puberty onset of gender differences in rates of...
TRANSCRIPT
Journal ofAffectiL)e Disorders, 29 (1993) 145-1.58
0 1993 Elsevier Science Publishers B.V. All rights reserved 0165.0327/93/$06.00
145
JAD 01045
Puberty onset of gender differences in rates of depression: a developmental, epidemiologic and neuroendocrine perspective
Adrian Angold a and Carol W. Worthman b a Decelopmental Epidemiology Program, Duke llnilrersity Medical Center, Durham, NC, 27710.3454, USA and
’ Department of Anthropology, Emory University, Atlanta, GA 30322, USA
Summary
A dramatic feature of the epidemiology of depression is the appearance of a 2: 1 female excess of depression during adolescence. In childhood, rates of depression either do not differ between boys and girls or show a slight excess in boys. In this paper we review a number of lines of evidence that implicate the physical and hormonal developments of puberty in this change. We also argue that the analysis of pubertal change in the etiology of depression must take into account the fact that the causation of depression is almost certainly a complex process. In understanding such a process we suggest that developmental epidemiological studies will be of particular value.
Key words: Depression; Adolescents; Puberty; Sex hormones
Introduction
One of the best replicated epidemiological findings in the psychiatric literature is that the prevalence of unipolar depression in women is about twice that found in men (Weissman and Klerman, 1977). However, studies of children have found either no gender differences in rates or a slight excess in boys. However by the age of 15 the female predominance is established (Guyer et al., 1989; Velez et al., 1989; McGee et al., 1990; Rutter et al., 1976; Anderson et al., 1987; Bird et
Corresponding authors: a Tel.: 919-687-4686; Fax: 919-687- 4737; b Tel.: 404-727-4489; Fax: 404-727-2860.
al., 1988; Cohen and Brook, 1987; Kashani et al., 1987; Velez et al., 1989; McGee et al., 1990; McGee and Williams, 1988; Costello, unpub- lished data; Nolen-Hoeksema et al., 1991; Angold and Rutter, 1992; Costello and DeRosier, 1993). It is as yet uncertain whether the excess of de- pressed teenage girls is due solely to an increase in the rates of depression in girls or is also
augmented by a decline in the rate of depression in teenage boys (Angold et al., 1993). A majority of studies indicate that gender-related differences are apparent by the age of 12 (and perhaps 2 or 3 years before that). There is also evidence that the rate of depression continues to rise in girls well beyond this age (see Angold and Rutter, 1992; Angold et al., 1993, for reviews). The change in
146
rates of depression therefore occurs at the same time as the hormonal and physical changes of puberty, and the many psychological, behavioral and social transitions that accompany them. This coincidence of timing has, not unnaturally, led to the suggestion that the physiological changes of puberty may be responsible in some way for the change in rates of depression. The purpose of this paper is to bring together several lines of research from child and adult psychopathology and endocrinology and to focus them on a re- search agenda for understanding the relationship between depression and the physiological changes of puberty, and to indicate the weaknesses in the literature to date.
Age changes in patterns of depressive symptoma- tology
Although sex-differentiated age trends in the prevalence of depression are clearly present, the patterns of symptomatology that contribute to those trends are less well explored. Unresolved issues include whether patterns of symptomatol- ogy shift over the course of puberty, and whether gender differences in those patterns first emerge at this time. Consideration of relevant findings should be prefaced by the recognition that, if the
base rates of individual symptoms in the popula- tion change both with age and in relation to each other, then their relative weights in contributing to the diagnosis of depression may also change. As yet we are very short of data on such changes in relation to other symptoms of depression (Mascie-Taylor, 1991), but there is evidence that they occur and that such developmental progres- sions may need to be incorporated into current nosologies (Digdon and Gotlib, 1985). For in- stance, McConville et al. (1973) have described three forms of depression characterized by ‘af- fectual symptoms’, guilt, and low self-esteem re- spectively. The ‘affectual type’ was more common in 6 to 8 year-olds, while the low self-esteem type appeared with increasing frequency at later ages. The guilt type, which resembled adult psychotic depression, mostly emerged after the age of 11. Inamdar et al. (1979) noted the absence of motor agitation or retardation, delusions of guilt, and hopelessness in their sample of 30 depressed ado-
lescents; though these symptoms have been re- ported in adolescents with depression (Strober et al., 1981; Chambers et al., 1982; Kazdin and Petti, 1982; Friedman et al., 1983a; Friedman et al., 1983b), they may be rarer in childhood and early adolescence than in adulthood.
