The study was conducted at Thika Level 5 Hospital (TL5H) in Kiambu County, Kenya at the Diabetes Comprehensive Care Centre (DCC). The clinic attends to approximately one hundred patients per week. The DCC is an out-patient clinic that operates on a daily basis. Diabetic patients from Kiambu County and nearby areas attend the clinic on appointment days for routine monitoring of blood glucose, blood pressure and nutrition status (body mass index; BMI), as well as for treatment and collection of medication. Newly diagnosed patients with either Type 1 or Type 2 Diabetes mellitus are also referred here from neighboring health facilities for further management. The clinic serves both male and female patients with Type 1 and Type 2 diabetes mellitus. The patients are mainly from low and middle income backgrounds.
Study design and ethics
This was a randomized controlled trial, with two intervention groups (nutrition education; NE and Nutrition education with peer to peer support; NEP) and a control group (C). The study was approved by the Kenyatta National Hospital-University of Nairobi Ethics and Research Committee (KNH-UoN-ERC), Permit No: KNH-ERC/A /232, and, the Kenya National Commission for Science, Technology and Innovation (NACOSTI); Permit No: NACOSTI /P/16/83452/10118. Study participants gave a written informed consent before the start of the study.
Study participants were men and women, aged 20–79 years, with Type 2 diabetes mellitus attending care at the Diabetes Comprehensive Care Centre (DCC) at TL5H. They were recruited during their daily clinic attendance while waiting to see a health professional. Recruitment was done over a period of 2 months from August 2016 to October 2016. All patients who met the following criteria were selected: patients suffering from Type 2 diabetes mellitus aged between 20 and 79 years, regular attendance at the DCC; not planning to move from the study area during the study period; not pregnant; with no complications such as renal failure, congestive heart failure, or stroke. A total sample size of 153 patients was recruited for the study.
Sample size determination
To confer 90% power at 5% level of significance, and to detect an absolute effect size of 30% improvement on metabolic syndrome (MetS) in Type 2 diabetes mellitus patients (i.e. a decline from 90 to 65% Mets prevalence with intervention), we needed to include 46 study participants in each study arm using the formula by Armitage et al.,  and Lwanga and Lemeshow . The sample size was subjected to a correction factor of 10% to cater for attrition, thus each arm had 51 participants making a total sample size of 153 patients.
The study consisted of two intervention groups and a control group. The Nutrition Education (NE) group received nutrition education; the Nutrition Education with Peer-to Peer support (NEP) group received nutrition education with peer to peer support; while the control group (C) received standard care. Participants were randomized to either NE or NEP or C groups by use of random numbers as shown in Fig. 1. To allow equal chances for participants, randomization was stratified on the basis of sex and age. Sealed sequentially numbered opaque envelopes per each stratum (1–3), mixed using the lottery method were used. The participants were requested to pick an envelope each and join their groups (1–3). A volunteer from each group was then requested to move forward and pick another envelop each, that contained their treatment allocation (NE, NEP and C). Upon confirmation of the treatment allocation, the participants were allocated to their treatment group by the principal investigator (PI), and the group members recorded. Each group was assigned 51 participants. After randomization baseline data was collected from all the participants. Randomization and flow of the participant throughout the study is as shown in Fig. 1.
Before random assignment to control or intervention groups, all study participants received standard education that covered content on diabetes pathophysiology, risk factors, symptoms, complications, hyperglycemia and hypoglycemia symptoms and foot care treatment goals and modalities. This was done by the principal investigator (PI) together with a clinician who runs the clinic (Registered Clinical Officer with a Bachelor of Science degree in Clinical medicine). The Standard Education relied on pictorial flip charts and additional learning material with diabetes management information. These were adapted from the diabetes prevention and management guidelines from the Ministry of Public Health and Sanitation (MoPHS), Kenya ;, the NorvoNodisk Changing Diabetes poster, as well as diabetes posters from the Ministry of Health (MOH), Kenya, with supplementary information provided by the PI obtained by a review of different literature. Different teaching methods including lectures, discussions, demonstrations, role plays and group work were used to deliver the information. The participants also received standard care that included blood glucose and blood pressure monitoring, treatment for those with problem as well as education on diabetes care by a clinician on monthly basis.
