. 2020 Mar 12;20(Suppl 1):140. doi: 10.1186/s12902-019-0468-y

Table 1.

Overview of the studies targeting participants aged ≥ 18 years and minimum follow-up of 12 months

	Reach		Implementation & adaptation		Efficacy & maintenance
Daqing Diabetes Prevention Study China Pan et al., 1997 [5] Li et al., 2008 [15]	People living in Daqing area, > 18 y old IGT	110,660 people screened with OGTT 577 randomized sex: 54% men age: 45 ± 9 y	Cluster randomized controlled study in four groups: control, diet only, physical activity (PA) only, or diet plus PA FU 6 years (n = 533) + 20 years. Goals: Diet (increase vegetables, decrease alcohol and sugar, caloric and weight reduction if overweight), PA (1–2 units/day; unit = 30 min of slow walking etc., 20 min of fast walking etc. or 5 min of jumping rope etc.) or combined both	Individual counselling + compliance evaluation by physician/nurse every 3 months + small groups weekly for 1 month, monthly for 3 months and every 3 months thereafter. Intervention duration 6 years.	HRs (adjusted for baseline BMI and f-Glu): HR = 0.69 for diet vs CG, p < 0.03; HR = 0.54 for diet + PA, p < 0.0005; 0.58 (diet+PA), p < 0.005; HR = 0.41 for PA vs CG p < 0.0005. The average number of PA units per day was significantly higher after 6 years. No significant changes in diet. ++
Diabetes Prevention Study, DPS Finland Tuomilehto et al., 2001 [6] Lindström et al., 2003 [16] Lindström et al., 2006 [17] Lindström et al., 2013 [18] Wikström et al., 2009 [19]	People with high risk for T2D IGT in two OGTTs IGT (according WHO 1980) BMI > 25 kg/m² age 40–64	Opportunistic screening based on age and BMI; previous study patient files; newspaper ads. Stepwise screening (1st OGTT - > 2nd OGTT). Approximately 10% of those tested were eligible. 522 randomized sex: 33% men age: 55 years	RCT in 5 study centres in Finland. Intensive, individualized intervention vs. general “mini-intervention” at baseline. FU mean 3.2 years (n = 482) + 7 years + 13 years Goals: < 30% of total energy from fat; < 10% energy from saturated fat; at least 15 g of fiber/1000 kcal; at least 4 h/week physical activity; > 5% weight reduction. PA sessions (2/week in free gym) were offered.	7 individual counselling sessions with nutritionist (at 2w, 5w, 3 m, 4 m, 6 m (first 1 h, later 30 min), every 3 months thereafter. Sessions included pre-defined topic + review of food and exercise diaries + goal setting with participants. 1 year intensive phase plus maintenance 1 to 5 years, mean duration 4 years. Optional very low caloric diet phase. Stages of change model, emphasizing self-efficacy, monitoring, feedback, behaviour planning, relapse management.	Diabetes incidence in 3.2 year (main results) reduced by 58%.Significant difference in changes of weight (− 4.5 kg in IG vs. -1 kg in CG at year 1), BMI, waist circumference, diet, physical activity, fasting and 2 h glucose in IG compared to CG. Effects of intervention sustained after 7 years and after 13 years. ++
Diabetes Prevention Program, DPP USA Knowler et al., 2002 [7] Rubin et al., 2002 [20] Fujimoto et al., 2000 [21]	aiming for 50% ethnic minorities IGT (WHO 1980) fasting glucose > 5.5 mmol/l age > 25 y, BMI > 24 kg/m²	Recruitment with mail, advertisements in media and work sites. 133,683 individuals pre-screened, 26,518 screened with an OGTT. 3.048 randomized sex: 33% men age: 50 + 11 years 45% from ethnic minority groups	RCT in 27 clinical centres Intensive lifestyle vs. metformin vs. placebo FU mean 2.8 years, drop-out 7.5% Weight reduction 7% Diet: fat 25 E% Physical activity (e.g. brisk walking) 150 min/week (700 kcal/week)	Main goal to achieve and maintain a weight reduction of > 7% through a healthy low-calorie, low-fat diet and to engage in physical activity of moderate intensity. Goal-based behavioural intervention lasting for 2.8 years (mean); case-managers (1 per 20–26 participants) held 16-session core curriculum in groups during the first 24 weeks; individual session monthly	T2D risk reduction 58% in lifestyle vs placebo after 2.8 years. 50% of IG met the weight reduction goal after 24 weeks. Mean weight loss 7 kg (7%) at 1 year. ++
Prevention of type 2 diabetes by lifestyle intervention: a Japanese trial in IGT males Japan Kosaka et al., 2005 [22]	30–60 year old men IGT (according WHO 1980)	Random selection of men with IGT from health-screening program for mostly government employees. 458 randomised n = 356 in CG, n = 102 in IG sex: 100% men age: mean NA	RCT intensive care vs. standard hospital care (1:4). FU 4 years. Drop-outs 5.6% in CG, 4,7% in IG BMI < 22; reduce energy intake by 10%; increase vegetable intake; fat < 50 g/day; alcohol < 50 g/day; eating out once /day or less; walking 30–40 min / day	Face-to-face counselling by nurse in hospital every 2–3 months, 4-year intervention. Regular weight self-monitoring. Concrete, standardised advices to reach the goals of the study.	Relative risk reduction 67% in 4 years. Body weight reduction 2.2 kg at 4 years vs. 0.39 kg in the control group ++
Indian Diabetes Prevention Programme, IDPP-1 India Ramachandran et al., 2006 [23]	Middle-class population; 35–55 years; IGT in two OGTT (WHO 1999)	Recruiting by workplace announcements and circulars. 10,839 subjects underwent initial screening using glucometer. n = 531; sex: 81% men age: 45.9 years	RCT in community-based setting in 4 groups: Control; metformin (MET); lifestyle modification (LSM); LSM + MET. FU 3 years (n = 502). Goals: > 30 min brisk walking daily: reduction in total calories, refined carbohydrates and fat; avoidance of sugar and inclusion of fibre-rich foods	Participants had a personal session at 6-monthly intervals and were contacted by phone every month. Diet modification was advised for each subject.	Absolute risk reduction at year 3 was 15.7% in LSM, 14,5% in MET and 15.5% in LSM + MET compared to control (all p-values for HR < 0.03). ++
N/A UK, Newcastle Oldroyd et al., 2006 [24] Oldroyd et al., 2001 [25]	European origin men and women aged 24–75 years IGT in 2 OGTT (WHO 1985)	Recruited people from previous research studies, local hospital and GP databases. n = 78 sex: 57% men age: 58.2 years	RCT in hospital setting; control vs intervention FU 2 years (n = 54) BMI < 25 kg/m², dietary fat < 30 E%, polyunsaturated to saturated fat (P:S) ratio ≥ 1.0; carbohydrate 50 E%, dietary fibre > 20 g per 4.2 MJ, 20–30 min aerobic exercise at least once a week	Individual counselling from a dietician and physiotherapist using the stages of change model. Intervention lasted 2 years and 12 sessions with duration of 15–20 min.	Meeting intervention goals was higher in IG for all other but P:S ratio. In IG weight change was − 1.8 kg vs + 1. 5 kg in CG at 24 months. Significant change in fasting serum insulin between groups at 12 months. +
Hoorn Prevention Study The Netherlands Lakerveld et al., 2012 [26] Lakerveld et al., 2013 [27]	Age 30–50 At least 10% risk for T2D and/or CVD estimated according formula of ARIC and SCORE projects.	A screening invitation to GP customers (n = 8193). 2401 respond, 921 eligible based on waist circumference. 772 were screened. n = 622 sex: 42% men age: 43.5 years	RCT in general practice. FU 12 months (n = 502). Goals were at least one fruit, at least 200 g vegetables and at least 30 min PA per day.	Participants were offered 6 face-to-face sessions (30 min) and 3 monthly telephone sessions with trained nurses. Methods were based on motivational interviewing, problem solving treatment, theory of planned behaviour and theory of self-regulation. CG got brochures of health guidelines.	No significant results in weight or fasting glucose or glucose tolerance. Increase in fruit intake between baseline and 6 months (1.1 - > 1.3 pieces per day) but not after 12 months. Median participation in sessions was 2. (−)
