Preventing Diabetes in the Clinical Setting


  • Deborah L. Burnet MD, MA,

    1. Section of General Internal Medicine, Department of Medicine. Diabetes Research and Training Center, The University of Chicago, Chicago, IL, USA
    Search for more papers by this author
  • Lorrie D. Elliott MD,

    1. Section of General Internal Medicine, Department of Medicine. Diabetes Research and Training Center, The University of Chicago, Chicago, IL, USA
    Search for more papers by this author
  • Michael T. Quinn PhD,

    1. Section of Endocrinology, Department of Medicine, Diabetes Research and Training Center, The University of Chicago, Chicago, IL, USA.
    Search for more papers by this author
  • Andrea J. Plaut MPH,

    1. Section of General Internal Medicine, Department of Medicine. Diabetes Research and Training Center, The University of Chicago, Chicago, IL, USA
    Search for more papers by this author
  • Mindy A. Schwartz MD,

    1. Section of General Internal Medicine, Department of Medicine. Diabetes Research and Training Center, The University of Chicago, Chicago, IL, USA
    Search for more papers by this author
  • Marshall H. Chin MD, MPH

    1. Section of General Internal Medicine, Department of Medicine. Diabetes Research and Training Center, The University of Chicago, Chicago, IL, USA
    Search for more papers by this author

  • Ms. Plaut is now a student at the University of Illinois at Chicago School of Medicine, Chicago, IL, USA.

  • Dr. Elliott is now on faculty at Northwestern University Feinberg School of Medicine, Chicago, IL, USA.

Address correspondence and requests for reprints to Dr. Burnet: University of Chicago, 5841 South Maryland Avenue, MC 2007, Chicago, IL 60637 (e-mail:


Objective: Translating lessons from clinical trials on the prevention or delay of type 2 diabetes to populations in nonstudy settings remains a challenge. The purpose of this paper is to review, from the perspective of practicing clinicians, available evidence on lifestyle interventions or medication to prevent or delay the onset of type 2 diabetes.

Design: A MEDLINE search identified 4 major diabetes prevention trials using lifestyle changes and 3 using prophylactic medications. We reviewed the study design, key components, and outcomes for each study, focusing on aspects of the interventions potentially adaptable to clinical settings.

Results: The lifestyle intervention studies set modest goals for weight loss and physical activity. Individualized counseling helped participants work toward their own goals; behavioral contracting and self-monitoring were key features, and family and social context were emphasized. Study staff made vigorous follow-up efforts for subjects having less success. Actual weight loss by participants was modest; yet, the reduction in diabetes incidence was quite significant. Prophylactic medication also reduced diabetes risk; however, lifestyle changes were more effective and are recommended as first-line strategy. Cost-effectiveness analyses have shown both lifestyle and medication interventions to be beneficial, especially as they might be implemented in practice.

Conclusion: Strong evidence exists for the prevention or delay of type 2 diabetes through lifestyle changes. Components of these programs may be adaptable for use in clinical settings. This evidence supports broader implementation and increased reimbursement for provider services related to nutrition and physical activity to forestall morbidity from type 2 diabetes.

Type 2 diabetes affects approximately 8% of U.S. adults.1 The prevalence is rising among adults and youth,2,3 paralleling the dramatic increase in obesity.4 Increased incidence of diabetes, especially among youth, portends a serious increase in early morbidity, health care costs, and lost productivity. Diabetes prevention has become a key target for clinicians, patients, and policymakers, as substantial evidence has accumulated that diabetes can be prevented or delayed in those at high risk. Presenting the results of the Diabetes Prevention Program (DPP) trial, Secretary of Health and Human Services Tommy Thompson declared, “In view of the rapidly rising rates of obesity and diabetes in America, this good news couldn't come at a better time …. By promoting healthy lifestyles, we can improve the quality of life for all Americans, and reduce health care costs dramatically.”5 A working group from the American Diabetes Association and the National Institute of Diabetes and Digestive and Kidney Diseases published a cogent position statement regarding the scientific findings and health policy implications from diabetes prevention trials.6 The current article provides practicing clinicians with a more detailed review of evidence regarding prevention of type 2 diabetes, insights into the components of successful interventions, and consideration as to which aspects of interventions are most adaptable for use in clinical practice.


A MEDLINE literature search from 1980 to 2004 was performed to identify articles about prevention or delay of type 2 diabetes in adults. Key phrases included diabetes prevention, type 2 diabetes, lifestyle intervention, pharmacologic prevention, nutrition and exercise, and combinations thereof. References of relevant articles were searched as well. The inclusion criteria were clinical trials including an active intervention with longitudinal follow-up to decrease the onset of type 2 diabetes. Six reviewers agreed upon inclusion of the studies identified. Three prevention studies utilizing prophylactic medication and 4 utilizing lifestyle changes to prevent diabetes were identified. As lifestyle interventions proved more efficacious and have been recommended for first line use,7 the main focus of this review is the 4 major studies that describe successful lifestyle interventions to prevent or delay the onset of diabetes.

