Complementary and alternative system of medicine for type 2 diabetes mellitus and its complications: A scoping review [version 1; peer review: awaiting peer review]

Background : This review aims to map the evidence on the effectiveness of Complementary and Alternative Medicine (CAM) for Type 2 Diabetes Mellitus (T2DM) and its associated complications and identify research gaps in key outcomes and CAMs. Methods : We followed the Johanna Briggs Institute guidelines to undertake this scoping review. We carried out a comprehensive search from inception to March 2020 in electronic databases and registers. Forward and backward citations of included studies were also searched. The randomized controlled trials (RCTs) and non-randomized studies with intervention and control arm were mapped based on the effectiveness of various CAM (Yoga, Ayurveda, Homeopathy, Siddha, Naturopathy, Unani, and Sowa-rigpa) on outcomes among individuals with prediabetes or metabolic quality RCTs and systematic reviews are needed to bridge the gap of evidence in CAM therapies. Future investments for research in CAM may focus on assessing the quality of existing trials and finding out an optimal dose, frequency, and duration of CAM, both as an alternative and complementary approach.


Introduction
Description of the condition Type 2 diabetes mellitus (T2DM) is one of the important public health problems and accounts for 90-95% of all diabetes. 1 T2DM is a long-lasting metabolic disease caused as a result of a combination of resistance to insulin action and an insufficient compensatory insulin secretory response. 2, 3 Globally, the number of people having diabetes mellitus is rising. 4,5 In 2017, 6.28% of the world's population had T2DM, which is equivalent to approximately 462 million individuals. Furthermore, these estimates are expected to rise to 7,079 individuals per 100,000 by the year 2030. 4 T2DM ranked seventh and ninth in the year 2017, among the leading causes of disability adjusted life years lost and deaths, respectively, as compared to the nineteenth and eighteenth position, respectively, in 1990. 4 Diabetes in general has a significant effect on individuals' functional capacities and quality of life (QoL). 4 Individuals with diabetes are vulnerable to many health-related complications such as cardiovascular diseases, kidney failure, delayed wound healing (leading to diabetic foot and ultimately to leg amputation if not treated), vision loss, nerve damage 3,6 and microvascular damage. 5,7 It is also known to reduce life expectancy. 3 Diabetes and its complications bring about significant financial loss to individuals with diabetes and their families through direct medical cost and loss of work and wages, and in some cases catastrophic health spending. 3 Increased duration of treatment, presence of complications, hospitalization, surgery and insulin therapy leads to increased expenditure 8 affecting the health system and national economy. 3 Description of the intervention Alternative treatments with potential anti-hyperglycemic effects are commonly used for chronic diseases such as T2DM. "Complementary and Alternative Medicine" (CAM) is defined by the World Health Organization as "broad set of health care practices that are not part of that country's own tradition or conventional medicine and are not fully integrated into the dominant health-care system". 9 CAM includes various types of therapies such as acupuncture, body work (e.g., massage), energy healing (e.g., reiki), herbal or traditional medicines (e.g., naturopathy, Ayurveda, Chinese medicine), mind body techniques (e.g., meditation, yoga), faith healing and music therapy. 10,11 The global use of CAM varies between 17% and 72.8% of the population. 12 Every four out of five individuals living in developing countries still depend on CAMs for treatment of certain health conditions. 10 A recently published meta-analysis found a global prevalence of CAMs to be 51% (95% confidence interval 43, 59). Use of CAM as a complementary medicine in addition to conventional therapy was found to be 78%. However, as an alternative to conventional medicine was found to be 21%. 9 A review also found that about 67% of the people who use CAMs did not disclose its use to healthcare professionals. 9 Furthermore, individuals with diabetes are 1.6 times more likely to use CAMs than non-diabetics as reported in 2014. 10 Similarly, diabetes is one of the main diseases, for which patients consult Ayurvedic practitioners and consume Ayurvedic medications. 