Comment on Harmful USPSTF Kids Weight Loss Recommendation - Part 2

This is the Weight and Healthcare newsletter! If you like what you are reading, please consider subscribing and/or sharing!

The US Preventive Services Task Force has put forth draft guidelines for higher-weight children and they are open for public comment until January 16. In Part 1, I gave a brief overview of the good (not much) and bad (too much) of these guidelines, and offered some sample answers to the Draft Evidence Review questions.

Today I’ll give sample answers to the Draft Recommendation Statement and do a deeper dive into the evidence they claim supports their choices.

Before I begin, I want to acknowledge that this piece is incredibly long. Normally I would break it up across two newsletters (or three!) but I know that organizations like NAAFA and FLARE are putting out their calls to action this week. I’m honored that they link to my work to help support their communities, and I wanted those folks to have all the information they could get. That said, I’m not going to publish a newsletter on Saturday to give y’all some extra time to read this before I send out something else.

Finally, I want to give a quick shout-out to my paid subscribers (understanding that not everyone can/wants to pay and that’s completely fine – I’m glad you are here!) Analyzing this information has been almost a full-time job for the last week. My ability to take that kind of time is thanks to subscriber support and I’m truly grateful. And thanks to my amazing partner Julianne for proofreading all of this!

Ok, let’s get to answering and analyzing! I’m going to start with Draft Recommendation Statement public comment questions (remember that the Draft Evidence Review questions are in Part 1 here.) My analysis of their evidence is below the draft answers.

DRAFT ANSWERS

I’m going to offer various points and you are welcome to use this information, but it’s important that the answers that are submitted are different – otherwise they can get flagged as duplicates or spam. So, feel free to use these points and include additional points, your experiences, and your own words.

You don’t have to answer all the questions. I think that question 6 is particularly important (“Please share any experiences that you think could further inform the USPSTF on this draft Recommendation Statement”) because they are claiming that putting children and adolescents on intensive weight loss interventions doesn’t cause harm based on the fact a few studies that did only tracked harm for a year at best. So if you are someone who was harmed by these interventions, I think it’s critical to send this information so that everyone who is making decisions around this has the opportunity to know! .

EDIT: In my original post I did not make it clear that the group who compiled the evidence synthesis (Kaiser Permanente Evidence-based Practice Center, Kaiser Permanente Center for Health Research) did not actually vote on the guidelines. Voting on the guidelines was only by USPSTF members. Also, I’ve learned that it is best to direct critiques of the *recommendation* to the recommendation statement rather than the evidence review for the greatest impact.

You can add your own answers here, the deadline is January 16.

Question 1:
Based on the evidence presented in this draft Recommendation Statement, do you believe that the USPSTF came to the right conclusions?

No; I do not believe the USPSTF came to the right conclusions.

NOTE: I feel that this is a poorly worded question. The USPSTF only made one recommendation which was to recommend comprehensive intensive behavioral interventions to children 6 and older (which I and, I would argue, the studies they used, do not support,) but in doing so they came to a lot of conclusions, including the choice not to recommend pharmacotherapy (which I agree with.)

While one option is “Somewhat; I believe the USPSTF came to the right conclusions in some ways but not in others,” it’s not clear (and not directly asked) which conclusions I agree with and which I don’t, so I will choose “no” to make sure that there isn’t a chance that my answer will be interpreted as supporting their weight loss recommendation in any way.

Question 2:
Please provide additional evidence or viewpoints that you think should have been considered.

An important viewpoint to consider is that the studies included, as well as the contextualizing information and evidence in the evidence synthesis, either fail to adequately support, or directly contradict, the recommendation of intensive weight loss interventions to children and adolescents ages 6 and up. The recommendations also fail to adequately assess the possibility of harm.

This recommendation is based on only 50 studies, most with quite small sample sizes, none of which included the entire population (children and adolescents ages 6 and up), none of which used the same intervention as another, and none of which offered data beyond about a year (which typically included the intervention time.)

Their recommendation is based only on changing the size of the children and adolescents’ bodies as they found that the intervention they are recommending leads to modest short-term weight loss with “small to no impact on health, behavioral, or intermediate cardiometabolic outcomes.”

This is not nearly enough data or time to capture the chance of weight regain and/or physical and psychological harms – both of which the contextualizing information in the evidence synthesis notes are a distinct possibility.

The recommendation claims, “The USPSTF found adequate evidence to bound the harms of comprehensive, intensive behavioral interventions in children and adolescents as no greater than small, based on the absence of reported harms in the evidence and the noninvasive nature of the interventions.”

Yet they discuss issues with weight regain and harm in section CQ3 of the evidence synthesis. In terms of weight regain the evidence synthesis says, “evidence suggests that adolescents who diet (i.e., change the way they eat to lose weight) tend to have a higher BMI five years later, compared with adolescents who do not diet.” The evidence synthesis offers no research that suggests that adolescents who undergo the recommended intensive weight management interventions will fare any better.

They provide research that shows that harm comes from labeling the weight of children and adolescents as too high (which is a necessary part of sending them to intensive interventions meant to make them smaller.)  From the evidence synthesis: “Labeling children and adolescents as having overw*ight or ob*sity is associated with poorer psychosocial outcomes, higher rates of unhealthy weight control behaviors, and appears to have either no impact on weight or is associated with greater future weight gain. Weight labeling may undermine health-promoting behaviors associated with body satisfaction among youth with higher BMI, and children who misperceive their weight as being lower than it is may experience less future weight gain, fewer depressive symptoms, and improved blood pressure.”

