More than 20 years ago, it was suggested that the normal reproductive development of the

Transcription

1 ARE PROBLEMS WITH MALE REPRODUCTIVE HEALTH CAUSED BY ENDOCRINE DISRUPTION? DESCRIPTIVE Correspondene to: Dr Mihael JoVe, Department of Epidemiology and Publi Health, Imperial College Shool of Mediine, St Mary s Campus, Norfolk Plae, London W2 1PG, UK m.joffe@i.a.uk Mihael Joffe More than 20 years ago, it was suggested that the normal reprodutive development of the male embryo may be aveted by exposure to oestrogens, the fous initially being on the mother s endogenous oestrogen. In 1993, Sharpe and Skakkebaek proposed the hypothesis that inreased oestrogen exposure in early life inreases the risk of the two genital malformations hypospadias and ryptorhidism, and of testis aner in adult life, as well as restriting the ahievable sperm onentration (also alled density, or more loosely, sperm ount). 1 In reent years, the fous has shifted to onern about exogenous (xenobioti) substanes. This has beome generalised beyond oestrogens to enompass evets relating to other types of hormone, leading to the onept of an endorine disrupter:...an exogenous substane that auses adverse health evets in an intat organism, or its progeny, onsequent on hanges in endorine funtion. w1 Some prefer the term endorine modulator, as some evets may be neutral or benefiial. It has been suggested that this type of mehanism has aused a deteriorating trend in male reprodutive health involving the same four end points of hypospadias, ryptorhidism, testiular aner, and sperm density throughout the world in reent deades, w1 4 although it is generally agreed that diret evidene is laking. Juxtaposition of the fragmentary human evidene with that on fish, reptiles, and other wildlife has fuelled onern in the media, and has led to a large researh evort. It is important to separate the two main questions the possible existene of a generalised deterioration in male reprodutive health, and the hypothesis that endorine disruption an ause these evets. There are other possible mehanisms: the most potent known testiular toxin in adult life is the nematoide dibromohloropropane (DBCP), an alkylating agent (like many pestiides). w5 Germ line geneti damage to either parent before oneption ould theoretially avet male reprodutive health, but this possibility has reeived little attention. The purposes of this paper are: (a) to summarise the desriptive epidemiology; and (b) to review ritially the plausibility of the endorine disruption hypothesis as a possible explanation. Evidene from toxiology and endorinology and from studies on wildlife are not diretly onsidered, although they inform the disussion. Prostati aner is exluded, as age distribution and other features indiate that it is epidemiologially distint from these four end points. Other human health end points are not reviewed, suh as thyroid funtion, female reprodution, and breast aner, whih have also been linked with endorine disruption. EPIDEMIOLOGY Oup Environ Med 2001;58: Timing of exposure For hypospadias and ryptorhidism, exposure would need to take plae before birth. Adult life end points suh as semen quality and testiular aner ould be aveted by exposures at any time over a broad span. The most likely possibilities are reent exposures, and those ourring in early life that avet long term development of the reprodutive system. The idea of an early impairment with long term evets aords well with what is known about embryoni, fetal, and infant development. Timing of an exposure is as important as its hemial, biologial or physial nature: windows of sensitivity our, whih may be relatively short lived. This formed part of the original oestrogen hypothesis, but it is also ompatible with non-endorine mehanisms. In epidemiologial terms, early impairment orresponds to a ohort evet for example, a birth ohort evet in the ase of exposure in utero. A ombination of short and long term exposures ould be relevant. The most obvious example is for testiular aner, as it is usual to distinguish initiators and promoters of arinogenesis. *281 Oup Environ Med: first published as /oem on 1 April Downloaded from on 19 August 2018 by guest. Proteted by opyright.

