To the best of our knowledge this is the first randomized trial evaluating the effect of pulsatile dry cupping in OA of the knee. We observed statistically significant differences in OA patients who received pulsatile dry cupping for 4 weeks compared to patients with no intervention in most of the outcome measures after 4 weeks. Our results are also of clinical importance: There are various minimal clinically relevant absolute changes in OA patients reported in literature: if we consider 9.1 (96% CI: 10.5; 7.5) on WOMAC as a clinically relevant change
, then our results showed a clinical relevant differences between both treatment groups after 4 weeks. After 12 weeks (including a no treatment period of 8 weeks in both groups) we still observed significant differences, but less prominent, the WOMAC subscores for pain and stiffness were not significant anymore.
In addition, the results of our trial are of high clinical interest regarding the low risk of side effects of the treatment compared to common OA treatments, for example, intra-articular injections of glucocorticosteroids
Due to the exploratory design of our trial, certain limitations have to be considered: We planned an exploratory trial without defining a primary outcome parameter or a formal sample-size calculation – both would be essential in a confirmatory design. Nevertheless, the results of this trial are robust and most of the outcome parameter were statistically significant after 4 and 12 weeks. Another limitation is that our patients were mainly recruited through newspaper advertisements and might not be representative for all patients with OA of the knee. Also, due to the nature of the intervention, a blinding of patients and study physicians was not possible. Because patients in the control group received cupping after 12 weeks of waiting, the long-term effects of cupping in OA could not be properly assessed. However, our results suggest that 8 sessions of cupping in 4 weeks still has a relevant effect after 12 weeks. Only mild side effects were observed, but from a total of only 21 patients in the cupping group, of course only preliminary conclusions can be drawn.
Only a few trials have already evaluated the effectiveness of cupping in various diseases so far. A trial at the University of Essen studied the effect of dry traditional as well as pulsatile dry cupping in chronic neck pain. Statistically significant and clinically relevant effects of cupping compared to waiting-list control were found after 5 sessions of dry cupping with both methods
[10, 11], but also an effect of traditional cupping on pain and quality of life already after one cupping session has been observed in this trial
Future studies should aim to assess the optimal frequency and application of cupping. As we combined cupping on the lower back and knee region, we are not sure if the cupping of the knee or the lumbosacral area alone or even sham cupping would be as effective.
To minimize various types of bias, blinded trials are needed. The development of a special cupping sham device would be helpful for distinguishing specific from unspecific or placebo effects and local from non-local effects in future studies. The design of our exploratory study with no intervention control group does not allow any distinction between specific and unspecific effects of cupping. Pulsatile cupping itself is already a complex intervention consisting of cupping, massaging and injuring the local (knee) and distant (lower back) skin areas, but also other unspecific factors contributing to placebo effects such as possible relaxing by hearing the rhythmic pulsations and noises of the cupping device, patient-practitioner interactions, empathy, therapeutic expectations and even suggestive effects of the information sheets may play an important role. Comparing cupping with other pharmacological or nonpharmacological therapies would be helpful to validate our results and to assess clinical importance and safety.
In a review about the neurobiological basis of naturopathic reflex therapies including massage, acupuncture cupping and other therapies, Musial et al.
 summarized three potential mechanisms of actions as hypothesis: Reflex therapies may firstly influence chronic pain locally by deforming or even injuring the skin which stimulates Aβ fibres in painful but also distal skin regions. Secondly, the level of the spinal cord may also be involved: Manipulations may stimulate inhibitory receptive fields of the multi-receptive dorsal horn neurons. Thirdly, therapeutic effects of the special naturopathic setting which may have a relaxing and social comforting effect on the patient and may imitate a “grooming situation” were discussed. The physiological effects of low-amplitude oscillation sucking of skin and underlying tissues have not yet been systematically investigated, but reproducibility has been demonstrated
From a pragmatic point of view, our results indicate that pulsatile cupping can be of valuable clinical use and help for patients with OA of the knee and may therefore be useful in ambulatory health-care services in addition to other analgesic treatment e.g. pain medication on demand.