A Single Day of Reduced Activity Alters the Next Day’s Transcriptomic and Metabolic Exercise Response

Abstract

Regular aerobic exercise decreases the risk of metabolic diseases; however, reduced physical activity may attenuate these beneficial adaptations. This study investigated the influence of a single day of reduced activity on the skeletal muscle transcriptome (RNASeq) and systemic metabolism surrounding an acute bout of aerobic exercise. Utilizing a crossover design, subjects (n=9 [7M, 2F], 24±1y) exercised at 65% VO₂max for 60-minutes following a day of reduced activity (IN; 3,581±1,185 steps) and normal activity (A; 11,069±3,631 steps). RNAseq from skeletal muscle biopsies were taken before and 4-hours post-exercise. Respiratory gas exchange analysis was completed at 20, 40, and 60 minutes of exercise. Blood samples for select metabolites were collected before, during, and throughout the 4-hour recovery. Inactivity resulted in 494 Differentially Expressed Genes (DEG) at rest, while exercise altered 793 and 1,403 genes in A and IN, respectively. Gene set enrichment analysis revealed an interaction of exercise and trial, which were related to inflammation, metabolism, cell cycle, and innate immunity. Respiratory exchange ratio (0.93±0.04 vs. 0.91±0.03; p<0.05) and carbohydrate utilization (2.16±0.79 vs. 1.96±0.72 g/min; p=0.09) were higher in the IN vs. A trial at 60-minutes of exercise. Triglycerides were higher in the IN trial (p<0.05), with no differences between trials observed in other measured metabolites. These findings indicate that a single day of reduced physical activity can elicit modest but significant changes in the metabolic and transcriptomic exercise response. Such alterations have implications for exercise guidelines and underscore the importance of controlling for daily step count when assessing exercise responses.

New and Noteworthy:

This investigation shows modest differences in the metabolic demands of an acute bout of exercise following a day of inactivity. Additionally, there are differences in the skeletal muscle transcriptome at rest and in the exercise response with decreased physical activity before exercise. This study highlights the need for more research on the role of physical inactivity in the exercise response, as well as the need to control for step count before exercise response studies.

INTRODUCTION

The poor health of the United States population cannot be overstated, as obesity rates increased 11% (30.5% to 41.9%) and severe obesity doubled (4.7% to 9.2%) in the last two decades (1). A key factor contributing to poor health is a sedentary lifestyle, as the average total sedentary time in the United States has nearly doubled from 5.5 to 9.5 hours per day between 2007 and 2019 (2). Recent data shows that sitting ~10 hours a day results in a ~25% increased risk of cardiovascular disease (CVD) (3). Together, these studies ultimately suggest the average adult in the United States now has a 25% elevated risk of CVD compared to 2007, solely due to an increase in sedentary lifestyle, which is a trend growing worldwide (4). The recognized adverse impact on cardiovascular health has led to the common public health slogan “sitting is the new smoking”. While epidemiological evidence indicates equating sitting with smoking may not be entirely accurate (5), the underlying message underscores the urgency to investigate and mitigate the effects of low physical activity on health.

The American Heart Association recommends a minimum of 150 minutes of moderate-to-vigorous aerobic exercise or 75 minutes of vigorous exercise per week. This type of exercise prescription has been shown to improve health outcomes, including cardiorespiratory fitness (VO₂max) and blood lipid profile, if completed consistently (6). The promotion of exercise is noble; however, prolonged sitting may interfere with the beneficial effects of exercise (7–11). Subjects who took <5,000 steps and then ran for 1 hour at 60% of VO₂max had blunted fat oxidation of post-prandial plasma triglycerides (TG) compared to those who took >8,500 steps and then exercised (8). The 1-hour run likely equates to ~10,000 steps by itself, which is a common recommendation for daily activity. With endurance exercise, fatty acid oxidation in skeletal muscle can increase up to 10-fold (12) and may remain elevated for ~18 hours after exercise (13). As lipoprotein lipase appears to be downregulated with inactivity (9, 14), skeletal muscle metabolism could be the root cause of the blunting of circulating TG’s with inactivity, which may affect the skeletal muscle’s response to exercise.

When single bouts of exercise have blunted acute responses, long-term adaptations will likely decrease. Previous studies have shown significant heterogeneity in the exercise response, which may be due to both non-modifiable (e.g., chronological age, sex, race/ethnicity) and modifiable (e.g., nutrition, obesity, medications, exercise capacity, and disease state) factors (15–18). Given the aforementioned role of sedentary time on blunted TG oxidation after exercise, physical activity around the exercise bout also appears to be a modifiable factor in the exercise response. As reduced activity before athletic competition (i.e., tapering) appears to enhance the skeletal muscle’s exercise responsiveness (19), the contrast of inactivity-induced exercise resistance and enhanced responsiveness during tapering highlights the complexity of physical inactivity and subsequent exercise responsiveness. Therefore, this study aims to identify the systemic exercise response of an acute aerobic exercise bout on the same individuals after a day of reduced activity (<5,000 steps) or normal activity (>8,500 steps) the day before exercise. These step thresholds were previously suggested by Burton and Coyle to induce metabolic exercise resistance (20). We hypothesized that individuals with a single day of reduced activity would have an altered transcriptome at rest (effect of inactivity) and in response to a bout of aerobic exercise (interaction of exercise and inactivity). We also hypothesized that a single day of reduced activity would decrease systemic fat metabolism during exercise and thus alter the blood lipid profile during and after exercise. This study may give insight into the role of inactivity on the exercise response and provide a framework of muscular adaptations that may impact CVD risk with inactivity-induced exercise resistance. Additionally, this study will have a wide-ranging impact on other areas, including determining whether steps per day should be controlled for in studies that measure the exercise response. If we determine that inactivity influences skeletal muscle exercise response, then future studies may need to consider controlling for step count when studying the acute exercise response.

