Health coaching research spans diabetes prevention, weight management, physical activity, and chronic disease support. The statistics below show the size and mix of the evidence, how National Diabetes Prevention Program organizations staff and deliver programs, and what selected studies measured over time.
Contents
- Health coaching evidence at a glance
- The health coaching research mix
- Who delivers coaching
- Where and how programs operate
- Diabetes prevention and metabolic outcomes
- Digital and telephone-supported coaching
- Weight maintenance and physical activity
Health coaching evidence at a glance
The 2023 addendum to the Health and Wellness Coaching Outcomes compendium summarized 212 health and wellness coaching studies. Of those, 108 were randomized clinical trials. The review is a secondary synthesis of mixed study designs, so these figures describe the research base rather than one pooled treatment effect.
Several topic-specific counts show where the literature is concentrated. The compendium identified 106 studies addressing diabetes, 109 addressing obesity, and 32 addressing heart disease. The diabetes set included 46 randomized controlled trials and 21 quantitative studies. The heart-disease set included 24 randomized controlled trials and one quality-improvement study.
The same synthesis reported outcome signals within the studies it reviewed:
| Evidence area | Reported statistic | Measurement period or scope |
|---|---|---|
| Diabetes studies | 106 studies | 2023 addendum |
| Diabetes studies with randomized controlled trials | 46 trials | 2023 addendum |
| Diabetes studies measuring A1c | 79% reported a decrease | Studies in the compendium |
| Obesity studies | 109 studies | 2023 addendum |
| Obesity studies with positive weight or BMI outcomes | 81.8% | Studies in the compendium |
| Heart-disease studies | 32 studies | 2023 addendum |
These percentages should be read as descriptive findings from the compendium, not as guarantees for an individual coaching client. The underlying studies may differ in population, intervention, comparison group, follow-up, and outcome definition.
The health coaching research mix
The diabetes evidence is especially substantial within the compendium. It includes 106 studies, but only some studies used randomized designs or quantitative measures. Among studies that measured glycated hemoglobin A1c, 79% reported a decrease. That statistic describes the direction reported in the reviewed studies; it does not provide a single average amount of change.
The obesity evidence set contained 109 health coaching studies. Positive body-weight or body-mass-index outcomes were observed in 81.8% of those studies. Again, the statistic is a study-level summary. It does not mean that 81.8% of people receiving any form of coaching will achieve a positive result.
Heart disease was represented by 32 studies. The evidence set included 24 randomized controlled trials and one quality-improvement study, alongside other study designs. This mix matters when comparing the apparent strength of evidence across topics: a count of studies alone does not establish that interventions, populations, or outcomes were comparable.
For readers evaluating a health coaching claim, the most useful questions are often practical ones: Was the result randomized? How many people took part? How long was follow-up? Was the outcome weight, BMI, A1c, activity, knowledge, or another measure? The studies below illustrate why those details change interpretation.
Who delivers coaching
The CDC analysis of National Diabetes Prevention Program delivery organizations surveyed 440 respondents in August–September 2021. Respondents could identify more than one role where the source allowed multiple selections. Among them, 399 people, or 90.7%, identified as lifestyle coaches. Another 182 respondents, or 41.4%, identified as program coordinators, and 48, or 10.9%, identified as master trainers.
Lifestyle coaches were on staff at 430 of the 440 organizations, or 97.7%. Those organizations reported a mean of 7.4 coaches per organization. The mean number of staff members dedicated 100% to the National DPP was 2.0.
The survey also describes the respondents’ demographic composition. Women represented 387 respondents, or 88.0%. Black or African American respondents represented 75 people, or 17.0%, while White or Caucasian respondents represented 264, or 60.0%. Respondents aged 25–34 accounted for 86 people, or 19.5%, and those aged 45–54 accounted for 92, or 20.9%.
These figures describe people responding for delivery organizations during the 2021 survey period. They are not a census of every health coach and should not be used to infer the composition of the entire profession.
Where and how programs operate
National DPP delivery organizations came from several organizational settings. Health care or hospital organizations accounted for 138 of 440 organizations, or 31.4%. Community-based health care organizations accounted for 104, or 23.6%, and government agencies accounted for 59, or 13.4%.
Delivery formats were similarly varied. Virtual delivery was reported by 343 organizations, or 78.0%. Live distance delivery by video or audio was reported by 252, or 57.3%. In-person small-group delivery with 20 or fewer participants was reported by 219 organizations, or 49.8%.
