General Discussion

Although I had no influence on the composition of the cases in my series, they did not constitute a representative sample of transsexuals. They consisted of transsexuals under such a strain because of the social or mental sequels of their anomaly that they were forced to seek medical help. What they had in common was a severe form of the anomaly, or a personality unable to cope with its social or mental consequences. Accordingly, one might say that they arc representative of transsexuals who go to a physician for help.

Because the series was selected and varied greatly in age, and came from different parts of the country, it was hard to find a control series suitable for all the variables I examined. For this reason, in the cases where I compared the transsexuals with controls, I chose controls examined the same way as the transsexuals for the variable in question and as representative as possible of the general population. It may be objected that doing this makes it impossible to draw any general conclusions, but I decided that it was best to use large, well-analyzed series from other sources for purposes of comparison for my study, whose aim was to throw light on certain variables to help provide a basis for further research. For the study of social data, I assembled my own control group.

There are no other prevalence figures calculated in the same way with which to compare mine - 1 per 37,000 men and I per 103,000 women. Otherwise Anchersen (1956) reported that he knew of 6 transsexuals in Norway's population of a little over 3 million; Hamburger et al (1953) reported 5 in Denmark's population of about 4 million; Turtle (1963) believed that there was roughly I transsexual in every 3,000 to 17,000 persons in Great Britain; and Pauly (1966) estimated from his study of the literature that I out of every 100,000 men and I out of every 400,000 women was a transsexual. My results indicate that transsexualism is more common than most of these figures indicate, both among men and women.

The male/female ratio in the series was 2.3:1. This agrees well with the ratio Randell (1959) found, but the proportion of men is lower than that given by other authors. The preponderance of men may be partly due to it being easier for women to compensate for the anomaly than men, with the result that men seek medical advice more often than women. It may also be that the ratio is affected by different attitudes to the anomaly in men and women on the part of both physicians at Ad laymen, as well as authorities. Thirdly, it is not inconceivable that there is some biologic reason for the preponderance of males among cases of transvestism/transsexualism and other forms of sexual deviation (Roth & Ball 1964).

My series of transsexuals did not differ much from the general population in external features, except for I case in which there was reason to suspect an endocrine disorder. Some of the patients showed an abnormal distribution of pubic hair, but so do subjects in the general population (Dupertius, Atkinson & Elftman, 1945). One of my male patients had no beard and 4 had only a sparse one, but these men, who had passed through puberty at a normal age (about 15), were mostly still too young at the time they were examined to be able to say for sure that they were abnormal in this respect.

There was I case of gynecomastia, but gynecomastia is not unusual at the age this patient was. On the other hand, the patient's urine contained an abnormally high content of pregnanediol and his transsexualism grew worse while the gynecomastia was developing.

Neither the men nor the women differed to any noteworthy extent from the normal in any of the body dimensions measured. True, the women often showed a larger than average biacromial breadth in proportion to the length of the radius and tibia, which to some degree bears out the observation that many female transsexuals have the wide shoulders of the young man. But there were only 13 women in the series, and their ages were not representative of the general population.

The mental characteristics of the patients fell into three groups: ones intimately linked with the transsexualism, ones resulting from the transsexualism, and ones independent of the transsexualism, and of the kind found in any group of persons with mental troubles.

One must be careful about drawing specific conclusions about the mental characteristics before puberty. The feeling of belonging to the opposite sex apparently sets in early in life, but anomalies of behavior seem to be more prominent to begin with. It is perhaps best to say that before puberty transsexualism is characterized by an extreme form of cross-gender behavior, but not by any particularly specific features. After puberty, more specific features can be made out. These, what I consider the cardinal features of transsexualism, are: (1) a conviction that one belongs to the other sex than indicated by one's body, (2) abhorrence of one's body, especially one's genitals, (3) demands for operation to change one's anatomic sex and/or legal measures to have one's sexual status changed, and (4) a desire to be accepted by everyone as a member of the opposite sex.