METHODS

Subjects and Overall Study Design

A crossover design was utilized to determine the effects of a single day of inactivity on the next day’s exercise response. Eleven recreationally active subjects (9 males, 2 females) were recruited from the greater Mequon, WI area. Subjects completed a total of five visits over six weeks. Evaluations included anthropometrics and body composition, muscle imaging of vastus lateralis (VL), cardiorespiratory fitness test (VO₂max), an exercise familiarization test to validate the workload associated with 65% of subject’s VO₂max, and two acute aerobic exercise trials (Visits 4 and 5) with biospecimen collections after a day of normal activity (>8,500 steps, A) and inactivity (<5,000 steps, IN). Visits 4 and 5 were completed in a randomized and counterbalanced fashion and separated by 13–28 days. The study design schematic is shown in Figure 1. Two subjects (2 males) did not complete either acute exercise bout due to scheduling conflicts and are not included in the analysis. One subject (1 male) only completed the IN trial due to an injury that was not associated with the study. Subject characteristics are presented in Table 1. All study procedures, risks, and benefits were explained to the subjects before giving written consent. This research was approved by the Institutional Review Board of Concordia University of Wisconsin and Ann Arbor (IRB #FY23–5).

Figure 1.

Study design overview. A) Subjects completed five visits over the course of 6 weeks. The first three visits were designed to collect anthropometric data and prescribe the workloads for the Acute bout day with biospecimens (visits 4 and 5). B) On Visits 4 and 5, subjects reported to the Exercise Physiology Lab at 0700 after a day of active (>8,500 steps) or inactivity (<5,000 steps). Strategically timed muscle biopsies, blood samples, and respiratory exchange analysis using the metabolic cart helped comprehensively characterize the exercise response.

Table 1.

Subject characteristics

n	9 (7 males, 2 females)
Age (years)	24 ± 1
Height (cm)	178 ± 3
Weight (kg)	80 ± 35
Body Mass Index (kg/m2)	28 ± 4
Body Fat (%)	26.3 ± 8.7
Vastus Lateralis Thickness (cm)	Right: 2.97 ± 0.47 Left: 2.88 ± 0.60
VO 2 max Test
VO2max (L/min)	3.42 ± 0.96
VO2max (ml/kg/min)	37.1 ± 7.1
Peak Workload (watts)	271 ± 61
Heart Rate, maximum (bpm)	191 ± 8.3
Respiratory Exchange Ratio	1.17 ± 0.08
Ventilation (L/min)	132 ± 38
Rating of Perceived Exertion	18 ± 1

Bpm, beats per minute

VO₂max test

Before the VO₂max test, subjects’ weight and body composition were measured via bioelectrical impedance analysis (mBCA Pro, SECA; Hamburg, Germany). Subjects then performed a continuous cycle ergometer test (Corival CPET, Lode; Groningen, Netherlands) with a 12-lead electrocardiogram (ECG) to volitional exhaustion. Respiratory gas exchange was measured and averaged in 15-s intervals using indirect calorimetry via an automated open-circuit system (Parvo Medics; Sandy, UT). Before each test, the gas analyzers were calibrated with standardized gases, and the pneumotach was calibrated using a standard volume of air. Subjects completed a 3-minute warm-up (100 W) at a minimum of 60 revolutions/min. Following the warm-up, the watt load on the cycle ergometer was increased in a ramped fashion (Females: 15 W per min; Males: 20 W per min) until volitional fatigue. Maximal heart rate was determined from the ECG, and peak workload was measured as the highest watt output completed. All subjects reached a maximal effort as evidenced by a plateau in oxygen uptake, >95% age-predicted maximal heart rate, and/or rating of perceived exertion ≥18. The average of at least two of the highest consecutive readings of the test was used for the measurement of VO₂max.

Vastus Lateralis Thickness

Muscle thickness of the vastus lateralis (VL) was measured as described previously (21). Briefly, a reference point was identified 20 cm superior to the patella, directed toward the anterior superior iliac spine. Medial and lateral borders of the VL were identified at the reference with a 12L linear probe (LOGIQ^™ P9, General Electric; Chicago, IL), and the midpoint was established. The probe was placed over the midpoint in a transverse position, with care taken to minimize compressive forces. Image depth was adjusted to visualize the cortical outline of the femur. At the reference point, two fixed images were captured independently for each leg. Using the GE LOGIQ^™ P9 system software, VL muscle thickness was measured for each image. The mean value was then calculated for the right and left leg.

Exercise Familiarization

At least 24 hours following the VO₂max test, subjects completed one familiarization session to determine the workload on the same cycle ergometer as the VO₂max test, corresponding to ~65% VO₂max (workload used for the acute bout). After a 10-minute warm-up at 50% of the prescribed workload, the workload was increased to the prescribed workload for 10 minutes while respiratory gas exchange analysis occurred. The workload was adjusted if needed to achieve a workload associated with 65 ± 5% of their VO₂max.

