Exercise and weight loss: long-term follow-up

This post is by Phil Price, not Andrew.

Waaaay back in 2010, I wrote a blog entry entitled “Exercise and Weight Loss.” I had added high-intensity interval training back into my exercise regime, and had lost 12 pounds in about 12 weeks; but around the same time, some highly publicized studies were released that claimed that exercise does not lead to weight loss in overweight people. I suggested that that claim was too strong: at best they had demonstrated that moderate-intensity exercise does not lead to weight loss in most overweight people. I am completely convinced that when I am slightly overweight, I lose weight when I do occasional high-intensity workouts.

Well, I’m back with another data point. After spending a month in hell earlier this year, during which I got no exercise, I had not only failed to lose my winter weight but had added a few pounds. When I was finally able to get to my usual spring activities, which include road biking — sometimes with high-intensity intervals — I quickly lost a couple of pounds. But then I crashed, nothing serious but enough to keep me off the bike and mostly sedentary for more than a month, and I put on some more weight, topping out at about 203 or 204 pounds, the heaviest I had been since I wrote that “Exercise and weight loss” blog post back in 2010. Already this experience would seem to contradict the suggestion that exercise doesn’t control weight: if I wasn’t gaining weight due to lack of exercise, why was I gaining it?

I was able to resume exercise in early May, and in the next six weeks I lost about six pounds. In the past few weeks I’ve lost a few more. Yesterday and today, I’ve weighed in at 193 pounds, ten pounds lighter than I was two months ago. Given past experience, I expect to remain somewhere in the 192- to 195-pound range until November, when I will start edging upwards.

So I’m reiterating the point of that eight-year-old blog post I mentioned at the top: maybe moderate-intensity exercise doesn’t lead to weight loss in most overweight people, but high-intensity exercise does lead to weight loss in me when I am somewhat overweight, and as long as I regularly do some high-intensity exercise I don’t tend to gain weight.

The broader point here is that I think researchers (and journalists) tend to over-generalize. If you do a test that subjects one group of people to one set of conditions, don’t assume the results will extend to a different set of people and/or a different set of conditions, even if the people and the conditions have some similarity to those used in the experiment. The differences can matter.

Reminder: this post is by Phil, not Andrew.

63 thoughts on “Exercise and weight loss: long-term follow-up

  1. Fitness studies in general are a prime example of fhe phenomenon Andrew complains about, where researchers throw away huge amounts of data because they’ve been taught only to test the p-value significance of differences in group averages. They should instead embrace variation and focus on trying to explain it. (Indeed, my understanding is that there almost always a couple of unlucky people in every sample who manage to gain weight even under laboratory conditions of caloric restriction and regular exercise.)

    Gretchen Reynolds of The NY Times wrote an excellent book on fitness studies yet still writes terrible columns in the Health section hyping small-n studies with ambiguous results.

    • One of the things I like to do with diet/fitness studies is go through them and, if scatterplots or slopegraphs aren’t provided, note which measures have ranges or standard deviations implying that at least some experimental subjects got fatter/less fit during the experiment.

    • Fitness studies in general are a prime example of the phenomenon Andrew complains about, where researchers throw away huge amounts of data because they’ve been taught only to test the p-value significance of differences in group averages./blockquote

      I think this is a profound and powerful point. Andrew summarized this in a list of points that I can’t find right now.
      He should should post a link to that list on his homepage. (It would be handy to be able to cite it by name rather than explain it anew.)

      Newspapers pretty much automatically transform very weak results into grossly inaccurate generalizations. A study of diet x might honestly be described as showing that “on diet x, people lost from 18 to -14 pounds with the average weight loss across all dieters being 1.9 pounds. Taking into account random variation, approximately 11% of dieters lost weight because of the diet”. Reporters want to transform this into the headline that “people lost weight on diet x.” But that is false because not everyone lost weight on the diet.

      An even better example was the recent post about how “anti-immigration extremists [AIE] were less supportive of democracy”. The actual results were that 89 out of 100 AIEs were highly supportive of democracy while 93 out of 100 AIEs were highly supportive of democracy. I spent a lot of time thinking about how to accurately boil these results down into an honest summary and found it very difficult. You can’t say that “AIEs are less supportive of democracy” because most AIEs are just as supportive of democracy as everyone else. (Most summaries slide into saying something about all AIEs and so are false.)

      You need to say something about how there are slightly fewer AIEs who are strongly supportive of democracy without implying anything about the AIEs that are strongly supportive of democracy. But anything accurate you say sounds like very weak tea. The problem for reporters is that the results are, in reality, terribly weak and uninteresting, and reporters are trying to make a mountain out of a molehill (they want to slam all AIEs). In reality, if you talked to a 100 AIEs and a 100 non-AIEs, you would be hard-pressed to notice the difference in support for democracy.