Ryan et al. (1987) compared the symptomatol- ogy of 95 pre-pubertal children and 92 adoles- cents with Major Depressive Disorder, as as- sessed by the K-SADS interview (Chambers et al., 1985). The prepubertal children had more somatic complaints, psychomotor agitation, sepa- ration anxiety, phobias and hallucinations, and presented a more depressed appearance, while the adolescents suffered from more anhedonia, hopelessness, hypersomnia, weight change, use of illicit drugs, and lethality of suicide attempts (though not more severe suicidal ideation or in- tent). However, these represented only a minority of the symptoms evaluated, and they concluded that the similarities between the depressions of the younger and older groups were more striking
than the differences. In contrast to the findings of Inamdar et al. (1979), psychomotor agitation and retardation and hopelessness were very common
in both the children and adolescents. Mitchell et al. (1988) compared the symptoms
of 45 depressed children with those of 50 adoles- cents, interviewed with the K-SADS. They found only one significant difference in symptom rates; hypersomnia was commoner in the adolescents. But comparison with the rates of symptoms in young adult depressives reported earlier by Baker et al. (1971) revealed more extensive differences. The combined child and adolescent group had higher rates of guilt, somatic complaints, low self-esteem, suicidal attempts, and hallucinations, while the adults had higher rates of early morning wakening, anorexia, weight loss, and delusions. However, there are a number of problems with this latter comparison. Mitchell et al’s young patients were mostly outpatients, while the Baker et al. study looked at inpatient adults detected in 1971, and used different assessment instruments.
Weissman et al. (1987) found that older chil- dren with Major Depressive Disorder (MDD) more often reported weight loss and insomnia than younger depressed children. Angold et al. (1987) followed up this finding in a larger subset
147
of children from the same study of children at low and high familial risk of depression who met the Part A DSM-III (1980) criterion for Major Depressive Disorder. Symptom data were col- lected with the K-SADS-E. These analyses were therefore based on the reports of the 89 children who met this criterion, and the 62 parents who said that their children met this criterion. Episodes of depressed mood or anhedonia occur- ring in older girls were marked by considerably more sleep and appetite disturbance, and an in- crease in cognitive symptoms, concentration prob- lems, and suicidality than those occurring in younger girls and boys of all ages. There was no change in the rates of symptoms defining melan- cholia. No such age effects appeared in the re-
ports of either the boys or the parent reports about boys or girls.
A number of problems arise in trying to com-
pare these sometimes divergent findings. All of Ryan et al.‘s subjects had been referred to a specialist child and adolescent depression clinic,
and were thus highly selected for the severity of their depressions. Subjects in the Angold et al. study were selected on the basis of their parent’s
psychiatric status, rather than on the basis of their own symptomatology. The availability of data about the full range of depressive symptoms de- pended only upon their reporting that they had an episode of depressed mood or anhendonia at some time. Ryan and his colleagues also pre- sented only combined analyses for boys and girls,
and this may have obscured age trends which Angold et al. found to be confined to the girls. Though both groups used versions of the K-SADS
for their psychiatric assessments, Ryan et al. and Mitchell et al. had the same interviewer interview both parents and children, and then reconcile disagreements in ratings with the interviewees, following which summary ratings were made. This procedure might also have obscured age effects that were limited to self-report data, since An- gold et al. found no age effects in parent reports.
Despite these problems, there is consistency in the finding that sleep and appetite disturbances are more common in adolescent than childhood depressives. But we do not know whether these changes are specific to depression or reflect more general changes in sleep and appetite that are
characteristic of adolescence itself. Further evi- dence of the need to consider developmental progressions in symptomatology is found in sleep studies of depressed children, which suggest that sleep disturbance only emerges in relation to depression in later adolescence (Puig-Antich, 1986) and from epidemiologic studies which show dramatic increases in rates of suicidal behavior in adolescence (Shaffer, 1982). It is also noteworthy that boys and girls show different patterns of change in suicidality, with boys having higher rates of completed suicide, while girls have much higher rates of non-fatal self-harm.
In summary, the current literature on patterns of depressive symptomatology indicates that there are changes in both patterning of depressive
symptoms and rates of depression between child- hood and adolescence, and that these pattern and rate changes are sexually dimorphic.
An overview of the physiology of puberty
Fig. 1 presents a visual overview of the mor- phological and hormonal changes of puberty (based in part on Lee, 1980, and references cited
in Worthman, 1992). Although the hypothalamo-gonadal-adrenal
axes show a brief burst of activity in the first
months after birth (reviewed in Skuse, 19841, circulating concentrations of gonadotrophins, and gonadal and adrenal steroids are very low in early to mid-childhood. Increases in adrenal androgen
output (adrenarche) occur at around ages 6-8 years (Sizonenko and Paunier, 1975; Ducharme et al., 1976; DePeretti and Forest, 1976; Reiter et
al., 1977; Parker et al., 1978; Lashansky et al., 1991). Adrenarche precedes the earliest changes of puberty on the hypothalamo-pituitary-gonadal
axis by about 2 years and was initially thought to act as a trigger for its onset (Collu and Ducharme, 19751, but clinical and other studies have subse-
quently shown this to be untrue (Korth-Schutz et al., 1976; Wierman et al., 1986; Counts et al., 1987). The precise developmental role of adrenarche remains unclear, but it is still sus- pected to play a facilitative role in initiation of puberty.