The recommendations offer no research that the children and adolescents subjected to these interventions won’t experience high rates of weight regain and the harms outlined in CQ3. Rather they make their claim of lack of harm by ignoring this evidence and basing their findings on only the 50 studies they chose to include, only 18 of which even tracked harm and, again, only for about a year. So they are basing their claims around a lack of evidence, not around an actual lack of harm.

Given the complete absence of data showing long-term weight loss maintenance, and the extreme paucity of data around harms (and complete lack of data on harms after a year,) in the included 50 studies, the only way to protect children is to make the assumption that, based on the evidence in children and adolescents (and also in adults,) those subjected to these interventions will likely regain the weight that they lost and may well regain more than they lost, and will likely experience the harms outlined in CQ3.

Finally, they were only able to reach any recommendation at all through pooled analysis. Pooled analysis of studies with heterogeneous study designs, populations, and statistical models (as these 50 studies have) can lead to unreliable conclusions. This  is not sufficiently addressed.

Another viewpoint that should have been included is weight-neutral interventions. Instead of purposefully excluding studies that focused on improving health to the exclusion of weight loss, the USPSTF should have considered weight-neutral interventions. The upholding of short-term, heterogenous RCTs as adequate evidence but observational studies (often including longer-term data as inadequate continues to prop up a failed paradigm and healthcare inequalities in which funding is relatively easy to acquire for RCTs focused on weight loss (many of which are funded and/or conducted by the weight loss industry itself) and nearly impossible to acquire for RCTs of weight-neutral interventions.

Question 3:
How could the USPSTF make this draft Recommendation Statement clearer?

Based on the short duration, small sample size, and heterogeneity of the 50 included studies, the use of pooled analysis, and the concerns raised by the contextualizing evidence, the USPSTF should not make a recommendation for comprehensive, intensive behavioral interventions. At the absolute least, the recommendation should be down-graded significantly (I would suggest a D) rather than as a “B” as it is currently graded.

The statement could be much more clear that the research found that these “intensive” interventions requiring 26 or more appointments (and the time and money required to travel to and pay for them) resulted in weight loss of only 4.4 to 6.6 pounds and that in a 24-month study published after the review, the children in the weight loss group had returned to their original BMI at 24 months.

The statement could be much more clear that the recommendations are using weight loss as a proxy for health improvement despite not having sufficient evidence to actually show that weight loss will lead to health improvements and, in fact, that they found “small to no impact on health, behavioral, or intermediate cardiometabolic outcomes.”

They could make it much more clear that, as they found in the “Prevalence” section of the evidence synthesis, weight loss attempts have been steadily increasing since 1991 while average weights have steadily increased during that same time, and they have no data to suggest that increasing weight loss attempts through their recommendations won’t continue this same trend. I disagree fundamentally with the idea that higher-weight children and adolescents should be eradicated, but given that the USPSTF has gone to great lengths to make a recommendation to attempt just that, it matters that their research supports the likelihood that their recommendation will have the opposite of its intended effect.

Question 4:
What information, if any, did you expect to find in this draft Recommendation Statement that was not included?

I expected an evaluation of the evidence that exists for weight-neutral health-supporting behavioral interventions (focusing on supporting health directly and without a weight loss component).

I expected at least an acknowledgment of the lived experiences of people of all sizes who have discussed the harm that these interventions have caused in their lives.

I expected an analysis (or, at the very least, an acknowledgment) that the recommended intensive behavioral interventions are cost- and labor-intensive such that, if the USPSTF truly believes that they are a good idea (and, as stated above, I don’t believe the evidence they provided supports that,) the recommendation can’t help but increase health inequalities, including for those of lower socioeconomic status and/or for whom transporting a child/children to an appointment once a week for six months or once every other week for a year is more difficult or impossible (for example, a single parent working two jobs.) Especially given the fact that there is no long-term data to suggest that the (admittedly) “moderate” amount of weight lost will be maintained, that it will result in actual health improvements, or that it won’t create significant harm in the children and adolescents subjected to these interventions.

Question 5:
What resources or tools could the USPSTF provide that would make this Recommendation Statement more useful to you in its final form?

An informed consent/refusal form for healthcare providers and parents to use (preferably required) explaining that the research that exists suggests that, despite the significant cost and labor involved in the intervention, weight loss is likely to be small and temporary, and that the recommended interventions expose those subjected to them to significant harm and, finally, that there is research to support weight-neutral, health-supporting behaviors as having the same or greater efficacy with less harm.

Question 6:
The USPSTF is committed to understanding the needs and perspectives of the public it serves. Please share any experiences that you think could further inform the USPSTF on this draft Recommendation Statement.

IN YOUR OWN WORDS: In this section I recommend adding personal experiences of weight regain and harm (everything from damage to relationships with food, movement, and/or your body, disordered eating, eating disorders etc.) that you experienced after being subjected to weight loss interventions as a child and/or adolescent. I believe that this is very important, even if it’s the only question you answer. 

Question 7:
Do you have other comments on this draft Recommendation Statement?

Again, use your own words here, don’t just copy and paste! 

What has happened here is common, it’s harmful, and it has to stop happening.

Organizations that are supposed to be respected and evidence-based must stop issuing guidelines for higher-weight people (of all ages) based on short-term studies with questionable research methods. Higher-weight people deserve ethical, evidence-based medicine, not to be subjected to what amounts to, at best, experimental medicine and, at worst, interventions that have been shown to have the opposite of the intended effect the majority of the time.

