Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Jun 22.
Published in final edited form as: J Appl Physiol (1985). 2025 Jan 7;138(2):439–449. doi: 10.1152/japplphysiol.00507.2024

Timing of Resistance Exercise and Cardiometabolic Outcomes in Adults with Prediabetes: A Secondary Analysis

Jason V Thomas 1, Brenda M Davy 2, Richard A Winett 3, Christopher M Depner 1, Micah J Drummond 4, Paul A Estabrooks 1, Sheetal Hardikar 5,6, Zhining Ou 7, Jincheng Shen 5,7, Tanya M Halliday 1
PMCID: PMC12182900  NIHMSID: NIHMS2086608  PMID: 39773011

Abstract

Objective:

The objective of this study was to explore if the time of day (AM vs PM) resistance exercise is performed influences glucose and insulin concentrations, body composition, and muscular strength in adults with prediabetes.

Methods:

A secondary data analysis was conducted using data from the "Resist Diabetes" study, a phase II exercise intervention. Participants (Age:59.9±5.4 yrs; BMI:33±3.7 kg/m2) with prediabetes and overweight or obesity were categorized into AM (N=73) or PM (N=80) exercisers based on when they completed all of their supervised exercise sessions during a 12-week, 2x/week resistance exercise intervention. Blood glucose and insulin derived from oral glucose tolerance tests, body composition, and muscular strength were assessed pre and post resistance exercise training. Inverse propensity score weighting approach was used to estimate the efficacy of AM/PM exercise on the change of clinical responses. Paired samples t-test was used to compare pre-/post outcomes within AM/PM group.

Results:

No differences between AM and PM exercisers were detected in the change in glucose or insulin areas under the curve (AUC), body composition, or muscular strength. When exploring within-group changes, PM exercisers reduced glucose AUC (change: −800.6 mg/dl*120 min; p=0.01), whereas no significant change was detected for AM exercisers (change: −426.9 mg/dl*120 min; p=0.26). Only AM exercisers increased fat-free mass (change: 0.6 kg; p=0.001).

Conclusions:

The time of day of resistance exercise is performed may have some impact on glucose concentrations and body composition response. Future randomized clinical trials are needed to understand how exercise timing influences cardiometabolic outcomes in at-risk adults.

Keywords: Aging, Exercise Timing, Prediabetes, Resistance Exercise

Graphical Abstract

graphic file with name nihms-2086608-f0001.jpg

New and Noteworthy:

In this secondary analysis, there was no difference between AM and PM exercisers in blood glucose, insulin, body composition, or muscular strength following 12 weeks of supervised exercise. However, examining within-group changes, glucose AUC was significantly reduced in PM exercisers, but not in AM exercisers.

INTRODUCTION

Type 2 Diabetes (T2DM) remains a significant public health concern, with an estimated ~500 million diagnosed and undiagnosed cases worldwide (1). Obesity and many lifestyle-related risk factors contribute to T2DM development, including dietary factors, sedentary behavior, and low quantities of physical activity (PA) (2, 3). Lifestyle modifications, including exercise interventions, are widely recognized as first-line non-pharmacologic strategies to prevent and treat T2DM (4, 5). Exercise can confer many short and long-term protective benefits, including improvements in body composition and glycemic control (6, 7). Many health organizations have established evidence-based aerobic and resistance exercise guidelines to help prevent and treat T2DM. Typically, these guidelines contain principles for frequency, intensity, duration, type, volume, and exercise progression (4, 8). However, no guidelines exist for the time of day that exercise occurs.

The time of the day exercise occurs may be an underappreciated factor for those at risk of T2DM due to the influence of endogenous circadian rhythms on insulin sensitivity and glycemic regulation (9, 10). Studies have demonstrated that whole-body glycemic regulation and insulin secretion exhibit rhythmic fluctuations over the ~24-hour day (11, 12). When assessed after 8-12 hours of fasting, glucose tolerance and insulin sensitivity are typically worse in the evening than in the morning, which can be attributed to reduced skeletal muscle insulin sensitivity later in the day (12, 13). Because exercise can stimulate glucose uptake through insulin-independent pathways, conducting exercise training at different times of day may differentially impact metabolic pathways and influence physiological response to exercise.

The effects of exercise at different times of the day (herein referred to as exercise timing) in clinical populations are poorly understood (14). Recent findings demonstrate that PM, compared to AM exercise, may have a more positive impact on glucose and insulin outcomes in adults with and without T2DM when assessed with continuous glucose monitoring and clamp techniques (15-18); although conflicting results have also been reported (19, 20). Additionally, secondary analyses suggest that early-day exercise may be more beneficial for weight loss, weight maintenance, and changes in body composition (21, 22). However, much of the existing exercise timing literature has utilized aerobic exercise. Few studies have assessed the impact of resistance exercise timing on clinically relevant outcomes for a population at risk for T2DM (16, 19). Resistance exercise is an important modality for preventing and managing T2DM, particularly in older adults who experience age-related muscle loss and worsening glucose tolerance (23). Understanding the relationship between resistance exercise timing and protective physiological adaptations may help inform exercise prescriptions for T2DM prevention.

Previously, we demonstrated that older adults with prediabetes and obesity who performed 12 weeks of 2x/week supervised resistance exercise increased muscular strength, reduced waist circumference and body fat percent, increased lean mass, and reduced prevalence of prediabetes (24). However, there was a wide variation in responsiveness to the exercise intervention (25). For example, while 120-min glucose following a 75g oral glucose tolerance test (OGTT) was reduced by 10.2 mg/dl across the entire sample at 3 months, 31% of the sample increased 120-min glucose levels, and the change scores ranged from −101.65 to 85.5 mg/dl. Similarly, while body fat percent decreased by 0.6% across the sample, 33% of the sample increased body fat percent, with change scores ranging from −4.1% reduction to 2.6% increase. Given the potential for exercise timing to influence physiological outcomes, the purpose of this analysis is to explore if the time-of-day resistance exercise was performed differentially impacted blood glucose concentrations, insulin sensitivity, body composition, and muscular strength.

Methods:

This study was registered at ClinicalTrials.gov (NCT 01112709). The Virginia Tech Institutional Review Board provided ethical approval for the original study. All study participants provided written informed consent prior to enrollment.

Participants:

This analysis utilized data from the "Resist Diabetes" phase 2 clinical trial in which participants underwent a 12-week supervised resistance exercise program, followed by a 6-month faded contact phase and a 6-month no-contact follow-up phase. Only data from the initial 12-week supervised exercise phase were utilized for the present analysis. Full inclusion and exclusion criteria and detailed methods for the "Resist Diabetes" trial have been previously published (26). Briefly, middle to older aged adult men and women (age: 50-69 years) with overweight or obesity (BMI: 25-39.9 kg/m2) who were sedentary or minimally active (<120 min/week moderate PA or <60 min/week vigorous PA), and had prediabetes (defined as: impaired fasting glucose [IFG; fasting glucose concentration= 95-125 mg/dl]; impaired glucose tolerance [IGT; 120-min glucose concentration= 140-199 mg/dl in response to a standard 75g 2-hour OGTT; or combined IFG and IGT), but otherwise were healthy (no indication of major disease and received physician clearance), were eligible for enrollment (5, 24).