The Joetsu Diabetes Prevention Trial Japan Kawahara et al., 2008 [28]	20–70 year old men and women IGT (ADA 2003)	Recruiting from 11 outpatient practices and health evaluation and promotion centres in Joetsu area. Patients with BMI 20–34 kg/m² and FPG ≥ 5.6 mmol/l or HbA1C 5.2–6.4% were screened n = 426 sex: 47% men age: 51.4 years	Community-based clinical trial comparing short-term hospital (STH, n = 143) or outpatient diabetes education support (DES, n = 141) to no-treatment (n = 142). Mean FU 3.1 years. Goals for STH were 25–30 kcal / ideal body weight kg / day; 20–25% E% of fat; at least 30 min/day walking or exercise at least 5 times a week. Goals for DES were to follow diabetes guidebook.	STH group had 2-day hospital stay with a course of nine group/individual lessons covering diet, exercise and behaviour modification. Subsequent sessions every 3 months were offered individually. DES group got written information and 3-monthly individual sessions of a healthy lifestyle. Lessons were 20–40 min and were taught by different medical specialists. Mean intervention duration was 3 years.	The incidence of diabetes was 42 and 27% lower in STH and DES groups compared with no-treatment group, and 21% lower in STH than DES. Also FPG, 2 h OGTT plasma glucose, HbA1C and weight changes between groups were significantly different. STH) was more cost effective than DES ++
EDIPS-Newcastle UK Penn et al., 2009 [29]	> 40 years, BMI > 25 kg/m², IGT in two OGTTs	Recruiting by invitation letter to eligible customers of primary care physicians. 1567 were contacted; 1084 replied; 682 agreed to testing; 482 completed at least one OGTT n = 102 sex: 41% men age: 57.1 years	RCT of two arms; intervention vs usual care. Mean FU 3.1 years (n = 42), up to 5 years > 50 E% carbohydrate; < 30 E% fat; reduce saturated fat intake; increase fiber intake; BMI < 25 kg/m2	Approx. 24 sessions with dietitian and physiotherapist as individual motivational interviewing for behavioural changes, including feedback from food diaries, weight and waist measurements. Cooking groups and discount of leisure service card was offered. Quarterly newsletter containing recipes, nutritional info and exercise suggestion was sent. Control group got usual care.	Weight reduction was 2.3 kg in IG vs. no change in CG, p = 0.007 at year 1 but no significant difference in consecutive years. The overall incidence of diabetes was non-significantly reduced by 55% in the intervention-group vs. the control group, with RR 0.45 (95% CI 0.2 to 1.2). (+)
PREDIAS Germany Kulzer et al., 2009 [30]	20–70 years BMI ≥ 26 IGT or IFG	Invitations by primary care physician based on FINDRISC > 10. n = 182 sex: 57% men age: 56.3	Two armed randomized control study in general practice. FU 12 months (n = 165). Goals based on DPP.	12 × 90-min group lessons in 12 months. Program was based on self-management theory and delivered by diabetes educator or psychologist. Participants got an exercise book, with diabetes prevention information and worksheets for lessons. Control group got written information about diabetes prevention.	There was significant difference in weight loss (−3.8 ± 5.2 vs. − 1.4 ± 4.09 kg), reduction in fasting glucose, increase in duration of PA and changes in eating behaviour after 12 months between intervention and control group. +
Telephone support in addition to Greater Green Triangle Diabetes Prevention Program Australia Dunbar et al., 2010 [31]	40–75 years FINDRISC > 12 points	Participants who completed the 12 month diabetes prevention program and were willing to participate in follow-up. n = 205 sex: 28% men age: 56.6 years (self-care); 57.1 (telephone support)	Telephone support vs self-care after 1 year lifestyle intervention pretest-posttest study for next 18 months FU 30 months (n = 164) < 30 E% fat; < 10 E% saturated fat; > 15 g fibre/1000 kcal; > 4 h/week moderate PA; > 5% weight reduction	Telephone support started after 12 month original intervention consisting of 6 group sessions. Telephone group got up to 12 calls following semi-structured interview with questions regarding personal goals.	There wasn’t significant difference between telephone support and self-care group. Original interventions improvements in 12 months were generally maintained to 30 months in both groups. (−) for telephone support
DE-PLAN, Greece Greece Makrilakis et al., 2010 [32]	FINDRISC ≥15 points	3240 individuals were screened with FINDRISC-questionnaires; 620 were eligible. n = 191 sex: 40% age: 56.3 y	Pretest-posttest study in community setting in Greece. FU 12 months (n = 125) Reduce saturated and trans fat, sugars and sweets, refined cereals; ≥ 5 portions of fruits and vegetables per day; PA ≥ least 30 min 5 times a week. Weight reduction was recommended if overweight.	A dietitian held 6 group intervention sessions during 1 year. Each session was focused on one of goals and included information, discussion and written material. Sessions were held in workplaces or near participants’ residence. Evaluation of achieving goals was discussed in the beginning of each session.	Weight, BMI, blood pressure, and total cholesterol reduced significantly in those who completed program. Also reduction in whole fat dairies, processed meats, sugars and refined cereals was significant. +
PREDIMED-Reus Spain Salas-Salvadó et al., 2011 [33] Martínez-González et al., 2012 [34]	Men 55–80 years; Women 60–80 years; At least 3 risk factors for cardiovascular disease: smoking; hypertension; dyslipidemia; BMI > 25 kg/m²; family history of cardiovascular disease	1125 participants were screened in primary care centre and 870 fulfilled inclusion criteria. Of these 452 were diabetics, so 418 were the final population in this sub-study. n = 418 sex: 42% men age: 67 y	RCT for primary cardiovascular prevention in three arms: “low-fat” –control; Mediterranean Diet (MedDiet) + oil; MedDiet + nuts. Median FU 4 years. MedDiet; use olive oil abundantly; increase consumption of fruits, vegetables, legumes, nuts and tomato sauce for cooking; reduce total and red meat use and use fish and white meat instead; avoidance of butter, cream, fast food, sweets, pastries and sugar-sweetened beverages; moderate use of red wine. Control: reduce all type of fat, no free foods. All groups: No energy restriction, no PA promoted	Dietitians gave personalized dietary advice to participants on basis of a 14-item (MedDiet) or 9- item questionnaire (control). At inclusion and quarterly there after dietitians administered individual and group sessions. Participants were offered written material including descriptions of seasonal foods, shopping lists, weekly meal plans and cooking recipes. In addition participants in MedDiet groups were given free virgin olive oil (1 l/week) or nuts (30 g/day).	Hazard ratio for diabetes was 0.55 (0.32–0.95 95%CI) for both MedDiets compared with control diet in crude model and 0.48 (0.27–0.86) in multivariate adjusted model. Diabetes incidence was lower in participants who complied with the MedDiet better. Largest risk reduction was seen in MedDiet in subgroups of women vs men, over 67 year-old vs under and with those whose fasting glucose > 6.1 mmol/l in baseline than those who had ≤6.1 mmol/l. ++
