TL;DR Science works, but good science is hard to find. I'm linking some of the more useful info I've found below.
This is year two of trying to get in shape. Year one failed, but I was determined to do better. I had a lot to learn, and I just didn't know it yet. I've done my best to add references where possible, but admittedly some of this is "common sense" advice that I can't justify with anything more than "it makes sense, doesn't it?"
#0. Measure everything you can think of, and then find more things to measure. You can't see changes you don't measure.
Before starting out, I had a sense of the fact that multiple factors would affect the weight on the scale not all of which would correlate to me losing fat, which was the real objective. Though my scale at home does BIA to measure body fat, after getting my body composition tested with an InBody 570 (maybe not the gold standard, but more reliable), I found that feature useless. It was off by 5% (and these things can be off by as much as 7% SEE [1]). Knowing this I was expecting to at least see trends, using the fact that it was off by the same amount every time, but the changes were negligible in the readings because of the underestimation. Body circumference measurements, measured at the widest point possible around my chest, upper arms, waist, hips, thighs and calves, turned out to be much more reliable. I've tried progress pics every few weeks, but I honestly don't see the differences between them, even though I made sure to use the same clothing, lighting, body position and camera angles, so I've ruled those out as too subjective.
#1. Exercise alone won't make a visible difference.
Initially, I didn't think to change my diet. It wasn't that bad to begin with. I've always made sure to eat fruits and vegetables, balance my meals between sources of carbohydrates, protein and fats, and don't indulge in junk food very often. I convinced myself that since my weight was stable (though bordering BMI 25) I was probably eating about maintenance and just needed to exercise consistently. I knew that if I went overboard too quickly I'd be too tired to keep it up so I made a plan to go out for a walk every morning for about an hour. I also made sure to keep records of not only my weight but my body circumference measurements (added up to be able to see trends).
I was able to keep this up for about 4 months, and then life got in the way, I needed the extra sleep, and I was too demotivated to start over. I had managed to lose 3lbs and 9in. It's not nothing, but it's little enough that I couldn't notice. Even though I had made measurements I could trust to tell me there was one, I couldn't see the difference. By the time I decided to try again this year, I knew I needed an approach that would yield results faster and more consistently. Had I found the NIIDK Body Weight Planner any sooner, I would have known that a small change in exercise wouldn't make a big dent. The math behind the model shows that exercise makes up a small part of total energy expenditure [2], [3].
When I picked up exercise again early this year, I opted for more intense exercise, having read about the benefits of HIIT [4] and also decided to exercise at least 3 times a week instead of everyday, considering it easier to stick to. I also added some strength/resistance training, changing the focus for the first time from simply losing weigh to changing my body composition. I told myself that if I stayed the same weight by losing fat and gaining muscle I could be still call it progress. In two months, I made no real progress. I "lost" 1.5in in body circumference overall (and it's easily just from starting to measure a week after starting weight training, when I was likely swollen and then going back to baseline). My weight was all over the place and showed no downward trends.
#2: You're not a freak of nature if you're struggling to lose weight. Thinking you are will keep you from figuring out why you're struggling.
I went away on vacation for 3 weeks and spent most of my time walking (averaging 8mi a day, including hiking). When I came back I finally saw some change: I'd lost 3lb and almost 4in. Had I not been in a depressive rut for the month that followed I maybe wouldn't have lost that progress. When I decided to start again, I kicked the exercise up another notch. In three weeks I lost 1lb and about 2in, but then I saw the numbers creeping back up. I was at my wit's end, ready to believe something was medically wrong with me and schedule a doctor's appointment. As far as my doctor could tell, I was perfectly healthy, and if anything my resting heart rate and cholesterol levels had both gone down since the last time I visited. No thyroid or insulin resistance problems (though they do run in my family), no PCOS (which turns out wasn't as likely as I thought [5]). The effects of exercise on my body were real (heart rate down to ~60bpm, cholesterol lower than it's been in years). They just weren't the effects I was looking for. My doctor assured me I looked fine and all my markers were in healthy ranges, but commended me for taking up the exercise and told me to keep it up.