Acute Exercise Trials

Subjects were instructed to refrain from COX-inhibitor consumption (e.g., acetaminophen, ibuprofen, aspirin) for seven days, and abstain from structured exercise, alcohol, and caffeine for at least 24 hours before the acute exercise bout. Subjects were asked to maintain their normal dietary habits and were instructed to record all food consumed (type, portion size, and time) during the two days before each trial (Visits 4 and 5). The dietary log was copied and returned to the subjects before Visit 5 (the second acute exercise trial) with instructions to replicate their diet as closely as possible. Two days before the first acute bout, subjects were randomized to complete either <5,000 steps (IN) or >8,500 (A) steps the day before the acute bout. Subjects were instructed to avoid any running, walks >10 minutes, and minimize stairs. Subjects were asked to wear an activity monitor (VivoSmart 5, Garmin; Olathe, Kansas) to measure steps upon waking and remove it upon going to bed.

At 6:00 AM the morning of the acute bout (two hours before the acute exercise bout), subjects consumed ~25% of their normal daily calories in the liquid form (Ensure Plus, Abbott Laboratories, Columbus, OH; 57% carbohydrate, 15% protein, and 28% fat) to mimic a standard breakfast. Estimated daily caloric consumption was calculated as 1.5 multiplied by the estimated resting energy expenditure (22). Subjects were then allowed to drink water but avoid other liquids or food until the visit was completed.

Subjects arrived at the Exercise Physiology Laboratory at 7:00 AM. Once in the laboratory, subjects rested quietly in the supine position before undergoing a skeletal muscle biopsy of the VL (23) and a blood sample for select metabolites. Subjects then warmed up at a light load (50% of the acute exercise workload) for 5 minutes on the cycle ergometer from the VO₂max test and familiarization session. After the warm-up, subjects completed a 60-minute bout of aerobic exercise at 65% of their VO₂max (as determined by the Exercise Familiarization Session), followed by a 2-minute cool-down at the same workload as the warm-up. If subjects were unable to complete the 60-minute workload, the workload was decreased to ensure completion of the 60-minute exercise bout. Any changes in workload were replicated in the second acute bout (Visit 5). Rating of perceived exertion was assessed at 20, 40, and 60 minutes. Heart rate was recorded at 5-minute stages during exercise (H10, Polar; Kimpele, Finland). Respiratory gas exchange data were collected during 15–20, 35–40, and 55–60 minutes (Parvo Medics; Sandy, UT). Respiratory gas exchange data were averaged over the last two minutes of each period and reported for minutes 20, 40, and 60. After exercise, subjects rested quietly in the laboratory for 4 hours for serial blood sampling and post-exercise muscle biopsy (24).

Substrate Utilization

Respiratory gas exchange data were used to calculate substrate utilization [carbohydrate (CHO) and fat] using the following equations (25):

1. CHO (g·min⁻¹) = 4.210 (VCO₂) – 2.962 (VO₂)

2. Fat (g·min⁻¹) = 1.695 (VO₂) – 1.701 (VCO₂)

3. CHO (kcal·min⁻¹) = CHO (g·min⁻¹) * 4.07 (kcals·g⁻¹)

4. Fat (kcal·min⁻¹) = Fat (g·min⁻¹) * 9.75 (kcals·g⁻¹)

5. Calorie consumption (kcal·min⁻¹) = Fat (kcals·min⁻¹) + CHO (kcals·min⁻¹)

Blood Samples

All blood was obtained via finger stick. Lactate was analyzed immediately (Lactate Plus Meter, Nova Biomedical; Waltham, MA). Total cholesterol, TG’s, HDL, and LDL were also analyzed immediately (Allegro, Nova Biomedical; Waltham, MA). Blood analyzers were calibrated each day according to the manufacturer’s instructions. Two 40 μL capillary tubes of blood were collected and stored in a −80°C freezer for future analysis of fatty acid methyl esters (FAME). Blood samples were obtained at baseline, minutes 40 and 60 of exercise, as well as at 30, 60, 90, 120, 180, and 240-minutes post-exercise. Lactate and FAME were only completed at time points of baseline, minutes 40 and 60 of exercise, as well as 30- and 60-minutes post-exercise.

Fatty Acid Methyl Esters

Fatty acid methyl esters (FAME) were prepared from extracted lipids using a 1 ethanol: 4 ethyl acetate: 1 water mixture followed by transesterification using a methanolic hydrochloric acid solution as previously described (26, 27). Fatty acids were quantified and analyzed using external standards and gas chromatography (Trace 1310, ThermoFischer; Waltham, MA) equipped with an Agilent DB23 column (Agilent; Santa Clara, CA). Data analysis was performed using Chromeleon 7.2.10. The limit of detection (LOD) and limit of quantitation (LOQ) were 0.34 μg/ml and 1.04 μg/ml for each FAME using the ratio of standard deviation to slope. The average FAME coefficient of determination was 0.998. Relative standard deviation was 6.6%. Due to technical issues, one subject (trial IN) was not analyzed.

Muscle Biopsy

Skeletal muscle biopsies were obtained from the VL following local anesthetic (1% lidocaine HCL) with a Bergstrom needle with suction pre-exercise and ~4 hours after exercise. Following the muscle biopsy, excess blood, visible fat, and connective tissue were removed from each muscle biopsy sample and were immediately frozen in liquid nitrogen and stored in a −80°C freezer until analysis.