DE-PLAN-Krakow Poland Gilis-Januszewska et al., 2011 [35]	FINDRISC points > 14; OGTT to exclude diabetics	Recruiting from primary health care centres. 566 completed questionnaire; 368 eligible; 275 underwent OGTT n = 186 sex: 22% men age: 56.1 y	Pretest-posttest study in primary health care in Poland FU 12 months (n = 175) Goals were weight loss, reduced intake of total and saturated fats, increased consumption of fruits, vegetables and fibre and increased PA.	Active phase of intervention consisted of 10 group sessions in first 4 months followed by 6-month maintenance with six motivational phone calls and two motivational letters. PA sessions were offered once or twice a week. Social and family involvement was encouraged. Prevention managers were educated.	Significant changes in weight, BMI and waist circumference. No significant changes in fasting glucose or glucose tolerance test results. +
Making the Connection Healthy Living Program, MTC HLP USA Ruggiero et al., 2011 [36]	underserved latino population; age 18–65 years; BMI > 24.9, normal glucose or prediabetes; Latin background	Community-based health screening events (schools, family centers, hospital etc). 1162 screened, 367 tested for eligibility, 244 eligible. n = 69 sex: 7% men age: 38 y	Single-group, non-randomized follow-up, community-based translation of DPP. FU 12 months (n = 57). DPP goals: Weight reduction 7%; Diet: fat 25 E% Physical activity (e.g. brisk walking) 150 min/week	Culturally specific intervention was developed and conducted in collaboration with the community to minimize barriers to participation education, literacy, language, income, transportation, lack of medical coverage. 22 group sessions during 1 year, delivered by trained community health workers. Cook books, pedometer, scales provided. Group walks arranged. Participants attended 57% of group sessions.	At 6 m, 20% achieved 7% weight reduction, and at 12 m 16% achieved. Moderate improvements in body weight (− 4.8 kg at 6 m, − 2.8 at 12 m), waist, fruit and vegetables, fat intake, PA were observed at 6 m but attenuated at 12 m. Forward movement in “stages of change” scale was observed. +
The Prevention of Diabetes and Obesity in South Asians: PODOSA UK Bhopal et al., 2014 [37] Douglas et al., 2013 [38] Douglas et al., 2011 [39]	South Asian families living in UK > 35 years; IGT and/or IFG	Recruiting via National health service, South Asian organizations and peer to peer. Pre-screening by waist circumference. 1319 screened with an OGTT, 196 (15%) were eligible. n = 171 (156 families) sex: 45.6% men age = 52.5 y	Family-based two armed RCT in South Asians living in UK FU 3 years Goals were calorie deficit diet, at least 30 min brisk walking per day and at least 2.5 kg weight reduction.	Intervention visits were done in participants’ homes, 15 for intervention group and 4 for control group in 3 years. Practical dietary counselling: cooking, food shopping, food labelling and recipes. Counterweight program was used to dietary counselling. Change management tools, self-reflection and cultural adaptations were included.	Weight change in the IG was − 1.13 kg (SD 4.12) and in CG + 0.51 kg (SD 3.65), an adjusted mean difference of − 1.64 kg (95% CI − 2.83 to − 0.44).No difference in changes of fasting plasma glucose, OGTT or physical activity. Progression to diabetes was observed less frequently in the IG than the CG (OR 0·68) but not statistically significant (p = 0·3705). (+)
Zensharen Study for Prevention of Lifestyle Diseases Japan Saito et al. 2011 [40]	30–60 years; BMI > 24 IFG	First screening from health check-ups and eligible people were invited to OGTT (diabetics were excluded). n = 641 sex: 78% men age: median 50 y in IG, 48 y in CG	Unmasked, multicenter RCT in health care setting: frequent intervention group (FINT n = 311) vs. control (n = 330). FU 36 months (n = 562). Goals: 5% weight loss; reduce total energy intake; fat 20–25 E%; carbohydrate 55–60 E%; increase fibre intake and moderate alcohol intake. Increase of incidental PA to 200 kcal/d.	FINT got individual instructions and follow-up support for lifestyle modification from mainly dieticians and nurses at least 9 times. Self-monitoring using pedometers and body weight recording sheets. Control group got similar individual instruction 4 times during first 12-months without follow-up support or self-monitoring tools.	The HR for T2D in FINT was 0.56 (95% CI 0,36-0,87) compared to control. In FINT IFG only group HR was 1.17 (0.50–2.74), and in FINT IFG + IGT 0.41 (0.24–0.69), compared to corresponding CGs. > 5% weight loss was significantly more achieved in FINT group during the first 12 months. Also PA goals and reducing energy intake goal were achieved more in FINT group. ++
APHRODITE The Netherlands Vermunt et al., 2011 [41] Vermunt et al., 2012 [42] Vermunt et al., 2012 [43]	FINDRISC ≥13; age 40–70 years	FINDRISC-questionnaire was sent to GP patients from 14 primary care practices (n = 16,032). Individuals with a score ≥ 13 were invited (n = 1533) for OGTT and diabetics were excluded. n = 925 sex: 38% men age: 58 y	RCT in Dutch primary care (IG n = 479, CG n = 446) FU 2.5 years (IG n = 368, CG n = 341) Weight reduction ≥5% if overweight, PA for at least 30 min a day / 5 days a week, fat intake < 30 E% and saturated fat < 10 E%, dietary fibre intake ≥3.4 g per MJ	Participants were offered 11 individual sessions with nurse or general practitioner, one with dietitian and 5 group sessions with dietitian and physiotherapist. Intervention lasted 30 months. Dietary advices were based on food diary. The intervention was based on trans-theoretical model (the stages of change).	Differences between groups were significant only for total physical activity and saturated fat and fibre intake. In the intervention group, self-efficacy was significantly higher in individuals successful at losing weight compared with unsuccessful individuals. (+)
Lawrence Latino Diabetes Prevention Project (LLDPP) US, Massachusets Ockene et al., 2012 [44]	Low-income latinos 25–79 y old latino / hispanic ethnicity; ≥ 25 y old; BMI > 24; > 30% increased according T2D risk algorithm	Recruiting (n = 949) from local health centre and local media. n = 312 FU: 289 (CG 142, IG 147) sex: 25.6% men age: 52 y	Randomised community-based, culturally tailored, literacy sensitive lifestyle intervention (n = 162) vs. usual care (n = 150). FU 1 year (n = 289) Increase intake of whole grains and vegetables; reduce sodium, total and saturated fat, portion sizes and intake of refined carbohydrates and starches; increases walking by 4000 steps/day	3 individual sessions at home (1 × 1 h, 2 × 0.5 h) and 13 group sessions (1 × 1.5 h, 12 × 1 h) over 12-month period. Participants got cash incentives at baseline, at 6- and 12-months. Participation was maximized with compensatory sessions and home visits. Practical, hands-on methods and demonstrations were used. Intervention was based on social cognitive theory and patient-centred counselling.	Participants lost more weight in IG (−2.5 lb) than in CG (+ 0.63 lb), effect of intervention − 2,5 lb. (p = 0.004). Also HbA1c reduced more in IG vs CG (effect of intervention − 0.10%, p = 0.009). Participants in IG reduced more energy intake from fat and saturated fat and increased dietary fibre intake. +
E-LITE US Xiao et al., 2013 [45] Ma et al., 2013 [46]	≥ 18 years BMI ≥ 25; IFG or metabolic syndrome (2005 AHA)	Recruiting from a primary care clinic: 3439 contacted, 752 screened BL n = 241 (81 cont; 79 coach-led; 81 self-directed) sex: 53% men age: 52.9	RCT in primary care with three arms: control, coach-led group intervention vs. self-directed DVD intervention. FU 15 months (n = 221) Weight loss goal based on DPP	Intensive intervention (3 months) included 12-session in groups or at home via DVD. Coach-led group had food tasting and guided PA and self-directed group got one face-to-face session, weight scale, pedometers and biweekly reminder e-mails during 12 month maintenance phase. All intervention participants were trained to use self-monitoring web portal.	At month 24 the mean change in BMI from baseline was −1.9 +/− 0.3 kg for coach-led group (p = 0,001 vs CG); − 1.6 +/− 0.3 kg for self-directed group (p = 0.003) and − 0.9 +/− 0.3 in the control group. Fasting plasma glucose was significantly more improved in IGs compared to CG. +