2 Eduation 282 * Testiular aner Epidemiologial information on aner of the testis is very reliable. As a disease of relatively young men that has unmistakable features, it is likely to be rarely missed or misdiagnosed, so that only an eyient ollating system is required to produe high quality asertainment. Good data on inidene have been available from aner registries in many developed ountries for some deades. Mortality data are also available for ertain ountries going bak 100 years, and sine the disease is invariably fatal if untreated, these are reliable for the early deades of the 20th entury. This is not true more reently, however, as ure rates are now high. In priniple, the desriptive epidemiology should be reviewed separately for eah histologial type, seminoma and non-seminoma. However, as the desriptive data are generally remarkably similar for both major types, this refinement will be ignored. Trends In reent deades, this disease has been inreasing in most if not all ountries for whih data exist. An important and often overlooked question is when this trend began. In England and Wales, mortality started rising around 1920, having been stable before the first world war (fig 1). 2 The age distribution also hanged, from one that inreased steadily with age throughout life to the now familiar pattern with muh lower rates over the age of 40 years. It may be that the disease present in the 19th entury was diverent from that whih inreased during the 20th entury, or at least that its aetiology divered. In Denmark, a ontinuous rise in age standardised inidene is observable sine aner registration began in A similar piture is seen throughout the developed world, with trebling of rates being ommon. The rise has been steady, exept for a few instanes for example, in Vitoria, Australia where the rate was transiently stable or fell slightly in the early 1970s. w6 Predisposition to testiular aner is present from an early age, probably in utero. w7 Environmental agents ould influene this predisposition. If these trends are examined in relation to the time of birth rather than of death or diagnosis, mortality started rising among men born before 1900 in England and Wales, for those who developed the young onset form of the disease (fig 1). 2 In Denmark, the inrease began in men born around In Denmark, Norway, and Sweden, rates stabilised or fell for men born during (slightly earlier for Sweden), whereas the rise was rapid and inexorable among men born from 1920 until at least 1960 in East Germany, Finland, and Poland. 4 Reent data suggest that the rates may be stabilising for Danish men born sine about w8 Spatial and ethni variation The highest inidene in the world is found in Denmark, where the lifetime risk is now almost 1%. However, the Nordi ountries do not have a uniformly high risk, as Finnish men have omparatively low rates, with Norway and Sweden in intermediate positions (fig 2). 3 The spatial pattern for testiular aner in the Nordi ountries does not resemble that of other hormone sensitive arinomas suh as those of the prostate or female breast, but is not dissimilar from that of oloretal aner in both sexes. 5 However, testiular aner is unusual in that the gradations in risk appear to follow national boundaries. 5 Other high risk populations inlude Switzerland 6 w9 and New Zealand (inluding Maoris), 6 w10 whereas the Balti Figure 1 Death rates per million from testiular aner in England and Wales, by age, in suessive deades by (A) year of death (period effet) and by (B) birth ohort. Reprodued from Davies 2 with permission of the publisher. states 3 and Afrian Amerians 6 have omparatively low rates. The tumour is rare among Chinese and Japanese men. 6 Interpretation Heredity may be relevant for international and inter-ethni diverenes, but annot explain rapid trends. Possible explanations inlude dietary hanges in maro- or mironutrients or in ontaminants or environmental exposure. The early part of the trend annot be aused by hemials introdued sine the mid-20th entury. During the earlier period there were major hanges in diet, suh as inreasing meat onsumption, and an inreasingly sedentary lifestyle. However, although these fit quite well with the epidemiologial trend, there is urrently no evidene to inriminate them ausally. Semen quality The literature on semen quality is muh more problemati. It is usually taken to inlude sperm onentration, motility, and morphology, but the time trend debate has foused mainly on the first of these. All are subjet to large degrees of within-person biologial variation or measurement error, or both. In addition, desriptive epidemiology annot be based on representative samples of the general population, as Oup Environ Med: first published as /oem on 1 April Downloaded from on 19 August 2018 by guest. Proteted by opyright.

