- Email: [email protected]
Psychiatric reactions to oral contraceptives
CURRENT DEVELOPMENTS An evaluation Psychiatric reactions to oral contraceptives FRANCIS
J.
KANE,
JR.,
M.D.
Chapel Hill, North Carolina
1 N I TIA L studies1-3of the oral contraceptive agents seemed to reveal few serious side effects of these hormonal combinations, while providing excellent conception control. More experience has raised the question of whether these early estimates reflected fully the problems which could and/or should have been anticipated with acute and chronic use. It seemslikely that the enormous social issuesof international scope to which these drugs are intimately bound have contributed to the difficulties of dispassionate and full assessmentof responseto these drugs. There would seem to be no reason to deny these drugs their contribution to “The Diseasesof Medical Progress,” an apt term quoted by Lasagna4 for those illnessesnewly generated by our potent therapeutic agents. He cites several studies5*6 indicating that 5 to 15 per cent of first admissions to some hospitals result from such aberrations of our therapeutic efforts. For the oral contraceptive agents, a place now seemsassured. Clinical and experimental reports indicate that thromboembolic disorder,r prolonged amenorrhea,s hypertensive reactions,g*lo neurological disorders,ll disturbance in carbohydrate metabolism,12and emotional disturbances13may accompany their use. In view of the large literature on mood and behavioral alterations in relation to gonadal hormone variation in women, such
From the Department of Psychiatry, University of North Carolina.
reactions should really come as no surprise. Psychiatric reactions,141 l5 especially depression, in the early postpartum period are commonplace in the practice of obstetriciangynecologists. In a recent series of 90 pregnant women studied at this hospital,la 28 per cent reported subjective symptoms of anxiety and depression, confirmed by psychological test materials. Somewhat more surprising, however, was the fact that this figure was lessthan half that of the changes reported in the last trimester of pregnancy. Rates of psychosisamong pregnant women are below those rates expected in nonpregnant women.l? In the early postpartum period, milder forms of illness become less frequent and major psychoses,which many psychiatrist?** I9 feel are unusual in their clinical phenomenology, dramatically increase in incidence. The fact that 80 per cent of these psychotic episodes occur within the first 30 days after delivery suggestsa relationship to the dramatically altered endocrine state occurring at the end of pregnancy.17 The menstrual cycle is also associatedwith alterations in mood and behavior, especially immediately prior to and during the menses. A number of studieszOl21 have documented the markedly increased incidence of psychiatric illness during this period, including suicides,suicide attempts, homicides, and admissions for acute psychosis. Some have attributed this to the decline in hormones taking place during this time. The relief of premenstrual symptomatology by the oral 1053
1054
Kane
contraceptive agents is striking in many studieP and may be related to the shortening of the period when there are low levels of hormones. Another period of major psychosocial endocrine change for a woman is the menopause, which also is associated with a marked increase in incidence of major and minor psychiatric illness. The incidence of such illness, especially so-called involutional psychotic reactions, is more than 3 to 1 in favor of women.23 The use of exogenous corticosteroid hormones has also been associated with notable mood and behavioral change in a certain number of people receiving these drugs.24* 25 A variety of studies have reported mild to moderate euphoria and increased well-being as well as tiredness, depression, and increased irritability. Frank manic and depressive psychotic episodes with higher dosages of steroids are not unusual. Previous studies Glick22 surveyed the available literature on the use of oral contraceptive agents. Nineteen studies, mostly by obstetrician-gynecologists, revealed sporadic reporting of emotional distress, with a 5 per cent incidence of depression being the highest reported. There were more favorable comments reported, though these comments were likewise largely anecdotal. .4 similar scattering of reports from the same clinical material showed increased “libido,” increased satisfaction with coitus, increased control of dysmennorrhea, and decreased premenstrual tension. Changes of an adverse kind were much less frequent, but they were reported for all except premenstrual tension symptoms. Wearingz6 reported a 16 per cent incidence of depression in 62 patients, and when it occurred, the depression became more prominent the longer the patient remained on drug. MooP reported a questionnaire survey of women asked to rate a variety of symptoms separately for menstrual, premenstrual, and inter-menstrual phases of their most recent cycle and for their worst cycle. He reported that most reported slight decrease in men-
Am.
December 1, 1968 J. Obst. & Gynec.
strual symptoms, while IO per cent had a significant increase. Sequential hormone users reported more symptoms of water retention and negative feeling than the combination drug group. Four reports surveyed by Glickz2 described drug use in psychiatric populations, where it was hoped their use would help patients who had exacerbation or worsening of psychiatric symptoms associated with the menses. These reported were generally favorable, though they lacked controls in most instances. Wez8 have reported previously upon one patient with marked premenstrual exacerbation of psychiatric symptoms who improved on Enovid and showed relapse on placebo in a double blind study. Bakker and DightmarP report a study of 100 women and conclude that while fluid retention was a significant problem in 30 per cent of the subjects studied, depression and changes in libido could not be related to medication as such. Murawski and associateP report a longitudinal study of 72 women over a 15 month period, using a variety of instruments for subjective and objective feelings and behavior. Marked behavioral change was not reported, and reactions seemed related to a number of factors (disappointment with magical expectations of drug, persistent frigidity in spite of relief of fear of pregnancy, etc.). Sedation was reported by some patients, with dramatic behavioral alterations in at least one patient. The investigators attributed this to disturbance of customary modes of adapting, based on compulsive working. However, similar kinds of indifference to usual concerns are common with the use of a variety of stimulant and depressant drugs. The formal report of the study did not include the reported exacerbation of psychosis reported elsewhere by SturgisIB1 a somewhat puzzling omission. Bakke32 reports a double blind cross-over study of an estrogen-progestin mixture, estrogen, and placebo. Twenty of 27 women preferred either estrogen or the oral contraceptive agent, while only 3 preferred placebo (reasons given, related in the patients’ words: “to feeling marvelous, with much more zest”
Volume Number
102 7
Psychiatric
and “having increased sexual drive”), The reasons for preference of the placebo related to relief of undesired hormone effects of the In other estrogen-progestin combination. words, the patient felt worse on the drug, with tender breasts and poor sleep, all of which improved when they stopped. The evaluation of the placebo period was very difficult because of the withdrawal effects which were so common. Complaints of moodiness, being cross and tired, alterations in sexual drive, weight gain, edema, and insomnia were commonest in the group using the estrogen-progestin group. Twelve women reported increased sexual drive. Six women who enjoyed the increase in drive continued on the drug, while 6 rejected the drugs because of this same change. Scott and Bra@ reported their experience in the use of massive doses of norethynodrel in the treatment of endometriosis on 20 patients. Daily dose varied from 40 to 100 mg. of norethynodrel. They commented that “mood and behavior changes were noted in all patients, and 3 developed moderate depressions that responded to antidepressant drugs. One patient also developed a tremendous increase in libido.” Psychotic
reactions
Our own interest in these drugs was stimulated by the observation of a psychotic episode upon withdrawal of 30 mg. daily of Enovid used to control endometriosis.84 The patient experienced relief of her psychotic behavior with drug replacement and intensification of her psychosis when placebo was substituted. We have seen 4 additional women since that time who have suffered psychotic episodes coincident with the use of oral contraceptive agents. In 2 instances35 these were associated with the use of the combined agents and in 2 with the use of sequential hormone combinations. One patient noted change with both, although she became hospitalized while using the sequential agent. Three of these women gave a history of postpartum disturbance in the past and one was receiving psychotherapy for a character disorder. One patient tolerated
reactions
to oral
contraceptives
1055
sequential agents very well for a year but suffered psychotic episodes with only one month of combination drug use on separate occasions before and after the use of the sequential agents. This would seem to raise a question as to whether there is not some increased risk in using these drugs in persons with known psychiatric illness. There have been two additional reports of psychotic reactions in the world literature. Idestriim36 reported another case of psychosis upon withdrawal of drug; in this case, norethindrone. Sturgis31 reports recurrence of psychosis in a patient participating in the prospective study of drug effects by Murawski and associate9O described earlier. This patient was able to tolerate continued use of drugs. It is not clearly stated whether she received psychopharmacologic agents during her illness. Further
clinical
studies
Our first study13 involved the use of c1inica.i interviews and personality testing on 50 women referred by physicians in the community and the medical school complex. Eleven of these patients were psychiatric inpatients at the time of interview and the patients had used only combination drugs. The clinical findings from the study are summarized in Table I. Twenty-eight patients reported undesirable effects while 11 felt better on drugs and 11 reported no change. Depression and decreased sexual interest were noted most frequently. Unusual stress was not reported but rather increased susceptibility to usual stress. Moral conflicts seemed negligible as a factor, since most previously had used mechanical contraceptive devices. Only the nymphomaniac patient reported an evidenced guilt over her increased sexual drive, and she was the only one who showed marked changes on psychological testing. This study had obvious defects, especially in the nature of the selectivity of the sample. A second study was therefore conducted, sampling widely from the mothers in our Obstetric-Gynecologic Clinic, Pediatrics Department, and a busy pediatric practice. Two sophomore medical
1056
December 1, 1968 Am. J. Obst. & Gynec.
