The amount of people with diabetes has now reached over 4 million in the United Kingdom. Type 2 Diabetes accounts for the majority of all cases of diabetes and increases the risk of many other diseases, such as heart disease.

Research has shown that individuals from South Asian descent have elevated risk of certain chronic diseases, such as type 2 diabetes and cardiovascular disease. The risk of developing these diseases can be reduced by engaging in a healthy lifestyle. One component of a healthy lifestyle is engaging in physical activity. However, previous research has shown that South Asian individuals engage in less physical activity, compared to other ethnicities. It is not yet wholly understood why South Asians engage in less physical activity, but it is vitally important to try and find new ways to increase South Asian physical activity levels. Therefore, the investigators have worked with South Asian communities in identifying culturally appropriate forms of physical activity. From this, yoga and walking were identified as two forms of culturally appropriate physical activity. This study will test and compare whether yoga and light intensity walking can be effectively used in the prevention of type 2 diabetes. The results of this study will help the investigators and health policy makers understand how South Asians and other ethnicities respond to exercise, the therapeutic benefits of yoga and help inform future diabetes prevention programmes within multi-ethnic communities.

Physical activity, commonly defined as any movement of skeletal muscle that requires low-mid level energy expenditure has long been associated with improving cardiometabolic health. Furthermore, increased PA has more recently been associated with decreased all-cause and chronic disease mortality. Evidence on lifestyle risk factors such as PA continues to grow and demonstrate associations with improved health. Interestingly, evidence has suggested type 2 diabetes is a 'lifestyle disease', with 80-90% of the prevalence in any given ethnic group explained by lifestyle and environmental factors. This suggests lifestyle factors, such as PA and diet may play a pivotal role in the prevention and management of T2D. The number of individuals with diabetes has risen from 108 million in 1980 to 425 million in 2017, with global prevalence in over 18 year olds rising from 4.7% to 8.5% in the same time period. This illustrates the importance of PA in preventing and managing T2D.

Diabetes is a disease characterised by chronic hyperglycaemia with disturbances to carbohydrate metabolism, resulting from the body's impaired ability to produce or respond to insulin. Diabetes can lead to a number of complications which can reduce a person's quality of life and life expectancy. The main cause of premature mortality with diabetes is cardiovascular disease. Insulin resistance is suggested to be the predominant factor in the aetiology of T2D. PA has been shown to be a stimulator of insulin sensitivity, both acutely and chronically. Epidemiological data on PA (and other lifestyle habits) and T2D is important as it provides associations between exposures and the disease. This research can give a sense of an association and where robust evidence reinforces these associations, it can be the starting point for initiating disease prevention programmes based on reducing/increasing exposure to the risk/preventative factor.

A systematic review and network meta-analysis of lifestyle, pharmacological and surgical interventions stated that lifestyle interventions, which included exercise, are beneficial in reducing the risk of developing T2D, compared to standard care. When focusing on lifestyle interventions, diet plus exercise plus pedometer had the highest probability of being effective (HR 0.35 95% CI [0.11-1.14]). Exercise had the second highest probability of being the most effective lifestyle intervention (HR 0.51 95% CI [0.33-0.82]). The Indian Diabetes Prevention Programme found improved lifestyle significantly improved cardiometabolic health in a 42-month study. Improved lifestyle accounted for educational sessions revolving around improving diet and then engaging in PA. When comparing control, metformin, lifestyle and lifestyle + metformin groups it was found metformin, lifestyle and lifestyle & metformin groups significantly delayed time to diabetes, compared to control. The two groups that involved positive lifestyle changes delayed time to diabetes marginally longer than the metformin group. The solely lifestyle group delayed time to diabetes the longest over the entire 42-month period. Similarly, The Diabetes Prevention Program found that incorporating positive lifestyle changes significantly reduced the cumulative incidence of T2D, compared to metformin and placebo groups. The lifestyle group were instructed to engage in 150min.week-1 of PA. There was a 58% reduction in incidence of T2D in the lifestyle group, compared to placebo, in a 4-year follow-up. When participants were split into their respective ethnicities, American Indians and Asians saw the greatest reductions in T2D incidence. At 10 year post trial follow up, the lifestyle group still had significantly lower cumulative incidence of T2D.

The molecular mechanisms by which PA affects insulin have been heavily researched. In brief, after one bout of exercise there is an immediate increase in glucose uptake to the skeletal muscle, which is in response to muscle contraction. This increase is thought to be due to increased GLUT-4 translocation to the cell surface, which increases insulin sensitivity. Additionally, exercise may increase oxidative metabolism in mitochondria in the skeletal muscle. An increase in the size and number of mitochondria in the cells improves oxidative capacity, decreasing the amounts of TG and fatty acids. TG and fatty acids interfere with the insulin signalling process, which may cause insulin resistance. This reiterates the importance of PA in T2D prevention and management. Additionally, physical inactivity and sedentary behaviour have both been noted to disrupt normal metabolic state and chronically increase insulin resistance/decrease insulin sensitivity.

