Sitting less elicits metabolic responses similar to exercise and enhances insulin sensitivity in postmenopausal women

Abstract

Aims/hypothesis: In our current society sedentary behaviour predominates in most people and is associated with the risk of developing type 2 diabetes. It has been suggested that replacing sitting time by standing and walking could be beneficial for individuals with type 2 diabetes but the underlying mechanisms are unknown and direct comparisons with exercise are lacking. Our objective was to directly compare metabolic responses of either sitting less or exercising, relative to being sedentary.

Methods: We performed a randomised, crossover intervention study in 12 overweight women who performed three well-controlled 4 day activity regimens: (1) sitting regimen (sitting 14 h/day); (2) exercise regimen (sitting 13 h/day, exercise 1 h/day); and (3) sitting less regimen (sitting 9 h/day, standing 4 h/day and walking 3 h/day). The primary outcome was insulin sensitivity measured by a two-step hyperinsulinaemic-euglycaemic clamp. We additionally performed metabolomics on muscle biopsies taken before the clamp to identify changes at the molecular level.

Results: Replacing sitting time by standing and walking over 4 days resulted in improved peripheral insulin sensitivity, comparable with the improvement achieved by moderate-to-vigorous exercise. Specifically, we report a significant improvement in peripheral insulin sensitivity in the sitting less (~13%) and the exercise regimen (~20%), compared with the sitting regimen. Furthermore, sitting less shifted the underlying muscle metabolome towards that seen with moderate-to-vigorous exercise, compared with the sitting regimen.

Conclusions/interpretations: Replacing sitting time by standing and walking is an attractive alternative to moderate-to-vigorous exercise for improving metabolic health.

Trial registration: ClinicalTrials.gov NCT03912922.

Keywords: Clinical trial; Exercise; Insulin sensitivity; Metabolomics; Muscle metabolism; Sedentary time; Sitting; Sitting less.

© 2021. The Author(s).

Figures

Fig. 1

Sitting, sitting less and exercise regimens to assess metabolic health. Regimens for sitting, sitting less and exercise are denoted as SIT, SL and EXE, respectively. (a) Visualisation of the three activity regimens. Each participant followed the activity regimens in a random order. A day in the SIT regimen consisted of 1 h walking, 1 h standing, 14 h sitting and 8 h sleeping. A day in the SL regimen consisted of 3 h walking, 4 h standing, 9 h sitting and 8 h sleeping. A day in the EXE regimen consisted of 1 h exercise, 1 h walking, 1 h standing, 13 h sitting and 8 h sleeping. (b) Each activity regimen lasted 4 days and measurements were performed on day 5. The activity regimens were separated by a washout period of 9–23 days

Fig. 2

The sitting less and exercise regimens improve insulin sensitivity compared with the sitting regimen. Regimens for sitting, sitting less and exercise are denoted as SIT, SL and EXE, respectively. Data are shown by boxplots in which each participant is represented by a particular symbol throughout the fig. A hyperinsulinaemic–euglycaemic two-step clamp was performed to assess insulin sensitivity. (a) Whole-body insulin-stimulated glucose disposal during high-dose insulin infusion expressed as Rd (n = 10). (b) NOGD during high-dose insulin infusion (n = 10). (c) Suppression of hepatic EGP during low-dose insulin infusion (n = 11). (d) IHL measured by 1H-MRS (n = 10). *p < 0.05 and **p < 0.01. Boxplots: box includes the IQR corresponding to 25th percentile (Q1, bottom of box), 50th percentile (Q2, bold line within box, i.e. the median) and 75th percentile (Q3, top of box) of the data. The bottom whisker is set at Q1 – 1.5 × IQR and the top whisker at Q3 + 1.5 × IQR

Fig. 3

Responses to sitting less are similar to those produced by exercise at the molecular metabolic level in skeletal muscle. Regimens for sitting, sitting less and exercise are denoted as SIT, SL and EXE, respectively. Metabolomics analyses were performed in skeletal muscle biopsy samples of participants (n = 11). Heatmap and box plots show stepwise metabolite abundance changes going from SIT to SL to EXE. (a) Heatmap of the top 25 metabolites, ranked by their VIP scores from the PLS-DA (ESM Fig. 1a). Relative abundance levels in each participant per regimen is scaled from low abundance (blue) to high abundance (red). (b) Comparison of the significance of differences that EXE induces (compared with sitting), relative to differences induced by SL (compared with sitting). Units on the axes are p values on a −log10 scale. Directionality of induced changes are represented as either negative values (decreased) or positive values (increased). Pearson’s r = 0.393, p = 2 × 10−6. (c–f) The metabolomic shift induced by SL and EXE is illustrated for malate (c), uridylic acid (d), deoxyadenosine monophosphate (e) and tryptophan (f). Data are shown by boxplots in which each participant is represented by a particular symbol throughout the figure. Significance was determined using an empirical Bayes moderated t test in a linear model framework. *p < 0.05 and **p < 0.01. AU, arbitrary units; CDP-choline, cytidine 5′-diphosphocholine; dAMP, deoxyadenosine monophosphate; Hexose-P, hexose phosphate; UMP, uridylic acid. Boxplots: box includes the IQR corresponding to 25th percentile (Q1, bottom of box), 50th percentile (Q2, bold line within box, i.e. the median) and 75th percentile (Q3, top of box) of the data. The bottom whisker is set at Q1 – 1.5 × IQR and the top whisker at Q3 + 1.5 × IQR