The aim of this randomized trial is to determine whether liver fat depletion via a short-term (i.e., two weeks) very-low calorie diet will restore the normal exercise-induced secretion of a signaling protein (fibroblast growth factor 21) from the liver in people living with type 2 diabetes. Participants will have their liver fat, body composition, and various markers of metabolic health assessed and then will be randomized to either the very-low calorie diet intervention or a free-living control group for two weeks. Upon completion of the two-week intervention period, participants will redo all of the pre-intervention assessments. The changes in the assessments from before vs. after the intervention period will be compared between the two intervention groups (i.e., the very-low calorie diet group vs. the free living control group).

Fibroblast growth factor 21 (FGF21) is a hepatokine that confers multiple beneficial effects when signaling to other tissues in the body. This is particularly true in adipose tissue, where it improves insulin sensitivity and leads to adiponectin production/secretion. In healthy individuals, FGF21 secretion is induced by exercise: however, this effect is absent in people living with type 2 diabetes (T2D). Aberrant secretion of FGF21 and other hepatokines is associated with increases in ectopic fat in the liver, but it is unknown if depletion of ectopic liver fat ameliorate this effect.

42 participants living with T2D will be recruited for this parallel randomized controlled trial. At baseline, participants will have their liver fat assessed via magnetic resonance imaging, exercise-induced FGF21 incremental area under the curve (following ~1 hour of exercise on a cycle ergometer at a workload equivalent to 60% VO2peak) as well as anthropometrics and outcomes relating to cardiometabolic health. Participants will then be randomized to either 1) two weeks of a very-low calorie diet (VLCD; ~800 kcal/day) or 2) two weeks of free living control (CON). Following the two week intervention period, participants will repeat the baseline outcome assessment.

The primary objective of this trial is to determine whether diet-induced liver fat depletion will restore exercise-induced FGF21 secretion in people living with T2D. Secondary objectives include determining the effect of diet-induced liver fat depletion on fasting levels of circulating FGF21 and expression of FGF21 receptor/co-receptors and downstream signals in adipose tissue, as well as other cardiometabolic outcomes.

Inclusion Criteria:

• Men and women 30-70 years of age
• The target population is persons with type 2 diabetes. I.e., persons are eligible if they are diagnosed with type 2 diabetes either only with metformin for managing glucose or without use of glucose lowering medications. Persons with a HbA1c ≥ 48 mmol/mol with or without the use of glucose lowering medications are also eligible. Any glucose lowering medications other than metformin are disallowed (described under "Exclusion criteria, below)
• Diabetes duration < 7 years
• Body Mass Index (BMI) ≥ 30 kg/m2 and ≤ 40 kg/m2
• Accepts medical regulation by the study endocrinologist
• Inactivity, defined as < 1,5 hours of structured physical activity pr. week at moderate intensity and cycling < 30 minutes/5 km pr. day at moderate intensity (moderate intensity = out of breath but able to speak)

Exclusion Criteria:

• HbA1c ≥ 75 mmol/mol with no glucose lowering medications
• HbA1c ≥ 64 mmol/mol with mono glucose lowering therapy (if compliant with the prescription)
• HbA1c ≥ 57 mmol/mol with ≥ dual glucose lowering therapy (if compliant with the prescription)
• Diagnosis of Type 1 diabetes, MODY-diabetes, Type 1½ diabetes or LADA-diabetes
• eGFR<60mL/min (assessed via screening blood sample)
• Treatment with any glucose-lowering medications other than metformin (e.g., insulin (long and/or short acting), sulphonylurea based drugs, glucagon-like peptide 1 receptor agonists, dipeptidyl peptidase 4 inhibitors, sodium-glucose co-transporter-2 inhibitors, thiazolidinediones, alpha-glucosidase inhibitors)
• Presence of metal in the body that would contraindicate an MRI scan
• Known or signs of intermediate or severe microvascular complications to diabetes (retino-, neuro- or nephropathy)
• Known cancer
• Lung disease, other than asthma that can be managed with beta2-agonists and does not exhibit seasonal variation
• Known cardiovascular disease
• Known hyperthyroid disease
• Clinical or biochemical signs of hypothyroid disease
• Changes in hypothyroid disease treatment within the last 3 three months prior to enrolment
• Known liver disease - defined as ALAT or ASAT elevated three times above upper limit
• Known autoimmune disease
• Psoriasis disease requiring systemic treatment or cutan elements bigger than a total area of 25 cm2
• Other endocrine disorder causing obesity
• Current treatment with anti-obesity medication
• Current treatment with anti-inflammatory medication
• Weight loss of > 5kg within the last 6 months
• Changes in symptoms or anti-depressive medication three months prior to enrolment
• Diagnosis of psychiatric disorder or treatment with anti-psychotic medication
• History of suicidal behavior or ideations within the last three months prior enrolment
• Previous surgical treatment for obesity (excluding liposuction > 1 year prior to enrolment)
• Pregnant/considering pregnancy, or lactating
• Functional impairments that prevent the performance of intensive exercise
• Participation in other research intervention studies
• Macroalbuminuria at pre-screening (assessed via screening blood sample)
• Biochemical sign of other major diseases
• Presence of circulating glutamatdecarboxylase anti body (GAD) 65 (assessed via screening blood sample)
• Objective findings that contraindicate participation in intensive exercise
• Incidental findings that contraindicate participation in the study
• Unable to allocate the needed time to fulfill the intervention
• Language barrier, mental incapacity, unwillingness, or inability to understand and be able to complete the interventions

If the data can be fully anonymized, the data can be shared.