3 Eduation Figure 2 Map of age standardised inidene of testiular aner in the ounties of the Nordi ountries, A frequeny diagram is also given. Reprodued from Møller Jensen et al, 5 with permission of the Danish Caner Soiety. partiipation rates are too low. The best evidene is from andidates for semen donation and for vasetomy; data from men in ontat with medial servies for a fertility related problem are unreliable. Trends The most ited paper is a 1992 review of the world literature that related sperm onentration to date of publiation irrespetive of loation, 7 and whih laimed a 50% deline in the mean onentration over 50 years, from /ml down to /ml. However, its major importane is that it stimulated analysis of more reliable data, in plaes where information on semen quality has been ontinuously available over a long period. Those data are less likely to have been distorted by possible hanges in the method of semen examination and/or in seletion proesses aveting the populations studied. The prinipal onlusions to emerge are that: (a) delines in semen quality have ourred in some plaes (for example, Paris (fig 3), Gent, Edinburgh) but not in others (Toulouse, Finland, and the five US ities with published data) 8 10 w11 15 ; (b) at most, the available data go bak to the early 1970s; and () where onentration has deteriorated, so usually have sperm motility and morphology. Where a deline has ourred, the findings are ompatible with a birth ohort evet epidemiologially, aveting men born sine about 1950, but this is not onlusive. There is no information on the year when the deline started, for any of the sites, nor what the pre-deline values were. As semen quality is inferior in humans ompared with other mammalian speies, it is possible that deterioration from a natural level has a muh longer history than we have the data to substantiate. A re-analysis of the hypothesis that formed the basis of the original meta-analysis attempted to address its suggested methodologial inadequaies. 11 The deline in sperm density was found to be muh steeper in Europe than in the USA; studies from elsewhere are too sparse and diverse to draw onfident onlusions. The prinipal remaining onfounding fators that ould at least partially aount for a downward trend were age, duration of abstinene, and method of speimen olletion. However, a question mark still remains over the plausibility of a hypothesis that involves grouping together all studies that originate from an area as large as the USA or western Europe, espeially as the spatial heterogeneity within eah is known to be onsiderable. w16 One way in whih this analysis an be reoniled with the more reliable single entre time trend studies is to aept that semen quality has delined in some parts of Europe, but that there has been no suh trend in the USA. Spatial variation Substantial spatial variations in sperm onentration have 10 w16 w17 been demonstrated, both in Europe and in the USA, being relatively high in New York and Finland and low in California and north western Europe, inluding Denmark and Britain. In the ase of Finland, ouple fertility is also high ompared with Britain, w18 suggesting that the higher sperm ounts there are not aused by longer abstinene (less frequent interourse). Figure 3 Linear regression of (A) sperm density, (B) mobility, and (C) morphology against year of donation for 1351 fertile men, Paris, The results were equally ompatible with a birth ohort effet. Reprodued from Auger 8 with permission of the publisher. *283 Oup Environ Med: first published as /oem on 1 April Downloaded from on 19 August 2018 by guest. Proteted by opyright.

4 Eduation 284 * Interpretation Variations in semen quality are poorly understood, and are hampered by lak of good quality data. Fousing on the hypothesis that there has been a roughly simultaneous worldwide deline obsures important spatial variation, and the ourrene of trends at diverent times in diverent plaes. If unlinked to other end points, possible explanations for these variations ould inlude an inreasingly sedentary way of life, possibly together with tight lothing, sine raising the intratestiular temperature has a potent evet on the quality as well as the quantity of sperm. w19 A dereasing duration of abstinene, perhaps as a result of inreasing aeptability of masturbation, ould aount for a deline in sperm quantity but not a deterioration in quality. Congenital abnormalities of the male genitalia: hypospadias and ryptorhidism Hypospadias is a ondition in whih the urethral opening is abnormal in position, in severe ases on the shaft of the penis. Cryptorhidism refers to the failure of the testes to desend into the srotum. Both are likely to be unreliably asertained at birth, partiularly in mild ases, and the study of ryptorhidism is further ompliated by the diyulty of distinguishing testes that have not desended from those that readily but reversibly retrat bak into the abdominal avity. The onsequene is that published data from ongenital malformation registries annot be relied upon to reflet real variations: reported time trends and diverenes between registries may both merely reflet diverenes in asertainment and reporting. 12 Self-reported data (by mothers) are similarly unreliable. Trends and spatial variation The only lear indiation of an upward trend in hypospadias is from Atlanta, Georgia, where a step inrease was observed between 1982 and 1985 in the severe form, whih is more reliably asertained. 