6 In the United States, around 2-3.6 million people depend on CAMs for the treatment of diabetes, whereas in Australia and the United Kingdom about 46% rely on it. India showed a very high use of CAMs among diabetics accounting for 67% in 2014. 10 The reasons for high proportion of individuals with diabetes using CAMs are multi-faceted. These reasons could be higher cost, high level of medication adherence and fear of side-effect of modern allopathic medicines, dissatisfaction with healthcare providers, and easy availability of CAMs without doctor's prescription. 10 How the intervention might work Diabetes, a complex disease, caused due to pathophysiological changes that affects the whole-body glucose homoeostasis, is a typical example requiring an integrative medicine method for its holistic (complete) management. The pathophysiological web of diabetes along with obesity, inflammation and insulin resistance is a challenge for its management. The current diabetes treatment based on modern medicine focusses on reducing hyperglycemia through targeted molecular medications, which needs long-term use and adherence, along with a continuous increase in dose. 13 CAMs may act as complementary to modern medicines and help in contributing to holistic approach. Table 1 provides the list of some of the CAMs and its possible effect in lowering blood glucose levels.
Why it is important to do this scoping review? Reducing premature mortality from non-communicable diseases including T2DM by one-third, achieving universal health coverage and providing access to affordable essential medicines to all by 2030 are the targets set by member states as a part of the agenda for 2030 Sustainable Development. 3 The emergence of T2DM into global epidemic and accessibility and availability issues of modern medicines to many individuals in low resource-settings have encouraged low-cost, easily available alternative therapies. Most of these alternative therapies claim to manage diabetes in no or minimal side-effects, due to which, most of the people using CAMs for T2DM use it along with conventional therapy. 18 Ayurveda Ayurveda includes all facets of life, be it physical, psychological, spiritual or social. 8 The main goal of Ayurveda is to maintain a balance between the five basic elements of life such as earth, water, fire, air and vacuum. 6 The management of disease by Ayurveda involves a personalized diet, lifestyle, medicines and systemic cleansing therapies. 14 It promotes the use of various herbal preparations such as decoctions, juices and powders, all of which are of plant origin, but may also contain animal and inorganic products. 8 Approximately 21,000 ayurvedic medicinal plants have been registered by the WHO around the world, out of which 2500 types are of Indian origin and 800 have an antidiabetic potential. 7 Antihyperglycemic medications of Ayurveda uses mixtures of different herbs that lower blood sugar level. 13 Systematic reviews of clinical trials and case reports of several Ayurvedic drugs reveals beneficial effects of Ayurvedic medications on T2DM related outcomes, including reducing blood sugar levels with no major sideeffects. 6,15 Ayurveda plant extracts may also protect the organs and eventually improving the overall health and wellbeing of an individual. 16 Some of the herbal products used in Ayurveda e.g., curcumin influences insulin resistance and hyperglycemia, and it has shown to prevent harmful complications of diabetes. 17 Bitter gourd juice or decoction are known to lower blood sugar levels. 1 Research suggests that some herbs may regenerate ß-cells and overcome insulin resistance. 18 Some of the herbs are known to lower cholesterol levels and act as antioxidant. 18 Yoga Yoga is a branch of CAM that is built on the principle that the body, mind and spirituality are closely related to each other. Yoga helps to balance and harmonize the body, mind and emotions of an individual. Yoga consists of various components such as cleansing processes, postures, controlled breathing, meditation, relaxation, chanting mantras, yogic diet and spirituality. 19 Yoga poses and breathing exercises enhance strengthening of muscle strength, flexibility, uptake of oxygen and blood circulation. 20 Yoga in daily life is known to help in the management of T2DM by tackling the pathophysiologic mechanism of diabetes. This helps in glucose utilization 19 and achieving glycemic control and reducing the risk of complications in T2DM. 9,21,22 It is hypothesized that yoga may influence stress and relaxation mechanism of the body thereby, benefiting an individual with T2DM. It may enhance insulin sensitivity and decrease insulin resistance. 18 Unani Unani system of medicine considers imbalance in certain fluids (e.g., blood and bile) as the main cause of diseases. 8 Unani medicines has shown significant reduction in blood sugar level in individuals with T2DM. 23 Siddha Siddha emphasizes on maintaining a balance between environment, climatic conditions, physical activity and stress to guarantee good health. 8 Herbal (plant based). Inorganic (metals and minerals) and animal products are the common elements used in the treatment of many diseases by the Siddha system of medicine.