      • Oops. I screwed up the blockquote.

        Fitness studies in general are a prime example of the phenomenon Andrew complains about, where researchers throw away huge amounts of data because they’ve been taught only to test the p-value significance of differences in group averages.

        I think this is a profound and powerful point. Andrew summarized this in a list of points that I can’t find right now.
        He should should post a link to that list on his homepage. (It would be handy to be able to cite it by name rather than explain it anew.)

        Newspapers pretty much automatically transform very weak results into grossly inaccurate generalizations. A study of diet x might honestly be described as showing that “on diet x, people lost from 18 to -14 pounds with the average weight loss across all dieters being 1.9 pounds. Taking into account random variation, approximately 11% of dieters lost weight because of the diet”. Reporters want to transform this into the headline that “people lost weight on diet x.” But that is false because not everyone lost weight on the diet.

        An even better example was the recent post about how “anti-immigration extremists [AIE] were less supportive of democracy”. The actual results were that 89 out of 100 AIEs were highly supportive of democracy while 93 out of 100 AIEs were highly supportive of democracy. I spent a lot of time thinking about how to accurately boil these results down into an honest summary and found it very difficult. You can’t say that “AIEs are less supportive of democracy” because most AIEs are just as supportive of democracy as everyone else. (Most summaries slide into saying something about all AIEs and so are false.)

        You need to say something about how there are slightly fewer AIEs who are strongly supportive of democracy without implying anything about the AIEs that are strongly supportive of democracy. But anything accurate you say sounds like very weak tea. The problem for reporters is that the results are, in reality, terribly weak and uninteresting, and reporters are trying to make a mountain out of a molehill (they want to slam all AIEs). In reality, if you talked to a 100 AIEs and a 100 non-AIEs, you would be hard-pressed to notice the difference in support for democracy.

  2. Agree based on identical experience. Furthermore, some interval training builds muscle, which weighs more than fat. So weight loss underestimates the beneficial transformation of tissue.

    As a cycling fan, I’m curious what intervals you do. I do these: https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&uact=8&ved=0ahUKEwjClJ-B85XcAhVrxYMKHXEAA9oQFggpMAA&url=http%3A%2F%2Fwww.lemondfitness.com%2Ftraining%2Fg-force-bikes%2F4-power-workouts&usg=AOvVaw30AoVdQMlfe4-A3N6vxG_W

    • Michael,
      The only time I do anything formal as far as intervals is when I’m doing a workout on my trainer (= stationary bike). I don’t have the inclination or discipline to do more than about an hour at a time on the trainer, and sometimes I do as little as half an hour, so I try to really make every minute of those trainer workouts count.

      I choose somewhat haphazardly from various ‘types’ of intervals. I have a feeling, not really backed up by data, that the classic ‘Tabata’ workout gives me the most bang per minute. As I do it, that’s 20 seconds at maximum effort, 10 seconds rest; repeat until I have done as many as eight, if I can. Max really means max: by the end of the third interval my heart rate is over 180 bpm (my max heart rate is about 185). By the end of the sixth interval, I feel like I’m going to die, and there is a nearly overwhelming tendency to cheat just a bit (take 12 seconds rest instead of 10; start the next interval with a 5-second ramp-up rather than going hard from the start; etc.). I rarely make it to 8 intervals, and frequently stop after 6. When I am not in great shape I can’t manage 6.

      I really have to psyche myself up to do the intervals described above. If I’m not quite feeling up to it, I’ll do multiple (3 to 5) 1-minute intervals with 30 seconds rest. Those 1-minute intervals are necessarily at substantially lower power (‘intensity’) than the 20-second ones I do in the Tabata workout. I try to gauge my intensity so that my heart rate hits 180 bpm at the end of each interval.

      But most of my riding is outdoors, not on the trainer, and I let the terrain and my mood dictate my intervals. For instance, I frequently climb a hill (“South Park”, in Berkeley’s Tilden Park, for those of you following along at home) that has four very steep sections of different length, separated by slightly less-steep sections. I will sometimes sprint the steep sections and take it relatively easy on the less-steep sections. In this case, I try to exert myself on each steep section so that my heart rate exceeds 180 at the end of the section. Basically I am unable (or at least unwilling) to maintain my heart rate above 180 bpm for very long, so I can’t do more than this, but if my heart rate doesn’t get to 180 in under a minute or so I don’t count it as a ‘high-intensity’ interval. In my mental accounting, any time I get my heart rate over 180 three times in less than 10 minutes, by doing efforts under 1 minute each, I count that as a ‘high-intensity interval workout’.

      No doubt I could improve my fitness more if I had a formal workout schedule, but I don’t have the motivation for that. The way I do it is good enough for me. If I ever decide to try racing, I suppose I’ll try to optimize.