Kane
Table I. Mood and behavioral change with the use of oral contraceptive agents Emotional
1 Patients
change
Depression, tearfulness, ( 1 hospitalized for depression)
19 type)
139
=
90 (64
per
cent)
Depression
48
(34
per
cent)
Irritability
41
(29
per
cent)
29
(20
per
cent)
per
cent)
Increased
nervous
Elevated
N = Symptomatic
irritability recurrent
Withdrawal psychosis (Schiziphreniform Less
Table III. Symptomatology associatedwith drug use
well-being
Lethargy
mood
Lethargy,
disinterest
in
Increased
well-being
and
environment vigor
Sexual behavior Decreased interest Decreased
orgasmic
Increased
interest
16 capacity
1
( 1 nymphomania)
Other phenomena Moon facies, hirsutism, gain 30 pounds
4
weight
32
(24
Decreased
sexual
desire
21
(15
per
cent)
Decreased
sexual
orgasm
14 (10
per
cent)
Increased
sexual
desire
10
( 7 per
cent)
Increased
sexual
orgasm
14
(10
per
cent)
14
(10
per
cent)
10 ( 7 per
cent)
14
per
cent)
3 ( 2 per
cent)
Felt
better
Felt
worse
off drugs off drugs
Increased
well-being
Increased
orgasm
only only
(10
216
Relief
of rheumatoid
Menopausal
phenomena
arthritic (hot
symptoms flashes,
Visual hallucinosis of geometric & migraine headache following No
changes Total
No.
etc.)
2
patterns 1
reported
1
of patients
11
Table II Age
27.3
yr.
Education
15.4
yr.
Parity
1.6
students were trained in conducting the interviews with the volunteer population. The interview was initially open-ended, asking only about the experience of the patient with whatever drug she used. After her spontaneous comments were recorded, a standard questionnaire was administered, which included questions about attitude toward, preparation for, and reaction to menarche, as well as length and regularity of the menstrual cycle. Nine symptoms that comm~&y occur at the menseswere rated on a none-
slight-severe basis, and impairment of activity at the menseswas scored on a 5 point scale. Information about frequency of sexual intercourse and orgasm was also elicited, and change with drug noted. An obstetric history and a history of postpartum reactions, if any, were recorded. Type of drug used was elicited, and questions relating to the presence of increased well-being, weight change, depression, anxiety, and irritability and reactions to stopping drug to permit menseswere scored on a none-slight-severe basis. Enquiry was made about a history of prior psychiatric care. The volunteer population (N=lOl) was compared with the data gathered from an obstetric sample of 90 pregnant women, of whom 38 reported using oral contraceptive agents. Since the two populations were similar in regard to demographic variables and showed no statistical differences in occurrence of complaints, they were treated as a single group. Table II showsthat our sample is a relatively young, highly educated group with relatively small families. This is probably accounted for by the fact that many of
Volume 102 Number 7
Psychiatric
reactions
to oral contraceptives
1057
Table IV. Relationship of symptoms to background variables Symptomatic Inadequate preparation for menses Untoward reaction to menarche Irregular periods Period length Irritability Cramps Depression Bodily swelling Anxiety Reported disability at menses Sexual intercourse l/week Menstrual symptoms relieved by medication Infrequent (‘/s of time or less orgasm) * = difference
during
these studies.
Depression
(34 per cent), irritability (29 per cent), and lethargy (23 per cent) were most common. Reports of decreased sexual desire (15 per cent) exceed the number of women reporting increased sexual desire (7 per cent). Changes in capacity for orgasm were equal in frequency of occurrence. Becausepatients had previously reported reactions to stopping drugs to permit their periods, we asked about this, and 17 per cent of the patients reported feeling better or worse when they went off the drugs to permit their menses.While 14 of the patients reported only that they felt increased well-being and vigor while on the medication, 15 reported feeling this way in addition
(27 per cent) (40 per cent) (30 per cent) (80 (67 (60 (80 (40 (17 (18 (42 (14
per per per per per per per per per
cent) cent) cent) cent) cent) cent) cent) cent) cent)
49 29* 14 17 5.2 34 30 25 34 23 2* O* IO* A* 33
(59 per cent) (28 per cent) (34 per cent) (69 (61 (51 (69 (47 ( 4
per per per per per
cent) cent) cent) cent) cent)
per
cent)
(20 per cent) ( 6 per cent)
significant at p 0.5 (x2).
these women are graduate student or young faculty wives. Table III showsthe incidence of reported subjective complaints in the 139 patients interviewed
90 25 36 27 5.1 73 61 54 74 36 16 17 38 10 ii
( Nonsymptomatic
to reporting
other
more
unpleasant
symptoms, which we think probably reflects the lability of feeling that characterizes the response of many women to these medications. In other words, it may reflect that there is no definitive fixed change in the way they feel, but rather much greater variability and increased susceptibility to stress. Table IV shows the comparison on other
in that (1) they reported more disability at the menses, (2) sexual intercourse was significantly less frequent in this group, (3) women who reported relief of premenstrual symptomatology with drug use also reported more complaints with drug use, and (4) decreased frequency of orgasm. Fifteen per cent of those women complaining about the effects of the drugs stopped the medication for this reason, while another 10 per cent reported that they had wanted to, but had been dissuadedfrom doing so by their physician. Perhaps the most extreme of these was the wife of a member of the health professions who complained to her husband that she had become depressed and was having suicidal ideas. She had never felt this way before and wondered if the oral contraceptive medication might be involved. He assuredher that this could not be so, becausesuch complaints were not included in the list of side effects on the package insert which accompanied the medication. The data presented do reflect more morbidity than has been reported previously. Sixty-four per cent of the sample studied perceived change in themselves, most often of an adverse kind. Becauseof the extremely
data collected, especially with relation to symptomatology at the menses. Those ad-
complex issues involved, it is hard to be certain just how serious the problems are. While 25 per cent of those complaining felt
versely
badly enough to stop the drugs, it was clear
affected
were
significantly
different
1058
Kane
from the clinical interviews that some other women felt they were choosing between poor alternatives. The wife of a faculty member had quite severe problems with breast pain and mood changes of irritability and depression, which alarmed her husband, When these were discussed with her, it was found that other contraceptive methods were unacceptable or unavailable to her for physical reasons and that she felt her present difficulties to be less than those that might be occasioned by another pregnancy at this time. Those who report themselves to be most disabled at the menses do seem to be more at risk, as do those with a lower rate of sexual intercourse, suggesting an interaction of endocrine and psychological factors. Relief from premenstrual symptomatology also seems to predict future difficulty with the medication. It would seem that their demonstrated sensitivity to vicissitudes of hormone levels, especially at the menses, predicts their response to any alteration in hormonal state. None of these women reported previous psychiatric illness or care, but menstrual disability and infrequent sexual intercourse are frequent correlates of neurotic disorders in women. This is strengthened somewhat by our previous finding that in the 5 women who have had psychotic episodes in temporal association with drug use 4 had previous psychiatric disease. In a pilot double blind study of drug effects on 3 women,“7 severity of symptomatology seemed related to the level of neurotic conflicts around femininity, though drug effects of an undesirable kind were present in al1 3. Another study has reported that depression with the menses is predictive of depression with drug use,38 a fact not borne out by our study. There is a trend in that direction in our data, but it was not statistically significant. Such a finding should not be entirely unexpected, however, since similar phenomena have been reported in connection with antihypertensive agents, especially those affecting catecholamine metabolism and, indirectly, central sympathetic functioning. Reserpine and alpha methyldopamine
(Aldomet) ‘among other:, have been associated with serious depressive illnesses, especially in persons who have suffered pmvious depressive illness.“3 Myerson,“” in some animal experiments, has suggested that the mechanism of action of estrogen and progesterone may be dependent on biogtmic amines. He has shown that estrous behavior in ovariectomized animals, normally activitated by estrogen and progesterone, will also be activated by estrogen in combination with tetrabenzine and reserpine. In previous experiments, 41 he had shown that estrogen plus inhibited monoamine oxidase inhibitors sexual behavior. He sees this as most easily explained by the assumption of monoaminedependent pathways mediating heat inhibition. Increased levels of monoamines produce decreased sexual response, while a decrease in the biogenic amines increases the heat response. While the presence of estrogen was necessary to elicit this response, it also Suggested that progesterone may act by altering amine levels in brain or receptor response to amines in some way. From other animal experimentation it would seem as though progesterone and guanethidine do exactly this in the uterus.“’ Our datas8 from a one month period of drug use indicated changes in normetanephrine and vanillyl mandelic acid excretion similar in kind, but lesser in degree, to those seen in depressive states.43 This study did not provide for dietary controls and is being repeated under more controlled conditions. A recent reporP4 confirms the influence of gonadal hormones in either sex on norepinephrine metabolism in the brain of rats. Endogenous norepinephrine levels do not change in brain with gonadectomy, but the rate of synthesis is increased. Removal of the pituitary gland did not influence catecholamine synthesis. They concluded that “the pituitary gonadal axis regulates the rates at which brain norepinephrine is synthesized and turns over in the rat,” and “that norepinephrine participates in the central regulation of gonadal function in the species.” While Myerson focuses on biogenic amines, especially catecholamines, it is clear that
Volume Number
102 7
serotonin is also altered under the experimental conditions he describes. This is of more than passing importance, since investigators have shown that serotonin is a precursor of melatonin, a neurohormone of pineal origin with importance for regulation of sexual behavior.45 Estrogen has been shown to depress the enzyme responsible for conversion of serotonin to melatonin.46 Paulson and Robson4? have reported that an MAO inhibitor, phenelzine, commonly used as an antidepressant in humans, prevented implantation in rats. Several related derivatives of phenelzine were effective in interrupting pregnancy. One of these derivatives interrupted the estrous cycle, delayed the onset of estrus in immature female mice, and depressed the weight of sex organs in immature mice. The effect of several of these agents used on pregnant mice could be partially reversed by progesterone. These studies seem to demonstrate that, in animals, the mechanism underlying sexual and reproductive processes in the species studied is intimately bound up with the function of biogenie amines. While less attention has been paid to the other behavioral effects of such manipulations on animals, alterations in biogenie amines, especially in the central nervous system, are seen as critical factors in onset and recovery from psychiatric illness, especially serious depressive illness.43 The experiments cited, plus others indicating altered autonomic response with the use of gonadal hormones,48 seem to indicate that emotional disturbance should be, in a certain number of the people using these hormones, an expectation rather than an occasion for surprise. Another area in which there exists unresolved controversy is that of the rise in corticosteroids seen with the use of these hormonal agents. The rises in total plasma corticosteroids are substantial, although most investigator?” believe these to be in bound and therefore physiologically inactive form. Other investigators50, 51 have cast doubt upon this, with one report citing a 2yz fold increase in unbound steroids at the tissue level. If this is so, a relative state of hypercorticism may exist in at least some of the patients. From