RNA Extraction

Pre- and post-exercise VL muscle biopsy samples were sent to the University of Wisconsin Genomic Expression Center (RRID: SCR_017757) for RNA extraction and RNA sequencing. Briefly, Trizol (Invitrogen; Waltham, MA) was added to each tissue and homogenized using Tissuelyser (Tissuelyser II, Qiagen; Venlo, The Netherlands), followed by organic extraction. RNA was DNase-treated on-column using the RNeasy Mini Kit (Qiagen; Venlo, The Netherlands) according to the manufacturer’s instructions. RNA yield, purity, and integrity were assayed on Agilent Plate Reader Take 3 (Agilent; Santa Clara, CA) and Agilent 4200Tapestation (Agilent; Santa Clara, CA). Samples with RIN values ≥ 6.3 were included in the study. Four samples did not meet this requirement and were excluded from RNA sequencing analysis.

Library Preparation

Directional mRNA libraries were prepared from samples that met the NEBNext Ultra II Directional RNA Library Prep Kit for Illumina (New England BioLabs, Inc.: Ipswich, MA) guidelines (Version 4.0_4/21) following the manufacturer’s protocol. Final libraries were assessed for size and quantity using an Agilent DNA1000 Screentape (Agilent; Santa Clara, CA) and BioTek Synergy | H1 (Biotek; Winooski, VT). Microplate Reader using QuantiT^™ PicoGreen^™ dsDNA Reagent (Invitrogen; Waltham, MA), respectively. Libraries were standardized to 2nM. Paired-end 150bp sequencing was performed, using standard SBS chemistry on an Illumina NovaSeq X Plus sequencer (Illumina Inc.; San Diego, CA).

DATA ANALYSIS AND STATISTICS

Statistical analysis for sequencing data

The authors utilized the University of Wisconsin-Madison Biotechnology Center Bioinformatics Core Facility (Research Resource Identifier – RRID:SCR_017799) for analysis. Briefly, reads were subjected to quality control using FastQc (28). The trimming software Skewer was used to preprocess raw fastq files. Following preprocessing, reads were aligned to GRCh38.p10 (genome-build-accession NCBI:GCA_000001405.25) using STAR aligner (v2.5.3a; (29)), employing a two-pass mapping strategy to improve splice junction detection. Transcript-level quantification was performed with RSEM (RNA-Seq by Expectation Maximization, v1.2.31; (28)), which estimates gene and isoform abundance using the STAR-aligned BAM files. To assess global gene expression patterns, a multidimensional scaling (MDS) plot was generated based on log2-counts-per-million (logCPM) values from TMM-normalized gene counts, illustrating the transcriptional relationships among the four experimental conditions (two exercise trials, each sampled at pre- and post-exercise time points). Three samples identified as biological outliers were removed based on this exploratory analysis, resulting in a final dataset of 25 samples: 6 Pre-exercise Trial A, 6 Pre-exercise Trial IN, 6 Post-exercise Trial A, and 7 Post-exercise Trial IN.

Differentially expressed gene (DEG) analysis was performed using the edgeR package (v4.0.15) in R (v4.3.3; (29)). Prior to statistical analysis with edgeR, genes with low expression were removed using independent filtering with the filterByExpr function. This function retains genes that have sufficiently large counts, defined as a count-per-million (CPM) value corresponding to at least 10 raw counts in the smallest library size, in a minimum number of samples. The minimum number of samples is set to the number of biological replicates in the smallest experimental group, ensuring that retained genes have enough power for detection. Library sizes were normalized using the trimmed mean of M-values (TMM) method (30). A negative binomial generalized linear model (GLM) was fitted to the count data, and contrasts were defined to test for differential expression between trials at baseline (Pre-exercise A vs. Pre-exercise IN), after intervention (Post-exercise A vs. Post-exercise IN), and within each trial across time (Post- vs. Pre-exercise in both A and IN arms). Significantly DEGs were defined using a false discovery rate (FDR; (31)) <= 0.05, corrected for multiple comparisons via the Benjamini-Hochberg method. The validity of the Benjamini-Hochberg FDR multiple testing procedure was evaluated by inspection of the uncorrected p-value distribution.

To explore functional enrichment, DEGs were subjected to Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis as previously described (26). Additionally, gene set enrichment analysis (GSEA; (32)) was conducted using the R package fgsea (v1.34.0; (33)) in conjunction with msigdbr (v25.1.0;(34)) to evaluate the coordinated expression of functionally related gene sets. For each contrast, a pre-ranked gene list was generated from the differential expression result by calculating −log10(uncorrected p-value) * sign(log2 fold-change), and gene set enrichment was tested using the fgsea Multilevel function.

Statistics for non-sequencing data

Statistical analyses were performed using GraphPad Prism (V10.5, GraphPad Software; Boston, MA). A two-way mixed effect method (main effects of trial and time) with a Tukey’s post hoc test was used to compare the effects of exercise and inactivity on data related to blood metabolites and expired gas analysis. A student’s paired t-test was used to compare the step count between trials and total energy expenditure. Statistical significance was set at p<0.05. Data presented are means ± standard deviation.

