The most viral testosterone claim in fitness runs on evidence older than most of the people sharing it. Here's what the data actually shows.
The six studies behind the testosterone scare were published between 1979 and 2005. Five were never randomized. The researchers' own conclusion: large trials are needed before anyone should change their diet.
A meta-analysis did find that low-fat diets lowered testosterone in men. Six crossover studies, 206 participants, and a result that crossed the statistical significance bar. That much is true, and it's the part the fitness internet kept.
Here's the part it didn't.
Every study behind that finding was published between 1979 and 2005. The newest evidence in the most-shared testosterone claim of the past five years is older than the people sharing it.
The testosterone scare that dominates fitness forums was built on evidence that applies to almost nobody reading them.
- The 'low-fat' diets in the studies averaged 54 grams of fat per day on a 2,500-calorie diet — with one study at just 19 grams.
- A larger 2025 review measured testosterone and six other sex hormones across 11 randomized trials — and found none of them changed significantly on low-fat diets.
- The evidence suggests caution only below about 20 to 25 percent of daily calories from fat — a range most people never reach during a normal cut.
- The research behind the scare tested only men, mostly of European descent, on short-term diets lasting two to ten weeks.
1979 to 2005
The meta-analysis was published in 2021 by Joseph Whittaker and Kexin Wu at the University of Worcester. They pooled every eligible crossover study they could find, six of them, spanning 206 men with a mean age of 46.
Those six studies shared something beyond the topic. Not one was published after 2005. Five of six were never randomized. And the diets they tested were nothing like the ones their results get applied to.
The average low-fat diet in those studies meant about 54 grams of fat per day, roughly 19.5 percent of total calories on a 2,500-calorie diet. That's not a moderate cut.
One study (Reed, 1987) dropped participants to 19 grams of fat per day. Just 6.8 percent of total calories. If you've ever tracked macros, you know how extreme that is.
The high-fat condition in those studies averaged about 40 percent of calories from fat. The testosterone finding rests on what happens when you swing 20 percentage points in one direction, from a fat intake most people would call normal to one most would struggle to maintain.
And the result the internet turned into a scare? It was a P value of 0.04, meaning the finding had roughly a 1-in-25 chance of being statistical noise. The confidence range ran so close to zero that a single additional study could tip the conclusion the other way.
The study's own authors said it plainly: large randomised controlled trials are needed to confirm these findings before practical recommendations can be made.
If your fat intake runs somewhere around 25 to 30 percent of your calories (the range most lifters land in on a cut), the diets that produced the testosterone drop in these studies look nothing like yours.
Half a Nanogram
While the 2021 meta-analysis was circulating through fitness forums, another team was asking the same question with more data.
In 2025, Soltani and colleagues published a systematic review of 11 randomized controlled trials with 888 participants [1]. They measured the effect of low-fat versus high-fat diets on testosterone and six other sex hormones.
The testosterone difference: 0.49 nanograms per deciliter. Not 10 percent. Not 15 percent. Half a nanogram, with a confidence range spanning nearly 28 nanograms in both directions.
Every other sex hormone they measured, all six of them, came back the same way. None moved significantly.
The finding that started the scare disappeared in broader data [1].
When the researchers looked at just European and North American men, the testosterone drop didn't weaken — it got stronger, and the disagreement between studies disappeared entirely. For the population most likely reading this page, the evidence for the effect at extreme fat intakes is actually the cleanest in the dataset.
What Neither Study Can Tell You
The 2021 meta-analysis wasn't fabricated. The testosterone drop across four hormone measurements was consistent. Four different measures of testosterone, from blood levels to what shows up in urine, all dropped in the same direction.
The explanation made biological sense: the body was making less testosterone, not clearing it faster. And when the analysis was restricted to European and North American men, the statistical inconsistency between studies vanished.
But the 2025 study wasn't wrong either. It included broader populations (only two of 11 trials enrolled men exclusively). The sample included people with conditions that change how hormones behave, which could dilute a male-specific effect [1].