      By the way: I’m 53 years old, so by the standard rules of thumb my maximum heart rate ‘should’ be much lower than it is: The rule “220 minus your age” would put it at 167, and “217 minus 0.8 * age” would yield 173. So this is another example of interpersonal variability being a big deal. My heart is slightly smaller than that of a median man of my height, and my maximum heart rate is much higher than that of a median man of my age. But I don’t care about the medians, I care about me!

  3. Andrew’s late friend Seth Roberts was the master of n=1 studies, including weight loss. He used self-experimentation to develop a diet called the Shangri-La diet. I’ve been following it for over ten years (it’s an easy diet to maintain) and have successfully lost about 15 pounds on it.

      • Seth had intended to write a book about self-experimentation, and wrote a chapter (maybe more) to show to potential publishers. The chapter he gave me had a bunch of examples of useful experiments that had been performed through the ages with n=1 or at least n very small. He had some examples from his own experience, too.

        Suppose I eat ice cream late one night, and sleep terribly that night. I think ‘I wonder if I slept badly because I ate ice cream just before bed?’ A month or two later I have the same experience. A few weeks after that, same thing. Of course I may have had some bad sleep experiences in the mean time, too…perhaps this is just a coincidence. Or maybe it’s a self-fulfilling prophecy: now that the idea is in my head that I won’t be able to sleep after eating ice cream, when I get in bed I am worried that I won’t be able to sleep, so I don’t.

        Still, if I repeatedly find that I can’t sleep after eating ice cream, then “I shouldn’t eat ice cream shortly before bedtime” is a good rule for me. Of course I haven’t ruled out other possibilities (e.g. maybe there’s something else special about the days that I choose to eat ice cream) but it would be silly to require or even promote the idea that you can’t learn anything without a formal experiment. People learn things all the time without formal experiments!

        • Of course it’s fine to learn from personal experiences, even in a uncontrolled, unblinded manner. I think everyone does that kind of thing, usually without even realizing. But then it should remain a private thing (or if communicated, taken with a grain of salt). Abstaining from ice cream helped you sleep better and drinking olive oil or whatever helped Seth Roberts lose weight, and that’s great for you and him, but it says very little (next to nothing) about what could help me or someone else sleep better or lose weight.

      • Adede:

        In all seriousness, I think that Seth fell for his own hype. It took him about 10 years to make his first set of discoveries using self-experimentation (various things related to sleep and depression) and about 5 years to make his next discovery (sugar water and weight loss). Then the discoveries started coming faster and faster. At the end, I think he was eating a stick of butter a day because he thought it helped his brain function. I suspect that his measurements of brain and body function were hopelessly compromised by his hopes and expectations, given the he knew exactly what treatments he was applying to himself each time.

        • Yeah, I’m definitely not recommending the specific things Seth did. I think he did a fair amount of noise-chasing, and he was very susceptible to believing the ‘just so stories’ that he constructed.

          But he did write a very good chapter on self-experimentation, with interesting historical examples of cases in which people learned from experimenting on themselves.

        • This has been the outcome of most self-experimenters that I know. They’ll cut out a food when they feel bad, and since they get better after, they’ll remove it form their diet. This continues every time they feel bad until they’re subsisting on raw pinecones. n=1 experimentation requires a huge amount of discipline and objectivity in exactly the place where we are least capable of being disciplined and objective… ourselves.

          I on the other hand have the opposite problem as a consequence of professional training.

          me: – throwing up after eating dodgy shellfish –

          wife: Did you get sick from dinner?

          me: It is theoretically possible but let’s not jump to conclusions.

        • LOL

          I don’t think I am in that category self experimenters as you describe them. For me there is merit to eating small portions regardless of the regime. We eat way too much to begin with. I have cut my portions in halves. And I have lost 5 lbs.

          Surprisingly, nearly all the athletes [hunks] I know drink a relatively large amount of alcohol. And we don’t really read or hear much on that in the news.

  4. What is the definition of a high-intensity workout? What pulse rate do you reach in high-intensity and for how long do you maintain it (I guess I’d need to know your age too to calibrate that)? I’m 54 and try to exercise about 8 hours a week, but I never push myself to the point of breathlessness. That’s because my cardiologist told me not to (I’m in my 8th year of hemodialysis, so in complete kidney failure); he said the best way is to exercise in a sustained way (over a long period) such that you are not gasping for breath. So I hold a pulse rate of about 105 for a sustained period. I guess this is only specific to my situation, but I’m curious about why high-intensity (I’m not sure what that means yet) is better than not-out-of-breath intensity.

    • I recall reading or being told that high-enough intensity exercise is needed to build heart muscle (but clearly this is counter-indicated in your case). But also, I think that what is recommended as high intensity for adding heart muscle depends on things like age, sex, size, and pre-existing level of fitness.