Resistance Exercise Protocol:

Following baseline testing, participants completed a 12-week, 2x/week resistance exercise program. The ~45-minute, full-body training protocol consisted of 1 set of 12 exercises using Nautilus machines (leg press, leg extension, seated leg curl, calf raise, chest press, latissimus dorsi pulldown, row, shoulder press, seated dip, lower back, abdominal crunch, and rotary torso) to target the major muscle groups. Participants completed 8-12 repetitions for each exercise and utilized a 3-second concentric/3-second eccentric contraction on all exercises with a high degree of effort (9-10/10 Rating of Perceived Exertion) or concentric failure. A minimum of one rest day occurred between resistance exercise sessions. At least 48 hours elapsed between completing the last exercise training session and the 12-week post-testing battery. All training sessions occurred between 0600 and 2000 and were supervised by American College of Sports Medicine (ACSM) certified personal trainers.

Exercise Timing Categorization:

Per the "Resist Diabetes" trial protocol, participants were required to complete at least 17 of the possible 24 training sessions (70% attendance) to stay enrolled in the trial. Only those meeting that requirement were included in this analysis. Participants were categorized into either AM (time of day: 0600-1159) or PM (time of day: 1200-1830) exercisers based on the time of day they started 100% of their exercise sessions. Exercise time was self-selected based on participant preference and trainer availability. 1200 (noon) was selected as the cutoff in this analysis to allow approximately equivalent time windows for AM and PM exercise, allow for a similar number of participants per group, and is consistent with previous approaches (27, 28). Participants who did not complete 100% of their exercise sessions within the same time category were excluded from the analysis (N=5), to avoid any potential effects of inconsistent exercise timing on outcome measures.

Outcome Measures:

The current analysis utilizes measurements collected at baseline (‘pre’) and following the 12 weeks of resistance exercise (‘post’) that were a part of the larger Resist Diabetes trial (24, 26). Following abstinence from exercise of at least 48 hours and an overnight fast, participants reported to the research facility between 0700-0900 to complete a 2 hour, 75-gram OGTT (29). Blood samples were collected in the fasted state and 10, 20, 30, 60, 90, and 120 minutes after a 75g glucose bolus was consumed. When catheter insertion was unsuccessful, a butterfly needle was used for blood collection at minutes 0 and 120 (n=56 pre, 52 post). Samples were analyzed for glucose (YSI 2700 Select Glucose Analyzer; YSI Life Sciences, Yellow Springs OH), insulin and c-peptide (ELISA, ALPCO IMMULITE, Siemens). Area under the curve (AUC) was calculated using the trapezoidal method for all blood markers (30). Insulin resistance was calculated using the Homeostatic Model of Insulin Resistance (HOMA-IR) (31) and insulin sensitivity was calculated using the Matsuda Insulin Sensitivity Index (ISI) (32). Absolute and relative fat mass and fat-free mass were assessed using dual-energy X-ray absorptiometry (DXA; GE Lunar Prodigy, Madison, WI). Upper and lower body muscular strength were tested using a 3-repetition max on chest and leg press, respectively (33).

Statistical Analyses:

Demographic and clinical outcomes were summarized using means and standard deviations (SD) for continuous variables and counts and percentages for categorical variables. We compared these variables between the two groups, AM or PM exercisers, using the nonparametric Wilcoxon rank sum test for continuous variables and chi-square or Fisher's exact test for categorical variables. We also presented descriptive statistics on the AM and PM groups, respectively. In within group analysis, we compared the continuous variables between baseline and 12-week time points using paired t-tests.

We conducted an inverse propensity score weighting based analysis to estimate the efficacy of AM/PM exercise on the change in clinical responses, where change is defined as the difference in the outcome measure at 12 weeks vs. baseline (12-week minus baseline). The first step was to obtain propensity scores of exercise timing (AM/PM) treatment from a multivariable logistic regression model, where the predictors included the following baseline variables: age (in years), glucose concentration (mg/dl), insulin concentration (uIU/mL), BMI (kg/m2), use of antidepressants or blood pressure medicine (Yes/No), biological sex (male/female), full-time employment status (fulltime/not fulltime), prediabetes status (IFG, IGT, IFT+IGT), body fat percent, chest press (kg), and leg press (kg). These propensity scores estimate the likelihood of someone being an AM vs. a PM exerciser and could be used to balance the groups to adjust for baseline differences and reduce the likelihood of selection bias given the non-randomized nature of this study. Inverse propensity score weight (IPW) were calculated for each exerciser (34). Balance was assessed using absolute standardized differences for each baseline variable, a threshold of 0.2 was employed to indicate if balance was achieved (35). Next, our outcome model estimated the impact of AM/PM exercise on the change in clinical responses in a weighted linear regression using the IPTW weights. We reported regression coefficients and their 95% confidence intervals (CIs) and p-values. For outcomes that have larger amount of missing data (>30%), we excluded subjects with missing values from the analysis and report the modeling results for the purpose of completeness instead of the primary results. Statistical significance was assessed at the 0.05 level. Analyses were conducted using R software v.4.1.2 (The R Foundation, Vienna, Austria).

RESULTS

Participant Characteristics and Resistance Exercise Session Adherence

153 subjects met the criteria to be included in the analysis. Baseline characteristics are presented in Table 1. AM (n=73) exercisers were slightly older than PM (n=80) exercisers (AM: 60.9 ±5.3yrs vs.PM: 58.7±5.2yrs; p= 0.008) and were less likely to be employed full-time (AM: 50% full-time employment; PM: 83.8% full-time employment; p<0.001), but the groups were otherwise well-balanced on baseline characteristics. Adherence to the 2x/week resistance exercise protocol did not differ between groups (AM: 91.3% vs PM 89.4%, p= 0.23). The frequency distribution of each participant's average starting time of their exercise sessions is displayed in Figure 1. The majority of AM exercisers (83%) started their sessions between 0700-1000, while the majority of PM exercisers (80%) started exercise between 1500-1800. Notably, there was high with-in participant consistency on the start time of their exercise, with >97% of both AM (72/73) and PM (78/80) exercisers starting all exercise sessions at the same clock hour.

Table 1:

Baseline Characteristics by Group

Variable AM (N=73) PM (N=80) P-value
DEMOGRAPHICS
Age, mean (SD), years 60.9 (5.3) 58.7 (5.2) 0.007w
Sex, n (%) Male 24 (32.9%) 19 (23.8%) 0.21c
Female 49 (67.1%) 61 (76.2%) -
Race White 69 (94.5%) 74 (92.5%) 1.00f
African American 4 (5.5%) 5 (6.2%) -
Asian 0 (0%) 1 (1.2%) -
Ethnicity Non-Hispanic 72 (98.6%) 79 (98.8%) 1.00f
Hispanic 1 (1.4%) 1 (1.2%) -
Employment Status Retired, part-time, or Unemployed 37 (50.7%) 13 (16.2%) <0.001c
Full-time 36 (49.3%) 67 (83.8%) -
BLOOD MEASURES
Fasting glucose, mean (SD), mg/dl 101.9 (8.7) 101.3 (7.9) 0.90w
Fasting insulin, mean (SD), uIU/dl 12.4 (12.4) 11.6 (12.7) 0.58w
Prediabetes Status, n (%) IFG 30 (41.1%) 42 (52.5%) 0.26c
IGT 9 (12.3%) 11 (13.8%) -
IFG+IGT 34 (46.6%) 27 (33.8%) -
ANTHROPOMETRICS
Weight, mean (SD), kg 93.2 (14.5) 93.2 (11.8) 0.84w
BMI, kg/m2 32.9 (3.9) 33.1 (3.5) 0.72w
Percent Body Fat, % 43.3 (7.1) 43.8 (6.5) 0.92w
MUSCULAR STRENGTH
Chest Press*, kg 73.8 (28.8) 75.4 (23.1) 0.20w
Leg Press, kg 311.2 (85.6) 314.2 (74.7) 0.89w
Exercise Adherence, % 91.3% (6.7) 89.4% (8.2) 0.23w
Open in a new tab

Variables expressed as means (SD) or frequency (%)

*

Missing values: baseline chest press= 1 PM.

w

Wilcoxon rank sum test

c

Chi-squared test

f

Fisher's exact test.