12 w20 Reent studies in Denmark and Finland using strit riteria have shown a higher rate in w21 w22 Denmark. In the ase of ryptorhidism, a study was arried out in Oxford during the 1950s using strit diagnosti riteria, and examination of the baby boys at 3 months when the diagnosis is more reliable. 13 A subsequent study also in southern England using the same riteria found an approximate doubling of the proportion of boys having ryptorhidism at 3 months. 13 Reent studies in New York w23 and in Finland, w24 again using the same riteria, found a similar proportion to the original Oxford estimate, whereas in Denmark it was lose to the later English value. w25 Interpretation The sarity of good data is a serious problem for both these onditions. Cryptorhidism may well have inreased in England, and appears to be more prevalent there and in Denmark than in New York and Finland. Hypospadias is also more frequent in Denmark than in Finland. The question of linkage If two onditions frequently our together, it is likely that they share ommon ausal agents. Thus, the observation that aner tends to arise in the testes of subfertile men more often than expeted by oinidene 14 enourages the searh for a ausal agent that underlies both end points. However, if one auses the other, as is likely for ryptorhidism and testiular aner, w26 this inferene does not follow (fig 4); ryptorhidism does not neessarily share determinants with testiular aner in non-ryptorhid testes (over 90% of ases), and other evidene is required. The prinipal risk Figure 4 In the left hand diagram, the apparent assoiation between A and B is brought about by the onfounding effet of C. In the right hand diagram, as C auses A and A auses B, there is no need to infer any relation between C and B. Box 1: Desriptive epidemiology key points It has been suggested that the health of the male reprodutive system has been deteriorating, affeting four end points: sperm density, testiular aner, hypospadias, and ryptorhidism Testiular aner has inreased almost everywhere, the upward trend beginning in the early 20th entury or before Predisposition to aner of the testis arises in early life, possibly in utero; this may be true of impaired semen quality as well The evidene for a trend in the other end points is inonlusive, but suggests deterioration in semen quality not only sperm density as well as in ryptorhidism and possibly hypospadias For end points other than testiular aner, evidene for a trend is weak outside north western Europe, and even there the date of onset is unlear There is notable spatial variation in all the end points. In eah ase the reprodutive system of Finnish men is healthier than that of Danish men, whih is onsistent with the idea that the four are linked fator known to be shared low birth weight w27 does not ontribute to the present disussion, as its spatial and temporal variation are insuyient. Common risk fators an be inferred from epidemiologial evidene if the onditions have a similar pattern of variation. For example, the sharp ontrast between Finland and Denmark for all four end points suggests possible linkage. Simultaneous time trends for diverent onditions may also suggest a shared ausal fator that varies in a orresponding manner, allowing for latent periods, et. Thus, for an exposure ating in utero, linked trends in ongenital malformations and adult life end points will tend to be a few deades apart. However, as many fators vary with time, an observation of simultaneous variation is unreliable evidene for linkage. Endorine explanations The oestrogen hypothesis A large number of substanes have at least some oestrogeni ativity. Lists of these are available (see box), but aution is required, as many of these ompounds have important biologial ations that are not mediated by this mehanism (and also many published lists are inaurate). In addition, one needs to onsider the mother s endogenous estradiol. The soure of exposure is related to the question of timing. In utero exposure (neessary for the two end points that are present at birth) ould be to endogenous or exogenous oestrogens; in addition, exogenous oestrogens ould be relevant in postnatal life. Oup Environ Med: first published as /oem on 1 April Downloaded from on 19 August 2018 by guest. Proteted by opyright.

5 Eduation Endogenous oestrogen Mammals are adapted to starting life inside their mothers, whose blood streams are rih in estradiol (even before the early pregnany surge). In ontrast to other vertebrates the default sex is female, and masulinisation of the gonads and entral nervous system depends on the presene of androgens. There is some evidene that variation in the onentration of maternal estradiol is assoiated with the risk of testiular aner w28 w29 and of ryptorhidism. w30 Nutritional intake is assoiated with estradiol onentration. It has been suggested that the modern low fibre, high fat western diet may inrease maternal onentrations, leading to problems with the reprodutive health of their male ovspring. w31 Exogenous substanes with oestrogeni evets The relative exposure of the fetus to exogenous and endogenous oestrogens is an important issue, whih depends both on poteny and on onentration in fetal tissues. Diethylstilboestrol (DES) is a syntheti non-steroidal substane that is more potent than estradiol and rosses the plaenta. Other oestrogeni substanes have muh lower poteny. Taking aount of onentrations likely to our, by far the most important are phyto-oestrogens, suh as the isoflavones whih are abundant in soya. They have mixed oestrogeni and anti-oestrogeni evets, as they bind to the reeptor but only ativate it weakly. Industrial hemials with known oestrogeni ativity similarly tend to have mixed agonist and antagonist evets. They are at least five orders of magnitude ( fold) less potent than estradiol and 100-fold less potent than phyto-oestrogens. They also our in far smaller quantities than dietary phyto-oestrogens. 