      Also, the definition of high-intensity is usually measured in terms of METS (“metabolic equivalents”) rather than pulse rate.

      What I consider high intensity exercise for me is at a pulse rate of about 130, which translates to a little over 5 METS. I know someone who exercises at about 10 METS for about 45 minutes most days to build/maintain heart fitness.

    • Yea, lot’s of potential definitions of “high intensity”. Heuristically, it’s anything where you cannot maintain a reasonable physiological steady-state- so above lactate threshold IOW. As for benefits, at very high intensities (around or greater than VO2max power) you are recruiting a large amount of your fast twitch fibers, in addition to the slow twitch. It is also time efficient, in that you can train your oxidative capacity with much lower volumes (Martin Gibala at McMaster has done a lot of the research showing this, along with guys like Jan Helgerud). Finally, maximal stimulation of eccentric cardiac hypertrophy (which increases stroke volume and heart wall elasticity) probably requires at least SOME work at 95+% of VO2max.
      For heart rate ranges, I like the Karvonen system (where you subtract your min from max heart rates to establish heart rate reserve). Exercise at 70-75% of hear rate reserve (i.e. 0.7*HRR + HRmin) seems to be a lovely sweet spot for steady aerobic exercise, but for high-intensity intervals, I go off of power rather than heart rate. But for me, VO2max intervals are done at around 15-16 METS…

    • Shravan,

      Here is an explanation of high intensity workout. The site contains many articles, along with video interviews with exercise gurus. There are various for calculating maximum and minimum heart rate for exercise purposes. Dr. Mercola cautions against marathon running, which can lead to prolonged injuries.

      https://articles.mercola.com/peak-fitness.aspx

      Obviously one has to get a clearance from a physician especially if one has heart issues.

      I myself have trained at a maximum rate of 170 with a minimum of 120. But takes a year to build up to it. I wouldn’t advise catapulting to a high rate.

      It is probably a good idea to lose weight before attempting high interval training. It was designed for already conditioned people, an alternative to long distance running. Now it is adapted to less conditioned people. The claim is that high interval training once or twice a week [in 20 minute routine] one can heart conditioning. Again each person has to experiment with it.

      Mercola website has great body of information. Probably one of the most popular health related sites in US at least.

    • Shravan,
      ‘High-intensity’ means different things to different people and in different contexts. In this post and my other post on the subject, I’m referring to an exercise intensity that I’m unable to maintain for more than 60 seconds maximum. As a practical matter this means anything that gets my heart rate over 180 bpm in under a minute, because I’m unable (or anyway unwilling) to keep my heart rate that high for more than a very short time. My max heart rate is somewhere around 185 bpm as far as I can tell. As I noted in another comment, my max heart rate is much higher than it is ‘supposed’ to be based on the standard age-based rules of thumb, but that’s just another illustration of the fact that different people are different and a rule that works for the ‘average’ person isn’t necessarily relevant to a specific person.

      As for working out at different intensity levels, I am no expert but I have done a little bit of reading (including the book ‘The Time-Crunched Cyclist’) so I will take a stab at it, acknowledging that I may not get this exactly right:
      If you work out just slightly below, or slightly above, the level at which you can control your breathing enough to maintain a conversation, you are near your ‘lactate threshold’. If you exert yourself above your lactate threshold, the concentration of lactic acid in your blood will increase and eventually force you to decrease your exercise intensity. By exercising very close to the threshold, you can increase the threshold, especially if you do occasional bursts of exercise above the threshold intensity. If you are trying to perform well at long-duration exercise, such as running a half-marathon, then increasing your power at lactate threshold is what you need to do.
      For much shorter durations, though, like sprinting a mile or biking at maximum speed up some of the short but steep hills here in the East Bay (of the SF Bay Area), the more important consideration is ‘maximal aerobic capacity’, often measured by the volume of oxygen consumption at maximum effort (“VO2 max”).

      One of the things I had not appreciated, until I read The Time-Crunched Cyclist, is the extent to which these two quantities are independently trainable. They aren’t _completely_ independent, but enough that if you want to improve one of them you really need to work on it specifically, not work on the other one and count on them both to improve. You can run as many 6-minute miles as you want, even consecutive 6-minute miles, and it won’t prepare you to run a single 4-minute mile. And if you can run a 4-minute mile, congratulations for being an elite middle-distance runner, but that doesn’t mean you can necessarily beat a decidedly non-elite runner in a half-marathon.

      Shravan, I am very sorry to hear about your kidney failure, I hope you are doing OK. Please please please do not add high-intensity exercise to your exercise regime unless your doctor says to do so! I’m sure there are very good reasons that your doctor recommends against it. Perhaps it’s simply a matter of removing the metabolic waste products. Good luck in all things.

