Abstract
Aims:
To describe National Diabetes Prevention Program (NDPP) uptake, retention, and outcomes by enrollee characteristics and program type.
Methods:
We studied 776 adult University of Michigan employees, dependents, and retirees with prediabetes and overweight or obesity who enrolled in one of four CDC-recognized NDPPs at no out-of-pocket cost. Programs included 1) an in-person classroom-based program led by certified diabetes educators in an endocrinology outpatient clinic; 2) an in-person classroom-based program led by trained peer instructors in community settings; 3) an in-person fitness-focused program led by trained lifestyle coaches; and 4) an online digital program led by personal health coaches with virtual group meetings. Data from the insurer and surveys were analyzed.
Results:
Older individuals with hypertension and cardiovascular disease were more likely to enroll in classroom-based programs. Program time, location, and perceived focus on diet or physical activity influenced program selection. Retention, weight loss, and physical activity were greater among enrollees in in-person classroom-based programs. Changes in blood pressure, lipid levels, self-rated health, and health-related quality-of-life did not differ by program, nor did Type 2 diabetes mellitus incidence.
Conclusions:
Individuals with prediabetes who enrolled in a NDPP achieved health benefits regardless of the type of program they chose.
Keywords: Prospective observational study, Prediabetes, Type 2 diabetes mellitus, Prevention
1. Introduction
In 2002, the Diabetes Prevention Program clinical trial demonstrated that an intensive behavioral lifestyle intervention that employed a low-fat, calorie-restricted diet and moderate physical activity reduced the incidence of Type 2 diabetes mellitus by 58% over 2.8 years in adults ≥ 25 years of age with BMI ≥ 24 kg/m2, elevated fasting glucose, and impaired glucose tolerance [1]. Based upon these findings, the Centers for Disease Control and Prevention (CDC) launched the National Diabetes Prevention Program (NDPP) to provide high-quality lifestyle change programs to people with prediabetes to reduce their risk of Type 2 diabetes mellitus and improve their overall health [2]. Between 2012 and 2020, over 2,000 CDC-recognized programs enrolled over 450,000 adults with prediabetes in the 1-year NDPP [3]. Between 2012 and 2016, 36% of enrollees achieved 5% weight loss and 42% reported at least 150 min of physical activity per week [4].
CDC-recognized lifestyle change programs differ in their format (in-person vs. online), leaders (medical personnel vs. trained peer leaders), and locations (hospital vs. community-settings). To date, CDC has not systematically described the characteristics of individuals with prediabetes who elected to participate in the various types of programs, assessed NDPP retention by program type, or reported outcomes according to participant-level characteristics and the types of programs in which the participants enrolled.
The purpose of this study was to describe and compare the baseline characteristics of individuals with prediabetes who enrolled in one of four different CDC-recognized NDPPs with distinct formats, leaders, and locations and to describe and compare program retention and 1- and 2-year outcomes according to the type of NDPP selected. The results are important to understand participant preferences, retention, and outcomes to better connect participants to programs and optimize their outcomes.
2. Subjects, materials and methods
2.1. Study setting
The University of Michigan is a large, public, research university in Ann Arbor, Michigan, with satellite campuses in Flint and Dearborn, Michigan. Beginning in 2015, the university offered the NDPP at no out-of-pocket cost to university employees, dependents, and retirees ≥ 18 years of age who had prediabetes and overweight or obesity and who were enrolled in the university’s self-funded health insurance program. The university selected four programs fully-recognized by CDC to provide the NDPP: 1) a health system-based program which provided in-person classes led by certified diabetes educators in an endocrinology outpatient clinic in Ann Arbor (health system); 2) a community-based program offered by a not-for-profit voluntary health organization which provided in-person classes led by trained peer instructors in community settings across southeastern Michigan (including Dearborn and Flint) (community); 3) a program offered by a not-for-profit charitable organization which provided in-person classes led by trained lifestyle coaches in a fitness facility in Ann Arbor (fitness-focused); and 4) an online digital program which provided flexibly-timed online classes and virtual group meetings led by personal health coaches (online). In some settings, the community-based program offered access to wellness centers. The fitness-focused program offered enrollees family memberships to the facility which included free fitness classes and free childcare. The online program provided wearable activity trackers and digital scales.
2.2. Study population
We identified 64,131 university employees, dependents, and retirees ≥ 18 years of age without evidence of Type 1 or Type 2 diabetes mellitus and identified 8,131 of them (12.7%) as having prediabetes and overweight or obesity based on health plan records [5]. Prediabetes was defined by either a health plan claims diagnosis of prediabetes or a hemoglobin A1c (HbA1c) level of 5.7 to 6.4% (39 to 46 mmol/mol). Overweight or obesity were defined by BMI ≥ 25.0 kg/m2 (≥23.0 kg/m2 for Asians). Individuals were excluded if they had histories of Type 1 or Type 2 diabetes mellitus. Individuals with prediabetes were identified and encouraged to enroll in a NDPP as previously described [5,6]. This study focused on the 776 people with prediabetes and overweight or obesity who enrolled in a NDPP between August 2015 and July 2019 and attended at least one session.
2.3. Data collection
We used data from the health plan, including data collected from NDPP providers, and data from surveys of enrollees to describe the characteristics of adults with prediabetes who enrolled in a NDPP [5,6]. Demographic (age, sex, race, and zip code) and clinical variables (body mass index (BMI), blood pressure, lipids, HbA1c, and medical history) were obtained from the health plan. Baseline values were defined as the last values recorded before the date of enrollment in a NDPP. One-year and 2-year follow-up values of clinical variables were defined as the last reported values in the time interval between the date of NDPP enrollment plus 365 days and between the date of enrollment plus 366 to 730 days, respectively. We defined incident Type 2 diabetes mellitus on the basis of HbA1c and self-report. The 1-year incidence of Type 2 diabetes mellitus based on HbA1c was calculated as the number of enrollees with HbA1c values ≥ 6.5% (48 mmol/mol) measured during year 1 divided by the number of enrollees who had at least one HbA1c value measured during year 1 multiplied by 100. The 2-year incidence of Type 2 diabetes mellitus was calculated as the number of enrollees with HbA1c values ≥ 6.5% (48 mmol/mol) during year 1 or year 2 divided by the number of enrollees who had at least one HbA1c value measured during year 1 or 2 multiplied by 100.
