Incidence of disorders of spermatogenesis in middle aged Finnish men, 1981-91: two necropsy series

BMJ 1997; 314 doi: (Published 04 January 1997) Cite this as: BMJ 1997;314:13
  1. Jarkko Pajarinen, doctora,
  2. Pekka Laippala, professorb,
  3. Antti Penttila, professora,
  4. Pekka J Karhunen, professorc
  1. a Department of Forensic Medicine, University of Helsinki, PO Box 40, 00014, Finland
  2. b Department of Biometrics, University of Tampere, Finland
  3. c School of Medicine, University of Tampere, Finland
  1. Correspondence to: Dr Pajarinen
  • Accepted 16 October 1996


Objective: To investigate if the incidence of disorders of spermatogenesis and testicular tissue morphology have changed in middle aged Finnish men over 10 years.

Design: Two necropsy series completed in 1981 and in 1991.

Setting: Department of Forensic Medicine, University of Helsinki, Finland.

Subjects: 528 men, aged 35 to 69 years, subjected to medicolegal necropsy.

Main outcome measures: Scoring of spermatogenesis and morphometric analysis of testicular tissue components. Individual risk factors for testicular disorders obtained by postmortem blind interviews with acquaintances.

Results: Normal spermatogenesis was found in 41.7% of the men (mean age 53.1 years). Between 1981 and 1991, the ratio of normal spermatogenesis decreased significantly (odds ratio 3.5; 95% confidence interval 2.5 to 5.1) from 56.4% to 26.9%, with a parallel increase in the incidence of partial and complete spermatogenic arrest (2.1; 1.4 to 2.9 and 2.9; 1.7 to 5.0, respectively). During this period, the size of seminiferous tubules decreased, the amount of fibrotic tissue increased, and the weight of testicles decreased significantly. Alterations in testicular characteristics over time could not be explained by changes in body mass index, smoking, alcohol drinking, or exposure to drugs.

Conclusions: The incidence of normal spermatogenesis decreased among middle aged Finnish men from 1981 to 1991, and the incidence of disorders of spermatogenesis and pathological alterations in testicles increased. Deteriorating spermatogenesis may thus be one important factor in the explanation of declining sperm counts observed worldwide.

Key messages

  • Several recent reports have suggested a significant decrease in human sperm counts over the past few decades

  • We used middle aged subjects, with no biasing selection as regards fertility or status of spermatogenesis, to evaluate changes in the incidences of disorders of spermatogenesis from 1981 to 1991

  • Normal spermatogenesis was observed significantly less often in the 1991 series than in 1981, whereas the incidences of disorders of spermatogenesis increased significantly during that time

  • Changes in the status of spermatogenesis coexisted with decreased testicular weight, smaller seminiferous tubules, and increased fibrosis of testicular tissue

  • Alterations in spermatogenesis could not be explained by a change in individual risk factors between the series, such as smoking, drinking, or use of medication, thus challenging further research to illuminate specific reasons for deteriorating spermatogenesis and declining sperm counts


In 1992 Carlsen et al reported a significant decrease in the quality of human semen during the past 50 years, including a deterioration in mean semen volume and mean sperm concentration in semen from voluntary sperm donors.1 Although contradictory and critical results have been published,2 3 4 5 this observation has subsequently been corroborated by several reports additionally suggesting that a similar deterioration may have taken place in the morphology and motility of sperm.6 7 8 The reasons for the declining quality of semen and sperm are subjects of current research. Several environmental toxins and chemicals such as alcohol,9 drugs,10 industrial solvents,11 and endogenous and exogenous oestrogen-like compounds12 have been suggested to alter the function of male reproductive organs via multiple mechanisms, including endocrinological disorders and direct toxic damages on gonadal cells.13 14

