Complementary and alternative medicine (CAM) has become increasingly popular over the last decades. Several surveys suggest an increasing demand and use of CAM in almost all western nations1. While in these surveys CAM was reported to be mainly used for common diseases like allergic reactions and common cold, we have to state a much broader variety of diseases covered by an expanding spectrum of CAM therapies.

One of the whole medical systems of CAM, which is quite popular in Germany, is the Anthroposophic Medicine (AM), introduced by Rudolf Steiner in the beginning of the 20^th century 2. Based on a comprehensive view of the human body, health in AM conceptually depends on a harmonious relationship and interaction between the physical body, vital force, soul, and ego 3. Therefore, therapies of AM may address each of these four entities to support the reestablishment of their harmonious interrelations (e.g., to enhance the salutogenetic capacities of the patients, or to strengthen their autonomy 4).

Over the time, AM has developed a variety of pharmacological and non-pharmacological therapies. With respect to pharmacological therapies of AM, interventions like mistletoe therapy for cancer treatment 567 or Cardiodoron, a composition of extracts of blossoms from Primula officinalis and Onopordon acanthium and from the herbs of Hyoscyamus niger for cardio-respiratory coordination 89, have been subject of several investigations from both basic research to systematic review or meta-analysis.

Among elderly patients with chronic diseases, 2% use AM applied by medical doctors, and at least 4% of patients with cancer 10 – although more 50–70% of cancer patients may use plant extracts from Viscum album which are part of the AM concept to treat cancer patients. Thus, these 2–4% stated above probably represent a lower bound of the factual usage by the population.

In addition, AM offers a variety of non-pharmacological therapies, but only few such therapies have been investigated extensively. It has been shown that Anthroposophic Speech Therapy has an impact on cardiovascular oscillations (e.g., cardiorespiratory synchronization during hexameter recitation 11), and may have beneficial effects on the cardiorespiratory interaction 12. Whether these effects may have a 'cardio-preventive effect' or not, remains to be clarified. On the other hand it has been shown that slow breathing improves baroreflex sensitivity and decreases blood pressure in essential hypertension 13. Because speech therapy employs similar breathing patterns with low breathing frequencies, such effects are supposable. Other non-pharmacological therapies such as Rhythmic Embrocation showed an improvement of pain perception and mood in patients with chronic pain 14, while Rhythmic Massage resulted in a long-term reduction of disease symptoms and quality of life in patients with chronic diseases 15 A recent systematic review investigated the clinical effects of Anthroposophic Eurythmy Therapy 16 and stated positive treatment effects with clinically relevant effect sizes in most cases.

In the context of balneology, AM offers a unique contribution, i.e., the Oil Dispersion Bath (ODB). In 1920, Rudolf Steiner, the founder of Anthroposophy suggested the use of finely-dispersed oil in water in 17. Analogous to the process of oil synthesis in plants and fruits (e.g. olives, 'oil forming process'), he suggested to enable the patients by activating his salutogenetic capacities of the organism with water as a mediator for the essential oils. According to Steiner, if a patient's 'warmth organism' is affected or weakened, one may connect it to the 'warmth processes' of nature, i.e., the 'oil forming process' of flowering plants, and thus introduce finely dispersed oil into a bath to strengthen the weakened 'warmth organism' of the patient 17. According to AM, the oils are related to the ethereal 'flowering process' of plants (in terms of a formation or accumulation of 'warmth') and address particularly the 'self organization' which in turn has an influence on its 'warmth organism'. Based on these recommendations, in the 1930s, the massage therapist and medical bath attendant Werner Junge began to develop a system for very fine oil dispersal in water and developed a vortex mechanism which churns water together with essential oils into a fine mist 18. The oil-covered water droplets empty into a 35 to 37°C (95 to 99°F) warm water tub, where the patient immerses for 15–20 minutes (Fig. 1).

Main indications are the regulation of the 'warmth organism', particularly in patients with arthrosis, gout, and arteriosclerosis 19, but also neurological conditions such as polyneuropathia 20, to improve hypotonic blood pressure 21, and improvement of the 'structuring metabolization' in terms of a tendency to deformation 2119. Patients with cardiovascular diseases or neurological diseases may use slightly colder bathes, i.e., 32 to 34°C (90 to 93°F).

The assortment of the oils, according to the criteria of AM are as follows: Oils from roots are used to strengthen the nervous system; oils from flowers are applied to enhance the capacities of the metabolic system; oils from seeds are supposed to impact the heart and oils from fruits are recommended for improving vascular circulation.

After the bath, the patients should be covered in dry and warm clothes and blankets for at least half an hour in a quiet atmosphere to let the applied warmth saturate the body 19.

Although experts often recommend ODB and despite of its frequent usage in the treatment of in-patients in anthroposophic hospitals (in most cases ODB are applied daily and later three times per week) together with other therapies of AM, still less is known about the physiological and clinical effects. We pragmatically intended to overview the current literature on this unique therapy in a systematic review.