FIG 1: Distribution Of Exercise Training Times.

FIG 1:

Open in a new tab

Distribution of exercise start time. The x-axis displays the time of day on a 24- hour clock (0600= 6am; 1800= 6pm). The y-axis displays the number of participants who started their exercise sessions during the specified hour. Black bars represent AM exercisers. Greys bars represent PM exercisers.

Glucose and Insulin:

Table 2 depicts the results of the inverse probability weighted univariable linear regression models for each outcome variable. There was no difference in the change in glucose AUC (β coefficient= 373.7; 95%CI: −562.6,1310; p= 0.43), fasting glucose (β coefficient= 1.4, 95%CI: −1.2, 4.1; p= 0.29), and 120-minute glucose (β coefficient= 0.7; 95%CI: −9.0,10.5; p= 0.88) between AM and PM exercisers. Likewise, there was no difference in the change in insulin AUC (β coefficient= −387.3; 95%CI: −2250.9, 1476; p= 0.68), fasting insulin (β coefficient= −0.8; 95%CI= −3.9,2.4; p= 0.64), 120-min insulin (β coefficient= 15.6; 95%CI:−11, 42.3; p= 0.25), Matsuda ISI (β coefficient=0.7; 95% CI: 0, 1.4; p=0.06) and HOMA-IR (β coefficient= −0.1; 95%CI: −1, 0.7; p=0.73) between AM and PM exercisers. There was no difference in c-peptide AUC (β coefficient= −73.2; 95% CI: −240.5, 94.2; p=0.39), fasting c-peptide (β coefficient= 0.2, 95% CI: −240.5, 94.2), or 120-min c-peptide (β coefficient= −0.3; 95%CI: −2.5,1.9; p= 0.81). Out of the participants with complete glucose AUC data from the OGTT, 90-minute blood glucose was reduced more in PM than in AM exercisers (β coefficient= 13.5, 95% CI:1.1, 25.8; p= 0.03). AM exercisers reduced 10-minute insulin more than PM exercisers (β coefficient= −6.9, 95% CI:−13.6, −0.2; p= 0.05). No other differences in the change in blood glucose or insulin concentrations at any other time point were detected between AM and PM exercisers.

Table 2:

Outcome models for continuous outcomes between AM and PM exercisers.

Outcome variable Number of
observations		 β-coefficient
(PM as
reference)		 95% CI p-value
GLUCOSE
Glucose AUC, mg/dl*120min 100 373.7 (−562.6, 1310.0) 0.43
Fasting Glucose, mg/dl 153 1.4 (−1.2, 4.1) 0.29
10 min Glucose, mg/dl 91 −2.6 (−8.4, 3.1) 0.36
20 min Glucose, mg/dl 92 −4.2 (−11.7, 3.4) 0.28
30 min Glucose, mg/dl 92 −1.0 (−10.3, 8.3) 0.83
60 min Glucose, mg/dl 92 8.0 (−4.2, 20.1) 0.20
90 min Glucose, mg/dl 92 13.5 (1.1, 25.8) 0.03
120 min Glucose, mg/dl 152 0.7 (−9.0, 10.5) 0.88
INSULIN
Insulin AUC, uIU/dl*120 min 95 −387.3 (−2250.9, 1476.4) 0.68
Fasting Insulin, uIU/dl 153 −0.8 (−3.9, 2.4) 0.64
10 min insulin, uIU/dl 67 −6.9 (−13.6, −0.2) 0.05
20 min insulin, uIU/dl 68 −8.0 (−23.2, 7.2) 0.30
30 min insulin, uIU/dl 68 −5.0 (−27.1, 17.1) 0.65
60 min insulin, uIU/dl 65 4.6 (−20.2, 29.4) 0.71
90 min insulin, uIU/dl 64 4.6 (−28.9, 38.1) 0.78
120 min insulin, uIU/dl 109 15.6 (−11.0, 42.3) 0.25
C-peptide AUC, ng/mL*120min 98 −73.2 (−240.5, 94.2) 0.39
Fasting C-peptide 153 0.2 (−0.3, 0.7) 0.43
120min C-peptide 153 −0.3 (−2.5, 1.9) 0.81
Matsuda Insulin Sensitivity Index (ISI) 58 0.7 (0.0, 1.4) 0.06
HOMA insulin resistance, mg 153 −0.1 (−1.0, 0.7) 0.73
BODY COMPOSITION
Weight, kg 153 −0.1 (−0.8, 0.5) 0.72
Fat Mass, kga 153 −0.2 (−0.7, 0.3) 0.47
Percent Fat Mass, %a 153 −0.3 (−0.7, 0.1) 0.19
Fat-Free Mass, kga 153 0.2 (−0.3, 0.8) 0.43
Percent Fat-Free Mass, %a 153 0.3 (−0.1, 0.7) 0.17
MUSCULAR STRENGTH
Chest Press, kgb 152 2.6 (−2.3, 7.6) 0.29
Leg Press, kgb 151 −2.8 (−18.3, 12.6) 0.72
Open in a new tab

Bolded values indicate a significant estimated effect of resistance exercise timing

a

Assessed via DXA

b

Assessed via 3-repetition maximum

Table 3 depicts the comparison of pre to post testing for AM and PM exercisers. When within-group changes in AM and PM participants were explored separately, there was no significant change in glucose AUC in the AM exercisers (pre= 19309.4±2635.4 mg/dl*120 min; post= 18874.7±3064.5 mg/dl*120 min; p= 0.24; Figure 2A). AM exercisers experienced a reduction in blood glucose at the 120-min time point (pre= 147.2±33.3; post= 137.4±36.9 mg/dl; p= 0.01), but not at any other time point after 12 weeks of resistance exercise (Figure 2B). In contrast, there was a significant reduction in glucose AUC in the PM exercisers (pre= 18730.8±3155.2; post= 17961.3±3098.6 mg/dl*120 min; p= 0.014), with significant reductions noted at the 60 (pre= 183±32 mg/dl, post= 171.2±36.9 mg/dl; p= 0.024), 90 (pre= 71.1 ±40; post= 152.5±39.6 mg/dl; p<0.001), and 120-minute timepoints (pre= 141.2±37.6; post= 130.7±34.5 mg/d; p<0.001) after 12 weeks of resistance exercise. (Figure 2C).

Table 3.