At this point I was more than a little upset. Okay, so maybe I was healthy, but that wasn't specifically the goal. I knew I was healthy enough (though I could of course be healthier). I wanted to be able to wear my clothes from a couple of years ago and know that "my size" will continue to fit when I go shopping for clothes. I had stayed around 138 +/- 3lb through ages 17-25 and at 29 I'm not so old that the goal was unrealistic. It's closer to the middle of the healthy BMI range for my height, and I wish the doctor hadn't made me feel like I was crazy to have this goal. Speaking of goals...
#3: "Healthy" can be a relative term.
My doctor was using BMI to determine that I was healthy. She wasn't completely wrong, as BMI ranges (from their inception) correlate to mortality rates [6]. However, this analysis is crude [7], and doesn't work for all populations [8]. Health and fitness blogs love to bash BMI because it places athletic, muscular people (epitome of healthy) in the overweight category. They tout body fat percentages as the gold standard, and yet the recommended ranges don't have the best explanations. Whenever the charts are attributed to anyone, many trails lead to the ACE or the ACSM but I can't seem to find an explanation for how they picked their ranges (possibly because I refuse to pay money to buy their manuals, but the references to "Cooper Clinic patients" they collected data from was a dead end). I have an idea, though, based on hat I could find: approximations for the body fat percentages that pair up with the BMI categories [9], and percentiles for body fat percentages in the US population [10,11]. They line up nicely enough with the "ideal body fat" charts, with caveats.
As mentioned before, the BMI as a measure of health is flawed, even if we only apply it to non-athletes (who make up the majority of the population anyway). The division of the cutoffs into ethnic groups is to account for the fact that the same BMI standards don't apply to everyone. And this makes the math in the simplified body fat charts fuzzy. Even big names like the WHO have questionable backing for some of their claims on body fat cutoffs [12]. So, my advice to you is to be clear about what your goal is and why.
#4: "Health" sounds nicer than "vanity," but vanity is a big driving force in efforts to lose weight.
Let's forget about BMI, then, and go for the vanity factor of placing yourself on a percentile. Take a look at the US population data [10,11]. If you have the time to kill and can dabble in SAS and SPSS, play with the percentiles (if you don't, but trust whoever made the applet, you can follow this link to see where you fall). Some ACSM charts mention percentiles, and "fair" begins at the 40th percentile, while the "good" category starts at the 60th. Ah, so we're playing the comparison game. Turns out you were doing statistics all along when you compared yourself to everyone else. Look carefully at the average population stats [9], and then look at the trends over time [13]. I haven't run into any updates to either BMI or body fat ranges, even though they're both based on data that goes back ~30+ years and is by now obsolete. I'll address proper health in my next point, but to me admitting that vanity played a role in my decision to lose weight was an eye opener.
It should go without saying that advertisements and social media are not representative of the average population, but it bears reminding. Near perfect models get photoshopped, clothes on mannequins are pinned to enhance an unreal silhouette, and everything you see on social media has been cherry picked. I have an issue with these last ones because they pretend to be the real deal, and the truth is they only show what they want you to see, using a number of tricks to exaggerate what they already have. This isn't even a new trend. If you set out trying to look like someone else, you should be aware of the fact that maybe even that person doesn't look like the picture you're aiming for. I'm personally shooting for a number I think is feasible (because I've been at that weight before), but I'll be happy at a different weight if my old clothes fit. I know it's a vanity thing, but I'm also fairly confident it's realistically attainable.
#5: You can improve your health without changing your weight much.
Back to my story, after being upset at my doctor and poring through mounds of information trying to prove her wrong... I have to admit she was mostly right. My cholesterol and resting heart rate were proof of it. The fact that I could walk up 7 flights of stairs and beat an elevator (though I had to stop to catch my breath at the top) was proof of it. The fact that I could sustain a running pace for more than 2min was proof of it. I don't want to start a debate about whether or not it is possible to be healthy at any weight, but I do believe you can be healthier at your current weight when you take up exercise and a sensible diet, even if it doesn't show on the scale. If you thought I'd reference this study, I have more to tell you.
#6: Top Google are the ones people like to click on the most, which means writers have an incentive to tell you anything you want to hear.