RESULTS

Exercise Test

Data from the acute exercise test is found in Table 2. Subjects walked 11,069 ± 3,631 or 3,581 ± 1,185 steps the day before the A or IN trial, respectively (p<0.05). There were no differences in exercise workload, heart rate, VO₂, or VCO₂ between trials or time points. Time influenced the Rating of Perceived Exertion as subjects had lower values at minute 20 compared to minutes 40 and 60 (p<0.05). Respiratory exchange ratio (RER) decreased at each time point (p<0.05). RER was greater in trial IN compared to trial A at minute 60 (p<0.05).

Table 2.

Data from the acute exercise bout

	Exercise Timepoint (minutes)
	20		40		60
Trial	IN	A	IN	A	IN	A
Workload (Watts)	142 ± 44	138 ± 47	141 ± 45	137 ± 47	133 ± 47	136 ± 48
Rating of Perceived Exertion	13 ± 2*	13 ± 1*	15 ± 3	15 ± 2	16 ± 2	15 ± 2
Heart Rate (bpm)	157 ± 14	155 ± 13	156 ± 13	162 ± 13	157 ± 17	162 ± 16
VO2 (L/min)	2.17 ± 0.53	2.11 ± 0.55	2.21 ± 0.56	2.16 ± 0.60	2.19 ± 0.62	2.16 ± 0.63
Percent VO2max (%)	64 ± 7	63 ± 5	65 ± 8	64 ± 4	66 ± 6	66 ± 6
VCO2 (L/min)	2.12 ± 0.54	2.08 ± 0.58	2.07 ± 0.53	2.05 ± 0.61	2.02 ± 0.63	1.99 ± 0.61
Respiratory Exchange Ratio	0.97 ± 0.04*	0.98 ± 0.04*	0.94 ± 0.03ǂ	0.94 ± 0.04*	0.92 ± 0.05†	0.91 ± 0.03

^*P<0.05 vs. other time points within trial;

^†P<0.05 vs. Active trial at the same time point.

^ǂP<0.05 vs. 20 minutes within trial; IN, Inactive Trial; A, Normal Activity Trial; bpm, beats per minute

Substrate Utilization

Energy expenditure and substrate utilization are presented in Figure 2. Fat utilization increased in minutes 40 and 60 of both trials. CHO utilization decreased in minutes 40 and 60 in trial A. In the IN trial, CHO utilization decreased in minute 40 (p<0.05) but showed no difference between minutes 40 and 60. Comparisons between trials showed a statistical trend in greater CHO utilization in the IN trial at minute 60 compared to A (p=0.09). No differences were observed between trials in total caloric consumption (642 ± 178 and 656 ± 167 kcals in trial A and IN, respectively).

Figure 2.

Substrate utilization during exercise. Respiratory gas analysis was done at minutes 15–20, 35–40, and 55–60 during exercise. The last two minutes of each time period were averaged. Calorie consumption, fat utilization, and carbohydrate (CHO) utilization were calculated at each time point (25). Each data point represents a subject (n=8 A; 9 IN). A) No differences for time of exercise or trial were observed in Calorie consumption. B) In both trials, fat utilization increased from minute 20. Negative fat utilization values indicate elevated VCO2 resulting from bicarbonate buffering. C) CHO utilization decreased at each reading in the Active trial; however, there was no change in minutes 40 and 60 in the Inactive trial. CHO utilization trended to be higher in the Inactive trial. *P<0.05 vs. other time points. ‡P=0.09 vs. Trial A.

Blood Metabolites

Blood metabolites are presented in Figure 3. TGs were highest at minute 40 of exercise and then decreased at the 60-minute post-exercise time point and beyond (p<0.05). Between trials, TG’s were higher in the IN trial (p<0.05). LDL was greater 90 minutes post-exercise than pre-exercise, with no other differences for time or trial were observed. No differences in time or trial were observed in HDL, total cholesterol, or total FAME. Lactate was greatest in minutes 40 and 60 of exercise and decreased at post-exercise time points (p<0.05). No differences in lactate were observed between trials.

Figure 3.

Blood metabolites response to exercise after a day of Inactivity (<5,000 steps) or normal activity (>8,500 steps). A) Lactate, B) Total Cholesterol, C) HDL, D) LDL, E) Triglycerides and F) Total Fatty Acid Methyl Esters (FAME) were measured at selected time points before, during, and after an acute bout of exercise. Mixed methods analysis with Tukey’s post hoc test was used for statistical analysis. P values for time, trial, and interaction are listed on each figure. ^AP<0.05 vs. 60, 90-, 120-, 180-, and 240-minute post exercise; ^BP<0.05 vs. 240-minute post exercise. ^CP<0.05 vs. pre-exercise. ^DP<0.05 vs. 30- and 60-minute post-exercise.

Gene-level differential expression

Pairwise contrasts revealed nine differentially expressed genes (DEGs) (FDR < 0.05) when comparing the pre-exercise (basal) samples between the IN and A trial. Of these, seven were upregulated (four ribosomal RNA transcripts, APLN, RN7SK, and PDP1). The genes MTND1P23 and TNFRSF12A were downregulated with inactivity (Figure 4). Exercise resulted in 793 and 1,403 DEGs (FDR<0.05) between post- and pre-exercise samples in trials A and IN, respectively. No genes exhibited significantly different expression between the groups post-exercise, and no genes reached significance in the interaction contrast. DEGs for the exercise response in trial IN and A are provided in Supplemental Table S1.

Figure 4.