RCT of SMS for Drivers With Pre-DM China Wong et al., 2013 [47]	Professional drivers IFG/IGT; had a mobile phone	Screened 3376 drivers identified by community screening and media advertisement. n = 104 subjects age: 53 y sex: 93.3% men	RCT of short message service (SMS) intervention on vs. control (leaflet) FU 24 months (IG n = 41;CG n = 29). “Diabetes-related information in reducing the risk of developing diabetes”	Participants got sms from 4 themes: diabetes information; lifestyle change; how others would appreciate the lifestyle modification; self-efficacy enhancing statements. In the first 3 months sms were sent 3 times a week, next 3 months weekly and last 18 months monthly. Both groups got leaflets about diabetes. Intervention was based on theory of planned behaviour and social cognitive theory.	No significant reduction in diabetes risk after 12 or 24 months. Significant mean differences in diastolic blood pressure and HDL-cholesterol over time between the groups. Intervention cost was 5.05 $/ subject (−)
DH!AAN The Netherlands Admiraal et al., 2013 [48] Vlaar et al., 2012 [49]	South Asian migrants Age 18–60 y; IFG, IGT, HbA1c > 6.0% or HOMA-IR > 2,39	2307 screened via general practices (invitation letter with reply card), followed by reminder and telephone call). n = 536 age: 44.9 y sex: 49.4% men	RCT in general practice among South Asian migrants in Netherlands getting a culturally targeted intervention or generic lifestyle advice (control). FU 2 years (n = 335). Goals according to SLIM study; based on current guidelines on diet and physical activity.	6–8 individual sessions in general practice during 6 months, 2 sessions during the next 6 months + 1 family session + two cooking classes. 20-week supervised exercise program was offered. Trained dieticians gave dietary counselling using motivational interviewing. Participants got a gift coupon for participating in baseline measurements. Control arm got 2 group sessions + 2 flyers.	No significant results. Median participation in 5 individual sessions. High drop-out and low participation 26% participated in family session, 26% in cooking sessions and 22% in PA sessions. (−)
“Group Lifestyle Balance™ program in the community setting USA Kramer et al. 2014 [50]	26–80 years old BMI ≥ 25 kg/m²; IFG / metabolic syndrome; ≥ 25 years old	Recruiting via GPs, information letters, e-mail contact and newspaper advertisement. n = 81 sex: 22% men age: 52.9 y	Group Lifestyle Balance (GLB) program in three outpatient-hospital in a pretest-posttest setting. FU 12 months (n = 52) 7% weight loss; 150 min PA/week	12-session group lifestyle intervention adapted from DPP delivered by trained diabetes educators in groups over 12 to 14 weeks and monthly thereafter. Each session lasted 1 h. Handouts, self-monitoring booklets, fat- and calorie-tracking book and a pedometer were given to participants.	Significant changes: weight loss was 5.1% (p < 0,001); decrease in waist circumference; fasting plasma glucose; LDL-cholesterol; triglycerides and blood pressure. +
Subanalysis of the Japan Diabetes Prevention Program Japan Sakane et al., 2014 [51]	30–60 years IGT	People measured in yearly health check-ups were recruited using posters, fliers and by word of mouth. 1279 were screened with OGTT. n = 304 sex: 50,0% men age: 59% men	RCT in 32 primary health care centres using existing resources in two arms. mean FU 2.3 years (n = 213) 5% reduction in body weight in overweight and obese; increase energy expenditure by 700 kcal per week; < 25 E% fat; < 160 kcal/day from alcohol	Study nurse held four 2–3-h group sessions during first 6 months followed by biannual 20–40 min individual sessions, intervention was 3 years. After first year individual sessions were held on phone. Personalised goals were set. A booklet was given and monthly tip cartoons were sent via fax. Self-efficacy, self-monitoring and trans-theoretical model was used.	No significant results in T2D risk. In IG participants with BL HbA1c levels ≥5.7% (n = 177) cumulative incidence was significantly lower. Significant results: at 1 year IG had improved body weight and daily non-exercise leisure time energy expenditure and at 3 years better Matsutada index. (+)
Use of Information Technology in the Prevention of Diabetes India Nanditha et al., 2014 [52] Ramachandran et al., 2013 [53]	Men No major illness; age 35–55 years; positive family history of T2D; BMI > 23; IGT	First screening with questionnaire (n = 8741) and then OGTT first with a glucometer and confirmatory venous blood glucose within 1 week. n = 537 sex: 100% men age: 46 y	RCT in industrial male workers lifestyle modification with SMS. FU 2 years (n = 517) Avoidance of simple sugars and refined carbohydrates; total fat intake < 20 g/day; restrict use of saturated fat; increase fibre; enhance aerobic exercise > 30 min brisk walk/day; walk 3–4 km in 30 min at least 5 days a week	All participants got personalised lifestyle modification in the beginning. SMS-group received mobile phone messages at frequent intervals (2–4 sms/week). SMS content was tailored according to participant’s stage of change in trans-theoretical model.	Risk reduction in sms-group compared to control was 9% (HR = 0.64 CI 0.446–0.917 p = 0.015). ++
“Lifestyle modifications in Chinese women who had gestational diabetes mellitus” China Shek et al., 2014 [54]	Age > 18 years; GDM history; IGT 6–8 weeks after delivery; excluded if insulin for GDM	Patients from hospitals were invited if criteria were fulfilled n = 450; sex: 100% women age: 39 y	RCT conducted in hospital in Hong Kong. Intensive lifestyle intervention vs. no intervention (control). FU 36 months (n = 423). Optimal caloric intake (based on Harris-Benedict) for ideal body weight	7 individual sessions in 3 years (3 m, 6 m 12 m, 18 m, 24 m 30 m, 36 m). Dietician and study nurse gave individual dietary and exercise advice based on food and exercise records (n = 7).	No significant difference in cumulative incidence of diabetes. In women > 40 years, difference between groups was significant. Significant differences between groups at 1 year in BMI and waist-hip ratios, but not significant in the end of the study. (+)
“Prevention of diabetes in Finnish airline” Finland Viitasalo et al., 2012 [55] Viitasalo et al., 2015 [56]	Airline workers (majority shift-work) FINDRISC > = 10 or IFG or IGT	Occupational health care check-up n = 2312, 657 had high risk and were offered intervention. n = 350 sex: 60% men age: 47 y	Work-site study targeted at identified high-risk workers of an airline. Average FU 2.5 years (n = 402). DPS goals and other goals according to risk factor levels (BP, cholesterol).	1–3 individual lifestyle counselling sessions in addition to the check-up by nurse/physician. Lifestyle sessions were delivered by diabetes nurse or nutritionist.	Among elevated risk men, body weight was slightly reduced and 14.3% lost > 5% of weight, and cholesterol and LDL decreased. Those men who attended more lifestyle sessions lost more weight. Fasting glucose increased in all groups. FINDRISC score increased, but less so among high-risk men. (+)