We classified residential address zip codes as being within or outside the city of Ann Arbor and merged zip codes with data from the U.S. Census Bureau and the National Neighborhood Data Archive (NaNDA) to describe neighborhood indicators of median household income, percent unemployment, percent of families receiving Supplemental Nutrition Assistance Program (SNAP) benefits, and to describe neighborhood access to public transportation, internet, and social services [7–10] . NaNDA is available through the U-M Institute for Social Research Survey Research Center at nanda.isr.umich.edu.
Information provided by the NDPPs to the health plan was used to calculate the percentages of enrollees who completed specific numbers of sessions, the mean number of sessions attended, weights, percent weight loss, and physical activity minutes. Three of the four programs assessed reported weight as measured with a calibrated scale at the in-person visit and reported by the enrollee at each session and physical activity minutes by self-report. The online program used digital scales that directly uploaded the weight each week. The online program used wearable fitness trackers to measure physical activity and estimated that 250 steps were equivalent to one minute of vigorous physical activity.
We also surveyed NDPP enrollees at baseline, 1 year, and 2 years [6,11]. We mailed surveys to all 776 enrollees and received 532 completed surveys (69% crude response rate). When we excluded those who had died, opted out of the research, or could not be contacted because of incorrect addresses, the adjusted response rate was 74%. As previously reported, there were only minor differences between survey respondents and non-respondents at baseline [6]. The surveys assessed clinical and social variables, reasons for choosing specific NDPP programs, and self-reported progression to Type 2 diabetes mellitus at 1 and 2 years follow-up [6]. Progression to Type 2 diabetes mellitus was ascertained with the survey question “Have you been diagnosed with diabetes?”. Clinical variables included self-rated health, EQ-5D-5L-derived health utility scores, and the EQ-5D visual analog scale [12]. Social variables included full or part-time employment and affiliation with the university as an employee, retiree, or dependent. Reasons for choosing a specific program included convenience (class time and location), groups of people attending the program (family, coworkers, or friends), and perceived focus of the program on diet or physical activity.
2.4. Data analysis
We compared the baseline characteristics of adults with prediabetes and overweight or obesity who enrolled in each of the four NDPPs using t-tests for continuous variables and chi-square tests for categorical variables. Changes in risk factors (BMI, blood pressure, lipids, and HbA1c) were calculated by subtracting baseline values provided by the health plan from 1-year and 2-year values. Changes in weight were calculated by subtracting the first values recorded by the NDPP program from the last values recorded at or before the 6-month and 1-year timepoints. Weekly minutes of physical activity was assessed as the mean of up to the last three non-missing non-zero values of self-reported or fitness tracker assessed minutes of physical activity at 6 month and 12 month visits. The percent of people with at least 150 min of physical activity per week at six and twelve months was calculated as the proportion with mean minutes of physical activity ≥ 150 per week.
We assessed program retention among enrollees as the percentage of enrollees who attended the program at a given week divided by the number of enrollees who attended during the first week of the program multiplied by 100. A participant was considered to be retained through a given week if he or she attended a session that week or during a subsequent week [3]. We also assessed the number of core sessions and maintenance sessions attended. Core sessions were defined as any sessions numbered 1 through 16 that were attended during the first 168 days after enrollment. Using this variable, we calculated the mean and median number of core sessions attended and the number and percent of enrollees who attended at least 75% (12 of 16) of the core sessions. Maintenance sessions, offered at least monthly, were defined as any sessions numbered 17+ that were attended between 112 days and 425 days of enrollment. Using this variable, we calculated the mean and median number of maintenance sessions attended and the number and percent of enrollees who attended at least 75% (6 of 8) of the maintenance sessions. Since the online virtual sessions were “unlocked” weekly and could be accessed at any time after they were unlocked, we calculated attendance of maintenance sessions as completing at least one session numbered 17+ in each 28-day period between 112 days and 425 days.
All analyses were performed using SAS 9.4 (SAS Institute, Cary, NC).
2.5. Human subjects approval
The study was reviewed and approved by the University of Michigan Institutional Review Board for Human Research (HUM#00108065) and was granted a waiver of documented informed consent. Study participants were mailed informed consent documents with the surveys and the return of a completed survey was considered to indicate consent.
3. Results
There were 776 individuals with prediabetes and overweight or obesity who enrolled in the NDPP (Table 1). NDPP enrollees were 53 ± 10 years of age (mean ± SD); 72% were women, and 82% were white. Most (86%) had made at least one primary care provider visit in the past year and 70% had visited a specialist. Mean BMI was 34.1 ± 7.0 kg/m2. Blood pressure and lipids were generally well-controlled. Mean HbA1c was 5.9 ± 0.3% (41 ± 0 mmol/mol). Medical claims diagnoses for overweight/obesity, hypertension, and dyslipidemia were common. More than one-third of participants (36%) were receiving antihypertensive medications and approximately one-quarter (24%) were receiving lipid lowering medications. Only 2% of participants had health plan claims diagnoses for smoking and 8% had claims diagnoses for cardiovascular disease.
Table 1.
Baseline characteristics of NDPP enrollees with prediabetes and overweight or obesity, stratified by CDC-recognized NDPP provider.