Preparations from Siddha elements have shown to have an antihyperglycemic effect. 24 Homoeopathy Homoeopathy is practiced globally 8 and millions of individuals use it in many diseased conditions as CAM. 25 Homoeopathy system of treatment considers a holistic approach towards the patient, to promote inner balance at the physical, mental, emotional and spiritual level. 8 The use of homoeopathy for the treatment of diabetes is common. 25 Mechanisms through which homeopathic medicines act on glycemic control are explained by using high dilution effects (physicochemical models). 26 Massage It is one of the relaxation therapies whereby muscles and connective tissues are manipulated, 20 which may decrease heart rate and blood pressure.
Massage may release muscle tension in based on patient reported outcome measures and an electromyographic testing. 18 Massage therapy may benefit in normalizing blood glucose and managing symptoms of diabetic neuropathy, by improving the blood circulation. Additionally, serum insulin action can be enhanced by massaging at the site of insulin injection. 18,20,27 Massage was found to be having stress reducing benefits, which in turn may cause body to use insulin in an effective way. 18 Naturopathy Naturopathy (the healing power of nature) refers to an integral, selforganizing curative process in living systems that creates, preserves and re-establishes health. 28 In aromatherapy, inhalation and topical application of essential oils over the skin has been used therapeutically. It is often used alongside massage. 18 Using water (cold or hot) for therapeutic purpose is referred to as hydrotherapy. It caused muscle relaxation and helps release toxins. 18 Naturopathy can have an anti-hyperglycemic effect in an individual with diabetes. Aromatherapy may act as coping mechanism in dealing with long standing disease like T2DM. Oils used in aromatherapy may influence the person to feel relaxed and get good sleep. 18 Hot tub therapy (a form of hydrotherapy) may increase the blood circulation to the muscles and hence can be beneficial to be used in T2DM. 18,20 Acupuncture It is one of the most commonly used CAMs for chronic pain. 18,20 In acupuncture, certain body points (acu-points) are pressed or penetrated to stimulate flow of energy or make the person sleep. 20,29 There are various types of acupuncture. 29 Mechanisms through which acupuncture act in T2DM is unknown. 20 However, studies on animal have demonstrated that acupuncture can activate glucose-6-phosphate. It may also influence pancreas to produce insulin and speed-up better use of insulin thereby lowering the blood glucose. 18 It is used mostly in the management of diabetic neuropathy. 20,29 Evidence on assessing the effectiveness of CAM has increased in recent years. It is recommended by the American Diabetes Association position statement that a CAMs can be used based on evidence from research. 30 To make the process of implementation of global partnership stronger, SDG 17 is persuading collection, monitoring and accountability of the data. 31 Therefore, timely, good quality and reliable data on CAM therapies for T2DM will help subsequent governments and organizations to take effective steps. Good quality data are also essential in framing the clinical practice guidelines. To inform policymakers, CAMs and the outcomes measured among T2DM and its complications should be known for implementing findings for policy, program development and direct future research. A scoping review can contribute to prioritizing evidence needs as suggested by SDG goal number 17. To the best of our knowledge, till date there is no scoping review in the area of CAMs and T2DM. Therefore, this scoping review intends to map the evidence on effectiveness of CAMs for T2DM and its associated complications. We will also identify research gaps in key outcomes and CAMs.