Study Interventions and Outcomes

The design and outcomes of the 4 major lifestyle intervention studies are summarized in Table 1. Studies ranged in size from 415 subjects (Malmö, Sweden, 1980s)8 to 3,234 (DPP, U.S., 1996 to 2001)9; the length of follow-up ranged from approximately 3 to 6 years. All the studies included a lifestyle intervention encouraging participants to improve nutrition, lose weight (for overweight subjects), and increase physical activity. Actual weight lost in these studies was modest, with about half the weight on average regained over the course of the studies. Nonetheless, significant decreases in diabetes incidence were demonstrated in the lifestyle intervention groups. The Finnish Diabetes Prevention Study10–12 and the U.S. DPP9 each demonstrated a relative risk reduction of 58% through lifestyle change compared with placebo.

Table 1. Lifestyle Trials to Prevent or Delay Diabetes
 Diabetes Prevention
Program (DPP)8,29
Diabetes Prevention
Study (DPS)10,30
Da-Qing IGT and
Diabetes Study32
Malmö Feasibility
  • *

    One study-specific physical activity unit=30 minutes of mild intensity (e.g., slow walking), or 20 minutes of moderate intensity (e.g., brisk walking), or 10 minutes of strenuous intensity (e.g., slow running), or 5 minutes of very strenuous intensity (e.g., jumping rope) exercise.

  • IGT, impaired glucose tolerance; FPG, fasting plasma glucose; DM, type 2 diabetes mellitus; BMI, body mass index [weight in kilograms/(height in meters)2]; RCT, randomized-controlled trial; kcal, indicates kilocalories.

 Years1996 to 20011993 to 20001986 to 19921974 to 1985
 Inclusion criteriaIGT and ↑FPGIGTIGTMild DM (no symptoms), IGT, normal controls
 Age (y, mean±SD)50.6±10.755±745.0±9.1Range=47 to 49
 BMI, (kg/m2 mean±SD)34.0±6.731.3±4.6 (intervention group)25.8±3.827.7±3.7 (group 1; DM, lifestyle)
 31.0±4.5 (control group) 26.6±3.1 (group 2; IGT, lifestyle)
   26.7±4.0 (group 3; IGT controls)
   24.3±2.8 (group 4; normal controls)
 Follow-up (mean y)2.83.265
 Race55% Caucasian100% Caucasian100% Asian100% Caucasian
20% African American   
16% Hispanic   
5% American Indian   
4% Asian American   
Study design
 TypeRCT; individuals randomizedRCT; individuals randomizedRCT; clinics randomizedNonrandomized feasibility study Baseline differences in groups
 Number of sites275331
 Arms4 arms:2 arms:4 arms:Lifestyle intervention
 Lifestyle intervention,  n=1079
 Metformin, n=1073
 Troglitazone  (discontinued 1998)
 Control, n=1082
 Lifestyle intervention, n=265
 Control group, n=257
 Diet alone, n=130
 Exercise alone, n=141
 Diet±exercise, n=126
 Control, n=133
 Group 1 (DM), n=41
 Group 2 (IGT), n=181
No intervention
 Group 3 (IGT), n=79
 Group 4 (normals), n=114
 Weight loss7% weight loss≥5% weight lossFor BMI <25: noneNot mentioned
  For BMI ≥25: 0.5 to 1.0 kg loss/mo until BMI=23 
 Diet<25% kcal from fat<30% kcal from fat
<10% saturated fat
≥15 g fiber/1000 kcal
 25 to 30 kcal/kg intake
 55% to 65% carbohydrates
 10% to 15% protein
 25% to 30% fat
 ↓kcal intake
↓Simple carbohydrates
↑Complex carbohydrates
↓Saturated fats
Substitute polyunsaturated fats
↓Kilocalories for obese subjects
 Physical activity150 min physical activity per week30 min moderate intensity physical activity per day↑Leisure physical activity by 1 to 2 study-specific units per day*Not mentioned
Intermediate outcomes
 Weight change
  kg (mean)Lifestyle   ↓5.6Lifestyle ↓4.2 at 1 yDid not develop DM  Did developLifestyle (groups 1 and 2)
Metformin  ↓2.1     ↓3.5 at 2 yControl:   ↑0.27  ↓1.55↓6 at 1 y
Placebo   ↓0.1Control ↓0.8 at 1 y
     ↓0.8 at 2 y
DM diet:   ↑0.93  ↓2.43↓2.0 to 3.3 at 5 y
  Exercise:   ↑0.71  ↓1.93Control (groups 3 and 4)
  Diet+exercise:↓1.77  ↓3.33↓0.2 to 2.0 at 5 y
  % subjects meeting weight loss goal50% in lifestyle armBy year 1: 43% lifestyle group  
 13% control group  
  % subjects maintaining weight loss goal38% in lifestyle arm  82% group 1 & 71% group 2 maintained overall weight reduction over 5 y
Physical activity
  % subjects meeting activity goal74% in lifestyle armBy year 1: 86% lifestyle group  
  71% control group  
  % subjects maintaining activity goal58% in lifestyle arm   
Diabetes outcomes
 IncidenceCumulative 3 y DM incidence:
 Control  28.9%
 Lifestyle  14.4%
 Metformin  21.7%
Cumulative 2 y DM incidence:
 Control  14%
 Lifestyle  6%
Cumulative 4 y DM incidence:
 Control  23%
 Lifestyle  11%
Cumulative 6 y DM incidence:
 Control  67.7%
 Diet  43.8%
 Exercise  41.1%
 Diet±exercise  46.0%
Cumulative 6 y DM incidence:
 Lifestyle (group 2)  10.6%
 Control (group 3)  28.6%
 Control (group 4)  0%
 Risk reduction in intervention vs control groupDM risk reduction over 3 y:
 Lifestyle  58%
 Metformin  31%
DM risk reduction over 6 y:
 Lifestyle  58%
DM risk reduction over 6 y:
 Diet  31%
 Exercise  46%
 Diet±exercise  42%
DM risk reduction over 6 y:
 Lifestyle  63%
 (group 2 vs group 3)