Basal gene expression changes with a day of inactivity (<5,000 steps) compared to normal activity (active; >8,500 steps) in skeletal muscle. The RNA profiles of the pre-exercise samples were compared. Differentially expressed genes (DEG) were defined as false discovery rate <0.05. Red indicates genes that were down-regulated with inactivity while green represents genes that were up-regulated with inactivity. In total, 494 DEG’s were found with 116 up-regulated and 378 down-regulated. The top 5 affected genes that were up-regulated and down-regulated are listed.

Pathway-level enrichment

To determine whether these gene-level responses reflected specific biological pathways, we performed KEGG enrichment analysis using up to 500 DEGs from each contrast. Significant enrichment was observed only in the within-group exercise contrasts, yielding 35 in the IN trial and 17 in the A trial (Table 3). Of the significantly enriched pathways with exercise, 14 were shared between trials. Neither baseline nor post-exercise between-group comparisons produced enriched KEGG categories.

Table 3.

Pathways that were significantly impacted by exercise inactive trial and normal activity trial from KEGG enrichment analysis

Inactive Trial
category	subcategory	Description	p.adjust
Environmental Information Processing	Signal transduction	PI3K-Akt signaling pathway	0.00157
Human Diseases	Cancer: specific types	Small cell lung cancer	0.00192
Cellular Processes	Cellular community - eukaryotes	Focal adhesion	0.00518
Human Diseases	Endocrine and metabolic disease	AGE-RAGE signaling pathway in diabetic complications	0.01036
Environmental Information Processing	Signal transduction	Rap1 signaling pathway	0.01431
Environmental Information Processing	Signal transduction	JAK-STAT signaling pathway	0.03247
Human Diseases	Infectious disease: parasitic	Malaria	0.03247
Organismal Systems	Endocrine system	Adipocytokine signaling pathway	0.03545
Organismal Systems	Aging	Longevity regulating pathway	0.03545
Human Diseases	Cancer: specific types	Thyroid cancer	0.03545
Metabolism	Carbohydrate metabolism	Amino sugar and nucleotide sugar metabolism	0.03665
Human Diseases	Cancer: specific types	Melanoma	0.03665
Human Diseases	Cancer: specific types	Non-small cell lung cancer	0.03665
Cellular Processes	Cell growth and death	Cellular senescence	0.03665
Cellular Processes	Cell growth and death	p53 signaling pathway	0.03947
Human Diseases	Infectious disease: viral	Human papillomavirus infection	0.03947
Cellular Processes	Cell motility	Cytoskeleton in muscle cells	0.03947
Human Diseases	Cancer: specific types	Bladder cancer	0.04071
Human Diseases	Cancer: specific types	Pancreatic cancer	0.04071
Organismal Systems	Endocrine system	Growth hormone synthesis secretion and action	0.04821
Organismal Systems	Digestive system	Mineral absorption	0.04821
Normal Activity Trial
Human Diseases	Cancer: specific types	Acute myeloid leukemia	0.04135
Human Diseases	Infectious disease: viral	Human T-cell leukemia virus 1 infection	0.04135
Human Diseases	Cancer: specific types	Renal cell carcinoma	0.04135
Shared Pathways Between Trials
Organismal Systems	Environmental adaptation	Circadian rhythm	3.1E-05
Environmental Information Processing	Signal transduction	MAPK signaling pathway	3.1E-05
Human Diseases	Cardiovascular disease	Fluid shear stress and atherosclerosis	0.00164
Human Diseases	Cancer: specific types	Chronic myeloid leukemia	0.01103
Environmental Information Processing	Signal transduction	NF-kappa B signaling pathway	0.02113
Environmental Information Processing	Signal transduction	HIF-1 signaling pathway	0.02558
Organismal Systems	Endocrine system	Insulin signaling pathway	0.03554
Environmental Information Processing	Signal transduction	TNF signaling pathway	0.03554
Human Diseases	Cancer: specific types	Breast cancer	0.04135
Human Diseases	Endocrine and metabolic disease	Insulin resistance	0.04135
Human Diseases	Cancer: overview	Transcriptional misregulation in cancer	0.04135
Human Diseases	Infectious disease: viral	Human cytomegalovirus infection	0.04135
Human Diseases	Cancer: overview	Proteoglycans in cancer	0.04164
Cellular Processes	Transport and catabolism	Efferocytosis	0.04164

Gene set enrichment analysis reveals broader effects

To capture more subtle and coordinated transcriptional changes, we applied GSEA using the Hallmark collection from MSigDB, testing all genes ranked by differential expression. This analysis uncovered a broader range of biological processes associated with both exercise and baseline activity status. Statistically significant gene sets (FDR < 0.05) were identified in all five contrasts: 26 in baseline inactive vs. normal activity comparison (Supplemental Table S2), 33 in inactive exercise response and 27 in normal activity exercise response (Supplemental Table S3), 18 in post-exercise group comparison (Supplemental Table S4), and 25 in the interaction (Supplement Table S5).