| Enrolled in NDPP | Health system program | Community program | Fitness-focused program | Online program | p-value | |
|---|---|---|---|---|---|---|
| N (%) | 776 | 277 (36%) | 45 (6%) | 206 (27%) | 248 (32%) | – |
| Age (years) | 53 ± 10 | 55 ± 9 | 57 ± 8 | 51 ± 11 | 51 ± 11 | <0.0001 |
| Sex | 0.1545 | |||||
| Women | 557 (72%) | 212 (77%) | 33 (73%) | 142 (69%) | 170 (69%) | |
| Men | 219 (28%) | 65 (23%) | 12 (27%) | 64 (31%) | 78 (31%) | |
| Race | 0.3975 | |||||
| Asian | 54 (8%) | 24 (10%) | 1 (3%) | 11 (6%) | 18 (9%) | |
| Black | 49 (8%) | 13 (5%) | 3 (8%) | 18 (10%) | 15 (7%) | |
| White | 538 (82%) | 200 (83%) | 34 (89%) | 140 (81%) | 164 (81%) | |
| Other | 12 (2%) | 3 (1%) | 0 (0%) | 3 (2%) | 6 (3%) | |
| At least 1 PCP visit in past year | 663 (86%) | 249 (90%) | 43 (98%) | 171 (84%) | 200 (82%) | 0.0041 |
| At least 1 specialist visit in past year | 539 (70%) | 206 (74%) | 33 (75%) | 144 (71%) | 156 (64%) | 0.0516 |
| Body Mass Index (kg/m2) | 34.1 ± 7.0 | 33.8 ± 6.3 | 33.6 ± 7.5 | 34.3 ± 7.1 | 34.3 ± 7.5 | 0.7589 |
| Blood pressure (mmHg) | ||||||
| Systolic | 126 ± 15 | 128 ± 15 | 128 ± 17 | 123 ± 14 | 126 ± 14 | 0.0154 |
| Diastolic | 74 ± 10 | 74 ± 10 | 73 ± 10 | 74 ± 9 | 75 ± 10 | 0.5546 |
| Lipids (mg/dl) | ||||||
| Total cholesterol | 196 ± 39 | 199 ± 37 | 185 ± 35 | 195 ± 40 | 196 ± 40 | 0.1772 |
| HDL-cholesterol | 53 ± 14 | 54 ± 14 | 53 ± 14 | 50 ± 12 | 53 ± 15 | 0.0697 |
| Triglycerides | 147 ± 96 | 152 ± 102 | 134 ± 80 | 143 ± 77 | 147 ± 105 | 0.6762 |
| LDL-cholesterol | 113 ± 32 | 114 ± 31 | 105 ± 26 | 115 ± 34 | 111 ± 32 | 0.2709 |
| HbA1c (%) | 5.9 ± 0.3 (41 ± 0 mmol/mol) | 5.9 ± 0.3 (41 ± 0 mmol/mol) | 5.8 ± 0.3 (40 ± 0 mmol/mol) | 5.8 ± 0.4 (40 ± 0 mmol/mol) | 5.9 ± 0.3 (41 ± 0 mmol/mol) | 0.1701 |
| Health plan claims diagnoses of … | ||||||
| Overweight/obesity | 355 (46%) | 126 (45%) | 17 (38%) | 102 (50%) | 110 (44%) | 0.4697 |
| Hypertension | 285 (37%) | 119 (43%) | 20 (44%) | 60 (29%) | 86 (35%) | 0.0100 |
| Any antihypertensive medication | 283 (36%) | 113 (40%) | 22 (49%) | 60 (29%) | 88 (35%) | 0.0175 |
| Dyslipidemia | 282 (36%) | 113 (41%) | 16 (36%) | 62 (30%) | 91 (37%) | 0.1180 |
| Any lipid lowering medication | 188 (24%) | 77 (28%) | 12 (27%) | 39 (19%) | 60 (24%) | 0.1566 |
| Smoking | 17 (2%) | 6 (2%) | 1 (2%) | 2 (1%) | 8 (3%) | 0.4451 |
| Cardiovascular disease | 61 (8%) | 29 (10%) | 6 (13%) | 13 (6%) | 13 (5%) | 0.0577 |
| Reside within city of Ann Arbor | 272 (35%) | 94 (34%) | 9 (20%) | 114 (55%) | 55 (22%) | <0.0001 |
| Neighborhood indicatorsa | ||||||
| Median income ($) | 71,287 | 71,784 | 65,204 | 71,546 | 71,603 | 0.1601 |
| Percent unemployment | 34.4 ± 4.6 | 34.6 ± 4.6 | 35.1 ± 4.9 | 33.4 ± 4.1 | 34.7 ± 4.7 | 0.0089 |
| Percent receiving SNAP benefits | 8.2 ± 6.0 | 7.7 ± 4.8 | 11.4 ± 10.1 | 7.7 ± 5.1 | 8.6 ± 6.7 | 0.0006 |
| Transit stops per square mile | 4.6 ± 9.0 | 3.9 ± 8.3 | 3.5 ± 8.6 | 6.7 ± 9.2 | 3.8 ± 9.4 | 0.0015 |
| Childcare services in zip code area | 24.0 ± 14.8 | 22.7 ± 13.6 | 21.0 ± 14.2 | 30.4 ± 15.2 | 20.5 ± 14.5 | <0.0001 |
| Child/youth services in zip code area | 1.3 ± 1.7 | 1.2 ± 1.7 | 1.0 ± 1.5 | 1.9 ± 1.9 | 1.1 ± 1.6 | <0.0001 |
| Elderly/disable services in zip code area | 3.9 ± 2.6 | 3.7 ± 2.5 | 4.1 ± 3.0 | 4.5 ± 2.5 | 3.6 ± 2.6 | 0.0009 |
| Percent of households with internet subscription | 89 ± 0.1 | 89 ± 0.1 | 86 ± 0.1 | 90 ± 0.0 | 88 ± 0.1 | 0.0002 |
| Percent of households with any computing device with internet | 88 ± 0.1 | 89 ± 0.1 | 85 ± 0.1 | 89 ± 0.0 | 87 ± 0.1 | <0.0001 |
| N (%) with baseline survey data | 532 | 201 (38%) | 37 (7%) | 130 (24%) | 164 (31%) | – |
| Employed by UM | 334 (87%) | 140 (88%) | 19 (86%) | 72 (85%) | 103 (87%) | 0.9361 |
| Reasons for choosing program | ||||||
| Convenient time | 400 (77%) | 155 (79%) | 28 (78%) | 75 (58%) | 142 (89%) | <0.0001 |
| Convenient location | 398 (76%) | 167 (85%) | 32 (89%) | 88 (68%) | 111 (69%) | 0.0002 |
| People attending | 44 (8%) | 15 (8%) | 3 (8%) | 18 (14%) | 8 (5%) | 0.0524 |
| Perceived focus on diet | 238 (46%) | 107 (54%) | 16 (44%) | 47 (36%) | 68 (43%) | 0.0115 |
| Perceived focus on physical activity | 235 (45%) | 81 (41%) | 11 (31%) | 86 (67%) | 57 (36%) | <0.0001 |
| Self-rated health | 0.2949 | |||||
| Excellent/Very good | 260 (49%) | 93 (47%) | 23 (62%) | 60 (47%) | 84 (51%) | |
| Good/Fair/Poor | 270 (51%) | 107 (54%) | 14 (38%) | 69 (53%) | 80 (49%) | |
| EQ-5D-5L health utility score | 0.86 ± 0.14 | 0.87 ± 0.13 | 0.89 ± 0.13 | 0.87 ± 0.13 | 0.85 ± 0.17 | 0.3708 |
| EQ-5D visual analog scale score | 76 ± 14 | 75 ± 14 | 76 ± 16 | 76 ± 14 | 77 ± 14 | 0.7496 |
NDPP: National Diabetes Prevention Program, HbA1c: Hemoglobin A1c, HDL: High-density lipoprotein cholesterol, LDL: Low-density lipoprotein cholesterol, PCP: Primary Care Provider, SNAP: Supplemental Nutrition Assistance Program.