Psychiatric
reactions
to oral
contraceptives
1059
our previous experience with the use of exogenous corticosteroids, there will be psychiatric morbidity in a certain number of the population from such use. The reports of the patients, both positive and negative, in our study are very similar to those reported in studies on patients using exogenous corticosteroids. Some of our own clinical data, especially those reflecting the relief of arthritic symptoms and the presence of hirsutism, moon facies, and weight gain, would certainly point to this possibility. The reduction, by 50 per cent, in the amount of corticosteroids used in arthritic patients reported in another study on the use of Enovid would also support this possibility.5” There also exists the possibility that the patient who suffered a psychosis upon withdrawal of Enovid may have been suffering a cortisone withdrawal state. Symptoms of similar but less intense distress were reported by some of our patients. Another problem area of possible relevance to the problems under discussion relates to the increasing difficulties posed by drug interaction. This can be especially problematical where raised levels of corticosteroids are involved, since many commonly used analgesic agents contain salicylates, which compete for binding sitesZ3 and unbind steroids, thereby raising tissue levels and producing a relative hypercorticism. Another area of considerable research interest, especially with regard to depressive states in psychiatric patients, concerns the possibility that alterations in the metabolism of tryptophan and its subsequent metabolism to a serotonin (5 hydroxytryptamine) may be important in the genesis of depressive illness.54 Studies in depressed patients have revealed that they excrete less tryptamine when depressed, and when they recover tryptamine excretion is increased. The administration of tryptamine also has beneficial effects upon the disturbed sleep of depressed patients,55 and the administration of tryptamine plus psychotropic drugs has been shown by some to have a beneficial effect upon the course of depressive illness.56 It is of interest that one metabolic studys7 shows that at least
1060
Kane
one of the combination oral contracepti\~t~ agents produces changes of possible importance in response to a loading dose of tryptophan, possibly through an alteration in tryptophan pyrrolase. The net effect of this is that less 5hydroxytryptamine would be made when the drug is being taken. While balance studies on normal women indicate there is no change in the basal excretion, but that the change is seen only in response to a loading dose of tryptophan, they do raise the question of whether some patients, especially those who have been previously depressed, may have deficiencies in tryptamine and 5-hydroxytryptamine as a result of taking this medication. There are also data that indicate that progesterone is a central nervous system depressant, and it may be that progestins share some of this capacity. Our data, and those of Murawski and associates3’ would seem to indicate that such is the case for at least some users. EEG slowing has been noted during the luteal phase of the menstrual cycle in humansSs Progesterone produces general anesthesia in various experimental animals in pharmacologic doses when administered intraperitoneally or intravenously.“” This effect seems more marked in females. In humans, 500 mg. intravenous doses in one hour were shown to be effective in inducing sleep of one to 2 hours’ duration.F0 In epileptic patients there was found ( 1) reduction of seizures during the luteal phase of the cycle, (2) increase in seizures immediately before, during, and just after menstruation, when progesterone levels are lowest, and (3) activation of abnormal EEG pattern and precipitation of seizure by exogenous estrogen.61 There was also reported a decrease in seizure with premenstrual progesterone therapy. One would wonder if the water retention of the premenstrual phase is a factor in reducing seizure threshold. Another variable in the complex matrix of factors that may be involved in the psychiatric reactions to oral contraceptive agents concerns the observations by a number of psychoanalytical investigators related to change in the persons’ attitudes toward them-
Svlvvs
and
others
urldc‘r
tlics imlm
1 ~$1 III~‘II-
strual cycle hormonal variation. Benc~lck and Rubenstein”’ showed impressive correlations between dream material and vaginal smears in a group of 15 women studied l)sychoanalytically. During the preovulatory phase, there was increased interest in the environment and increased heterosexual interest on the part of the woman, followed in the postovulatory phase by decreased heterosexual interest, decreased interest in the environment, and increased preoccupation with themselves. Benedek and Rubenstein”’ see this as part of the biologic response in anticipation of pregnancy. Until recently, there have been few studies supporting the changes in sexual behavior, since they did not collect systematic data on sexual behavior from the patients being studied. However, more recently, Michael”” has demonstrated a postluteal fall in the sexual behavior of primates, and Udry and Morris”“ have reported similar data relating to decreased sexual activity and capacity for orgasm in the postovulatory period in a group of women studied. It is of interest that in our double blind study of 3 women using oral contraceptives the psychiatrist doing objective ratings noted a decreased concern with grooming and appearance. This decreased interest in being attractive to the opposite sex may be the parallel in humans of the altered responsiveness of male primates to their female partners during the luteal phase of the cycle, reported in the study of Michael. Benedek and Rubenstein”” also pointed out that, because of these ho’rmonal changes, there was an intensification of conflicts related to their own mother-child relationship. Insofar as these agents do produce a state of pseudopregnancy, it is conceivable that, as a result of what is essentially a markedly prolonged luteal phase with the combined agents, there would be even further intensification and involvement of the individual with conflicts. Another variable contributing to the overall adjustment to oral contraceptive agents is the use to which they may be put in the complex relationship between husband and
Volume Number
102 7
wife. Zell and Crispe5 categorized the fears associated with drug use : 1. Fears about bodily damage. While there can be some legitimate concern about such a possibility, these may also be a more superficial manifestation of resentment related to castration conflicts in women. Anything that might provoke physical change would be seen as a threat, especially if it were more likely to remind one more forcibly of their deprived female status. Such drugs might also tend to diminish the possibility of using sexuality to manipulate a relationship with a man. 2. Fear of loss of control of sexual impulses and latent fantasies of prostitution. Such fears may account for some of the loss of sexual desire and capacity for orgasm seen in our sample, as a further defense against the feared sexual impulses which can no longer be suppressed or otherwise controlled by concerns about pregnancy. These had previously been useful as a defense against guilt-provoking sexual experiences. Our finding that people who have less frequent sexual intercourse and less frequent orgasm report more emotional disturbance would support such a premise. 3. Fears about future fertility. Here again a seemingly realistic fear may reflect concern about interference with an important source of self-esteem for all women; their ability to bear children. In certain cultural groups, this is an important source of self-esteem to the woman and her partner and has actually contributed to the non-use of any method of contraception, often where they are most needed. The fears about altered meno’pause may also be related to such an unconscious attitude. Psychiatrists are especially aware of such phenomena, since it so often is a factor of importance in those middle life depressions and has been shown to be an important source of morbidity in reaction to operationss6 (salpingectomies, hysterectomies, etc.) where loss of procreative function is a byproduct, whether intended or not. Similar considerations apply to those women who have strong unconscious wishes for children and respond to the thwarting of these wishes
Psychiatric
reactions
to oral contraceptives
1061
with depression and other manifestation of psychological conflict. To my knowledge, only one prospective study of oral contraceptive users has been done which compared drug users to users of other forms of contraception, This study indicated a 10 per cent incidence of depression serious enough to cause cessation of use of drug. 38 Data analysis is incomplete on this study at present. Murawski and associates30 questioned the ethics and propriety of doing systematic double blind studies with oral contraceptive agents. Such studies are difficult and hard to manage, but they are not impossible, especially if the drugs are not being relied upon for contraceptive purposes. While this may not tell you as much about the interaction of drug and personality factors as when they are being used for contraception, I think it can give you a great deal of information about the effects of the drugs themselves, which will be important knowledge indeed. I commented earlier on our findings on 3 such patients studied for a 3 month period, one month of which involved the use of a combination oral contraceptive agent. These studies are laborious and difficult, but they should and must be done. Even here, we found an interaction between drug and personality in that those who seem most “conflicted” about their femininity seemed to experience the most symptomatology. Careful prospective studies, including double blind studies, remain to be done. For the present, it would seem important to evaluate carefully complaints of emotional change associated with oral contraceptive use, especially if these are persistent. Psychiatric consultation can be helpful in evaluating which of these reactions may be oaly incidentally connected with drug use or where the psychologic make-up of the patient is such that characterologic alteration by psychotherapeutic means may be necessary before the drugs can be tolerated without guilt or anxiety. Summary
The available literature on emotional disturbance concurrent with the use of oral
1062
Kane
contraceptive agents has been reviewed, which indicates a low incidence of such disturbance as well as a low incidence of altered sexual desire and orgasmic response. Our own data indicate these figures may be too conservative, since more than 50 per cent of the population studied reported adverse effects, at least one quarter of whom felt badly enough to wish to stop the drug. Feelings of depression, irritability, and lethargy are most common, while disturbances in desire for and enjoyment of sex are somewhat less frequent.
