasab M. The effect of 12 weeks aerobic, resistance, and combined exercises on omentin-1 levels and insulin resistance among type 2 diabetic middle-aged women. Diabetes Metab J 41: 205–212, 2017. doi: 10.4093/dmj.2017.41.3.205. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Bharath LP, Choi WWCJ, Min Skobodzinski AA, Wong A, Sweeney TE, Park SY. Combined resistance and aerobic exercise training reduces insulin resistance and central adiposity in adolescent girls who are obese: randomized clinical trial. Eur J Appl Physiol 118: 1653–1660, 2018. doi: 10.1007/s00421-018-3898-8. [DOI] [PubMed] [Google Scholar]
- 8.Liu SX, Zheng F, Xie KL, Xie MR, Jiang LJ, Cai Y. Exercise reduces insulin resistance in type 2 diabetes mellitus via mediating the lncRNA MALAT1/MicroRNA-382-3p/resistin axis. Mol Ther Nucleic Acids 18: 34–44, 2019. doi: 10.1016/j.omtn.2019.08.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Musial B, Fernandez-Twinn DS, Duque-Guimaraes D, Carr SK, Fowden AL, Ozanne SE, Sferruzzi-Perri AN. Exercise alters the molecular pathways of insulin signaling and lipid handling in maternal tissues of obese pregnant mice. Physiol Rep 7: e14202, 2019. doi: 10.14814/phy2.14202. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.US Department of Health and Human Services. Physical activity guidelines for Americans (Online), 2nd ed. https://health.gov/paguidelines/second-edition, 2020.
- 11.ACSM. American College of Sports Medicines’s (ACSM’s) Guidelines for Exercise Testing and Prescription. Philadelphia: Lippincott William and Wilkins, 2014. [Google Scholar]
- 12.Hamarsland H, Nordengen ALN, Aas S, Holte K, Garthe I, Paulsen G, Cotter M, Børsheim E, Benestad HB, Raastad T.. Native whey protein with high levels of leucine results in similar post-exercise muscular anabolic responses as regular whey protein: a randomized controlled trial. J Int Soc Sports Nutr 14: 43, 2017. doi: 10.1186/s12970-017-0202-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.HOMA2. HOMA2 calculator: overview (Online) https://www.dtu.ox.ac.uk/homacalculator/ [2020 6 Feb].
- 14.Van Poppel MNM, Oostdam N, Eekhoff MEW, Wouters MGAJ, Van Mechelen W, Catalano PM. Longitudinal relationship of physical activity with insulin sensitivity in overweight and obese pregnant women. J Clin Endocrinol Metab 98: 2929–2935, 2013. doi: 10.1210/jc.2013-1570. [DOI] [PubMed] [Google Scholar]
- 15.Davenport MH, Ruchat SM, Poitras VJ, Jaramillo Garcia A, Gray CE, Barrowman N, Skow RJ, Meah VL, Riske L, Sobierajski F, James M, Kathol AJ, Nuspl M, Marchand AA, Nagpal TS, Slater LG, Weeks A, Adamo KB, Davies GA, Barakat R, Mottola MF. Prenatal exercise for the prevention of gestational diabetes mellitus and hypertensive disorders of pregnancy: a systematic review and meta-analysis. Br J Sports Med 52: 1367–1375, 2018. [30337463] doi: 10.1136/bjsports-2018-099355. [DOI] [PubMed] [Google Scholar]
- 16.Davenport MH, Sobierajski F, Mottola MF, Skow RJ, Meah VL, Poitras VJ, Gray CE, Jaramillo Garcia A, Barrowman N, Riske L, James M, Nagpal TS, Marchand AA, Slater LG, Adamo KB, Davies GA, Barakat R, Ruchat SM. Glucose responses to acute and chronic exercise during pregnancy: a systematic review and meta-analysis. Br J Sports Med 52: 1357–1366, 2018. doi: 10.1136/bjsports-2018-099829. [DOI] [PubMed] [Google Scholar]
- 17.Song C, Li J, Leng J, Ma RC, Yang X. Lifestyle intervention can reduce the risk of gestational diabetes: a meta-analysis of randomized controlled trials. Obes Rev 17: 960–969, 2016. doi: 10.1111/obr.12442. [DOI] [PubMed] [Google Scholar]
- 18.Wang C, Wei Y, Zhang X, Zhang Y, Xu Q, Sun Y, Su S, Zhang L, Liu C, Feng Y, Shou C, Guelfi KJ, Newnham JP, Yang H. A randomized clinical trial of exercise during pregnancy to prevent gestational diabetes mellitus and improve pregnancy outcome in overweight and obese pregnant women. Am J Obstet Gynecol 216: 340–351, 2017. doi: 10.1016/j.ajog.2017.01.037. [DOI] [PubMed] [Google Scholar]
