Background: Exercise is first line treatment for patients with type 2 diabetes (T2D), but most
patients with T2D do not achieve a satisfying glycemic control with exercise alone. The reason for
this is unknown, but one explanation is the intensities with which T2D patients perform exercise,
as exercise intensities may be too low to induce metabolic improvements. Thus, pharmacological
treatment with metformin is most often initiated. It is known that metformin and exercise both
activate AMPK, resulting in many different metabolic effects. Because of this similarity, an
interaction between metformin and exercise has been suggested, but knowledge in the area is
sparse. Some studies have reported that metformin blunts the training-induced improvements in
glycemic control, while others have shown an additive effect on glycemic control when metformin
and exercise are combined.
Aim: With this PhD thesis we aimed to investigate the individual and combined effects of
metformin and exercise in order to see whether an interaction occurs, and to investigate the
underlying mechanisms.
Methods: In order to fulfill our aims, two studies were conducted: study 1, a randomized, doubleblinded,
cross-over study in healthy young males, in which we investigated 1) the effect of
metformin on self-selected exercise intensity, and 2) the effect of an acute bout of exercise (+/-
metformin treatment) on various mechanistic variables in skeletal muscle, previously shown to
contribute to the cellular exercise training adaptations (maximal respiration, complex-I linked
activity, ROS emission and AMPK activation); study 2, a double-blinded, randomised, parallelgrouped,
controlled trial with two arms (metformin/placebo) in patients with impaired glucose
tolerance, in which we examined the effect of a 12 week supervised training intervention on
postprandial glucose, as well as the effect of an acute bout of exercise and a 12 week training
intervention on the same mechanistic parameters as in study 1.
Results: We found that metformin did not affect self-selected exercise intensity in healthy males
(paper I). In addition, we found that metformin in combination with exercise training was not
superior to exercise training alone in improving postprandial glucose (paper II). Finally, we found
no evidence to the fact that a mechanistic explanation for an interaction should be found in skeletal
muscle (paper III).
Conclusion: There is an interaction between metformin and exercise on training adaptations in
relation to glucose metabolism. Our results suggest that the explanation behind this interaction
should not be found in skeletal muscle.
Place of employment
Principal supervisor:
Bente Klarlund Pedersen, MD, DMSc, Professor, Centre for Physical
Activity Research, Rigshospitalet, University of Copenhagen,
Denmark
Primary co-supervisor:
Henriette Pilegaard, MSc, PhD, Professor, Department of Biology,
University of Copenhagen, Copenhagen, Denmark
Co-supervisors:
Kristian Karstoft, MD, DMSc, Department of Clinical Pharmacology,
Bispebjerg Hospital, University of Copenhagen, Denmark.
Katrine Bagge Hansen, MD, PhD, Steno iabetes Center Copenhagen,
Gentofte, Denmark