I’sychotic reactions do occur. sccmin~ly mom often in patients who are known to have suffered emotional disturbances in the past. One psychosis occurred upon withdrawal of large amounts of hormone used to control endometriosis. The evidence to date suggests an interaction of psychological and endocrine factors in the genesis of these disorders which may necessitate careful psychiatric evaluation before continued drug use can be advised.
REFERENCES
1.
2. 3. 4.
Food and Drug Administration Report on Oral Contraceptives by the Advisory Committee on Obstetrics and Gynecology, FDA, Aug. 1, 1966. Available from the United States Government Printing Office, Washington, D. C. 20402. Pincus, G.: Science 153: 493, 1966. Ryder, N. B., and Westoff, C. F.: Science 153: 1199, 1966. Lasagna, L.: Perspect. Biol. & Med. 7: 457,
17.
Paffenbarger, 161, 1961.
la.
Hamilton, J. A.: Postpartum Psychiatric Problems, St. Louis, 1962, The C. V. Mosby Com-
19. 20.
21.
1964, 5. 6. 7.
a.
9. 10. 11.
12.
13. 14. 15. 16.
Barr, D. P.: J. A. M. A. 159: 1452, 1955. Schimmel. E. M.: T. Chron. Dis. 16: 1, 1963. Hertz, Roy: Epidemiologic and Experimental Aspects of Oral Contraceptives? Presented to the New England Obstetrical and Gynecological Society,-Boston, November, 1967. Whitelaw. M. Tames: Should Nullinara and Infertility Patiints Be Given Oral Contraceptives, presented to the New England Obstetrical and Gynecological Society, Boston, November, 1967. Laragh, J. H., et al.: J. A. M. A. 201: 918, 1967. Woods, J. W.: Lancet 2: 653, 1967. Hornestein, Simon: Serious Neurological Problems Associated with Oral Contraceptives, presented to the New England Obstetrical and Gynecological Society, Boston, November, 1967. Spellacy, William N.: Carbohydrate Metabolism and the Pill, presented to the New England Obstetrical and Gynecological Society, Boston, November, 1967. Kane, ‘F. J., Daiy, R. J., Ewing, J., and Keel&, M.: Brit. J-Psych&. 113: 265, -1967. Bovd. D. A.. Tr.: AM. T. OBST. & GYNEC. 43: 143; 43: 335, 1942.” Jannson, B.: Acta psychiat. scandinav. 39: 10, 1963. (Suppl. 172.) Jarrahi, A., Kane, F. J., Van de Castle, R., and Ewing, J.: Emotional and Cognitive Disturbance During the Childbearing Period, presented at the APA Annual Meeting, Detroit, May, 1967.
22. 23.
24. 25. 26. 27.
R.
S., Jr.:
J.
Chron.
Dis.
13:
pany, p. la. Tetlow, C.: J. Ment. SC. 101: 629, 1955. Dalton, K.: The Premenstrual Syndrome, Springfield, Illinois, 1964, Charles C Thomas, Publisher. Mandell, A. J., and Mandell, M. P.: J. A. M. A. 280: 792, 1967. Glick, I. D.: Psychopharmacologia 10: 363, 1967. Bigelow, N.: American Handbook of Psychiatry, New York, 1960, Basic Books, p. 540. Rome, H. P., and Braceland, F. J.: Am. J. Psvchiat. 108: 641. 1952. Glaser, G. H.: Psychosom. Med. 15: 280, 1953. Wearing, M. P.: Canadian M. A. J. 89: 239, 1963. Moos, R.: Psychologic Aspects of Oral Contraceptives, presented to the American Psychosomatic Society, April, 1967.
28.
Kane, F., Daly, R., Wallach, M., and Keeler, M.: Dis. Nerv. System 27: 339, 1966.
29.
Bakker, Cl. B., and Dightman, C. R.: Obst. & Gvnec. 28: 373. 1966. Murawski, B. J., et al.: Fertil. & Steril. 19: 50, 1968. Sturgis, S. H.: M. Aspects Hum. Sexuality 2: 4, 1968. Bakke, J. L.: Pacific Med. & Surg. 73: 200, 1965. Scott, J., and Brass, P.: AM. J. OBST. & GYNEC. 95: 1166, 1966. Keeler, M. H., Daly, R., and Kane, F. J.: Am. J. Psvchiat. 120: 1123. 1964. Daly,-R. J., Kane, F. J., and Ewing, J. A.: Lancet 2: 444. 1967. Idestrijm, Cl. M.: Lancet 1: 718, 1966. Marcotte, D., Kane, F. J., Van de Castle, R., and Lipton, M.: Unpublished data. Coppen, A.: Personal communication.
30. 31. 32. 33. 34. 35.
it 38.
Volume Number
39. 40. 41. 42.
43. 44. 45. 46.
47. 48. 49. 50. 51. 52.
102 7
Psychiatric
Bunney, W. E., and Davis, John M.: Arch. Gen. Psychiat. 13: 483, 1965. Myerson, B.: Psychopharmacologia 6: 2 10, 1964. Myerson, B.: Arch. internat. Pharmacodyn. et therap. 150: 4, 1964. Keug-Sup, Cha, Woo-Choo, Lee, Rudizk, A., and Miller, J. -W.: J. Pharmacol. & Exper. Theran. 148: 9. 1965. Schildkraut, J.‘J., and Kety, S. S.: Science 156: 21, 1967. Anton-Tay, F., and Wurtman, R. J.: Science 159: 1245, 1968. Relkin, R.: New England J. Med. 274: 944, 1966. Wurtman, R. J.: Isolation and Synthesis of Melatonin, presented at the Symposium on Indole Alkylamines, New York, New York, May, 1967. Paulson, E., and Robson, J. M.: J. Endocrinol. 30: 205, 1964. Fullerton, A., and Morrison, J. F.: J. Endocrinol. 33: 75, 1965. Plager, J. E., Schmidt, K. G., and Staubitz, W. J.: J: Clin. Invest. 43: 1066, 1964. Wvnn. V.. and Doar. 1. W. H.: Lancet 2: 715, i966: Zinneman, H. H., Seals, U. S., and Doe, R. P.: J. Clin. Endocrinol. 27: 397, 1967. Gilbert, M., Rotstein, J., Cunningham, C., I
53. 54.
55. 56. 57. 58. 59. 60. 61. 62. 63. 64.
”
65. 66.
reactions
to oral
contraceptives
1063
Estrin, I., Davidson, A., and Pincus, G.: J. A. M. A. 190: 235, 1964. Brodie, B. B.: Proc. Roy. Sot. Med. 58: 946, 1965. Coppen, A.: Depressed States and Indole Alkylamines, presented at the Symposium on Indole Alkylamines, New York, New York, May, 1967. Oswald, I.: Psychosom. Res. 9: 111, 1965. Pare, C. M. B.: Lancet 2: 527, 1963. Price, J. M., et al.: Am. J. Clin. Nutrition 20: 452, 1967. Gibbs, F. A., and Reid, D. E.: AM. J. OBST. & GYNEC. 44: 672, 1942. Selye, H. J.: J. Pharmacol. & Exper. Therap. 73: 127, 1941. Merryman, W.: J. Clin. Endocrinol. 14: 1957, 1954. Morrell, F.: Neurology 9: 352, 1959. Benedek, T., and Rubenstein, G. B.: Psychosom. Med. 1: 245. 1939. Michael, R.: Proc. Roy. Sot. Med. 58: 595, 1965. Udry, J. R., and Morris, N. M.: Distribution of Coitus in the Menstrual Cycle, presented at the Meetinas of the Ponulation Association of America, Cincinnati, dhio, April, 1967. Zell, J., and Crisp, W. E.: Obst. & Gynec. 23: 657, 1964. Ellison, R. M.: M. J. Australia 2: 625, 1964.
