Idiopathic Male Infertility: A Quarter Have Low Testosterone

In about four out of ten cases of male infertility, no cause is ever found, and the file is closed with a single word: idiopathic. A 2024 case-control study in Endocrine examined 214 such men and found that almost a quarter had a total testosterone below the level scientific societies use to call a man normal.

What the Word Idiopathic Is Actually Doing

Male factor is involved in about 50 percent of infertile couples of reproductive age in Western countries, and in roughly 40 percent no cause is ever identified. Those men become idiopathic.

The definition is precise and entirely negative. At least one semen parameter must fall below the World Health Organization decision limits, no recognizable cause may be found, and the gonadotropins, meaning the two pituitary hormones that drive the testis, follicle stimulating hormone (FSH) and luteinizing hormone (LH), must sit inside the reference range. A diagnosis of exclusion is only ever as good as the exclusion.

Idiopathic is not a mechanism. It is a record of where the search stopped.

How the Study Was Built

Cases were 214 men with idiopathic infertility, mean age 38.2 years with a standard deviation of 6.2, seen at the Andrology Unit in Modena, Italy between June 2016 and June 2023 in a retrospective real-world study. Each was the male partner of a couple who had not conceived after twelve months of unprotected intercourse, with at least one abnormal semen parameter, FSH between 1 and 12 IU/L, LH between 1 and 9 IU/L, and testosterone above 2.1 ng/mL. Men below 2.1 ng/mL were excluded, since that value identifies hypogonadism beyond doubt: the authors wanted the ambiguous middle, not the obvious cases. Chromosomal alterations, Y chromosome microdeletions, CFTR mutations, varicocele, urogenital infection, obstruction and other endocrine disease were excluded as well. Of the 214, 190 men (88.8 percent) had semen abnormalities and 24 (11.2 percent) were azoospermic, meaning no sperm at all in the ejaculate.

Controls were 224 men with normozoospermia, mean age 33.7 years with a standard deviation of 7.5, referred for screening semen analysis between September 2010 and May 2022.

The Comparison That Should Change the Conversation

Blood was drawn fasting at 8:00 am. Mean total testosterone was 5.2 ng/mL in cases, standard deviation 2.0, against 5.7 ng/mL in controls, standard deviation 1.5, p = 0.002. That gap alone is a footnote. The distribution is the story: neither group was normally distributed, and the case curve was shifted left.

Then the authors applied the threshold scientific societies suggest for a normal level, 3.5 ng/mL, equivalent to 12.1 nmol/L.

• 23.8 percent of cases, 51 of the 214 infertile men, fell below 3.5 ng/mL
• 4.5 percent of controls, 10 of the 224 normozoospermic men, fell below the same line
• The difference reached p < 0.001

Read that with the entry criteria in mind. All 51 had already been declared free of any identifiable cause, and all had a testosterone above 2.1 ng/mL, so none was an obvious hypogonadal case. They sat in the gray zone between 2.1 and 3.5 ng/mL, that is 7.3 to 12.1 nmol/L, and nobody had called it anything.

Why a Gray Zone Level Matters Inside the Testis

The testosterone in your blood is not the testosterone that makes sperm. Sperm production depends on intratesticular testosterone, the concentration inside the testis itself, far higher than the level in the bloodstream.

The compartments are driven separately. LH stimulates the Leydig cells, which make testosterone. FSH stimulates the Sertoli cells, which nurse the developing sperm. Three findings point the same way:

• FSH was higher in cases, 5.2 against 4.1 IU/L, p < 0.001, suggesting the sperm producing compartment is struggling and the pituitary is pushing harder
• LH did not differ, 4.0 against 3.8 IU/L, p = 0.324
• The testosterone to LH ratio was lower in cases, 1.5 against 1.9, p < 0.001, meaning the testis delivered less testosterone for the same pituitary drive

Within the cases that ratio separated more sharply still, 0.9 below 3.5 ng/mL against 1.7 at or above, p < 0.001. This is what the authors mean by functional hypogonadism: the pituitary signal is nominally within range, but the gland it is aimed at is not answering properly.

The Correlation That Appeared Only in the Low Group

Splitting the cases at 3.5 ng/mL produced the most interesting result. In men at or above the threshold, testosterone correlated with no semen parameter at all. More testosterone bought nothing.