  5. Maybe using a sample mean as our Polaris is just silly. Maybe we’ve many more dimensions to consider than Gosset’s barley yield. Maybe we’re all stuck being our own guinea pigs. In my case cycling and weightlifting in numberless variations never worked. I finally tried powerlifting, experimented with numerous routines that weren’t quite right, and finally came up with a bespoke routine that works for me.

    It’s n-of-1 I’ve decided; and anything else is just wishful thinking.

    • One has to experiment with different exercise routines. I follow Dr. Mercola’s site for interviews with exercise and nutrition gurus. mercola.com

      For staying limber stretching & dancing are beneficial, especially if engaged in weights training b/c weights do constrict muscles and body postures.

      I have benefited from bursts of high interval training.

      Diet wise it is helpful to curb high carbs and sugar. Most of us consume far too much of both. Sidney Wolfe one told me that small food portions is a factor in longevity. Of course there are individual genetic and biological variations to be accounted for.

      • Darn I was going to appear to make that mistake.

        More seriously, I have given up on trying to make sense of individual published results unless I get access to the raw data, know the researchers involved (and they are in the small subset of terribly careful researchers) or some how get some other assurance the results are not just being primarily driven by miss-recorded assessments, data entry errors or other sloppy research practices.

        Hey almost every is underfunded, over-stretched and extending beyond their reach with inadequate training and understanding of how easily any sloppiness destroys the value of randomized experiments.

        Early in my career (1985/6), I contrived a small fake two group survival study where miss-entering one date the result changed dramatically and it was barely perceptible in the usual survival plots of the time, to try and make this point. I believe 20% of the clinical researchers got it and the rest just didn’t (I guess they thought it was like worrying about being struck by lightening).

        Sorry Phil that this comment is getting so long. But there have been a number of occasions where I have gotten to see behind the curtain of important studies and more than half the time there was data entries that had a big impact. One a colleague got to see, had numerous impossible weights, BMI, labs values even for elephants! Unfortunately there are often laws and agreements that prevent this from being made public and the publications can stand as is.

        Later in my career, I hoped meta-analysis would pick this up but discrepant results between very similar studies are a hard this to puzzle out. Who was sloppy (if anyone) and how does one discern this. Now with gene expression studies you can check for correlations between expressions other that those of interest and if 5 studies are highly correlated and 2 are not – the 2 a good bet as being the sloppy ones (may not be the investigators fault – it may have been beyond their control). A case of nuisance being an actual advantage.

        So one explanation is simply that the published results do not actually reflect what happened in the trial.

  6. Phil, don’t you need a baseline condition? Gain weight, and exercise using low-intensity rather than high-intensity to see what works better. Then repeat several times.

    • See my post from 2010. I was getting quite a bit of low- and medium-intensity exercise and gradually edging upward in weight and downward in sprinting ability. Then I started doing some high-intensity exercise and I quickly lost twelve pounds.

      I’m not alone. If this interests you, you might read that earlier blog post and the comment thread.

  7. oh yay, I remember that post. Gosh that was a while ago.

    I think what you are saying is reasonable. You have an intervention, which when you are able to apply it rigorously leads to some weight control. This makes sense to me, and doesn’t contradict any of the literature that I am aware of (though I’m not in the field). There are a lot of other interventions will yield guaranteed weight loss results, for example, stopping eating.

    Another broad point that can be made is to measure the outcome you are interested in under the attempted treatment. You intend to treat (ITT!) yourself to a high intensity regiment, and are presumably a motivated participant physically primed to undertaking it. Yet you find that your weight dips and rebounds for months at a time due to time constraints and physical strain. Even on an individual level, whether you count the experiment a “success” or not depends on whether you measure the endpoint during one of the dips or rebounds.

  8. There is a very nice book all about how we should stop looking at averages: https://www.amazon.com/s/ref=nb_sb_ss_i_10_7/138-3211731-1318726?url=search-alias%3Dstripbooks&field-keywords=the+end+of+average&sprefix=average%2Caps%2C171&crid=3JE5V3090WKPE. I think the general point is correct – averages tell you one thing, but not much about the individual effects. I think this over-emphasis on averages is largely the result of past technology – when things needed to be calculated by hand (or on hand-held calculators), the data needed to be simplified in order to say anything. A natural simplification is to look at averages. The fact that technology has changed – for decades now! – is a telling example of just how slowly academic habits change. One of my favorites is to review statistics textbooks – it is amazing that they still market books that have the “big” data set printed in the back of the book. It makes it easy to just throw that book in the pile to not consider any further.

  9. Exercise is great, but diet is also important. Here’s a nice study,

    Vuksan V, Choleva L, Jovanovski E, et al. Comparison of flax (Linum usitatissimum) and Salba-chia (Salvia hispanica L.) seeds on postprandial glycemia and satiety in healthy individuals: A randomized, controlled, crossover study. Eur J Clin Nutr. 2017;71(2):234-238.

    https://www.ncbi.nlm.nih.gov/pubmed/28000689

  10. Some health gurus, including Mercola, recommend fasting periodically. Mercola generally features statistics on various diet regimes. I recommend green juices which contain k2. In fact I also recommend food based organic vitamins. The nutritional quality of mos foods is questionable due to soil quality itself.

