STUDY QUESTION
Are infertile men with reduced semen quality at risk of a further decrease in testicular function?

SUMMARY ANSWER
Infertile men with severely reduced semen quality risk further deterioration of semen quality 15 years after treatment for infertility, and a lower baseline sperm concentration was associated with a more pronounced increase in LH and decrease in testosterone/LH ratio at follow-up.

WHAT IS KNOWN ALREADY
Male factors account for up to 50% of human infertility. The most common finding is spermatogenic failure (SgF) yet the life course of semen quality and testosterone production in such men has not been described.

STUDY DESIGN, SIZE, DURATION
A follow-up study of men with SgF was performed 15 years after the initial infertility assessment between January 1995 and December 2000.

PARTICIPANTS/MATERIALS, SETTING, METHODS
Hospital records were used to identify potential participants in the study. A total of 137 men with primary male infertility due to SgF and 70 controls with good semen quality from couples with female factor infertility who attended a tertiary referral centre were included: the participation rate was 31% and 26%, respectively. The men provided semen samples and underwent a physical examination. Blood samples were taken to measure levels of reproductive hormones (FSH, LH, testosterone, sex hormone-binding globulin, estradiol and inhibin B). Current results were compared with results from the initial assessments.

MAIN RESULTS AND THE ROLE OF CHANCE
At the time of follow up the SgF men had significantly lower Leydig cell capacity than the control group as well as much lower semen quality. For the SgF men, between baseline sampling and follow up, the median sperm concentration decreased from 1.9 to 0.6 mill/ml and total sperm count from 7.7 to 2.0 million (P = 0.019 and 0.012, respectively), and 10% developed azoospermia. Calculated free testosterone (cFT), but not total testosterone (tT) decreased in the SgF group by ~0.6% (95% CI 0.1–1.2%) per year. In the SgF group, LH increased by 1.6% (CI 0.9–2.3%) annually, and consequently tT/LH and cFT/LH ratios had decreased by 1.3% (CI 0.5–2.1) and 2.1% (CI 1.2–3.0%), respectively. The increase in LH and the decreases in tT/LH and cFT/LH ratios were more pronounced in men with lower baseline sperm concentrations.

LIMITATIONS, REASONS FOR CAUTION
We consider the case group as representative of infertile men not in need of testosterone treatment at baseline investigation, but do not have information on those that chose not to participate in the follow-up study. There were alterations in some hormone analysis methods during the follow-up period that may introduce uncertainty in interpretation of long-term changes in hormone levels despite rigorous quality control. The validity of the control group suffers from a lack of hormone values at baseline. Also, at follow-up, for practical reasons only one semen sample could be obtained, which makes the effect estimate more uncertain and there is a risk of non-differential misclassification.

WIDER IMPLICATIONS OF THE FINDINGS
Without being able to predict individual outcomes, it is prudent to consider sperm cryopreservation or advise not to postpone fertility treatment when men present with infertility due to impaired semen quality. Whether partly compensated Leydig cell insufficiency in men with SgF will eventually develop into overt testosterone deficiency cannot be determined from our study.

STUDY FUNDING/COMPETING INTEREST(s)
Aase and Einar Danielsen (Grant no. 10-001053), Nordic Research Committee (Grant no. 5109), The Kirsten and Freddie Johansen Fund, and Rigshospitalet’s Research Fund (grant no. R24-A812). There are no competing interests.