If robust, long-term evidence for a weight loss intervention does not exist (and it does not,) then that intervention should not be recommended.

I don’t think that the people who are making these recommendations are doing so maliciously. I think that they are caught up in a paradigm that has, for decades, propped up weight loss as a healthcare intervention based on the sincerely held (but unsupported by evidence) belief that a small amount of short-term weight loss is worth any amount of money, labor, or risk extracted from higher-weight people.

It is my most sincere hope that the USPSTF will take this opportunity to change their recommendation and become a clarion voice for the physical and psychological well-being of higher-weight children, rather than just another part of the cacophony of voices encouraging the  risking of the lives and quality of life of higher-weight children to make them a few pounds lighter for a few months.

Again, You can add your own answers here, the deadline is January 16.

RESEARCH ANALYSIS

Ok, let’s get into a longer analysis of the research they are using to support this recommendation. I want to be clear that, while I’ve tried to be thorough and make a clear case that the evidence does not support the conclusions, this is not an exhaustive analysis. The synthesis is 381 pages long and explaining all of the issues would make this piece longer than that! The items in italics are quotes from the USPSTF.

Their stated objective:
“To examine the benefits and harms of weight management interventions in healthcare settings among children and adolescents with high body mass index (BMI).”

Note that their objective isn’t to find the best way to support health, they are starting from the perspective of studying weight management exclusively. In fact, they exclude any research that focuses on supporting the health of higher-weight kids rather than manipulating their body size.

I’m only going to delve into their analysis of the research around behavioral interventions since that is the only recommendation they made (“comprehensive intensive behavioral interventions” for higher-weight children ages 6 and up.)

Their recommendation is based on 50 trials of behavioral interventions with a total of 8,798 subjects. They admit that:
“Data on behavioral interventions were extremely limited beyond 12 to 13 months.”
“In the short term, weight management interventions led to [modest] lower weight in children and adolescents with…small to no impact on health, behavioral, or intermediate cardiometabolic outcomes.” “Maintenance of weight changes beyond one year are unknown.”

This is clearly not adequate data to make a recommendation for a cost- and labor-intensive intervention for about 20% of US children.

Terminology and Measures Section

They acknowledge the preference of people with lived experience to not use “ob*se” or “overw*ight,” which I appreciate, and they make some effort not to use those terms. Still the issue isn’t simply the use of the terms. The true problem is the pathologization of bodies based on size, and that’s exactly what they are doing and perpetuating.

They also say “Being comprised only of height and weight, BMI is a crude measure that does not account for different distributions in fat or fat-free mass.12 On the other hand, the convenience of BMI measurement makes it suitable for use in a variety of settings and BMI is the accepted clinical standard measure of excess fat in the United States13”

This is a perfect example of the problem here. BMI is not just a crude measure, it’s actually NOT a measure of health. (You can have two people of the exact same BMI with very different health statuses, and you can have two people with very different BMIs with the exact same health status.) The continued justification for its “convenience” and the tragedy that this mistake has been made for a long time and on a large scale should NOT override the fact that BMI is not fit for purpose as a method of classification for health/healthcare. The insistence and justification of its use just keeps kicking the same dented, broken can down the road, to the great detriment of people who find themselves on the “wrong” side of BMI-based classification.

Prevalence

“National Health and Nutrition Examination Survey (NHANES) data from 2017 to 2020 (pre-pandemic) show that 19.7 percent of children and adolescents in the United States have BMI ≥95th percentile for age and sex, based on the 2000 CDC growth charts.15”

Beyond the problem that there is a natural diversity of body sizes and there is no such thing as a “healthy weight” (since people of all weights find themselves at all points across the health spectrum,) as I mentioned in Part 1, the idea of judging children as being the “wrong” weight based on how they compare to the weights of children over twenty years ago does not have the ring of sound science. Also, in the comments in Part 1, Katherine Flegal (who was a reviewer for these guidelines) points out:

“The growth charts were intended as a way to monitor children's growth and the 95th percentile of the growth charts was not called "obesity"; rather it was called "overweight" and not considered as a disease but supposed to be just an indicator for a further assessment, serving as a screening level. Then without any real justification, a group formed by Bill Dietz then at CDC, decided to change the terminology without justification. So a screening level got transformed into a "disease." I and my colleague at CDC fought hard and long against this change. The decision went way up the chain and was eventually decided at a higher level, above CDC We lost. See Moynihan R. Expanding definitions of obesity may harm children. BMJ. 2006;332(7555):1412 This is available for free.”

[Editor’s note – Dr. William Dietz is a reviewer for these guidelines. He is also the director of the STOP Obesity Alliance, which is an astroturf organization similar to the OAC and TOS. Their main corporate funders are companies from the weight loss industry including Novo Nordisk (who, as we know, have faced sanctions in multiple countries for deceptive practices.) In 2021, corporate members included Novo Nordisk, WW, Currax, and Pfizer, who contributed a total of $120,000 of the organization’s total budget of $142,000. Novo Nordisk also gave them $200,000 in 2021 for “research project funding.”]