Reach

Implementation & adaptation

Efficacy & maintenance

Name of study, Acronym
Country
References

Target group, Inclusion criteria

Screening, recruitment, study population (n, sex, mean age), drop-outs

Study design, follow-up (FU) duration, lifestyle goals/targets

Intervention delivery, intervention duration, change theories

Results
Clinical significance estimate*

Daqing Diabetes Prevention Study

China

Pan et al., 1997 [5]

Li et al., 2008 [15]

People living in Daqing area, > 18 y old

IGT

110,660 people screened with OGTT

577 randomized

sex: 54% men

age: 45 ± 9 y

Cluster randomized controlled study in four groups: control, diet only, physical activity (PA) only, or diet plus PA

FU 6 years (n = 533) + 20 years.

Goals: Diet (increase vegetables, decrease alcohol and sugar, caloric and weight reduction if overweight), PA (1–2 units/day; unit = 30 min of slow walking etc., 20 min of fast walking etc. or 5 min of jumping rope etc.) or combined both

Individual counselling + compliance evaluation by physician/nurse every 3 months + small groups weekly for 1 month, monthly for 3 months and every 3 months thereafter.

Intervention duration 6 years.

HRs (adjusted for baseline BMI and f-Glu): HR = 0.69 for diet vs CG, p < 0.03; HR = 0.54 for diet + PA, p < 0.0005; 0.58 (diet+PA), p < 0.005; HR = 0.41 for PA vs CG p < 0.0005. The average number of PA units per day was significantly higher after 6 years. No significant changes in diet.

Diabetes Prevention Study, DPS

Finland

Tuomilehto et al., 2001 [6]

Lindström et al., 2003 [16]

Lindström et al., 2006 [17]

Lindström et al., 2013 [18]

Wikström et al., 2009 [19]

People with high risk for T2D

IGT in two OGTTs IGT (according WHO 1980)

BMI > 25 kg/m²

age 40–64

Opportunistic screening based on age and BMI; previous study patient files; newspaper ads.

Stepwise screening (1st OGTT - > 2nd OGTT). Approximately 10% of those tested were eligible.

522 randomized

sex: 33% men

age: 55 years

RCT in 5 study centres in Finland.

Intensive, individualized intervention vs. general “mini-intervention” at baseline.

FU mean 3.2 years (n = 482) + 7 years + 13 years

Goals: < 30% of total energy from fat; < 10% energy from saturated fat; at least 15 g of fiber/1000 kcal; at least 4 h/week physical activity; > 5% weight reduction. PA sessions (2/week in free gym) were offered.

7 individual counselling sessions with nutritionist (at 2w, 5w, 3 m, 4 m, 6 m (first 1 h, later 30 min), every 3 months thereafter. Sessions included pre-defined topic + review of food and exercise diaries + goal setting with participants. 1 year intensive phase plus maintenance 1 to 5 years, mean duration 4 years. Optional very low caloric diet phase. Stages of change model, emphasizing self-efficacy, monitoring, feedback, behaviour planning, relapse management.

Diabetes incidence in 3.2 year (main results) reduced by 58%.Significant difference in changes of weight (− 4.5 kg in IG vs. -1 kg in CG at year 1), BMI, waist circumference, diet, physical activity, fasting and 2 h glucose in IG compared to CG. Effects of intervention sustained after 7 years and after 13 years.

Diabetes Prevention Program, DPP

USA

Knowler et al., 2002 [7]

Rubin et al., 2002 [20]

Fujimoto et al., 2000 [21]

aiming for 50% ethnic minorities

IGT (WHO 1980) fasting glucose > 5.5 mmol/l

age > 25 y, BMI > 24 kg/m²

Recruitment with mail, advertisements in media and work sites. 133,683 individuals pre-screened, 26,518 screened with an OGTT.

3.048 randomized

sex: 33% men

age: 50 + 11 years

45% from ethnic minority groups

RCT in 27 clinical centres

Intensive lifestyle vs. metformin vs. placebo

FU mean 2.8 years, drop-out 7.5%

Weight reduction 7%

Diet: fat 25 E%

Physical activity (e.g. brisk walking) 150 min/week (700 kcal/week)

Main goal to achieve and maintain a weight reduction of > 7% through a healthy low-calorie, low-fat diet and to engage in physical activity of moderate intensity. Goal-based behavioural intervention lasting for 2.8 years (mean); case-managers (1 per 20–26 participants) held 16-session core curriculum in groups during the first 24 weeks; individual session monthly

T2D risk reduction 58% in lifestyle vs placebo after 2.8 years. 50% of IG met the weight reduction goal after 24 weeks. Mean weight loss 7 kg (7%) at 1 year.

Prevention of type 2 diabetes by lifestyle intervention: a Japanese trial in IGT males

Japan

Kosaka et al., 2005 [22]

30–60 year old men

IGT (according WHO 1980)

Random selection of men with IGT from health-screening program for mostly government employees.

458 randomised

n = 356 in CG, n = 102 in IG

sex: 100% men

age: mean NA

RCT intensive care vs. standard hospital care (1:4).

FU 4 years. Drop-outs 5.6% in CG, 4,7% in IG

BMI < 22; reduce energy intake by 10%; increase vegetable intake; fat < 50 g/day; alcohol < 50 g/day; eating out once /day or less; walking 30–40 min / day

Face-to-face counselling by nurse in hospital every 2–3 months, 4-year intervention. Regular weight self-monitoring. Concrete, standardised advices to reach the goals of the study.

Relative risk reduction 67% in 4 years. Body weight reduction 2.2 kg at 4 years vs. 0.39 kg in the control group

Indian Diabetes Prevention Programme, IDPP-1

India

Ramachandran et al., 2006 [23]

Middle-class population; 35–55 years;

IGT in two OGTT (WHO 1999)

Recruiting by workplace announcements and circulars. 10,839 subjects underwent initial screening using glucometer.

n = 531;

sex: 81% men

age: 45.9 years

RCT in community-based setting in 4 groups: Control; metformin (MET); lifestyle modification (LSM); LSM + MET.

FU 3 years (n = 502).

Goals: > 30 min brisk walking daily: reduction in total calories, refined carbohydrates and fat; avoidance of sugar and inclusion of fibre-rich foods

Participants had a personal session at 6-monthly intervals and were contacted by phone every month. Diet modification was advised for each subject.

Absolute risk reduction at year 3 was 15.7% in LSM, 14,5% in MET and 15.5% in LSM + MET compared to control (all p-values for HR < 0.03).

N/A

UK, Newcastle

Oldroyd et al., 2006 [24]

Oldroyd et al., 2001 [25]

European origin men and women aged 24–75 years

IGT in 2 OGTT

(WHO 1985)

Recruited people from previous research studies,

local hospital and GP

databases.

n = 78

sex: 57% men

age: 58.2 years

RCT in hospital setting; control vs intervention

FU 2 years (n = 54)

BMI < 25 kg/m², dietary fat < 30 E%, polyunsaturated to saturated fat (P:S) ratio ≥ 1.0; carbohydrate 50 E%, dietary fibre > 20 g per 4.2 MJ, 20–30 min aerobic exercise at least once a week

Individual counselling from a dietician and physiotherapist using the stages of change model. Intervention lasted 2 years and 12 sessions with duration of 15–20 min.

Meeting intervention goals was higher in IG for all other but P:S ratio. In IG weight change was − 1.8 kg vs + 1.

5 kg in CG at 24 months. Significant change in fasting serum insulin between groups at 12 months.

Hoorn Prevention Study

The Netherlands

Lakerveld et al., 2012 [26]

Lakerveld et al., 2013 [27]

Age 30–50

At least 10% risk for T2D and/or CVD estimated according formula of ARIC and SCORE projects.

A screening invitation to GP customers (n = 8193). 2401 respond, 921 eligible based on waist circumference. 772 were screened.

n = 622

sex: 42% men

age: 43.5 years

RCT in general practice.

FU 12 months (n = 502).

Goals were at least one fruit, at least 200 g vegetables and at least 30 min PA per day.

Participants were offered 6 face-to-face sessions (30 min) and 3 monthly telephone sessions with trained nurses. Methods were based on motivational interviewing, problem solving treatment, theory of planned behaviour and theory of self-regulation. CG got brochures of health guidelines.

No significant results in weight or fasting glucose or glucose tolerance. Increase in fruit intake between baseline and 6 months (1.1 - > 1.3 pieces per day) but not after 12 months. Median participation in sessions was 2.