15 It has often been argued that maternal estradiol does not reah the fetus, beause it is bound to plasma proteins. Up to 99% maybeevetively removed in this way, although this still leaves 1% of irulating hormone free. It is also said that estradiol does not ross the plaenta. However, although the estradiol onentration in the embryo during sexual diverentiation is unknown, it seems unlikely that there is lose enough to an absolute barrier to overome the vast diverene in poteny between endogenous oestrogens and exogenous agents, exept possibly in the ase of phyto-oestrogens. DES was presribed for millions of pregnant women in the 1940s, 1950s, and 1960s, in the mistaken belief that it ould prevent threatened misarriage. Pharmaologial doses were given in early pregnany, and in priniple this is therefore a good test of whether the male embryo is sensitive to oestrogens. Of the four end points disussed in this paper, only ryptorhidism was learly inreased. w32 Other genital malformations also ourred, notably epididymal ysts, hypoplasti testes, and urethral stenosis, but not hypospadias w32 (a mis-reading of stenosis as hypospadias has unfortunately been propagated erroneously in the literature). One study of sperm density found a slight redution among those who had been exposed in utero ( /ml versus /ml in the plaebo exposed group). w33 This fall was smaller than that observed in Paris, for example, from /ml in 1973 to /ml in Not all studies have found a redution. w34 The fertility and sexual funtion of men exposed in utero were unaveted when they were studied at almost 40 years of age. w35 The situation with testiular aner is unlear, with both positive and negative studies. A review by the US National Caner Institute onluded that there was no link, w36 but others estimate that the relative risk may be as high as 2.0 w2 (whih is less than the inrease in inidene in many ountries). There may have been some underestimation of assoiation in some studies beause exposure ourred after Box 2: Prinipal exogenous substanes that may affet sex hormone funtion A Oestrogeni and anti-oestrogeni effets (1) High poteny DES (diethylstilboestrol) Ethinyl estradiol (omponent of ontraeptive pill) (2) Medium poteny Phyto-oestrogens isoflavones (for example, genistein, daidzein) oumestans (for example, oumestrol) lignans (3) Low poteny Bisphenol A Otylphenol and nonylphenol Pestiides, inluding hlordeone, DDT, dieldrin, endosulfan, p,p -methoxyhlor, toxaphene B Anti-androgeni effets p,p -DDE Certain phthalates (for example, DBP, DEHP) Pestiides, inluding linuran, proymidone, metabolites of vinlozolin Hydroxyflutamide C Others Dioxins, furans, and dioxin-like PCBs (polyhlorinated biphenyls) This lassifiation is an over simplifiation: it onflates reeptor mediated evets with those aused by other mehanisms for example, interferene with hormone synthesis. Moreover, several of the oestrogens show onsiderable aynity for the androgen reeptor For the reasons given in the text, ompounds in the A3 ategory annot plausibly be onsidered responsible for the types of impairment of the male reprodutive system onsidered in this paper the ritial developmental period in about half of the subjets, together with problems in asertaining exposure, w37 but on the other hand the positive ase ontrol studies may have been influened by reall bias. The learest epidemiologial evidene on the evet of phyto-oestrogens omes from populations with a high soya intake, notably Chinese and Japanese men. The inidene of testiular aner is lower even than in Finland, 6 but semen quality appears to be omparable with that in western Europe. w38 There is no epidemiologial evidene on industrial ompounds with oestrogeni ativity (other than DES), suh as bisphenol A, otylphenol, and nonylphenol. Clear ut and repeatable toxiologial results are also laking for the end points onsidered here. As these ompounds are pharmaologially similar to phyto-oestrogens but far less potent, it is diyult to believe that there is a harmful evet. Other endorine mediated effets Anti-oestrogeni evets As mentioned above, many exogenous oestrogens also have anti-oestrogeni evets. One way in whih the original hypothesis has beome broadened is the idea that oestrogen antagonism an generate the evets that were originally attributed to oestrogeni stimulation. Empirially, the evidene on phyto-oestrogens is equally relevant to this hypothesis. Anti-androgeni evets An alternative version of the endorine disruption hypothesis is androgen antagonism. Anti-androgens do not fae the same problem of ompetition with endogenous estradiol, and the idea is plausible beause the development of maleness in mammals depends on androgens. *285 Oup Environ Med: first published as /oem on 1 April Downloaded from on 19 August 2018 by guest. Proteted by opyright.

