Of Course Exercise is Effective For Weight Loss, You Low-Expectation-Having Science-Allergic Triumphalist Donkeys

Note: This is long-ish. If you have a shit attention span, or you’re in a hurry, or you’re lazy, skip to the end — there you will find a 150 word summary.

Nothing Works, Give Up, Go Home

Often, a good topic to write about presents itself at an inopportune moment. That’s what happened here.

Ideally, what I’d prefer to do with a topic like this is:

  1. do a lot of work on it quietly
  2. wait until it becomes relevant again
  3. release a hastily finished pre-written piece at that exact moment
  4. recieve accolades for looking like I had some weighty and extensive thought at the precise moment it became relevant, and not six months previously when I was half-drunk at 1 in the morning

We’re not doing this today, because I can’t be bothered trying to look clever any more. So consider this a pre-response to the next time someone, anywhere, makes the claim that exercise is ineffective for weight loss.

We won’t have to wait long, because people say this all the time.

Well, science has some bad news for you. More and more research in both the UK and the US is emerging to show that exercise has a negligible impact on weight loss. (Guardian, 2010)
The truth, they say, is that while physical activity is useful in reducing the risk of developing heart disease, dementia and other conditions, it “does not promote weight loss”. (Guardian, 2015)
Why you won’t lose weight with exercise alone (ScienceDaily, 2016)
Okay Don’t Freak Out But Exercising Doesn’t Make You Lose Weight (NYmag, 2016)

(You can go and look for yourself if you need more examples. That was just the first page of Google.)


Many of these articles are OK. Some of them contain nuance. Some of them even attempt to discuss evidence.

But if you stick them all together, and leave them on the Internet — a place that never saw a convenient fiction it wasn’t willing to embrace — a straightforward and absolute opinion metastasizes out of the various articles: don’t even bother going for a jog, Chuck, it don’t do nothing.

Anyone who writes this, or thinks this, is doing a weird kind of cognitive dissonance. We very obviously live in a world full of people who aren’t fat and are transparently doing a lot of exercise. Are you telling me every time you go outside, and a group of people jog past, you actually think there is no connection between the running and body composition?

Now, if you’ve just thought “correlation doesn’t equal causation”, save me the trouble of coming to your house to sort you out, and slap yourself gently in the face. The key qualifier here is necessarily. Things that are related are very often causal, there is just no implication that this MUST be the case.

Look at her wasting her time. Do you even PubMed, happy lady?

Are these people dieting themselves thin and then just running, cycling, lifting, dancing around for fun? This is deeply silly, and I don’t think anyone believes it. So, I think there’s a little bit of coded language at work in ‘exercise doesn’t help you lose weight’.

What’s implied is something normative. A ‘regular’ person. A ‘normal’ amount of exercise. Some level where practical and physical barriers apply. In other words, what people mean is:

“a regular person is unlikely to lose weight with a normal amount of a normal type of exercise”

That seems a lot more reasonable, doesn’t it?

And for good reason. Let’s look at the evidence.

Touch Me Gently on the Cochrane

The Cochrane Database is a big aggregator of clinical evidence reviews. They produce authoritative papers on clinical questions.

They have one on exercise and weight loss in fat people. It combines the results of 43 other studies into a single analysis.

What did it find?

“When compared with no treatment, exercise resulted in small weight losses across studies. Exercise combined with diet resulted in a greater weight reduction than diet alone (WMD — 1.0 kg; 95% confidence interval (CI) -1.3 to -0.7). Increasing exercise intensity increased the magnitude of weight loss (WMD — 1.5 kg; 95% CI -2.3 to -0.7).”

Now, that’s not very much goddamn weight. So, while we can conclude there is a reliable, small effect it’s not one that’s worth a good goddamn. A single solitary kilo?