J.
KANE,
JR.,
M.D.
Chapel Hill, North Carolina
1 N I TIA L studies1-3of the oral contraceptive agents seemed to reveal few serious side effects of these hormonal combinations, while providing excellent conception control. More experience has raised the question of whether these early estimates reflected fully the problems which could and/or should have been anticipated with acute and chronic use. It seemslikely that the enormous social issuesof international scope to which these drugs are intimately bound have contributed to the difficulties of dispassionate and full assessmentof responseto these drugs. There would seem to be no reason to deny these drugs their contribution to “The Diseasesof Medical Progress,” an apt term quoted by Lasagna4 for those illnessesnewly generated by our potent therapeutic agents. He cites several studies5*6 indicating that 5 to 15 per cent of first admissions to some hospitals result from such aberrations of our therapeutic efforts. For the oral contraceptive agents, a place now seemsassured. Clinical and experimental reports indicate that thromboembolic disorder,r prolonged amenorrhea,s hypertensive reactions,g*lo neurological disorders,ll disturbance in carbohydrate metabolism,12and emotional disturbances13may accompany their use. In view of the large literature on mood and behavioral alterations in relation to gonadal hormone variation in women, such
From the Department of Psychiatry, University of North Carolina.
reactions should really come as no surprise. Psychiatric reactions,141 l5 especially depression, in the early postpartum period are commonplace in the practice of obstetriciangynecologists. In a recent series of 90 pregnant women studied at this hospital,la 28 per cent reported subjective symptoms of anxiety and depression, confirmed by psychological test materials. Somewhat more surprising, however, was the fact that this figure was lessthan half that of the changes reported in the last trimester of pregnancy. Rates of psychosisamong pregnant women are below those rates expected in nonpregnant women.l? In the early postpartum period, milder forms of illness become less frequent and major psychoses,which many psychiatrist?** I9 feel are unusual in their clinical phenomenology, dramatically increase in incidence. The fact that 80 per cent of these psychotic episodes occur within the first 30 days after delivery suggestsa relationship to the dramatically altered endocrine state occurring at the end of pregnancy.17 The menstrual cycle is also associatedwith alterations in mood and behavior, especially immediately prior to and during the menses. A number of studieszOl21 have documented the markedly increased incidence of psychiatric illness during this period, including suicides,suicide attempts, homicides, and admissions for acute psychosis. Some have attributed this to the decline in hormones taking place during this time. The relief of premenstrual symptomatology by the oral 1053
1054
Kane
contraceptive agents is striking in many studieP and may be related to the shortening of the period when there are low levels of hormones. Another period of major psychosocial endocrine change for a woman is the menopause, which also is associated with a marked increase in incidence of major and minor psychiatric illness. The incidence of such illness, especially so-called involutional psychotic reactions, is more than 3 to 1 in favor of women.23 The use of exogenous corticosteroid hormones has also been associated with notable mood and behavioral change in a certain number of people receiving these drugs.24* 25 A variety of studies have reported mild to moderate euphoria and increased well-being as well as tiredness, depression, and increased irritability. Frank manic and depressive psychotic episodes with higher dosages of steroids are not unusual. Previous studies Glick22 surveyed the available literature on the use of oral contraceptive agents. Nineteen studies, mostly by obstetrician-gynecologists, revealed sporadic reporting of emotional distress, with a 5 per cent incidence of depression being the highest reported. There were more favorable comments reported, though these comments were likewise largely anecdotal. .4 similar scattering of reports from the same clinical material showed increased “libido,” increased satisfaction with coitus, increased control of dysmennorrhea, and decreased premenstrual tension. Changes of an adverse kind were much less frequent, but they were reported for all except premenstrual tension symptoms. Wearingz6 reported a 16 per cent incidence of depression in 62 patients, and when it occurred, the depression became more prominent the longer the patient remained on drug. MooP reported a questionnaire survey of women asked to rate a variety of symptoms separately for menstrual, premenstrual, and inter-menstrual phases of their most recent cycle and for their worst cycle. He reported that most reported slight decrease in men-
Am.
December 1, 1968 J. Obst. & Gynec.
strual symptoms, while IO per cent had a significant increase. Sequential hormone users reported more symptoms of water retention and negative feeling than the combination drug group. Four reports surveyed by Glickz2 described drug use in psychiatric populations, where it was hoped their use would help patients who had exacerbation or worsening of psychiatric symptoms associated with the menses. These reported were generally favorable, though they lacked controls in most instances. Wez8 have reported previously upon one patient with marked premenstrual exacerbation of psychiatric symptoms who improved on Enovid and showed relapse on placebo in a double blind study. Bakker and DightmarP report a study of 100 women and conclude that while fluid retention was a significant problem in 30 per cent of the subjects studied, depression and changes in libido could not be related to medication as such. Murawski and associateP report a longitudinal study of 72 women over a 15 month period, using a variety of instruments for subjective and objective feelings and behavior. Marked behavioral change was not reported, and reactions seemed related to a number of factors (disappointment with magical expectations of drug, persistent frigidity in spite of relief of fear of pregnancy, etc.). Sedation was reported by some patients, with dramatic behavioral alterations in at least one patient. The investigators attributed this to disturbance of customary modes of adapting, based on compulsive working. However, similar kinds of indifference to usual concerns are common with the use of a variety of stimulant and depressant drugs. The formal report of the study did not include the reported exacerbation of psychosis reported elsewhere by SturgisIB1 a somewhat puzzling omission. Bakke32 reports a double blind cross-over study of an estrogen-progestin mixture, estrogen, and placebo. Twenty of 27 women preferred either estrogen or the oral contraceptive agent, while only 3 preferred placebo (reasons given, related in the patients’ words: “to feeling marvelous, with much more zest”
Volume Number
102 7
Psychiatric
and “having increased sexual drive”), The reasons for preference of the placebo related to relief of undesired hormone effects of the In other estrogen-progestin combination. words, the patient felt worse on the drug, with tender breasts and poor sleep, all of which improved when they stopped. The evaluation of the placebo period was very difficult because of the withdrawal effects which were so common. Complaints of moodiness, being cross and tired, alterations in sexual drive, weight gain, edema, and insomnia were commonest in the group using the estrogen-progestin group. Twelve women reported increased sexual drive. Six women who enjoyed the increase in drive continued on the drug, while 6 rejected the drugs because of this same change. Scott and Bra@ reported their experience in the use of massive doses of norethynodrel in the treatment of endometriosis on 20 patients. Daily dose varied from 40 to 100 mg. of norethynodrel. They commented that “mood and behavior changes were noted in all patients, and 3 developed moderate depressions that responded to antidepressant drugs. One patient also developed a tremendous increase in libido.” Psychotic
reactions
Our own interest in these drugs was stimulated by the observation of a psychotic episode upon withdrawal of 30 mg. daily of Enovid used to control endometriosis.84 The patient experienced relief of her psychotic behavior with drug replacement and intensification of her psychosis when placebo was substituted. We have seen 4 additional women since that time who have suffered psychotic episodes coincident with the use of oral contraceptive agents. In 2 instances35 these were associated with the use of the combined agents and in 2 with the use of sequential hormone combinations. One patient noted change with both, although she became hospitalized while using the sequential agent. Three of these women gave a history of postpartum disturbance in the past and one was receiving psychotherapy for a character disorder. One patient tolerated
reactions
to oral
contraceptives
1055
sequential agents very well for a year but suffered psychotic episodes with only one month of combination drug use on separate occasions before and after the use of the sequential agents. This would seem to raise a question as to whether there is not some increased risk in using these drugs in persons with known psychiatric illness. There have been two additional reports of psychotic reactions in the world literature. Idestriim36 reported another case of psychosis upon withdrawal of drug; in this case, norethindrone. Sturgis31 reports recurrence of psychosis in a patient participating in the prospective study of drug effects by Murawski and associate9O described earlier. This patient was able to tolerate continued use of drugs. It is not clearly stated whether she received psychopharmacologic agents during her illness. Further
clinical
studies
Our first study13 involved the use of c1inica.i interviews and personality testing on 50 women referred by physicians in the community and the medical school complex. Eleven of these patients were psychiatric inpatients at the time of interview and the patients had used only combination drugs. The clinical findings from the study are summarized in Table I. Twenty-eight patients reported undesirable effects while 11 felt better on drugs and 11 reported no change. Depression and decreased sexual interest were noted most frequently. Unusual stress was not reported but rather increased susceptibility to usual stress. Moral conflicts seemed negligible as a factor, since most previously had used mechanical contraceptive devices. Only the nymphomaniac patient reported an evidenced guilt over her increased sexual drive, and she was the only one who showed marked changes on psychological testing. This study had obvious defects, especially in the nature of the selectivity of the sample. A second study was therefore conducted, sampling widely from the mothers in our Obstetric-Gynecologic Clinic, Pediatrics Department, and a busy pediatric practice. Two sophomore medical
1056
December 1, 1968 Am. J. Obst. & Gynec.