The feeling or conviction of belonging to the other sex, existing both before and after puberty, is to my mind the central problem for these people, around which all their other troubles cluster. In my opinion, this is the only real primary disorder, but the three other features just listed are so intimately connected with it that I have called them cardinal. This feeling of belonging to the opposite sex corresponds to what Stoller (1964 a, b, c) called inverted core gender identity, the basis of the whole transsexual complex. It was the first subjective intimation of the anomaly in every one of my cases. To my mind, an inverted core gender identity is what mainly differentiates transsexualism from other disorders associated with cross-dressing.

In no case in the present series was the cross-dressing associated with sexual excitement or satisfaction. The women cross-dressed more than the men, the sexes differing significantly in this respect, both before and after puberty. External circumstances seemed to make more difference to the amount of cross-dressing in the case of the men than in the women. The men seemed to be more emotionally concerned about their situation than the women; they were more envious of women than the female patients were of men, and they were more persistent in their efforts to undergo a change in sex.

All this was reflected in the difference between the men and women in frequency of suicidal thoughts and a feeling of being persecuted, which were both more common in the men than in the women, though the differences were not statistically significant. Twenty per cent of the men had tried to commit suicide against 8 per cent of the women.

A group with psychiatric problems must yield a higher percentage for attempted suicide than a Population chosen at random. Nevertheless, the high rate in the present series - 16 per cent against 0. 1 per cent in Parkin & Stengel's general population -shows, even though the latter was not calculated in exactly the same way, that more attention than hitherto should be given to the risk of transsexuals committing suicide.

That the transsexuals required more -social assistance than controls is probably due, partly at least, to the great strain under which they lived on account of their anomaly, no doubt largely because it was hard for them to get work that suited them.

The great amount of sick-listing in their case can be traced to a small minority-the ones who took long sickness absences or several short ones, for surgery for example, and for which they were generally classi

fied under mental disorder. Thus the large number of absences for "mental disorder" does not necessarily mean that the patients often got attacks of mental disorder in the ordinary sense of the term.

Most of the patients, both men and women, had sexual relations with their own sex. But only a few had had purely heterosexual or homosexual experience, the proportion between the two varying from case to case. As Kinsey et al. (1948) and others have noted, whether or not one has sexual relations with members of one's own or the other sex sometimes depends on circumstances, and even heterosexually oriented persons may become. homosexual for a while if they get into a situation where it is impossible to get a partner from the opposite sex. Thus it is impossible to draw any conclusions about heterosexuality and homosexuality from the point of view of the kind of sexual experience the patients had had. An interesting observation in this series is that 12 patients, 8 men and 4 women, had not had any form of genital activity with either sex. Most of them said that they had never felt a desire for it. Some said that they wanted to wait until they could get their bodies changed, but closer analysis revealed that they probably had a basically weak sexual drive. Thus 28 per cent had not had any form of sexual activity, and equally many denied having masturbated. Twenty-three per cent had a low sexual drive, as defined here, women more often than men. This is in keeping with the observation that only about 16 per cent wanted a change in sex for sexual reasons. As far as I could see, sex, as the word is usually used, did not play a prominent part in any case. Whether this was because of a low sexual drive, or because the drive was sublimated in the psychoanalytical sense, is impossible to determine from this investigation. Three observations, howeverabsence of sexual motives for the cross-dressing, in many cases below average sexual urge, and the fact that conversion operations were seldom wanted for purely sexual reasons -indicate that the sexual component is weaker than it seems to be in other kinds of sexual "aberration", such as fetishism, and simple transvestism.

As in the cases reported by other authors, none of the transsexuals in my series regarded having sexual relations with members of their own sex as homosexual. In fact, they were shocked at the idea that others thought they were homosexual. To them sexual relations with a member of their own anatomic sex were heterosexual, not homosexual. In keeping with this, the male transsexual sought the company of virile heterosexual men, and refused to have anything to do with homosexual men. Judging by the interviews I had with the male sexual partners of the men, the partners regarded them as women. Thus the terms heterosexual and homosexual acquire another meaning in connection with transsexualism than they have otherwise.

To sum up, at the time I saw the patients, they were all sexually aroused mentally by persons of their own sex. This was the most consistent sex variable found, and only to be expected, seeing that transsexuals feel that they belong to the opposite sex. The findings regarding sexual activity were not consistent, varying from case to case, and from one time to another in the same case, apparently depending largely on the circumstances in which the patients lived. About every fourth to fifth patient had only a weak sexual urge or little interest in sex, the women more often than the men. The observations indicate how important it is to distinguish between the partner used for sexual activity, and the sex exciting the subject mentally.