(−)

The Joetsu Diabetes Prevention Trial

Japan

Kawahara et al., 2008 [28]

20–70 year old men and women

IGT (ADA 2003)

Recruiting from 11 outpatient practices and health evaluation and promotion centres in Joetsu area. Patients with BMI 20–34 kg/m² and FPG ≥ 5.6 mmol/l or HbA1C 5.2–6.4% were screened

n = 426

sex: 47% men

age: 51.4 years

Community-based clinical trial comparing short-term hospital (STH, n = 143) or outpatient diabetes education support (DES, n = 141) to no-treatment (n = 142).

Mean FU 3.1 years.

Goals for STH were 25–30 kcal / ideal body weight kg / day; 20–25% E% of fat; at least 30 min/day walking or exercise at least 5 times a week. Goals for DES were to follow diabetes guidebook.

STH group had 2-day hospital stay with a course of nine group/individual lessons covering diet, exercise and behaviour modification. Subsequent sessions every 3 months were offered individually. DES group got written information and 3-monthly individual sessions of a healthy lifestyle. Lessons were 20–40 min and were taught by different medical specialists. Mean intervention duration was 3 years.

The incidence of diabetes was 42 and 27% lower in STH and DES groups compared with no-treatment group, and 21% lower in STH than DES. Also FPG, 2 h OGTT plasma glucose, HbA1C and weight changes between groups were significantly different. STH) was more cost effective than DES

EDIPS-Newcastle

Penn et al., 2009 [29]

> 40 years, BMI > 25 kg/m²,

IGT in two OGTTs

Recruiting by invitation letter to eligible customers of primary care physicians. 1567 were contacted; 1084 replied; 682 agreed to testing; 482 completed at least one OGTT

n = 102

sex: 41% men

age: 57.1 years

RCT of two arms; intervention vs usual care.

Mean FU 3.1 years (n = 42), up to 5 years

> 50 E% carbohydrate; < 30 E% fat; reduce saturated fat intake; increase fiber intake; BMI < 25 kg/m2

Approx. 24 sessions with dietitian and physiotherapist as individual motivational interviewing for behavioural changes, including feedback from food diaries, weight and waist measurements. Cooking groups and discount of leisure service card was offered. Quarterly newsletter containing recipes, nutritional info and exercise suggestion was sent. Control group got usual care.

Weight reduction was 2.3 kg in IG vs. no change in CG, p = 0.007 at year 1 but no significant difference in consecutive years. The overall incidence of diabetes was non-significantly reduced by 55% in the intervention-group vs. the control group, with RR 0.45 (95% CI 0.2 to 1.2).

(+)

PREDIAS

Germany

Kulzer et al., 2009 [30]

20–70 years

BMI ≥ 26 IGT or IFG

Invitations by primary care physician based on FINDRISC > 10.

n = 182

sex: 57% men

age: 56.3

Two armed randomized control study in general practice.

FU 12 months (n = 165).

Goals based on DPP.

12 × 90-min group lessons in 12 months. Program was based on self-management theory and delivered by diabetes educator or psychologist. Participants got an exercise book, with diabetes prevention information and worksheets for lessons. Control group got written information about diabetes prevention.

There was significant difference in weight loss (−3.8 ± 5.2 vs. − 1.4 ± 4.09 kg), reduction in fasting glucose, increase in duration of PA and changes in eating behaviour after 12 months between intervention and control group.

Telephone support in addition to Greater Green Triangle Diabetes Prevention Program

Australia

Dunbar et al., 2010 [31]

40–75 years

FINDRISC > 12 points

Participants who completed the 12 month diabetes prevention program and were willing to participate in follow-up.

n = 205

sex: 28% men

age: 56.6 years (self-care); 57.1 (telephone support)

Telephone support vs self-care after 1 year lifestyle intervention pretest-posttest study for next 18 months

FU 30 months (n = 164)

< 30 E% fat; < 10 E% saturated fat; > 15 g fibre/1000 kcal; > 4 h/week moderate PA; > 5% weight reduction

Telephone support started after 12 month original intervention consisting of 6 group sessions. Telephone group got up to 12 calls following semi-structured interview with questions regarding personal goals.

There wasn’t significant difference between telephone support and self-care group. Original interventions improvements in 12 months were generally maintained to 30 months in both groups.

(−) for telephone support

DE-PLAN, Greece

Greece

Makrilakis et al., 2010 [32]

FINDRISC ≥15 points

3240 individuals were screened with FINDRISC-questionnaires; 620 were eligible.

n = 191

sex: 40%

age: 56.3 y

Pretest-posttest study in community setting in Greece.

FU 12 months (n = 125)

Reduce saturated and trans fat, sugars and sweets, refined cereals; ≥ 5 portions of fruits and vegetables per day; PA ≥ least 30 min 5 times a week. Weight reduction was recommended if overweight.

A dietitian held 6 group intervention sessions during 1 year. Each session was focused on one of goals and included information, discussion and written material. Sessions were held in workplaces or near participants’ residence. Evaluation of achieving goals was discussed in the beginning of each session.

Weight, BMI, blood pressure, and total cholesterol reduced significantly in those who completed program. Also reduction in whole fat dairies, processed meats, sugars and refined cereals was significant.

PREDIMED-Reus

Spain

Salas-Salvadó et al., 2011 [33]

Martínez-González et al., 2012 [34]

Men 55–80 years; Women 60–80 years; At least 3 risk factors for cardiovascular disease: smoking; hypertension; dyslipidemia; BMI > 25 kg/m²; family history of cardiovascular disease

1125 participants were screened in primary care centre and 870 fulfilled inclusion criteria. Of these 452 were diabetics, so 418 were the final population in this sub-study.

n = 418

sex: 42% men

age: 67 y

RCT for primary cardiovascular prevention in three arms: “low-fat” –control; Mediterranean Diet (MedDiet) + oil; MedDiet + nuts.

Median FU 4 years.

MedDiet; use olive oil abundantly; increase consumption of fruits, vegetables, legumes, nuts and tomato sauce for cooking; reduce total and red meat use and use fish and white meat instead; avoidance of butter, cream, fast food, sweets, pastries and sugar-sweetened beverages; moderate use of red wine. Control: reduce all type of fat, no free foods. All groups: No energy restriction, no PA promoted

Dietitians gave personalized dietary advice to participants on basis of a 14-item (MedDiet) or 9- item questionnaire (control). At inclusion and quarterly there after dietitians administered individual and group sessions. Participants were offered written material including descriptions of seasonal foods, shopping lists, weekly meal plans and cooking recipes. In addition participants in MedDiet groups were given free virgin olive oil (1 l/week) or nuts (30 g/day).

Hazard ratio for diabetes was 0.55 (0.32–0.95 95%CI) for both MedDiets compared with control diet in crude model and 0.48 (0.27–0.86) in multivariate adjusted model. Diabetes incidence was lower in participants who complied with the MedDiet better. Largest risk reduction was seen in MedDiet in subgroups of women vs men, over 67 year-old vs under and with those whose fasting glucose > 6.1 mmol/l in baseline than those who had ≤6.1 mmol/l.

DE-PLAN-Krakow

Poland

Gilis-Januszewska et al., 2011 [35]

FINDRISC points > 14;

OGTT to exclude diabetics

Recruiting from primary health care centres. 566 completed questionnaire; 368 eligible; 275 underwent OGTT

n = 186

sex: 22% men

age: 56.1 y

Pretest-posttest study in primary health care in Poland

FU 12 months (n = 175)

Goals were weight loss, reduced intake of total and saturated fats, increased consumption of fruits, vegetables and fibre and increased PA.

Active phase of intervention consisted of 10 group sessions in first 4 months followed by 6-month maintenance with six motivational phone calls and two motivational letters. PA sessions were offered once or twice a week. Social and family involvement was encouraged. Prevention managers were educated.

