Intermittent Fasting vs a Calorie Deficit: What Actually Drives Fat Loss

Intermittent fasting is the most popular thing I get asked about that nobody actually needs. Every second new client tells me they are “doing 16:8” before they can tell me what they ate yesterday.

I am not against fasting. I have clients who eat in a window and do very well on it. What I am against is the story that fasting is a separate mechanism, some metabolic switch that burns fat in a way that eating three normal meals cannot. That is not what the research says, and the research on this is messier than either side of the argument admits. So let me walk you through both sides properly, including the parts that do not support my usual advice.

The evidence genuinely disagrees with itself

If someone tells you the science on fasting is settled, they have read one paper.

Start with Kim, Kang and Kim, who pooled 16 randomised trials covering 1,438 participants in 2022 and found no significant difference between intermittent energy restriction and continuous energy restriction in weight, waist circumference or body fat. Dropout was similar in both, about 26.6 percent versus 24.1 percent.

Then read He and colleagues, who pooled 11 randomised trials in 850 people in 2021 and found the opposite. In their analysis, intermittent restriction had greater effects on weight loss and insulin sensitivity than continuous restriction. Crucially, that advantage was concentrated in the short-term studies, the ones running two to three months.

Then read Ezzati and colleagues, who in 2024 looked specifically at the cleanest possible test: seven energy-matched trials in 579 people where time-restricted eating was added on top of calorie restriction. Their conclusion was that the majority of studies did not find additional benefits from adding the eating window.

Three credible reviews. Three different-sounding answers. Anyone quoting only one of them is selling something.

Where the biggest and longest trials land

When reviews disagree, I look at the individual trials that ran longest with the most people, because those are the ones that reflect a real year of somebody’s life rather than an eight week burst of novelty.

Trepanowski and colleagues ran 100 adults for a full year, comparing alternate-day fasting with plain daily calorie restriction. At twelve months, the fasting group had lost 6.0 percent of bodyweight and the daily restriction group 5.3 percent. Their own summary sentence is worth reading twice: alternate-day fasting did not produce superior adherence, weight loss, weight maintenance, or cardioprotection.

Liu and colleagues ran the longest good test of time-restricted eating, 139 patients over 12 months published in the New England Journal of Medicine in 2022. One group ate inside an 8am to 4pm window and counted calories. The other group just counted calories. The window group lost 8.0 kg, the calorie-only group lost 6.3 kg, and the difference of 1.8 kg was not statistically significant.

The honest summary is this. When calories are matched, fasting does not reliably beat continuous restriction, and the largest and longest trials found no significant difference. It is also not useless, and the short-term data in He’s review is real. It is a tool, not a tier above.

So why do people lose weight on it?

This is the part that clears up most of the confusion, and it is well established. Varady and colleagues, reviewing the field in the Annual Review of Nutrition in 2021, reported that all forms of intermittent fasting produced weight loss in the range of 1 to 8 percent, and that all of them reduced energy intake by 10 to 30 percent.

There it is. Fasting works, when it works, because it makes you eat less. Skip breakfast and you are down one eating occasion. Close the kitchen at 8pm and the after-dinner snacking stops. That is a calorie deficit arrived at by a structural rule instead of by counting, and for some people a rule is far easier to follow than a number.

That is a mechanism through calories, not a mechanism instead of them. Which is the whole point.

The thing that actually predicts results

Look at what the diet-comparison literature keeps landing on. A network meta-analysis of 48 randomised trials in 7,286 people found low-carbohydrate diets at 7.25 kg and low-fat diets at 7.27 kg at twelve months, and the authors concluded that this supports the practice of recommending any diet that a patient will adhere to.

The POUNDS LOST trial makes the same point even more bluntly. 811 adults were assigned to four different macronutrient combinations for two years. Everyone lost about 6 kg by six months and 2.9 to 3.6 kg by two years, with no significant differences between the diets. What did predict results was attendance at the group sessions, at about 0.2 kg lost per session attended.

Turning up beat the macros. It usually does.

Two cautions before you commit to a fasting window

Dropout is higher on the harder protocols. In the Trepanowski year-long trial, 38 percent of the alternate-day fasting group dropped out compared with 29 percent on daily calorie restriction. A plan you quit is worse than a plan you find boring.

Watch your protein and your training if you narrow the window. In the TREAT trial run by Lowe and colleagues in 2020, 116 adults did 16:8 with no calorie counting for 12 weeks and lost 0.94 kg, with no significant difference from the three-meal group. In the in-person subgroup of 50 participants, a secondary measure of appendicular lean mass index came out worse in the fasting group. That is one secondary outcome in a small subgroup and I would not turn it into a headline, but it is a sensible reason to keep lifting and keep your protein up if you compress your eating hours.

If you work shifts, read this bit

A lot of my clients are nurses, drivers, security, F&B. The advice you get online about eating windows is written for people who work office hours, and until recently there was almost nothing tested on real shift workers over a real timeframe.

There is now. The SWIFt trial, published in eBioMedicine in 2025, randomised 250 real night-shift workers with obesity across 24 weeks, comparing continuous calorie restriction against two different intermittent fasting schedules, one on days off and one on shift nights. Among those who completed, weight change was 5.3 kg on continuous restriction, 7.9 kg on the days-off fasting protocol and 5.2 kg on the shift-night protocol. But the between-group analysis across everyone randomised found no significant differences at all, with fasting coming out 0.2 kg apart from continuous restriction.

So no, fasting through your night shift is not the evidence-based answer. The only long real-world trial in your situation found it did not beat plain daily calorie restriction.

One more thing, and then my actual advice

Please do not treat fasting as medicine. An umbrella review by Sun and colleagues in eClinicalMedicine in 2024, covering 23 meta-analyses and 351 associations, found that only 10 percent of the significant associations were supported by high-certainty evidence, and that compared with continuous restriction, intermittent fasting came out worse for systolic blood pressure. Nothing in this article treats, prevents or manages any medical condition. If you are on medication, pregnant, living with diabetes or kidney disease, or you have any history of disordered eating, talk to your doctor before you change when or how much you eat.

My actual advice is unglamorous. Pick the eating pattern you can run for a year. If skipping breakfast makes your day simpler and stops you snacking at 11pm, fast, and I will help you get your protein and your training right inside that window. If skipping breakfast makes you eat a whole packet of biscuits at 3pm, do not fast, and we will build the deficit a different way.

The window is a tool for getting to the deficit. The deficit is the thing.