Table 2 describes the design and outcomes of diabetes prevention trials using medication. The DPP8 demonstrated a 31% reduction in diabetes risk in subjects receiving metformin, compared with placebo. The Study to Prevent NIDDM (STOP-NIDDM)13,14 was an international study of 1,429 overweight adults with impaired glucose tolerance (IGT), who were followed for an average of 3.3 years. Subjects were randomized to receive acarbose or placebo. Compared with placebo, subjects receiving acarbose were 25% less likely to develop diabetes. The Troglitazone in Prevention of Diabetes Study,15 randomized 266 Hispanics with gestational diabetes to 400 mg of troglitazone daily or placebo. After a median follow-up of 30 months, the annual diabetes incidence was 12.1% with placebo and 5.4% in the drug arm, a risk reduction of over 50%. The DPP troglitazone study arm was terminated when a patient on troglitazone died from liver failure. In the DPP, the relative advantage of lifestyle intervention over metformin was greater in older subjects, those with lower baseline body mass index, and those with lower baseline fasting glucose.9 Prophylactic medication clearly reduces diabetes risk; however, lifestyle changes are more effective overall and are recommended as first-line strategy.7

Table 2. Drug Trials to Prevent or Delay Diabetes
 Diabetes Prevention Program (DPP)8Study to Prevent Noninsulin-Dependent
Diabetes Mellitus (STOP-NIDDM)12,13
Troglitazone in Prevention of
Diabetes (TRIPOD)14
  1. IGT, impaired glucose tolerance; FPG, fasting plasma glucose; DM, type 2 diabetes mellitus; OGTT, indicates oral glucose tolerance test; BMI, body mass index [weight in kilograms/(height in meters)2]; RCT, randomized-controlled trial.

 CountryUSACanada, Germany, Austria, Norway, Denmark, Sweden, Finland, Israel, SpainUSA (Los Angeles county)
 Years1996 to 20011995 to 20011995 to 2000
 Inclusion criteriaIGT and↑FPGIGT and↑FPGHispanic women with history of gestational DM
  High risk by 5 h OGTT
 Age (y, mean±SD)50.6±10.754.3±7.9 (intervention)34.9±6.6 (intervention)
 54.6±7.9 (control)34.3±6.5 (placebo)
 BMI (kg/m2, mean±SD)34.0±6.731.0±4.3 (intervention)30.6±6.1 (intervention)
 30.9±4.2 (control)30.3±5.3 (control)
Follow-up (mean y)
DrugBiguanide antihyperglycemic (metformin) Thiazolidinedione (troglitazone—stopped early secondary to liver failure)α-glucosidase inhibitor (acarbose)Thiazolidinedione (troglitazone). Study terminated early secondary to liver failure
Study design
 TypeRCT; individuals randomizedInternational, multicenter double-blind RCTDouble-blind RCT
 Number of sites279 countries1
 Arms4 arms:2 arms:2 arms:
 Lifestyle intervention, n=1079
 Metformin (850 mg twice daily), n=1073
 Troglitazone (400 mg/d, discontinued 998)
 Placebo, n=1082
 α-glucosidase inhibitor titrated to  100 mg 3 times daily or maximum tolerated dose, n=714
 Placebo, n=715
 Thiazolidinedione 400 mg/d, n=133
 Placebo, n=133
DietStandard lifestyle recommendations for med arms; written information on diet.Instruction in weight-reducing dietDietary advice at annual visits
Annual counseling on healthy lifestyle.Yearly visits with dietitian 
ExerciseEncouraged to↑physical activity.Encouraged to exercise regularlyAdvised to walk 30 min, 3 d/wk
Adherence to medication77% in placebo30% of treatment group discontinued early, most because of GI side effects11% (30 women) lost to follow-up (11 placebo, 19 drug)
72% in metformin  
Diabetes outcomes
 IncidenceCumulative incidence DM over 3 y:Cumulative incidence DM at 3.3 y:Average annual DM incidence:
 Placebo  28.9% Acarbose  32.4% Placebo  12.1%
 Metformin  21.7% Placebo   41.5% Troglitazone  5.4%
 Lifestyle  14.4% Annual incidence rates posttrial:
   Placebo   21.2%
   Troglitazone  3.1%
 Risk reduction in intervention vs control groupDM risk reduction over 3.3 y:DM risk reduction over 3 y: 
 Lifestyle  58% Acarbose  25%Hazard ratio=.45
 Metformin  31%  


Who Was Targeted in the Diabetes Prevention Trials?