Onset of puberty has long been thought to be characterized by the appearance and progressive
148
amplification of sleep-entrained pulsatile release hormone (LH). Current evidence indicates that of gonadotropins (Kulin et al., 1976; Judd et al., luteinizing hormone releasing hormone (LHRH) 1977; Beck and Wuttke, 1980). But recent ad- pulse amplitude and frequency increase across vances in immunometric methods have led to the the late prepubertal to early and mid stages of recognition that a circadian pattern of pulsatile puberty (Hale et al., 1988; Landy et al., 1992), gonadotropin release of low frequency and ampli- and show progressive diminution of diurnal varia- tude appears in late childhood (Dunkel et al., tion in early to mid puberty. Pulse frequency 1990; Wu et al., 1990), and that the transition to decreases in late puberty (although amplitude puberty is a gradual rather than acute event me- continues to increase in males) as LHRH release diated through physiological mechanisms that are becomes more sensitive to negative feedback con- probably operative at ages 8-12 years, or l-3 trol from gonadal steroids (Wennink et al., 1989; years prior to the onset of morphological puberty Dunger et al., 1991; Marshall et al., 1991). An (Wennink et al., 1989; Wennink et al., 1990; Wu important sex difference established over the et al., 1990). Endocrinologic onset of puberty is course of puberty is that females develop pul- now seen as involving increased gonadotropin satile Gonadotrophin Releasing Hormone pulse frequency as well as amplitude, and is (GnRH) release along with fluctuating estradiol marked by the appearance of frequent, closely and progesterone (Marshall et al., 1991). Matura- sleep-entrained nighttime pulses of luteinizing tion of this pattern extends from before menar-
Age 6- 10 11 12 13 14 15 16 years
MALES A PHV
growth
physical strength
sex characters
genital stage I 2 . . . . . . . 3 . . . . . . 14...5 . . . . . . . . . .
penis growth
pubic hair stage
axillarv hair
first emission
other
endocrine change
LH. FSH
testosterone
._... gynaecomast~ Xchange &r
adrenal androgens < adrenarche > 1 mm1111111 linear gain
FEMALES
growth
A PHV
-
fat in 8
sex characters
breast stage
pubic hair stage
axillary hair
menarche
endocrine change
LH, FSH
estradiol
adrenal androgens
night bulse I II II! IIll 14 daytime levels increase
I.,,. a ~~~BI.B~.I~~I~.~B~ gradualgainkyc~ic
<adrenarche>~- I I I I I I I I I I I linear gain
Fig. 1.
149
the to several months or years beyond in the course of establishing regular ovulation and luteal function (Vihko and Apter, 1980; Wennink et al., 1990).
These neuroendocrine changes are reflected in morphological signs of puberty such as growth and the development of secondary sex character- istics, such as those rated by Tanner’s pubertal stages (see Fig. 1). These morphological changes
have consistently been shown to reflect, albeit grossly, underlying endocrine change (Burr et al., 1970; Sizonenko et al., 1970; Lee and Migeon, 1975; Sizonenko and Paunier, 1975; Lee et al., 1976; Apter, 1980). Indeed, both chronological age and pubertal stage correlate with endocrine parameters, but hormone levels explain on aver- age less than half the variance in morphological pubertal development and growth in girls, but as much as 80% in boys (Nottelmann et al., 1987a).
The reasons for this difference in hormone-mor- phology correlations are unknown. Thus morpho- logical pubertal stage is an imprecise index of
individual endocrine status, especially in females.
Implications of the physiology of puberty for psychopathological research
Consideration of the biology of puberty in both sexes suggests several points that need to be incorporated into a developmental account of the relationship between puberty and depression:
1. Some hormone change is cryptic; that is, adrenarche in both sexes and the earliest en- docrine changes of puberty (especially in boys) are not reflected in visible morphologic change. The possible psychological effects of these early endocrine changes have been little studied, both because studies of adolescents generally do not
commence until ages 9 or 10 or even later, and because much of the research has been focused on the visible morphological events. However, intensive research that examines both visible markers and endocrine status has fairly consis- tently demonstrated that it is the endocrine, rather than the visible pubertal changes, that best predict negative/ depressive affect at puberty (Paikoff et al., 1991; Susman et al., 1991).