6 Eduation 286 * Most of the available information on anti-androgens omes from toxiology. Bloking of androgen ation has been found by p,p'-dde, a stable breakdown produt of DDT (dihlorodiphenyltrihloroethane), 16 and other substanes for example, metabolites of the pestiide vinlozolin. 17 Several phthalates are known to inhibit testosterone synthesis. w39 There is some experimental evidene in rats and/ or rabbits that suh substanes an produe hypospadias and ryptorhidism, early malignant hange in ryptorhid testes, and redued sperm ounts. w39 Expeted spetrum of effets A diverent approah is to ask, what epidemiologial hanges would be expeted a priori from exposure to endorine disrupters of various types: whih end points would be aveted, and in whih sex? In general, females are more sensitive than males to oestrogeni evets. Female ovspring were far more aveted than male by maternal DES exposure, 18 and the evets of phyto-oestrogens are prinipally evident in females. Many other aners are hormone sensitive, and ould potentially be influened by xenobioti endorine disrupters. One would expet a rise in oestrogeni or a fall in androgeni ativity to go with an inrease in male breast aner and a derease in prostate aner. Among women, endometrial and breast aner would be expeted to rise. Only the latter has ourred, and its inidene is muh lower in populations who onsume high quantities of soya. 6 Toxiologists routinely use end points involving growth and development as markers of endorine ativity. Any endorine mediated disturbane would be expeted to avet age at puberty. No suh hange has ourred among boys, either in western Europe or the USA, w40 w41 in ontrast with suggestive evidene of widespread preoious puberty among girls in the USA, espeially Afrian Amerian girls. w42 Disturbane of other seondary sexual features might also be expeted for example, inappropriate body hair or gynaeomastia and disturbanes of growth inluding timing of epiphysis losure. These have not been reported. Interpretation Variation in endogenous estradiol is a possible andidate to explain the observed temporal and spatial variations. However, it is unlear whether maternal estradiol onentrations are suyiently sensitive to nutritional fators to explain the observed trends and/or spatial variations, as there has been insuyient researh in this area. Substanes that blok androgen or oestrogen reeptors in the hypothalamo pituitary axis ould inrease maternal gonadotrophin and therefore endogenous oestrogen seretion. Evidene that the onentration of maternal estradiol is assoiated with biologial hanges undermines the argument that the embryo is not exposed to endogenous oestrogen. For exogenous agents, the DES episode provides evidene that in utero oestrogeni exposure is unlikely to influene strongly the end points disussed in this paper other than ryptorhidism. This ontrasts with its muh more devastating impat on female ovspring, whih inluded vaginal arinoma in young women. w43 The low inidene of testiular aner in Japanese and Chinese men undermines the idea that exposure to phyto-oestrogens (at any stage of life) or to any mehanistially similar ompounds inrease the risk of the disease. Overall evaluation of the evets of phyto-oestrogens needs to take aount of their benefiial role in relation to numerous diseases, inluding oronary heart disease and various aners. 19 Box 3: Endorine explanations key points A prevalent hypothesis is that the male reprodutive system is suseptible to damage by substanes with endorine ativity, suh as oestrogens The mother s endogenous oestrogen may possibly affet the male fetus, but there is insuffiient evidene to support this as an explanation of the epidemiologial findings The human evidene on relatively strong exogenous oestrogens does not support the hypothesis: DES (see text) has been onlusively linked only to ryptorhidism, and some populations whose diets ontain abundant oestrogens of plant origin have low testiular aner rates Environmental pollutants with weaker oestrogeni ativity annot plausibly be responsible for the observed trends or spatial variation Anti-androgens are a more likely ause, but the earliest known example is DDE (from DDT use) whih was introdued after the trend in testiular aner began, and widespread exposure in the developing world has not led to an epidemi of this disease The observed spetrum of effets differs from that whih would be expeted a priori for an endorine mehanism The timing of the testiular aner trend is onsistent with a dietary origin, and the searh for andidates should extend beyond hormonal agents to inlude those apable of ausing geneti damage Environmental pollutants annot plausibly be responsible for adverse evets on the health of the male reprodutive system aused by oestrogeni ativity, on grounds of poteny and mehanisti analogy. In ontrast, it is plausible that anti-androgens ould play a role. However, the substanes urrently known to have anti-androgeni ativity have been introdued sine the seond world war, and therefore annot explain the earlier rise in testiular aner. Exposure to p,p'-dde also annot explain the observed epidemiologial diverenes between Finland and Denmark, as DDE onentrations in human milk have been shown to be similar in all the Nordi ountries; they have delined sharply sine However, even if spatial diverenes and past trends are not expliable in this way, urrent/reent exposure to DDE and to phthalates (whih are now ubiquitous) may possibly have adverse evets. It is reassuring that very high exposure to DDE in developing ountries sine the 1950s, as a result of malaria eradiation programmes, has not resulted in a major epidemi of testiular aner in young men. 6 This suggests that anti-androgeni substanes may not be suyient to ause this disease in non-ryptorhid testes. A broader argument against the importane of endorine disruption for problems with the male reprodutive system is that the observed spetrum of evets divers in important respets from that expeted on biologial grounds. The trends in a variety of endorine sensitive aners do not form a oherent pattern, and assoiated disturbanes in boys pubertal development and seondary sexual harateristis have not been observed. Conlusions The aumulation of evidene on diverent end points suggests a real deterioration in the health of the male reprodutive system, at least in some populations. There has been a widespread inrease in testiular aner, starting 100 years ago in some plaes. There has probably been a deterioration in semen quality in north western Europe in Oup Environ Med: first published as /oem on 1 April Downloaded from on 19 August 2018 by guest. Proteted by opyright.

