Aim: Tactile massage (TM) is a gentle and superficial form of massage. A pilot study of patients with type 2 diabetes in primary care reported a reduction of 0.8% in glycosylated haemoglobin (HbA1c), whereas a randomized study comparing the effects of 10 weeks of TM once per week with relaxation exercises performed once per week as per instructions on a CD found no effects of TM on HbA1c in an intention-to-treat analysis. However, a significant reduction in waist circumference (WC) was found between the groups. Methods: This was a secondary per-protocol analysis of the effect of TM (n = 21) compared with relaxation (n = 25) on other metabolic biomarkers. Anthropometrics (BMI and WC) and metabolic factors (B HbA1c, S IGF, fS insulin, S adiponectin, S leptin and fP ghrelin) were assessed, insulin resistance (IR) was determined by modified homoeostasis model assessment (HOMA2-IR) using fP glucose and fS insulin, and ratios of adiponectin-to-leptin, adiponectin-to-HOMA-IR, adiponectin-to-WC and adiponectin-to-HbA1c were calculated at baseline, and at 10 weeks and 6 months after the intervention. Results: Significant results adjusted for age, gender and changes in lifestyle and medical factors were shown for WC in women (-6.2 cm [95% CI: -10.4, -1.9]), but not in men. In addition, improvements in the TM group were found for adiponectin and ratios of adiponectin-to-leptin and adiponectin-to-HbA1c levels. Conclusion: Our data indicate that TM therapy may affect metabolic markers in type 2 diabetes despite the lack of significant effects on HbA1c. The clinical implications of our findings need to be evaluated in further studies. © 2013 Elsevier Masson SAS. All rights reserved.