Kane
Table I. Mood and behavioral change with the use of oral contraceptive agents Emotional
1 Patients
change
Depression, tearfulness, ( 1 hospitalized for depression)
19 type)
139
=
90 (64
per
cent)
Depression
48
(34
per
cent)
Irritability
41
(29
per
cent)
29
(20
per
cent)
per
cent)
Increased
nervous
Elevated
N = Symptomatic
irritability recurrent
Withdrawal psychosis (Schiziphreniform Less
Table III. Symptomatology associatedwith drug use
well-being
Lethargy
mood
Lethargy,
disinterest
in
Increased
well-being
and
environment vigor
Sexual behavior Decreased interest Decreased
orgasmic
Increased
interest
16 capacity
1
( 1 nymphomania)
Other phenomena Moon facies, hirsutism, gain 30 pounds
4
weight
32
(24
Decreased
sexual
desire
21
(15
per
cent)
Decreased
sexual
orgasm
14 (10
per
cent)
Increased
sexual
desire
10
( 7 per
cent)
Increased
sexual
orgasm
14
(10
per
cent)
14
(10
per
cent)
10 ( 7 per
cent)
14
per
cent)
3 ( 2 per
cent)
Felt
better
Felt
worse
off drugs off drugs
Increased
well-being
Increased
orgasm
only only
(10
216
Relief
of rheumatoid
Menopausal
phenomena
arthritic (hot
symptoms flashes,
Visual hallucinosis of geometric & migraine headache following No
changes Total
No.
etc.)
2
patterns 1
reported
1
of patients
11
Table II Age
27.3
yr.
Education
15.4
yr.
Parity
1.6
students were trained in conducting the interviews with the volunteer population. The interview was initially open-ended, asking only about the experience of the patient with whatever drug she used. After her spontaneous comments were recorded, a standard questionnaire was administered, which included questions about attitude toward, preparation for, and reaction to menarche, as well as length and regularity of the menstrual cycle. Nine symptoms that comm~&y occur at the menseswere rated on a none-
slight-severe basis, and impairment of activity at the menseswas scored on a 5 point scale. Information about frequency of sexual intercourse and orgasm was also elicited, and change with drug noted. An obstetric history and a history of postpartum reactions, if any, were recorded. Type of drug used was elicited, and questions relating to the presence of increased well-being, weight change, depression, anxiety, and irritability and reactions to stopping drug to permit menseswere scored on a none-slight-severe basis. Enquiry was made about a history of prior psychiatric care. The volunteer population (N=lOl) was compared with the data gathered from an obstetric sample of 90 pregnant women, of whom 38 reported using oral contraceptive agents. Since the two populations were similar in regard to demographic variables and showed no statistical differences in occurrence of complaints, they were treated as a single group. Table II showsthat our sample is a relatively young, highly educated group with relatively small families. This is probably accounted for by the fact that many of
Volume 102 Number 7
Psychiatric
reactions
to oral contraceptives
1057
Table IV. Relationship of symptoms to background variables Symptomatic Inadequate preparation for menses Untoward reaction to menarche Irregular periods Period length Irritability Cramps Depression Bodily swelling Anxiety Reported disability at menses Sexual intercourse l/week Menstrual symptoms relieved by medication Infrequent (‘/s of time or less orgasm) * = difference
during
these studies.
Depression
(34 per cent), irritability (29 per cent), and lethargy (23 per cent) were most common. Reports of decreased sexual desire (15 per cent) exceed the number of women reporting increased sexual desire (7 per cent). Changes in capacity for orgasm were equal in frequency of occurrence. Becausepatients had previously reported reactions to stopping drugs to permit their periods, we asked about this, and 17 per cent of the patients reported feeling better or worse when they went off the drugs to permit their menses.While 14 of the patients reported only that they felt increased well-being and vigor while on the medication, 15 reported feeling this way in addition
(27 per cent) (40 per cent) (30 per cent) (80 (67 (60 (80 (40 (17 (18 (42 (14
per per per per per per per per per
cent) cent) cent) cent) cent) cent) cent) cent) cent)
49 29* 14 17 5.2 34 30 25 34 23 2* O* IO* A* 33
(59 per cent) (28 per cent) (34 per cent) (69 (61 (51 (69 (47 ( 4
per per per per per
cent) cent) cent) cent) cent)
per
cent)
(20 per cent) ( 6 per cent)
significant at p 0.5 (x2).
these women are graduate student or young faculty wives. Table III showsthe incidence of reported subjective complaints in the 139 patients interviewed
90 25 36 27 5.1 73 61 54 74 36 16 17 38 10 ii
( Nonsymptomatic
to reporting
other
more
unpleasant
symptoms, which we think probably reflects the lability of feeling that characterizes the response of many women to these medications. In other words, it may reflect that there is no definitive fixed change in the way they feel, but rather much greater variability and increased susceptibility to stress. Table IV shows the comparison on other
in that (1) they reported more disability at the menses, (2) sexual intercourse was significantly less frequent in this group, (3) women who reported relief of premenstrual symptomatology with drug use also reported more complaints with drug use, and (4) decreased frequency of orgasm. Fifteen per cent of those women complaining about the effects of the drugs stopped the medication for this reason, while another 10 per cent reported that they had wanted to, but had been dissuadedfrom doing so by their physician. Perhaps the most extreme of these was the wife of a member of the health professions who complained to her husband that she had become depressed and was having suicidal ideas. She had never felt this way before and wondered if the oral contraceptive medication might be involved. He assuredher that this could not be so, becausesuch complaints were not included in the list of side effects on the package insert which accompanied the medication. The data presented do reflect more morbidity than has been reported previously. Sixty-four per cent of the sample studied perceived change in themselves, most often of an adverse kind. Becauseof the extremely
data collected, especially with relation to symptomatology at the menses. Those ad-
complex issues involved, it is hard to be certain just how serious the problems are. While 25 per cent of those complaining felt
versely
badly enough to stop the drugs, it was clear
affected
were
significantly
different
1058
Kane
from the clinical interviews that some other women felt they were choosing between poor alternatives. The wife of a faculty member had quite severe problems with breast pain and mood changes of irritability and depression, which alarmed her husband, When these were discussed with her, it was found that other contraceptive methods were unacceptable or unavailable to her for physical reasons and that she felt her present difficulties to be less than those that might be occasioned by another pregnancy at this time. Those who report themselves to be most disabled at the menses do seem to be more at risk, as do those with a lower rate of sexual intercourse, suggesting an interaction of endocrine and psychological factors. Relief from premenstrual symptomatology also seems to predict future difficulty with the medication. It would seem that their demonstrated sensitivity to vicissitudes of hormone levels, especially at the menses, predicts their response to any alteration in hormonal state. None of these women reported previous psychiatric illness or care, but menstrual disability and infrequent sexual intercourse are frequent correlates of neurotic disorders in women. This is strengthened somewhat by our previous finding that in the 5 women who have had psychotic episodes in temporal association with drug use 4 had previous psychiatric disease. In a pilot double blind study of drug effects on 3 women,“7 severity of symptomatology seemed related to the level of neurotic conflicts around femininity, though drug effects of an undesirable kind were present in al1 3. Another study has reported that depression with the menses is predictive of depression with drug use,38 a fact not borne out by our study. There is a trend in that direction in our data, but it was not statistically significant. Such a finding should not be entirely unexpected, however, since similar phenomena have been reported in connection with antihypertensive agents, especially those affecting catecholamine metabolism and, indirectly, central sympathetic functioning. Reserpine and alpha methyldopamine
(Aldomet) ‘among other:, have been associated with serious depressive illnesses, especially in persons who have suffered pmvious depressive illness.“3 Myerson,“” in some animal experiments, has suggested that the mechanism of action of estrogen and progesterone may be dependent on biogtmic amines. He has shown that estrous behavior in ovariectomized animals, normally activitated by estrogen and progesterone, will also be activated by estrogen in combination with tetrabenzine and reserpine. In previous experiments, 41 he had shown that estrogen plus inhibited monoamine oxidase inhibitors sexual behavior. He sees this as most easily explained by the assumption of monoaminedependent pathways mediating heat inhibition. Increased levels of monoamines produce decreased sexual response, while a decrease in the biogenic amines increases the heat response. While the presence of estrogen was necessary to elicit this response, it also Suggested that progesterone may act by altering amine levels in brain or receptor response to amines in some way. From other animal experimentation it would seem as though progesterone and guanethidine do exactly this in the uterus.“’ Our datas8 from a one month period of drug use indicated changes in normetanephrine and vanillyl mandelic acid excretion similar in kind, but lesser in degree, to those seen in depressive states.43 This study did not provide for dietary controls and is being repeated under more controlled conditions. A recent reporP4 confirms the influence of gonadal hormones in either sex on norepinephrine metabolism in the brain of rats. Endogenous norepinephrine levels do not change in brain with gonadectomy, but the rate of synthesis is increased. Removal of the pituitary gland did not influence catecholamine synthesis. They concluded that “the pituitary gonadal axis regulates the rates at which brain norepinephrine is synthesized and turns over in the rat,” and “that norepinephrine participates in the central regulation of gonadal function in the species.” While Myerson focuses on biogenic amines, especially catecholamines, it is clear that
Volume Number
102 7
serotonin is also altered under the experimental conditions he describes. This is of more than passing importance, since investigators have shown that serotonin is a precursor of melatonin, a neurohormone of pineal origin with importance for regulation of sexual behavior.45 Estrogen has been shown to depress the enzyme responsible for conversion of serotonin to melatonin.46 Paulson and Robson4? have reported that an MAO inhibitor, phenelzine, commonly used as an antidepressant in humans, prevented implantation in rats. Several related derivatives of phenelzine were effective in interrupting pregnancy. One of these derivatives interrupted the estrous cycle, delayed the onset of estrus in immature female mice, and depressed the weight of sex organs in immature mice. The effect of several of these agents used on pregnant mice could be partially reversed by progesterone. These studies seem to demonstrate that, in animals, the mechanism underlying sexual and reproductive processes in the species studied is intimately bound up with the function of biogenie amines. While less attention has been paid to the other behavioral effects of such manipulations on animals, alterations in biogenie amines, especially in the central nervous system, are seen as critical factors in onset and recovery from psychiatric illness, especially serious depressive illness.43 The experiments cited, plus others indicating altered autonomic response with the use of gonadal hormones,48 seem to indicate that emotional disturbance should be, in a certain number of the people using these hormones, an expectation rather than an occasion for surprise. Another area in which there exists unresolved controversy is that of the rise in corticosteroids seen with the use of these hormonal agents. The rises in total plasma corticosteroids are substantial, although most investigator?” believe these to be in bound and therefore physiologically inactive form. Other investigators50, 51 have cast doubt upon this, with one report citing a 2yz fold increase in unbound steroids at the tissue level. If this is so, a relative state of hypercorticism may exist in at least some of the patients. From