Summary of within-group outcomes for AM and PM exercise

AM EXERCISE (N=73) PM EXERCISE (N=80)
Variable Mean (SD) Mean (SD) Mean (SD) Mean (SD)
Pre (N=73) Post (N=73) Change P-value Pre (N=80) Post (N=80) Change P-value
GLUCOSE
Glucose AUC, mg/dl*120min 19309.4 (2635.4) 18874.7 (3064.5) −426.9 (2515.7) 0.26 18730.8 (3155.2) 17961.3 (3098.6) −800.6 (2225.8) 0.01
Fasting Glucose, mg/dl 101.9 (8.7) 101.7 (9.9) −0.17 (8.78) 0.87 101.3 (7.9) 99.8 (10.0) −1.6 (7.6) 0.07
120min glucose, mg/dl 147.2 (33.3) 137.4 (36.9) −9.8 (33.3) 0.01 141.2 (37.6) 130.7 (34.5) −10.6 (26.7) <0.001
INSULIN
Insulin AUC, uIU/dl*120 min 10944.5 (6944.1) 11652.9 (7591.0) 164.7 (3594.7) 0.76 10793.0 (7791.7) 11786.5 (8604.2) 551.9 (5479.2) 0.48
Fasting Insulin, uIU/dl 12.4 (12.7) 15.5 (11.4) 3.1 (7.9) 0.005 11.6 (12.7) 15.4 (12.5) 3.8 (11.7) 0.005
120 min insulin, uIU/dl 149.3 (124.4) 150.4 (120.7) 2.7 (65.3) 0.76 140.6 (130.9) 131.3 (121.2) −12.9 (75.9) 0.21
C-peptide AUC (ng/mL*120min) 1542.2 (648.8) 1556.5 (602.4) 34.7 (413.5) 0.56 1460.242 (677.7) 1583.586 (618.8) 107.9 (426.6) 0.09
Fasting C-Peptide (ng/mL) 4.0 (1.8) 4.0 (1.7) 0.0 (1.1) 0.87 4.0 (2.4) 3.8 (1.7 −0.2 (2.0) 0.42
120 min C-peptide (ng/mL) 16.8 (7.9) 17.2 (7.8) 0.4 (6.1) 0.60 15.7 (9.6) 16.4 (7.5) 0.65 (7.7) 0.45
Matsuda ISI 3.1 (2.0) 3.4 (2.6) 0.27 (1.3) 0.26 3.0 (1.8) 2.7 (1.8) −0.4 (1.3) 0.13
HOMA IR, mg 3.2 (3.3) 4.0 (3.2) 0.8 (2.3) 0.003 2.9 (3.2) 3.9 (3.2) 1.0 (2.9) 0.005
BODY COMPOSITION
Weight, kg 93.1 (14.5) 93.0 (14.5) −0.1 (1.9) 0.77 93.2 (11.8) 93.2 (12.3) 0.1 (2.3) 0.83
Fat Mass, kg 40.1 (8.4) 39.3 (8.7) −0.8 (1.7) <0.001 40.4 (7.1) 39.9 (7.2) −0.6 (1.7) 0.004
Percent Fat Mass, % 43.3 (7.1) 42.6 (7.0) −0.7 (1.3) <0.001 43.8 (6.5) 43.3 (6.6) −0.4 (1.3) 0.003
Fat-Free Mass, kg 52.5 (10.8) 53.1 (10.9) 0.6 (1.4) 0.001 52.3 (9.8) 52.6 (10.3) 0.4 (2.0) 0.10
Percent Fat-Free Mass, % 56.7 (7.1) 57.4 (7.0) 0.7 (1.3) <0.001 56.2 (6.5) 56.7 (6.6) 0.4 (1.2) 0.003
MUSCULAR STRENGTH
Chest Press, kg 73.8 (28.8) 95.4 (38.3) 21.6 (16.7) <0.001 75.4 (23.1) 94.4 (27.5) 19.6 (12.7) <0.001
Leg Press, kg 311.2 (85.6) 365.8 (95.4) 54.6 (45.8) <0.001 314.2 (74.7) 370.0 (78.0) 57.4 (50.4) <0.001
Open in a new tab

Missing values:

AM Pre-Intervention: Glucose AUC= 21, Insulin AUC = 25, 120 min Insulin= 18, C peptide AUC= 21, Matsuda ISI=41

AM Post-Intervention: Glucose AUC= 23, Insulin AUC= 24, 120 min Insulin= 5, C peptide AUC= 23, Matsuda ISI=28

AM Change Scores: Glucose AUC=23, Insulin AUC=27, 120 min Insulin = 17, C peptide AUC= 23, Matsuda ISI=43

PM Pre-Intervention: Glucose AUC= 28, Insulin AUC=29, Chest Press=1, 120 min Insulin= 21, C peptide AUC= 29, Matsuda ISI=50

PM Post-Intervention: Glucose AUC= 30, Insulin AUC= 30, Chest Press= 1, Leg Press= 2, 120 min Glucose=1, 120 min Insulin= 10, C peptide AUC= 31, Matsuda ISI=37

PM Change Scores: Glucose AUC= 30, Insulin AUC= 31, Chest Press=2, Leg Press= 2, 120 min Glucose= 1, 120 min insulin= 27, C peptide AUC= 32, Matsuda ISI=52

Fig 2: Glucose and Insulin After 3 Months of AM or PM Resistance Exercise.

Fig 2:

Open in a new tab

Blood glucose and insulin concentrations in response to 12 weeks of AM and PM resistance exercise. Total AUC change from pre to post test for glucose (Fig 2A) and insulin (Fig 2B). Graphs comparing the pre and post test OGTT outcomes for AM glucose (Fig 1C), AM insulin (FIG 1D), PM glucose (Fig 2E) and PM Insulin (Fig 2F). Black colors represent AM exercise. Grey colors represent PM exercise. Shapes represent the mean value. Solid lines represent baseline values. Dashed lines represent the post-test values. *p<0.05 within-group comparison. Whisker bars represent SEM.

There was no within-group change in insulin AUC from baseline to post-testing for either AM or PM exercisers (Figure 2D). Fasting insulin increased in both AM (pre= 12.4±12.4; post= 15.5±11.4 uIU/dl; p= 0.002) and PM exercisers (pre= 11.6±12.7; post= 15.4±12.5 uIU/dl; p= 0.005). No other within-group changes from pre to post were detected for insulin at any time point for either AM or PM exercisers (Figure 2E & 2F). Due to the slight increase in fasting insulin, a concomitant increase in HOMA IR was seen within both AM (pre= 3.2±3.3; post= 4.0±3.2; p= 0.003) and PM groups (pre= 2.9±3.2; post= 3.9±3.2; p= 0.005; Table 3). No within-group changes from pre to post were detected for the Matsuda ISI or c-peptide (Table 3).

Body Composition and Muscular Strength

There was no difference in the change in fat mass (β coefficient= −0.2, 95%CI: −0.7,0.3; p= 0.47), percent fat mass (β coefficient= −0.3; 95%CI: −0.7,0.1; p= 0.19), fat-free mass (β coefficient= 0.2, 95%CI: −0.3,0.8; p= 0.43), and percent fat-free mass (β coefficient= 0.3; 95%CI: −0.1,0.7; p= 0.17; all Table 2) between the AM and PM exercisers. When within-group changes in AM and PM participants were explored separately, only the AM group experienced an increase in fat-free mass (pre= 52.5±10.8kg; post= 53.1±11.0kg; p= 0.001; Table 3). Similarly, there was no difference in the increase in both 3-repetition maximum measured upper (β coefficient= 2.6, 95%CI: −2.3,7.6; p= 0.29) and lower body (β coefficient= −2.8, 95%CI: −18.3,12.6; p= 0.72) strength between AM and PM exercisers following the 12-week training period.