Lifestyle interventions are most effective in patients at high risk for disease.16 Accordingly, all 4 diabetes prevention lifestyle studies enrolled subjects with IGT as evidenced by oral glucose tolerance testing (OGTT). Impaired glucose tolerance is defined as a 2-hour postprandial glucose level between 140 and 199 mg/dL on standard OGTT. Persons with IGT are known to be at high risk for progression to diabetes.17

How Should We Identify Patients in Clinical Practice?

Patients at risk for diabetes are asymptomatic; reliable methods are needed to identify those at high risk. Hemoglobin A1C is not recommended for screening or diagnosis18 because of nonstandardized methods of testing. Impaired glucose regulation can be identified by documenting impaired fasting glucose (IFG) or IGT, although some patients exhibit 1 abnormality without the other. By definition, IGT requires glucose tolerance testing for identification. Impaired glucose tolerance is more strongly associated with cardiovascular risk than IFG19,20; however, both are markers for microvascular risk.21 Although controversial, glucose tolerance testing is not generally recommended for screening in clinical practice22,23 as it is costly, inconvenient, and less reproducible than fasting plasma glucose (FPG). The American Diabetes Association (ADA) Expert Committee18 recommended decreasing the lower limit for IFG from 110 to 100 mg/dL to optimize sensitivity for predicting future diabetes. This change also increases the proportion of persons with IGT who can be identified by the fasting blood test, making this a rational screening strategy.

Clinical characteristics also predict risk of diabetes.24 The clinical characteristics associated with type 2 diabetes risk include obesity and overweight, age (risk rises steadily from puberty into geriatric years), a history of gestational diabetes, polycystic ovary syndrome, a family history of type 2 diabetes, and membership in certain high-risk minority groups: African American, Hispanic, Native American, and Asian-Pacific Islanders.25–27 The U.S. Preventive Services Task Force (USPFTF) finds “insufficient evidence” to recommend screening all asymptomatic adults for diabetes because “It has not been demonstrated that … screening provides an incremental benefit compared with initiating treatment after clinical diagnosis.” They do, however, recommend screening persons at high risk, including those with hypertension or hyperlipidemia (“B” recommendation: good evidence).28 The ADA recommends screening youth and adults with multiple risk factors for type 2 diabetes; FPG is the preferred first-line test.3,29 Emerging evidence suggests that youth-onset type 2 diabetes is an aggressive disease associated with increased risk of morbidity.30 In sum, patients with multiple risk factors are logical targets for diabetes prevention efforts, especially if risk is confirmed through finding of IFG.

What Preventive Strategies Should We Use to Decrease Diabetes Risk?

In the DPP, lifestyle changes were more effective than medication, and lifestyle changes do not involve exposure to medications and risk of side effects. This approach is embodied at the level of the general population in the Surgeon General's Call to Action31 and the 2005 USDA Dietary Guidelines for Americans.32 The studies reviewed here support directing intensive lifestyle intervention efforts toward those at highest risk based on clinical characteristics, IFG, and/or IGT.

What are the Components of Successful Diabetes Prevention Strategies?

These studies set goals for modest weight loss for overweight participants, and for increased physical activity of moderate intensity.The recommended dietary content was similar throughout these studies, comprising less than 25% to 30% of caloric intake from fat. The DPS additionally encouraged high fiber intake; the Da Qing study specified 55% to 60% of caloric intake from carbohydrates and 10% to 15% from protein.

The DPP and DPS set weight loss goals of 7% and 5% of body weight, respectively.33,34 The Malmö35 and Da Qing36 studies called for decreased caloric intake with a gradual weight loss in overweight subjects. The DPP and DPS set physical activity goals of 150 minutes/week, or 30 minutes/day, 5 days/week, of moderate intensity physical activity (DPP recommended brisk walking). These recommendations are consistent with the 2005 USDA Dietary Guidelines for Americans,32 which promote “at least 30 minutes of moderate-intensity physical activity … most days of the week.” Da Qing recommended increased leisure physical activity defined in study-specific units, and Malmö provided 2-hour-long sessions per week of various physical activities.