  11. Looking at averages in a diet (or workout) study doesn’t seem out of place to me. Yes, some people will gain or lose weight independent of any diet or exercise regime: maybe they start biking to work, they get lunch delivered from a different place, or they divorced their spouse who was cooking healthy meals for them. Yes, some people will fail to adhere to the diet or the exercise regimen, although you can account for the latter with some tracking (e.g. gym studies that use card-swipe data to measure attendance). You’d need massive sample sizes to do any sensible subgroup analyses, which just isn’t feasible.

    But at the end of the day, if you don’t see an effect of an intervention with all the institutional support that can come with being part of one of these studies, how could you recommend it as a solution to someone who doesn’t have that support in place? How should it influence, for example, an employer’s (or a health insurer’s) decision to pay for gym memberships? “Hey, exercising helped me lose weight!” is a perfectly valid statement, but it’s also purely anecdotal. Plenty of successful entrepreneurs dropped out of college, too, but that doesn’t mean we should encourage everyone to drop out and start their own company.

    For what it’s worth, my understanding of some of these studies is that people tend to offset their exercising with extra consumption. E.g. they reward their trip to the gym (where they burned 300 calories) with a Frappucino (350 calories). But we could also imagine other offsetting behavior that is harder to avoid. It makes sense to think about whether it’s preferable to spend an hour in the gym and getting takeout, or spending that hour cooking a healthy meal.

  12. I don’t really understand this post and the discussion about exercise and weight loss. Maybe I‘m missing something, but in my world there is a direct relation between weight loss, exercise and nutrition: if you burn much more calories than you eat and if you don’t concentrate on building muscles than you will loose weight. For US citizens this task may be difficult to achieve (because of what you eat), but in general this should work for nearly everybody. It just requires a certain amount of discipline.

    • Paul, there is nothing wrong with the concept of “calories in minus calories out”: if you take in more calories than you burn, you will gain weight; take in fewer and you lose weight. There’s no fallacy there.

      Unfortunately, the simplicity of “calories in minus calories out” has historically led people to fallacious thinking. For instance, people think “I drink a Coke every day, and that’s 150 calories; if I switch to Diet Coke that will cut out 150 calories per day, I’ll lose a pound every few weeks.” That might be true if people made no other changes to diet or activity when they made the switch, but in fact they do both: they eat more other stuff to make up those 150 calories, and they become very slightly less active. Same thing if people increase their exercise: they eat more, too.

      And it’s not just Americans (anymore): obesity rates in Europe are increasing rapidly, and about 20% of people in Western Europe are now obese. That’s less than the US but not by all that much.

      Anyway, sure, people could lose weight if they were willing to be a little hungry all the time. Few people want to do that; certainly I don’t.

      But this is all very off-topic when it comes to this blog post. When I stop doing high-intensity exercise I gain weight, and when I start doing it again I lose it. That’s all I’m saying. I’m not saying that this applies to you, or that this fact is relevant to the ‘obesity epidemic’ in the US or elsewhere, or whatever. I’m just saying that reports and studies that claim that “exercise doesn’t lead to weight loss” are over-generalized, because that claim is not true when it comes to me doing certain types of exercise.

      • Thanks for your response, Phil. You write

        “When I stop doing high-intensity exercise I gain weight, and when I start doing it again I lose it.”

        But doesn’t this mean that when you stop doing high-intensity exercise you take in more calories than you burn? All I want to say is that we can replace “high-intensity exercise” with an objective measurement (calories in minus calories out), i.e. if you want to lose weight, you need to choose an exercise where you burn more calories than you eat. Don’t get me wrong, I’m not a fan of counting calories. But since this is a blog about statistics I thought we can be a bit more analytical.

        And yeah, there are a lot of people who think they can lose weight by just drinking some magical soup without changing anything else. However, I hope that you and other readers of this blog do not belong to this group of people so that we can have a meaningful discussion about this topic.