DISCUSSION

Due to an increased interest in how the timing of health-related behaviors impacts metabolic health outcomes, coupled with emerging evidence indicating exercise timing may differentially impact health outcomes, we conducted a secondary analysis of the previously completed Resist Diabetes study (24) to examine if differential responses were detected between AM and PM exercisers. Specifically, we examined if exercise timing over 12-weeks impacted the following outcomes in adults with overweight or obesity and prediabetes: blood glucose and insulin responses in the fasted state, as well as in response to a 2-hour OGTT, body composition, and muscular strength. Using a propensity score-matched model to account for the non-randomized nature of the data, we did not detect any statistically significant differences between AM and PM groups for the change in glucose or insulin AUC, body composition, or muscular strength. However, when exploring within-group changes, PM exercisers reduced glucose AUC, whereas AM exercisers experienced no significant change in glucose AUC. In addition, while both groups reduced fat mass, only the AM group increased fat-free mass.

To our knowledge, no other study has utilized a clinically relevant measure such as the OGTT to assess the impact of resistance exercise timing on changes to blood glucose and insulin concentrations. Overall, the time of day of exercise was performed did not influence the change in glucose area under the curve, fasting, or 120-minute glucose. However, there was a statistically significant within-group reduction in glucose AUC in the PM exercisers but not in AM exercisers. This decrease in AUC in the PM exercisers was primarily driven by reductions in blood glucose at 60, 90, and 120 minutes. This finding is consistent with prior work indicating exercise later in the day could result in greater glycemic-related health benefits. For example, in patients with T2DM, 2 weeks of HIIT cycling improved 24-hour blood glucose profiles when done in the afternoon but not in the morning (17). Additionally, analyses of the LOOK-AHEAD trial and UK biobank indicate that obtaining the majority of physical activity later in the day is associated with a 30-50% larger reduction in HbA1C (18) compared to early day exercise, as well as a lowest risk of mortality and cardiovascular disease (28). Thus, the reductions in glucose AUC seen within the PM group in this analysis may potentially impact long-term health and disease progression.

Fasting and 120-minute glucose are used as diagnostic thresholds for prediabetes and T2DM and, as such, were of interest in this analysis. Our finding that resistance exercise timing did not influence the change in fasting blood glucose is contrary to one other study on resistance exercise timing. In a randomized controlled trial of healthy older women (N=31; 66±4yrs), both morning (0730) and evening (1800) exercise groups reduced fasting glucose, and the evening exercisers had a greater reduction than the morning exercisers. In contrast, we only saw a trend towards improving fasting blood glucose in the PM group (p= 0.07) and no improvements in the AM group. This discrepancy could be due to differences in the sample population, exercise timing windows, and/or differences in the exercise protocol. The Resist Diabetes study enrolled men and women with prediabetes and as it was not initially designed to evaluate exercise timing, participants exercised across broad windows, whereas the trial by Krčmárová et al. enrolled healthy older women and as exercise timing was the intervention of interest, exercise occurred at singular AM and PM timepoints. While both exercise programs utilized a 2x/week full-body protocol done to concentric failure, the Resist Diabetes protocol required participants to perform 1 set for 12 exercises, and the study by Krčmárová et al. required participants to perform 3 sets of 8 exercises. It is plausible that a greater number of sets or greater exercise duration may confer a more pronounced impact on fasting blood glucose (36, 37).

Exercise typically improves insulin sensitivity in adults with prediabetes (38). However, only a few randomized controlled trials on exercise timing have evaluated insulin and insulin sensitivity as an outcome. Considering this gap, we also examined if 12-weeks of AM vs. PM resistance exercise impacted insulin and insulin sensitivity. We found that the time of day resistance exercise was performed had no impact on the change in insulin AUC, fasting insulin, 120-minute insulin, c-peptide AUC, Matsuda ISI, nor HOMA-IR. These findings align with a meta-analysis and other studies that report that fasting insulin and HOMA-IR are not differentially impacted by AM vs PM exercise (17, 19, 39). Since HOMA-IR estimates hepatic insulin resistance and glucose output from fasting glucose and insulin values alone (31), it is unlikely to be impacted by a resistance exercise timing intervention. Thus, we also calculated the Matsuda ISI, which is a measure of whole-body insulin sensitivity. While no difference between AM and PM exercise, nor change within AM or PM groups, was detected in our sample, this data is presented cautiously due to a high level of missingness (calculated on 58 of 153 participants) as a result of being unable to insert an IV catheter on all participants for complete blood collection throughout the OGTT. However, other investigations which utilized a hyperinsulinemic-euglycemic clamp to assess peripheral insulin sensitivity have found that skeletal muscle and adipose tissue insulin sensitivity improved more after combined aerobic and resistance exercise performed in the afternoon (1500-1800) compared to the morning (0800-1000) (16). As exercise has been shown to impact skeletal muscle clocks and glucoregulatory gene expression (40, 41), improvements in insulin sensitivity may be more detectable when evaluating skeletal muscle insulin sensitivity rather than whole body insulin sensitivity. Future trials evaluating how exercise timing impacts insulin sensitivity may benefit from assessments of both hepatic and peripheral insulin sensitivity to better evaluate any potential effects of exercise timing.

Our results on resistance exercise timing and body composition are similar to the previously discussed work of Krčmárová et al (42). They reported no difference in reductions in body fat between morning (0730) and evening (1800) exercise groups. They also reported that the AM group, but not the PM group, increased their percentage of fat-free mass. Likewise, our analysis shows fat mass was reduced by similar amounts in both AM and PM exercisers. The reduction in body fat in our analysis was accompanied by a significant increase in absolute fat-free mass of 0.6kg within AM exercisers and a non-significant 0.3kg increase in fat-free mass within PM exercisers. The magnitude of body weight and composition changes in our sample were expected, as it is well documented that resistance exercise typically results in less weight loss, less fat mass loss, and greater increases in fat-free mass than aerobic exercise (43). Thus, while the timing of resistance exercise may be important to consider in increasing fat-free mass, it may not have as great of an impact on total body weight and fat mass as the timing of aerobic exercise (21, 44). Additionally, the duration that exercise timing is maintained may also influence body composition. One study showed that 12 weeks of combined aerobic and resistance training in the AM or PM did not differentially impact vastus lateralis cross-sectional area in healthy young adults. Despite no difference at 12 weeks, when training continued out to 24 weeks, PM exercise (1630-2000) resulted in greater increases in vastus lateralis cross sectional area than morning exercise (0630-1000).(45). Thus there may be differences in body composition depending on the age of participants and/or duration that exercise timing is maintained. The potential benefit of exercise timing in older adults may have practical application in countering age-related loss of muscle mass and warrant further exploration.

Lastly, the improvements seen in upper and lower body muscular strength did not differ between AM and PM exercisers in our sample. Some evidence indicates that neuromuscular performance and anaerobic power are higher later in the day (46). This could allow for a higher training stimulus, and thus greater physiological adaptations to improve strength. Despite increases in fat-free mass seen in the AM exercisers, the time of day that resistance exercise occurred in this study appears to have no impact on the change in upper and lower body strength in an adult population with prediabetes and overweight/obesity. While we used a 3-repetition maximum test, others have reported similar changes in strength with a 6-repetition maximum (42) and 1-repetition maximum (45) muscular strength tests. Notably, there may be an effect between timing of training and timing of performance, or in this study, the 3 repetition max test, where increases in strength may more readily be observed at the same time of day that coincides with coincides with training time (47-49). Unfortunately, the time of day of the 3-repetition maximum test was conducted is not available for analysis.

