Keto made easy

Do Calories Still Matter on Keto? Deficit Yes, Restriction No

2017-08-04

In this video

Four minutes, one question, and the answer comes in the first sentence: yes, calories still matter on a ketogenic diet.

Watching this in 2026

The distinction I was reaching for is the most useful thing in the video. A deficit is the outcome. Restriction is one way of getting there, the one where you white-knuckle a number all day.

I was arguing for reaching the same outcome by another route: let appetite come down, and the deficit follows. That idea has aged fine.

What has not aged well is how I backed it up. I described "a survey of diets" without naming it, and gave no author or year (video 8Jk5ShSUpFY). So here is what the named trials show.

The claim worth unpacking: does keto beat calories?

The strictest test of "calories are irrelevant" is a metabolic ward, where researchers control every bite. Two ward studies cover it.

With calories held equal, keto did not produce extra fat loss. That supports the 2017 answer: calories still matter.

Outside a ward, people eat what they like, and the averages come out close. DIETFITS (Gardner et al., 2018) followed 609 adults for 12 months: −5.3 kg on healthy low-fat, −6.0 kg on healthy low-carb, not significantly different. Across 13 trials of a year or more, Bueno et al., 2013 found very-low-carb ketogenic diets ahead of low-fat by 0.91 kg (95% CI −1.65 to −0.17).

Here is where my explanation and the literature sit apart. None of these papers tests my mechanism, insulin down so hunger down, directly. They measure the result. If keto works for you, the likeliest reason is that you ate less without a fight, which is the adherence argument I made a month later (video 9AGpKAfs3Yk).

Then vs now

The core line held. In 2019 I was still saying weight loss comes down to eating less, and that being less hungry is what makes eating less comfortable (video 1JbK1_ClDmY).

The carb number moved. In two videos uploaded the same day in July 2017, I told people to work their way down to 20 g (video DoSUgIr2rNo) and said under 50 g and nothing else matters (video xyEZL0HuejI). By 2018 I was giving 50 as the answer while calling it unpopular (video UygKGoP_BtA).

The line that holds up, and the one to check

Holds up: "calorie deficit matters but calorie restriction is not advisable" (from the video). Eating to full while cutting carbs is a reasonable first month for most people.

Worth checking: "protein requirements are based on lean muscle mass." Some sports-nutrition reviews do work that way. Helms et al., 2014 gives 2.3–3.1 g per kg of lean body mass, but for lean, resistance-trained people in a deficit. The reference intakes for ordinary adults are per kg of body weight. Both ways are laid out in how much protein and body weight vs lean mass.

Who should not just "eat until full" on their own: if you have a history of an eating disorder, bring in your therapist and your doctor before changing how you eat. I said that in another video too (video xyEZL0HuejI). If you take medicine for diabetes, talk to your doctor first. Cutting total carbohydrate has the most evidence for improving blood sugar (American Diabetes Association, Evert et al., 2019), and your doses were set for the way you ate before.

Key points