        • Paul,
          Yes, if I cut out high-intensity exercise and _also_ dieted — in the sense of eating fewer calories than I am inclined to eat — I could maintain or even reduce my weight. In the vast parameter space spanned by all of the possible diets and all possible exercise regimes there are big regions that would involve maintaining or losing weight.
          Supposedly the vast majority of people will not stay on a diet for more than a few months. Maybe I’d be different if I were to try it…hey, everyone is different, maybe I could be the king of dieting! But I have no reason to believe I’d be on the good side of the curve, and actually if I had to guess I’d guess I’d be on the bad side: I love to eat. Indeed, one of the great things about biking is that it lets me eat more. Once, after a big ride (108 miles, 10K feet of climbing), my wife and I went to a restaurant that turned out to have a choice of two different prix fixe dinners, both of which looked good…so I got both. Winning! I have a pair of socks that are embroidered with “Ride to Eat. Eat to Ride.” Exactly!
          Anyway, for me, if I do some high-intensity exercise I get down to a good weight and stay there, _without_ having to exert any conscious control over my diet. I much prefer (do some high-intensity exercise, eat whatever I want) to (don’t do high-intensity exercise, spend my whole life dieting) so I do the former. I’m not saying the latter wouldn’t work too!

        • Calories in – Calories out is obviously true but not very helpful. We have a basal metabolic rate that is highly regulated by something like an “energystat” in the brain. The setpoint for this energystat can be reset given a certain diet or weight or exercise history. If we gain weight and then lose it and return to where we were last year, out energystat will have been modified (and probably not return to where it was last year) and even though we look the same (as one year ago), it takes more (or less) food/activity to maintain the same weight we had one year ago. Also, the energystat is part of a complex feedback for a highly regulated appetite (we think of this as “will power” — it’s not, or put differently “will power” is probably not what people think it is) that includes endocrine signaling (leptin, ghrelin, and others). Because of interindividual variation in these systems, people have very different responses to starting something new – like an exercise regime. one person may start training 8 hours per week and their energystat amps up their appetite but to a point that doesn’t quite make up for extra calories used and this person loses weight. Another person starts the exercise and the energystat amps up their appetite to a point that is just above what is needed to maintain weight and they gain weight. This research is far from my wheelhouse so probably better to take what I’ve said as “kinda right” and see some of the posts at this blog http://www.stephanguyenet.com, for example.

      • Anyway, sure, people could lose weight if they were willing to be a little hungry all the time. Few people want to do that; certainly I don’t.

        Isn’t this the real issue? Why are people more hungry than they need to be to maintain the supposed healthy weight? Obviously you stop being hungry after eating some amount right? Why doesn’t this happen a little sooner?

        After my own self experimenting I came to the same conclusion as many others. Its that when my diet contains a large amount of carbohydrates (lets say approximately over 50 grams or 200 calories per day, about what you would get from eating about 1 kg of vegetables, eg a couple large salads), my appetite increases. I eat more to be comfortable. Alcohol does the same.

        Also, it seems likely to me that people have found out that by putting sugar in food they sell more of it. If not why is sugar added into almost every processed food? Refined sugar used to be considered a drug, apparently (I have not traced to the original source) it was even ok to consume while fasting:

        In fact, Mintz quotes from Thomas Aquinas discussing the delicate issue of whether one could eat sweets during fast without breaking religious rules. The answer is very clearly in favor of sugar, thanks to its medical value:

        ” Though they are nutritious themselves, sugared spices are nonetheless not eaten with
        the end in mind of nourishment, but rather for ease in digestion; accordingly, they do
        not break the fast any more than taking of any other medicine”.

        https://www.researchgate.net/publication/258051415_Attitudes_Towards_Sugar_and_Sweetness_in_Historical_and_Social_Perspective

  13. The bottom line question is, how much does it actually matter to health?

    I will grant that severe obesity is a serious health hazard. But, I keep seeing things saying that being modestly or even a fair bit overweight is not terribly harmful.

    Is this true? Does anyone here actually know?

    • how much does it actually matter to health?

      The most obvious thing is you are carrying around extra mass and need more blood vessels, immune cells, etc to support that tissue for what practical purpose? So using up (admittedly cheap today) energy and (precious, it seems to be unsafe for any stem cell to be more than ~60 divisions separated from the zygote[1,2]) cell divisions for no advantage would seem to be a negative.

      On the other hand, even without something obvious like periodic famine to use it for, there could be advantages to slight amounts of extra fat such as resistance to lipophilic toxins[3], or just “hormones” somehow.

      1. https://en.wikipedia.org/wiki/Hayflick_limit
      2. https://www.ncbi.nlm.nih.gov/pubmed/25459141
      3. https://en.wikipedia.org/wiki/Volume_of_distribution

    • I don’t know the answer to your questions, but would like to point out that we need to distinguish between answering them “on average” and “for a given individual.”

      For example (based on the experience of someone I know), if someone of normal weight for their height and build has congestive heart failure in part because of weak heart muscle, increasing their expected longevity probably needs to include intensive exercise to strengthen the heart muscle and then to maintain the strength. The additional exercise burns more calories, so will lead to weight loss — and probably weakness and risk of fainting and falling — unless they also increase their caloric input. But this needs to be done in a way that excludes bad-for-heart-function foods. So it’s not just a matter of weight.