- 19.Sonagra AD, Biradar SM, Dattatreya K, Murthy DS J. Normal pregnancy—a state of insulin resistance. J Clin Diagn Res 8: CC01–CC03, 2014. doi: 10.7860/JCDR/2014/10068.5081. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Lefort N, Glancy B, Bowen B, Willis WT, Bailowitz Z, De Filippis EA, Brophy C, Meyer C, Højlund K, Yi Z, Mandarino LJ. Increased reactive oxygen species production and lower abundance of complex I subunits and carnitine palmitoyltransferase 1B protein despite normal mitochondrial respiration in insulin-resistant human skeletal muscle. Diabetes 59: 2444–2452, 2010. doi: 10.2337/db10-0174. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Yamada C, Kondo M, Kishimoto N, Shibata T, Nagai Y, Imanishi T, Oroguchi T, Ishii N, Nishizaki Y. Association between insulin resistance and plasma amino acid profile in non-diabetic Japanese subjects. J Diabetes Investig 6: 408–415, 2015. doi: 10.1111/jdi.12323. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Yoon M-S. The emerging role of branched-chain amino acids in insulin resistance and metabolism. Nutrients 8: 405, 2016. doi: 10.3390/nu8070405. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Zhang J, Light AR, Hoppel CL, Campbell C, Chandler CJ, Burnett DJ, Souza EC, Casazza GA, Hughen RW, Keim NL, Newman JW, Hunter GR, Fernandez JR, Garvey WT, Harper ME, Fiehn O, Adams SH. Acylcarnitines as markers of exercise-associated fuel partitioning, xenometabolism, and potential signals to muscle afferent neurons. Exp Physiol 102: 48–69, 2017. doi: 10.1113/EP086019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Lindsay KL, Hellmuth C, Uhl O, Buss C, Wadhwa PD, Koletzko B, Entringer S. Longitudinal metabolomic profiling of amino acids and lipids across healthy pregnancy. PLoS One 10: e0145794, 2015. doi: 10.1371/journal.pone.0145794. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Ryckman K, Donovan B, Fleener D, Bedell B, Borowski K. Pregnancy-related changes of amino acid and acylcarnitine concentrations: the impact of obesity. Am J Perinatol Reports 6: e329–e336, 2016. doi: 10.1055/s-0036-1592414. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Zhao H, Li H, Chung ACK, Xiang L, Li X, Zheng Y, Luan H, Zhu L, Liu W, Peng Y, Zhao Y, Xu S, Li Y, Cai Z. Large-scale longitudinal metabolomics study reveals different trimester-specific alterations of metabolites in relation to gestational diabetes mellitus. J Proteome Res 18: 292–300, 2018. doi: 10.1021/acs.jproteome.8b00602. [DOI] [PubMed] [Google Scholar]
- 27.Duggleby SL, Jackson, AA. Higher weight at birth is related to decreased maternal amino acid oxidation during pregnancy. Am J Clin Nutr 76: 852–857, 2002. doi: 10.1093/ajcn/76.4.852. [DOI] [PubMed] [Google Scholar]
- 28.Biolo G, Maggi SP, Williams BD, Tipton KD, Wolfe RR. Increased rates of muscle protein turnover and amino acid transport after resistance exercise in humans. Am J Physiol Endocrinol Metab 268: E514–E520, 1995. doi: 10.1152/ajpendo.1995.268.3.E514. [DOI] [PubMed] [Google Scholar]
- 29.Allman BR, Andres A, Børsheim E. The association of maternal protein intake during pregnancy in humans with maternal and offspring insulin sensitivity measures. Curr Dev Nutr 3: nzz055, 2019. doi: 10.1093/cdn/nzz055. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Rousseau M, Guénard F, Garneau V, Allam-Ndoul B, Lemieux S, Pérusse L, Vohl MC. Associations between dietary protein sources, plasma BCAA and short-chain acylcarnitine levels in adults. Nutrients 11: 173, 2019. doi: 10.3390/nu11010173. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Merz B, Frommherz L, Rist MJ, Kulling SE, Bub A, Watzl B. Dietary pattern and plasma BCAA-variations in healthy men and women—results from the KarMeN Study. Nutrients 10: 623, 2018. doi: 10.3390/nu10050623. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Burton DA, Stokes K, Hall GM. Physiological effects of exercise. Contin Educ Anaesth Crit Care Pain 4: 185–188, 2004. doi: 10.1093/bjaceaccp/mkh050. [DOI] [Google Scholar]
- 33.Beetham KS, Giles C, Noetel M, Clifton V, Jones JC, Naughton G. The effects of vigorous intensity exercise in the third trimester of pregnancy: a systematic review and meta-analysis. BMC Pregnancy Childbirth 19: 281, 2019. doi: 10.1186/s12884-019-2441-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Hinman SK, Smith KB, Quillen DM, Smith MS. Exercise in pregnancy: a clinical review. Sports Health 7: 527–531, 2015. doi: 10.1177/1941738115599358. [DOI] [PMC free article] [PubMed] [Google Scholar]