Methods
We followed the Johanna Briggs Institute guidelines to undertake the scoping review and adhered to "PRISMA-ScR (Preferred Reporting Items for Systematic review and Meta-Analyses extension for Scoping Reviews) Checklist". 32 The protocol was registered with Open Science Framework (DOI 10.17605/OSF.IO/6CNH7).

Eligibility criteria
Evidence source: Publications that were eligible for inclusion were randomized controlled trials (RCTs) and quasi randomized controlled trials (QRT) (i.e., interventional studies having two groups assessed pre and post and might not have adhered to strict randomization procedures) published until March 2020. Single group pre-post studies, systematic reviews and observations studies were excluded.
Population: Adults (>/= 18 years of age) with a confirmed diagnosis of T2DM (based on appropriate standards such as American Diabetic Association), metabolic disorder or prediabetic condition (marked by elevated levels of blood glucose) were eligible for inclusion. Additionally, individuals experiencing complications, which are associated with DM were included. We excluded adults living with T2DM with multiple organ dysfunction and failure, and those with life support. Additionally, we excluded studies involving type-1 DM, gestational diabetes or diabetes insipidus. In case of mixed population, we coded information of individuals with population of this review interest, if information on the subgroup was provided. If information on the subgroup was not available or in the case of missing information, the study was excluded.
Concept or interventions: Considering the authors' experience and knowledge we considered specific types of CAM therapies viz. Ayurveda, herbal medications, homeopathy, yoga, Unani, Siddha, massage therapy and naturopathy involving multiple modalities. Although CAMs included a wide range of strategies, we did not include vitamin and mineral nutritional supplementations and other CAMs such as Chinese traditional medicines, qigong, tai-chi and reiki. Additionally, a combination of CAM or CAM administered along with conventional medications and lifestyle modifications were eligible. However, these interventions had been equally distributed in all the arms of clinical trials. For orally administered therapy, a drug could be mono-or combi-preparation of any dose, frequency or duration, but it should have been restricted to one type of therapy (e.g., Ayurveda). Similarly, for yoga and massage therapy there was no restriction on type, length per session, frequency and duration. Any mode of administration of these therapies were eligible to be included such as oral or topical application.
Comparison: No intervention, placebo or lower dose of the same intervention, comparison between different therapy and active medications of conventional medicines (e.g., oral hypoglycemic medications) were eligible for inclusion. In case of co-intervention, one arm should have received at least one CAM.
Outcome measures: We considered the following outcomes after reviewing the literature, after consulting the CAM practitioners (who are known to authors).
a. Clinical outcomes: • Glycemic control: glycosylated hemoglobin (HbA1c), blood sugar levels: fasting blood sugar levels (FBSL) and 2-hour post-prandial blood sugar levels (PPBSL) • Serum insulin • Anthropometric measures such as body mass index, body weight, waist circumference • Adverse effects (e.g., hypoglycemia) • Lipid profiles Context: We did not impose any geographic or setting restriction. Studies could have been hospital or community-based.

Source of evidence selection
A comprehensive search was conducted to locate relevant records in electronic bibliographic databases, using a comprehensive search strategy. The scoping review included published studies on CAMs. However, due to time and financial constraints we considered only English publications. Search was undertaken in the following list of databases from their inception till March, 2020.
2. Database of trial registry.

Reference list of included studies was searched for eligible records (backward and forward citations in April 2021).
Search terms such as "Alternative Medicine", "Alternative Therapies", "Complementary Medicine", "Complementary Therapy", "Ayurveda", "Yoga", "Homoeopathy", "Massage therapy", "Diabetes" etc. were used to identify relevant literature. A pre-set search strategy was formed by referring to previous studies and trial search carried out initially in PubMed, which was modified with incorporating more keywords and MESH terms. A search string that was resulted from the process was altered to fit the identifiers of each database separately. The PubMed search strategy as an example can be found in the Extended data -DOI: 10.6084/m9.figshare.19512349. 33

Data management
Search results of all databases and records identified from other sources were exported to EndNote X7 software and duplicates were removed. Screening, coding and presentation of data was undertaken in Microsoft Excel.

Screening: applying inclusion and exclusion criteria
Screening was undertaken at three stages (title, abstract and full texts) by two review authors, independently. A thorough discussion was held between the review authors until consensus, in the case of any discrepancies on the exclusion of eligible records. We followed an extensive screening protocol. The review team did not have expertise in Naturopathy, Unani and Sowa-Rigpa therefore, once the screening was completed, a list of medicines was prepared to be confirmed by professional working in the field (through our extensive contacts). Based on the experts' recommendations CAMs were categorized.

Data charting
A pretested coding list was used that consisted of the following variables in details; citation details, study design, country, population (number, age, duration of diabetes, history of oral hypoglycemics and details on the complications), CAM therapies (type, dose, frequency, duration), comparator and outcomes. Data charting was done by six authors. As data charting was not done independently, each study extraction was cross-checked by second author. Any discrepancy on charted data was resolved with discussion until consensus. We did not carry out critical appraisal of included studies.

Analysis of the evidence
The extracted information was mapped descriptively using tables. Frequencies of population, intervention, study details, and country has been reported.