They go on:

“There are statistically significant differences in the prevalence of BMI ≥95th percentile by age, race and ethnicity, and socioeconomic status (Figure 1). The prevalence of high BMI increases with age, decreases with higher income, and is the highest in Hispanic/Latino and nonHispanic Black children and adolescents”

So the takeaway here is that they are recommending time-consuming, expensive interventions to about 20% of children, and disproportionately to children from low-income families and Black and Hispanic/Latinx children and adolescents. (Critically, while they say that they are concerned about health inequities, their recommendations do not take cost into account.)

“Overall, there has been an upward trend over time in the prevalence of BMI ≥95th percentile among children and adolescents in the United States, with some periodic variation in trends in particular age group…

Data from the Youth Risk Factor Behavior Surveillance System show that 48.3 percent of US-based 9 to 12th graders were trying to lose weight in 2019, with dramatically higher rates among females (59.8%) than males (37.0%).17 The proportion of high school students attempting to lose weight has increased since these data began being monitored in 1991”

So, let’s examine the situation. Weight loss attempts among children have steadily increased since 1991. The weight of children has steadily increased in that same time frame. So their solution is…more weight loss attempts? Seriously?

This is especially ridiculous given the body of research that shows that weight loss attempts in adults almost always lead to weight regain and often (up to 66% of the time in some studies) end in people regaining more weight than they lost, and the evidence that they provide (in the harms section of the evidence synthesis) says, “it has been estimated that at least two-thirds of adolescents restrict their food intake with the intent of losing weight, and the use of diet and exercise behaviors to lose weight has increased over time.44, 196 Such weight loss behaviors (whether unhealthy or not) are likely to be counterproductive, as evidence suggests that adolescents who diet (i.e., changed the way they eat to lose weight) tend to have higher BMI five years later, compared with adolescents who do not diet”

This one doesn’t even require an understanding of research methods or statistics, it’s an obvious failure in basic logic.

Association of Weight Loss and BMI With Health Outcomes

This is the section where they try to explain why the lack of proof that weight loss leads to improved health outcomes should be ignored.

“There are no randomized controlled trials (RCTs) of weight loss interventions in children and adolescents with sufficient followup length to provide direct evidence about whether weight loss is potentially associated with outcomes such as cardiovascular disease (CVD) events or mortality. In the absence of evidence in children and adolescents, studies in adults may inform the evidence landscape. However, evidence about the association between weight loss and health outcomes from RCTs in adults are inconclusive and limited by short followup and low event rates.21, 22 Given the lack of robust RCT evidence in either pediatric or adult populations, observational literature has been used to assess the relationship between BMI and health outcomes more broadly.”

Translation: there isn’t good data to suggest that weight loss will improve health for the children and adolescents for whom we are recommending weight loss. We would use adult data, except there isn’t good data for adults either. So we’re just going to use really questionable data. Best of luck kids!

“In the adult literature, evidence suggests that lifestyle behaviors and cardiorespiratory fitness play an important role in the association between BMI and mortality.32, 33 When these factors are taken into account, behavior appears to modify the association and vastly mitigates or eliminates the role of higher BMI in findings of higher mortality risk. Unfortunately, all of these data are from studies of associations, so cannot demonstrate that weight loss will results in similar effects”

Here they admit that the literature supports weight-neutral interventions with no reason to believe that weight loss would offer the benefits that weight-neutral interventions show. And yet they refused to study weight-neutral interventions in kids and are recommending weight loss?

It feels like this would be a helpful time for a reminder that their recommendation could have been that they didn’t recommend weight loss attempts given the lack of data on its impacts (positive or negative) on health.

Labeling children and adolescents as having overweight or obesity is associated with poorer psychosocial outcomes, higher rates of unhealthy weight control behaviors, and appears to have either no impact on weight or is associated with greater future weight gain, at least in the context of school and the home.34-44 Weight labeling may undermine health-promoting behaviors associated with body satisfaction among youth with higher BMI, as under-perception of weight is associated with positive outcomes such as lower future weight gain, fewer depressive symptoms, and improved blood pressure

Wait, what? They admit that labeling kids’ weight as too high creates negative health and psychological impacts (not to mention being associated with greater future weight gain which is the opposite of their intent.) They admit that higher-weight kids who don’t see their weight as too high have fewer depressive symptoms and improved blood pressure.

Then they recommend… labeling kids’ weight as too high?  Again I say, seriously?

Etiology and Risk Factors

They go through a number of factors that they claim may be “risk factors” for kids being higher-weight. Then say:

“The evidence to support many of these risk factors, however, is almost entirely observational in nature, and cannot establish causality. The role of the microbiome in high BMI is an active area of research and may have the potential to uncover some causal mechanisms.”

So they admit they don’t know why some kids are higher-weight. They also fail to list previous weight loss attempts as a “risk factor” for being higher weight (despite having pointed out how weight loss attempts predict higher BMI) and, perhaps because this report is fully committed to pathologizing higher-weight bodies, they do not explore the idea that there is a natural diversity of BMIs similar to the diversity we see in height, hand and foot size, etc., choosing instead to double down on pathologizing body size by looking for “risk factors” for these kids existing based on the belief that they shouldn’t exist.

Prevalence of Potentially Weight-Related Behaviors in Children and Adolescents

They begin by saying, “Data suggest that children in the United States are generally not meeting recommendations for healthy diet and activity behaviors.” They go on to discuss diet, physical activity, sleep, and screen time.

Despite the fact that the data they share applies to children of all sizes, they go on to discuss weight loss recommendations (by organizations that either directly profit from weight loss interventions or are funded by companies that do.)