After the standard education, the intervention groups (NE and NEP) underwent a nutrition education programme for 8 weeks, which also covered the importance of physical activity (NE group). The curriculum for this programme is provided in the Appendix. In addition, the NEP group was trained on peer-to-peer support. The nutrition education given to the NE and NEP intervention groups included weekly (120 min each) nutrition classes conducted over 8 weeks by the PI. The nutrition education curriculum was developed by the PI after review of related literature on nutrition management of Type 2 diabetes mellitus. The PI also applied her experience gained from practice as a nutritionist. The NE curriculum was written in English and supplemented by photos and illustrations to help the patient understanding the content better. It focused on nutrition in relation to diabetes; food portion control for weight reduction; healthier food choices; individualized meal planning,; glycemic index and glycemic loads of different food and their importance in blood glucose control; the food pyramid, and its use together with food exchange list in meal planning. Patients learnt about the basics food groups, the difference between simple and complex carbohydrates and their relation to glycemic index and glycemic load, fibre content of different cereals and starches, the difference between saturated and unsaturated fats and their relation to diabetes management; sources of protein and the different nutrient content of each, hidden calories contained in beverages, and the micronutrient and fiber values of fruits and vegetables. The nutrition education content was presented using lectures, demonstrations, discussions, and other participatory methods. The nutrition education curriculum was first tested in a subgroup (10% of sample) of patients not involved in the study before the actual implementation. The physical activity lesson was given to the intervention groups (NE and NEP group) in the last week of the education programme. The aim of the physical activity was to ensure that patients accumulate a minimum of 150 min of moderate intensity exercise each week from personal activity at home that includes walking, digging, jogging, cycling, house hold duty, aerobics and sport activities. The participants were encouraged to perform the exercise at least 3 days each week with no more than two consecutive days without exercise. During the physical activity lesson, the patients were led through the importance of physical activity in management of Type 2 diabetes. Additionally, demonstrations on activities they can do at home were done by a physiotherapist experienced in diabetes management t. The participants were encouraged to continue with the exercises at home in addition to normal routine work.
Participants in the NEP group were divided into small support groups (5–10 participants); depending on the location they came from as well as their age. After each education session, members of the support groups were encouraged to set and share with one another other weekly goals for specific changes in their eating and physical activity behavior. The goals were aimed at making healthy food choices, reduction of portion sizes and being active. The participants reported on their progress to the group members at the beginning of the next session. After the 8 week training, participants were followed monthly, and they presented their progress and new goals to the group members, for a period of 6 months. A trained peer educator living with diabetes for 13 years from Kenya Defeat Diabetes Association (KDDA) joined the PI during the monthly meetings and encouraged the participants in the peer support groups by sharing his experiences. Together with the PI he also assisted them review and adjust their goals during monthly meetings. Also, group counseling was done on each visit for participants requiring more support.
The follow up was done monthly after the intervention period. After the end of the 8 weeks intervention the patient were requested to be coming to the hospital monthly on selected days for follow up. At the start of the study the patient were given appointment cards developed by the PI indicating the day they were supposed to come for the appointment. The PI also got phone numbers for the participants which assisted in follow up. A call was given to the participant reminding them on the appointment day 1 week to the appointment day and 2 days to the appointment day to ensure they avail themselves. Those who did not turn up would be given another day and be reminded again of their appointment. For those who could not make to come after second reminder, they were followed in their home and requested to come for the appointment. This prevented loss to follow up. Patient in the NEP group continued with peer to peer support during the follow up period.
Measurements were taken on anthropometry and clinical data, blood pressure, blood glucose and lipid profile, as well as physical activity levels and food intake. A physician and clinical officer were also present during the study period to manage any patient requiring medical treatment.
