To compare metformin (2 g daily), or maximum tolerated dose, for 16 weeks with placebo as a treatment of hand osteoarthritis symptoms.

Inclusion Criteria:

1. Age ≥18 years
2. Hand OA according to the ACR criteria
3. Average finger (2 to 5) pain ≥4 on a 0-10 numeric rating scale (NRS where 10 is worst pain) over the past 30 days
4. Metformin naive

Exclusion Criteria:

Comorbidities

1. History of, or current signs of medical disease that may affect joints, e.g. rheumatoid arthritis, gout, psoriatic arthritis

2. Psoriasis

3. Known malignancy (except successfully treated squamous or basal cell skin carcinoma)

4. Drug or alcohol abuse in the last year

5. Existing nerve entrapment syndromes (e.g. carpal tunnel syndrome)

6. Known diabetes

7. Generalised pain syndromes such as fibromyalgia

8. Known peripheral neuropathies

9. Known allergies towards the interventions

10. Gastric bypass or other malabsorption syndrome

11. In case of pharmacological weight loss medication (e.g. glucagon like peptide-1 (GLP-1) analogues) or pharmacological osteoporosis medication, dosage must have been stable for 3 months without any plan of up-titration during the study period

12. Any other condition or impairment that, in the opinion of the investigator, makes a potential participant unsuitable for participation or obstructs participation e.g. psychiatric disorders.

    Surgical history

13. History of hand surgery in the target hand within 12 months prior to enrolment

14. History of arthroplasty, arthrodesis or surgical treatment of thumb base osteoarthritis in the target hand

    Management strategies

15. Use of systemic corticosteroids equivalent of ≥ 7.5 mg prednisolone daily within 3 months

16. Treatment with denosumab (Prolia/Xgeva)

17. Participation in experimental device or experimental drug study 3 months prior to enrolment

18. Intra-articular treatments of any kind of any joint of the target hand 3 months before inclusion

19. Current use of synthetic or non-synthetic opioids

20. Planning to start other treatment for hand OA in the study participation period

21. Planned CT scan with iodine contrast

22. Scheduled surgery on upper extremity of the target hand during study participation

23. Scheduled surgery requiring pause of metformin, e.g. surgery in general anaesthesia, during study participation

    Reproductive system

24. Pregnancy

25. Planned pregnancy within the study period, 3 months after end of study treatment for female fertile participant and 6 months after end of study treatment for male participant

26. Insufficient anti-conception therapy for female fertile participants within the study period and 3 months after end of study treatment

    • Sufficient anti-conception therapy consists of intra-uterine device (coil), hormonal anti-conception (birth control pills, implant, intra-uterine system, dermal patch, vaginal ring, or injections) or sexual abstinence
    • Female participants are considered infertile if they are postmenopausal or if they have undergone surgical sterilisation (bilateral salpingectomy, hysterectomy or bilateral oophorectomy)
    • Postmenopausal state is defined as no menses for 12 months without alternative medical cause before inclusion in the study

27. Insufficient anti-conception therapy for male participants within the study period and 6 months after end of study treatment

    • Sufficient anti-conception therapy consists of condom or sexual abstinence
    • Male participants are considered sterile if they have undergone surgical sterilisation (vasectomy)

28. Breast-feeding

    Blood analysis

29. Positive anti-cyclic citrullinated peptide (>10 kU/L)

30. eGFR <60 ml/min/1.73 m2

31. Vitamin B12 deficiency < 200 pmol/L

32. Hba1c ≥ 48 mmol/mol

Data sharing plan Individual participant data that underlie the results reported in this study and analytic code will be availa-ble from Henning Bliddal (henning.bliddal@regionh.dk) once all planned analyses have been completed and published. The request will be considered on individual basis. Consent for data sharing was not obtained, but the dataset is anonymised, and risk of reidentification is very low. The study protocol and statistical analysis plan are part of the manuscript. The informed consent form is available upon re-quest.