Both sets of researchers arrived at the same place: the evidence isn't strong enough for a confident answer. Whittaker called for large randomised trials. Soltani acknowledged the small number of studies and low certainty of evidence.
What does that mean for anyone trying to decide? The testosterone drop is probably real in men who push fat below 20 percent of calories for several weeks. It probably doesn't matter above that threshold.
And if that feels frustratingly open-ended, that's because the honest answer is open-ended. The researchers who built both datasets landed there. The internet's certainty was never earned.
That honesty matters. Somewhere between the researchers' caution and the internet's certainty, something broke. The finding didn't change. The way it was packaged did.
Millions of Followers, Zero Citations
Studies analyzing testosterone content on TikTok and Instagram found that the accounts driving the conversation had a combined audience in the millions. Hundreds of thousands of likes. Content dominated by supplement companies. Almost no physicians involved.
Not a single post cited scientific evidence.
206 men. Six studies. 1979 to 2005. A result that barely cleared the significance line. Authors who wrote, in their own conclusion, that the evidence wasn't strong enough for dietary recommendations.
That's the gap between the scale of the fear and the weight of the evidence behind it. A scare manufactured at industrial scale from a foundation thinner than anyone absorbing it ever knew.
If you spent time in that ecosystem (scrolling the feeds, watching the videos, internalizing the certainty), you weren't being gullible. You were caught in a machine with no incentive to show you the fine print.
A larger 2025 review with 888 participants found a testosterone difference of half a nanogram per deciliter — with a confidence range so wide the number is statistically meaningless.
The Threshold You're Already Above
If the testosterone effect exists at extreme intakes (and the evidence says it probably does, somewhere below 20 to 25 percent of total daily calories), it shows up at levels most people never reach on purpose.
On a 2,500-calorie diet, that floor sits around 56 to 69 grams of fat. On 2,000 calories, about 44 to 56 grams. If you're eating anywhere near 25 to 30 percent of your calories from fat, you're already above the line the evidence draws.
That's the resolution the internet never delivered. Not a new number to chase. Not a new rule to follow. Just context that turns a late-night scare into a non-issue for anyone whose fat intake isn't clinically extreme.
That floor is one of three separate answers to the fat question — weight, body composition, and hormones — that land on the same range from completely independent evidence.
If the follow-up worry is weight — whether eating enough fat to keep testosterone stable means gaining fat — a separate question has an answer. Thirty-seven trials across 57,079 adults found that the total effect of dietary fat on weight is roughly 1.4 kg. The testosterone floor and the weight-loss ceiling land in the same place: dietary fat is a weak lever for both.
There's a question this meta-analysis opens but can't answer: does the type of fat matter?
The low-fat diets in these studies didn't just cut total fat. They changed which kinds of fat participants ate. Whether that shift in fat types, not just the total amount, drove the hormonal change is something different research, using body-composition imaging, was built to explore.
The evidence behind the testosterone scare is real — but it only showed up at fat intakes most people never reach on purpose.
If your fat intake sits anywhere near 25 to 30 percent of your calories, the studies that found the testosterone drop tested diets that look nothing like yours. The gap between the internet's certainty and the evidence's actual scope is the entire story.
The next time a forum post or a viral video tells you low fat is destroying your hormones, the question isn't whether the research exists. The question is whether the research tested anything resembling your diet. For most readers, the answer is no.
What other research found
What this means for you
The low-fat diets in these studies averaged about 19.5 percent of calories from fat. On a 2,500-calorie diet, that's roughly 54 grams. On a 2,000-calorie day, about 43 grams.
If your fat intake runs anywhere near 25 to 30 percent of your calories, you're eating 70 to 83 grams of fat on 2,500 calories — well above the range where the testosterone drop appeared in the research. The studies that produced the scare tested conditions that don't resemble a typical cut.
The data in this meta-analysis simply doesn't speak to your diet.