While the timing of exercise may have had some impact on the assessed outcomes, other relevant but unmeasured factors, such as chronotype, and sleep may have influenced the outcomes in this secondary analysis. Participants' chronotype, or preference for early or later awakening and activity, could have influenced the selection of exercise timing and resultant classification into AM vs. PM groups. Further, chronotype has also been shown to influence insulin sensitivity, where morning types display better hepatic and peripheral insulin sensitivity(50), higher fat oxidation at rest and during aerobic exercise (51), glycemic control (52), and higher physical activity (53) compared to late types. Physical activity timing and chronotype may also interact to promote circadian phase shifts and alignment or misalignment of circadian rhythms (54) and sleep structure (53, 55). Poor sleep has been shown to blunt weight loss, reduce insulin sensitivity, and ultimately increase the risk of T2DM Poor sleep has been shown to blunt weight loss (56), reduce insulin sensitivity (57), and ultimately increase the risk of T2DM (58). While exercise in general is thought to improve sleep (59-61), it is unknown if participants sleep patterns were changed and moderated the outcomes. Lastly, other behavioral factors, including non-exercise physical activity and energy intake, may interact with exercise timing (62) and impact cardiometabolic outcomes in this analysis. Non-exercise physical activity and energy intake were subjectively assessed via online self-report and recall at baseline and 12 weeks in the Resist Diabetes study, but were not objectively measured (63, 64). Given the limitations of self-report, non-exercise physical activity and energy intake were not included in this analysis. The interactions between exercise timing, sleep, chronotype, and other modifiable behaviors require further investigation to fully appreciate how timing of behaviors impacts cardiometabolic health

Strengths of this analysis include: the use of a supervised resistance exercise program that met muscle-strengthening recommendations; inclusion of an at-risk population including men and women; and utilization of gold-standard and clinically-relevant outcome measures such as fasting and 120-minute glucose and insulin derived from a standard OGTT, and leveraging DXA to assess change in body composition. Lastly, participants were highly consistent in their exercise timing, allowing for 153 participants to be included in this analysis.

Despite these strengths, we acknowledge the following limitations: As this is a secondary analysis, participants were not randomized into AM and PM exercise groups. However, there were minimal differences between groups at baseline, and the use of the IPW analysis allows the estimation of intervention effect even with observational data. Second, our classification of AM vs PM exercise, as in prior investigations, is based upon clock time and not individualized markers of circadian timing (such as habitual wake/sleep cycles or dim-light melatonin onset). Participant chronotype nor other health related behaviors which may influence physiological response to exercise and exercise timing were not accounted for in this analysis.. While adherence to ACSM guidelines for resistance exercise (2x/week, full body, 1 set to failure) is a strength of the present study, it is also possible that the exercise protocol was an insufficient stimulus to determine if the habitual timing of resistance exercise influences blood glucose, insulin, body composition, and muscular strength outcomes. Lastly, we acknowledge that the differences in outcomes seen within-group may indicate a difference in nominal significance. This trial was not designed nor powered to detect differences in AM vs. PM exercise. Therefore, results should be interpreted with caution.

CONCLUSION

Resistance training is beneficial for older adults with overweight/obesity and prediabetes, regardless of exercise timing. No differences were detected between AM vs PM exercisers in cardiometabolic outcomes. However, there was a significant within group decrease in blood glucose AUC in the PM exercisers. This provides additional evidence to the literature that consistent exercise at a time when insulin sensitivity is reduced could be beneficial for glycemic control. However, given that the present manuscript presents a secondary analysis from a trial in which exercise timing was not randomized, results should be interpreted cautiously. Future, prospective, randomized clinical trials are needed to understand if the time of day that resistance exercise is conducted impacts cardiometabolic health outcomes.

ACKNOWLEDGEMENTS:

We would like to acknowledge and thank Resist Diabetes study team members and participants for providing the data for this analysis. This investigation was supported by the University of Utah Study Design and Biostatistics Center, with funding in part from the National Center for Research Resources and the National Center for Advancing Translational Sciences, National Institutes of Health, through Grant UL1TR002538 (formerly 5UL1TR001067-05, 8UL1TR000105 and UL1RR025764). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of Health. Figures were created using GraphPad Prism version 10 (GraphPad Software, Boston, Massachusetts, USA). The graphical abstract was created using Biorender.com (Created in BioRender. T, J. (2024) https://BioRender.com/e82a360).

GRANTS:

NIH NIDDK R01DK082383 (BMD, RAW), NIH NCATS 5KL2TR002539 (TMH), NIH NIDDK K01DK134800 (TMH), NIH NHLBI 5K01HL145099 (CMD), NIH NCI K07CA222060 (SH), NIH NCATS UL1TR002538

Footnotes

DISCLOSURES: None

DISCLAIMERS: None

CLINICAL TRIAL REGISTRATION: NCT 01112709

DATA AVAILABILITY:

Source data for this study are not publicly available due to privacy or ethical restrictions. The source data are available to verified researchers upon request by contacting the corresponding author. At the time of data collection for the Resist Diabetes study, data availability was not a prior publishing or NIH requirement. Data exists and is available, but not in a format that is easy to share or deposit.

