Search Weight Loss Topics:




May 3

Effects of different doses of exercise and diet-induced weight loss on … – Nature.com

Diet Comments Off on Effects of different doses of exercise and diet-induced weight loss on … – Nature.com

Study design

The study was a 16-week, parallel-group, four-arm, assessor-blinded, randomized clinical trial conducted between February 2019 and October 2021 at the Centre for Physical Activity Research (CFAS), Rigshospitalet, Copenhagen, Denmark. The study was preregistered at ClinicalTrials.gov (NCT03769883) and was approved by the Scientific Ethical Committee of the Capital Region of Denmark (approval number H-18038298) before the commencement of any study procedures. Guidelines from the Helsinki Declaration were followed, and the data are reported following the CONSORT guideline for multi-arm trials43 and the REPORT standards43. The study protocol for this clinical trial is available in the Supplementary Information and has been published previously22. The prespecified full statistical analysis plan (SAP) was completed and uploaded to our website before commencing any statistical analyses (https://aktivsundhed.dk/images/docs/SAP_doseex_nov21.pdf).

Participants were recruited through the media, municipalities and the Danish Health Data Authorities. The potential participants contacted the study nurse and completed the screening process before the medical examination. The main inclusion criteria were (1) men and women aged 1880years, (2) diagnosed with T2D within <7years, (3) no current treatment with insulin and (4) BMI>27kg/m2 and <40kg/m2. All participants provided written and oral informed consent before any testing.

CON received standard care and was encouraged to maintain habitual physical activity and dietary habits throughout the study. DCON received standard care and dietary intervention. MED received standard care, dietary intervention and an exercise intervention with two aerobic training sessions per week and one combined aerobic and resistance training session per week, totaling 150165min of exercise training per week. HED received the standard care and dietary interventions as described above but had twice as much exercise as MED, with a total of four aerobic training sessions per week and two combined aerobic and resistance training sessions per week, totaling 300330min of exercise training per week.

Standard care included pharmacological management of blood glucose, blood lipids and blood pressure according to a prespecified algorithm and was managed by an endocrinologist who was blinded for participant allocation22. To minimize an influence on the findings of poor glucose control upon study entry, medical standardization was introduced according to the prespecified treat-to-target algorithm for 6weeks before the baseline measurements. Furthermore, the pharmacological treatment was evaluated according to the algorithm following baseline measurements and at week 12 of the intervention. The treatment targets were in line with current guidelines. In adjunct to the algorithm, pharmacological treatment was adapted to mitigate subjective signs of hypotension or hypoglycemia. Blood lipids, blood pressure and blood glucose were measured before the intervention and 4, 12 and 16weeks into the intervention. In case of any adverse events, the participants were advised to contact the study nurse. At each visit, the study nurse interviewed all participants about potential adverse events. The adverse events definition followed ICH E2A guidelines44.

Daily energy requirements were estimated using the age-adjusted Oxford equation45. The dietary intervention aimed at ~2530% energy deficit per day with a macronutrient distribution within the range of 4560 energy percent (E%) carbohydrate, 1520E% protein and 2035E% fat (<7E% saturated fat). The intervention consisted of individualized recommendations and recipes. A clinical dietician implemented the plan at three sessions during the intervention, and adjustments were performed based on self-reported, 3-day food records.

The exercise intervention consisted of both aerobic and resistance training, and the first 2weeks served as a familiarization period. The aerobic training sessions of 30-min duration had a target intensity of 60100% of maximal heart rate (HRmax). Throughout the intervention, the relative time spent exercising in intensity zone 80100% of HRmax was increased, and the relative time spent in the intensity zone 6079% of HRmax was reduced accordingly. Resistance training was added in combined sessions with 30min of aerobic training and 3045min of resistance training. The resistance training consisted of three sets in the main muscle groups, for example, chest press, leg press, back row, and leg extension. The 812 repetitions aimed at a resistance consistent with 03 repetitions in reserve46. All heart-rate profiles were recorded during the exercise interventions (Polar V800), and all training sessions were supervised by educated trainers.