This is another clear example of the issues with the weight loss paradigm on which these recommendations are based. Children of all sizes are not meeting these recommendations, but somehow that justifies expensive and dangerous interventions for kids who happen to be larger?

Previous USPSTF Recommendation

“In 2017, the USPSTF concluded with moderate certainty that screening for obesity in children and adolescents 6 years and older and offering or referring them to comprehensive, intensive behavioral interventions to promote improvements in weight status is of moderate net benefit (B recommendation).1”

So, they recommended this same thing in 2017 and between then and now the number of higher-weight kids has increased. So they are recommending it again? Because…why?

Study Selection

This section outlines what studies were included and why. It gives some hints as to the issues with the studies they used to support their recommendations.

“The inclusion criteria for individual studies most commonly were children with BMI at or above the 85th or 95th percentile for age and sex; however, other definitions of higher BMI were also eligible, such as those based on the percent of ideal weight, or with cut-offs other than the 85th or 95th percentiles. We did require overall cut-offs to be no lower than the 75th percentile.”

So there was no standardization of what even constituted “High BMI” in these studies, but their recommendation was for kids 6 and older who are “≥95th percentile for age and sex” (based on a comparison to the weight of kids over 20 years ago.)

“Studies included for KQ1 to KQ3 had to report weight or adiposity change at least 6 months following the start of the intervention to be included”

This is important to understand. Often these studies will be referred to as having “6 months of follow-up” which can create the (mis)impression that there was an intervention and then a follow-up period. Rather, they included studies that only offered 6 total months from the start of the intervention to the end of the study. Given the body of research in adults that shows that the vast majority of people regain weight after year 1, and the research they point to that shows weight loss attempts predict higher BMIs in 5 years, this is obviously not sufficient time to determine long-term outcomes in terms of weight regain (or any other harms.)

“Studies where followup for all outcomes was below 60 percent were excluded for quality as described further below.”

With a reminder that we are talking about short-term studies, they were fine with studies where 40% of the study population dropped out, even though dropouts in these kinds of studies can indicate that following the intervention was impossible, harmful, and/or that weight was either not being lost or they were regaining already.

“We also included harms associated with labeling; stigma or increased body image concerns or negative mental health effects; negative impacts on provider-patient relationship (e.g., care avoidance or dissatisfaction with care); unhealthy weight management efforts (e.g., using laxatives or self-induced vomiting) or eating patterns (excessive fasting, overly restrictive eating, or binging); suppressed growth; exercise-induced injury; other serious treatment-related harms at any time after initiation of intervention (i.e., death or medical issue requiring hospitalization or urgent medical treatment) or other treatment-related harms reported in trials meeting inclusion criteria for intermediate or health outcomes. For self-esteem and body satisfaction outcomes, we acknowledge these are problematic because improvement would represent a benefit and decline would represent a harm. We captured these under KQ4 for consistency but acknowledge this complexity.”

It's important to first note that when discussing harms, the people conducting the studies (who will often personally benefit financially from the recommendation of the weight loss intervention they are studying) are the ones tasked with identifying harms. As you might imagine, in my experience they often either don’t try, or do an incredibly poor job of tracking these harms. And, again, even if they really gave it the old college try, we are talking about a year or less of tracking here.

I’ll also point out that “unhealthy weight management” is a made-up term that has been used to try to obfuscate/erase the harms of these weight management interventions. The definition of “unhealthy weight management” often has significant overlap with behaviors recommended in “intensive” weight management interventions. For example, the use of “excessive” to modify “fasting” and “overly” to modify “restrictive eating” without any clear definition gives the people creating the interventions wide latitude to decide what an “acceptable” amount of fasting or restrictive eating is. Further, things like “laxative effect” and “vomiting” are often known side effects (sometimes actually considered benefits) of interventions like pharmacotherapy and weight loss surgeries.

“Hours of contact were estimated based on number of planned treatment sessions and the length of each session, or average contact time if that was reported…

So they didn’t have clear data for their weight recommendation and, here again, their intervention recommendation is based on estimates and pooled data.

Again, they had the option of saying that there wasn’t sufficient data to recommend subjecting children and adolescents to these interventions.

Our primary analyses present studies stratified by whether the intervention involved fewer than 26 contact hours versus 26 or more contact hours. This cut-point was based both on the logic that this corresponds to weekly one-hour meetings for six months, and because the current USPSTF recommendation recommends referral to weight management interventions that include at least 26 hours of contact.

So, their recommendation that kids and adolescents (and their parents!) engage in expensive, time-consuming interventions is based on the convenience of round numbers and a previous recommendation that was also not based on robust data? C’mon.

Estimated hours of contact in the first 12 months only are shown on the forest plots because the primary outcome was weight change at 12 months (or closest followup available).

I don’t mean to sound like a broken record, but 12 months is not remotely long enough to track long-term weight regain or harm. Subjecting kids to these interventions without knowing the likelihood of long-term weight regain or harm (including disordered eating, eating disorders, negative health and psychological impacts – either of the diets themselves or the weight regain) for a little bit of (very possibly short-term) weight loss is neither ethical nor responsible.