Thus far, most reports suggesting a deterioration of human semen and sperm during the past decades have analysed the quality of semen from voluntary sperm donors and men attending infertility clinics. Nothing is known, however, about the changes in quality of sperm from middle aged or older men, who may have been exposed to environmental and other risk factors for a longer time than those younger subjects in previous studies. In addition to high selection of subjects in previous studies, which possibly biases the outcome, semen analysis may not relate directly to the severity of testicular lesion at the level of spermatogenesis,15 16 although it reflects well the overall function of the male reproductive tract. Therefore, to what extent changes in semen quality can be explained by a deterioration of spermatogenesis or some other disorder(s) in male sex organs is undetermined. Alterations in semen quality of donors may even be unrelated to spermatogenesis, reflecting other factors such as changes in the frequency of coitus, or be due to non-toxic exogenous means such as variations in scrotal temperature because, for example, differences in types of clothing. Moreover, it is not known if changes in some known individual risk factors might be associated with declining sperm counts. In this study, we investigated possible changes in incidence of disorders of spermatogenesis and testicular morphology in two necropsy series between 1981 and 1991, comprising middle aged men with no selection as regards spermatogenesis and fertility. Data on exposure to risk factors for testicular disorders were obtained by interviewing close relatives or friends of the deceased.

Subjects and methods

Necropsy series

We examined two necropsy series comprising middle aged men who were subjected to medicolegal necropsy at the Department of Forensic Medicine, University of Helsinki. The capital Helsinki and the surrounding Uusimaa province in southern Finland are the most densely populated parts of Finland, with nearly one fifth (1 million) of the population of the country. About 2500 medicolegal necropsies are performed annually in this area. The first series was collected in 1981 and the second in 1991, both as substudies of a study of changes in the risk factors for sudden and violent death. Both series initially comprised 264 consecutive men aged 35 to 69 years (mean 53.1). More than half of the men (65%) died from disease (table 1), cardiovascular diseases being the most common. One third (33.7%) died violently or due to intoxication, most of them accidentally or by committing suicide. In seven cases (1.3%) the cause of death remained unknown. By covering 42% of all deaths of men aged 65 years or less in this area, this series is the most representative sample of middle aged men that can be obtained.

Table 1

Features of necropsy series. Figures aremeans (SD; 95% confidence intervals) unless stated otherwise

View this table:


Testicles were weighed at necropsy, and the middle section of one testis per man was fixed in 10% buffered formalin solution or in Bouin's solution. Histological sections (5 µm) were prepared after samples were dehydrated and embedded in paraffin and were then visualised by Herovic staining.

Spermatogenesis score

In each slide spermatogenesis was scored in 25 randomly chosen cross sections of seminiferous tubules by one member of the study group (JP) with a light microscope at a magnification of x200. The analysis was performed without knowledge of interview data or other features and series of the subjects. Each sample was allocated to one of the following groups.

Normal spermatogenesis–Germinal epithelium was normal in most tubules. Sertoli cells, spermatogonia, spermatocytes, spermatids, and spermatozoa were all apparent in the tubules with the epithelium appearing thick and condensed. Men showing modest hypospermatogenesis, characterised by a thin germinal epithelium with diffusely reduced cell quantity at all stages of spermatogenesis, were included in this group.

Spermatogenic arrest–All tubules contained at least Sertoli cells and spermatogonia. No mature spermatozoa were observed, suggesting an arrest in the normal spermatogenesis. Complete spermatogenic arrest was defined as a state in which all tubules exhibited arrest of spermatogenesis, whereas for partial spermatogenic arrest some of the seminiferous tubules (> 5%) showed normal spermatogenesis.

The Sertoli cell only syndrome–A complete absence of germinal cells was observed. Only Sertoli cells remained in most tubules, and in some areas tubules were obliterated by fibrosis. In some cases with complete spermatogenic arrest some of the tubules showed only Sertoli cells, but these cases were grouped on the basis of the main features of the section.

Fibrotic tubules–All seminiferous tubules were totally obliterated by fibrotic tissue with both germinal cells and Sertoli cells missing.