I could make this post so much longer, because the bad science abounds and annoys me. Some rules of thumb:
- if an article lists no sources or references (even if they allude to unnamed studies), be skeptical
- if the sources/references are other articles by the same author, watch out for circular reasoning or a lack of sources/references in the cited article
- if there is a very real study listed, don't be afraid to read it and question it (beware of small sample sizes, sketchy methods, and sources of funding)
- if they're oversimplifying something (giving you a quick rule of thumb like "losing 10lb equals one dress size" when there's so much that goes into clothing sizes), it's probably wrong
- ALWAYS ask yourself if they're trying to sell you something (especially true of those suspenseful articles that promise results and the information you are looking for if only you sign up for their program, buy their book/supplements, etc.)
Now, some mythbusting:
- Burning 3500 calories won't make you lose a pound of fat. When it's quoted this way, the authors conflate weight with fat (because when we set out to lose weight we usually want to lose fat), but the study that landed on this number spoke only of weight [14]. This rule is especially problematic when used to determine the time it will take to lose a given amount of weight [15]. The 3500 cal rule proposes a linear model, which is admittedly easier to understand if you're a little apprehensive about math and accurate enough in short time scales, but ultimately wrong. Better models use differential equations and multiple variables to make accurate predictions within a chosen margin of error. The NIIDK Body Weight Planner I mentioned above is user friendly and has been a source of comfort when I feel like I'm stalling: I'm still within the margin of error and I invariably go back to bouncing around the trendline.
- CICO (Calories In, Calories Out) works. Macros, you can fudge a little [16]. Diet plans that claim to ignore calorie counting but restrict your intake otherwise (low carb, low fat, eliminate sugars, eat only whole/natural foods, eat only in small time frames) ultimately aim to reduce the number of calories you consume but trick you into thinking that you're otherwise free to do whatever you want. The fact is that if you were to follow any of them and go over, you wouldn't lose weight, so don't try to game the system by pretending anything goes within the limits of your restrictions. If you replace breads, pasta and sweets with avocados, nuts and cheese (which you will likely consume in equivalent volumes to stave off hunger) you will probably still be eating more than you need.
- Protein matters, but not as much as they advertise. This post does a pretty good job of explaining why, with references I won't bother repeating in my already very long list. The takeaway is that most people are already meeting minimum protein requirements. Higher protein intake is recommended to prevent muscle loss and modestly increase the rate of weight loss (though it doesn't work on everyone). You still probably don't need the 1g of protein per lb of body weight. There are more, but here's one study making slightly lower suggestions of up to 0.7g of protein per lb of body weight for people who exercise a lot. If you weigh over 100lb and are both trying to cut calories and increase your protein to 1g per lb of body weight, you will most likely end up eating more meats or taking more supplements (because they're protein-dense) than dietitians would recommend (they favor a higher plant-based food intake). It's also hard to sustain long term. If you try to go the natural route and use plant based proteins, like nuts, you may go over your calorie intake. Moderation is key. At 150lb, I'm shooting to average 90-100g of protein a day and it seems to be working fine.
- There's no science to justify the whoosh effect, or that you can use cheat meals/days to "trick your body into losing weight." There, I said it. This falls under the list of things people love to read because it tells them what they want to hear. Don't get me started on how little sense it makes that your fat cells would hold on to water that you pee later. Big swings in body weight, once calories are accounted for, are most likely due to fluctuations in water retention, and that depends on your sodium levels, how intensely you've been exercising, water intake... Expect the fluctuations, and play with them if it makes you happy. Just remember the changes on the scale are likely just water weight, and you're interested in fat loss. Taking a break can have a positive psychological impact, and is good practice in learning to forgive yourself if you eat a bit more than you planned.. You may not make new progress, but you won't lose much either. Allow yourself to have a big meal with friends and family for special occasions because it will make you happy and because it's not a permanent switch to caloric excess. Just make sure to go back to your plan and the math of averages will be on your side: if you have 100 days eating at a deficit and one day eating above, it averages out to a deficit anyway. All you need is to watch your deficit days:surplus days ratio.
