← Longevity & Supplement Guides
One Workout a Week: What the New Trial Actually Prescribed
A study from Hong Kong went around the world this week wearing a very good disguise. One workout a week melts belly fat. Once-a-week exercise matches three times a week. The trial underneath those sentences is real, it is randomised, and it is genuinely interesting — which is exactly why it deserved better than the sentence it got.
The headline travelled faster than the protocol
Here is what actually happened. Researchers at the University of Hong Kong put 315 previously inactive adults through sixteen weeks of interval training and published the result in Nature Communications. Both training groups did the same 75 minutes of hard work per week. One spread it across three days. The other did all of it on a single day. A third group sat through health-education classes.
At sixteen weeks, body composition measured by DXA, both training groups had lost fat and gained cardiorespiratory fitness, and the researchers could not tell the two schedules apart.
That is a real, useful, well-run result. It is not the result the headlines described.
What "one workout" actually involved
The published protocol is specific, and the specificity is the whole story. A single 25-minute bout was four four-minute intervals at 85–95% of peak heart rate, each separated by a three-minute active recovery at 50–70%. On a motorised treadmill. Supervised by a certified trainer.
The thrice-weekly group did one of those bouts, three times a week. Reasonable.
The once-weekly group did that same bout three times in a row, in one visit, with fifteen to thirty minutes of rest between each one.
Do the arithmetic no headline did. That is twelve four-minute intervals at 85–95% of maximum heart rate in a single afternoon, plus the recoveries, plus two long breaks — roughly two hours in the building, most of it spent either near-maximal or waiting to go near-maximal again. Performed by people who, at enrolment, were doing under 150 minutes of moderate activity a week.
"One workout a week" is true in the way that "one meal" is true of a wedding banquet.
Anatomy of the session
The compressed week, drawn to scale
Drawn from the published trial protocol. The thrice-weekly group did one 25-minute bout per visit; the once-weekly group did three, back to back. Same 75 minutes of training either way.
Eight hundred grams
Then there is the size of what all that bought. Against the control group at sixteen weeks, the once-weekly arm lost 0.8 kg of total body fat (95% CI −1.4 to −0.2) and the thrice-weekly arm lost 1.0 kg (95% CI −1.6 to −0.5).
Over sixteen weeks, that is roughly fifty grams of fat a week. Nothing melted.
This is not a criticism of the trial. It is a criticism of the verb. About a kilogram of fat over four months, from exercise alone, in people who were not asked to change how they ate, is almost exactly what the exercise literature has reported for forty years. Exercise is a mediocre instrument for moving the scale and an excellent one for moving nearly everything else.
Which is why the fitness result is the one worth your attention. Both groups improved cardiorespiratory fitness — the measure with by far the strongest link to how long and how well people live, and the one that tracks with brains that age better. It got a clause. The 800 grams got the headline.
Week 16, versus control
What “melts belly fat” looked like on the scale
The numbers, precisely
Adjusted mean difference in total body fat mass versus control at week 16: once weekly −0.8 kg (95% CI −1.4 to −0.2, P = 0.0107); thrice weekly −1.0 kg (95% CI −1.6 to −0.5, P = 0.0003). Bars are scaled to a 1.2 kg maximum for legibility. Both groups also reduced body fat percentage and waist circumference and improved cardiorespiratory fitness relative to control. Participants were aged 18 and over, BMI 23 or higher, waist 90 cm or more in men and 80 cm or more in women, and physically inactive at enrolment.
Illustrative. Fat-mass change is one outcome among several, and the trial did not include a diet arm.
"Similar" describes the study, not the body
The word carrying the most weight in the coverage is similar. It is worth being precise about what it means.
Both training arms beat the control group. Neither beat the other. But the trial was designed and powered to test each schedule against a control — not to demonstrate that the two schedules are equivalent to each other. Failing to find a difference is not the same as finding there is none, especially when the point estimates are −1.0 and −0.8, in that order, and their confidence intervals overlap almost completely.