Data are expressed as N (%) or mean ± standard deviation. P-values indicate significance between NDPP program types.
Neighborhood indicators determined by merging residential zip codes with data reported from U.S. Census Bureau and the National Neighborhood Data Archive.
Of the 776 NDPP enrollees, 277 (36%) enrolled in the health system program, 45 (6%) in the community program, 206 (27%) in fitness-focused program, and 248 (32%) in the online program (Table 1). In general, enrollees in the in-person classroom-based health system and community programs were older than enrollees in the fitness-focused and online programs (Table 1). There were no differences in the sex or race distributions of enrollees in the individual programs although men tended to be more likely to enroll in the fitness-focused and online programs and whites tended to be more likely to enroll in the classroom-based health system and community programs. Mean BMI levels did not differ by program. Systolic blood pressure was lowest in fitness-focused enrollees. This may in part reflect the younger ages of those enrollees. Diastolic blood pressure, lipid, and HbA1c levels did not differ across programs. Enrollees in the classroom-based health system and community programs were more likely to have made at least one primary care provider visit in the past year and to have claims diagnoses for hypertension and cardiovascular disease, perhaps related to their older ages. Individuals who lived in Ann Arbor were more likely to enroll in the health system and fitness-focused programs which were offered in Ann Arbor. Enrollees in the community program were more likely to live in neighborhoods with lower median neighborhood incomes, higher percentage unemployment, and higher percentage of individuals receiving federal Supplemental Assistance Program (SNAP) benefits. Those who enrolled in the community program were also the least likely to live in households with internet subscriptions and computing devices. Individuals who enrolled in the fitness-focused program were the most likely to have available public transit and social services, perhaps related to the fact that those enrollees were more likely to live in the city of Ann Arbor which provides many social services.
We used surveys to further assess the characteristics of enrollees and their program preferences (Table 1). Eighty-seven percent of NDPP enrollees were university employees: only 13% were dependents or retirees. University of Michigan employment status was not associated with program selection. Enrollees were significantly more likely to choose the health system and community programs based on “convenient location” and the online program based on “convenient time”. Enrollees were most likely to choose the fitness-focused program because of “people attending” and its perceived “focus on activity” and the health system program because of its perceived “focus on diet”. Program enrollees did not differ by self-reported health or by health-related quality-of-life as assessed by the EQ-5D-5L and its visual analog scale (Table 1).
In general, attendance during the core phase of the NDPP was good: 73% of enrollees attending at least 75% of core sessions in the first 6 months (Table 2). Attendance for the core sessions was greatest for the health system program (81% attending ≥ 75% of core sessions) followed by the community program (78%) and the online program (77%). Attendance of the core sessions was lowest for the fitness-focused program (58%). Forty-seven percent of enrollees attended at least 75% of maintenance sessions in the last 8 months. Attendance of at least 75% of maintenance sessions was greatest for the community program (69%), intermediate for the fitness-focused (48%) and health system program (46%) and lowest for the online program (43%). This might be due in part to differences in the way attendance was calculated for the online program. Overall, the median numbers of core and maintenance sessions attended were 14 and 5, respectively. The median number of core sessions attended was greatest for the online enrollees (16), followed by the community (15), health system (14), and fitness-focused programs (13). Attendance of maintenance sessions was greatest for community program enrollees (7), followed by health system (5) and fitness-focused (5), and online program enrollees (4).
Table 2.
Participation in NDPP stratified by provider.
| Enrolled in NDPP | Health system program | Community program | Fitness-focused program | Online program | p-value | |
|---|---|---|---|---|---|---|
| N (%) | 776 | 277 (36%) | 45 (6%) | 206 (27%) | 248 (32%) | – |
| Core sessions (sessions 1–16) | ||||||
| Mean number of core sessions attended | 13 ± 4 | 13 ± 3 | 13 ± 4 | 12 ± 4 | 14 ± 4 | <0.0001 |
| Median number of core sessions attended | 14 (11–16) | 14 (12–15) | 15 (13–16) | 13 (9–15) | 16 (13–16) | – |
| Attended 75% of core sessions in first 6 months | 570 (73%) | 225 (81%) | 35 (78%) | 119 (58%) | 191 (77%) | <0.0001 |
| Maintenance sessions (sessions 17 +) | ||||||
| Mean number of maintenance sessions attended | 5 ± 3 | 5 ± 2 | 7 ± 2 | 5 ± 2 | 4 ± 3 | <0.0001 |
| Median number of maintenance sessions attended | 5 (3–8) | 5 (3–7) | 7 (5–9) | 5 (3–7) | 4 (1–8) | – |
| Attended 75% of maintenance sessions in last 8 months | 277 (47%) | 102 (46%) | 24 (69%) | 78 (48%) | 73 (43%) | 0.0515 |
NDPP: National Diabetes Prevention Program.
Data are expressed as N (%), mean ± standard deviation, or median (interquartile range). P-values indicate significance between NDPP program types.
As previously described, overall retention was good among NDPP enrollees (median number of sessions attended 18, median retention 38 weeks) with 77% retained through 18 weeks and 40% retained through 44 weeks [11]. Fig. 1 shows enrollee retention by program and week. Retention in the community program at 44 weeks was better, but not significantly better, than the health system program (OR 1.61 (95% CI 0.95, 2.74)). Retention in the fitness-focused program was somewhat worse (OR 0.79 (95% CI 0.62, 1.01)) and retention in the online program was significantly worse than in the health system program (OR 0.61 (95% CI 0.49, 0.77)). When we considered the in-person classroom-based health system and community programs together and compared retention in them to retention in the fitness-focused and online programs, retention was significantly worse for the fitness-focused (OR 0.75 (95% CI 0.59, 0.95) and online programs (OR 0.58 (95% CI 0.47, 0.72)).