Description of the Included Studies

It should be very clear that the 50 studies that they included had significant differences (“heterogeneity” in research methods language.) It might be misunderstood that the 50 studies all apply to the entire population for whom the recommendations are being made (i.e., kids ages 6 or older with a BMI ≥95th percentile compared to kids from over 20 years ago.) In fact, many of the studies included only a small part of this group. For example just in terms of age they explain:

“Eighteen (36%) of the trials primarily included elementary-aged children, generally from age 6 to 8 years up to age 12 years, and an additional 13 (26%) included a wider range of ages covering elementary-aged and either preschool-aged children or adolescents. Twelve trials targeted adolescents and seven targeted preschool- to kindergarten-aged children. Three trials were limited to girls (all were studies of adolescents) and the remaining trials included both boys and girls (percent female across all trials, 57.2%).”

When it came to the actual components of the interventions there were, again, significant differences in each study (The 50 studies included 62 active intervention arms.) Recall that the only recommendation they made was more than 26 visits. We’ve already learned that they based that on the idea of a round number (1 visit per week for 6 months,) it’s also because it’s basically the only thing that the studies had (mostly) in common when they pooled the data.

They explain some of the vast differences between them (content note for inclusion of specific diet behaviors):

“Forty-seven of the 50 included trials provided at least dietary counseling and some information about behavior change principles. Most of these trials also provided counseling regarding physical activity or sedentary behavior. Two trials used interpersonal therapy as the primary treatment approach, linking overeating and loss-of-control eating to interpersonal functioning with apparently little to no counseling or education about eating a healthy diet and increasing physical activity.  Another study also did not appear to provide content on healthy diet or physical activity, but instead examined a “regulation of cues” intervention based on appetite awareness and handling cues that trigger the desire to eat other than hunger. Another trial tested the impact of a calibrated dinner plate and breakfast bowl to aid in portion control. Most trials focused on initial weight management, but one trial addressed longer-term weight maintenance after participation in a weight management program.”

The studies failed to even include a standardized participation measure. Some used percent of sessions attended, others included the percent who attended “some threshold” number of sessions. Some didn’t report participation at all. And even where it was reported the numbers, again, varied widely.

“Where reported, the average percent of sessions completed generally ranged from the mid-60s to low 80s. The percentage of participants who attended all planned sessions ranged from 31 to 93 in nine trials that reported this outcome”

I don’t want anyone to be misled into thinking that 50 studies of an intervention showed (at least short-term) weight loss. Rather 50 studies of varying interventions showed varying degrees of short-term weight loss.

They used a rating scale of good, fair, and poor. Poor studies were excluded. Of the 50 studies included, only 9 managed a “good” rating, with 41 (or 82%) being only “fair.” They explain:

“Among the fair-quality trials, three reported generally good methods but were downgraded to fair because attrition was greater than 15 percent.113, 114, 136 More typically, there was more than one concern if studies received a fair rating. Aside from attrition, common concerns included unblinded outcomes assessment, or failure to report one or more of: allocation concealment, randomization methods, outcomes assessment blinding, information about intervention fidelity, or patient adherence or attendance”

So 41 had a fair rating, of those 38 (or 76% of the studies) had “more than one concern.” Yikes.

Another huge issue is the sample size of the individual studies. While all together the 50 studies included a total of 8,798 subjects, they explain, “Many trials had small sample sizes; the median sample size was 108 (IQR, 70 to 206), equating to approximately 50 participants per study group”

So, here again, it’s not like 8,798 participants received an intervention and these guidelines are based on that. Rather, small numbers of participants received varied interventions and they’ve cobbled together the results to create a recommendation based on pooled data.

When I was a student learning research methods and statistics, this right here is the kind of thing that taught us never to do.

Are There Harms Associated With Weight Management Interventions for Children and Adolescents?

Of the 50 studies, only 18 (36%) actually reported any potential harms, and none of those “found an increase in the risk of any adverse event or serious adverse events, or decreases in self-esteem, body satisfaction, or disordered eating after 6 to 12 months.” Even if we ignore the fact that these studies often do a very poor job at tracking harm and accept this premise, let’s remember that in studies of adults, the vast majority of people start regaining the weight they lost after 12 months so these 6-12 months studies are still looking at the “honeymoon” period where participants are being “successful” at weight loss and are likely being rewarded (socially, familially, and even by their trusted healthcare provider) for more closely approximating the current cultural stereotype of beauty. Absent longer-term research, this is not a comforting number, even without the fact that almost two-thirds of the included studies didn’t even bother to try to find out about harms, and no data about harm was available past 12 months.

They claim that, “Two trials of interpersonal therapy with limited counseling to change diet and physical activity found reductions in disordered eating, suggesting a benefit rather than a harm of behavioral interventions.“

This is much more complicated than this one-sentence treatment makes it appear. I wrote about this here.

To review, their claim that, “The USPSTF found adequate evidence to bound the harms of comprehensive, intensive behavioral interventions in children and adolescents as no greater than small, based on the absence of reported harms in the evidence and the noninvasive nature of the interventions.” is based only on this small amount of shoddy data and ignores the research they themselves found about the harms.

I also want to point out that within online fat liberation spaces, adults talk a lot about their recent realizations of the harm that weight loss interventions, to which they were subjected as children, have done to them. It is impossible to overstate that 12 months of data from 18 studies is not NEARLY enough to make any claims about harm.