Morphometric analysis of testicular tissue component

Measurement of the percentage area covered by fibrotic tissue and Leydig cells was performed with a light microscope equipped with an ocular grid with 100 defined elements. The number of these elements situated on each of the testicular tissue components was then calculated for one of three randomly chosen cross sections of testicular tissue from each man at a magnification of x40. An average percentage value for tissue components was calculated. Mean diameters of seminiferous tubules were measured in a proportioned segment of line on the microscope view on a computer screen. Ten randomly chosen circular cross sections of seminiferous tubules were measured, and an average diameter was calculated.

Scoring of spermatogenesis, morphometric analysis of areas of tissue components of testis, and measurement of the diameter of seminiferous tubules were performed blind to interview data.

Interview study and risk factors for testicular disorders

To evaluate the role of various individual risk factors for altered spermatogenesis a relative or a close acquaintance of the subject was contacted and interviewed (table 2). The interview comprised a structured set of more than 50 detailed questions on occupation, use of medications, and smoking and drinking habits.17

Table 2

Details of risk factors for testicular disorders obtained through interviews with acquaintances in two necropsy series. Figures are numbers (percentages) of subjects

View this table:

A complete forensic toxicological examination was performed for all men, including a determination of blood and urine alcohol concentration at the time of necropsy and analysis of the presence of drugs and chemicals in blood, stomach, and liver.

Statistical analysis

The data on spermatogenesis were analysed in cross tabulated form by using odds ratios with 95% confidence intervals (CIA software). Confidence intervals excluding the value of 1 were considered to be significant. When analysing testicular features we used two way analysis of covariance. We used covariate structure and adjustment of means to eliminate the confounding effect, when necessary. Post-hoc comparisons were made with Sheffe's test. The data were analysed in logarithmic and square root transformed form, but the results were identical with those based on crude data. The analysis was carried out with Statistical for Windows (version 5.0) on a 486 PC.


Necropsy series

There were no significant differences in the causes of death between the two series. The interval between death and necropsy was slightly longer in the 1991 series, in which the men were also slightly younger and had a significantly higher mean body mass index compared with those in the 1981 series. No significant differences were observed between the occupations of men in 1981 and 1991.

Changes in spermatogenesis score and testicular morphometric characteristic

Normal spermatogenesis was found in 41.7% (220) of all the 528 men. In 1981, 56.4% (148) of the men showed normal spermatogenesis, whereas in 1991 the incidence had decreased by more than a half (29.5% percentage points) to 26.9% (71) (odds ratio 3.5; 2.5 to 5.1; table 3). Simultaneously, the incidence of partial spermatogenic arrest increased from 31.4% (83) to 48.5% (128) (2.1; 1.4 to 2.9) and that of complete spermatogenic arrest from 8.0% (21) to 20.1% (53) (2.9; 1.7 to 5.0). There were 15 cases of the Sertoli cell only syndrome and eight cases of fibrotic testicles. No significant change was observed, however, in their incidences between 1981 and 1991.

Table 3

Comparison of status of spermatogenesis (numbers (percentages)) between series (1981 v 1991) among all subjects

View this table:

Testicular weight showed a significant decline from an adjusted mean of 18.9 g in 1981 to 17.8 g in 1991 (table 4). The adjusted mean diameter of seminiferous tubules and morphometrically measured percentage area of seminiferous tubules were also smaller in 1991, with a corresponding increase in the percentage area of fibrotic tissue in the 1991 series.

Table 4

Comparison of status of testicular features between series (1981 v 1991) among all subjects. Figures are adjusted means (SD; 95% confidence intervals) unless stated otherwise

View this table:

Risk factors for testicular disorders

Alcohol–The proportion of men with reported moderate and heavy drinking did not change significantly between 1981 and 1991. Disorders of spermatogenesis increased significantly between 1981 and 1991 among both moderate and heavy drinkers (table 5). In both series normal spermatogenesis was less common in heavy drinkers than in moderate drinkers in corresponding series. Testicular weight showed a slight decrease in both consumption groups in both series (table 6).