Read honestly, the finding is this: across 315 people and sixteen weeks, the frequency you pick was not the biggest thing in the room. That is genuinely worth knowing. It is also a smaller claim than "identical," and the gap between those two sentences is where most health journalism lives.
Frequency was never the interesting variable
Here is my position, and the trial has now made it easier to hold: we spent years arguing about the calendar when the argument was always about the dose.
The observational data said as much first. Among 89,573 UK Biobank adults wearing accelerometers, Khurshid and colleagues compared people who got their 150 weekly minutes in one or two days against people who spread them out. Heart failure: hazard ratio 0.62 for the concentrated pattern, 0.64 for the distributed one. Atrial fibrillation, 0.78 against 0.81. Stroke, 0.79 against 0.83. The schedule was interchangeable. The total was not. We walked through that evidence in the weekend-warrior post.
But observational data about scheduling has an obvious weakness. People who compress their exercise into Saturday differ from people who don't — different jobs, different families, different constraints, possibly different health to begin with. You can adjust for what you measured. You cannot adjust for why someone's week looks the way it does.
Randomisation fixes precisely that, and that is the real contribution here. A question that previously had only correlational answers got its schedule assigned by coin flip, and the answer held. What survives are the two variables we keep pointing at: how hard, and how much.
What spreading it out still buys you
Now the other half, which the trial cannot speak to because it did not measure it.
A hard session leaves effects behind that do not wait politely for the next one. Insulin sensitivity rises after a bout and stays measurably elevated for somewhere between a few hours and a couple of days, depending on intensity. Blood pressure typically runs a little below baseline for some hours afterwards. Those are per-session effects. Three sessions collect them three times a week; one session collects them once, and then six days pass.
Body composition at sixteen weeks will never show you that difference. Your Tuesday might.
Second, the days you are not training are not neutral. Compressing your week does nothing about the other eleven waking hours a day, and the shape of sedentary time carries its own signal, largely independent of whether you exercised.
Third, the unglamorous one: a schedule with one slot has no redundancy. Miss it and your week is zero. Miss one of three and you still banked two-thirds of the dose. The trial escorted its participants onto supervised treadmills for sixteen weeks. Your calendar will do no such thing.
If one slot a week is genuinely what you have
Then take it. That is the honest, useful message underneath the noise, and it is the researchers' own framing — for people whose real barrier is finding three separate evenings, a single concentrated session is a legitimate option rather than a consolation prize. A few things to carry with it:
- Earn the intensity before you compress it. These participants ramped up under supervision. Twelve near-maximal intervals is a destination, not a starting point — and if you have any cardiovascular history at all, this is a conversation for your doctor before it is a plan for Saturday.
- The structure is the active ingredient, not the day. Four minutes hard, three minutes easy, repeated four times — that pattern is what produced the result, whether you run it once or three times a week.
- Notice what a compressed session costs afterwards. A very hard block close to bedtime is one of the few reliably documented ways to disturb the night that follows. If your one slot lands on a Thursday evening, put it as early in that evening as you can — the timing evidence is thinner than people claim, but that particular finding holds up.
- Do not let the other six days go empty. Not training and not moving are two different failures, and only one of them was tested here.
- Measure the dose, not the attendance. Hard minutes per week is the number that moved. Days per week never was. In the Agen app, that means watching your weekly minutes by intensity and your heart-rate recovery trend, not counting green squares on a calendar.
The bottom line
The trial is good and the randomisation is the point: schedule your 75 hard minutes however your life allows, because across sixteen weeks the distribution did not appear to matter much and the total did. The headline that travelled was a compression of a compression — 75 minutes became "one workout," twelve intervals at 85–95% of maximum heart rate became "a brisk walk," and 800 grams became "melts."
Use numbers to correct fantasy, not to replace experience. The fantasy on offer this week was that a single easy weekly session reshapes your body. The number that corrects it is twelve.