  14. Phil:

    I agree the whole thing is puzzling. Consider this: About a year ago, my appetite diminished a lot. I still enjoy food as much as ever, but while in the old days I was always hungry, this is no longer the case. In the old days I might go to Num Pang and eat 4 sandwiches, while now I’ll be satisfied with 1, and be very full after 2. I’ll often eat just one meal in the day, whereas I used to eat two meals every day. I’m not quite sure how it adds up—there’s no way I’m eating half as many calories per day as I used to—but I do feel I’m eating a lot less.

    Around the same time as this change in appetite, I gained about 10 pounds.

    From one perspective, this makes sense: When I was thin, I was hungry all the time. Now that I’m fatter, I’m not so hungry. My body is doing its job.

    The thing is, though . . . when I was thin and hungry all the time, I ate my fill. It’s not like I stayed thin by eating a low-carb diet, or by some sort of effort of willpower. I was hungry all the time, but pleasantly so. When I was hungry enough, I ate whatever I wanted.

    As Seth would say, my “setpoint” has changed, but a lot of other things have changed too. You’d kind of think that, if I’m less hungry and I eat less, my weight should go back down to what it was before, but it hasn’t done so. I haven’t put in any effort to reduce my weight—but, again, that’s part of the point of the story.

    As I said, the whole thing baffles me. And that’s just the within-person comparison. The between-person comparisons baffle me too. All of it almost makes sense, but not quite.

    • I gained about 10 pounds.

      In Phil’s OP post he talks about +/- 10 pounds as well. From what I’ve seen this could be a normal variation, just your body retaining more or less water for “some reason”. Maybe you’ve been sleepwalking to the sauna every night?

      • Anon:

        No, I don’t think so. My weight has been pretty stable. The +10 points seems like a real change in level. And the change in appetite is real too. Of course, as noted above, given the change in weight, the change in appetite makes perfect sense.

    • I didn’t mention, though perhaps I should have, that when I had my crash and had to spend five weeks off the bike (and couldn’t run or do much other exercise either) my appetite plummeted, much like you describe. And I, too, gained weight. In my case that’s easy to understand: I was exercising a lot less and eating a lot less, but I wasn’t eating _enough_ less, so I gained weight even though eating less.

      As for you, though…hey, how would I know? Somehow you are burning fewer calories. I don’t know why.

    • Hi Andrew, one self-experiment worth doing may be to switch to eating very simple (almost bland), minimally processed foods. My bet is that you will then be able to ‘eat your fill’ *and* move set-point downwards. Stephan Guyenet has written a lot about this whole nexus of food reward, and dopamine signaling in the brain: https://www.amazon.com/Hungry-Brain-Outsmarting-Instincts-Overeat/dp/125008119X/ref=sr_1_1?ie=UTF8&qid=1531398136&sr=8-1&keywords=Stephan+Guyenet

      • Chris:

        I eat some simple bland minimally processed foods, but I wouldn’t want to eat only that! I mean, sure, if otherwise I would die, I’d do it. But, like Phil, I get a lot of joy from eating, and I’d like to keep as much of that joy as I can.

        Also, there’s something weird about eating one way for over 50 years and then changing. I guess I can accept that my body has changed, and so this changes what I need to eat. It’s still kind of hard for me to understand. As I wrote above, one funny thing about all this is that I seem to be so much less hungry than I used to be, just a couple years ago.

        • Yep, bodies do weird things- constantly adjusting to a fluctuating internal and external milieu :) I agree, overall satisfaction is important. I’ve never been overweight but I DO gain a couple pounds whenever I switch to eating a lot of convenience foods, instead of my usual simple, plant-heavy fare. I rapidly shed when I cut convenience foods to 3 hours/week, I drop a few (despite eating what feels like more than enough to compensate), and gain a few when I do less. These patterns are, at least so far, weirdly predictable for me.

        • lol, blog destroyed that comment! “When I cut convenience foods to <20%, I rapidly lose those couple pounds. Likewise, when aerobic exercise volume exceeds 3 hours/week (despite eating to compensate)".

  15. A decade ago for a TV show a dozen random sedentary adults, men and women, trained to run a marathon. They included a lean 22-year-old woman, a 40-year-old woman who weighed 250 pounds, and all sorts in between. Only one of them had ever run competitively, twenty years prior in high school. At the end eleven of them ran, AND FINISHED, the New York marathon. One woman dropped out of the training after five weeks from stress fractures, and because she was diabetic she couldn’t recover in time. The relevant point here is that the program mentioned that after about two months every one of them lost 12 to 15 pounds. For the next four months that didn’t vary. Diet was entirely up to the individuals. So this one experiment confirms that heavy exercise alone works up to a point, and then no more. BTW, the big gal kicked ass.

Leave a Reply to Martha (Smith) Cancel reply

Your email address will not be published. Required fields are marked *