When it comes to associations, causality matters – the importance of which is easily understood when considering childhood obesity and inactivity. Whether inactivity leads to kids to gain weight, or whether weight leads kids to become inactive have very different implications
My very publicly expressed bias is that childhood obesity drives inactivity, and there’s data to support that assertion including this study whereby when observed over time in 8-11 year olds, weight gain predicted inactivity, while inactivity did not predict weight gain, leading the authors of that study to conclude,
“adiposity is a better predictor of PA and sedentary behavior changes than the other way around.”
As to what’s going on, if I were to venture a guess, it’d be some combination of obesity related abject bullying (multiple studies demonstrate bullying is associated with decreased physical activity and obesity has been shown to be the number one target of schoolyard bullying, not to mention the fact that bullies may make fun of heavier kids when exercising explicitly), being one of the worst/slowest on a team, and increased effort involved, that lead heavier kids to decreased MVPA (moderate-to-vigorous physical activity).
Now I’ve blogged before about the Coca-Cola funded ISCOLE trial almost entirely ignoring the possibility that childhood obesity drives inactivity rather than the other way around, and recently, another study came out of the ISCOLE group that did the same.
(and if you’re interested, here’s a published discussion of the emails between ISCOLE investigators and Coca-Cola that not surprisingly suggests that these relationships have the very real potential to influence the framing of results even if funders not involved in study design)
The study, Joint associations between weekday and weekend physical activity or sedentary time and childhood obesity, published in the International Journal of Obesity, looked at weekday and weekend levels of MVPA and sedentary time in 9-11 year old children in 12 countries and their associations with obesity.
Yes, they were found to be associated.
Though they do have a single throwaway line speaking to causality,
“It is not known whether lower levels of physical activity are the cause or the consequence of obesity”,
that did not stop them from writing this as their final line and conclusion,
“Since children have more discretionary time during weekend days than weekdays, children should be encouraged to increase physical activity during weekend days, especially a high level of MVPA“
which while true on the basis of health as a whole, when framed in the explicit context of childhood obesity as it is in this paper, seems to deny the much more likely case that weight simply slows kids down.
And honestly, that matters.
It matters because one of the most predominant stereotypical narratives of obesity is that it is a disease of laziness and ISCOLE’s recurrently utilized framing supports that stigmatizing message. It also matters because, if considered at least by the parents I regularly see in my office, inactive children with obesity are being regularly judged (and sometimes even shamed or bullied) by their well-intentioned parents for not being active enough to lead them to lighter weights and I can’t help but wonder if this would still be the case if the meat of the discussions in these sorts of papers focused on the barriers to physical activity erected by childhood obesity itself?
I would love to see more research done on the various mechanisms by which childhood obesity might contributes to inactivity, and more longitudinal studies designed to test causality, rather than paper after paper with conclusions that to my confirmation bias at least, ignore the many reasons why kids with obesity are understandably less likely to be physically active, and in so doing, fail those poor kids.
I’ve long called for a rebranding of exercise to promote it on the basis of all of its incredible benefits, and explicitly not in the name of weight loss.
While on paper there’s no doubt that people can lose weight through exercise (and in research studies too), in practice they generally don’t. And though there’s also definitely the suggestion that exercise helps to keep weight off (or serves as a marker or inspiration for maintaining a whole slew of weight responsive behaviour changes), when it comes to public health, I believe focusing on weight loss as the outcome of choice in exercise interventions risks those interventions’ dissolution when weight loss doesn’t occur.
Helping to make my point is a recent study published in the American Journal of Preventative Medicine. The study, Implementing School-Based Policies to Prevent Obesity: Cluster Randomized Trial, looked at the impact school based nutrition and physical activity policies had on weight.
The study found that while school based nutrition policies seemed to have an impact on weight over time, school based physical activity policies didn’t.
Not measured of course, or at least not mentioned, were the impacts those physical activity policies might have had on other health related parameters (blood pressure, blood sugar, non-alcoholic fatty liver disease, mood, sleep, attention, learning, physical literacy, and more) – things that I think the literature would support as being far more likely to see exercise-related improvements.
But it’s the study’s abstract’s conclusion that got me, as I think it does a great job of highlighting the risk of clinging to exercise as an important driver of weight loss. Here it is in its entirety (highlighting mine),
“This cluster randomized trial demonstrated effectiveness of providing support for implementation of school-based nutrition policies, but not physical activity policies, to limit BMI increases among middle school students. Results can guide future school interventions.”
Suffice to say I think it’d be an incredible shame if results like these guide any future school related physical activity interventions, as the benefits of exercise are myriad, something these results wholly ignore, and if these results guide anything, they’d guide the avoidance or elimination of school based physical activity policies which would let kids down on so many levels.
Last week saw the publication of a new study in the BJSM entitled (highlighting mine), “Is interval training the magic bullet for fat loss? A systematic review and meta-analysis comparing moderate-intensity continuous training with high-intensity interval training (HIIT)“. Understandably intrigued given a prominent medical journal was suggesting there was a magic bullet for fat loss, I clicked through, and then reading the piece I learned that the amount of fat lost that the BJSM was calling a “magic bullet” was a 1 pound difference, one which the study’s abstract’s conclusion described as, “a 28.5% greater reductions in total absolute fat mass (kg)”. Duly surprised, I then took to Twitter to poke around and found that one of the study’s authors, James Steele, was tweeting out a corrective thread to his own study’s hype – hype which understandably and predictably led to an onslaught of media overreach. Intrigued, I approached him directly to ask about the discordance in tone between his tweets and his study’s title and conclusion, and he sent me such a thorough and thoughtful response (explaining how it was the BJSM’s editor who’d changed both), that I asked him if he’d mind my sharing his thoughts here as a guest post. Suffice to say, in my opinion, medical journals and their editors shouldn’t be in the business of clickbait hype, as it diminishes themselves, research, and furthers societal scientific illiteracy by suggesting that such things as “magic bullets” for weight or fat loss can conceivably exist.