The personality analysis of the patients revealed an unusually high frequency of psychoinfantilism especially among the men. Skoog (1959) found this trait in only about 16 percent of patients admitted to the psychiatric department of a general hospital, against about 48 per cent in the present series. The men in my series were also probably significantly more often asthenic and hysteroid than the women. The sex ratio for asthenic and hysteroid personalities in Lindegard's (1966) normal series were the reverse to those in mine. But the men and women were more often schizothymic than in Skoog's (1959) and Jansson's (1964) series. The greater intensity of the transsexualism among the men was probably due to the greater frequency of asthenic, hysteroid and psycho-infantile traits among them.

The men seemed to suffer more from their situation than did the women. They were more insistent in their demands for a change in sex, they were more often depressed, they more often tried to commit suicide, they tired more easily, they gave way more easily under strain, and they were more maladjusted on the whole. The women more often planned their future a long way ahead, they seemed to suffer less from their situation, they were more determined to solve their problems as best they could, and they succeeded better in coming to terms with life.

The results of the personality analysis make it easier to understand the attitude of the patients when they come for consultation. If one keeps in mind how often transsexuals are psychoinfantile, it is easier to understand why they strike us as more intense, more insistent in their demands, and more helpless than patients with other disorders. This is particularly true of the men.

The frequency of both schizothymia and psychoinfantilism, also indicate that the way transsexuals behave and talk when they come to a physician do not necessarily reflect their original personality. Some of the traits they show then may partly be the result of the conflicts they have had both with their fellow-men and the authorities because of their anomaly. One might say that they suffer from "social adjustment insufficiency" in the sense of Lindegard (1962).

The transsexuals gave the opposite results to representative normal populations on the Marke-Gottfries verbal measure of masculinity-femininity. It may be objected that one cannot compare their scores on this test with the 15 year old population. But other investigations (Brown & Tolor, 1957; Lynn, 1959) have shown that one rarely changes the sex with which one identifies oneself after the age of 15, and this was borne out by Marke and Gottfries' (1966) study of 20 to 25-year-old Students. Where the 15-year-olds differed most from the older controls was in interests, which were more masculine in the 15 year olds. In emotions the 15-year-olds w.-re more feminine. Otherwise they gave much the same results as the older subjects. The older male students probably had more feminine interests than men of similar age in other walks of life.

The subjects might be able to reckon out to a certain degree how to answer the questions in this questionnaire so as to appear masculine or feminine, and the same is true of the other verbal test used-the MF scale from the MMPI. But the aim of both these tests is to have the subjects describe themselves. They measure the answers (behavior) which the stimuli (questions) provoke. Thus, in this respect the patients behaved quite differently from normal subjects.

As to the nonverbal tests used, it has been suggested that they measure unconscious rather than conscious identity, and Lynn (1959) concluded that the Draw a Person Test, for example, was a measure of sexual role identification. The male and female transsexuals did not differ significantly in their results on the Frank Drawing Completion Test, though they both gave results contrary to those of normal members of their sex. Nevertheless, the results from the transsexuals give the impression that, if the series had been larger, a significant difference would have emerged between the sexes.

Many factors may affect the results on the Draw a Person Test-for instance, the sex of the examiner, the way the instructions are given, what the subjects are thinking of when they are drawing, and what associations take place-and there is also the problem of choice of object versus identification. But the results from earlier research indicate that, normally, both men and women tend to draw a person of their own sex first, especially men. The transsexuals showed a decided inclination to draw a person of the opposite sex first. Otherwise the same was true of the transsexual series as of normal series, that more women than men drew the opposite sex first.

In view of the technical difficulties connected with hormonal analysis and the uncertainty as to what is normal or abnormal at various ages (e.g., Diczfalusy, 1960), one must not attach too much significance to values obtained, especially when as in the present series they were based on single tests, not serial testing. The only cases worth discussing are: case 3, because of the suspiciously high value for total gonadotropins together with suspiciously high estrogen fractions; case 6, because of the high level of pregnanediol excreted; and case 30, because of the high level of estrogens excreted. All three patients said that they had not taken any form of hormones or other drugs. It was not possible to make any follow-up analyses in these cases. But, all in all, the hormone analyses in this series did not point distinctly to transsexualism being associated with abnormalities in the forms of steroid metabolism studied.