Fig. 1.
Enrollee retention by NDPP program and week. NDPP: National Diabetes Prevention Program.
Table 3 shows changes in body weight and physical activity at 6 months and 12 months by program. Average percent weight loss at 6 months was 4.1% ± 4.6% and at 12 months was 4.8% ± 5.5%. Thirty-seven percent of enrollees achieved ≥ 5% weight loss at 6 months and 44% achieved the CDC goal of ≥ 5% weight loss at 12 months. Nine percent of enrollees achieved ≥ 10% weight loss at 6 months and 15% achieved ≥ 10% weight loss at 12 months. Fifty-six percent reported ≥ 150 min of physical activity per week at 6 months and 63% at 12 months.
Table 3.
Six-, 12-month, and 2-year changes in risk factors and outcomes for NDPP enrollees overall and stratified by provider.
| Enrolled in NDPP | Health system program | Community program | Fitness-focused program | Online program | p-value | |
|---|---|---|---|---|---|---|
| N (%) | 776 | 277 (36%) | 45 (6%) | 206 (27%) | 248 (32%) | – |
| 6-months | ||||||
| Weight | ||||||
| N | 774 | 277 | 45 | 206 | 246 | – |
| Weight (lbs) | −8.6 ± 10.4 | −9.2 ± 9.6 | −13.0 ± 11.3 | −7.5 ± 10.3 | −8.0 ± 10.9 | 0.0063 |
| Percent weight loss | 4.1 ± 4.6 | 4.5 ± 4.3 | 6.4 ± 5.2 | 3.4 ± 4.5 | 3.8 ± 4.9 | 0.0004 |
| ≥5% weight loss | 37% | 40% | 58% | 34% | 33% | 0.0098 |
| ≥10% weight loss | 9% | 10% | 16% | 6% | 9% | 0.1496 |
| Physical activity | ||||||
| N | 662 | 270 | 42 | 169 | 181 | – |
| Physical activity change (minutes) | 13 ± 119 | 11 ± 123 | 33 ± 81 | 29 ± 115 | −5 ± 122 | 0.0348 |
| ≥150 min/week physical activity | 56% | 69% | 67% | 60% | 32% | <0.0001 |
| 12-months | ||||||
| Weight | ||||||
| N | 514 | 213 | 35 | 142 | 124 | – |
| Weight change (lbs) | −10.1 ± 12.1 | −10.3 ± 12.3 | −16.1 ± 14.0 | −9.5 ± 11.3 | −8.6 ± 11.7 | 0.0126 |
| Percent weight loss | 4.8 ± 5.5 | 5.0 ± 5.4 | 7.9 ± 6.6 | 4.3 ± 5.0 | 4.3 ± 5.4 | 0.0030 |
| ≥5% weight loss | 44% | 45% | 71% | 38% | 40% | 0.0038 |
| ≥10% weight loss | 15% | 15% | 29% | 11% | 17% | 0.0589 |
| Physical activity | ||||||
| N | 329 | 190 | 30 | 59 | 50 | – |
| Physical activity change (minutes) | 24 ± 168 | 8 ± 147 | 98 ± 193 | 78 ± 209 | −24 ± 146 | 0.0004 |
| ≥150 min/week physical activity | 63% | 64% | 73% | 70% | 47% | 0.0407 |
| Body Mass Index | ||||||
| N | 653 | 243 | 38 | 170 | 202 | – |
| Change in BMI (kg/m2) | −1.2 ± 2.4 | −1.3 ± 2.4 | −2.3 ± 2.1 | −0.9 ± 2.2 | −1.0 ± 2.6 | 0.0104 |
| Blood pressure (mmHg) | ||||||
| N | 641 | 240 | 35 | 169 | 197 | – |
| Change in systolic | −2 ± 15 | −4 ± 16 | −1 ± 18 | 0 ± 14 | −3 ± 15 | 0.1191 |
| Change in diastolic | −1 ± 15 | −2 ± 11 | 0 ± 10 | −1 ± 11 | −1 ± 11 | 0.5641 |
| Lipids (mg/dl) | ||||||
| N | 240 | 108 | 16 | 49 | 67 | – |
| Change in total cholesterol | −6 ± 32 | −7 ± 34 | 6 ± 20 | −4 ±30 | −8 ± 32 | 0.4469 |
| N | 238 | 107 | 16 | 48 | 67 | – |
| Change in HDL-cholesterol | 1 ± 8 | 0 ± 8 | 1 ± 7 | 1 ± 9 | 2 ± 7 | 0.4706 |
| N | 234 | 106 | 16 | 47 | 65 | – |
| Change in triglycerides | −27 ± 103 | −39 ± 113 | −3 ± 37 | −9 ± 81 | −25 ± 111 | 0.2993 |
| N | 352 | 135 | 26 | 83 | 108 | – |
| Change in LDL-cholesterol | −4 ± 25 | −3 ± 28 | −1 ± 20 | −5 ± 22 | −6 ± 24 | 0.6464 |
| HbA1c | ||||||
| N | 475 | 197 | 26 | 118 | 134 | – |
| Change in HbA1c (%) | −0.1 ± 0.4 (0 ± 0 mmol/mol) | −0.1 ± 0.3 (0 ± 0 mmol/mol) | −0.1 ± 0.3 (0 ± 0 mmol/mol) | −0.1 ± 0.4 (0 ± 0 mmol/mol) | −0.1 ± 0.4 (0 ± 0 mmol/mol) | 0.7143 |
| Incidence of diabetes | ||||||
| N | 546 | 222 | 35 | 140 | 149 | – |
| Incidence of diabetes by HbA1c | 43 (8%) | 19 (9%) | 1 (3%) | 12 (9%) | 11 (7%) | 0.6820 |
| N | 439 | 171 | 31 | 105 | 132 | – |
| Incidence of diabetes by self-report | 14 (3%) | 6 (4%) | 1 (3%) | 2 (2%) | 5 (4%) | 0.8564 |
| Two-years | ||||||
| Body Mass Index | ||||||
| N | 600 | 216 | 38 | 165 | 181 | |
| Change in BMI (kg/m2) | −0.6 ± 2.7 | −0.8 ± 2.9 | −1.7 ± 3.1 | 0.0 ± 2.3 | −0.6 ± 2.6 | 0.0019 |
| Blood pressure (mmHg) | ||||||
| N | 597 | 220 | 36 | 164 | 177 | |
| Change in systolic | 0 ± 16 | −1 ± 17 | −1 ± 20 | 2 ± 14 | −2 ± 15 | 0.0970 |
| Change in diastolic | −1 ± 11 | −2 ± 11 | −1 ± 9 | 0 ± 11 | −2 ± 11 | 0.3010 |
| Lipids (mg/dl) | ||||||
| N | 160 | 64 | 9 | 39 | 48 | |
| Change in total cholesterol | −2 ± 31 | −5 ± 33 | 6 ± 11 | −5 ± 33 | 3 ± 30 | 0.4261 |
| N | 161 | 64 | 9 | 39 | 49 | |
