Idiopathic precocious puberty in girls: Psychosexual development

Journal of Youth and Adolescence, VoL 14, No. 4, 1985

Idiopathic Precocious Puberty in Girls: Psychosexual Development H e i n o F. L. Meyer-Bahlburg, 1''° A n k e A . Ehrhardt, 2 Jennifer J. Bell, 3 Susan F. Cohen, 4 Jane M. Healey, 5 Judith F. F e l d m a n , e Akira Morishima, 7 Susan W . Baker, e and Maria I. N e w 9

A promising model syndrome for the exam&ation o f the role of physical maturation in the development of female sexuality is idiopathic precocious puberty (IPP). In this first controlled study o f psychosexual development in IPP females, 16 females between 13 and 20 years o f age with a history of lPP were compared to 16 control subjects with a history of normal puberty pair-matched to the index subjects on the basis o f sex, race, age, socioeconomic level, and menarcheal status. The psychosexual history and the current psychosexual status were assessed by a systematic half-structured interview. The IPP females on average passed the psychosexual milestones at an earlier age than their normal maturing peers, with a particularly early

This work was supported by NIMH Center Grants MH 30906-01A and MH 30906-03 and by a grant from the Spencer Foundation. 'New York State Psychiatric Institute and Department of Psychiatry, Columbia University, College of Physicians and Surgeons. 2New York State Psychiatric Institute and Department of Psychiatry, Columbia University, College of Physicians and Surgeons. 3Department of Pediatrics, Columbia University, College of Physicians and Surgeons. 'New York State Psychiatric Institute and Department of Psychiatry, Columbia University, College of Physicians and Surgeons. SNew York State Psychiatric Institute. 8New York State Psychiatric Institute and Department of Psychiatry, Columbia University, College of Physicians, and Surgeons. 'Department of Pediatrics, Columbia University, College of Physicians and Surgeons. "Department of Pediatrics, Cornell University Medical College. gDepartment of Pediatrics, Cornell University Medical College. '°Correspondence should be addressed to H. F. L. Meyer-Bahlburg, Department of Psychiatry, Columbia University, 722 West 168 Street, New York, New York 10032. 339 0047-2891185/0800.0339504.50/0 © 1985 Plenum Publishing Corporation

340

Meyer-Bahlburg, Ehrhardt, Bell, Cohen, el al.

onset o f masturbation. Those who were sociosexually active tended to report a higher total orgasmic outlet and a higher sex drive. There was no increase in homosexuality among I P P girls. The timing o f puberty has a (modesO influence on psychosexual development in females.

INTRODUCTION The role of puberty in the development of female sexuality is poorly understood. Puberty denotes the hormone-based development of secondary sex characteristics and reproductive capacity. The endocrine and neuroendocrine mechanisms underlying this phenomenon are being unravelled by joint efforts of investigators of animal and human development. That hormone treatment can prematurely induce the sexual behavior characteristic of the pubertal animal is well known (e.g., Beach, 1942). Yet animal research has provided little guidance for human behavioral studies of spontaneous puberty, especially for females. One difficulty stems from the fact that female sexual behavior in subprimate mammals changes with the estrus cycle (Feder, 1984), but numerous attempts to establish analogous menstrual cycle variations in human sexual behavior have led to little consistency in results (Sanders and Bancroft, 1982). By and large, it seems that most women experience only modest variations, if any, of sex drive and sexual expression during the course of the menstrual cycle. And the existing data on primates are of little help: The correlation of menstrual cycle with sexual behavior shows an enormous variability across the primate species (Feder, 1984), which precludes a simple extrapolation to the human situation. A second difficulty is that even in lower mammals, correlative studies of endocrine and behavioral development in puberty are rare. Moreover, sex-behavioral development in relation to spontaneous puberty has not been thoroughly studied in female primates except for the recent study by Roy et al. (1984) on rhesus monkeys. Their data show that the frequency of spontaneous sexual solicitation behaviors ("presentations") of maturing females was significantly correlated with the pubertal status of their male peers in the group, but not measurably influenced by the females' own pubertal ovarian activity. Concerning hormone effects on female sexuality in humans, hormone replacement studies in oophorectomized adult women suggest a positive influence of sex hormones on sexual motivation and behavior (Dennerstein and Burrows, 1982; Sherwin, 1983). From these findings, one can derive a plausible working hypothesis regarding the role of puberty. To the extent that the hormonal and physical changes of puberty are causally involved in human psychosexual development, variations of the timing of puberty should correlate with differences in the timing of psychosexual milestones. That this