According to Frey and his co-workers, a low alpha frequency in the EEG is common among subjects with mild mental disorders of external origin (Frey, 1946; Frey & Steinwall, 1953; Forssman & Frey, 1953; Frey & Sjbgren, 1959; Frey, 1961). But the present series showed no tendency to a low alpha frequency. Nor were the abnormal EEG findings consistent from case to case.

Judging from figures for normal series, there was an unusually high incidence of rhythmic low-frequency activity in the posterior leads in the present series (Aird & Gastaut, 1959; Peters6n & S6rbye, 1962; Selld6n, 1964); this abnormality was noted in 5 cases (4, 7, 15, 19, 29). But figures from series with different disorders with which this can be compared are lacking.

Twelve of the 42 transsexuals in the present series examined electroencephalographically had abnormal EEG's, and 14 out of the 42 similarly examined cases in my literature series. Together, this gives 26 abnormal records out of 84, or almost one-third. Thus, both separately and combined, the two series had an unusually high number of abnormal EEG's, judging by figures from normal series (Selld&n, 1964). In view of this, and the over-representation of epilepsy in my literature series of 207 transvestites/transsexuals, further research on this question is now in progress at our institute.

No evidence of an intersex state was forthcoming from the study of the sex chromatin pattern and karyotype. This agrees with the observations in other large series (Burchard, 1963; Benjamin, 1964c; Ball, 1966). Combining my cases with all those I could find in the literature gives 199 male and 42 female cases in which the sex chromatin or karyotype was examined; 10 of the 199 men had positive sex chromatin, and all 42 women were normal in this respect. Except for 2 cases (Money & Pollitt, 1964), all 10 men were from single case reports. Obviously cases in which the sex chromatin did not agree with the phenotype would be more apt to be reported than ones in which it did. Hambert's (1966) series of 75 men with positive sex chromatin contained no instance of transvestism or transsexualism. Nor did Lindsten (1963) find any case in a series of Turner subjects (49 cases between 16 and 30 years of age) or Hampson (1965) find any in 13 other Turner subjects. But before one can say what these negative findings signify, one must know how often transsexualism occurs in cases of chromosomal deviation, and this remains to be discovered.

Judging by the information received, none of the mothers of the 43 transsexuals had shown any form of abnormality during pregnancy. In 2 of the 28 cases of hospital delivery, the amniotic fluid was discolored. Two of the subjects born at home were of twin birth (bi-ovular). Otherwise the deliveries were apparently normal. In other words, 39 of the deliveries, or 91 per cent, were uncomplicated in the usual sense of the term; 26 of the 28 hospital deliveries that could be checked in hospital records, or 93 per cent, were uncomplicated. The corresponding figure for Roth & Ball's series (1964) was 94 per cent.

The 43 subjects apparently showed nothing out of the way during the first two weeks of life. The abnormalities noted later, during the first three years of life, may or may not have involved a cerebral lesion-it is hard to tell. In case 1, in which there was reason to suspect an endocrine disorder, the findings may all point to early brain damage. The same is true of the severe scarlet fever and slowness in learning in case 29. But the familial occurrence of mental retardation and epilepsy in this case could also point to a hereditary disposition. After the age of 3, only 3 of the histories contained events of particular interest to this study: the head injury followed by epilepsy in case 25; the onset of the transsexualism at adult age a few years after a severe head injury in case 15, and the juvenile diabetes and frequent attacks of coma in case 16. The 5, at least six cases, all male, had a history of circumstances which might have affected cerebral functioning (cases 1, 13, 15, 16, 17, 25). Of these, case 15 is particularly interesting from the point of view of etiology. All in all, however, the possibility of coincidence in these 6 cases cannot be excluded. The observations seem to indicate, nevertheless, that one should pay more attention than has been done hitherto to the subject's physical state in early childhood.

