The Health Coaching Research Has a Problem: What Does “Health Coaching” Actually Mean?

Sep 7

6 min read

You've probably seen some version of this statement: “Research shows that health coaching works.”

But there's a surprisingly difficult question hiding inside that sentence: What exactly is health coaching? Is it a 45-minute conversation using motivational interviewing? A nurse calling a patient every two weeks? A structured diabetes-education program? Goal setting through an app? A clinician giving lifestyle advice? A trained coach helping a client develop their own goals?

Researchers have been publishing studies labeled “health coaching” for years, but the interventions being studied can look remarkably different from one another which creates a fundamental scientific problem. If researchers aren't consistently studying the same thing, how confidently can we talk about whether that thing “works”?

A recently published study tried to answer exactly that question, and its findings are worth understanding.

Researchers Examined 54 Health Coaching Studies

Researchers writing in the International Journal of Behavioral Medicine developed a new tool called the Health Coaching Quality Index, or HCQI. Their objective wasn't primarily to determine whether health coaching works. They asked something that logically comes first:

When a research study says it delivered health coaching, how closely does the intervention actually resemble high-quality health coaching?

The researchers reviewed 54 published intervention studies involving health coaching. They evaluated the interventions against theory-driven criteria intended to capture core characteristics of coaching, including areas such as:

coach training;
theoretical foundations;
collaborative goal setting;
strategic questioning;
client-centeredness;
intervention consistency; and
supervision and fidelity.

They then classified the coaching intervention in each study as high, medium or low quality.

The results were striking.

Only 24% Were Classified as High-Quality Coaching

Of the 54 interventions:

13 (or 24%) were classified as high quality.

22 (41%) were classified as medium quality.

19 (35%) were classified as low quality.

So more than three-quarters of the published interventions reviewed by the researchers did not meet their threshold for high-quality health coaching.

That doesn't mean 76% of the studies were bad studies, and it doesn't necessarily mean the interventions were ineffective. The HCQI evaluates the characteristics and reported quality of the coaching intervention vs. the overall scientific quality of the study which is an important distinction.

Still, it exposes a problem that anyone interested in evidence-based coaching should understand. Two papers can both say they're studying “health coaching” while studying meaningfully different interventions.

Imagine This Problem in Another Field

Suppose researchers wanted to determine whether strength training improves health. One study defines strength training as supervised weightlifting three times per week, and another gives participants resistance bands and an instruction sheet. Another asks participants to do 20 push-ups once a week, and yet another provides general advice to “exercise more.”

Then researchers combine all four studies and ask: "Does strength training work?"

The question has become difficult to answer because the intervention itself isn't consistent.

Health-coaching research faces a version of this problem. Some interventions involve highly trained professionals receiving extensive coaching-specific preparation and supervision, while others involve comparatively limited training. Some are explicitly grounded in behavior-change theories, and others are not. Whether it is emphasis on client-generated goals, prescribed goals, strategic questioning, or collaborative exploration, there are inconsistencies. Yet they can all appear in the scientific literature under the same broad label: Health coaching.

What Did “High-Quality” Coaching Look Like?

The 13 interventions receiving high HCQI scores shared some interesting characteristics:

All reported coach training.
Ten of the 13 used a training manual.
Nine reported extensive training of 16 hours to more than 100 hours, together with protocol checks and supervision.
Twelve of the 13 used licensed professionals, including nurses, physicians, psychologists, social workers, medical assistants and dental professionals.

High-quality interventions also tended to have clearly described theoretical foundations and specific coaching strategies.

Goal setting was particularly interesting. In seven of the 13 high-quality interventions, goals were either initiated by the patient or collaboratively established, rather than simply prescribed to the participant. That distinction gets at one of the central questions surrounding the definition of coaching. If a professional tells a patient exactly what to eat, how much to exercise and what goals to achieve, is that coaching? It may be excellent healthcare or education, and t may produce better outcomes, but how do we define whether this is coaching? A standardized research definition matters precisely because those aren't necessarily the same intervention.

Higher-Quality Coaching Studies Had Better Results

Among the studies with high-quality coaching interventions, 12 of 13 (~92%) reported significant improvement or treatment effects in at least one primary outcome.