Evidence from relevant epidemiologic studies provides an additional insight into dietary factors likely to lower diabetes risk. Several cohort studies showed that diabetes risk was highest for sedentary individuals who ate a “western diet” (red meat, processed meats, French fries, high fat dairy products, refined grains, sweets, and desserts).37,38 Conversely, a “prudent diet” emphasizing fruits, vegetables, legumes, fish, and whole grains was associated with a lower risk. A number of prospective studies found diets high in whole grains or cereal fiber to be associated with a reduced risk for type 2 diabetes.39–42

While the “glycemic index (GI)” has gained popular attention recently, consistent evidence to support its use is lacking. The GI and its derivative, the glycemic load (GL), have been proposed as physiologic ways to categorize carbohydrates; foods are classified by how rapidly they are digested and absorbed compared with a standard food (commonly, glucose or white bread). A high GI characterizes readily digestible starch, refined grain products, and potatoes, while foods with a low/moderate GI include legumes, unprocessed grains, and nonstarchy fruits and vegetables. A few studies have shown increased diabetes risk in patients ingesting high GL; however, this finding has not been universal, and the clinical utility of the GI in diabetes prevention remains unproven.43–46

All 4 studies featured intensive interaction with staff and individualized counseling, with group sessions on a voluntary basis. Table 3 lists the key components of the lifestyle interventions, including staff qualifications and training. Each DPP participant in the lifestyle arm was assigned a master's level case manager who provided individual counseling sessions based on behavior change theory. Malmö participants could choose individual or group sessions; most opted for individual counseling.

Table 3. Key Components of Lifestyle Interventions
 Diabetes Prevention
Program (DPP)29
Diabetes Prevention Study
Da-Qing IGT and Diabetes
Malmö Feasibility Study31
Staff training
 StaffMDs, nurses, techniciansNot mentionedMDs, nurses, techniciansStaff included dietitian, nurse, physiotherapist, and MD
 TrainingBehavior change training 2 d/yNot mentionedBehavior change training 2 d/yNot mentioned
 Format16 core curriculum sessions on nutrition, physical activity, and behavioral self-managementFood records used as basis for tailored dietary advice 4 times/y in individualized sessionsIndividual counseling on diet and exercise by physicians at 3-mo intervalsSubjects in the intervention group could choose small group or individual counseling
Individual and group elements7 visits with nutritionist in year 1Small group counseling sessions weekly for 1 mo, monthly for 3 moSubjects received dietary information at monthly group meetings for 6 mo
Individualized plansIndividual guidance on↑physical activity to improve cardiovascular fitness 60-min activity sessions 2 times/wk (e.g., calisthenics, walking-jogging, soccer, badminton) under the guidance of a physiotherapist
Optional physical activity sessions led 2 times/wk by DPP staffVoluntary group walking, biking, and supervised resistance training Smokers were advised to stop or reduce smoking
Brisk walking recommended   
 Follow-upFollow-up sessions every 2 mo with phone calls between visits1 session every 3 mo1 group session every 3 mo for remainder of study 
 If weight goal not achieved in 6 to 12 mo, a very low calorie diet (VLCD) was considered  
Social supportSpouses invited to join sessionsSpouses invited to join sessions, especially if responsible for shopping/cooking Spouses invited to monthly group meetings

The Finnish DPS drew upon Prochaska's Trans-theoretical (Stages of Change) Model.47–49 The DPP invoked similar principles, and also sought to address the cultural background of the participants. Staff used 5 different ethnic versions of the DPP curriculum and helped participants individualize goals within their particular cultural context. Table 4 lists the topics covered in the DPP's 16 individual sessions. Complete curricular contents are available at Voluntary group sessions augmented individual counseling in the DPP and DPS, including lectures, cooking lessons, supermarket visits, and exercise sessions.

Table 4. Curriculum for the Diabetes Prevention Program
Session 1AWelcome to the Lifestyle Balance ProgramReasons for joining DPP, benefits, goalsBuild commitment, heighten awareness of risk, and increase awareness of benefits. Begin to set personal goals
Session 1BGetting Started Being ActiveParticipants choose intervention goal to begin with: increasing physical activity or losing weightIncrease commitment and ownership by encouraging patient to choose own goals
Session 1BGetting Started Losing Weight  
Session 2Move Those MusclesPersonal experience, preferences, self-monitoringBuild awareness of habits and preferences by self-monitoring of activity. Increase self-efficacy by reviewing past successes
Session 3Being Active: A Way of LifeFinding time for physical activity; safetyBegin to schedule physical activity to fit it into patient's lifestyle
Session 4Be a Fat DetectiveSources of fat, self-monitoring, goal settingLearn to identify fat sources. Begin to set personal fat goals
Session 5Three Ways to Eat Less FatMeasuring portionsLearn to weigh and measure foods and estimate appropriate portion size
Session 6Healthy EatingMeal planningLearn the importance of planning for timing and content of meals and helpful eating behaviors (e.g., eating slowly)
Session 7Take Charge of What's Around YouCues at home; stimulus control; choicesLearn cues in environment that prompt unhealthy food and activity choices; learn to alter cues
Session 8Tip the Calorie BalanceWhat it takes to lose 1 to 2 pounds/wkLearn energy balance and what it takes to lose 1 to 2 pounds a week
Session 9Problem SolvingIdentify problems, brainstorm solutions, plan steps, evaluate outcomesLearn 5-step problem-solving approach: describe problem, brainstorm solutions, pick solution, create action plan, and evaluate success
Session 10Four Keys to Healthy Eating OutPlanning, assertion, stimulus control, choicesDevelop healthy dining out approach: anticipate and plan, assertion, stimulus control, and healthy food choice
Session 11Talk Back to Negative ThoughtsSubstituting positive thoughtsIdentify common pattern of negative thoughts and practice countering them with positive statements
Session 12The Slippery Slope of Lifestyle ChangeTriggers for slip-ups; strategies for recoveryRecognize that slips are normal; identify personal triggers for slips, reactions, and strategies for recovery
Session 13Jump Start Your Activity PlanHeart rate, fitness, variety of physical activityIntroduce aerobic fitness: measure heart rate and perceived exertion, add variety to fitness plan
Session 14Make Social Cues Work for YouDealing with social pressureManaging problematic social cues; increasing helpful social cues
Session 15You Can Manage StressAssertion, social supports, problem solvingAcquire stress management techniques: assertion, social support, problem solving, planning, countering negative thoughts
Session 16Ways to Stay MotivatedOngoing goals and support strategiesAcquire relapse prevention skills/maintain motivation: review personal reasons for joining, personal successes, setting new goals, seeking social supports