Forty-three per cent of the men and 23 per cent of the women had shown at least one neurotic trait during childhood - against about 77 per cent of Roth & Ball's male transsexuals. Six men (cases 3, 11, 15, 18, 19, 23), or 20 per cent, had shown two or more of these traits and 2 of the women (cases 33, 43), or about 15 per cent. The figures do not appear to differ much from those in children in general. For example, Jonsson & KäIvesten's (1964) figure for nail-biting in 222 Stockholm boys was 21 per cent, and for sleep disorders 15 per cent as against 20 and 13 per cent respectively in the present male series. Most authors are cautious about drawing any conclusions about the significance of neurotic traits in childhood. These symptoms are often only superficial and disappear after a time. Both cases of persistent bed-wetting were girls, while this was not mentioned in any of the male cases; it is generally the other way about. Thus, according to Hallgren (1957), 13 per cent of boys wet their beds at night against 9 per cent of girls; from his figures one would expect about 4 bed-wetters among the boys in my series and only I among the girls, instead of none and 2, respectively.

Jonsson & Kälvesten (1964) found that 25 per cent of a group of normal city boys needed some kind of help or guidance from child psychiatrists. Altogether 29 per cent of the patients in the present series had been taken to a psychiatrist either by their parents or on the initiative of the child welfare boards. This is a surprisingly low percentage considering that the feeling of belonging to the opposite sex which most of the transsexuals had already as a small child must cause a great deal of mental conflict, especially during adolescence. Thus the present study does not point to any particular amount of nonspecific behavior disorder or nervous disease in the early life of transsexuals.

It appeared from the analysis of the family background that the transsexuals more often grew up in an insecure and unstable environment than did the control series with which they were compared. Thus many more complaints had been made to the authorities about the way they were being brought up, and many more of their fathers had been reported to the officials for repeated drunken misconduct. Fifty-four per cent of the transsexuals, significantly more than in the control group, had suffered from some form of parental deprivation in their youth, if by parental deprivation is meant death or divorce of the parents before the subject was 15, placement away from home and illegitimacy. The corresponding figure for the 207 cases I analyzed from the literature was 37 per cent. These figures agree well considering that the figures for illegitimacy and placement away from home were probably higher in the literature cases than shown by their case histories -11 per cent of the transsexuals had a history of illegitimacy, against 4 per cent of

the other cases. The corresponding figures for placement away from home were 27 and 7 respectively. Others report higher figures for parental deprivation (e.g., Roth & Ball, 1964) but this is probably because they analyzed their cases in different ways and used different operational definitions.

It may be assumed that losing a parent at an early age affects a child's sexual identification and other factors in psychosexual development. The only reliable figures one can get for the onset of parental deprivation are for parental death, divorce and illegitimate birth. Half the subjects in the present 16 cases of parental deprivation were deprived of their parents in these ways before they were 3, and the other half after they were 3. The corresponding figures for the controls were 43 and 57 per cent. Consequently it is impossible to draw any general conclusions on the significance of the onset of parental deprivation from these cases.

Lack of opportunity for proper sexual identification, because of the same-sexed parent being away from home most of the time the child was growing up has been suggested as causing transsexualism (e.g., Lukianowicz, 1959a). But there was little evidence to support this theory in the histories of the present cases.

Psychoanalytically inclined researchers assume that cross-dressing is a way of mitigating castration anxiety (and thus exclude the possibility of a female transsexual) but most of my male patients wanted to have their gonads and penis removed. Moreover, after they undergo sexchanging surgery, many patients become more stable and have fewer attacks of anxiety and depression. Thus it is hard to understand how castration anxiety could lead to the development of transsexualism.

The significance of parental rejection, which some authors believe is the cause, is hard to judge from my series. In the 19 cases in which it was possible to interview the relatives, they said nothing indicating that the parents had wanted a child of the opposite sex. Three patients said that their parents had said so now and then, but it was impossible to interview these parents. In no case had the parents deliberately dressed their children in the clothes of the opposite sex. Thus my series did not support the theory of "psychologic conditioning" of this-nature.

As many as 46 per cent of the transsexuals had been brought up in an apparently normal atmosphere. This would indicate that some cases of transsexualism at least must stem from other factors than those connected with the family background.