DISCUSSION

Elevated sedentary time (i.e., reduced step count) has been shown to alter TG metabolism, promote insulin resistance, and ultimately increase the risk of cardiovascular and metabolic diseases (3, 14, 30, 31). Inactivity (<5,000 steps/day) even blunts the beneficial metabolic effects of exercise (8, 20), a phenomenon called exercise resistance. The dysregulation of the metabolic exercise response is also likely to alter the skeletal muscle response to exercise and has a significant impact on exercise guidelines to enhance exercise adaptations in individuals. With this in mind, this study sought to characterize the systemic metabolic and skeletal muscle transcriptomic exercise response in the same subjects to a single day of inactivity (<5,000 steps) compared to normal activity (>8,500 steps). A single day of inactivity was chosen as this is likely a common occurrence, given that many exercise guidelines do not provide specific guidance on physical activity (i.e., step count thresholds) on days without exercise (32). Additionally, it is important to establish a timeline on the impact of reduced activity given that tapering (reduced exercise before an athletic event) can increase the exercise response (19), highlighting the complexity of reduced activity on the acute exercise response. The main findings from this study show 1) a single day of inactivity alters substrate utilization in a 60-minute aerobic exercise bout, and 2) inactivity resulted in differences at the RNA level at rest and post-exercise, which are related to inflammation, metabolism, cell cycle, and innate immunity. These findings further highlight the need to investigate the nuances of physical activity’s role in enhancing the exercise response.

Inactivity altered systemic metabolism during exercise

In the current study, systemic fat metabolism increased over the 60-minute exercise bout while CHO utilization decreased, a result which is to be expected with aerobic exercise (33). However, it appeared that this systemic shift from CHO to fat was less towards the end of exercise in the IN trial, as CHO utilization and respiratory exchange ratio at minute 60 were higher than in the normal activity trial. Past studies have shown an attenuation of TG oxidation after an hour-long aerobic exercise bout in response to a high-fat/glucose tolerance test (7, 9, 20), this study also highlights altered systemic metabolism within aerobic exercise. Respiratory gas exchange was not monitored after exercise; however, TGs were statistically higher in the IN trial. Given that the current study had only one day of inactivity, this study highlights a shift in elevated TG beginning with a single day of inactivity. No other differences between trials were found in blood metabolite analysis, highlighting the unique effect of inactivity-induced exercise resistance on TG oxidation.

Basal RNA expression with a single day of inactivity

Comparing the pre-exercise RNA profiles provides insight into the impact of a single day of inactivity on the skeletal muscle, which may explain the systemic metabolic differences seen during exercise. To explore potential mechanisms, 494 DEGs were found by comparing the pre-exercise biopsies. Seven of these nine were upregulated with inactivity, including four ribosomal RNA transcripts, RN7SK, PDP1, and APLN. RN7SK is a small nuclear transcript that appears to be related to cell differentiation and cell cycle (34). PDP1, Pyruvate Dehydrogenase Phosphatase 1, is upregulated by calcium during muscle contraction and responds to metabolic demands (35). APLN encodes for the protein Apelin, which has been shown by others to be a potent myokine that stimulates a broad range of mostly beneficial effects, including muscle function and activation of AMP-activated protein kinase (AMPK) (36, 37). AMPK upregulates lipoprotein lipase activity, which is a key enzyme involved in TG oxidation that is downregulated with inactivity (9, 38). Also noteworthy is that other studies have also shown APLN is slightly downregulated in skeletal muscle with inactivity (39), suggesting Apelin may be involved in the blunted TG oxidation with inactivity.

Other examples of DEGs are MTND1P23 and TNFRSF12A, which are downregulated with inactivity. MTND1P23, mitochondrially encoded NADH:ubiquinone oxidoreductase core subunit 1 pseudogene 23, has recently been shown to be related to cancer (40) and epilepsy (41), suggesting a potential function that is unknown. TNFRSF12A, TNF Receptor Superfamily Member 12A, has previously been implicated as responsive to resistance exercise and regulates muscle growth (42–44), and therefore a noteworthy DEG which may impact skeletal muscle health with inactivity. Together, these DEGs provide an interesting starting point to mechanistically discover the role of inactivity on altered metabolic function.

With the 494 DEGs found in the baseline comparisons, 24 enriched pathways were significantly altered. These pathways were largely related to immune signaling (allograft rejection, interferon responses, complement, IL6-JAK-STAT3 signaling, inflammatory response) as well as metabolic and growth-related programs (Oxidative Phosphorylation, mTORC1 signaling, PI3K-AKT-mTOR signaling). The transcriptomic profile results suggest that subtle, coordinated molecular differences are detectable after a single day of inactivity at rest, even when only nine genes were classified as statistically significant by the DEG analysis. Several of the DEGs and enriched pathways are important factors in the exercise response and therefore warrant more research on the mechanistic influence of these genes on the altered fat metabolism with exercise.

Transcriptomic Response to Exercise Between Inactive and Normal Activity

A primary aim of this study was to investigate how a single day of inactivity affects the skeletal muscle’s response to exercise. RNA sequencing revealed that the IN trial nearly doubled the amount of DEGs compared to the A trial. To further characterize the exercise response in these two trials, KEGG analysis revealed 35 and 17 pathways in the IN and A trials, respectively. In addition, GSEA Hallmark analysis revealed 33 and 24 enriched pathways, respectively. Many of these pathways are shared between trials, including stress responses, inflammatory signaling, metabolic remodeling, and tissue adaptation. These shared pathways highlight a conserved core of exercise-responsive biology across activity conditions, which is consistent with what we would anticipate with aerobic exercise (39). The inactive trial showed a broader array of responses in cell signaling, inflammation, and metabolic pathways. On the other hand, the normal activity trial also had unique pathways, including adipogenesis.