7 Eduation reent deades. There may well have been an upward trend in ryptorhidism, at least in England, but lear trends in hypospadias have not been established. Large spatial variations our in all of these four end points. There is suggestive evidene for linkage, in that the reprodutive system of male Finns is onsistently healthier than that of Danes and (probably) others in north western Europe. Explanations are muh less ertain. Extrapolation from egg laying reatures to mammals is unhelpful. The auses may be diverent for eah end point, but some shared determinants appear to be likely. Exogenous oestrogens are not strong andidates, but inreased maternal estradiol annot be ruled out. Anti-androgens are more likely, aording to findings in experimental animals, but for testiular aner this hypothesis would predit a major epidemi in large parts of the developing world whih has not ourred. The historial timing of the trend in testiular aner implies that any agent introdued sine the mid-20th entury annot be responsible for the earlier hanges. This inludes all industrial hemials urrently known to be endorine disrupters. A hange in dietary intake ould be responsible, possibly one assoiated with inreasing prosperity. The searh for ausal fators should not be onfined to endorine disrupters. Geneti damage ould predispose to testiular aner and to impaired apaity for spermatogenesis. A genotoxi agent would have to be absorbed by the gut, to reah the embryo during sexual diverentiation, and to loalise within the testis. Mutagens in food inlude heteroyli amines, whih our in meat (espeially if well ooked), and have been impliated in oloretal aner. It is plausible that they meet these riteria; in partiular, their struture suggests that ativated heteroyli amine moleules ould bind to androgen and/or oestrogen reeptors, whih would deliver them to the testiular DNA, where they ould ause divuse genomi damage. I would like to thank Sue Barlow, Jens Peter Bonde, Cate Boyle, and Tina Kold Jensen for omments on earlier drafts. The views expressed are entirely my own. Referenes 1 Sharpe RM, Skakkebaek NE. Are oestrogens involved in falling sperm ounts and disorders of the male reprodutive trat? Lanet 1993;341: This sets out the original hypothesis. Although it is tehnial, and more reent evidene has ast doubt on the partiular mehanism suggested, it repays some attention. 2 Davies JM. Testiular aner in England and Wales: some epidemiologial aspets. Lanet 1981;i: A lear analysis of the beginning of the upward trend in aner of the testis, using mortality data. 3 Adami H-O, Bergstrom R, Mohner M, et al. Testiular aner in nine northern European ountries. Int J Caner 1994;59:33 8. One of the lassi papers of testiular aner epidemiology, it ompares time trends in the Nordi ountries, Balti states, Poland, and Germany. It gives the best introdution to omparative trend analysis for this disease. 4 Bergström R, Adami H-O, Möhner M, et al. Inrease in testiular aner inidene in six European ountries: a birth ohort phenomenon. J Natl Caner Inst 1996;88: This artile reveals interesting features of the time trends in testiular aner inidene, fousing the analysis on birth ohorts, in a omparative framework. 5 Møller Jensen O, Carstensen B, Glattre E, et al. Atlas of aner inidene in the Nordi ountries. Nordi Caner Union, Parkin DM, Whelan SL, Ferlay J, et al, eds. Caner inidene in five ontinents vol. VII. Lyon: IARC Sientifi Publiations No 143, Carlsen E, Giwerman A, Keiding N, et al. Evidene for dereasing quality of semen during past 50 years. BMJ 1992;305: Auger J, Kunstmann JM, Czyglik F, et al. Deline in semen quality among fertile men in Paris during the past 20 years. N Engl J Med 1995;332: A single entre paper on trends in semen quality, whih is well written. It shows a deterioration in the different aspets of semen quality, whih are ompatible with a period or a birth ohort effet. 9 Fish H, Goluboff ET, Olson JH, et al. Semen analyses in 1,283 men from the United States over a 25-year period: no deline in quality. Fertil Steril 1996;65: Three entres are ompared, to see whether a deterioration in semen quality is evident, and to what extent the findings are similar in the different loations. It learly shows the importane of geographial differenes. 