Results
On conducting searches in various databases, registers and reference searches, we retrieved 3245, 105 and 448 citations, respectively. Of which 249 records are included finally. The study selection process is shown in Figure 1 with reasons for exclusion.
As per Figure 2, the published studies assessing the effect of CAM on diabetes/prediabetes have gradually increased over the years. Additionally, there are three protocols 34-36 that we did not consider while summarizing the findings. Table 2A contains detailed characteristics of included studies and citations and is provided in the Extended data-DOI: 10.6084/m9.figshare.19512349. 33 In the below text, we have summarized the characteristics of included studies.

Characteristics of included studies
Study design: Majority of the studies were RCTs, however we also included quasi-randomized/non-randomized studies having an intervention and control group. Twenty-one studies were efficacy, feasibility or pilot interventional studies.
Country: Studies were carried out in various parts of the world. The geographical distribution of the studies is depicted in Figure 3. As seen in the figure, most of the studies were conducted in India and Iran; both contributing to three-fourths of the identified studies. About four-fifth studies were conducted in Asia compared to other continents.
Intervention details: Herbal or Ayurveda products were assessed by majority of the studies (n=146), followed by yoga therapies (n=52), other CAM including massage and topical applications of herbal products (n=52), Unani medicines (n=2), Sowa-Rigpa (n=1) and Homeopathy (n=1). Within these broad categories various unique complementary therapies were identified and are listed using botanical or trade names in the Extended data -DOI: 10.6084/m9. figshare.19512349. 33

Mapping the interventions and outcomes
With the help of tables, we have mapped the studies that measured the effectiveness of various CAMs on outcomes among individuals known to have prediabetes or metabolic syndrome, T2DM without any comorbidities and T2DM with comorbidities.
b) T2DM without any comorbidity: Most measured outcome was FBSLs (n=31), followed by HbA1c (n=21), lipid profile (n=16), QoL (n=16), PPBSL (n=15), anthropometric measures (n=12), insulin or insulin resistance (n=7) and adverse events (n=2). Table 2 depicts the number of studies that measured the effect of yoga on various health outcomes.    Table 3A in Extended data 33 provides the citation details of these studies.       Table 3B in Extended data 33 provides the citation details of these studies. 1 Ayurveda poly herbal product.    Table 3C in Extended data 33 provides the citation details of these studies.      Table 3D in Extended data 33 provides the citation details of these studies.

Ayurveda or herbal products for prediabetes and T2DM
a) Prediabetes or metabolic syndrome: Curcuma longa (turmeric) was studied by three studies, followed by Cuminum cyminum (cumin), Ayurveda polyherbal medication and other products.
b) T2DM without any associated comorbidity: T2DM population were further classified as taking hypoglycemic modern medicines or not. Momordica charantia L. (bitter melon), Trigonella foenum-graecum (fenugreek seed), Cuminum cyminum (cumin), as compared to other products were mostly studied among individuals having T2DM who were newly diagnosed or had no history of oral hypoglycemic drugs as standard therapy. Whereas cinnamon, Crocus sativus L. (saffron), Juglans regia (walnut), Momordica Charantia, and Trigonella foenum-graecum were majorly identified as herbal products that assessed glycemic indices, lipid profile, anthropometric measures and other outcomes among individuals with T2DM on hypoglycemic agents. Table 3 depicts the number of studies that measured the effect of Ayurveda or herbal products on various health outcomes.

Other CAMs for prediabetes and T2DM
Effect of acupuncture on glycemic indices was mostly studied among those identified and studied (See Table 4).

CAM for T2DM with associated comorbidity
We identified studies that assessed the effect of various CAMs on comorbidities that are associated with T2DM. These are mapped in Table 5. Massage for diabetic peripheral neuropathy, laser therapy and honey dressing for diabetic foot ulcer were most studied CAMs.