Two experimental days were conducted at baseline and repeated at 16-week follow-up. Forty-eight hours before the experimental days, the participants were instructed to discontinue glucose-lowering medication use and refrain from any exercise. Moreover, no alcohol or caffeine was permitted 24h before the visits, and the participants were instructed to maintain their habitual diet. The participants arrived at the testing facilities at 07:30am after an overnight fast (10h fasting). Experimental days 1 and 2 were planned to be separated by 1week.

The participants completed a 3-h MMTT. The liquid meal was prepared using 400ml of Nestl Resource with an additional 36g of dextrose (total energy content, 735kcal; E%, 64/24/12 carbohydrate/fat/protein). Paracetamol (1.5g) was added to assess gastric emptying. Body weight was measured with an electronic scale, and height was measured with a Holtain stadiometer according to standard procedures. VO2max was assessed using indirect calorimetry (Quark CPET, Cosmed) on a Monark LC4 bicycle (Monark Exercise). The test was performed with a 5-min warm-up followed by increases of 20watts/min until exhaustion. Maximum muscle strength was assessed by two exercises performed in resistance training machines (chest press, leg extension) via estimating the maximum weight (kg) that could be lifted once with a full range of motion with proper form (that is, 1RM).

A three-stage hyperglycemic clamp was performed. After baseline blood sampling, a priming bolus of [6,6-2H2]glucose was injected intravenously and a continuous tracer infusion was initiated. The bolus dose and infusion rate of the tracer depended on the participants fasting glucose level and body weight as described elsewhere5. After 2h of tracer infusion, hyperglycemia was introduced by clamping glucose at 5.4mM above fasting glucose (whereas the absolute postintervention clamp glucose level was equal to the preintervention clamp level). An initial increase in blood glucose was brought about by a square-wave glucose infusion lasting 15min. After this, the glucose concentration was kept constant by adjusting GIRs based on blood glucose measurements (ABL 8 series, Radiometer) performed every 5min according to an automated algorithm5. After 2h of hyperglycemia, a continuous GLP-1 infusion was initiated at a rate of 0.5pmol/kg/min, and after 1h of hyperglycemia+GLP-1 infusion, an intravenous bolus of arginine hydrochloride (5g given over 30s) was administered to provide a maximal stimulus to the beta cells, leading to secretion of remaining intracellular vesicles of insulin. Before baseline sampling, the participant voided. Every time the participant voided during the clamp, the urine was accumulated, and urinary glucose concentration was measured at the end of the procedure.

Assessments of free-living physical activity and blood pressure were recorded by the participants between the 2 study days. Physical activity was also assessed with physical activity monitors (AX3, Axivity) for 7 consecutive days. Blood pressure was assessed with home-based resting measurements across 3days, including three measurements morning and evening. Furthermore, a 3-day record of total dietary intake was completed at baseline, during the intervention period (at weeks 4 and 12), and during the 3days leading up to follow-up testing.

Blood samples (plasma insulin, C-peptide, glucose, HbA1c, LDL-C, triglycerides and paracetamol) were analyzed at the Department of Clinical Biochemistry, Rigshospitalet, using standard procedures. GLP-1 and GIP were analyzed using in-house carboxy-terminal radioimmunoassays. The total GLP-1 assay (codename 89390) is based on the amidated COOH terminus and therefore measures GLP-1(736)NH2 and GLP-1(936)NH2. The assay results, therefore, reflect the secretion rate of GLP-1 (refs. 47,48). The total GIP assay (codename 80867) reacts fully with intact GIP and amino-terminally truncated forms49. The glucose tracer [6,6-2H2]glucose was used for whole-body measurements of Ra and Rd of glucose during steady-state hyperglycemia and was calculated using non-steady-state equations50 adapted for stable isotopes51,52.

All participants received up to DKK6,000 (800) in total to cover lost earnings, transport and discomfort. The transaction was completed upon completion of the study (all four full laboratory days (V1, V2, V6 and V7) or upon withdrawal). For every completed day of laboratory testing, participants received DKK1,000. Moreover, DKK500 in compensation was added per biopsy (up to four in total). To prevent loss to follow-up in the CON group, we offered three supervised training sessions and a free 16-week membership in a fitness center following final testing.

