Comment on “Health Coaching: 100 Strategies for Weight Loss: A Systematic Review and Meta-Analysis”

Dear Editor:

In their meta-analysis on health coaching strategies for weight loss, Sieczkowska et al. (1) concluded that available evidence is not of “sufficient quality” to support the use of self-reported health coaching as an intervention for weight loss. Given their chosen methods and data interpretation, we argue this conclusion is premature and possibly inaccurate. The authors’ choice to include any published article self-reported as “health coaching” is problematic and not acceptable at this point in the development of the field of health coaching.

We have published 2 papers (2, 3) using a standardized definition and criteria for health coaching. This definition agrees with the one used by the National Board of Health and Wellness Coaches (NBHWC), which is the leading body certifying health coaches. NBHWC is a nonprofit affiliate of the National Board of Medical Examiners, responsible for the development of physician licensing examinations in the United States since 1915. The NBHWC is referred to in the Current Procedural Terminology (for health care services recording and billing) codes for health and well-being coaching services recently approved by the American Medical Association (AMA). Accordingly, health and wellness coaching is a patient-centric process whereby coaches assist clients to use insight, personal strengths and resources, goal setting, action steps, and accountability to achieve healthy lifestyle change. There is an accepted knowledge base for health coaching—an “episteme”—that should be recognized and applied.

Sieczkowska et al. provided a definition of health coaching in their Introduction, and this is similar to the NBHWC definition. Ideally, Sieczkowska et al. would have used that definition to provide objectivity for inclusion criteria in their study. Instead, Sieczkowska et al. collected “coaching” studies and stated, “given the lack of a consensual definition of health coaching, and to better capture all the possible ways this intervention has been employed in the literature, we included any study described as ‘coaching’ by the authors.” In our compendium (2, 3) we identified 58 peer-reviewed papers as health coaching to treat obesity. In this process, >100 papers did not meet standardized criteria, although many of these papers claimed to be “coaching.” In fact, of 58 articles meeting criteria for a standard health coaching definition, only 8 were included in the primary meta-analysis by Sieczkowska et al. Not using standardized criteria for inclusion is likely to result in acts of omission and commission when selecting papers, leading to a poor representation of the health coaching literature.

Another potential shortcoming of this paper is the strict interpretation of Cohen's d and quick dismissal of significant, yet small, effect sizes as “trivial.” A small effect size should always be interpreted with consideration for the underlying outcome measure. Glass et al. (4) decried that there is no wisdom in associating effect size metrics with small, medium, large unless the context is fully understood. These authors further argued that a huge effect size (e.g., 2.0) might be “poor” while one of 0.1 might be “good.” Glass is one of the creators of the meta-analysis technique. A similar conclusion was reached by Kraft (5) of Brown University when describing results from educational interventions—effects that are small by Cohen's standards are large relative to the impacts of most field-based interventions. In this context, he went on to further define an effect size of 0.05–0.20, such as those seen in the Sieczkowska et al. (1) paper, as moderate.

Weight loss is extremely difficult to achieve and only a 5% decline in body mass is considered clinically significant (6). Therefore, a rather small change or difference in weight, relative to the SD of body weight, can be rather important. If Sieczkowska et al. (1) had provided raw mean differences in addition to effect size then readers could understand if a very low, but significant calculated d represented >5% weight loss and had clinical relevance. Without knowing mean weight loss achieved in these studies it is difficult to judge practical importance. When 15 of 16 studies consistently find a positive intervention effect there is something to be further investigated and not dismissed as trivial. Glass might say this is context for discussion.

The question of interpreting effect size aside, the paper in question simply does not present a representative look at the true health and wellness coaching literature. With published papers and the NBHWC providing standards for health and wellness coaching, the lack of defining criteria to select studies for a meta-analysis cannot be overlooked. There was a call to better define the field of health coaching (7); that call was heard, answered (8), and essentially adopted by the NBHWC and AMA. Those currently publishing health and wellness coaching research should accept and make use of this definition and not fall back on the claim that health coaching lacks guiding concepts and a clear definition. When doing a meta-analysis, failure to adopt established criteria to select health-coaching papers will result in an unrepresentative literature sample and potentially faulty conclusions.

Notes

The authors reported no funding received for this letter.

Author disclosures: Potential conflicts of interest: MM is CEO of Wellcoaches Corp and GEM is CMO of Wellcoaches Digital. GAS is a research advisor for Wellcoaches. SH reports no potential conflict of interest.

All authors read and approved the final manuscript.