Association of Labeling With Psychosocial Well-Being

Here they list more harms associated with the recommendation that they are making. Including that weight-based labeling “is associated with children’s physical self-perception and emotional wellbeing. For example, children labeled as having overw*ight* or ob*sity may feel less smart, less physically capable, and less positive about themselves in general.41, 42 Children report that weight-based labels contribute to feelings of embarrassment and sadness, and youth who perceive themselves to be overweight are more likely to experience depressive symptoms than children who do not perceive themselves to be overweight, even when controlling for actual BMI.43, 44 Internalization of thin-body ideal and body dissatisfaction are both associated with an increased risk of onset of binge- and purge-spectrum eating disorder”

And yet they claim that their intervention (which includes labeling kids as too heavy) has a chance of harm that is “no greater than small.” If you’re confused about how they are claiming little to no harm in the recommendation, despite continuously listing harms in the evidence synthesis, I completely understand. It’s because their recommendation is based only on the 50 (small, short-term) studies that their evidence review included. If you think it’s dangerous and irresponsible to recommend an intervention to children 6 and older when there are known correlations to harm, just because 50 small, short-term studies either didn’t track harm or failed to find it, I agree with you completely. 

What Level of Weight or Relative Weight Decrease in Children or Adolescents With Obesity Reduces the Likelihood of Obesity in Adulthood or Health Outcomes Associated With Obesity?

They start by pointing out that evidence of how much weight loss would actually create health benefits (or prevent higher-weight kids from becoming higher-weight adults – which is a problematic goal outcome based in weight stigma) is “limited.” They point out that the “most robust” evidence suggests a zBMI reduction of at least 0.7. Then they point out that the reductions in the trials that they reviewed ranged from 0.2 to 0.4 (far less than “at least 0.7.)

They also point out that the evidence that claims benefits from weight change doesn’t even attempt to distinguish the benefits from behavior changes from that of body size changes, so they don’t even know if weight loss confers any health benefits at all (which is also called into question by research that shows health benefits from behavior changes regardless of body size or body size changes.)

So why are they recommending expensive and time-consuming weight loss interventions that only produce about half as much weight loss as is required to maybe not create any health benefit?  

After previously explaining that they don’t actually know what amount of adiposity might impact health (or, honestly, if it does at all) they posit:

“simply arresting the gain in excess weight may constitute a clinically important benefit for many of these interventions, assuming the changes in weight trajectory are maintained in the long term.”

Here again we have a lowering of the bar for weight loss interventions to including…checks notes…not producing much weight loss. They reinvest in pathologizing body size, and recommending interventions for body size reduction even though they don’t reduce body size, because they might stop higher-weight people from becoming higher-weight adults, though there aren’t remotely adequate controls for whether the kids who don’t become higher-weight at the same pace were never going to do that in the first place, whereas the kids who do become higher weight as they get older were always going to do that.

They spend some time discussing the idea of 5-10% weight loss in adults creating health benefits, without adequately discussing the fact that this number was arrived at by attrition and fails to separate the impact of behaviors from body size change, and they don’t mention that research has shown that, in fact, the weight loss isn’t associated with the health changes.

They do admit that, “We found limited direct evidence to justify any specific threshold for percent weight reduction or percent BMI reduction for children or adolescents.”

CQ5. What Are the Inequities in Factors That Support Healthy Eating and Physical Activity in Youth (e.g., Food Insecurity, Financial Security, or Neighborhood-Level Factors)?

Their summary here states:

“Diet and physical activity are heavily influenced by the local community environment and by family-level economics, which are also shaped by larger structural forces that have systematically disadvantaged communities that have experienced discrimination, such Black, Hispanic/Latino, and Native American communities.213, 214,215 For example, neighborhood environment has an impact on physical activity.213 Further, financial insecurity is an important driver of dietary behaviors; families with limited financial resources must spend a higher proportion of their household income on food, and insufficient food budgets can drive families to prioritize cost-effectiveness over healthfulness in order to help reduce financial strain.214 Additionally, stress, uncertainty, and long work hours for parents constrain their ability to prepare and serve healthy foods, and are associated with haphazard meal planning, emotional eating, and snacking on sweets among adults with higher BMI.213”

Note that these effects happen to people of all sizes. And here we have another problematic use of data around the impacts of things like racism and other systemic oppression, and make the focus whether or not it might make those who are subjected to it higher-weight. This distracts from larger conversations including the health of people of all sizes in these communities.

For example, they say, “Encouraging families with higher-weight children to change their diet and physical activity forms the foundation of most weight management interventions.” Then go on to say how these structural inequities can make this difficult.

Previously they have made clear that they don’t have any evidence to suggest that these interventions will create anything more than small, short-term weight loss and that, in fact, adolescents who participate in them are likely to have higher BMIs in 5 years. While, again, I don’t think there is anything wrong with higher-weight people existing, I do think there is something wrong with an intervention that is likely to have the opposite of the intended effect. So the idea that the issue is that inequalities create issues with weight loss interventions becomes a red herring (since weight loss interventions by and large don’t work under any circumstances) distracting from important conversations about the ways that oppression harms marginalized people of all sizes.

CQ6. What Are the Inequities in Access to or Participation in Weight Management Interventions?

Here they point out that, “One important source of inequity in access to and participation in professionally supported weight management interventions is inequities in access to healthcare in general”

This is true, and focusing access to healthcare on expensive, time-consuming weight loss interventions which likely produce only small amounts of short-term weight loss will actually exacerbate this problem as well as create delays in actual evidence-based care while higher-weight kids are sent away for these “intensive” weight loss interventions for six months to a year. So here again these recommendations could very well increase healthcare inequities.

What Are the Harms of Stigma and Weight Bias? What Is the Extent of Weight Bias in Healthcare?