10 Vierula M, Niemi M, Keiski A, et al. High and unhanged sperm ounts of Finnish men. Int J Androl 1996;19: The higher level of sperm onentration in Finland, and the absene of a trend over time, is learly demonstrated. 11 Swan SH, Elkin EP, Fenster L. Have sperm densities delined? A reanalysis of global trend data. Environ Health Perspet 1997;105: The same researh question as the original paper by Carlsen and olleagues, but arefully analysed and written. The paper provides a disussion of residual onfounding effets that annot be ompletely resolved. 12 Paulozzi LJ. International trends in rates of hypospadias and ryptorhidism. Environ Health Perspet 1999;107: John Radliffe Hospital Cryptorhidism Study Group. Cryptorhidism: a prospetive study of 7500 onseutive male births, Arh Dis Child 1992;67: Møller H, Skakkebaek NE. Risk of testiular aner in subfertile men: ase-ontrol study. BMJ 1999;318: Safe SH. Environmental and dietary estrogens and human health: is there a problem? Environ Health Perspet 1995;103: This is still the most useful paper on the relative potenies and abundane of different oestrogeni ompounds. 16 Kele WR, Stone CR, Laws SC, et al. Persistent DDT metabolite p,p -DDE is a potent androgen reeptor antagonist. Nature 1995;375: The paper that established anti-androgeni ativity as an important fous, whih shows that a widely dispersed pollutant has this type of ativity. It does not deal with the effets on reprodutive end points, however. 17 Gray LE, Ostby JS, Kele WR. Developmental effets of an environmental antiandrogen: the fungiide vinlozolin alters sex differentiation of the male rat. Toxiol Appl Pharmaol 1994;129: Mittendorf R. Teratogen update: arinogenesis and teratogenesis assoiated with exposure to diethylstilbestrol (DES) in utero. Teratology 1995;51: Bingham S. Dietary phyto-oestrogens and aner. Pure Appl Chem 1998;70: Ekbom A, Wiklund-Glynn A, Adami H-O. DDT and testiular aner. Lanet 1996;347: website extra Additional referenes appear on the Oupational and Environmental Mediine website QUESTIONS (see answers on page 260) (1) Testiular aner: (a) is assoiated with male fator infertility (b) inidene has inreased in some plaes and fallen in others () is espeially ommon in Finland (d) is espeially ommon in Denmark (e) is mainly a disease of old age (2) Sperm onentration: (a) is the only way of assessing semen quality (b) is reliably measured from a single sample () has been shown to deline over time in Paris (d) has been shown to deline over time in New York (e) tends to be high among Finnish men ompared to Danes *287 Oup Environ Med: first published as /oem on 1 April Downloaded from on 19 August 2018 by guest. Proteted by opyright.

8 Eduation 288 * (3) Whih of the following statements on ryptorhidism and hypospadias are true/false? (a) There is good evidene for an inrease in ryptorhidism in England (b) There is good evidene for an inrease in hypospadias in England () Geographial differenes in hypospadias are diffiult to establish beause of problems with asertainment and reporting (d) Hypospadias is more ommon in Denmark than in Finland (e) Cryptorhidism is more ommon in Finland than in Denmark (4) Whih of the following statements are true/false? (a) In utero DES exposure inreases the risk of ryptorhidism (b) In utero DES exposure inreases the risk of hypospadias () Japanese men have a low risk of testiular aner (d) There has been a generalised inrease in all hormone sensitive aners (e) Preoious puberty has beome inreasingly ommon among teenage boys (5) Whih of the following statements on oestrogeni/ anti-androgeni ativity are true/false? (a) Endogenous oestrogens are bound to plasma proteins (b) Oestrogeni ativity has been found in numerous industrial ompounds () Soya ontains substanes that have mixed oestrogeni and anti-oestrogeni effets (d) Anti-androgeni ativity has been found in breakdown produts of ertain pestiides (e) p,p -DDE is the only known anti-androgen that was present in the environment in the early 20th entury Oup Environ Med: first published as /oem on 1 April Downloaded from on 19 August 2018 by guest. Proteted by opyright.