Psychiatric
reactions
to oral
contraceptives
1059
our previous experience with the use of exogenous corticosteroids, there will be psychiatric morbidity in a certain number of the population from such use. The reports of the patients, both positive and negative, in our study are very similar to those reported in studies on patients using exogenous corticosteroids. Some of our own clinical data, especially those reflecting the relief of arthritic symptoms and the presence of hirsutism, moon facies, and weight gain, would certainly point to this possibility. The reduction, by 50 per cent, in the amount of corticosteroids used in arthritic patients reported in another study on the use of Enovid would also support this possibility.5” There also exists the possibility that the patient who suffered a psychosis upon withdrawal of Enovid may have been suffering a cortisone withdrawal state. Symptoms of similar but less intense distress were reported by some of our patients. Another problem area of possible relevance to the problems under discussion relates to the increasing difficulties posed by drug interaction. This can be especially problematical where raised levels of corticosteroids are involved, since many commonly used analgesic agents contain salicylates, which compete for binding sitesZ3 and unbind steroids, thereby raising tissue levels and producing a relative hypercorticism. Another area of considerable research interest, especially with regard to depressive states in psychiatric patients, concerns the possibility that alterations in the metabolism of tryptophan and its subsequent metabolism to a serotonin (5 hydroxytryptamine) may be important in the genesis of depressive illness.54 Studies in depressed patients have revealed that they excrete less tryptamine when depressed, and when they recover tryptamine excretion is increased. The administration of tryptamine also has beneficial effects upon the disturbed sleep of depressed patients,55 and the administration of tryptamine plus psychotropic drugs has been shown by some to have a beneficial effect upon the course of depressive illness.56 It is of interest that one metabolic studys7 shows that at least
1060
Kane
one of the combination oral contracepti\~t~ agents produces changes of possible importance in response to a loading dose of tryptophan, possibly through an alteration in tryptophan pyrrolase. The net effect of this is that less 5hydroxytryptamine would be made when the drug is being taken. While balance studies on normal women indicate there is no change in the basal excretion, but that the change is seen only in response to a loading dose of tryptophan, they do raise the question of whether some patients, especially those who have been previously depressed, may have deficiencies in tryptamine and 5-hydroxytryptamine as a result of taking this medication. There are also data that indicate that progesterone is a central nervous system depressant, and it may be that progestins share some of this capacity. Our data, and those of Murawski and associates3’ would seem to indicate that such is the case for at least some users. EEG slowing has been noted during the luteal phase of the menstrual cycle in humansSs Progesterone produces general anesthesia in various experimental animals in pharmacologic doses when administered intraperitoneally or intravenously.“” This effect seems more marked in females. In humans, 500 mg. intravenous doses in one hour were shown to be effective in inducing sleep of one to 2 hours’ duration.F0 In epileptic patients there was found ( 1) reduction of seizures during the luteal phase of the cycle, (2) increase in seizures immediately before, during, and just after menstruation, when progesterone levels are lowest, and (3) activation of abnormal EEG pattern and precipitation of seizure by exogenous estrogen.61 There was also reported a decrease in seizure with premenstrual progesterone therapy. One would wonder if the water retention of the premenstrual phase is a factor in reducing seizure threshold. Another variable in the complex matrix of factors that may be involved in the psychiatric reactions to oral contraceptive agents concerns the observations by a number of psychoanalytical investigators related to change in the persons’ attitudes toward them-
Svlvvs
and
others
urldc‘r
tlics imlm
1 ~$1 III~‘II-
strual cycle hormonal variation. Benc~lck and Rubenstein”’ showed impressive correlations between dream material and vaginal smears in a group of 15 women studied l)sychoanalytically. During the preovulatory phase, there was increased interest in the environment and increased heterosexual interest on the part of the woman, followed in the postovulatory phase by decreased heterosexual interest, decreased interest in the environment, and increased preoccupation with themselves. Benedek and Rubenstein”’ see this as part of the biologic response in anticipation of pregnancy. Until recently, there have been few studies supporting the changes in sexual behavior, since they did not collect systematic data on sexual behavior from the patients being studied. However, more recently, Michael”” has demonstrated a postluteal fall in the sexual behavior of primates, and Udry and Morris”“ have reported similar data relating to decreased sexual activity and capacity for orgasm in the postovulatory period in a group of women studied. It is of interest that in our double blind study of 3 women using oral contraceptives the psychiatrist doing objective ratings noted a decreased concern with grooming and appearance. This decreased interest in being attractive to the opposite sex may be the parallel in humans of the altered responsiveness of male primates to their female partners during the luteal phase of the cycle, reported in the study of Michael. Benedek and Rubenstein”” also pointed out that, because of these ho’rmonal changes, there was an intensification of conflicts related to their own mother-child relationship. Insofar as these agents do produce a state of pseudopregnancy, it is conceivable that, as a result of what is essentially a markedly prolonged luteal phase with the combined agents, there would be even further intensification and involvement of the individual with conflicts. Another variable contributing to the overall adjustment to oral contraceptive agents is the use to which they may be put in the complex relationship between husband and
Volume Number
102 7
wife. Zell and Crispe5 categorized the fears associated with drug use : 1. Fears about bodily damage. While there can be some legitimate concern about such a possibility, these may also be a more superficial manifestation of resentment related to castration conflicts in women. Anything that might provoke physical change would be seen as a threat, especially if it were more likely to remind one more forcibly of their deprived female status. Such drugs might also tend to diminish the possibility of using sexuality to manipulate a relationship with a man. 2. Fear of loss of control of sexual impulses and latent fantasies of prostitution. Such fears may account for some of the loss of sexual desire and capacity for orgasm seen in our sample, as a further defense against the feared sexual impulses which can no longer be suppressed or otherwise controlled by concerns about pregnancy. These had previously been useful as a defense against guilt-provoking sexual experiences. Our finding that people who have less frequent sexual intercourse and less frequent orgasm report more emotional disturbance would support such a premise. 3. Fears about future fertility. Here again a seemingly realistic fear may reflect concern about interference with an important source of self-esteem for all women; their ability to bear children. In certain cultural groups, this is an important source of self-esteem to the woman and her partner and has actually contributed to the non-use of any method of contraception, often where they are most needed. The fears about altered meno’pause may also be related to such an unconscious attitude. Psychiatrists are especially aware of such phenomena, since it so often is a factor of importance in those middle life depressions and has been shown to be an important source of morbidity in reaction to operationss6 (salpingectomies, hysterectomies, etc.) where loss of procreative function is a byproduct, whether intended or not. Similar considerations apply to those women who have strong unconscious wishes for children and respond to the thwarting of these wishes
Psychiatric
reactions
to oral contraceptives
1061
with depression and other manifestation of psychological conflict. To my knowledge, only one prospective study of oral contraceptive users has been done which compared drug users to users of other forms of contraception, This study indicated a 10 per cent incidence of depression serious enough to cause cessation of use of drug. 38 Data analysis is incomplete on this study at present. Murawski and associates30 questioned the ethics and propriety of doing systematic double blind studies with oral contraceptive agents. Such studies are difficult and hard to manage, but they are not impossible, especially if the drugs are not being relied upon for contraceptive purposes. While this may not tell you as much about the interaction of drug and personality factors as when they are being used for contraception, I think it can give you a great deal of information about the effects of the drugs themselves, which will be important knowledge indeed. I commented earlier on our findings on 3 such patients studied for a 3 month period, one month of which involved the use of a combination oral contraceptive agent. These studies are laborious and difficult, but they should and must be done. Even here, we found an interaction between drug and personality in that those who seem most “conflicted” about their femininity seemed to experience the most symptomatology. Careful prospective studies, including double blind studies, remain to be done. For the present, it would seem important to evaluate carefully complaints of emotional change associated with oral contraceptive use, especially if these are persistent. Psychiatric consultation can be helpful in evaluating which of these reactions may be oaly incidentally connected with drug use or where the psychologic make-up of the patient is such that characterologic alteration by psychotherapeutic means may be necessary before the drugs can be tolerated without guilt or anxiety. Summary
The available literature on emotional disturbance concurrent with the use of oral
1062
Kane
contraceptive agents has been reviewed, which indicates a low incidence of such disturbance as well as a low incidence of altered sexual desire and orgasmic response. Our own data indicate these figures may be too conservative, since more than 50 per cent of the population studied reported adverse effects, at least one quarter of whom felt badly enough to wish to stop the drug. Feelings of depression, irritability, and lethargy are most common, while disturbances in desire for and enjoyment of sex are somewhat less frequent.