References

  • 1.Ong KL, Stafford LK, McLaughlin SA, Boyko EJ, Vollset SE, Smith AE, et al. Global, regional, and national burden of diabetes from 1990 to 2021, with projections of prevalence to 2050: a systematic analysis for the Global Burden of Disease Study 2021. The Lancet 402: 203–234, 2023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Kolb H, and Martin S. Environmental/lifestyle factors in the pathogenesis and prevention of type 2 diabetes. BMC Med 15: 131, 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Weisman A, Fazli GS, Johns A, and Booth GL. Evolving Trends in the Epidemiology, Risk Factors, and Prevention of Type 2 Diabetes: A Review. Can J Cardiol 34: 552–564, 2018. [DOI] [PubMed] [Google Scholar]
  • 4.Kanaley JA, Colberg SR, Corcoran MH, Malin SK, Rodriguez NR, Crespo CJ, et al. Exercise/Physical Activity in Individuals with Type 2 Diabetes: A Consensus Statement from the American College of Sports Medicine. Med Sci Sports Exerc 54: 353–368, 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Knowler WC, Barrett-Connor E, Fowler SE, Hamman RF, Lachin JM, Walker EA, and Nathan DM. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med 346: 393–403, 2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Kirwan JP, Sacks J, and Nieuwoudt S. The essential role of exercise in the management of type 2 diabetes. Cleve Clin J Med 84: S15–s21, 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Colberg SR, Sigal RJ, Fernhall B, Regensteiner JG, Blissmer BJ, Rubin RR, et al. Exercise and Type 2 Diabetes: The American College of Sports Medicine and the American Diabetes Association: joint position statement. Diabetes Care 33: e147–e167, 2010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Colberg SR, Sigal RJ, Yardley JE, Riddell MC, Dunstan DW, Dempsey PC, et al. Physical Activity/Exercise and Diabetes: A Position Statement of the American Diabetes Association. Diabetes Care 39: 2065–2079, 2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Mancilla R, Krook A, Schrauwen P, and Hesselink MKC. Diurnal Regulation of Peripheral Glucose Metabolism: Potential Effects of Exercise Timing. Obesity 28: S38–S45, 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Gabriel BM, and Zierath JR. Circadian rhythms and exercise - re-setting the clock in metabolic disease. Nat Rev Endocrinol 15: 197–206, 2019. [DOI] [PubMed] [Google Scholar]
  • 11.Van Cauter E, Blackman JD, Roland D, Spire JP, Refetoff S, and Polonsky KS. Modulation of glucose regulation and insulin secretion by circadian rhythmicity and sleep. J Clin Invest 88: 934–942, 1991. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Sonnier T, Rood J, Gimble JM, and Peterson CM. Glycemic control is impaired in the evening in prediabetes through multiple diurnal rhythms. J Diabetes Complications 28: 836–843, 2014. [DOI] [PubMed] [Google Scholar]
  • 13.Verrillo A, De Teresa A, Martino C, Di Chiara G, Pinto M, Verrillo L, et al. Differential roles of splanchnic and peripheral tissues in determining diurnal fluctuation of glucose tolerance. Am J Physiol 257: E459–465, 1989. [DOI] [PubMed] [Google Scholar]
  • 14.Janssen I, Campbell JE, Zahran S, Saunders TJ, Tomasone JR, and Chaput JP. Timing of physical activity within the 24-hour day and its influence on health: a systematic review. Health Promot Chronic Dis Prev Can 42: 129–138, 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Moholdt T, Parr EB, Devlin BL, Debik J, Giskeødegård G, and Hawley JA. The effect of morning vs evening exercise training on glycaemic control and serum metabolites in overweight/obese men: a randomised trial. Diabetologia 64: 2061–2076, 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Mancilla R, Brouwers B, Schrauwen-Hinderling VB, Hesselink MKC, Hoeks J, and Schrauwen P. Exercise training elicits superior metabolic effects when performed in the afternoon compared to morning in metabolically compromised humans. Physiol Rep 8: e14669, 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Savikj M, Gabriel BM, Alm PS, Smith J, Caidahl K, Björnholm M, et al. Afternoon exercise is more efficacious than morning exercise at improving blood glucose levels in individuals with type 2 diabetes: a randomised crossover trial. Diabetologia 62: 233–237, 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Qian J, Xiao Q, Walkup MP, Coday M, Erickson ML, Unick J, et al. Association of Timing of Moderate-to-Vigorous Physical Activity With Changes in Glycemic Control Over 4 Years in Adults With Type 2 Diabetes From the Look AHEAD Trial. Diabetes Care 2023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Teo SYM, Kanaley JA, Guelfi KJ, Marston KJ, and Fairchild TJ. The Effect of Exercise Timing on Glycemic Control: A Randomized Clinical Trial. Med Sci Sports Exerc 52: 323–334, 2020. [DOI] [PubMed] [Google Scholar]
  • 20.Morales-Palomo F, Ramirez-Jimenez M, Ortega JF, and Mora-Rodriguez R. Effectiveness of Aerobic Exercise Programs for Health Promotion in Metabolic Syndrome. Med Sci Sports Exerc 51: 1876–1883, 2019. [DOI] [PubMed] [Google Scholar]
  • 21.Willis EA, Creasy SA, Honas JJ, Melanson EL, and Donnelly JE. The effects of exercise session timing on weight loss and components of energy balance: midwest exercise trial 2. Int J Obes (Lond) 44: 114–124, 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Alizadeh Z, Younespour S, Rajabian Tabesh M, and Haghravan S. Comparison between the effect of 6 weeks of morning or evening aerobic exercise on appetite and anthropometric indices: a randomized controlled trial. Clinical Obesity 7: 157–165, 2017. [DOI] [PubMed] [Google Scholar]
  • 23.Fragala MS, Cadore EL, Dorgo S, Izquierdo M, Kraemer WJ, Peterson MD, and Ryan ED. Resistance Training for Older Adults: Position Statement From the National Strength and Conditioning Association. J Strength Cond Res 33: 2019–2052, 2019. [DOI] [PubMed] [Google Scholar]
  • 24.Davy BM, Winett RA, Savla J, Marinik EL, Baugh ME, Flack KD, et al. Resist diabetes: A randomized clinical trial for resistance training maintenance in adults with prediabetes. PLoS One 12: e0172610, 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Winett RA, Davy BM, Savla J, Marinik EL, Winett SG, Baugh ME, and Flack KD. Using response variation to develop more effective, personalized behavioral medicine?: evidence from the Resist Diabetes study. Transl Behav Med 4: 333–338, 2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Marinik EL, Kelleher S, Savla J, Winett RA, and Davy BM. The resist diabetes trial: Rationale, design, and methods of a hybrid efficacy/effectiveness intervention trial for resistance training maintenance to improve glucose homeostasis in older prediabetic adults. Contemp Clin Trials 37: 19–32, 2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Van Der Velde JHPM, Boone SC, Winters-Van Eekelen E, Hesselink MKC, Schrauwen-Hinderling VB, Schrauwen P, et al. Timing of physical activity in relation to liver fat content and insulin resistance. Diabetologia 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Sabag A, Ahmadi MN, Francois ME, Postnova S, Cistulli PA, Fontana L, and Stamatakis E. Timing of Moderate to Vigorous Physical Activity, Mortality, Cardiovascular Disease, and Microvascular Disease in Adults With Obesity. Diabetes Care 2024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Dalla Man C, Campioni M, Polonsky KS, Basu R, Rizza RA, Toffolo G, and Cobelli C. Two-hour seven-sample oral glucose tolerance test and meal protocol: minimal model assessment of beta-cell responsivity and insulin sensitivity in nondiabetic individuals. Diabetes 54: 3265–3273, 2005. [DOI] [PubMed] [Google Scholar]
  • 30.Pruessner JC, Kirschbaum C, Meinlschmid G, and Hellhammer DH. Two formulas for computation of the area under the curve represent measures of total hormone concentration versus time-dependent change. Psychoneuroendocrinology 28: 916–931, 2003. [DOI] [PubMed] [Google Scholar]
  • 31.Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, and Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 28: 412–419, 1985. [DOI] [PubMed] [Google Scholar]
  • 32.Matsuda M, and DeFronzo RA. Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp. Diabetes Care 22: 1462–1470, 1999. [DOI] [PubMed] [Google Scholar]
  • 33.ACSM's Guidelines for Exercise Testing and Prescription Philadelphia: Lippincott, Williams & Wilkins, 2005. [Google Scholar]