Here’s why. The following are some of the exercise protocols, chose at random, which are listed in the 43 studies aggregated:

Aggel-Leijssen 2001

“exercise (cycle ergometer, walking or aqua jogging) 4 times a week for 60 minutes a session at 40% VO2 max intensity”

Aggel-Leijssen 2002

“cycle ergometry 3 days a week for 33 minutes at 70% VO2max, or
cycle ergometry 3 days a week for 57 minutes at 40% VO2max”

Balkestein 1999

“4 days a week for 60 minutes a session at 40% of maximum heart rate”

Cox 1996

“vigorous intensity stationary cycling exercise 3 days a week for 30 minutes at 60–70% maximum heart rate”

Hellenius 1993

“walk / jog 2–3 days a week at 60–80% of maximum heart rate for 30–45 minutes”

Jackicic 2003

“vigorous intensity high duration exercise (5 days a week of brisk walking to expend 2000 kcal / week), or
moderate intensity high duration exercise (5 days a week of slower walking to expend 2000 kcal / week), or
moderate intensity moderate duration exercise (5 days a week of slower walking to expend 1000 kcal / week), or
vigorous intensity moderate duration exercise (5 days a week of brisk walking to expend 1000 kcal / week)”

On and on it goes in that vein.

What do you notice?

ONE: These protocols suck

No coach without profound cognitive impairment would ever propose such basic, low-expectation dreck as a meaningful exercise program.

Basically, these protocols range between a gussied-up version of ‘go for a walk’ through to ‘come on, shift yourself about a bit I guess’. They have no skill development component, no increase in work capacity to match progress, no anaerobic work of any kind, no mobility or movement pattern work of any kind, basically nothing. Even the most potently inbred polo-shirt-wearing chain-gym rep jockey would take better care of you.

TWO: These programs HAVE to suck in order to do science

Don’t get angry with the scientists. The reason these programs are identical and as bland as plain white toast is that it is necessary. Having variables sufficiently controlled enough to study demand that we heavily compromise on all the things that make fitness ‘work’. Everyone has to do the same program, for the same length of time, with no thought to any other factors whatsoever. We are trying to observe a controlled stimulus in order that we can derive information from it… but competent training (and the improvement you see in muscle mass, bone density, body composition, vascular health etc. that you see with it) is wildly nonlinear.

These programs in these studies aren’t designed to cause maximum weight loss. They’re designed to create an environment where we can demonstrate a meaningful difference between our set independent variables, if one exists.

And, now, we see the whole story.

Remember before when I said the qualified version of ‘exercise does nothing’ is “a regular person is unlikely to lose weight with a normal amount of a normal type of exercise”? Well, these programs — crap by necessity as they are — meet people’s definition of normal.

That is the key sentence. People writing about exercise being ineffective consider this evidence to be ‘reasonable’ examples.

And we reach the core of my dissatisfaction and anger with the whole premise.

The idea that this is a ‘normative’ amount of exercise is a total failure to communicate the correct expectations.

Crossfit might be occasionally goofy and insane, and if you want to read me popping off at it in detail that’s here, but they’ve certainly built a culture around aggressively negating this low-expectation fitness industry BS. They may have their own problems with culture, crap injury management, terrible programming, and many other things besides but they certainly do show up and bang.

And that works a damn sight better than ‘30 minutes at 50% of maximum heart rate’, as otherwise known as a LITERAL WALK IN THE PARK.

Here’s another personal example. I’ve been doing some body composition work recently, and so I’ve cranked up activity as much as possible. What does *really* active look like?

(A quick primer: in epidemiology, we can’t follow people around with a stopwatch and a bag of scientific equipment all day to measure how much energy they expend. Instead, various occupations, exercises, sports and other activities are counted in METs. This isn’t a basketball team, it stands for metabolic equivalent of task.

One MET is both a) the equivalent of 1 calorie per kilo per hour and b) about equivalent to your caloric expenditure at rest while sitting. Naturally, some tasks are harder and/or more active than others, and they have different MET scores. Here’s a quick list of tasks and their MET equivalents:

Standing up and fidgeting is 1.8 MET
Yoga is 2 to 4 MET.
Shovelling coal is 6.3 MET.
Walking up a steep hill at a brisk pace is 8 MET.
Competitive soccer is 10 MET.
Top shelf cross country skiiing is 15 MET.
Cutting wood Paul Bunyan style (i.e. 50 axe strokes per minute) is 17.5 MET.
Running absolutely Roger Bannister flat-out is 23 MET.

