Male Fertility After COVID-19: What Recovers, and When

Probably not permanently. Sperm numbers do fall after COVID-19, and the fall is real enough to show up when infected men are compared with uninfected ones. But it behaves like the dip that follows any illness with a high fever, and in the men who have been followed for long enough it comes back.

The practical version: expect a temporary dip, wait about three months, then get a semen analysis. If the result is still abnormal after that, the cause is more likely to be something that was already there than the infection.

If you are reading this because you have seen claims that a COVID-19 vaccine caused the problem, the pooled evidence does not support that, and the section below sets out what was actually measured and in how many men.

Mostly the fever, as far as anyone can tell. A high temperature suppresses sperm production whatever causes it, and this was measured carefully long before COVID-19 existed.

Twenty-seven healthy men in Copenhagen, median age 24.4, gave monthly semen samples and kept a daily record of fever for 16 months. In that cohort, fever during meiosis reduced sperm concentration by 32.6% (95% CI -49.9 to -9.2). In the same cohort, fever during spermiogenesis, the final stage of sperm formation, reduced it by 35.0% (95% CI -50.5 to -14.6). Morphologically normal sperm fell and immotile sperm rose. The effect scaled with the number of days of fever, and varied a great deal between individual men.

The professional bodies made the same point at the start of the vaccine rollout. A joint statement from the Society for Male Reproduction and Urology and the Society for the Study of Male Reproduction, dated 9 January 2021, noted that about 16 in every 100 men in the Pfizer/BioNTech vaccine clinical trial had fever after the second dose, that fever can temporarily reduce sperm production, and that any such decline would be "similar to or less than if the individual experienced fever from developing COVID-19 or for other reasons".

About three months, because that is roughly how long it takes to make a sperm. This has been measured rather than estimated: 11 men with normal sperm concentrations drank deuterated water for three weeks and gave semen samples every fortnight, and labelled sperm first appeared after a mean of 64 plus or minus 8 days, with a range of 42 to 76 days.

What has been observed after COVID-19 fits that clock. In 41 men who had recovered, sampled a median of 56 days after hospital discharge, total sperm count, sperm concentration and the proportion of motile and progressively motile sperm were all lower than in men who had not had COVID-19, while vitality and morphology were unaffected. Twenty-two of those men gave a second sample a median of 29 days later: total sperm count, concentration and the number of motile sperm per ejaculate had all risen significantly, and the proportion of abnormally shaped sperm had fallen.

Further out, the picture is quieter. Twenty men who happened to have semen analyses both before COVID-19 and a mean of 8.3 months afterwards showed no significant difference in any semen parameter once age, days of abstinence, frequency of ejaculation and the presence of fever were accounted for. In the eight couples in that series who had fertility treatment both before and after the man's infection, the embryology and clinical outcomes did not differ.

There is Indian data, and it is worth reading for its limits as much as its findings. Thirty men aged 19 to 45 at AIIMS Patna gave a semen sample during COVID-19 and a second 74 days later, an interval chosen deliberately to span a cycle of sperm production. Every sample at both time points was RT-PCR negative for SARS-CoV-2. Semen parameters and the DNA fragmentation index improved at the second sampling but had not returned to the expected range. There was no uninfected comparison group, and the authors describe it as a pilot study.

No, on the evidence that exists — and this is one of the better-answered questions in the whole area, because it was asked early and answered in several independent ways.

The first prospective look was small and direct. Forty-five healthy men aged 18 to 50, median age 28, had a semen analysis before an mRNA vaccine and again a median of 75 days after the second dose. Every parameter was higher afterwards, not lower: in these participants (n = 45) median sperm concentration rose from 26 to 30 x10^6/mL, total motility from 58% to 65% of sperm (n = 45), and total motile count from 36 to 44 million. Eight of the 45 men had oligospermia at baseline and seven of those did not at follow-up. There was no control group, so the right reading is that nothing was harmed, not that anything was improved.

The pooled picture agrees. Twelve cohort studies of 914 men, comparing vaccinated against unvaccinated men, found no significant difference in semen volume, sperm concentration, total motility, progressive motility, total sperm count, total motile sperm count, total progressively motile sperm count, or morphology. The finding held across mRNA, viral-vector and inactivated vaccines, which matters, because the mRNA vaccines that dominate the Western literature are not the only types people were given.

The strongest evidence measures conception rather than semen numbers. In a preconception cohort of 2,126 couples in the United States and Canada trying to conceive, vaccination was not associated with fecundability, the per-cycle probability of conception: fecundability ratio 1.08 (95% CI 0.95 to 1.23) for the female partner and 0.95 (95% CI 0.83 to 1.10) for the male partner. In the same cohort, male SARS-CoV-2 infection within the previous 60 days gave a fecundability ratio of 0.82 (95% CI 0.47 to 1.45), and infection more than 60 days earlier 1.16 (95% CI 0.92 to 1.47). The confidence interval on the first of those crosses 1, so it is a hint of a short-term dip after infection rather than a demonstration of one.

