I remember the first time I came across the “Twinkie diet” story and thinking there had to be a catch. A nutrition professor ate snack cakes, chips and sugary cereal for weeks, lost weight, and somehow ended up with better cholesterol numbers. It sounds like the kind of story made for an argument about whether calories are the only thing that matters.
The experiment really did happen. But the version that gets repeated online tends to leave out a few important details. Mark Haub did not simply eat unlimited Twinkies and become healthier. He deliberately cut his calories, lost a substantial amount of weight, and ran a short experiment on exactly one person: himself.
What Mark Haub actually did in 2010
Mark Haub was a professor of human nutrition at Kansas State University when he began the experiment in 2010. His question was fairly simple: if someone consistently eats fewer calories than they use, how much does the nutritional quality of those calories matter for weight loss?
For about 10 weeks, Haub replaced much of his usual food with products most people would put firmly in the junk-food category. Contemporary reports mention Twinkies, Little Debbie snack cakes, Doritos, Oreos and sugary cereals. He typically ate one of these foods every few hours.
The important part was the amount. Haub kept his daily intake below roughly 1,800 calories. Reporting at the time estimated that someone of his size would have needed around 2,600 calories per day to maintain his previous weight. In other words, whatever was on the menu, he was deliberately eating in a substantial calorie deficit.
It was never literally a Twinkies-only diet
This detail gets lost surprisingly often. Haub was eating a lot of highly processed snack foods, but he was not surviving entirely on sponge cake and cream filling.
He also consumed milk, a protein shake and one or two servings of vegetables. Earlier reporting said he used a multivitamin as well.
- Snack cakes, chips and sweetened cereals made up much of the diet.
- Daily calories were deliberately restricted.
- He still included some protein, milk and vegetables.
- He used vitamin supplementation during the experiment.
That does not suddenly make it a balanced eating pattern. It does mean that “a man ate nothing but Twinkies for 10 weeks” is not an accurate description of what happened. So, please, read everything carefully before you draw conclusions.
The numbers really did move in an unexpected direction
Haub started at a body mass index of about 28.8, which falls within the overweight BMI category, and ended at about 24.9. His reported weight loss was 27 pounds.
His blood lipid results attracted even more attention. Rather than worsening during the experiment, several measurements improved.
| Measurement | Reported change |
|---|---|
| Weight | Down 27 pounds |
| Total cholesterol | 214 to 184 mg/dL |
| LDL cholesterol | Down about 20% |
| HDL cholesterol | Up about 20% |
| Triglycerides | Down about 39% |
| BMI | 28.8 to 24.9 |
Haub himself was cautious about interpreting those numbers. He questioned whether improving a handful of biological markers was enough to describe the diet as healthy. That is still the right question to ask.
The cholesterol change is less mysterious than it first appears
When I first heard this story, the cholesterol part seemed much harder to explain than the weight loss. Eat fewer calories and lose weight? That makes sense. Eat snack cakes and lower LDL cholesterol? That sounds contradictory.
But weight loss itself can affect blood lipids, which made this part of Haub’s experiment make a little more sense to me. I came across a 2020 review that looked at 73 randomized trials and more than 32,000 people, and the overall pattern was pretty clear: when people lost weight, triglycerides and LDL cholesterol tended to come down, while HDL often moved up. So Haub’s cholesterol results were unusual enough to get attention, but they weren’t completely out of nowhere.
The exact response varies between people and depends on more than body weight:
- diet composition,
- starting health,
- medications,
- activity,
- and genetics.
Still, Haub losing 27 pounds gives us a very plausible reason his lipid profile could improve even though many of the foods he chose were not foods normally associated with heart-healthy eating.
Weight loss and diet quality can affect the same health marker at the same time. Seeing a cholesterol number improve does not tell us that every part of the diet producing that change was beneficial.
What the experiment did demonstrate about a calorie deficit
I think this is the part of Haub’s experiment that is actually useful, because it clears up something that confused me for years. One particular food does not decide whether you lose weight.
If you are eating fewer calories than your body is using, you have created a calorie deficit. That is the basic idea behind weight loss, and the CDC explains it in much the same way. Eating less, moving more, or some combination of the two can create that gap.
So if you eat a cookie, some bread, or a bowl of pasta, you have not suddenly ruined your deficit. I know how easy it is to think that way, especially when certain foods are constantly being blamed for weight gain.
But there is an important second part to this. Just because a food can fit into your calories does not mean nutrition isn’t important. Calories help explain what happens to your weight. Things like protein, fiber, vitamins, minerals, and the overall mix of foods you eat tell us much more about the quality of your diet.
I find it much easier to think about those as two connected questions instead of trying to make one answer everything.
What the Twinkie experiment did not prove
Haub showed that he could lose weight while eating a calorie-restricted diet that included a lot of snack foods. Even Haub himself said he was not claiming that the foods he ate were healthy.
There are some pretty big limits to keep in mind:
- It was one person. We cannot assume everyone would see the same changes in weight, cholesterol, hunger or energy.
- It lasted only about 10 weeks. That tells us very little about eating this way for years.
- There was no comparison group. This was a personal experiment, not a controlled clinical trial.
- Better cholesterol numbers do not tell us everything about his health. Nutrient intake, muscle, digestion and long-term heart health still matter.
So I would not read this as proof that food quality does not matter. To me, it simply shows that you can lose weight in a calorie deficit even when the foods you eat are far from what most of us would call a balanced diet.
Calories matter, but they are not the only information on the label
I understand why stories like this appeal to people who are tired of being told that every food needs to be perfectly healthy. I don’t want to turn the lesson into the opposite extreme either.
If I am comparing two foods, calories tell me something important. They do not tell me how much protein or fiber I am getting, whether the serving contains much saturated fat or sodium, or whether I will still be hungry shortly afterward.
The FDA’s guidance for reading Nutrition Facts labels specifically encourages looking beyond calories at nutrients such as:
- dietary fiber,
- saturated fat,
- sodium,
- added sugars,
- vitamins and minerals.
It also gives the useful rule of thumb that 5% Daily Value or less is considered low and 20% or more is considered high for a nutrient.
That is much closer to how I think about food now: not as good or bad, but as a collection of tradeoffs.
Weight loss doesn’t always mean better health
There is another reason I would be careful about turning Haub’s experiment into diet advice. Losing excess weight can itself produce meaningful health benefits. The CDC notes that even modest weight loss may improve blood pressure, cholesterol and blood sugar.
At the same time, the 2025-2030 Dietary Guidelines for Americans recommend an eating pattern centered on nutrient-dense foods such as vegetables, fruits, protein foods, dairy, healthy fats and whole grains while reducing highly processed foods.
Those two ideas do not conflict.
You can lose weight on a diet with poor food variety if calories are low enough. You can also construct a calorie deficit using foods that provide more fiber, protein, vitamins and minerals. The second approach answers more questions than simply, “Will my weight go down?”
The part of this experiment I actually find useful
I would not copy Haub’s menu, and he did not present it as a diet everyone should follow. What I would keep from the experiment is the reminder not to give individual foods magical powers.
I used to fall into that trap myself. If something looked healthy on the package, I assumed it belonged in a weight-loss diet. If something looked like junk food, I assumed eating it meant I had somehow ruined the day. Neither idea helped much.
Haub’s experiment shows why that thinking is too simple. Portions and total calorie intake still count. So do protein, fiber, micronutrients and the overall pattern you repeat week after week.
A Twinkie can fit into a calorie deficit. That sentence is very different from saying a Twinkie-based diet is the best way to create one.