Each of these studies emphasized behavioral contracting around self-derived goals. While investigators set diet and weight goals for the studies overall, participants used individualized counseling sessions to set their own goals. Cognizant of various stages of change, study staff helped individual participants tailor and modify goals progressively to achieve success. Participants documented their goals in concrete terms reinforced by behavioral contracting.

Patient empowerment and self-efficacy were further enhanced through promotion of self-monitoring through use of scales and measuring cups; subjects recorded their own diet and physical activity levels and maintained charts documenting their progress.

These studies acknowledged the importance of family and social context in targeting diabetes prevention efforts. Spouses of study subjects were encouraged to participate in the individualized counseling sessions in the DPP, DPS, and Malmö studies. Fisher et al.50 identified the family as the primary social context for the recognition, understanding, and management of diabetes and other chronic health conditions. Family is a key source of social supports and stresses, and interventions that target 1 member necessarily affect others. Therefore, these studies sought to engage family members whenever possible to optimize outcomes.

These studies incorporated vigorous follow-up efforts, especially for subjects having less success. Subjects failing to meet initial goals were actively encouraged by staff. The DPP used computer monitoring to track program adherence and trigger actions for “recovery” of participants failing to reach goals. DPP staff used a stepped strategy to optimize outcomes, with a “toolbox” of problem-solving strategies valued at $100 annually per participant, including exercise tapes and classes, personal trainers, cookbooks, and other resources.

Provider profiling was utilized as a means for quality assurance at the systemic level. Each of the 27 DPP sites received monthly feedback on their performance in attaining weight and activity goals relative to other sites.

How Can a Busy Clinician Incorporate Practical Strategies to Promote Lifestyle Change into an Office Visit?

These trials utilized behavior change strategies in the context of ongoing relationships with trial staff. Similarly, brief behavior-change counseling strategies can be effectively incorporated into patient encounters in the office setting.51,52 Research on lifestyle change has shown that individuals progress through 5 sequential stages in making changes, and that different strategies are useful at various stages.47–49,53–55 Counseling messages individualized to the patient's readiness to change are more successful, while mismatched messages often lead to patient and clinician frustration. By recognizing that many patients are in the early stages of change, clinicians can modify their expectations and redefine success as helping patients move along the continuum of change rather than as reaching a desired final outcome.

Behavioral counseling to decrease diabetes risk can be guided by the Five A's model, which was adapted by the USPFTF56 from the National Cancer Institute's model for physician counseling of smokers,57 and has been studied in a variety of brief primary care interventions.58–60 The Five A's involve assessing the patient's lifestyle risk factors and readiness to change, advising specific behavioral change, agreeing on behavior change goals, assisting the patient in acquiring information, skills, and confidence required to progress toward goals, and arranging follow-up. Asking patients nonjudgmentally about current diet and exercise behaviors, the physician can readily assess current practices, knowledge of risks, and readiness to change. Subsequent behavior change advice should be clear, strong, and personalized: “As your doctor, I think it's important for you to change your diet and increase your physical activity so you can reduce your risk for developing diabetes in the next few years.”

Within the Five A's model, brief counseling approaches can be guided by the principles of motivational interviewing,61 a patient-centered approach that elicits behavior change by helping patients address their ambivalence regarding recommended change. Brief versions of motivational interviewing developed for primary care settings emphasize building rapport, assessing patients' beliefs about the importance of behavior change, and their self-efficacy for change.62 Counseling strategies for patients who do not believe that health behavior change is important include providing information, giving feedback, and exploring the patient's ambivalence. Counseling strategies for patients with low self-efficacy include reviewing prior successful change attempts, focusing on manageable steps, and enhancing problem-solving skills.

The importance of arranging follow-up underscores the ongoing nature of behavior change and the role of the clinician-patient relationship in supporting changes. Follow-up on progress within regularly scheduled clinic appointments is essential, but may not be sufficient. Additional follow-up with a dietitian, nurse, or behavioral expert may help some patients make and sustain meaningful health behavior changes. Examples of brief counseling interventions matched to stages of change are given in Table 5.