Discussion
This scoping review provides an up-to-date evidence on studies conducted to identify the effectiveness of CAMs on various health outcomes among T2DM and individuals who had prediabetes. Multiple CAMs were identified, of which yoga and Ayurveda or herbal products were most studied. Majority of these studies were conducted in India and Iran. One of the reasons for identifying higher number of studies from these countries on Ayurveda, herbal products and yoga could be that these products are deeply rooted in the tradition and thus are culturally accepted by the people. Additionally, there might be higher awareness among academic and research communities and promotion of these products by some of the international organization such as the WHO. For instance, one-hundred and seventy WHO members have admitted of promoting and integrating traditional and complementary medicines into their health services. 37 Furthermore, yoga has also been promoted by the United Nations, which in its general assembly established, June 21 st as International Yoga Day. 38 Although we identified a lot of literature on the effectiveness of CAMs for T2DM, there is a lack of standardization of preparation of these medicines and interventional modalities. This leads to a great setback in the acceptance of CAM for the management of T2DM among the scientific fraternity. Unlike modern medicines, the concept of most of CAMs' mode of action over any disease is not solely based on the active principles. Each CAM is, however, based on its own principles and approach through a personalized medicine. For example, in Ayurveda, parameters like Rasa (taste of substance), Guna (properties), Veerya (active principle), Vipaka (final transformation), Prabhava (specific or special power of the drug) are important while deciding the medicines; and each patients gets a tailored medicine, diet and lifestyle advice. 6,15 May be because of the tailored approach of many CAMs, it is difficult to design a robust RCT, which might be the reasons for not identifying or limited evidence on some of the CAMs in the current scoping review. We did not identify any studies on the Siddha system of medicine. Also, the evidence on Unani, Homeopathy and Sowa-Rigpa is limited.
Although some of the CAMs are found to have used complementary to conventional therapies, 39 there are many CAM formulations available in the market that haven't been tested for efficacy or effectiveness. 20 Lack of high standard, quality CAM clinical trials with inadequate suitable methodology and systematic documentation and reporting have led to lack of credibility of evidence-based science. Owing to the poor quality of some of the CAM trials, they are often published in journals having lower impact. 40 It is one of the reasons we identified almost three-fifth studies through reference searching of included studies. Furthermore, almost two-fifths of identified studies employed a 50 or less sample size and many studies were of short duration, which may be due to lack of adequate funding. Despite this, as discussed previously in the introduction section, vast majority of people (e.g. Asian) use CAMs and, in many cases, do not disclose its use 20 with the fear of disapproval by the conventional healthcare professional. 41 To validate the potential benefits of CAM, robust research with suitable methodology such as high quality RCTs and systematic reviews can bridge the gap of CAM therapies in delivering a holistic approach to the wellbeing of humankind. Future investments for research in CAM should be focused on assessing the quality of existing trials and finding out optimal dose, frequency, and duration of CAM, both as an alternative and complementary approach. It is also important to carry out rigorous systematic reviews of CAMs and our scoping review can provide the base for such reviews. Also, researchers should consider carrying out cost-effectives studies on CAM. In this scoping review, we did not identify any of the studies that measured economic outcomes of CAM. In the recent past the "National Health and Medical Research Council, Australia" has banned practice of homeopathy in Australia as there was no evidence to demonstrate the effectiveness of Homeopathy and this was achieved through conducting a robust systematic review. 42

Strengths and limitations
Our extensive scoping review provided the groundwork for a systematic review in coming times to investigate the effect of CAMs on important health outcomes. Our intent was to map the evidence hence, we did not critically appraise the included studies. We also did not group the multiple publications arising from same studies. However, future researchers intending to do systematic reviews could overcome these shortcomings. Although we carried out extensive exercise of identifying the literature, important subject specific databases were not freely accessible, and had language barriers. Additionally, while searching we did not use unique names (or botanical names) of the individual drugs. Due to the aforementioned reasons, we do not refute the possibility of missing important publications.
Changes from protocol: We changed the title and throughout manuscript the terminology 'AYUSH system', which has restricted usage in India to 'Complementary and alternative system' of medicine so as to focus globally. At the protocol stage, we had an intent of mapping systematic reviews, however, due to large volumes of studies we excluded at a full text screening stage.

Conclusion
This scoping review summarized Ayurveda and herbal products, yoga, naturopathy and other complementary and alternative therapies for prediabetes and T2DM by including 249 interventional studies. Most studies were conducted in India and Iran. This scoping review also identified gaps in evidence in effectiveness of CAMs and outcomes. Designated ministries and governmental agencies must train CAM practitioners in clinical trial methodologies as it would help in thoroughly appraising the evidence to guide future practice. Research of this magnitude should be collaborative in nature where physicians from allopathic systems of practice, methodologists and practitioners of CAM should work in parallel to systematically synthesize literature on clinical conditions of interest. Although funding for evidence synthesis for CAM is generally low across countries, it is a need of the hour to direct resources to support evidence-based practice.