These MET values are divvied up into minutes (metabolic equivalent of task minutes). Sedentary is often regarded as less than 500 MET minutes per week. Active is over 1000 MET minutes per week.

So, here’s mine. I’ve guessed where there were no guidelines.

A Week’s Worth of Heavy Exercise, If You Are A Bit Mad
[data give as: minutes exercise (METS)]


30 weights (5)
30 heavy carry/pull/drag (8)
30 arm cycle (4)


60 basketball (6)
30 bike (6)
45 weights (5)
10 minute swing test (8)


55 light cardio (4)
15 stretching, active (4)
45 weights (5)


55 very heavy carries (10)
15 smashy ball (17.5)
15 weights (5)
30 walking (4)


10 rowing (7)
30 kettlebells (6)
45 weights (5)
45 cardio, brisk (5)


20 stair climbing / 20 stair descending (9 / 3.5)
20 bike (6)
30 very heavy carry/pull/drag (10)
20 medium carries (8)


20 smashy ball (17.5)
20 medium carries (8)
40 weights (5)
10 incline walking w. weights (6)

Yes, that’s all a real thing which happened. No, I didn’t get hurt. No, I didn’t ‘wreck my metabolism’. I’m fine. And a lot leaner.

If you remember from before, ‘active’ starts at over 1000 MET minutes per week. At this point in many studies, we stop counting, because you’ve met your ‘quota’.

However, the above is 5167.5

That’s more than five times the official designation of ‘active’.

Here’s the truly terrible part:

  1. I’m just some guy. I’m not an athlete. I’m a scientist who’s worried about having to buy new pants if he eats too many cream cakes, and therefore is doing a bit of body composition work. I’m not in the military, I don’t get paid to ride a bike in my underwear, I don’t have a sponsor who sends me ‘free supplements’. Just some guy.
  2. These are all conservative estimates. I have removed rest periods and been reasonably strict with what time counts. For instance, an hour of ‘stair climbing’ is more like 20 minutes of climbing stairs, 20 minutes of descending stairs, and 20 minutes of wanting to kill yourself. I have used the lower estimate wherever possible.
  3. I was just re-learning training management again during this process. That is, I was getting better, not worse, at handling the load over time. The reason everything is so scattergun was primarily to allow me to do the parts that say ‘weights’.
  4. People with seriously active jobs (say, forestry, or mining, or teaching PT) with more than four hours daily activity would do something equivalent. People with active jobs AND training would easily do more.
  5. Someone with a sedentary job and the right kind of hobby would do this much activity as well. Vigorous cycling would be about 10 METS, and road cyclists traditionally race or execute longer rides on a weekend schedule. That could be up to 2500 METS *in a single session*. And you wondered why they eat all that pasta!

And now, my own version of our original premise: CRAP exercise is ineffective for weight loss.

The focus at this point immediately shifts to two much more important questions.

  1. how do we change the focus away from weight loss, and to all the things exercise does much better than weight loss which also stop you dropping dead — glucose metabolism, arterial health, blood pressure, and so on?
  2. how do we compel people to learn to perform exercise which isn’t crap; specifically a) how do we communicate the proper expectations and b) how do we get people the physical skills they need in order to be able to do the right kind of work?

These are infinitely more serious and difficult questions. And the public health of most of the developed world is waiting for your answers.


Clinical trials on weight loss or body composition often compare a diet group to a diet+exercise group.

Frequently, exercise makes little / no difference when it comes to body composition changes.

This has led a variety of people to conclude that exercise is broadly ineffective for weight loss.

This conclusion is stupid.

Two main reasons:

  1. Most research on this topic is on obese people. The exercise protocols administered are a gussied-up version of ‘walking around’. On a 1–10 intensity scale, I give them about a 1.5
  2. It’s a dead issue. The focus on weight loss — and there’s so much — sucks oxygen from equally important issues (exercise capacity, cardiac and vascular health, glucose metabolism, muscle mass, bone density, tissue quality, etc.)

Better conclusion: CRAP exercise is ineffective for weight loss.

The lesson from that: learn to do exercise that isn’t crap.