I’sychotic reactions do occur. sccmin~ly mom often in patients who are known to have suffered emotional disturbances in the past. One psychosis occurred upon withdrawal of large amounts of hormone used to control endometriosis. The evidence to date suggests an interaction of psychological and endocrine factors in the genesis of these disorders which may necessitate careful psychiatric evaluation before continued drug use can be advised.
REFERENCES
1.
2. 3. 4.
Food and Drug Administration Report on Oral Contraceptives by the Advisory Committee on Obstetrics and Gynecology, FDA, Aug. 1, 1966. Available from the United States Government Printing Office, Washington, D. C. 20402. Pincus, G.: Science 153: 493, 1966. Ryder, N. B., and Westoff, C. F.: Science 153: 1199, 1966. Lasagna, L.: Perspect. Biol. & Med. 7: 457,
17.
Paffenbarger, 161, 1961.
la.
Hamilton, J. A.: Postpartum Psychiatric Problems, St. Louis, 1962, The C. V. Mosby Com-
19. 20.
21.
1964, 5. 6. 7.
a.
9. 10. 11.
12.
13. 14. 15. 16.
Barr, D. P.: J. A. M. A. 159: 1452, 1955. Schimmel. E. M.: T. Chron. Dis. 16: 1, 1963. Hertz, Roy: Epidemiologic and Experimental Aspects of Oral Contraceptives? Presented to the New England Obstetrical and Gynecological Society,-Boston, November, 1967. Whitelaw. M. Tames: Should Nullinara and Infertility Patiints Be Given Oral Contraceptives, presented to the New England Obstetrical and Gynecological Society, Boston, November, 1967. Laragh, J. H., et al.: J. A. M. A. 201: 918, 1967. Woods, J. W.: Lancet 2: 653, 1967. Hornestein, Simon: Serious Neurological Problems Associated with Oral Contraceptives, presented to the New England Obstetrical and Gynecological Society, Boston, November, 1967. Spellacy, William N.: Carbohydrate Metabolism and the Pill, presented to the New England Obstetrical and Gynecological Society, Boston, November, 1967. Kane, ‘F. J., Daiy, R. J., Ewing, J., and Keel&, M.: Brit. J-Psych&. 113: 265, -1967. Bovd. D. A.. Tr.: AM. T. OBST. & GYNEC. 43: 143; 43: 335, 1942.” Jannson, B.: Acta psychiat. scandinav. 39: 10, 1963. (Suppl. 172.) Jarrahi, A., Kane, F. J., Van de Castle, R., and Ewing, J.: Emotional and Cognitive Disturbance During the Childbearing Period, presented at the APA Annual Meeting, Detroit, May, 1967.
22. 23.
24. 25. 26. 27.
R.
S., Jr.:
J.
Chron.
Dis.
13:
pany, p. la. Tetlow, C.: J. Ment. SC. 101: 629, 1955. Dalton, K.: The Premenstrual Syndrome, Springfield, Illinois, 1964, Charles C Thomas, Publisher. Mandell, A. J., and Mandell, M. P.: J. A. M. A. 280: 792, 1967. Glick, I. D.: Psychopharmacologia 10: 363, 1967. Bigelow, N.: American Handbook of Psychiatry, New York, 1960, Basic Books, p. 540. Rome, H. P., and Braceland, F. J.: Am. J. Psvchiat. 108: 641. 1952. Glaser, G. H.: Psychosom. Med. 15: 280, 1953. Wearing, M. P.: Canadian M. A. J. 89: 239, 1963. Moos, R.: Psychologic Aspects of Oral Contraceptives, presented to the American Psychosomatic Society, April, 1967.
28.
Kane, F., Daly, R., Wallach, M., and Keeler, M.: Dis. Nerv. System 27: 339, 1966.
29.
Bakker, Cl. B., and Dightman, C. R.: Obst. & Gvnec. 28: 373. 1966. Murawski, B. J., et al.: Fertil. & Steril. 19: 50, 1968. Sturgis, S. H.: M. Aspects Hum. Sexuality 2: 4, 1968. Bakke, J. L.: Pacific Med. & Surg. 73: 200, 1965. Scott, J., and Brass, P.: AM. J. OBST. & GYNEC. 95: 1166, 1966. Keeler, M. H., Daly, R., and Kane, F. J.: Am. J. Psvchiat. 120: 1123. 1964. Daly,-R. J., Kane, F. J., and Ewing, J. A.: Lancet 2: 444. 1967. Idestrijm, Cl. M.: Lancet 1: 718, 1966. Marcotte, D., Kane, F. J., Van de Castle, R., and Lipton, M.: Unpublished data. Coppen, A.: Personal communication.
30. 31. 32. 33. 34. 35.
it 38.
Volume Number
39. 40. 41. 42.
43. 44. 45. 46.
47. 48. 49. 50. 51. 52.
102 7
Psychiatric
Bunney, W. E., and Davis, John M.: Arch. Gen. Psychiat. 13: 483, 1965. Myerson, B.: Psychopharmacologia 6: 2 10, 1964. Myerson, B.: Arch. internat. Pharmacodyn. et therap. 150: 4, 1964. Keug-Sup, Cha, Woo-Choo, Lee, Rudizk, A., and Miller, J. -W.: J. Pharmacol. & Exper. Theran. 148: 9. 1965. Schildkraut, J.‘J., and Kety, S. S.: Science 156: 21, 1967. Anton-Tay, F., and Wurtman, R. J.: Science 159: 1245, 1968. Relkin, R.: New England J. Med. 274: 944, 1966. Wurtman, R. J.: Isolation and Synthesis of Melatonin, presented at the Symposium on Indole Alkylamines, New York, New York, May, 1967. Paulson, E., and Robson, J. M.: J. Endocrinol. 30: 205, 1964. Fullerton, A., and Morrison, J. F.: J. Endocrinol. 33: 75, 1965. Plager, J. E., Schmidt, K. G., and Staubitz, W. J.: J: Clin. Invest. 43: 1066, 1964. Wvnn. V.. and Doar. 1. W. H.: Lancet 2: 715, i966: Zinneman, H. H., Seals, U. S., and Doe, R. P.: J. Clin. Endocrinol. 27: 397, 1967. Gilbert, M., Rotstein, J., Cunningham, C., I
53. 54.
55. 56. 57. 58. 59. 60. 61. 62. 63. 64.
”
65. 66.
reactions
to oral
contraceptives
1063
Estrin, I., Davidson, A., and Pincus, G.: J. A. M. A. 190: 235, 1964. Brodie, B. B.: Proc. Roy. Sot. Med. 58: 946, 1965. Coppen, A.: Depressed States and Indole Alkylamines, presented at the Symposium on Indole Alkylamines, New York, New York, May, 1967. Oswald, I.: Psychosom. Res. 9: 111, 1965. Pare, C. M. B.: Lancet 2: 527, 1963. Price, J. M., et al.: Am. J. Clin. Nutrition 20: 452, 1967. Gibbs, F. A., and Reid, D. E.: AM. J. OBST. & GYNEC. 44: 672, 1942. Selye, H. J.: J. Pharmacol. & Exper. Therap. 73: 127, 1941. Merryman, W.: J. Clin. Endocrinol. 14: 1957, 1954. Morrell, F.: Neurology 9: 352, 1959. Benedek, T., and Rubenstein, G. B.: Psychosom. Med. 1: 245. 1939. Michael, R.: Proc. Roy. Sot. Med. 58: 595, 1965. Udry, J. R., and Morris, N. M.: Distribution of Coitus in the Menstrual Cycle, presented at the Meetinas of the Ponulation Association of America, Cincinnati, dhio, April, 1967. Zell, J., and Crisp, W. E.: Obst. & Gynec. 23: 657, 1964. Ellison, R. M.: M. J. Australia 2: 625, 1964.