Why Calorie Counting Stops Working

Confession time: I used to love tracking my food in an app on my phone. I’d quickly add what I ate and drank, and would sport a oh-so-very-smug grin on my face when I “hit my numbers” for the day. There’s something deeply satisfying about reducing something as messy as eating to a tidy number. Which is exactly why it’s worth sitting with the question properly: if calorie counting is so logical, why does it so reliably stop working for so many women, myself included, at some point?

The short answer is that calories are real and they do matter, but the “calories in, calories out” model we were sold is an over-simplified version of a much more nuanced, much more interesting system.

Let’s get nerdy about why.

Calories matter

First, let’s not throw the baby out with the bathwater. Energy balance is real. Calories are a genuine unit of energy, and your body genuinely does need a certain amount of them to maintain, gain or lose weight over time. Nothing in this article is an argument against that basic physics.

Where the model falls apart isn’t the physics, it’s the assumption that both sides of that equation are stable, precise numbers you can simply plug in and trust.

“Calories out” isn’t fixed. Or easy to calculate. It shifts based on your genetics, your muscle mass, your sleep, your stress levels, even how much you fidget (yes, genuinely, this has a name: non-exercise activity thermogenesis, and it varies enormously between people, and it matters). Even if online calorie calculators were all exact and precise, everybody is different, so every body is different.

Everyday is different too. Your need for energy will adjust based on your level of stress, the quality of your sleep, how much you fidget that day etc. etc.

And I’m not even going into the seasonality of our energy needs, which is real. The outside temperatures, for starters, will have an impact on how many calories we burn, but the changes in sunlight between winter and summer also affects our metabolism.

“Calories in” isn’t fixed either, for reasons we’re about to get into. Treating this as a simple maths problem, X in, Y out, ignores that both X and Y are constantly moving targets, influenced by dozens of variables most tracking apps don’t account for at all.

The thing with “calories in”

Only processed food really allows you to track anything

Here’s a delightfully inconvenient truth: calorie counting only really works, even in theory, for foods that come with a label. Also know as “processed food”.

Think about it: a packet of crisps has a number on it. A homemade chicken and vegetable curry, cooked from scratch with a “glug” of olive oil and a “handful” of this and that, does not. Well not accurately, anyway.

This creates a strange, largely unspoken bias in calorie-counting culture: it’s far easier to track a protein bar than a plate of roast vegetables, which quietly nudges people toward more packaged, processed foods simply because they’re trackable.

Which is the opposite of what most of health-conscious folks are actually trying to achieve.

Food quality matters too

The quality of your food changes whether your body is even looking for more (aka cravings), or satisfied with the meal you’ve just had.

This connects directly to something else I’ve written about: in a tightly controlled NIH study, people eating a diet built mostly from ultra-processed foods spontaneously consumed around 500 more calories a day than people eating a whole food diet, even though both diets were matched on paper for calories, sugar, fat, salt and fibre. The food itself was influencing intake, independent of any conscious counting at all.

In other words, the quality of what’s on your plate can do a lot of the “appetite regulation” work that a calorie target is trying to force through willpower alone. So if you choose the calorie-counting option for weight loss, chances are you’re quietly sabotaging your efforts by needing to rely on processed food.

The food industry gets 20% leeway on the numbers, anyway

Wait, what?

Yep, even when you are tracking a labelled, packaged food, diligently logging every gram, the number on the label isn’t guaranteed to be accurate. In the US, food manufacturers are legally permitted a 20% margin of error on calorie labelling. 

Independent testing of packaged snack foods has found real world discrepancies too, with one widely cited study finding calorie content averaging 8% higher than labelled, and some products considerably more. 

Let that sink in for a moment. You could be logging your calories with the precision of an accountant, and the underlying numbers themselves could legally be off by a fifth.

That’s not a small rounding error, that’s the difference between “on track” and “quietly over” every single day, through no fault of your own tracking.

Appetite matters just as much as arithmetic

This is the piece that gets left out of most calorie conversations entirely: how hungry you feel isn’t a side issue, it’s a huge part of whether any eating plan is sustainable at all. Since it’s almost impossible to precisely calculate the calories your own body needs in general, let alone on specific days, appetite and energy are your body’s normal, natural feedback system to let you know it’s being underfed. Regardless what your numbers should be.

We’ve covered this properly in this article, but it bears repeating here: research following people for a full year after a structured low-calorie weight loss programme found that hunger hormones remained elevated well beyond the diet itself, meaning appetite genuinely increases as a biological response to calorie restriction. 