Table 5

Comparison of status of spermatogenesis among moderate and heavy drinkers between series (1981 v 1991). Values are numbers (percentages) of subjects

View this table:
Table 6

Comparison of status of testicular features among moderate and heavy drinkers between series (1981 v 1991). Figures are adjusted means (SD; 95% confidence intervals) unless stated otherwise

View this table:

Drugs–Of the men with interview data available, 122/260 (46.9%) used sedative or tranquillising drugs, whereas 20 (7.7%) were users of antihypertensive medication. The proportion of men using these drugs, however, did not change from 1981 to 1991.

Smoking–There were 184 (70.8%) reported smokers among the 260 men for whom data were available. Smoking was slightly more common in 1991, but no difference was observed in the status of spermatogenesis between smokers and non-smokers.

Body mass index–There was a significant increase in the mean body mass index of the men between 1981 and 1991. Body mass index, however, showed no statistical association with the occurrence of testicular disorders, although it was a significant covariate for testicular weight, diameter of seminiferous tubules, and testicular morphology.

Pooled risks–Disorders of spermatogenesis were more common in men with pooled risk factors for altered spermatogenesis (heavy drinking and smoking and use of tranquillisers) than in men with none of the mentioned risk factors (odds ratio 3.2). This difference, however, was not significant because of the small numbers of cases in both groups.


Declining sperm counts

Carlsen et al recently reviewed several reports published from 1938 to 1991 on semen quality of voluntary sperm donors and suggested that sperm counts may have decreased by 42% during the past five decades with a concurrent slight decrease in volume of semen.1 Although criticism has been directed against statistical methods used in that work 2 3 and controversial reports have been published,4 18 19 including on the semen quality of Finnish men,5 20 several results corroborating the observation of Carlsen et al have been published in recent years.6 7 8 The hypothesis of deteriorating function of the male reproductive tract has thus far been based on findings observed in quality of semen and sperm, which reflect well the overall function of male reproductive organs. Meinhard et al, however, reported normal spermatogenesis or blockage of seminiferous tubules, a status in which spermatogenesis proceeds normally, in half of 100 infertile men who were oligospermic.15 Similarly, in a recent study, fine needle cytology indicated normal spermatogenesis in almost one third of 534 azoospermic men.16

Thus, semen analysis may, in fact, not be a good guide to the severity of testicular lesion at the level of spermatogenesis, and therefore it is questionable whether or not the declining changes in sperm counts are due to increases in disorders of spermatogenesis or whether other disorders in the male reproductive tract are involved. In the present study, we found that the incidence of normal spermatogenesis has decreased significantly among middle aged men, with a parallel increase in the rate of disorders of spermatogenesis during an interval of 10 years between 1981 and 1991. This finding suggests that the quality and dispatch of spermatogenesis are deteriorating in middle aged men and also confirms earlier presumptions on deteriorating spermatogenesis being the main cause of decreasing sperm counts.

Sperm counts in Finland

Vierula et al reported recently that sperm density and total sperm counts of Finnish men have not gone through any changes in the past 28 years,5 confirming earlier findings on good semen quality of Finnish men.20 Although we found a significant decrease in the incidence of normal spermatogenesis in the past 10 years, our results do not inevitably disagree with those of Vierula et al. By definition, testes expressing partial spermatogenic arrest, a slight and probably reversible disorder which was most commonly found among men in 1991 series, still produce various amounts of mature spermatozoa. Therefore, if semen analysis is performed after a few days of abstinence before sperm donation, men with partial spermatogenic arrest may not differ significantly from those with completely normal spermatogenesis. The subjects in our study, however, were on average 20 years older than those of Vierula et al, and thus the results are not fully comparable. In addition, our material came from the densely populated province of Uusimaa, mostly in and around Helsinki, whereas that of Vierula et al originated from less densely populated and industrialised communities several hundred kilometers north east of Uusimaa province. Recent work from France has shown a significant decline in sperm counts during an interval of 20 years between 1973 and 1992,6 possibly suggesting that geographical variations do exist in the abundance and distribution of factors causing disorders of male reproductive tract, as also hypothesised by Vierula et al. Two recently published reports also found unchanged sperm counts in the United States over the past few decades, suggesting that deterioration of semen quality is not geographically uniform.18 19 Also, the fertility of Finnish couples was recently observed to be significantly greater than that of British couples, further supporting this hypothesis.21