Anthropometry and clinical data
Anthropometric measurement that includes weight, height, waist and hip were collected using standard protocols [38, 39] at baseline, during monthly follow up and post evaluation after 6 months. Height and weight were measured using standard methods with the participants wearing light clothes and no shoes . The weight was determined to the nearest 0.1 kg using a calibrated electronic weigh scale (Seca) and height to the nearest 0.1 cm using a stadiometer attached to the weighing scale. Body mass index (BMI) was then be calculated as weight (kilograms)/height (meters) 2 and classified as per WHO classification . The waist circumference and hip circumference were measured according to standard guideline . Waist circumference was measured mid-way between the lower rib margin and the iliac crest with flexible anthropometric tape to the nearest 0.5 cm while hip circumference was measured as the maximal circumference around the buttocks posteriorly and pubic symphysis anteriorly.
Blood pressure of the patient was also taken monthly. It was measured in the supine position using, a mercury sphygmomanometer (model: Autortensio® noSPG440) by trained nurses with at least a 10-min rest period before the measurement.
Blood samples were collected from each participant while in a seated position after fasting for 8-12 h for determination of serum triglycerides (TG), total cholesterol (TC), high density lipoprotein (HDL-c), low-density lipoprotein cholesterol (LDL-c), glycated hymoglobin (HbA1c) at baseline and 6 month post intervention. Fasting blood glucose was determined monthly. Levels of serum triglycerides (TG), total cholesterol (TC), high density lipoprotein (HDL-c), low-density lipoprotein cholesterol (LDL-c), were determined by enzymatic method [40,41,42,43,44,45,46]. Glycated Hemoglobin (HbA1c) and blood glucose were determined using high-performance liquid chromatography and glucose oxidase method respectively [47, 48].
Metabolic syndrome definition
Metabolic syndrome in the study was defined according to the definition of WHO  and “Circulation for Harmonizing the Metabolic Syndrome” criteria [2, 21]. The latter requires the presence of at least three of the following five components: Fasting blood sugar of 100 mg/dl or 5.6 mmol/l or drug treatment of elevated glucose, central obesity for Africans (waist circumference ≥ 94 cm in males and ≥ 80 cm in females), elevated triglycerides (≥1.7 mmol/l or 150 mg/dl and/or the use of triglyceride-lowering drugs), reduced HDL cholesterol (< 1.0 mmo/l or < 40 mg/dl in males and < 1.3 mmol/l or 50 mg/dl in females) and elevated blood pressure (systolic blood pressure ≥ 130 mmHg and/or diastolic blood pressure ≥ 85 mmHg and/or the use of antihypertensive drugs).
World Health Organization criteria also requires the presence of Type 2 diabetes mellitus, impaired glucose tolerance or insulin resistance, and any two of the following:(1) body mass index (BMI) ≥ 30 kg/m2 and/or waist-to-hip ratio > 0.90 (male), > 0.85 (female); (2) blood pressure ≥ 140/≥90 mmHg or on hypertension medication; and (3) triglyceride ≥1.7 mmol/Land/or HDL-C < 0.91 mmol/L (male), < 1.01 mmol/L (female).
Physical activity data was collected using a physical activity questionnaire. It included questions asking the participants the type of activities they did, the time spent on each activity and number of days per week on each activity. The metabolic equivalent for each physical activity was tabulated and recorded. This was done at baseline, month 1, month 3 and month 6 post intervention.
This was collected by asking the participants 12 questions on healthy dietary choices adapted from perceived dietary adherence questionnaire (PDAQ) , dietary approach to stop hypertension questionnaire (DASH)  and medical nutrition therapy (MNT) [51, 52]. These questions sought to inquire whether the participants followed their commendation of; health diet plan, diet rich in fruits and vegetables, complex carbohydrates high in fibre, low glycemic index food that included whole grains, reduced intake of saturated fat and overall fat, included fish or fish products in their meal, reduced intake of sugars and sugar sweetened products, spaced carbohydrate intake, reduced intake of salt, included low fat food in the meal as well as, uptake of monosataurated and polysaturaed fat. The responses to the questions were based on a 7-likert scale.
The data was analyzed using statistical package for social science (SPSS version 20). Data are present as means ± SD or SE for continuous variables and percentages for categorical variables. Chi square test and multinomial regression was used to compare groups for categorical variables and Analysis of Co-variance (ANCOVA) was used to compare difference of means between groups. Statistical significance was considered for p value < 0.05.