Table 5. Brief Counseling Messages Tailored to Stages of Change
 Precontemplation (No
Intention of Making
(Considering Making
Change, But
Preparation (Intends to
Take Action Within Next
Action (Has Changed
Behavior Less Than 6 mo)
Maintenance (Has
Changed Behavior More
Than 6 mo)
AgreePatient-Centered GoalsEvaluate Pros and Cons of Behavior ChangePersonal CommitmentStimulus ControlReinforce Self-Efficacy
“Would you be willing to think about the benefits of weight loss and exercise, and we can talk more the next time you're in?”“While you see some obstacles to exercise and weight loss, you also see some benefits. What are some of those benefits for you?”“Have you thought about setting a date to start changing your diet and exercise?”“Do you think it would help if you replaced some of the cookies and ice cream in your house with healthier snacks?”“It sounds like you got a little off track over the holidays. How confident are you that you can get your exercise and diet back on track?”
AssistEncourage Increased AwarenessSelf-ReevaluationSelf-MonitoringFeedback, Self-RewardPlan for Relapse
“Would you be interested in learning more about some of the personal health benefits of just modest exercise and weight loss?”“Can you picture yourself as a more active, healthier person? What would that be like for you?”“It might be helpful to keep track of what you're eating now, so when you start to change your diet, you'll know what to change.”“Do you think a pedometer might give you some helpful feedback on how far you're walking each day?”“Can you anticipate any obstacles or situations that would keep you from exercising and eating the way you have been?”
ArrangeValidate Lack of ReadinessBrainstorm Obstacles and SolutionsEstablish Social ResourcesBolster Self-EfficacyPlan for Follow-up Social Support
“At your next appointment, after you've given it some thought, we can talk more about whether you think exercise and weight loss are the right thing for you now.”“At your next appointment, let's talk more about some of those barriers to your exercise and weight loss.”“If you'd like, I could refer you to our dietitian. You could meet with her regularly to get a better understanding of your diet, what you might change, and what you don't need to change.”“At your next appointment, bring in your food records so we can see how much you've reduced your calorie intake.”“At your next appointment, could you bring in your wife? I'd like to share with her all the progress you've been making in reducing your diabetes risk.”

What Specific Recommendations Should Practitioners Make Regarding Physical Activity?

To decrease the risk of developing diabetes, patients should engage in moderate-intensity physical activity most days of the week. The type of exercise must be tailored to the patient's ability and preferences. To increase sustainability, the patient should enjoy the activity and be willing to make it a priority. For most patients, brisk walking is an appropriate start; those with arthritis may prefer water-based exercise or nonweight-bearing activities like bicycling. In addition, patients should be counseled to increase physical activity in daily routines, such as taking the stairs or parking farther away from buildings. Patients should accumulate at least 150 minutes/week of physical activity. Exercise should be of moderate intensity; patients may feel slightly out of breath and feel their heart beating more quickly, but they should not feel exhausted or unable to sustain the activity. Further counseling suggestions are given in Table 6.

Table 6. Counseling on Physical Activity and Nutrition
Physical activity
 Goal of 150 min of moderate-intensity exercise weekly
 Tailor physical activity to individual's ability and interest
  Walking for most; bicycling or water-based for those with arthritis
 Encourage increased activity in daily routines
  Take the stairs; park farther away; get off bus 1 stop early
 Previously inactive individuals should begin with short amounts of moderate-intensity exercise (for example, 10 min) and gradually increase the duration and/or intensity
 Goal-set with individual on preferred way to accrue 150 min weekly
  For example, 30 min of walking 5 d weekly or 50 min of walking 3 d weekly
  Make goals specific in time, amount, and activity
 Encourage self-monitoring of activity by keeping written records, using a pedometer, or using a heart rate monitor
 Emphasize that total calories matter
 Goal of fat intake less than 25% of total calories; minimize intake of saturated fats and trans fats (red meat, deep fried foods, oils solid at room temperature)
 Encourage portion size awareness and reading food labels
 Increase dietary fiber to 20 to 30 g/d
 Diet should be high in whole grains, fruits and vegetables, beans, and nuts
 Goal-set with individual on preferred initial changes to diet
  For example, piece of fruit at lunch each day, or red meat no more than once a week
 Make goals specific in time, amount, and type
 Encourage self-monitoring by keeping food logs
 Encourage self-reward for meeting goals
 Enlist family members to help with goals if acceptable to patient
 Help patient to anticipate potential barriers to exercise and solutions to those barriers
 Let patient know that relapse is the norm; rather than being discouraged, encourage them to think about what led to the relapse and how to overcome that in the next try
 Arrange close follow-up

Should Persons with IFG Who Plan to Begin Exercising Undergo Exercise Stress Testing?

The American College of Cardiology/American Heart Association recommend testing asymptomatic individuals with multiple cardiac risk factors, or men over age 45 and women over age 55 who plan to begin a vigorous exercise program. However, this recommendation is rated level IIB (conflicting evidence).63

What Specific Recommendations can Practitioners Make Regarding Dietary Change and Composition?