We intended to search for English, French or German language case reports, studies and controlled trials which address specific treatment effects of ODB in a clinical setting without any restrictions to indications or outcome measures. Comments, opinions, and theoretical considerations were excluded. To get a first overview, the following databases were used: Medline, PubMed, Embase, AMED and CAMbase. The search terms were "oil dispersion bath", "oil bath" and their respective translations in German and French. Moreover, an internet search was performed using Google Scholar adding the search terms "study" and "case report" to the above search terms. Finally, we screened the literary estate of Werner Junge and asked several experts for gray literature, i.e., material not found in conventional channels such as publishers with a lack of strict bibliographic control, and thus not listed in the above mentioned databases.

All articles found this way were fully read and their reference lists were checked for further relevant publications. The complete search was performed between June and October 2007. Report identification was performed by GB, TO and AB for the clinical studies, and as an additional reviewer for the experimental studies by DC. Data extraction was primarily performed by GB, and checked by TO and AB. We analyzed the full length articles rather than just the abstracts. Disagreements were resolved by consensus.

As the studies were quite heterogeneous including a variety of perspectives, but not any randomized controlled trial, we decided to classify them with respect to indication, treatment setting, research design, the number of patients involved and the outcome measures/results given in the studies/reports. The reporting of the results adhered, if possible and appropriate at all to the MOOSE guidelines 22. Relying on a checklist for the assessment of methodological quality of the identified studies was dispensable, because the identified studies were almost weak in methodological quality at all. Nevertheless, we intended to refer to the clear description of aim and objective, outcome measures, characteristics of included patients, follow up/drop outs, interventions, principal confounders; participants representative; blinding, randomization, multicenter, same recruitment period. Because most of these issues were not fulfilled in the identified studies, we decided to describe the main relevant flaws in the results.

After excluding theoretical work, descriptive literature and brochures and leaflets of pharmaceutical industries with limited scientific background, we identified 12 relevant articles. Among them, 1 prospective clinical study, 3 experimental studies (enrolling healthy individuals), 5 case reports, and 3 field-reports. The methodological quality of most studies was weak (i.e., presentation of inconclusive results, limited description of patients' characteristics, unclear process of patient selection, unclear inclusion-/exclusion criteria, unclear therapeutic aims, small number of enrolled patients, no control group, etc.). Main indications were internal/metabolic diseases and psychiatric/neurological disorders. Table 1 gives a detailed overview on indications, setting, study type, and outcome parameters.

This article aims at overviewing the existing literature on the effects of ODB. We found a small number of clinical and field studies which mainly focused on the exploration of physiological effects of ODB. The studies can roughly be subdivided into two groups: (1) mechanisms of transdermal resorption of the oil, and (2) regulatory effects of ODB. While there are good clinical studies which confirm that conventional balneotherapy is an effective treatment modality in elderly patients with osteoarthritis of the knee or with chronic low back pain 36, or can alleviate low back pain 37, there is an obvious lack of clinical studies on ODB which address the unique contribution of the oil dispersal in water.

Several effects could be depending on the medium used. Balogh et al. 37 found that bathing in mineral water resulted in a statistically significant improvement of pain, mitigation of muscle spasm, alleviation of local tenderness, enhanced flexion-extension and rotation of the spine, which persisted as long as 3 months after the completion of balneotherapy. In contrast, hydrotherapy with tap water resulted only in the temporary improvement of just a single parameter: the visual analogue scale score improved significantly.

With respect to the usage of distinct oils dispersed in water, one may refer to findings of Römmelt 2425, who found in healthy individuals an increased absorption of terpenes from bath additives containing pine needle oils or rosmary oil. Terpenoids are a subclass of the prenyllipids (i.e., terpenes, prenylquinones, and sterols) and represent the oldest group of small molecular products synthesized by plants. The major components of rosemary oil include monoterpene hydrocarbons (alpha and beta-pinene), camphene, limonene, camphor, borneol, cineole, linalool, and verbinol. Rosemary oil is known to possess marked antibacterial, antifungal, and antiviral properties; moreover, its exerts significant antioxidant activities depending on the terpenes 383940. Nevertheless, the clinical significance of a higher terpene concentration in the blood of patients or healthy volunteers using ODBs still remains unclear and has to be investigated in clinical studies.

The cardiovascular responses during water immersion show similar effects compared to ODB: a decrease of heart rate and blood pressure 41. Hence, although the cardiovascular responses during and after ODB indicate beneficial effects, it remains to be explored if the dispersed oil has specific effects on cardiovascular regulation.

Beyond the obvious beneficial effects of warm baths and the therapist's attention on the subjective well-being, it remains to be clarified what the unique contribution of the distinct essential oils dispersed in the water can be. In addition to further basic work that explores how the oil dispersion bath may influence physiological parameters beyond the effects of warm bath alone, well designed controlled clinical studies exploring the efficacy of ODB are strongly recommended to create a profound evidence base for this unique form of treatment.

The Chair of Medical Theory and Complementary Medicine received financial support with a grant of the Raphael Medical Centre, Hildenborough, Kent, GB.

AB, GB, TO and DC analyzed and reviewed the data, AB, TO and DC drafted the manuscript. PFM and FE contributed to draft the manuscript. AB and TO were principle authors of the paper, and had full access to the data. All authors contributed to manuscript revision and approved the final manuscript.