In men below it, testosterone correlated directly with the percentage of sperm of normal shape, known as normal morphology. On multivariate stepwise linear regression this gave R = 0.430, standard error 0.3, p = 0.020, and it survived adjustment for cryptorchidism, varicocelectomy, comorbidities, drug use, smoking and alcohol (R = 0.390, standard error 0.5, p = 0.025).

An R of 0.430 is a moderate association in a subgroup of 51 men. Treat it as a signal, not a law. But the shape matters: below the threshold testosterone tracks with sperm morphology, above it it does not. For scale, normal morphology averaged 2.0 percent in cases against 10.0 percent in controls, p < 0.001, and a normal result requires above 4 percent.

Where the Study Stops, Stated Plainly

The design is retrospective, so it shows association and cannot show cause. A low testosterone here may be contributing to the poor semen quality, may be a consequence of the same testicular problem, or may be incidental.

More awkward than that, the two subgroups did not differ on semen parameters. Splitting at 3.5 ng/mL produced no significant difference in sperm concentration, total sperm number or motility, and normal morphology was slightly higher, not lower, in the low testosterone subgroup, 2.3 percent against 1.3 percent, p = 0.064. Only the testosterone to LH ratio separated them. Across the whole cohort, bivariate correlation found no significant relationship between testosterone and any hormone or semen parameter once corrected for multiple testing, and neither logistic regression predicted who had a low level, whether from clinical characteristics (Chi-squared 6.5, p = 0.588) or from hormones and semen parameters (Chi-squared 128.9, p = 0.063).

That last failure is a limitation and also the paper's practical argument. You cannot work out from the history or the semen analysis which of these men has the low level. You have to measure it.

Three further caveats the authors raise themselves. Testosterone was assayed by immunoassay, less accurate than mass spectrometry and least accurate at low values, which is where this finding lives. The 3.5 ng/mL threshold is guideline supported but unproven as the most accurate line between eugonadal and hypogonadal. And the controls had normal semen analyses but unproven fertility, and were younger than the cases, 33.7 against 38.2 years.

The Move That Looks Obvious and Is Wrong

A man reads that a quarter of these cases have low testosterone and reaches the conclusion the internet hands him within thirty seconds: get testosterone. That is precisely backwards, and it is the most important sentence here.

Exogenous testosterone suppresses sperm production. Taking testosterone raises the level in the bloodstream, and the pituitary answers by switching off LH and FSH. Once LH falls, the Leydig cells stop making testosterone inside the testis, which is what spermatogenesis actually depends on. The blood number rises while the machinery that makes sperm winds down. In a man who wants children, testosterone therapy works against the goal, which we set out separately in what testosterone therapy does to fertility.

So what does a low level change? The assessment, not the prescription. The authors argue that a testosterone in the gray zone may mark the subgroup in whom stimulating the testis with gonadotropins, rather than replacing testosterone, has a mechanistic rationale, while the other 76 percent probably carry a defect further down the spermatogenesis cascade. They are candid that this remains a hypothesis requiring purpose-designed studies.

Current practice sits loosely by comparison. Across 21 trials FSH raised pregnancy rates overall, but 10 to 18 men had to be treated for a single pregnancy: a great deal of treatment for a category the authors themselves call heterogeneous.

What This Means If You Are the Man in Question

If your semen analysis is abnormal and the cause is said to be unknown, three questions are worth asking before you accept the word idiopathic.

1. Was total testosterone measured at all, and drawn fasting in the morning? Testosterone follows a daily rhythm, which is why the samples here were taken at 8:00 am.
2. Where did the number fall? Above 3.5 ng/mL this study offers you nothing. Between 2.1 and 3.5 ng/mL you are the man it describes.
3. Was the exclusion work genuinely done? Karyotype, Y chromosome microdeletions, CFTR, varicocele, infection, obstruction, other endocrine disease.

We will not promise more than the data allow, and neither does the paper. A testosterone of 3.1 ng/mL is not a fertility treatment waiting to happen. It moves a man out of a category with no mechanism into one with a testable hypothesis, which is the difference between being managed empirically and being investigated. The groundwork of sleep, body composition, metabolic health, alcohol and the medications that quietly interfere matters wherever the number lands, and we cover it in what actually improves male fertility.

Reading the whole hormonal picture alongside the semen analysis and the history, rather than glancing at one number and filing it, is the work our specialist does, double board-certified in Internal Medicine and Endocrinology. If you were handed a word instead of an explanation, that is worth a conversation.