| Change in HDL-cholesterol | 1 ± 7 | 0 ± 8 | 3 ± 6 | 2 ± 8 | 2 ± 6 | 0.3158 |
| N | 234 | 106 | 16 | 47 | 65 | |
| Change in triglycerides | −27 ± 103 | −39 ± 113 | −3 ± 37 | −10 ± 81 | −25 ± 111 | 0.2993 |
| N | 341 | 126 | 24 | 88 | 103 | |
| Change in LDL-cholesterol | −4 ± 28 | −5 ± 31 | −8 ± 16 | −6 ± 25 | −1 ± 27 | 0.5197 |
| HbA1c | ||||||
| N | 414 | 163 | 21 | 104 | 126 | |
| Change in HbA1c (%) | 0.0 ± 0.5 (0 ± 0 mmol/mol) | 0.0 ± 0.6 (0 ± 0 mmol/mol) | 0.1 ± 0.5 (0 ± 0 mmol/mol) | 0.0 ± 0.5 (0 ± 0 mmol/mol) | −0.1 ± 0.3 (0 ± 0 mmol/mol) | 0.2654 |
| Incidence of diabetes | ||||||
| N | 653 | 251 | 39 | 169 | 194 | |
| Incidence of diabetes by HbA1c | 73 (11%) | 32 (13%) | 3 (8%) | 18 (11%) | 20 (10%) | 0.7301 |
| N | 402 | 151 | 30 | 98 | 123 | |
| Incidence of diabetes by self-report | 25 (6%) | 12 (8%) | 1 (3%) | 5 (5%) | 7 (6%) | 0.6892 |
BMI: Body Mass Index, NDPP: National Diabetes Prevention Program, HbA1c: Hemoglobin A1c, HDL: High-density lipoprotein cholesterol, LDL: Low-density lipoprotein cholesterol, PCP: Primary Care Provider.
Data are expressed as N (%) or mean ± standard deviation. P-values indicate significance between those who enrolled in NDPP program types.
In general, percent weight loss at 12 months was greatest in community program enrollees (7.9% ± 6.6%), intermediate in health system program enrollees (5.0% ± 5.4%), and least in the online (4.3% ± 5.4%), and fitness-focused program enrollees (4.3% ± 5.0%). Percent weight loss was greater for men than women (5.9% vs 4.4% at 12 months respectively, p = 0.0060) and men were more likely to achieve 5% and 10% weight loss at 12 months (50% of men and 41% of women achieved 5% weight loss at 12 months, p = 0.0442, and 23% of men and 12% of women achieved 10% weight loss at 12 months, p = 0.0022). The percentage of enrollees who reported ≥ 150 min of physical activity per week at 12 months was greater for enrollees in the in-person community (73%), fitness-focused (70%), and health system (64%) programs, and least for enrollees in the online program (47%). Men were more likely than women to report ≥ 150 min of physical activity at 6 months but there was no difference at 12 months (at 6 months 67% of men and 52% of women reported achieving the physical activity goal, p = 0.0008, and at 12 months 59% of men and 65% of women reported achieving the physical activity goal, p = 0.3773).
Table 3 summarizes the 1- and 2-year changes in risk factors and outcomes for NDPP enrollees overall and stratified by NDPP provider. Overall, there was a decrease in BMI of 1.2 ± 2.4 kg/m2 at 1 year. The reduction in BMI was greatest among participants in the community program (−2.3 ± 2.1 kg/m2) followed by the health system program (−1.3 ± 2.4 kg/m2), the online program (−1.0 ± 2.6 kg/m2), and the fitness-focused program (−0.9 ± 2.2 kg/m2). There were also improvements in systolic and diastolic blood pressure, lipid profiles, and HbA1c. The improvements in these risk factors did not differ among the enrollees in the four programs. The reduction in BMI persisted at 2 years (−0.6 ± 2.7 kg/m2) and remained significantly greater in those who had enrolled in the community, health system, and online programs compared to the fitness-focused program. Improvements in diastolic blood pressure, total cholesterol, HDL cholesterol, triglycerides, and LDL cholesterol persisted with no differences among enrollees in the different programs. At 1 year and 2 years, 8% and 11% of participants reported being diagnosed with Type 2 diabetes mellitus. Incident Type 2 diabetes mellitus tended to be less frequently reported among community program enrollees at 1 year and 2 years but the differences among enrollees in the different programs were not statistically significant.
There were only slight changes in health-related quality-of-life overall. Most participants had no change in self-rated health at 1 year or 2 years. EQ-5D-5L scores were unchanged although nearly half of participants (45% at 1 year and 46% at 2 years) indicated at least a 3-point increase in their health-related quality-of-life as assessed by the EQ-5D visual analog scale. There were no differences in the change in health-related quality-of-life as assessed by the EQ-5D visual analog scale among enrollees in the different programs.
