journal article review: "Long-term results of an Obesity Program in an Ethnically Diverse Pediatric Population."

One of the best parts of being in the Ellyn Satter Institute is the monthly Journal Club where we review relevant articles. I nominated this one for its promising conclusion, of course with a healthy dose of curiosity and skepticism (as we should all approach all journal articles...)Here are the authors' words from the article in Pediatrics:"... benefits of an intensive lifestyle program can be sustained 12 months after completing the active intervention phase." and, "this gives us hope that behavior change is possible, even in the most challenging populations."Several points came up after our reading and discussion that have me less than enthused about the intervention and the conclusions. In other words, I am not convinced that this study proved the program is effective or better than nothing, or better than other interventions. (It did seem to prove that it is better than the standard care at the Yale Pediatric Obesity Center...)Here are the issues I have with this study, in random order...• the age range is from 8-16, the mean was 12 (this is a significant age spread, and the children were lumped in together for analysis. It is hard to compare an 8 year old boy with a 14 year old girl in pretty much every way I can imagine, but particularly with many of the blood tests/hormone levels. Also, is a curriculum developmentally appropriate for 8 year to 16 year-old span?)• dropout rate was significant. Is any behavioral intervention "successful" if more than half quit? (Granted, it is exceedingly difficult to do follow-up for two years. It just is.)-105 started for intervention group, 45 completed-control group 86 started, 31 finished• the authors claimed there was "no difference" between those who completed and those who dropped out based on initial objective characteristics. There was no explanation as to why they dropped out. Were there unusual eating habits? Stress or stigma around weight? Were they losing weight on their own, a natural process for many after an initial weight gain pre-puberty? Were there objections to the study design? Were older children dropping out more, less? I would have liked more information. Are the study designers OK with such a high rate? They didn't seem particularly troubled, or wonder if it was something about the program that caused dropout rates to be so high. (Of note, most weight loss studies have high drop out rates, HAES studies tend to have far lower drop-out rates and have also shown improved bio-markers in studies with adults- Linda Bacon just did a nice review.)• Also, the initial intervention group had a "structured meal plan" arm, which was discontinued because over 80% of those families dropped out in the first 6 months. I would have LOVED the conclusion to also come out and say, "Our study showed that structured meal plan approach to childhood obesity appeared particularly ineffectual and should not be used." What is harmful or ineffectual is also critically helpful information, and shouldn't just be glossed over.• the findings were lumped together for all the children, whether they were 8, 12 or 16, ignoring or not addressing that major physiological differences, particularly for insulin and related tests, exist between an 8 year-old and a pubescent young person.• BMI Z-score is controversial to follow populations in studies . Flegal (CDC) recommends using BMI% change from 95%. I don't know that I have enough info to form an opinion on this, but would like to see the data as % change as well.• the graph for BMI Z-score change starts at 6 months, not showing the trend for the initial 6 months. Also, the trend for the control group declined significantly from 12-24 months, while, though small, the trend for the intervention group increased from the 12-24 month group. Get that? The trend in the control group was for decreasing BMI, and for the intervention group the trend was increasing. Are we seeing the beginning of lost weight being regained that is such a common phenomena?• there was some question that fasting insulin is not a great measure of insulin tolerance, particularly if a portion of the children are experiencing the normal peaks and valleys of adolescence. May confound results if again, an 8 year old is compared to say, a 13 year old.• the intervention group looks like they were dumped into the same treatment as the control group from 12-24 months, thought this wasn't specifically addressed.• the "control" group again was not a strict "control group." (A major beef I have with most childhood obesity trials I have read.) They had every 6 months meetings at the Yale Pediatric Obesity Clinic with "general diet and exercise counseling" (about 30 minutes.). How can we know if the intervention helped, or the control treatment (standard of care?) harmed?• they did not separate  or control for exercise effect. Perhaps the markers improved in the group that was more active vs those that had BMI changes? Many studies now try to control for this, and those that do often show that fitness is a better predictor of health than BMI alone.• if traditional care was so ineffectual, why did the conclusion not challenge traditional care more?• they claim it is a  "non-diet approach that emphasized low-fat, nutrient-dense foods of moderate portions. Topics included "Determining portion sizes," and "Better Food Choices: A NonDiet Approach." I am intrigued by their definition of "non-diet."*• the program used on the intervention families in the study is for sale, developed by dietitian, and lead study author (Mary Savoye) at Yale, though they claim "no financial relationships relevant to this article." I don't understand how there is no conflict if the author is then selling the program? Would that not cause a bias for her to look for more favorable results for her work, if not the question of a direct financial gain? I really don't understand this. Maybe someone can explain it to me. Ideally, shouldn't the study be authored by someone other than the person who developed the intervention?• they talk about "behavior change" in their final article, but don't actually measure any behaviors via intake or exercise logs, or other means...• gold-standard for weight loss follow-up up is five years ideally, though again, VERY difficult to sustain a study that long.• they did not look for disordered behaviors pre or post intervention to my knowledge. Were kids dieting? Fasting? Binging? Vomiting? Recent reports state that bulimia is more common than previously thought in African American women. Were the interventions helping, harming?positives: The exercise for the kids was based on "games" and play (though the video I saw online showed kids standing around in a gym doing jumping-jacks), they attempted long-term follow-up, they involved the families, they are looking for solutions for an at-risk population, they showed how ineffectual structured meal plan approach is (but didn't stress it in the findings.)What do you all think? Also, what is your definition of "non-diet?*" Note: these comments are wholly my own, and don't represent any official commentary from ESI.