Precocious Puberty

341

is so for males is suggested by the available empirical evidence (MeyerBahlburg, 1980). For females, the (mostly recall-based) evidence is more tentative. Concerning normal variations of pubertal timing, the data collected by Kinsey et a/.(1953) on 5940 females showed only a slight tendency for early maturing girls to start their sexual activity (including masturbation, petting, and premarital coitus) earlier than their peers. An area sample survey in low-income neighborhoods of 16 United States cities, covering 3149 women, showed modest correlations of age at menarche with age at first intercourse, of age at first intercourse with age of first pregnancy, and of age at menarche with age at first pregnancy (Udry, 1979). Age at first pregnancy also was found to correlate with age at menarche in 2062 women of the Menstrual and Reproductive Health Study (Sandier, et al., 1984). Presser (1978), in a study of 408 New York City women, who had singleton first births, found similar but much weaker relationships. Age at menarche correlated with age at first date, but only for Blacks, while age at first date was related to age at first sexual intercourse for Blacks and Whites. On the other hand, Dornbusch et al. (1981), analyzing data from the U.S. National Health Examination Survey of 12-17-year-old youths, found that dating experience was much more closely linked to chronological age within either sex and that individual levels of sexual maturation added very little to the explained variance in dating after age had been taken into account. More extreme variations of pubertal timing can be found in clinical syndromes such as precocious puberty. The examination of such syndromes permits a quasi-experimental separation in time of the hormonal from some of the institutionalized social changes that coincide with normal puberty, especially the transition from elementary to junior high school. The basic endocrinology of idiopathic precocious puberty (IPP, or precocious puberty without a known cause) does not seem to differ markedly from that of normal puberty. IPP children show a hormonal pattern similar to that of normal children during puberty (Radfar et al., 1976), including the pubertal pattern of sleep-associated LH (luteinizing hormone) release (Boyar et al., 1973), and, in girls, an enhanced (in some cases even exaggerated) gonaddotropin response to LH-RF (releasing factor) administration (Reiter et ak, 1975). Thus, the IPP syndrome is a heuristically useful model for the psychoendocrine study of puberty, even though there may be some other hormonal differences between precocious, early, and late maturers (e.g., Bidlingmaier et al., 1977; Vihko and Apter, 1984). Very few data are available on the sexuality of girls with precocious puberty. Reuben and Manning (1922, 1923) cited 83 cases of pregnancy below age 15 years, with delivery occurring in 16 cases between ages 6 and 10 years, and claimed that the majority of the total sample had a history of precocious puberty. Additional cases of pregnancy in girls with precocious

342

Meyer-Bahlburg, Ehrhardt, Bell, Cohen, el al.

puberty were listed by Seckel (1946). The youngest mother recorded to date (Escomel, 1939) was a Peruvian girl who conceived at 5 years and 2 weeks and gave birth by Caesarean section at 5 years and 8 months. While such cases demonstrate a definite risk of early pregnancy in girls with precocious puberty, it is unclear whether the pregnancies recorded reflect a spontaneously increased sexual interest (or subjectively experienced "drive") of precociously maturing girls or are mainly the result of sexual seduction or abuse by sexually mature males. In the few detailed case reports published more recently that include any indication of sexual behavior, most researchers found no instances of sexual advance, flirtatiousness, or attempts at close physical erotic contact (Connor and McGeorge, 1965; Hampson and Money, 1955; Solyom et al., 1982), although some girls were occasionally seen to exhibit their breasts or genital area to their peers (e.g., Levitzky and Domash, 1978). Two clinical studies on groups of girls with precocions puberty (Thamdrup, 1961; Money and Walker, 1971) led to the conclusion that most of the sexual behavior found does not differ from the normal sexual activity of children of the same age, which confirms earlier reports (Kinsey et al., 1953, p. 746). All available information combined indicates that girls with precocious puberty are at risk of very early pregnancy if inseminated. However, the majority of published cases do not show clear indications of spontaneous premature sexual interest and activity. The lack of systematic assessment in most of these reports and the absence of any controlled study do not permit any more stringent conclusions concerning the psychosexual development in precocious puberty. The purpose of our paper is to present the findings of the first controlled sexological study on females with idiopathic precocious puberty. In line with the existing experimental animal literature, the survey data on women, and the pregnancy data for girls with precocious puberty, we hypothesized that (1) girls with precocious puberty on average pass their psychosexual milestones earlier than normal controls; (2) when compared in their teens, girls with a history of precocious puberty show higher rates of current sexual activity. An additional hypothesis concerns whether patients with precocious puberty show an increased incidence of homosexual orientation. This question is of interest because of Storms's (1981) theory explaining the development of homosexuality in terms of pubertal timing. According to Storms, homosexuals seem to attain their "sex drive development" earlier than heterosexuals, at a time when they are still mainly affiliated with peers of the same sex. Thus, while their sexual motivation increases presuamably in association with the endocrine changes of an early puberty, they are more likely to develop sexual arousal in response to the same-sex peer group. If Storms's theory is valid, children with precocious puberty should be prone to develop a homosexual orientation.