Significant changes in weight, BMI and waist circumference. No significant changes in fasting glucose or glucose tolerance test results.

Making the Connection Healthy Living Program, MTC HLP

USA

Ruggiero et al., 2011 [36]

underserved latino population; age 18–65 years;

BMI > 24.9, normal glucose or prediabetes; Latin background

Community-based health screening events (schools, family centers, hospital etc). 1162 screened, 367 tested for eligibility, 244 eligible.

n = 69

sex: 7% men

age: 38 y

Single-group, non-randomized follow-up, community-based translation of DPP.

FU 12 months (n = 57).

DPP goals: Weight reduction 7%; Diet: fat 25 E%

Physical activity (e.g. brisk walking) 150 min/week

Culturally specific intervention was developed and conducted in collaboration with the community to minimize barriers to participation education, literacy, language, income, transportation, lack of medical coverage. 22 group sessions during 1 year, delivered by trained community health workers. Cook books, pedometer, scales provided. Group walks arranged. Participants attended 57% of group sessions.

At 6 m, 20% achieved 7% weight reduction, and at 12 m 16% achieved. Moderate improvements in body weight (− 4.8 kg at 6 m, − 2.8 at 12 m), waist, fruit and vegetables, fat intake, PA were observed at 6 m but attenuated at 12 m. Forward movement in “stages of change” scale was observed.

The Prevention of Diabetes and Obesity in South Asians: PODOSA

Bhopal et al., 2014 [37]

Douglas et al., 2013 [38]

Douglas et al., 2011 [39]

South Asian families living in UK

> 35 years;

IGT and/or IFG

Recruiting via National health service, South Asian organizations and peer to peer. Pre-screening by waist circumference.

1319 screened with an OGTT, 196 (15%) were eligible.

n = 171 (156 families)

sex: 45.6% men

age = 52.5 y

Family-based two armed RCT in South Asians living in UK

FU 3 years

Goals were calorie deficit diet, at least 30 min brisk walking per day and at least 2.5 kg weight reduction.

Intervention visits were done in participants’ homes, 15 for intervention group and 4 for control group in 3 years. Practical dietary counselling: cooking, food shopping, food labelling and recipes. Counterweight program was used to dietary counselling. Change management tools, self-reflection and cultural adaptations were included.

Weight change in the IG was − 1.13 kg (SD 4.12) and in CG + 0.51 kg (SD 3.65), an adjusted mean difference of − 1.64 kg (95% CI − 2.83 to − 0.44).No difference in changes of fasting plasma glucose, OGTT or physical activity. Progression to diabetes was observed less frequently in the IG than the CG (OR 0·68) but not statistically significant (p = 0·3705).

(+)

Zensharen Study for Prevention of Lifestyle Diseases

Japan

Saito et al. 2011 [40]

30–60 years;

BMI > 24

IFG

First screening from health check-ups and eligible people were invited to OGTT (diabetics were excluded).

n = 641

sex: 78% men

age: median 50 y in IG, 48 y in CG

Unmasked, multicenter RCT in health care setting: frequent intervention group (FINT n = 311) vs. control (n = 330).

FU 36 months (n = 562).

Goals: 5% weight loss; reduce total energy intake; fat 20–25 E%; carbohydrate 55–60 E%; increase fibre intake and moderate alcohol intake. Increase of incidental PA to 200 kcal/d.

FINT got individual instructions and follow-up support for lifestyle modification from mainly dieticians and nurses at least 9 times. Self-monitoring using pedometers and body weight recording sheets.

Control group got similar individual instruction 4 times during first 12-months without follow-up support or self-monitoring tools.

The HR for T2D in FINT was 0.56 (95% CI 0,36-0,87) compared to control. In FINT IFG only group HR was 1.17 (0.50–2.74), and in FINT IFG + IGT 0.41 (0.24–0.69), compared to corresponding CGs. > 5% weight loss was significantly more achieved in FINT group during the first 12 months. Also PA goals and reducing energy intake goal were achieved more in FINT group.

APHRODITE

The Netherlands

Vermunt et al., 2011 [41]

Vermunt et al., 2012 [42]

Vermunt et al., 2012 [43]

FINDRISC ≥13; age 40–70 years

FINDRISC-questionnaire was sent to GP patients from 14 primary care practices (n = 16,032). Individuals with a score ≥ 13 were invited (n = 1533) for OGTT and diabetics were excluded.

n = 925

sex: 38% men

age: 58 y

RCT in Dutch primary care (IG n = 479, CG n = 446)

FU 2.5 years (IG n = 368, CG n = 341)

Weight reduction ≥5% if overweight, PA for at least 30 min a day / 5 days a week, fat intake < 30 E% and saturated fat < 10 E%, dietary fibre intake ≥3.4 g per MJ

Participants were offered 11 individual sessions with nurse or general practitioner, one with dietitian and 5 group sessions with dietitian and physiotherapist. Intervention lasted 30 months. Dietary advices were based on food diary. The intervention was based on trans-theoretical model (the stages of change).

Differences between groups were significant only for total physical activity and saturated fat and fibre intake. In the intervention group, self-efficacy was significantly higher in individuals successful at losing weight compared with unsuccessful individuals.

(+)

Lawrence Latino Diabetes Prevention Project (LLDPP)

US, Massachusets

Ockene et al., 2012 [44]

Low-income latinos

25–79 y old

latino / hispanic ethnicity; ≥ 25 y old; BMI > 24; > 30% increased according T2D risk algorithm

Recruiting (n = 949) from local health centre and local media.

n = 312

FU: 289 (CG 142, IG 147)

sex: 25.6% men

age: 52 y

Randomised community-based, culturally tailored, literacy sensitive lifestyle intervention (n = 162) vs. usual care (n = 150).

FU 1 year (n = 289)

Increase intake of whole grains and vegetables; reduce sodium, total and saturated fat, portion sizes and intake of refined carbohydrates and starches; increases walking by 4000 steps/day

3 individual sessions at home (1 × 1 h, 2 × 0.5 h) and 13 group sessions (1 × 1.5 h, 12 × 1 h) over 12-month period.

Participants got cash incentives at baseline, at 6- and 12-months. Participation was maximized with compensatory sessions and home visits. Practical, hands-on methods and demonstrations were used. Intervention was based on social cognitive theory and patient-centred counselling.

Participants lost more weight in IG (−2.5 lb) than in CG (+ 0.63 lb), effect of intervention − 2,5 lb. (p = 0.004). Also HbA1c reduced more in IG vs CG (effect of intervention − 0.10%, p = 0.009). Participants in IG reduced more energy intake from fat and saturated fat and increased dietary fibre intake.

E-LITE

Xiao et al., 2013 [45]

Ma et al., 2013 [46]

≥ 18 years

BMI ≥ 25; IFG or metabolic syndrome (2005 AHA)

Recruiting from a primary care clinic: 3439 contacted, 752 screened

BL n = 241 (81 cont; 79 coach-led; 81 self-directed)

sex: 53% men

age: 52.9

RCT in primary care with three arms: control, coach-led group intervention vs. self-directed DVD intervention.

FU 15 months (n = 221)

Weight loss goal based on DPP

Intensive intervention (3 months) included 12-session in groups or at home via DVD. Coach-led group had food tasting and guided PA and self-directed group got one face-to-face session, weight scale, pedometers and biweekly reminder e-mails during 12 month maintenance phase. All intervention participants were trained to use self-monitoring web portal.

At month 24 the mean change in BMI from baseline was −1.9 +/− 0.3 kg for coach-led group (p = 0,001 vs CG); − 1.6 +/− 0.3 kg for self-directed group (p = 0.003) and − 0.9 +/− 0.3 in the control group. Fasting plasma glucose was significantly more improved in IGs compared to CG.