One study did find a change after vaccination, and it is the one quoted most often out of context, so here it is in full. A retrospective multicentre cohort of 37 semen donors at three sperm banks, supplying 216 samples, found sperm concentration 15.4% lower (95% CI -25.5 to -3.9) than baseline. In that same cohort of 37 donors, total motile count was 22.1% lower (95% CI -35 to -6.6) at 75 to 125 days after a second BNT162b2 dose than at baseline. Semen volume and sperm motility were not affected at any point. By beyond 145 days the measurements had recovered. The authors attributed the dip to the systemic immune response — the same fever mechanism described above — and concluded that "long-term prognosis remains good".

It is rarely in semen. Across 24 studies of 1,589 men, SARS-CoV-2 RNA was detected in semen in 3 studies and 8 individuals, a pooled prevalence of 1.76% (95% CI 0.72 to 3.21). Every sample in the Indian series described above was negative at both time points. Transmission through semen is therefore unlikely to be a practical concern.

The claim that the virus attacks the testicles comes from receptor expression data, and that is a weaker thing than it sounds. Single-cell sequencing of adult human testes found the ACE2 receptor enriched in spermatogonia and in Leydig and Sertoli cells, which is where the idea started. But the virus needs both ACE2 and the protease TMPRSS2 to enter a cell, and a separate analysis of the same kind of data found no co-expression of the two in testicular cells, including sperm — its authors concluded that long-term effects on male reproductive function are unlikely. Both are analyses of gene expression, not observations of infection.

The one set of direct observations comes from men who died. In 11 fatal cases, testicular tissue sampled post mortem showed mild interstitial orchitis in 8, with congestion, oedema, thickening of the tubular basal membrane, fewer Leydig and Sertoli cells and reduced spermatogenesis in all. Viral antigen was found by immunohistochemistry in all 11, viral particles by electron microscopy in 4, and viral RNA by RT-PCR in only 3. That is genuine testicular damage. It is also a description of severe fatal disease, and it does not describe a man who had a week of fever at home.

A semen analysis, at least three months after the infection. Sooner than that and the sample is still reporting the illness rather than the recovery. Two samples a few weeks apart are more informative than one, because sperm counts vary substantially between samples from the same man.

If the result is normal, the question is closed. If it is abnormal after three months, the next steps are a scrotal examination and hormone tests — FSH, LH and testosterone — and a urologist or andrologist rather than more speculation about the infection. An abnormal result three months on is more often explained by something that predated COVID-19 than by it.

Many men reading a page like this have been quietly carrying the idea that they did this to themselves, either by catching the virus or by taking the vaccine. A semen analysis is an inexpensive test that usually answers the question inside a fortnight, and the answer is more often ordinary than dramatic.

If a result does come back abnormal, that still does not mean IVF is the next step. Sometimes the answer is further testing, timing, or a simpler treatment.

A good deal of what has been written about COVID-19 and male fertility, including much of what this page previously said, belongs here.

• It does not establish that COVID-19 causes lasting infertility. The longest follow-up retrieved here found no difference in any semen parameter a mean of 8.3 months after infection, in 20 men who had pre-infection baselines.
• It does not establish that the virus reaches the testis in ordinary illness. The receptor argument rests on expression data that two analyses read in opposite directions, and the only direct tissue evidence comes from men who died of COVID-19.
• It does not establish the size of the effect reliably. The meta-analysis previously cited on this page reported I-squared of 95-96% for concentration and motility, and on sensitivity analysis its sperm-concentration effect was no longer significant (SMD -1.02, 95% CI -2.16 to 0.12). The better-controlled pooled estimate still carried I-squared of 87.1% for concentration, and its authors recorded that the comparability domain of the risk-of-bias assessment was the most biased of all.
• It does not establish that COVID-19 vaccines harm male fertility. It equally does not establish that they improve it — the improvement seen in 45 men had no control group and is better read as an absence of harm.
• It does not establish whether anything persists at a molecular level in a way that matters. A proteomics study of 41 Indian men who had recovered reported 69 significantly dysregulated semen proteins against 10 control samples, and concluded an effect persists after recovery. It measured proteins, not pregnancies.
• It does not establish that the severity of the illness predicts the size of the fertility effect. The live version of this page asserted that it does; no source retrieved here tested it directly.
• It does not establish that antioxidants, supplements, a particular diet, exercise or sleep speed recovery of sperm after COVID-19. Those claims were on this page and have been removed rather than left standing.
• It does not establish anything about hormone therapy, stem cell therapy or gene editing for fertility after COVID-19. The live version of this page listed all three; none is supported by a source retrieved here.

• Infections That Harm Sperm: Which Ones Actually Matter
• How to Prepare for a Semen Analysis: Abstinence, Collection and Timing
• Semen Analysis Report — How to Read Yours, Line by Line
• What Causes Low Sperm Count?
• Heat and Male Fertility: Everyday Sources That Matter Most
• HPV and Male Fertility