A calorie target that ignores appetite is a calorie target that’s fighting your own physiology, quietly, in the background, every single day. But forcing yourself into a number can have a long-term negative impact on your metabolism.

Protein and fibre: the appetite regulation duo

Here’s where the nerdy bit gets genuinely useful. In a randomised, double blind trial, adults who drank a high protein, high fibre beverage before a meal reported feeling significantly less hungry and went on to eat meaningfully less at an all-you-can-eat meal afterward, compared with an identical calorie beverage that was low in both. Same calories in the preload. Very different appetite response, and a very different result on the plate that followed.

Protein and fibre slow digestion, trigger fullness signals, and keep your blood sugar steadier, all of which adds up to eating a satisfying amount without needing a app to confirm it.

Unfortunately for calorie-counting folks out there, the majority of high-protein and high-fibre foods don’t come in a packet with handy numbers on.

Sustainability: the boring word that actually matters

Calorie-counting, done diligently, is genuinely effective in the short term. The research consistently shows people can and do lose weight this way. The problem has never really been effectiveness, it’s been durability. 

Meticulous tracking (assuming those numbers were exact, which they’re not) is a habit that’s hard to sustain for years, especially alongside a job, a family, a social life and the general chaos of being a human being. 

When the tracking stops (and for most people, it eventually does), there’s often nothing else in place to guide eating, because the whole system was outsourced to an app or a set of numbers.

Your relationship with food

There’s also a quieter cost worth naming honestly. For some women, calorie-counting becomes less of a helpful tool and more of a constant, low grade background hum of judgement:
good day # bad day, good food # naughty food, on track # off track.

That relationship with food, tense, numbers-driven, very black or white, a bit joyless if we’re honest, can outlast the diet itself, long after any weight has been lost or regained.

And the same as with step-counters, blood pressure monitors or sleep trackers, for some of us it creates stress and anxiety in its own right. Relying on numbers makes us feel more in control of our activities, at least for a while. The issue comes when the outcome doesn’t match our expectations, even though, on the screen, everything looks “right”.

Or when life gets in the way (because it does, at some point or another), and we just can’t seem to hit the right numbers.

And we end up trusting the screen more than our own body and ignoring our body’s signals (has anybody ever woken up nice’n rested, but your health app tells you otherwise, and somehow you start to worry you didn’t sleep all that well?).

Whereas it’s usually this ignorance of what our body is trying to tell us that landed us with a few extra kgs in the first place.

We are not “anti-calorie”. We are “pro-understanding”

To be really clear: none of this is an argument that calories don’t matter, or that you should never track anything. Tracking can be a genuinely useful, short-term learning tool, particularly for understanding portion sizes, figuring out the energy density of the food you eat, or spotting patterns you didn’t know were there.

And I love the fact that those apps can quickly tell you all that by just snapping a pic of your plate.

The shift I’d encourage is this: instead of treating calorie-counting as a permanent, precise control system (which, as we’ve just seen, it structurally can’t be, thanks to inconsistent labels, unmeasurable homemade food, and a body that adjusts its own hunger in response to just about everything), treat it as one input among several. 

That’s not a less rigorous approach than counting. But given everything above, it’s arguably a more sustainable one.

---

Want a bit of help identifying which foods you love, but don’t love you back?
Download my free guide, “5 Foods In” for a quick’n easy way to upgrade your diet. Family-friendly and all.

References

U.S. FDA, 21 CFR 101.9 Nutrition Labeling regulations (20% compliance margin), with real world testing data from Urban LE, et al., “The Accuracy of Stated Energy Contents of Reduced-Energy, Commercially Prepared Foods,” Journal of the American Dietetic Association (2010) — https://www.fda.gov/regulatory-information/search-fda-guidance-documents/guidance-industry-guide-developing-and-using-data-bases-nutrition-labeling

Sumithran P, et al. “Long-Term Persistence of Hormonal Adaptations to Weight Loss,” New England Journal of Medicine (2011) — https://www.nejm.org/doi/full/10.1056/NEJMoa1105816

Sharafi M, et al. “Effect of a High-Protein, High-Fiber Beverage Preload on Subjective Appetite Ratings and Subsequent Ad Libitum Energy Intake,” Current Developments in Nutrition (2018) — https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6016687/

Hall KD, et al. “Ultra-Processed Diets Cause Excess Calorie Intake and Weight Gain,” Cell Metabolism (2019), NIH Clinical Center RCT — https://www.cc.nih.gov/news/2019/summer/story-01