Precocious Puberty

343 METHOD Sample Selection

The current study is part of a comprehensive research project on precocious puberty (for the overall design of the project and details of the sample selection, see Ehrhardt et al., 1984). Targeted were female subjects of mid-adolescent to late adolescent age (13-20 years) with a history of idiopathic precocious puberty. That is, signs of sexual maturation- usually pubic hair and/or breast development-had appeared before 9 years, had progressed beyond premature adrenarche and premature thelarche, and could not be explained by organic pathology. Exclusion criteria were major congenital anomalies, severe chronic diseases, and neurological disorders. All IPP subjects were English-speaking. The sample included in the study was drawn from the Columbia-Presbyterian Medical Center and New York Hospit a l - Cornell Medical Center. Sixteen female adolescents with a documented history of IPP participated in this study. For the recruitment of an appropriate control g r o u p - pair-matched to the patients by sex, race, age, socioeconomic level, and menarcheal s t a t u s two sources were used. One was the Collaborative Perinatal Project (CPP; Niswander and Gordon, 1972), which had a branch at Columbia University in New York City. The second source of controls, for the younger age group, consisted of three junior high and high schools. One additional control, the match for the youngest IPP subject of this report, was indentified in her local community by a staff member. Assessment Methods

All subjects participated in a two-day evaluation protocol. Psychosexual development was assessed by means of a detailed systematic halfstructured interview with preformulated rating scales. All interviews were audiotaped to permit supervision of interviewers and clarification of rating problems, if necessary. In addition, the interviewer and the subject both rated the subject's apparent age and physical attractiveness to the opposite sex. The same staff was involved in recruitment of subjects and interview assessment, so that it was impossible to keep the procedure blind. However, the ratings required involved mainly statements concerning the age at or frequency of certain sexual activities, and the inter-rater reliability of these scales was found to be exceedingly high (Meyer-Bahlburg and Ehrhardt, 1983). Given these facts, and because of the careful interviewer training and monitor-

344

Meyer-Bahlburg, Ehrhardt, Bell, Cohen, et ai.

ing involved, effects of potential interviewer bias on the results appear unlikely.

Data Analysis Case-control comparisons were performed with standard statistical procedures appropriate to the characteristics of the dependent variables. Comparisons of current characteristics and behavior were done by matched-pair procedures, and comparisons of past milestones by independent groups methods. Because the rare I P P syndrome represents the extreme variation of timing o f endocrinologically normal puberty, such a sample is valuable in spite of its small size. In order not to overlook potentially replicable differences in this small sample, we report results as "significant" that meet the conventional 0.05 level and as "trends" at p < 0.10.

RESULTS

Sample Characteristics Details of the sample characteristics are reported in Ehrhardt et al. (1984). Twelve I P P subjects and their 12 controls were Black, 2 subjectcontrol pairs were White, one of Hispanic, and one of mixed Hispanic/Black background; all grew up in this country and were English-speaking. The mean age at the time o f the study was 17 1/2 years in both the subject and control groups, with an age range o f 13 years 2 months to 20 years 10 months for the I P P females and t3 years 3 months to 20 years 7 months for the controis. The case-control correlation for age was r = 0.98. Parental socioeconomic status (SES) was assessed by the Four Factor Index of Social Status (Hollingshead, 1975), which is based on occupation and income of both parents and ranges f r o m 8 to 66. The subjects were distributed over the entire SES range, with a case-control correlation of r = 0.61 and almost identical means o f 33.2 for the patients and 36.5 for the controls. Excluding one discrepant pair (IPP subject's SES = 55, control subject's SES = 19), the casecontrol correlation was r = 0.83. With regard to IQ, the means of both groups were in the average category. All adolescents were either in junior high, high school, or college except for 2 I P P subjects and 1 control subject who had left school. These 3 were working and had regular jobs. According to the medical records, onset of first pubertal signs as observed by the I P P girls' mothers ranged f r o m 2 1/2 years to 8 years, with a median around age 6 1/2. Median age at menarche was about 10 years

Precocious Puberty

345

Table !. Physical Characteristics

Pubic hair, Tanner (5 +6):(