The survey reported organizational reach and setting as well. Medium organizations serving 1,000–50,000 people annually across all programs represented 151 respondents, or 34.3%. Rural organizations represented 181 respondents, or 41.1%, while urban organizations represented 182, or 41.4%.
Because the source notes that some questions allowed multiple selections, delivery-mode and related percentages should not automatically be added together. An organization could report more than one format or characteristic.
Diabetes prevention and metabolic outcomes
The CDC’s Prevent Type 2 Diabetes: Talking to Your Patients About Lifestyle Change page reported an estimated 115.2 million U.S. adults with prediabetes in the 2024 information cited on the page. It also reported that 8 in 10 U.S. adults with prediabetes do not know they have it. These are U.S. estimates and are not measures of the number of people currently receiving health coaching.
The Diabetes Prevention Program research basis linked lifestyle change after 5%–7% body-weight loss with a 58% lower incidence of type 2 diabetes over the trial’s 3-year outcome period. For people aged 60 and older, the same lifestyle intervention was associated with a 71% lower incidence over that period. In 15-year follow-up research, National DPP lifestyle-change participants had a 27% lower incidence of type 2 diabetes.
The CDC program description states that the National DPP lifestyle-change program targets at least 150 minutes of physical activity per week alongside calorie reduction. This is a program target, not an outcome showing that every participant reached the target.
The distinction between program goals and measured outcomes is important. A coaching program may provide education, activity targets, tracking, and support, while a study may measure weight, A1c, diabetes incidence, or attendance at a particular follow-up point. Those measures answer different questions.
Digital and telephone-supported coaching
The CDC’s Evaluation of a Digital Diabetes Prevention Program Adapted for Low-Income Patients, 2016–2018 evaluated a digital DPP that provided 52 weeks of education, health coaching, and peer support. The evaluation enrolled 227 low-income adults with prediabetes. Mean baseline BMI was 34.6, and mean baseline HbA1c was 5.8%.
At 12 months, mean weight loss in the digital-DPP intervention group was 4.4%. At the first follow-up, described as approximately 6 months, the intervention-versus-control difference in mean BMI percentage change was 3.3%, with P=.001. The cohort, intervention, and follow-up period are specific to this 2016–2018 evaluation; the results should not be treated as a universal estimate for all digital coaching programs.
A separate Sydney health-coaching program report covered a 2013 cohort of 238 patients. Of these, 178 participated in face-to-face sessions with telephone follow-up, while 60 received telephone-only coaching. Among patients whose baseline HbA1c exceeded 7%, mean HbA1c fell from 8.5% to 7.9% over 12 months.
Mean diabetes-knowledge scores in the Sydney cohort rose from 24.4 to 25.2 from baseline to 12 months. The study reported no BMI change in either coaching format over 12 months. Together, these results show why health coaching outcomes should be separated by endpoint: a program can report a change in A1c or knowledge without reporting a BMI change.
Weight maintenance and physical activity
The EHR-based weight-maintenance randomized trial enrolled 194 adults who had previously lost at least 5% of body weight. The coaching group received 24 scheduled coaching contacts over 24 months. At the 24-month follow-up, 157 of 194 randomized participants, or 81%, completed the trial.
Mean baseline weight was 85.8 kg and mean baseline BMI was 30.4 kg/m². At 24 months, mean weight regain was 2.1 kg with coaching compared with 4.9 kg with tracking tools alone. The adjusted between-group difference in 24-month weight change was −2.86 kg, with a 95% confidence interval from −4.60 to −1.11 kg. At that point, 65% of the coaching group versus 50% of the tracking group maintained at least 5% weight loss.
The result is specifically about maintenance after an earlier loss, not initial weight loss in an unselected population. It also comes from an EHR-based intervention with scheduled contacts over two years, so the contact schedule and follow-up duration are part of the context.
Another randomized trial of health coaching in obese adults randomized 44 middle-aged White adults: 22 to coaching and 22 to control. Over 12 weeks, percent excess-weight loss was 15.7% with health coaching versus 2.5% in controls. Moderate-to-vigorous physical activity increased by 50.3 minutes per week with coaching versus 7.1 minutes per week in controls.
That trial’s small sample and specific participant description limit how broadly its findings can be applied. Across the studies summarized here, the clearest interpretation is therefore conditional: health coaching has been studied in multiple formats and populations, with outcomes varying by endpoint, intervention design, comparison group, and follow-up period.