The primary outcome was the change in late-phase DI from baseline to the 16-week follow-up, reflecting the beta-cell response during the last 30min of the hyperglycemic stage15. DI was calculated as the product of late-phase ISR and late-phase ISI (designated secondary outcomes, see below).

Secondary outcomes were prespecified in the SAP (designated Major secondary outcomes in the SAP) and included the late-phase ISR, late-phase ISI derived during the last 30min of the hyperglycemic stage, and the oral DI, oral ISI and oral ISR derived from the MMTT53. Late-phase ISR was calculated from the deconvoluted C-peptide measurements54 and subsequently normalized to ambient blood glucose concentrations. Late-phase ISI was calculated as the GIR divided by the product of insulin and glucose39. Oral DI was calculated as the product of oral ISI and oral ISR. Oral ISI (the Matsuda index) was calculated as 10,000/(fasting glucosefasting insulin)(mean glucose0120minmean insulin0120min), and oral ISR was calculated as the tAUC for glucose divided by the tAUC for insulin from time 0 to 120min during the MMTT53.

The exploratory outcomes (designated Other secondary outcomes in the SAP) included the change (baseline to 16-week follow-up) in first-phase ISR, EGP, first-phase DI, ISI and ISR, as well as HbA1c, LDL-C, fasting glucose, fasting insulin, fasting C-peptide, fasting triglycerides, systolic blood pressure, diastolic blood pressure, body weight, absolute VO2max, relative VO2max, 1RM for chest press and leg extension (both absolute and relative to body weight), and tAUC and iAUC in glucose, insulin, C-peptide, GLP-1, GIP and paracetamol from the MMTT. AUCs for the different time periods were calculated using the trapezoidal rule. Ra and Rd were calculated from glucose tracers during clamp-induced steady-state hyperglycemia. Adverse events were self-reported.

Post hoc outcomes included intensification (yes or no), reduction (yes or no) and discontinuation (yes or no) for glucose-lowering and blood pressure-lowering medications. Due to restrictions in our pharmacological treatment algorithm regarding lipid-lowering medications, only intensifications were assessed for this outcome.

The participants were randomly allocated to the four intervention arms upon successful completion of the baseline measurements. An independent statistician (author R.C.) prepared a computer-generated randomization schedule in a ratio of 1:1:1:1, stratified by sex. To ensure concealment, the (permuted) block sizes were not disclosed. The schedule was forwarded to a secretary who was not involved in any study procedures and stored on a password-protected computer. Sequentially numbered, opaque, sealed envelopes were prepared and stored in a locked cabinet before commencing the recruitment. The envelopes were lined with aluminum foil to render the envelope impermeable to intense light. Following the conclusion of the hyperglycemic clamp, the appropriate envelope was opened by a study nurse, and the participant was informed about the allocation stated on the card inside the envelope. The participant received the allocation in a closed room. As such, the participants were blinded for treatment allocation until after the completion of the hyperglycemic clamp. Following the baseline assessment, blinding of the participants was no longer possible. Both study personnel involved with the data collection and the study endocrinologist managing pharmacological treatment and safety were blinded to allocation. The clinical results used for pharmacological management and safety assessment were presented to the endocrinologist by the study nurse without disclosing participant allocation.

We expected that an exercise intervention would increase the late-phase DI by 1.5 arbitrary units (a.u.) more than the control group, with a standard deviation of 1.5a.u. of the change in the exercise and 1.0a.u. in the control group5. For a contrast in a one-way analysis of variance (ANOVA) with four means (1.5, 1.0, 0.5, 0.0) and contrast coefficients (1, 0, 0, 1) using a two-sided significance level of 0.05, assuming an error standard deviation of 1.5 and a balanced design, a total sample size of 80 participants in the PP population (approximately 20 participants in each group) would yield statistical power of 87.7%.