I was first slightly concerned that the findings would be overhyped and potentially misrepresented when I saw the press release that was sent to the media. I was forwarded various requests by our institutions news team and saw the wording of the first line of which was
“Short bursts of high intensity exercise are better for weight loss than longer sessions in the gym, research suggests.”
My colleague James Fisher noted to me that he also thought the press release didn’t reflect the findings accurately and wondered whether the title change resulted in the perception of a different finding.
The original title in our submission to the journal was
“Comparing the effects of interval training versus moderate-intensity continuous training on body adiposity: is it possible to find a signal in the noise? A systematic review and meta-analysis“
which was chosen as an homage to Nate Silvers’ book and the use of meta-analysis to find the ‘signal’ from among the ‘noise’ of conflicting findings in smaller studies The paper underwent peer review as normal and we made changes suggested by the reviewers to improve the manuscript; but, none of the reviewers commented on the title if I recall. After the reviewers were happy with the paper and had no further changes they wanted we received a recommendation that it be published, but with minor revisions which were suggested by the editor. Most of the revisions suggested where helpful as they seemed to be aimed at improving readability of the manuscript. However, it was also suggested that the title was changed, as well as the addition of the percentage difference to the conclusion of the abstract. This was suggested to be intended to attract more attention to the article, make it seem more compelling, and ensure recognition was received for the work. I didn’t particularly like the newly suggested title, nor did some of my co-authors, but it was not strictly saying that anything ‘was’ a ‘magic bullet’ and so I did not push the issue. I must confess I did not at the time notice the seemingly minor change to the abstract conclusion though. I personally dislike the presentation of % values in this manner as to me they are often misleading and detract from whether the absolute values are really meaningful or not (a big problem in sport and exercise IMO wherein a lot of studies make interventions seem better than they are by reporting % values). The value is not inaccurate, but it does lead the less wary reader to potentially draw the wrong conclusions.
I did suspect that the changes were suggested because the paper would likely be selected for a press release which turned out to be correct. I’m glad the paper got some wide coverage, but wanted to make sure it was covered in a nuanced manner. So I tweeted a little thread to try and provide some balance and when I was interviewed about it on BBC World Service I also made sure to provide as balanced a commentary as I could in the time permitted.
It doesn’t surprise me that the media initially interpreted things to be saying that ‘HIIT’ (high-intensity interval training) was better than ‘MOD’ (moderate-intensity continuous training) for fat loss without considering all the nuance… that’s just how it goes sadly. I also can empathize with the journal and publisher in wanting to try and increase the reach of the work that they publish. To my mind if we can widen the reach of good science, and raise appreciation of its importance, then that’s a good thing. This is something I’d like to be able to do more of. But, though this is good in principle, in execution it proves to be difficult. It’s tough to get the nuance across because science is hard and most people aren’t really able to understand it. I guess it’s part of the media cycle though. The wider media wants ‘stories’ and just regular boring old science doesn’t make for a good story. So to get the wider media’s attention journals and academic publishers need to try and make things seem more exciting. In that process though nuance gets lost. However, I can’t think of any other way to communicate science more widely at the moment. I guess what we need to ensure is that, once the media get hold of a story and want to run it, the actual scientists themselves are the ones they speak to and interview so they end up with a platform and captive audience to explain the nuance and implications in an understandable manner. At least, that’s what I’ve tried to do and hope I achieved.
I think if I was able to ‘do over’ this example specifically then I would have likely pushed back more on the issues. I would like to have kept the original title and would have argued for this position as I suspect my co-authors likely would have too. I definitely would have pushed back on the change to the abstract conclusion and will be more vigilant to these issues in future. In likelihood this might have meant the paper would have been less ‘impactful’ as a story for the media. But it would have meant that the paper itself didn’t contribute to any potentially misleading publicity. The publisher could have still put out the press release as they desired… Can’t stop them from doing that. But at least the paper would have better reflected what we found in full. I think I would advise authors who face similar situations to make sure they think and have a conversation about this. We all want our work to reach the widest audience to hopefully have the biggest impact possible. But we don’t in the process want it to distort in terms of its message. Make sure to discuss it with your co-authors and the journal/publisher and find the right balance so that scientific integrity is retained, whilst reach is maximized. It’s tough to do, but worth striving for IMO.
Dr James Steele is the Principal Investigator at the ukactive Research Institute, and Associate Professor of Sport and Exercise Science at Solent University. James completed his BSc (Hons) in Applied Sport Science in 2010, and hid PhD examining the role of lumbar extensor resistance training in chronic low back pain in 2014. He has extensive experience of research and applied consultancy in the area of physical activity, exercise, and sport from over the past decade, working with a wide range of populations ranging from elite athletes across a range of sports, to the general population across the lifespan, and both those who are healthy and diseased. James has published numerous peer-reviewed articles and delivered several invited talks at international conferences on a variety of areas relating to sport, strength and conditioning, physical activity and exercise, health and fitness. He was appointed to the Expert Working Group revising the Chief Medical Officers Physical Activity Guidelines for the UK and is a Founding Member of the Strength and Conditioning Society, as well as the Society for Transparency, Openness, and Replication in Kinesiology, and member of both the British Association of Sport and Exercise Sciences and the American College of Sports Medicine.