Other mechanisms that PA may improve T2D by are improving lipoprotein profile, endothelial function (and subsequently hypertension), reducing obesity and increasing anti-inflammatory effects.

Evidently, PA is vitally important in the prevention and management of T2D. However, rates of PA are worryingly low in the UK and globally. Globally, it is stated 1 in 4 adults are not active enough and more than 80% of world's adolescent population are insufficiently physically active (WHO). Within the UK, 39% (20 million) of adults do not meet recommended guidelines for PA. Importantly, PA levels in UK South Asian populations are reported to be lower, compared to White European populations. This trend is particularly evident among women and older individuals. Consequently, specific ethnicities may need culturally appropriate and alternative forms of PA to engage them. To date, public health guidelines have largely been based on results from studies in Caucasian populations. It is essential that governments and public health organisations engage individuals in more novel, appropriate and culturally specific practices.

Therefore, researching culturally appropriate and alternative forms of exercise is important in understanding whether different types of exercise will engage individuals more and improve adherence, ultimately improving health. For example, SAs form the largest ethnic minority group in the UK and 1 in 5 individuals globally and while it is known PA can positively impact the health of SAs, it is not fully understood what types of PA can engage SA ethnic groups most effectively. This is of particular importance because the prevalence of T2D is 2 to 4 times higher in SAs, compared to WEs. India is estimated to have a diabetes prevalence of 134.3 million by 2045, with Pakistan and Bangladesh estimated at 16.1 and 13.7 million, respectively.

Accordingly, culturally appropriate forms of exercise are needed. Intervention trial data has suggested that yoga may yield similar health benefits to conventional exercise. Yoga is an ancient Indian tradition that embodies unity of body, mind, emotion and energy and has been part of traditional Indian spiritual practice for millennia. There are different forms of yoga that focus on slightly different aspects of one's self, with these respective philosophies varying focus on meditation (dhyana), body postures (asana) and breathing techniques (pranayama). Regardless of these different philosophies, yoga has over recent years become more prominent in improving physical and mental wellbeing worldwide.

Two recent systematic reviews on yoga and T2D state yoga improves HbA1c, fasting blood glucose and postprandial blood glucose. Additionally, significant improvements were found in lipid profile, blood pressure and body composition. Overall, the metabolic benefits of yoga are not fully understood, particularly in those at high-risk of T2D. Further research is required on the metabolic benefits of yoga. Additionally, it is suggested that yoga may also have benefits to mental health, quality of life and emotional state. Data are needed to evaluate how individuals may respond psychologically to yoga, compared to traditional exercise, especially when used as a therapy to improve metabolic health in those at high diabetes and cardiovascular risk.

There is a gap in the literature investigating the metabolic benefits of culturally appropriate and alternative forms of exercise in different ethnicities. Specifically, there is a lack of high-quality research investigating the health benefits of yoga in relation to T2D prevention, and comparing health benefits of yoga and traditional forms of PA. Therefore, this research will investigate the effects of yoga on glycaemic control in a multi-ethnic population at risk of type 2 diabetes and to compare these effects against matched traditional exercise and a non-exercise control.

Statistical design Based on a previous study undertaken by the investigators group in a multi-ethnic population (https://clinicaltrials.gov/ct2/show/NCT02453204), 24 are needed to complete the trial. This is based on estimating a reduction in incremental area under the insulin curve from 76 to 53 mU/l x hr (30% reduction) in either exercise intervention group, a standard deviation of 38 mU/l x hr, a within person correlation of 0.05, a power of 80% and a significance of 0.05.

Study Setting The study will be co-ordinated within the Leicester Biomedical Research Centre (Leicester Diabetes Centre) at the Leicester General Hospital. Clinical measurement sessions will be carried out by the appointed research team. Participants will be asked to visit the study centre on six occasions.

Study design This study is a three period crossover design. There are three experimental conditions (yoga, continuous exercise and control) and participants will be randomised to one of six sequences requiring them to complete all three conditions.

Inclusion Criteria:

• Male or female
• ≥18 to ≤75 years of age
• HbA1c >5.7% and BMI ≥23kg/m2* / ≥25kg/m2 (WE) OR BMI ≥27.5kg/m2* / ≥30kg/m2 (WE)
• Do not engage in regular sports or strenuous physical activity
• Able to walk and use a treadmill (to engage in light-moderate physical activity)
• No medical conditions that affect balance and ability to undertake yoga postures
• No other current medical conditions
• Ability to communicate in and understand English to participate in the informed consent process *Cut off points for BME background individuals

Exclusion Criteria:

• Engage in regular purposeful sport or strenuous leisure time exercise (>120 minutes self-reported exercise per week)
• HbA1c >8.0%
• Use of glucose lowering medication
• Inability to stand or undertake light-moderate physical activity.
• Diagnosed psychological condition that limits the psychological outcome component of the study (e.g. depression)
• Ongoing CVD
• Steroid abuse
• Current smoker
• Pregnant/lactation
• Inability to understand English
• Inability to give informed consent

There is no plan to make individual participant data (IPD) available to other researchers