  • 34.Austin PC, and Stuart EA. Moving towards best practice when using inverse probability of treatment weighting (IPTW) using the propensity score to estimate causal treatment effects in observational studies. Stat Med 34: 3661–3679, 2015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Lanza ST, Moore JE, and Butera NM. Drawing Causal Inferences Using Propensity Scores: A Practical Guide for Community Psychologists. Am J Community Psychol 52: 380–392, 2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Black LE, Swan PD, and Alvar BA. Effects of Intensity and Volume on Insulin Sensitivity During Acute Bouts of Resistance Training. J Strength Cond Res 24: 1109–1116, 2010. [DOI] [PubMed] [Google Scholar]
  • 37.Houmard JA, Tanner CJ, Slentz CA, Duscha BD, McCartney JS, and Kraus WE. Effect of the volume and intensity of exercise training on insulin sensitivity. J Appl Physiol (1985) 96: 101–106, 2004. [DOI] [PubMed] [Google Scholar]
  • 38.Huang L, Fang Y, and Tang L. Comparisons of different exercise interventions on glycemic control and insulin resistance in prediabetes: a network meta-analysis. BMC Endocr Disord 21: 181, 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Galan-Lopez P, and Casuso RA. Metabolic Adaptations to Morning Versus Afternoon Training: A Systematic Review and Meta-analysis. Sports Med 53: 1951–1961, 2023. [DOI] [PubMed] [Google Scholar]
  • 40.Dyar KA, Ciciliot S, Wright LE, Biensø RS, Tagliazucchi GM, Patel VR, et al. Muscle insulin sensitivity and glucose metabolism are controlled by the intrinsic muscle clock. Mol Metab 3: 29–41, 2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Ezagouri S, Zwighaft Z, Sobel J, Baillieul S, Doutreleau S, Ladeuix B, et al. Physiological and Molecular Dissection of Daily Variance in Exercise Capacity. Cell Metab 30: 78–91.e74, 2019. [DOI] [PubMed] [Google Scholar]
  • 42.Krčmárová B, Krčmár M, Schwarzová M, Chlebo P, Chlebová Z, Židek R, et al. The effects of 12-week progressive strength training on strength, functional capacity, metabolic biomarkers, and serum hormone concentrations in healthy older women: morning versus evening training. Chronobiol Int 35: 1490–1502, 2018. [DOI] [PubMed] [Google Scholar]
  • 43.Schwingshackl L, Dias S, Strasser B, and Hoffmann G. Impact of different training modalities on anthropometric and metabolic characteristics in overweight/obese subjects: a systematic review and network meta-analysis. PLoS One 8: e82853, 2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Ma T, Bennett T, Lee CD, and Wicklow M. The diurnal pattern of moderate-to-vigorous physical activity and obesity: a cross-sectional analysis. Obesity 31: 2638–2647, 2023. [DOI] [PubMed] [Google Scholar]
  • 45.Küüsmaa M, Schumann M, Sedliak M, Kraemer WJ, Newton RU, Malinen J-P, et al. Effects of morning versus evening combined strength and endurance training on physical performance, muscle hypertrophy, and serum hormone concentrations. Applied Physiology, Nutrition, and Metabolism 41: 1285–1294, 2016. [DOI] [PubMed] [Google Scholar]
  • 46.Knaier R, Qian J, Roth R, Infanger D, Notter T, Wang W, et al. Diurnal Variation in Maximum Endurance and Maximum Strength Performance: A Systematic Review and Meta-analysis. Med Sci Sports Exerc 54: 169–180, 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Bruggisser F, Knaier R, Roth R, Wang W, Qian J, and Scheer FAJL. Best Time of Day for Strength and Endurance Training to Improve Health and Performance? A Systematic Review with Meta-analysis. Sports Medicine - Open 9: 2023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Grgic J, Lazinica B, Garofolini A, Schoenfeld BJ, Saner NJ, and Mikulic P. The effects of time of day-specific resistance training on adaptations in skeletal muscle hypertrophy and muscle strength: A systematic review and meta-analysis. Chronobiol Int 36: 449–460, 2019. [DOI] [PubMed] [Google Scholar]
  • 49.Chtourou H, and Souissi N. The effect of training at a specific time of day: a review. J Strength Cond Res 26: 1984–2005, 2012. [DOI] [PubMed] [Google Scholar]
  • 50.Remchak M- ME, Heiston EM, Ballantyne A, Dotson BL, Stewart NR, Spaeth AM, and Malin SK. Insulin Sensitivity and Metabolic Flexibility Parallel Plasma TCA Levels in Early Chronotype With Metabolic Syndrome. The Journal of Clinical Endocrinology & Metabolism 107: e3487–e3496, 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Malin SK, Remchak MME, Smith AJ, Ragland TJ, Heiston EM, and Cheema U. Early chronotype with metabolic syndrome favours resting and exercise fat oxidation in relation to insulin-stimulated non-oxidative glucose disposal. Exp Physiol 107: 1255–1264, 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Anothaisintawee T, Lertrattananon D, Thamakaison S, Knutson KL, Thakkinstian A, and Reutrakul S. Later chronotype is associated with higher hemoglobin A1c in prediabetes patients. Chronobiol Int 34: 393–402, 2017. [DOI] [PubMed] [Google Scholar]
  • 53.Henson J, Rowlands AV, Baldry E, Brady EM, Davies MJ, Edwardson CL, et al. Physical behaviors and chronotype in people with type 2 diabetes. BMJ Open Diabetes Res Care 8: 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Thomas JM, Kern PA, Bush HM, McQuerry KJ, Black WS, Clasey JL, and Pendergast JS. Circadian rhythm phase shifts caused by timed exercise vary with chronotype. JCI Insight 5: 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Kim N, Ka S, and Park J. Effects of exercise timing and intensity on physiological circadian rhythm and sleep quality: a systematic review. Physical Activity and Nutrition 27: 052–063, 2023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Creasy SA, Ostendorf DM, Blankenship JM, Grau L, Arbet J, Bessesen DH, et al. Effect of sleep on weight loss and adherence to diet and physical activity recommendations during an 18-month behavioral weight loss intervention. Int J Obes (Lond) 46: 1510–1517, 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Buxton OM, Pavlova M, Reid EW, Wang W, Simonson DC, and Adler GK. Sleep Restriction for 1 Week Reduces Insulin Sensitivity in Healthy Men. Diabetes 59: 2126–2133, 2010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Anothaisintawee T, Reutrakul S, Van Cauter E, and Thakkinstian A. Sleep disturbances compared to traditional risk factors for diabetes development: Systematic review and meta-analysis. Sleep Med Rev 30: 11–24, 2016. [DOI] [PubMed] [Google Scholar]
  • 59.Burgess VN, Antonio J, Bland HW, Wagner R, Tartar JL, and Melton BF. The Effect of Timing and Type of Exercise on the Quality of Sleep in Trained Individuals. Int J Exerc Sci 13: 837–858, 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60.Saidi O, Colin E, Rance M, Doré E, Pereira B, and Duché P. Effect of morning versus evening exercise training on sleep, physical activity, fitness, fatigue and quality of life in overweight and obese adults. Chronobiol Int 38: 1537–1548, 2021. [DOI] [PubMed] [Google Scholar]
  • 61.Alley JR, Mazzochi JW, Smith CJ, Morris DM, and Collier SR. Effects of Resistance Exercise Timing on Sleep Architecture and Nocturnal Blood Pressure. The Journal of Strength & Conditioning Research 29: 1378–1385, 2015. [DOI] [PubMed] [Google Scholar]
  • 62.Blankenship JM, Rosenberg RC, Rynders CA, Melanson EL, Catenacci VA, and Creasy SA. Examining the Role of Exercise Timing in Weight Management: A Review. Int J Sports Med 42: 967–978, 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 63.Halliday TM, Savla J, Marinik EL, Hedrick VE, Winett RA, and Davy BM. Resistance training is associated with spontaneous changes in aerobic physical activity but not overall diet quality in adults with prediabetes. Physiol Behav 177: 49–56, 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 64.Halliday TM, Davy BM, Clark AG, Baugh ME, Hedrick VE, Marinik EL, et al. Dietary intake modification in response to a participation in a resistance training program for sedentary older adults with prediabetes: findings from the Resist Diabetes study. Eat Behav 15: 379–382, 2014. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

Source data for this study are not publicly available due to privacy or ethical restrictions. The source data are available to verified researchers upon request by contacting the corresponding author. At the time of data collection for the Resist Diabetes study, data availability was not a prior publishing or NIH requirement. Data exists and is available, but not in a format that is easy to share or deposit.

RESOURCES