4. Discussion
We observed clearcut differences in the characteristics of enrollees who participated in the four NDPP programs offered by their employer at no out-of-pocket cost. Enrollees in the in-person classroom-based health system and community programs were older than enrollees in the fitness-focused and online programs. Enrollees in the community and health system programs were also more likely to have received medical care in the past year as evidenced by at least one visit to a primary care provider, claims diagnoses for hypertension and cardiovascular disease, and treatment for hypertension. The latter findings may in part reflect their older ages. Enrollees in the health system and community programs were more likely to report choosing these programs based on their convenient locations. Those who enrolled in the community program were least likely to live in households with internet subscriptions and computing devices. Enrollees in the online program were more likely to report choosing it because of its flexible time. Enrollees were most likely to choose the fitness-focused program because of other people attending the program and its perceived focus on physical activity. Individuals who lived in Ann Arbor were more likely to enroll in the fitness-focused and health system programs which were offered only in Ann Arbor. Enrollees who chose the health system, community, fitness-focused, and online programs did not differ by self-reported health or health-related quality-of-life.
Retention also varied by program (Fig. 1). Despite having the greatest attrition before week 16, the community program had the highest retention at week 44. The health system and fitness-focused programs had slightly greater late attrition than the community program. The difference in longer-term retention between the community program and the health system and fitness-focused programs must be interpreted with caution because of the small number of enrollees in the community program. The online program had substantial attrition between weeks 16 and 20 but thereafter, had better retention. Some of the difference in retention between the in-person programs and the online program may be related to differences in how retention was assessed. Other studies have, however, suggested that retention is lower, and outcomes are worse among enrollees in programs that do not provide in-person contact [13,14].
The CDC reported that between 2012 and 2016, 36% of NDPP enrollees nationwide achieved the weight loss goal of ≥ 5% at 1 year and 42% met the physical activity goal of 150 min per week [4]. In our study, 44% of enrollees achieved the 5% weight loss goal and 63% reported achieving the physical activity goal. Weight loss and physical activity at 1 year differed across the programs. The percentage of NDPP enrollees achieving ≥ 5% weight loss at 1 year was 71% for community program enrollees, 45% for health system enrollees, 40% for online enrollees, and 38% for fitness-focused enrollees. In contrast, 73% of the community, 70% of the fitness-focused, 64% of the health system, and 47% of online program enrollees achieved the physical activity goal.
Older age has previously been associated with greater NDPP retention [3] and greater retention has been associated with greater weight loss [4]. This may explain the greater weight loss among community program and to a lesser extent health system program enrollees. Although prior studies have reported greater weight loss among NDPP enrollees who were white than among participants who were Hispanic or Asian, we did not observe differences in percent weight loss by race [4]. Men were more likely than women to achieve both the weight loss and physical activity goals. These differences in outcomes by sex have been described for the NDPP [4]. The fact that men were more likely to select the fitness-focused program and the fact that enrollees perceived that the program had a greater focus on physical activity may also explain the higher percentage of fitness-focused program enrollees who achieved the physical activity goal. The lower percentage of online program enrollees achieving the physical activity goal may be related to differences in how physical activity was assessed. However, lower self-reported physical activity for the online program enrollees may also be related to the higher attrition as attrition has been associated with less self-reported physical activity [4].
A recent meta-analysis of 63 studies that tested lifestyle modification strategies for Type 2 diabetes mellitus prevention examined weight loss outcomes according to the intervention strategy, provider type, and delivery setting of the programs. Programs were classified as including group education from healthcare professionals in clinical settings, group education from community members in community settings, and education/counseling delivered remotely using technology. The weight loss observed in group education programs in clinical settings led by healthcare professionals was 1.1 kg (95% CI −2.3, 0.1 kg). Weight loss in group education programs in community settings led by community members was −2.1 kg (95% CI −3.1, −1.0 kg). Education/counselling using remote technology was associated with a 0.1 kg weight loss (95% CI −2.8, 2.5 kg). Only the programs that employed group interventions in community settings led by community members achieved significant weight loss [15].
In our study, enrollees in the community program conducted by peer instructors in community settings had the best outcomes. This suggests that trained non-medical personnel are effective intervention delivery agents potentially achieving similar or even better weight loss outcomes. Interestingly, we found that participants in the community program likely had lower socioeconomic position than enrollees in other NDPPs based on geocoding data. Both Cannon et al. [3] and Galaviz [15] reported that racial minority groups and those with lower socioeconomic position had lower NDPP retention and less favorable outcomes. In contrast, a survey of NDPP lifestyle coaches reported that lower-income groups were more likely to report having motivation and family support as facilitators to NDPP participation and adherence compared to higher-income groups [16,17]. Despite historically lower participation and poorer outcomes among minority populations, it is possible that community program enrollees were more motivated and able to achieve greater weight loss. They may have also benefited from the fact that programs were offered in their communities and lead by trained peer instructors.
Despite differences in the likelihood that enrollees in the four NDPPs achieved the CDC weight loss and physical activity goals, we saw improvements in blood pressure, lipid, and HbA1c levels across programs and did not observe differences among programs at 1 year or 2 years. Few previous studies have reported HbA1c outcomes among NDPP participants, and those that did, did not demonstrate significant improvements in HbA1c [15]. We also found no differences in self-rated health across programs at 1 and 2 years but found that almost one-half of enrollees indicated at least a 3-point increase in the EQ-5D visual analog scale for health-related quality-of-life at 1 year and 2 years. Previous research suggests that a 3-point change is clinically significant [18].
In this study, we also found that despite differences in the achievement of weight loss and physical activity goals by enrollees across the four programs, the incidence of Type 2 diabetes mellitus did not differ by NDPP program at 1 or 2 years. Type 2 diabetes mellitus incidence was 8% at 1 year and 11% at 2 years, without significant differences among the programs. Previously, we demonstrated that after adjusting for age group, sex, and race, NDPP enrollees with prediabetes reduced their incidence of Type 2 diabetes mellitus by 40% at 1 year and 20% at 2 years compared to non-enrollees based on HbA1c criteria, and by 57% at 1 year and 46% at 2 years based on self-report [11].
Our study has limitations. First, program uptake was low. Although our total study population was smaller than previous studies of the NDPP [4], retention was good and we were able to describe differences in outcomes by program type. Second, there was heterogeneity in the assessment of retention and physical activity. Fortunately, the CDC has established standard procedures for program evaluation as a condition for accreditation as NDPP program providers that accommodate various methods of physical activity assessment [19]. Third, although enrollees were free to choose the NDPP in which they enrolled, relatively few enrolled in the community program. This may have resulted in both limited power to detect associations and spurious chance associations. Finally, the generalizability of our results may be limited as this study was performed with a single employer group with limited diversity.