- If it hasn't been 6 months on your fitness plan, it's most likely not a true plateau, but the result of you cutting yourself too much slack [17]. For people looking to lose a significant amount of weight who started out at a modest deficit, it's worth reassessing intake every few months. This is true for significant amounts of weight because your daily energy expenditure depends on your weight, so a big weight change will have a big impact in your expenditure. Other than that, you're most likely either being impatient in riding out the water weight fluctuations or lying to yourself about the size of your deficit.
- Your math won't be perfect, even if you try very hard, and that's okay. You've probably read somewhere that food labels are allowed to be off by as much as 20% and the FDA isn't even checking for compliance [18] (though name brands are closer to a 5% error [19]). You won't find it hard to believe that since portion sizes are hardly standard at restaurants you could be eating up to twice as much as it says you are in their nutritional information table. You're likely aware of the fact that most people underestimate their calorie intake and overestimate their exercise calorie expenditure. The machines and gadgets are probably lying to you (listing individual margins of error depends on the model of gadget, but they all have them). You can still work with numbers that are consistently off by the same amount. If your elliptical machine says you burned 400cal on Monday and 450cal today, it's still fair to say you burned a little more today. As long as most of your meals are not at restaurants (where it's hard to measure and get a full ingredients list to make your own estimates), even if your numbers are off, it won't make a big dent in your average accuracy (assuming the rest of your math is done with the information from calorie databases for each ingredient and those are accurate enough). Make sense of the numbers as you measure them. If my theoretical maintenance is 2300cal but my calorie intake averages 2100cal on a 3 week period with no weight change, then 2100cal is maintenance the way I'm measuring it and that's the number I want to stick to. I'm likely measuring wrong, but at least I know the baseline.
I know it's not short, but reading this summary already takes a lot less time than it took me to find these things out on my own and write them, so hopefully I'm saving someone else the trouble. Hope it's useful to you. Feel free to add your own resources or let me know if you found data that disagrees with what I'm saying. I'm open to editing and fixing any mistakes I may have made.
References:
#0:
BIA scales can be off by up to 7%
[1] Accuracy of Consumer Grade Bioelectrical Impedance Analysis Devices Compared to Air Displacement Plethysmography
#1:
Reasonably accurate mathematical model for weight loss based on energy imbalance
[2] Quantification of the effect of energy imbalance on bodyweight
[3] Dynamic Mathematical Model of Body Weight Change in Adults
Effects of HIIT
[4] High-Intensity Intermittent Exercise and Fat Loss
#2:
PCOS affects maybe 15-20% of women, and only 50-70% of those are affected by insulin resistance
[5] Epidemiology, diagnosis, and management of polycystic ovary syndrome
#3:
BMI and mortality
[6] Body-Mass Index and Mortality among 1.46 Million White Adults
BMI - history and use
[7] Body Mass Index
BMI cutoffs should be different in Asian populations
[8] Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies
Healthy BMI to body fat percentage conversions
[9] Healthy percentage body fat ranges: an approach for developing guidelines based on body mass index.
CDC Body composition data in the US
[10] Body Composition Data for Individuals 8 Years of Age and Older: U.S. Population, 1999–2004
NHANES Data from 1999-2000 (warning: SAS or SPSS is needed to open the data files)
[11] NHANES 1999-2000 Examination Data
Body fat cutoffs that are cited often have questionable reasoning behind them
[12] More on Body Fat Cutoff Points, More on Body Fat Cutoff Points - Reply
#4:
NHANES BMI trends
[13] Healthy weight, overweight, and obesity among U.S. adults
#6:
Caloric equivalent of 1lb of weight
[14] Caloric equivalents of gained or lost weight.
Predicting weight loss
[15] Time to Correctly Predict the Amount of Weight Loss with Dieting
Different macros don't seem to make a huge difference weight loss over a 2 year period
[16] Comparison of weight-loss diets with different compositions of fat, protein, and carbohydrates.
Your metabolism didn't adapt, you just stopped sticking to the plan
[17] Effect of dietary adherence on the body weight plateau: a mathematical model incorporating intermittent compliance with energy intake prescription
FDA rules for nutrition labels allow up to 20% error, and depend on the honor system
[18] Guidance for Industry: Nutrition Labeling Manual - A Guide for Developing and Using Data Bases
Common snack foods aren't off by that much, though
[19] Food Label Accuracy of Common Snack Foods