This is the one intake range where the meta-analysis findings apply directly. The testosterone drop appeared at fat intakes averaging 19.5 percent of calories, with one study pushing down to 6.8 percent.
Contest prep diets and aggressive cuts can land in this territory. The research found a consistent pattern across four different testosterone measurements at these extreme intakes, all moving in the same direction.
The researchers noted the effect appeared to come from reduced testosterone production rather than faster clearance. The fat floor around 20 to 25 percent of calories is where the evidence starts to apply.
This meta-analysis studied 206 men exclusively. Every finding on this page — the testosterone drop, the fat floor, the threshold where the effect appears — comes from male participants with a mean age of 46.
The larger 2025 study included mixed-sex populations, but only two of its 11 trials enrolled men exclusively. Neither study was designed to answer the testosterone-and-dietary-fat question for women.
The honest answer for female readers: the research that fueled the internet's testosterone scare was never about you, and no equivalent evidence exists to apply it to you.
Before you change anything
This meta-analysis pooled six studies of healthy men with an average age of 46, mostly of European and North American descent. Nobody with low testosterone, nobody on medications, nobody using supplements.
The diets tested were extreme by real-world standards — fat intakes averaging 19.5 percent of calories, with one study at 6.8 percent. Most people who reduce fat for a diet never go anywhere near these levels.
Women were excluded entirely. The one non-Western sample showed opposite results. If you don't match this specific population eating at these specific extremes, the evidence wasn't collected on someone like you.
Every included study lasted between two and ten weeks. Whether the testosterone drop persists, gets worse, or disappears over months is unknown — no long-term data exists.
The researchers couldn't separate the effect of total fat from the effect of fat type. The low-fat diets changed which kinds of fat participants ate, not just how much. The amount of fat and the kind of fat changed at the same time, and nobody can tell you which one drove the result.
Five of six studies weren't randomized. The one study with proper randomization showed a smaller effect than several of the non-randomized studies.
The testosterone drop is real enough to take seriously at extreme fat intakes — four independent hormone measurements moved in the same direction, and the European/North American subgroup showed a clean, consistent effect.
But it's built on six small studies from before 2006. The statistical significance line was barely crossed. The confidence range nearly touched zero. And a larger, newer review with more than four times the participants found the effect had vanished.
This is a finding that earns caution at the extreme and nothing more. It does not earn the certainty the internet attached to it.
The low-fat diets in these studies changed more than just the total amount of fat. Some types of fat dropped sharply while others rose — and whether that shift in fat balance drove the testosterone change is something this meta-analysis couldn't separate.
Separate research took a different approach entirely. Same total calories. Two different cooking oils. MRI scans tracking exactly where the extra fat went inside the body. The answer surprised researchers who had been thinking about dietary fat as a single category.
What This Study Found
All findings from this paper, in plain language.
- Low-fat diets produced a small but real drop in testosterone in men — though the result barely crossed the line for statistical significance.
- In European and North American men specifically, the testosterone drop was stronger and more consistent across every study.
- The testosterone your body can actually use — not just the total amount — also dropped on low-fat diets.
- The body wasn't clearing testosterone faster — it was producing less of it on low-fat diets.
- A related hormone that testosterone turns into also dropped on low-fat diets, pointing to lower production rather than faster breakdown.
- Two hormones that help regulate testosterone — one that signals production and one that carries it in the blood — didn't change significantly.
- The researchers concluded that low-fat diets reduced how much testosterone the body made rather than speeding up how fast it was used.
- The low-fat diets shifted which types of fat people ate, suggesting the kind of fat might matter for hormones — not just the total amount.
- Studies with better research methods showed smaller, more consistent effects — suggesting the true testosterone drop may be at the lower end.
- One study from South Africa showed the opposite result — testosterone went up on low fat — and removing it made all other studies agree.
- The low-fat diets didn't produce meaningful weight change — less than one kilogram difference — so the testosterone drop wasn't caused by losing weight.
- The study's own authors said large modern trials are needed before anyone should change what they eat based on these findings.