There is a number attached to this practice. Forty-five percent. It has been repeated in enough comment sections and enough community posts that it now travels without a source, the way numbers do once they get comfortable.
It comes from a study of twenty-eight men. In 2021, the journal withdrew it.
Before anyone runs with that sentence in either direction, here is the precise version: it was retracted for duplicate publication, not for fraud. Nobody caught anybody inventing data. That distinction is going to matter in a minute, and this desk is not going to blur it.
But if you have been running a streak because of that number, you should know where the number came from. And you should know what it actually described.
Three promises, and the man making them
Semen retention makes three physiological promises. Holding it raises your testosterone. Holding it protects your prostate. Holding it brings your drive back.
Before touching any of the three, something needs saying, because almost nobody in this conversation says it. The men doing this are not lazy. They are the opposite of lazy. They looked at a stretch of life where several things were visibly slipping, found one domain they could exert total control over, and exerted it. Privately. Without a prescription, without a purchase, without having to say anything out loud to anybody.
That instinct is correct. That instinct is most of the work.
What follows is not an argument against the discipline. It is an argument about where the discipline was pointed.
He was right that he had a discipline problem to solve. He was wrong about which system it was for.
Claim one: the testosterone number
Nearly the entire testosterone argument traces back to one paper, published in 2003 in the Journal of Zhejiang University. Twenty-eight volunteers. Serum testosterone measured daily across an abstinence period.
What it reported: essentially flat readings from day two through day five, a peak on the seventh day at 145.7 percent of baseline, and no regular fluctuation after the peak. The popular “forty-five percent boost” is that 145.7 percent figure, restated.
In 2021 — eighteen years after publication — the editor-in-chief issued a retraction notice. The reason was that the article substantially overlapped a paper the same authors had already published in Chinese in 2002. The authors agreed to the retraction.
Say that reason precisely, because it gets distorted in both directions. This was a retraction for duplicate publication. It was not a finding of fabrication or methodological fraud. The result was never demonstrated to be false.
It was also never replicated. It rested on twenty-eight men. And what it actually described was a one-day fluctuation that returned to baseline — not a plateau, not a new level, not something a man could hold. A spike on day seven, and then nothing.
So a man citing it as settled science is citing a withdrawn paper describing a transient blip in twenty-eight people. That has consequences, because he is building a strategy on top of it.
Grade: clearly misleading. Not because the paper claimed too much — it did not. Because the claim as it circulates asserts a large, sustained, strategy-worthy elevation, and the source supports none of those three properties.
Claim two: the prostate
This is the reason a great many men started. It also needs the most careful handling on this page, because the evidence points in an uncomfortable direction and this desk is going to refuse to draw the conclusion you are about to reach for.
The largest relevant evidence is the Health Professionals Follow-up Study: 31,925 men who answered an ejaculation-frequency questionnaire in 1992 and were followed through 2010. That is 480,831 person-years and 3,839 prostate cancer diagnoses.
Men reporting twenty-one or more ejaculations per month, compared with four to seven per month, had a hazard ratio of approximately 0.81 at ages 20 to 29 and approximately 0.78 at ages 40 to 49, with statistically significant trends holding after an additional decade of follow-up.
Now the limitations, in the same breath rather than buried at the bottom. This is observational data and it cannot establish causation. Frequency was self-reported, including recall of behaviour from decades earlier. The associations were driven by organ-confined, low-grade disease — precisely the category where screening-detection bias operates most strongly, since men who report different habits may also differ in how often they see a doctor. And reverse causation is entirely plausible, because early undiagnosed prostate disease can itself reduce frequency.
So here is what is not being said. This is not a claim that ejaculating prevents prostate cancer. That is a disease claim, it is not what the study shows, and this page does not make it.
Here is the narrower sentence, and it is enough: the best available data does not support the idea that retention protects the prostate, and if anything it points the other way. A man who took this practice up as prostate care is not getting what he was told he was getting.
Grade: unsupported.
There is a second thing worth flagging here. Retention practice rarely arrives alone. It usually shows up bundled with heavy kegel work and long edging sessions. A chronically over-tightened pelvic floor is a recognised contributor to urinary symptoms, pelvic pain and erectile difficulty. That ground is already covered in full on this site and there is no reason to re-argue it — but a man who has been squeezing for months and noticed his stream get worse should not treat that as a coincidence.
Claim three: the mechanism most men swap to
When the first two claims come apart, almost every man reaches for the same replacement. Fine, maybe it was not the semen. It was the pornography. The pornography broke my erections.
It would be a strange thing to spend two sections correcting one unsupported mechanism and then hand over a second one without checking it. So it got checked.