This is a decent section, but it fails to adequately address that language and action around pathologization, prevention, and eradication of higher-weight people (all of which their recommendations are based in, and perpetuate) constitute weight stigma.

In the section about the harms of weight stigma in healthcare they, somewhat extraordinarily, admit:

“the belief that voluntarily improving diet and/or increasing exercise can entirely prevent or reverse ob*sity is not well supported by biological or clinical evidence”

And yet they are recommending attempting to voluntarily improve diet and/or increasing exercise to prevent or reverse being higher-weight?

C1Q8. Are There Long-Term Harms of Weight Loss Attempts During Childhood or Adolescence?

Summary:

“Studies show that weight loss is typically followed by weight gain. This is consistent with physiologic adaptations in hormonal signaling and energy expenditure in response to weight loss and calorie restriction that promote weight regain.”

Then why in the world are you recommending weight loss?

In their discussion of the harms of weight cycling (for which most of the research is around adults,) which is, by far, the most common outcome of weight loss interventions, they hyper-focus on the idea that it might create weight gain (which continues to perpetuate the weight loss paradigm) but fail to discuss things like increased inflammation, cardiovascular disease, overall mortality et al. (I wrote about this here.)

“The RCTs of weight interventions included in our systematic review were of short duration, and therefore could not report on potential longer-term harms such as hormonal or metabolic adaptations that may promote weight regain, disordered eating tendencies, or psychological harms that appeared in the longer term or after weight re-gain. Therefore, we sought information on longer-term harms from the broader literature.”

It's good that they sought that information, but that doesn’t do any good if (as happened here) that information is ignored in the final recommendation.

They also point out that “We found no qualitative studies of adults reflecting on their experiences with weight management programs during their childhood or adolescence, but anecdotal first-person accounts among people with lived experience suggest that weight loss interventions during childhood, in the context of the larger culture of weight stigma, were damaging to their psychosocial health.”

This is an important point and I’m glad that they made it. It’s also a tragedy that there aren’t studies of this, and they should absolutely be done. (It’s not surprising to me that they don’t exist, and I would point out that a possible confounder here would be that the weight loss industry has worked so hard to convince us to blame ourselves, not just for the near-total lack of long-term success of their interventions, but for any harm that comes from them.)

Chapter 4: Summary of Evidence’

Alright, let’s bring this home.

“The evidence in this review demonstrated that structured behavioral weight management interventions in children and adolescents typically resulted in modest reductions in weight (~2-3 kg) compared to control groups for up to one year, particularly interventions with at least 26 hours of contact and that included physical activity sessions (Table 23). The clinical significance of this amount of weight loss is unclear.”

So, they are recommending that about 20% of kids over 6 and their parents find the money and time for 26 hours of healthcare interventions in order to lose 4.4 to 6.6 pounds. And they don’t know if that will have any impact on their health. That seems absolutely ludicrous to me.

“A fairly large (n=452) trial of a 26-session family-based behavioral intervention published after our search was completed supports our finding that behavioral interventions are associated with small weight changes.98 In this study, group differences were maintained for up to 24 months with continued intervention; children in the intervention group had returned to baseline BMI percentile after 24 months while those in the usual care group had a 6.5 percent increase their percent above median BMI. There was a larger effect among White children than among Black children in this study. The impact on quality of life, cardiometabolic parameters or longer-term followup were not reported”

And here it is. The results DO NOT “support our finding that behavioral interventions are associated with small weight changes.” In fact, what they found was that after 24 months, the kids who were put on the type of intervention that they recommended, based on 12 months of data, had returned to their original BMI! While they point out that the group who weren’t given the intervention had a slight increase in their BMI (note that it’s a 6.5% increase in their percent above median, not a 6.5% increase in their BMI) it’s not clear if that’s just a normal growth trajectory (which has been interrupted in the weight loss group) and/or that if the weight loss group will continue to gain weight more rapidly than they otherwise would have as a result of the weight loss intervention.

I cannot state clearly or loudly enough that I do not believe that this research remotely supports the recommendation of intensive weight loss interventions in any population. This recommendation puts kids at risk of significant harm with literally no evidence of long-term benefits.

Again, I don’t think that the people who are making these recommendations are doing so maliciously. I think that they are caught up in a paradigm that has, for decades, propped up the idea of weight loss based on the sincerely held (but unsupported by evidence) belief that a small amount of short-term weight loss is worth any amount of money, labor, or risk. This has done, and will continue to do unfathomable harm, and I deeply hope that the USPSTF will take this opportunity to change their recommendation to become a clarion voice for the physical and psychological well-being of higher-weight children, rather than another part of the cacophony of voices encouraging the risking of the lives and quality of life of higher-weight children in an ill-advised effort to eradicate them from existence.

Did you find this post helpful? You can subscribe for free to get future posts delivered direct to your inbox, or choose a paid subscription to support the newsletter (and the work that goes into it!) and get special benefits! Click the Subscribe button below for details:

Please consider sharing this piece to help get the word out:

More research and resources:
https://haeshealthsheets.com/resources/

*Note on language: I use “fat” as a neutral descriptor as used by the fat activist community, I use “ob*se” and “overw*ight” to acknowledge that these are terms that were created to medicalize and pathologize fat bodies, with roots in racism and specifically anti-Blackness. Please read Sabrina Strings’ Fearing the Black Body – the Racial Origins of Fat Phobia and Da’Shaun Harrison’s Belly of the Beast: The Politics of Anti-Fatness as Anti-Blackness for more on this.