Several authors have suggested that transsexualism is of organic origin, that it is due to genetic, hormonal or cerebrolesional mechanisms. Data pointing to an organic factor in my series were: (1) The large number

of abnormal EEG's. Epilepsy was over-represented in the cases I collected from the literature, and I of the 43 transsexuals in my own series was epileptic and another got a grand mal attack during photostimulation. I have already reported (Walinder, 1965) that cerebrolesional factors have been noted in cases of different kinds of sexual aberration. (2) In I of the present cases the transsexualism started some years after a severe head injury, no signs of d6iation being observed before; in this case the transsexualism disappeared on anticonvulsant medication (given because of an abnormal EEG) and reappeared when the medication was stopped. (3) The familial occurrence in 4 cases of mental retardation, cerebrolesional signs and abnormal EEG's, pointing to the possibility of a hereditary disorder in cerebral functioning. (4) Definite evidence of an early cerebral lesion in I case (case 1) and the possibility of such in case 16. Adding together these cases gives 15, or about 35 per cent, with evidence of an organic disorder (cases 1, 4, 7, 12, 13, 15, 16, 19, 23, 25, 29, 31, 33, 38, 42).

It is unlikely that the same mechanism lies back of every case of transsexualism. On the other hand, disorders in cerebral functioning may cause a wide variety of mental disorders, the kind probably depending on the site of the injury, and the age at which it occurs. In view of NS, and the usually early onset of transsexualism, the injury must occur early in life if transsexualism is of organic origin. One can influence the sexual behavior of animals by giving hormones prenatally (e.g., Young, 1961, 1963; Young, et al., 1965). My study of prenatal and perinatal factors, however, did not reveal any circumstances of note.

Of particular interest when discussing the possibility of an organic factor are the cases in which treatment of a hormonal disorder (Routier et al., 1964) or treatment of cerebrolesional disorders (e.g., Hunter et al., 1963) eliminated or lessened the intensity of the transsexualism" transvestism. My case 15 is another example. In all these 3 cases the symptoms were reversible and, in my case at least, they began later than in most cases. In the majority of cases, however, the transsexualism begins early in life and does not respond -to treatment. The consistency from case to case is compatible with some form of organic disposition.

My investigation has shown that it is hardly possible to attribute transsexualism to only psychologic or only organic causes. Circumstances pointing to organic origin were present in some cases, and circumstances pointing to environmental origin were present in others. It is reasonable to assume that the two kinds of factors interact, that environmental factors in the wide sense shape and determine how the transsexualism develops, and that some unfavorable external factors precipitate the transsexualism, or turn what was only a disposition to transsexualism into a permanent, fixed form of the anomaly. It is also possible that psychologic factors affect the fixity of the transsexualism, and help to make it irreversible after puberty.

Whether people are psychosexually neutral at birth, and their psycho-sexuality is determined through imprinting during childhood, is a question which cannot be answered by a retrospective study like the present one. But even if this is true, it does not gainsay the possibility of interaction between nature and nurture. Sometimes the environmental factors are most prominent, as when the child is brought up to feel as though it belonged to the opposite sex (e.g., Money et al., 1957; Hampson & Hampson, 1961; Hampson, 1965); sometimes constitutional factors are most prominent (e.g., Stoller, 1965; Barton & Ware, 1966). To repeat, however, none of my patients had been brought up as though they belonged to the opposite sex.

Both homosexuals and transvestites may show signs of cross-gender behavior in early years (e.g., Sturup, 1956; Green, 1966), but they identify themselves with their own anatomic sex. One of the most interesting tasks for further research in this field is to follow the development of children characterized by cross-gender behavior. Transsexualism seems to make greater inroads on personality than transvestism, for instance. If one can find factors positively correlated with the feeling of belonging to the opposite sex, it should be easier to determine the causes of transsexualism.

Transsexualism is a separate disease entity, at least in the great majority of cases, and it is manifested early in life. In my opinion, it is best described as a syndrome with three essential characteristics: a conviction of belonging to the opposite sex, disgust at the signs of one's anatomic sex, and the desire for a change in sex. These factors are specific of transsexualism, and because of them it is not difficult to delimit,transsexualism from other forms of sexual aberration.