Disorders of male reproductive tract

In addition to declining sperm counts, the incidences of testicular cancer and specific disorders of male reproductive tract have increased in recent decades.22 23 24 25 Both observations have provoked an active discussion of a possible common aetiology. Several environmental factors are known to cause alterations in the male reproductive tract, such as certain drugs, chemicals,10 11 and heavy metals as well as excessive drinking.26 The role of the increase in environmental oestrogen as a possible common denominator to the adverse effects on male reproductive tracts has achieved much emphasis during recent years.12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 The suggested increase of oestrogens or oestrogen-like compounds in past decades may originate from several sources, such as from diet,30 which has gone through many changes in industrialised countries, or from the increasing use of many organochlorides that act like oestrogens31 and may accumulate in fat tissue. It has also been hypothesised that fat tissue may convert certain steroids to oestrogens and that increasing body fat content may lead to an increase in the bioavailability of oestrogens through a decrease in the concentration of sex hormone binding globulin. In our work, men in the 1991 series had a significantly higher mean body mass index compared with men 10 years earlier, possibly indicating higher body fat content of men in the latter series. Body mass index, however, showed no association with disorders of spermatogenesis.

The mechanisms of declining sperm counts induced by oestrogen or toxins may entail a disturbance in prepubertal multiplication of Sertoli cells, possibly inducing a decrease in their number. Sertoli cells are involved in the formation of the blood-testis barrier and are known to provide nutritional and mechanical support to spermatogenic cells, thereby sustaining normal spermatogenesis.32 In the present study, small seminiferous tubules were associated with disorders of spermatogenesis, corroborating our own earlier findings9 and suggesting a disturbance in the function of Sertoli cells or a decrease in their number as a possible aetiological mechanism for declining sperm counts.

Risk factors for altered spermatogenesis

Excessive use of alcohol is often associated with disorders of spermatogenesis and testicular damage.33 34 In the present study, we analysed time dependent changes of spermatogenesis over a 10 year period among moderate and heavy drinkers and examined changes in smoking and use of drugs to evaluate the effect of possible changes in the exposure to these individual risk factors. The time differences in spermatogenesis scores were of greater extent among moderate drinkers than among heavy drinkers. This observation is likely to exclude increasing alcohol consumption in Finland between 1981 and 199135 as an explanatory factor to widespread deterioration of spermatogenesis. In 1991 moderate drinkers showed incidences of scores of spermatogenesis almost equal to those among heavy drinkers 10 years earlier. Similarly, there was a highly significant increase in the amount of fibrotic tissue in testes of men with moderate drinking habits, whereas no such time dependent change was observed among heavy drinkers. These results suggest that heavy drinkers showing normal spermatogenesis and testicular morphology might be more tolerant to toxic stress originating from the environment. We have also recently reported a possible involvement of a genetic component in the development of alcohol induced disorders of spermatogenesis.36

There were more smokers in the 1991 series compared with 1981. A large number of studies have produced controversial results on the effects of smoking on semen and sperm.37 38 39 40 41 In the present study, no association was observed between smoking and status of spermatogenesis. Additionally, there were no significant changes in the ratio of men who used drugs between 1981 and 1991.


In conclusion, we report a deterioration in male reproductive function at the level of spermatogenesis among middle aged Finnish men between 1981 and 1991. This finding supports several previous observations on declining sperm counts, suggesting that both events might be explained by common genotoxic factors that affect spermatogenesis. Declining spermatogenesis was not explained by different exposure to drugs or incidence of smoking or alcohol consumption, and thus more research is clearly needed to evaluate the roles of different factors for declining sperm counts and deteriorating spermatogenesis.


Funding: Yrjo Jahnsson Foundation, Finnish Foundation for Alcohol Studies, and Medical Society of Finland.

Conflict of interest: None.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
View Abstract