Does Calorie Counting Work? The Diet Study Evidence

People who count calories tend to lose more weight in the research, and the proof behind that is thin. A 2011 review of 22 studies found that self-monitoring, mostly of food, was consistently tied to weight loss, and it rated the evidence weak. The earlier ones used descriptive designs, which watch people instead of assigning them to anything. Here is what a calorie counting diet study can and can’t tell you.

What does a calorie counting diet study actually show?

In the 2011 review, 15 of the 22 studies were about tracking food, and the paper diary was the most common method. In all but two, the samples were mostly white women. The data relied on self-report, and the authors asked for studies that establish how much tracking is enough.

The Harvey study puts numbers on the habit. It followed 142 adults through a 24-week online program with a web-based food journal. At six months, people who had lost at least 5% of their weight were logging in to the journal 2.4 times a day. Those who lost less logged in 1.6 times. Time per day fell from 23.2 minutes in month 1 to 14.6 minutes in month 6. Among people still logging at month 6, time spent didn’t differ by weight loss.

Motivation clouds the result, since motivated people log more often and also do other things that help. The data can’t separate the two.

Decide before you start how many entries a day you can keep up, and choose a number you could still meet on your worst day.

Does the type of diet matter?

Less than diet books suggest. The DIETFITS trial randomized 609 adults to a low-fat or a low-carbohydrate diet for 12 months. Weight change was -5.3 kg on the low-fat diet and -6.0 kg on the low-carbohydrate one. The 0.7 kg difference was not statistically significant. Nobody was given an explicit calorie limit, yet both groups reported eating roughly 500 to 600 kcal a day less than at the start.

A second trial asked whether eating only between 8 a.m. and 4 p.m. adds anything. In the Liu trial, 139 adults followed a calorie-restricted plan for 12 months, with the eating window or without it. Calorie restriction alone gave a mean loss of 6.3 kg. The window group lost 8.0 kg, and the 1.8 kg gap was not statistically significant. The researchers set the calorie targets in that trial, so don’t read them as advice for you.

What I take from these two trials: when the food mix and the clock move the result this little, the part every group shared, eating less than before, carries the effect. That is my inference, and neither trial tested it. Low-carb trials land in the same place, as Atkins vs calorie counting diet shows, and calorie counting vs the Mediterranean diet finds the two work together rather than compete.

How wrong is a calorie count?

Sometimes far off, sometimes close. One lab study followed 10 adults who said they ate under 1,200 kcal a day and weren’t losing weight. Over 14 days it worked out what they really ate. They had underreported their intake by 47% on average and overreported their exercise by 51%. The group was tiny and picked for a history of “diet resistance”, so it doesn’t describe everyone. It does show how far a self-report can sit from the measurement.

Menus are a different case. A lab study of 269 restaurant food items found no significant gap between stated and measured calories on average, only 10 kcal per portion. But 50 of the items, 19%, held at least 100 kcal more than the menu said. That’s the tail to watch.

Your own scoops are an error you control. Here is what eyeballing can cost, worked from USDA records for three foods that are easy to overpour.

Food	Weighed portion	Eyeballed portion	Extra
Peanut butter, chunk style, 589 kcal per 100 g	32 g (2 tbsp), 188 kcal	50 g, 295 kcal	+107 kcal
Olive oil, 884 kcal per 100 g	13.5 g (1 tbsp), 119 kcal	27 g, 239 kcal	+120 kcal
Brown rice, cooked, 123 kcal per 100 g	202 g (1 cup), 248 kcal	250 g, 308 kcal	+60 kcal

The eyeballed weights are my own picture of a generous scoop, not measured averages. Add the three and the day is 287 kcal off. Weigh the oils, spreads and starches for one week, because a small slip in grams becomes a big one in calories there. How to track calories shows a routine for it that doesn’t eat your evening.

Does an app, a paper diary or a website work better?

In one small pilot, the app kept people logging longest. A 6-month pilot trial gave 128 adults a phone app, a website or a paper diary. The app group recorded food on a mean of 92 days. The website group managed 35 days and the diary group 29. At six months, 93% of the app group was still in the trial, against 55% of the website group and 53% of the diary group.

Weight didn’t line up neatly with logging days. Mean change was -4.6 kg for the app, -2.9 kg for the diary and -1.3 kg for the website. The website group logged more days than the diary group and still lost the least. It was a pilot with one app built by the researchers, so it can’t rank products. It does suggest the format changes how long people keep going.

If you’ve logged fewer than half the days in your first month, I’d change the format before I blamed the plan. The pilot hints at that and doesn’t prove it.

For a starting calorie figure from your age, height, weight and activity, use the calorie calculator. It has a safe minimum built in.

Who should skip calorie counting?

Anyone for whom numbers become pressure. A survey of 493 college students found that those who used calorie trackers reported more eating concern and more dietary restraint, after accounting for BMI. It was a snapshot, so it can’t say which came first. The authors called it preliminary and suggested that for some people the devices might do more harm than good.

Pregnant, ill, on medication that affects appetite or under 18? Counting isn’t something to start without your doctor. The same goes for anyone who has ever had a hard time with food. If numbers feel like pressure, calorie counting vs intuitive eating diet sets out what each approach is linked with.

If counting calories has started to feel hard to stop, or eating feels out of control, talk to someone: ANAD in the US (1-888-375-7767), Beat in England (0808 801 0677), NEDIC in Canada (1-866-633-4220) or the Butterfly Foundation in Australia (1800 33 4673). If you’re in danger right now, call your local emergency number. Other UK numbers and crisis lines are on our crisis support page.

So does it work?

For many adults, as a short experiment, yes. The studies back a narrow claim: people who record what they eat tend to lose more. The dependable move, I think, is to make recording easy enough to repeat. None of these studies tests counting against every other approach, and the student survey shows it doesn’t suit everyone.

Try it for two to four weeks. Weigh the few foods that carry most of your calories and compare the result with your weigh-ins. If the numbers start to feel like pressure, stop and put the log down.