RCT of SMS for Drivers With Pre-DM

China

Wong et al., 2013 [47]

Professional drivers

IFG/IGT; had a mobile phone

Screened 3376 drivers identified by community screening and media advertisement.

n = 104 subjects

age: 53 y

sex: 93.3% men

RCT of short message service (SMS) intervention on vs. control (leaflet)

FU 24 months (IG n = 41;CG n = 29).

“Diabetes-related information in reducing the risk of developing diabetes”

Participants got sms from 4 themes: diabetes information; lifestyle change; how others would appreciate the lifestyle modification; self-efficacy enhancing statements. In the first 3 months sms were sent 3 times a week, next 3 months weekly and last 18 months monthly. Both groups got leaflets about diabetes. Intervention was based on theory of planned behaviour and social cognitive theory.

No significant reduction in diabetes risk after 12 or 24 months. Significant mean differences in diastolic blood pressure and HDL-cholesterol over time between the groups.

Intervention cost was 5.05 $/ subject

(−)

DH!AAN

The Netherlands

Admiraal et al., 2013 [48]

Vlaar et al., 2012 [49]

South Asian migrants

Age 18–60 y; IFG, IGT, HbA1c > 6.0% or HOMA-IR > 2,39

2307 screened via general practices (invitation letter with reply card), followed by reminder and telephone call).

n = 536

age: 44.9 y

sex: 49.4% men

RCT in general practice among South Asian migrants in Netherlands getting a culturally targeted intervention or generic lifestyle advice (control).

FU 2 years (n = 335).

Goals according to SLIM study; based on current guidelines on diet and physical activity.

6–8 individual sessions in general practice during 6 months, 2 sessions during the next 6 months + 1 family session + two cooking classes. 20-week supervised exercise program was offered. Trained dieticians gave dietary counselling using motivational interviewing. Participants got a gift coupon for participating in baseline measurements. Control arm got 2 group sessions + 2 flyers.

No significant results. Median participation in 5 individual sessions. High drop-out and low participation 26% participated in family session, 26% in cooking sessions and 22% in PA sessions.

(−)

“Group Lifestyle Balance™ program in the community setting

USA

Kramer et al. 2014 [50]

26–80 years old

BMI ≥ 25 kg/m²; IFG / metabolic syndrome; ≥ 25 years old

Recruiting via GPs, information letters, e-mail contact and newspaper advertisement.

n = 81

sex: 22% men

age: 52.9 y

Group Lifestyle Balance (GLB) program in three outpatient-hospital in a pretest-posttest setting.

FU 12 months (n = 52)

7% weight loss; 150 min PA/week

12-session group lifestyle intervention adapted from DPP delivered by trained diabetes educators in groups over 12 to 14 weeks and monthly thereafter. Each session lasted 1 h. Handouts, self-monitoring booklets, fat- and calorie-tracking book and a pedometer were given to participants.

Significant changes: weight loss was 5.1% (p < 0,001); decrease in waist circumference; fasting plasma glucose; LDL-cholesterol; triglycerides and blood pressure.

Subanalysis of the Japan Diabetes Prevention Program

Japan

Sakane et al., 2014 [51]

30–60 years

IGT

People measured in yearly health check-ups were recruited using posters, fliers and by word of mouth. 1279 were screened with OGTT.

n = 304

sex: 50,0% men

age: 59% men

RCT in 32 primary health care centres using existing resources in two arms.

mean FU 2.3 years (n = 213)

5% reduction in body weight in overweight and obese; increase energy expenditure by 700 kcal per week; < 25 E% fat; < 160 kcal/day from alcohol

Study nurse held four 2–3-h group sessions during first 6 months followed by biannual 20–40 min individual sessions, intervention was 3 years. After first year individual sessions were held on phone. Personalised goals were set. A booklet was given and monthly tip cartoons were sent via fax. Self-efficacy, self-monitoring and trans-theoretical model was used.

No significant results in T2D risk. In IG participants with BL HbA1c levels ≥5.7% (n = 177) cumulative incidence was significantly lower. Significant results: at 1 year IG had improved body weight and daily non-exercise leisure time energy expenditure and at 3 years better Matsutada index.

(+)

Use of Information Technology in the Prevention of Diabetes

India

Nanditha et al., 2014 [52]

Ramachandran et al., 2013 [53]

Men

No major illness; age 35–55 years; positive family history of T2D; BMI > 23; IGT

First screening with questionnaire (n = 8741) and then OGTT first with a glucometer and confirmatory venous blood glucose within 1 week.

n = 537

sex: 100% men

age: 46 y

RCT in industrial male workers lifestyle modification with SMS.

FU 2 years (n = 517)

Avoidance of simple sugars and refined carbohydrates; total fat intake < 20 g/day; restrict use of saturated fat; increase fibre; enhance aerobic exercise > 30 min brisk walk/day; walk 3–4 km in 30 min at least 5 days a week

All participants got personalised lifestyle modification in the beginning. SMS-group received mobile phone messages at frequent intervals (2–4 sms/week). SMS content was tailored according to participant’s stage of change in trans-theoretical model.

Risk reduction in sms-group compared to control was 9% (HR = 0.64 CI 0.446–0.917 p = 0.015).

“Lifestyle modifications in Chinese women who had gestational diabetes mellitus”

China

Shek et al., 2014 [54]

Age > 18 years; GDM history; IGT 6–8 weeks after delivery; excluded if insulin for GDM

Patients from hospitals were invited if criteria were fulfilled

n = 450;

sex: 100% women

age: 39 y

RCT conducted in hospital in Hong Kong. Intensive lifestyle intervention vs. no intervention (control).

FU 36 months (n = 423).

Optimal caloric intake (based on Harris-Benedict) for ideal body weight

7 individual sessions in 3 years (3 m, 6 m 12 m, 18 m, 24 m 30 m, 36 m). Dietician and study nurse gave individual dietary and exercise advice based on food and exercise records (n = 7).

No significant difference in cumulative incidence of diabetes. In women > 40 years, difference between groups was significant. Significant differences between groups at 1 year in BMI and waist-hip ratios, but not significant in the end of the study.

(+)

“Prevention of diabetes in Finnish airline”

Finland

Viitasalo et al., 2012 [55]

Viitasalo et al., 2015 [56]

Airline workers (majority shift-work)

FINDRISC > = 10 or IFG or IGT

Occupational health care check-up n = 2312, 657 had high risk and were offered intervention.

n = 350

sex: 60% men

age: 47 y

Work-site study targeted at identified high-risk workers of an airline.

Average FU 2.5 years (n = 402).

DPS goals and other goals according to risk factor levels (BP, cholesterol).

1–3 individual lifestyle counselling sessions in addition to the check-up by nurse/physician. Lifestyle sessions were delivered by diabetes nurse or nutritionist.

Among elevated risk men, body weight was slightly reduced and 14.3% lost > 5% of weight, and cholesterol and LDL decreased. Those men who attended more lifestyle sessions lost more weight. Fasting glucose increased in all groups. FINDRISC score increased, but less so among high-risk men.

(+)

Clinical significance estimate* the scoring is marked as follows: ++ significant reduction in DM risk; + significant improvement in (most) target risk factors; (+) significant improvement in some/few risk factors; (−) no effect

Abbreviations: AHA American heart association, BL baseline, BMI body mass index, CG control group, CVD cardio vascular disease, DPP Diabetes Prevention program, E% percentage energy from, f-Glu fasting plasma glucose, FU follow-up, GDM gestational diabetes mellitus, GP general practice, HDL high density lipoprotein, HR Hazard ratio, IFG Impaired fasting glucose, IG intervention group, IGT impaired glucose tolerance, LDL low density lipoprotein, OGTT oral glucose tolerance test, PA physical activity, RCT randomised controlled trial, SMS short message service, T2D Type 2 diabetes, WHO World Health Organization;