Integrative reviews of the observational literature report little or no evidence that pornography use as such induces erectile dysfunction or delayed ejaculation. There is a consistent cross-sectional association between self-reported problematic use and erectile difficulty — but cross-sectional data cannot establish which direction the arrow points, and no causal link has been demonstrated for any pornography variable. Reviewers note that recognised confounders, including medical conditions, medications, substance use and social factors, plausibly explain the outcome better, and that longitudinal work controlling for them does not yet exist at the necessary quality.
And then there is the finding almost nobody circulates, which is the one that matters most for the man reading this. Approaches that conceptualise pornography as causing erectile dysfunction appear to paradoxically worsen distress and erectile problems. The sexual-medicine literature describes this as an iatrogenic effect of reboot-style programmes — harm produced by the explanation itself. A man is told his hardware is broken by a specific mechanism, he starts watching for evidence of it, anxiety does what anxiety reliably does to an erection, and the belief confirms itself.
That is not a small correction. That is a correction that plausibly leaves a man better off, not merely better informed.
To be unambiguous: this is not “pornography is harmless, carry on.” Compulsive sexual behaviour is real, is associated with genuine distress, and is worth addressing with a qualified professional. The claim being graded here is the mechanical one — that it broke his erections — and that specific claim is not supported.
Grade: weak evidence. No causal link established in either direction.
Why the whole structure feels true anyway
Every one of these claims is generated by a single assumption sitting underneath them, and it is worth naming directly: that sexual function is a stock to be conserved rather than a readout of systems.
It is intuitive folk physics. Semen is a substance. Substances are finite. Conserve a finite thing and there is more of it.
That is not what is happening. An erection is a hydraulic event. It depends on endothelial health, on nitric oxide availability, on arterial inflow, on whether blood pressure and blood sugar are letting those vessels do their job. Nothing a man retains touches any of those inputs.
An erection is not a balance. It is a report. Which is why the reading changes when those underlying systems change, and why it does not change when nothing changes except what is kept.
And yet plenty of men do feel better when they start. So be exact about that, because the point here is not to tell anyone it was in his head.
A man taking this up almost never changes one thing. He changes several at once. He is sleeping earlier because he is not awake at two in the morning. He is often training again. He is expecting improvement, and expectancy is a real physiological input rather than a fake one. And he usually started from a low point, which means some of the movement is regression to the mean doing its ordinary work.
The improvement was probably real. The explanation was borrowed. Two things changed, and one of them received all of the credit.
Where the discipline actually pays
What is left is a man with a demonstrated capacity for discipline, which is not a small asset. The task is to point it at systems capable of responding.
Load-bearing training. Squat, hinge, press, carry. Two to four sessions a week, tracked and progressed. Muscle is metabolic insurance — it is where glucose goes when it is not going somewhere less useful. Of every lever on this list, this one has the strongest claim on both hormonal and vascular function at the same time.
Sleep duration, before any supplement conversation. Testosterone production is tied to sleep architecture. A man running on six hours is arguing with his own endocrine system every night, and no protocol layered on top of that gets to win the argument.
Visceral fat. The fat around the organs is hormonally active tissue and it works against him. This is a waist measurement, not a scale number. It is also the lever most men skip, because it moves slowly and nothing about it feels like a streak.
Food that supports blood flow. Nitrate-rich vegetables — beets, arugula, leafy greens — support nitric oxide availability. Dark berries and pomegranate for polyphenols. Fatty fish. Protein at every meal to hold onto muscle. This is nutritional support for the systems described above, and nothing more than that.
Blood pressure, measured at home. Seated, twice, in the morning. Erectile difficulty and blood pressure occupy the same plumbing, and most men have never seen their own numbers outside a doctor’s office.
A real testosterone number. Two draws, early morning, fasted, on a certified assay, interpreted alongside symptoms rather than instead of them. A single random afternoon test is not a measurement. This is how a man replaces a withdrawn study with his own data.
The close
Nothing here was a failure. The aim was correct and the target was wrong, and those are very different situations to be in.
The instinct — that there was a domain here worth controlling, and that controlling it mattered — was right. Most men never get that far. Keep it. Move it. Put it into the barbell, the bedtime, the blood pressure cuff and the waistline, because those systems answer back.
The body keeps score. It is simply keeping score on a different ledger than the one being audited.
Stop with the excuses. Let’s get healthy.
The vascular side of this is where the real leverage sits, and it is covered in full here — start with the free guide: Get the free blood flow guide
This article is educational content and is not medical advice. It does not diagnose, treat, cure or prevent any disease. Foods and habits described here are discussed as nutritional and lifestyle support only. Anyone experiencing symptoms, or considering a change to medication or treatment, should speak with a qualified healthcare provider.