The interaction contrast yielded 22 significantly enriched Hallmark gene sets, despite showing no significant DEGs or KEGG pathways at the single-gene level. Enriched gene sets included: Immune and inflammatory signaling, Cell growth and cell cycle regulation, and Metabolism and differentiation. This profile indicates that while both groups responded to exercise, the magnitude and coordination of response differed between activity groups in immune-related and cell growth gene sets. These differences were not large enough at the gene level to yield statistically significant interaction terms, but pathway-level analysis revealed distributed shifts across many genes belonging to shared biological themes related to exercise.

Integrating the Global Exercise Response

The greater reliance on CHO at the end of exercise in the IN trial and altered RNA profile both at rest and post-exercise highlight the connection between systemic metabolism and exercise response. While this study did not measure metabolic substrates within the muscle, it is possible that a day of reduced activity resulted in higher muscle glycogen in the IN trial than the A trial. Glycogen availability has been shown to impact exercise metabolism and the global exercise response (45, 46) and therefore needs to be considered in the current study.

Subjects were instructed to maintain a similar diet before both trials, while taking 7,438 fewer steps in trial IN. This decreased step count is ~5.6 kilometers (assuming average stride length of 0.75 meters) and similar to ~30 minutes a day of moderate to vigorous physical activity (39). Additionally, the subjects likely decreased their non-exercise physical activity (i.e., non-exercise activity thermogenesis, NEAT) the day before trial IN. NEAT can increase resting energy expenditure by ~10% (47, 48). Given that the subjects were free-living, their 210% step reduction between the trials likely resulted in a similar percentage decrease of NEAT. Adults typically spend 6 hours doing non-exercise physical activity (48), thus subjects in the current study may have decreased their NEAT by ~4 hours the day before the IN trial.

The decreased caloric expenditure with the reduction in step count (~200 kilocalories) and NEAT (~50 kilocalories) likely resulted in a ~250 kilocalorie surplus in the inactive compared to the active trial (47, 48). This kilocalorie surplus may have resulted in augmented intramuscular glycogen storage. It is possible that the difference in muscle glycogen influenced the basal transcriptome and did not affect systemic metabolism until late in exercise (>40 minutes). Stephens et al. reported insulin dysregulation with one day of inactivity, which was moderately ameliorated when the subjects were isocaloric (36). Together, this highlights the link between inactivity, caloric consumption, and inactivity-induced exercise resistance, which warrants more research.

Limitations

This study sought to comprehensively characterize the exercise response after a single day of inactivity. We designed the study similarly to other exercise response studies, which controlled for diet and similar muscle biopsy time courses (7, 24, 49). As such, limitations of this study exist, including controlling diet to minimize a caloric surplus during inactivity (31). Additionally, it is apparent that molecular events regarding metabolism likely existed during exercise; therefore, timed muscle biopsies immediately following exercise would provide a better snapshot of these intramuscular metabolic events. This study also did not measure other hormonal and molecular events within the blood (i.e., insulin, myokines, and inflammatory markers) and muscle (metabolomics, protein regulation, phosphoproteome), which may provide greater insight into understanding the comprehensive effects of inactivity on the exercise response. Lastly, due to the low number of female subjects, sexes were combined. While the females in the study had data similar to males, others have published sex differences in molecular response to exercise (17, 50), and therefore it would be interesting to see if sex differences are apparent with inactivity. Nonetheless, this study provides a building block for future studies as we indeed demonstrate differences in the exercise response solely due to decreased step count the day prior. While step count is most commonly used to measure physical activity, other descriptors to measure sedentary time (i.e., caloric expenditure, NEAT, or step intensity (31, 48, 51)) would be of value to determine the effects of reduced activity on the exercise response.

Summary

In conclusion, this study highlights the influence that a single day of inactivity has on the acute exercise response. Our data highlights differences in the RNA profile at rest and after exercise in pathways related to inflammation, metabolism, cell cycle, and innate immunity that warrant further research. These transcriptomic differences may be a result of the altered systemic metabolism that occurred between trials. While this data has significant implications for exercise prescription, it also highlights the need to control for step count prior to exercise response investigations and can shed light on the role of physical activity in studies related to aging, obesity, and diseases. All of these factors are associated with decreased physical activity and therefore warrant more research on whether decreased physical activity is primary or secondary to metabolic or exercise responsiveness differences in varying physical states.

Supplementary Material

Supplementary Tables S1-S5: https://figshare.com/s/efa9b7e6cf2c20e86f3e

Figure 5.

Gene expression changes within skeletal muscle 4-hours after a bout of aerobic exercise. The RNA profile was compared post vs. pre-exercise in the Inactive (<5,000 steps) and Active (>8,500 steps) trials. Differentially expressed genes (DEG) were defined as false discovery rate <0.05. In total, 1403 and 793 genes were altered with exercise in the inactive and active trials, respectively (Supplemental Table S1). Of these genes, 397 were shared between the trials (A). Volcano plots are shown in the Active (B) and Inactive (C) trials. Red indicates down-regulated genes, while green represents genes that were up-regulated with exercise. The top five genes that were up-regulated and down-regulated are listed. We then compared gene expression with nine of the most common genes to be influenced by acute aerobic exercise that have been shown by the MetaMEx database (39) to provide context of the exercise response in the current study.

DATA AVAILABILITY

RNA sequencing data are available at the GEO repository under accession number GSE305038.