For medium-quality interventions, it was 17 of 22 (77%). For low-quality interventions, it was 12 of 19 (63%). At first glance, that looks like a remarkably clean pattern: Higher-quality coaching means a greater likelihood of positive outcomes.

That said, the study does not establish that increasing an intervention's HCQI score causes better health outcomes. This wasn't a randomized experiment in which patients were assigned to high-, medium- or low-quality coaching, and there may be other differences between the studies. Also, as the researchers explicitly note, the HCQI evaluates coaching quality rather than overall methodological quality. A study could theoretically contain an excellent coaching intervention but weak research methodology (or vice versa). So the the actual conclusion is generally characterized as “The relationship is interesting enough to investigate further” which is how evidence should work.

Another 2026 Review Found a Similar Problem

The HCQI study isn't the only recent research highlighting inconsistencies in health coaching.

A separate 2026 scoping review examined how researchers assess whether health coaching for people with or at risk of type 2 diabetes is actually delivered as intended (a concept researchers call intervention fidelity).

The authors reviewed 15 eligible studies. Twelve used checklists or study-specific forms to evaluate coaching fidelity. But only four of those 12 tools were validated. The researchers also found substantial differences in health-coach roles, training and implementation.

Their conclusion was fairly straightforward: the lack of standardized and validated tools makes it harder to determine whether health coaching is being delivered consistently, and therefore harder to interpret its impact. This is not a minor methodological issue!

Why “Fidelity” Matters

Imagine researchers design a study of motivational interviewing. They carefully specify how the intervention should work, but nobody checks whether the people delivering it actually use motivational interviewing. Some do, some mostly give advice, some lecture, some ask open-ended questions, some follow the intended intervention closely, and others barely follow it at all.

At the end of the study, researchers measure the results. What exactly did they test?

That's the problem intervention fidelity attempts to solve. Before concluding that an intervention works (or doesn't!) researchers need some confidence that participants actually received the intervention being studied.

For health coaching, that can be particularly difficult because coaching is conversational, individualized and inherently less standardized than swallowing a 10-milligram tablet.

This Changes How We Should Read Headlines About Coaching Research

None of this means health-coaching research should be dismissed, but it does mean we should read it more carefully. The next time you encounter a study claiming that “health coaching” improved weight, HbA1c, blood pressure, medication adherence or quality of life, consider asking:

Who delivered the coaching?
What training did they receive?
Was the intervention actually client-centered?
Were goals prescribed or collaboratively developed?
Which behavior-change or coaching techniques were used?
How frequently did coaching occur?
Was there supervision?
Did researchers check whether coaches actually delivered the intervention as designed?

And perhaps most importantly: Was this actually coaching? Or was “health coaching” simply the label researchers gave the intervention? Those questions don't undermine the evidence, but hey make us better consumers of it.

The Bigger Scientific Problem

Health coaching is still a relatively young field, and young fields often face a difficult transition on the path to standardization.

Researchers need common terminology, replicable interventions, validated measurement tools, clearer descriptions of coach training, ways to assess fidelity, and enough consistency that a researcher studying health coaching in Boston and another studying health coaching in Sydney are at least talking about approximately the same intervention.

The Bottom Line

The health-coaching evidence base has an unusual problem: Before researchers can definitively answer whether health coaching works, they need to get better at defining what health coaching actually is.

A recent analysis of 54 interventions found that only 24% met the researchers' threshold for high-quality health coaching. Intriguingly, those higher-quality interventions were also more likely to report positive outcomes. That doesn't prove causation, but it gives researchers an important hypothesis to test:

We can ask ourselves the questions:

Which components of coaching work?
How well must they be delivered?
How much training does the coach need?
Which patients benefit?
And under what circumstances?

Those questions are less satisfying than a headline declaring that coaching either “works” or “doesn't work", and they are much more scientifically interesting.

Sources

Brooks JM, et al. Development of the Health Coaching Quality Index: A Proposed Method for Evaluating Health Coaching Interventions. International Journal of Behavioral Medicine. The study reviewed 54 health-coaching interventions and developed the HCQI as a proposed tool for assessing intervention quality.

Assessing the fidelity of health coaching for adults with or at risk of developing type 2 diabetes: A scoping review. Patient Education and Counseling, 2026. The review examined 15 studies and identified substantial variation in coaching roles, training, implementation and fidelity measurement.

Blog