According to the protocol and the SAP, the analysis of the primary outcome was based on the as-observed population (missing data were not imputed in the primary analysis)55,56, as well as the PP population. The Full Analysis Set for the ITT population included all randomized participants irrespective of their compliance with the interventions. The PP population criteria included (1) completion of the primary outcome assessment (all groups), (2) compliance with the diet protocol defined as being within 30% of the prescribed energy intake (DCON, MED and HED), and (3) compliance with the exercise training protocol defined as completing 70% of the prescribed exercise volume across the intervention period (from weeks 2 to 16) (MED and HED). Missing data were assumed to be missing at random. Continuous data, including the primary, secondary and exploratory outcomes, were analyzed using constrained baseline longitudinal analysis via a linear mixed model57. As the baseline value is a part of the outcome vector, all participants with at least one measurement (baseline or follow-up) were included in the analyses57. The model included fixed effects for time (two levels), treatment (coded 0 for all groups at baseline and coded 0, 1, 2 or 3 at follow-up for CON, DCON, MED and HED, respectively) and sex (two levels), as well as the unique patient identifier as a random effect. The potentially biased PP population analysis was further adjusted for putative confounders: diabetes duration and baseline maximal oxygen consumption (ml O2/kg/min). Data are presented as the difference in the mean changes with 95% confidence intervals unless stated otherwise. The adequacy of the models was investigated via the predicted values and residuals. If the model assumptions were violated, the analyses were conducted using the log-transformed data and subsequently exponentiated for interpretation. Back-transformed data were expressed as the ratio of the geometric mean and interpreted as either percent change from baseline (within group) or difference in change between groups. A linear trend (interpreted as a linear doseresponse relationship) was examined by treating each treatment category as a continuous variable in the main model and tested using a Wald test (Pvalue reported). Linearity was inspected visually, and the P for trend was calculated only for the primary and secondary outcomes to the extent that the relationship was linear (that is, for late-phase DI and late-phase ISI). Sensitivity analyses were performed using multiple linear imputation procedures with the change in outcomes (post-pre values)55. The model included all covariates included in the main model, and beta coefficient and standard errors were based on 30 imputed data sets and adjusted for between-imputation variability58. Dichotomous outcomes were analyzed using logistic regression analyses. As sparsity of dichotomous outcomes (as expected for medications) invalidates the confidence intervals, exact logistic regression (exlogistic in Stata) was used when cases were <559,60. A post hoc statistical mediation analysis was performed to examine the extent to which the observed treatment effect (in the intervention groups) on the primary and secondary outcomes was mediated by the change in body weight. An exploratory statistical mediation analysis was performed in R61 to examine the extent to which the observed treatment effect (in the intervention groups) on the primary and key secondary outcomes was mediated by the change in body weight. The lme4 package was used to construct the linear mixed models for the analysis62. This simple mediation analysis partitions the total causal effect into average direct effects (ADE) and average causal mediation effects (ACME; otherwise known as indirect effects). Bias-corrected and accelerated 95% confidence intervals were generated via nonparametric bootstrap analysis (2,000 resamples with replacement).

All non-hypothesis-based comparisons (that is, on the secondary and exploratory outcomes) are per definition considered exploratory and supportive of the interpretation of the primary outcome. If the global test of significance indicated between-group differences (P<0.1)63, all outcomes (primary, secondary, exploratory and post hoc) on pairwise comparisons were explored. Although no corrections for multiplicity were performed, family-wise type 1 error rate on the primary outcome was retained by using a hierarchical analytic approach63. In accordance with our prespecified SAP, the six prespecified hierarchical hypotheses (based on a superiority assumption) were tested using the prespecified sequence: (1) CON versus HED, (2) CON versus MED, (3) CON versus DCON, (4) DCON versus HED, (5) DCON versus MED, (6) MED versus HED. If we failed to progress from any of the prior between-group comparisons (P>0.05), the subsequent Pvalues and confidence intervals were regarded as indicators of associations rather than causality. The statistical significance level (for superiority) was set at <0.05 (two-sided). The statistical analyses were performed using Stata/SE (StataCorp), version 17.1.

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Link:
Effects of different doses of exercise and diet-induced weight loss on ... - Nature.com

Related Posts

    Your Full Name

    Your Email

    Your Phone Number

    Select your age (30+ only)

    Select Your US State

    Program Choice

    Confirm over 30 years old

    Yes

    Confirm that you resident in USA

    Yes

    This is a Serious Inquiry

    Yes

    Message:



    matomo tracker