The overall goal for diabetes prevention is to reach and maintain an active, healthy weight with a tendency toward a hypocaloric diet. As summarized in Table 6, evidence supports limiting total calories and fat (<25% of caloric intake) and increasing dietary fiber (20 to 30 g/day). Essential skills include understanding portion sizes and reading food labels. Involvement of a dietitian is optimal to assess dietary history, navigate challenges inherent in change, and prevent relapse.

Are Diabetes Prevention Efforts Cost-Effective?

The DPP Research Group conducted detailed cost-effective analyses from both the health system and societal perspectives.64,65 From a societal perspective, lifestyle and drug interventions cost $24,400 and $34,500, respectively, for each case of diabetes prevented or delayed within the 3-year time horizon of the study. These costs are well within the generally accepted range for preventive strategies, and would be relatively lower if benefits were to persist beyond the study period. Sensitivity analysis estimating societal costs for lifestyle and drug interventions as they might be implemented in clinical practice projected $13,200 and $14,300, respectively, per case prevented. Lifetime cost-utility analysis65 projected costs per quality-adjusted life year of $1,100 and $8,800 for the lifestyle intervention from the health care and societal perspectives, respectively. Cost-effectiveness simulations for diabetes prevention are limited in that they are based on experience with research subjects, which may not generalize to the broader population, and such models are based on assumptions regarding long-term health outcomes. We do not yet have direct evidence from studies with long-term follow-up as to whether diabetes prevention efforts represent a cost-effective way to prevent or delay the clinically important complications of diabetes.6

How Can Health Care Systems Help Prevent Diabetes?

Although we have focused on practitioners, health systems may have an important role in diabetes prevention. We lack data on system-based approaches to diabetes prevention, but a variety of such techniques improve outcomes for patients already diagnosed with diabetes. Examples include computerized reminders and provider feedback,66 multidisciplinary teams providing patient education and follow-up,66 self-management education in community settings,67 disease management (organized, multicomponent approach to diabetes care), and case managers coordinating care.67 Some of these approaches may be adaptable for diabetes prevention. Multidisciplinary care teams consisting of nurses, dietitians, and health educators may provide more intensive counseling and increase the contact that a patient has with the health care system. Printed materials or interactive computer programs in offices can reinforce counseling efforts. Telephone support can be brief and effective. Group classes may help selected patients. Public health interventions are also needed to create safe environments for exercise and promote healthy lifestyles in schools and workplaces. Future studies judging the effectiveness of such interventions for diabetes prevention should focus on patient outcomes as well as process measures.68


Clinical trials have shown conclusively that diabetes can be prevented by lifestyle modification, at costs generally considered acceptable to society. Evidence from these trials suggests that clinicians should recommend behavior changes for asymptomatic patients at high risk for diabetes. High-risk patients can be identified through clinical characteristics augmented with judicious screening by fasting glucose. Although the diabetes prevention trials used intensive strategies for effecting lifestyle change, clinicians can translate key elements from those strategies into brief, office-based counseling on physical activity and dietary change.

Implementing diabetes prevention will require significant paradigm shifts for both patients and clinicians. Modest goals for weight loss and physical activity are appropriate; behavioral contracting and self-monitoring may enhance self-efficacy and outcomes for patients. We must educate clinicians in training and in practice about the potential benefits of diabetes prevention and strengthen training for behavioral change within medical education.

Diabetes prevention efforts need to be tailored for particular participants and settings. Despite implementation across very different cultures, however, these lifestyle prevention studies demonstrated remarkably consistent outcomes. Cultural adaptations for office-based counseling may be challenging in diverse communities; enlisting community resources may enhance these efforts.

Relationships and social context are key factors for diabetes prevention. In these trials, close coaching relationships with study staff facilitated lifestyle change by participants. Successful diabetes prevention efforts will likely require enlisting important family members, enhancing clinician-patient relationships, practice innovations facilitating feedback to clinicians and patient follow-up, and broader societal changes supporting healthy lifestyles in the context of schools, communities, and workplaces.

The rigorous cost-effectiveness analyses of the DPP provide a compelling case for increased insurance coverage of nutrition and physical activity interventions in persons at high risk for diabetes. Even in an era when patients switch insurance carriers every few years, savings may accrue rapidly through prevention or delay of diabetes. Less costly group intervention in clinical settings bears further investigation, and studies of the effects on complication rates are needed.

These diabetes prevention trials have shown dramatically how diabetes can be prevented or delayed through lifestyle changes. Many aspects of these prevention programs appear adaptable for use in clinical settings at present. Successfully implementing diabetes prevention on a large scale will require improved clinician-patient communications as well as innovative systems of care, making further translational research a priority.


This work was supported by the National Institute of Diabetes and Digestive and Kidney Diseases Diabetes Research and Training Center (P60 DK20595). Dr. Burnet is supported by a Mentored, Patient Oriented, Career Development Award (K23 DK064073-01), and Dr. Chin was a Robert Wood Johnson Foundation Generalist Physician Faculty Scholar.