In summary, our results suggest that enrollees in in-person programs are more likely to achieve the CDC weight loss and physical activity goals than enrollees in the online program. We also confirmed that weight loss outcomes are noninferior or even better in group interventions led by trained community members in community settings. Finally, we found that when provided a choice of NDPPs in which to enroll, individuals with prediabetes and overweight or obesity can achieve overall Type 2 diabetes mellitus prevention regardless of the program type or delivery method of the NDPP in which they enroll. The challenge remains to increase the diagnosis of and awareness of prediabetes at a population level and to increase program enrollment.
Acknowledgments
The authors thank Marsha Manning, former Manager, Medical Benefits and Strategy at U-M; Ashley Weigl, Associate Director, MHealthy; and Dr. Marc D. Keshishian and Dawn Beaird, Blue Cross Blue Shield of Michigan, for their contributions to this project.
Funding
This work was supported by National Institutes of Health, National Institute of Diabetes, Digestive, and Kidney Diseases grant number R01 DK109995.
Footnotes
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Duality of Interest
No potential conflicts of interest were reported.
Prior Presentation
None
References
- [1].Knowler WC, Barrett-Connor E, Fowler SE, Hamman RF, Lachin JM, Walker EA, et al. Reduction in the Incidence of Type 2 Diabetes with Lifestyle Intervention or Metformin. N Engl J Med 2002. Feb 7;346(6):393–403. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [2].Albright AL, Gregg EW. Preventing Type 2 Diabetes in Communities Across the U.S. Am J Prev Med 2013;44(4):S346–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [3].Cannon MJ, Masalovich S, Ng BP, Soler RE, Jabrah R, Ely EK, et al. Retention Among Participants in the National Diabetes Prevention Program Lifestyle Change Program, 2012–2017. Diabetes Care 2020. Sep 1;43(9):2042–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [4].Ely EK, Gruss SM, Luman ET, Gregg EW, Ali MK, Nhim K, et al. A National Effort to Prevent Type 2 Diabetes: Participant-Level Evaluation of CDC’s National Diabetes Prevention Program. Diabetes Care 2017;40(10):1331–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [5].Herman WH, Joiner KL, Hurst T, McEwen LN. The effectiveness of a proactive, three-level strategy to identify people with prediabetes in a large workforce with employer-sponsored health insurance. Diabetes Care 2021;44:1532–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Joiner KL, McEwen LN, Hurst TE, Adams MP, Herman WH. Domains from the health belief model predict enrollment in the National Diabetes Prevention Program among insured adults with prediabetes. J Diabetes Complications 2022;36(7):108220. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].US Census Bureau Data [Internet], [cited 2022 Dec 20]. Available from: https://data.census.gov/.
- [8].Finlay J, Li M, Esposito M, Gomez-Lopez I, Khan A, Clarke P, et al. National Neighborhood Data Archive (NaNDA): Social Services by ZIP Code Tabulation Area, United States, 2003-2017. Ann Arbor, MI. Inter-university Consortium for Political and Social Research. 2020-November-10. [Google Scholar]
- [9].Melendez R, Li M, Khan A, Gomez-Lopez I, Clarke P, Chenoweth M. National Neighborhood Data Archive (NaNDA): Public Transit Stops by ZIP Code Tabulation Area, United States, 2016–2018. Ann Arbor, MI: Inter-university Consortium for Political and Social Research; 2021-January-4. [Google Scholar]
- [10].Li M, Gomez-Lopez I, Khan A, Clarke P, Chenoweth M. National Neighborhood Data Archive (NaNDA): Internet Access by ZIP Code Tabulation Area, United States, 2015-2019. Ann Arbor, MI. Inter-university Consortium for Political and Social Research. 2021-November-19. [Google Scholar]
- [11].Herman WH, Villatoro C, Joiner KL, McEwen LN. Retention and real-world outcomes of enrollees and non-enrollees with prediabetes in the National Diabetes Prevention Program. Submitted for publication, Journal of Diabetes and its Complications. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Pickard AS, Law EH, Jiang R, Pullenayegum E, Shaw JW, Xie F, et al. United States Valuation of EQ-5D-5L Health States Using an International Protocol. Value Health 2019;22(8):931–41. [DOI] [PubMed] [Google Scholar]
- [13].Reed VA, Schifferdecker KE, Rezaee ME, O’Connor S, Larson RJ. The effect of computers for weight loss: a systematic review and meta-analysis of randomized trials. J Gen Intern Med 2012;27(1):99–108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Kodama S, Saito K, Tanaka S, Horikawa C, Fujiwara K, Hirasawa R, et al. Effect of web-based lifestyle modification on weight control: a meta-analysis. Int J Obes (Lond) 2012;36(5):675–85. [DOI] [PubMed] [Google Scholar]
- [15].Galaviz KI, Weber MB, Straus A, Haw JS, Narayan KMV, Ali MK. Global Diabetes Prevention Interventions: A Systematic Review and Network Meta-analysis of the Real-World Impact on Incidence, Weight, and Glucose. Diabetes Care 2018;41:1526–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [16].Skrine Jeffers K, Castellon-Lopez Y, Grotts J, Mangione CM, Moin T, Tseng C-H, et al. Diabetes Prevention Program attendance is associated with improved patient activation: Results from the Prediabetes Informed Decisions and Education (PRIDE) study. Prev Med Rep 2019;16:100961. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Baucom KJW, Bauman T, Gutierrez Chavez M, Nemirovsky Y, Aguirre MC, Ramos C, et al. Barriers to participation and lifestyle change among lower versus higher income participants in the National Diabetes Prevention Program: lifestyle coach perspectives. Transl Behav Med. 2022. Aug 17;12(8):860–869. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].McClure NS, Sayah FA, Ohinmaa A, Johnson JA. Minimally Important Difference of the EQ-5D-5L Index Score in Adults with Type 2 Diabetes. Value Health 2018;21(9):1090–7. [DOI] [PubMed] [Google Scholar]
- [19].CDC. Centers for Disease Control and Prevention Diabetes Prevention Recognition Program. Standards and Operating Procedures. 2021.