Zero in on a ShapeUp Obesity Study

I’m stepping aside from the series about the ShapeUp Employer Wellness Survey results to take a knee-deep dive into the obesity study published in the journal Obesity in 2009.

Bottom line: ShapeUp met a standard for excellence by publishing, in a peer-reviewed journal, the positive outcomes generated by it’s social-based wellness platform. The study should be a key consideration for any prospective purchaser who seeks evidence-based solutions (and that should be all of us). But…while the study, despite some limitations, gives ShapeUp the grounds on which to argue, “Anyone serious about controlling behavior-driven health care costs in America would be wise to zero in on weight loss interventions,” those of us in the field should take pause before accepting this conclusion in the face of years of experience in which effective employer-based behavioral solutions to obesity have eluded us.

Key points of the study:

  • It included 4,717 adults enrolled in a 16 week ShapeUp program, of which 3,311 completed at least 12 weeks.
  • 83.6% of the subjects were women.
  • Participants self-reported their weight on ShapeUp’s tracking system. In addition, they were e-mailed a 6-month follow-up questionnaire to assess weight loss over time.
  • This was a statewide program, not an employer group. The majority (62%) of participants, however, had their 15-dollar registration fee paid by their employer.
  • Objective data was collected from a sub-group of 201 participants by weighing them at the beginning and after completion of the program.
  • 30% of those who completed the program achieved a “clinically significant” weight loss of 5% or greater.
  • The study population that was obese declined from 39% to 31%.
  • To their credit, the study authors conducted an intent-to-treat analysis of the study results, deeming subjects who did not complete 12 weeks of the program to be noncompleters and assuming they remained at their baseline weight.

This was a reasonably rigorous and very promising study. But it’s premature to call the matter settled. Some limitations include:

  • The study population was not representative of most employer populations, as it was predominated by females who self-selected into the study (though the male participants lost more weight, on average, compared to the females).
  • The authors stated that a strength of the program was that body weight was reported six months post-intervention (albeit via self-report; the study does not mention any of the objectively measured group as being weighed more than four weeks post-intervention). But one also could argue that a weakness of the study was that measurements were collected no more than 6 months after the program. Study author Rena Wing, herself, has proposed “defining successful long-term weight loss maintenance as intentionally losing at least 10% of initial body weight and keeping it off for at least 1 year.”
  • The study does not mention adverse effects. We do not know whether subjects initiated unhealthful activities, such as fad diets, weight loss supplements, weight cycling (yo-yo dieting), purging, disordered eating, and dehydration, or whether there was a concurrent increase in body image disorders. But behavioral weight loss studies almost never address adverse affects.
  • There was no control group. Better to have no control group rather than no study. But we also must understand this limitation before drawing conclusions. The ShapeUp program studied was accompanied by a broad-based social marketing campaign (as it should be in the workplace, too). Was the weight loss in the studied population, in fact, a result of the ShapeUp program? Did nonparticipants lose weight, too? We don’t know.

Easy for me to sit back in the blogosphere and poke holes in studies. The ShapeUp study was good. I wish every vendor measured their products’ effectiveness with this level of rigor. What else can you ask? And ShapeUp recently published another study and probably has more in the pipeline. In the end, ShapeUp has grounds to advise employers to focus on weight loss. But employers have grounds to be skeptical.

Perhaps after we see repeated independent, controlled studies in a variety of employer settings consistently demonstrate the long-term effectiveness of ShapeUp (or comparable programs), we should all, indeed, flock to adopt these approaches. Until then, we have to at least ask whether taking ShapeUp’s advice to zero in on weight loss may take place at the expense of interventions for other conditions that lead to more beneficial health and economic outcomes.

(This post is fourth in a series. For the full series, click here).

, , ,

  1. #1 by Rajiv Kumar, M.D. on March 23, 2012 - 9:44 am

    Thanks for diving so deeply into our study here Bob. You’re absolutely right that the holy grail for all of us will be the ability to show sustained outcomes over time in randomized controlled trials. We’ll be publishing some longitudinal data soon showing sustainability of weight loss outcome generated from the ShapeUp social model for health improvement — and the numbers look pretty darn good. The challenge with RCTs is that in a risk-averse employer environment, it’s very hard to get an HR department to agree to let us randomize their population and only offer an intervention to 50% of their employees. Especially when that intervention creates a lot of buzz, like the team-based wellness challenges that we run. If you know of any large employers who are amendable to this type of approach to studying the effectiveness of wellness interventions within their population, please let me know!

    • #2 by Bob M. on March 24, 2012 - 10:07 pm

      Thank you for the feedback, Dr. Kumar. I should mention that after I published the original post I became aware, via the ShapeUp blog, of the fact that there was, indeed, follow up to the original study. I haven’t yet been able to access the full article, but I see, according to the blog (and if I am correctly associating the blog posts with the studies and their findings) that that “73% of weight loss was maintained at the 10-month mark.” That’s very encouraging.

      I understand the challenge of conducting a randomized, controlled, trial in an employer environment. Employers cannot expect randomized worksite studies, although it seems some studies have managed to use nonrandomized control groups (in which, say, employees in Building One are the treatment group and those in Building Two are the control group). Even as I’m writing this, I’m imagining all of the operational issues — from segmenting program promotion to the treatment group only, to creating a highly unusual eligibility file.

      You definitely have me thinking about what is reasonable to expect in terms of evidence of employer-group success. It’s a quandary. :)

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 1,090 other followers