2. The neuroendocrine process that initiates puberty occurs in the age window over which
rates of depression probably begin to rise in girls (and perhaps fall in boys), that is age 9-11. This underscores the need to study endocrine change in late childhood and to use earliest elevation of gonadotropins and gonadal steroids along with morphological progression into puberty as mark- ers of developmental maturation.
3. The pattern of relationship between en- docrine and morphologic change is different in
males and females: most growth and sex charac- ter development occurs in girls in the first half of puberty and before menarche, whereas extensive endocrine change occurs in boys before the exter- nal signs of androgenization appear (see Fig. 1). Thus, rapid physical transformations co-occur with the period of early, most rapid endocrine
change in girls, but are delayed and temporally removed in boys. Thus it is often stated that puberty occurs about two years earlier in girls than in boys in Western countries, but while this is true of a marker such as peak height velocity, it is not true of the relative timing of menarche in girls and first emission in boys, or of the onset of sleep-entrained pulsatile gonadotrophin secre- tion, which occurs only a little later in boys than
in girls (see Fig. 1). 4. ‘Pubertal status’ is not adequately described
by a single parameter relating to degree of pro- gression on a single developmental path. Even Tanner’s well-known stages of morphological pu- berty require separate ratings on a number of dimensions. Brooks-Gunn and Warren (1985) pointed to five parameters of pubertal change that could be relevant to psychopathological re- search:
(a) Maturational status: Where in development is each individual, and how does being at this point (as opposed to other points) impact on psychopathology?
(b) Maturational timing: Where in relation to others is each individual, and what impact does this relative status have on psy- chopathology?
(c) Maturational rate: How rapidly are the changes of puberty proceeding? The time taken to progress from Tanner Stage 1 to Stage 5 may be anything from a few months to several years.
(d) Maturational synchrony: To what degree are
the hormonal, physical, psychological and so- cial processes of puberty synchronized in an individual?
(e) Dff t’ 1 2 eren la salience of maturational erlents: From the point of view of an adolescent’s social and psychological understanding, not all events of puberty may be equally salient. The development of breasts or menarche would seem to be highly personally salient, but the same cannot be said of the bone changes that are proceeding concomitantly with more obvious effects of puberty. Con- centration on single salient events is not suf- ficient. The substantial literature on the im- pact of menarche has predictably led to no very satisfactory conclusions (Greif and Ul- man, 1982; Moffitt et al., 19921, since to divorce one aspect of a complex process from its developmental matrix is often a recipe for confusion. Two further parameters of pu- berty are also worthy of consideration, though the first can be seen as a qualitative compo- nent of the differential salience of matura- tional events.
(f) The meaning of maturational events at the personal lerlel. The onset of puberty seems likely to mean something very different to the shy, thin 12-year old who experiences greasy, spotty skin as a major manifestation of rising testosterone levels, and the outgoing 12-year old athlete who puts on muscle instead. It is also reasonable to suppose that such manifes- tations of hormonal activity will have rather different effects on their relationships with girls. Simmons (1992) found that pretty girls were more likely to regard the change in body shape that accompanies puberty nega- tively, and suggested that this was because
puberty for them meant the loss of a valued asset, while for less attractive girls it offered new hope for the future. Given that negative self-evaluation is part of the symptomatology of depression it would be worthwhile to ad- dress the personal meaning of the events of puberty more directly in future research.
(g) Cultural and social effects on puberty. Puberty is often treated as though it were independ- ent of cultural or psychological factors, but strong evidence suggests otherwise. Social ar-
rangements and practices largely determine environmental quality (nutrition, family size and spacing, child care> which, in turn, has been repeatedly found to affect rate of growth and timing of puberty (Eveleth and Tanner, 1991; Bielicki, 1986). The secular trends to reduced age at menarche and increased height for age in childhood that have consis- tently paralleled improved health care, indus- trialization and regularization of diet around the world vividly illustrate the influence of social arrangements on maturation patterns. Environmental effects operate, moreover, on the intrafamilial level: for instance, sibling height for age correlates inversely with sib order, even when the effect of socioeconomic status is controlled (Mascie-Taylor, 1991). Saugstad (1989a; 1989b) has employed Scan- dinavian and other European data to suggest that trends to earlier maturation have been associated with increased rates of manic-de- pressive disorder, and decreased rates of schizophrenia.
Similarly, the association between puberty and depression may not be one-way. Psychopathology and stress can alter the pace of growth and matu- ration. Psychogenic growth retardation is a dra- matic and severe instance (Pollitt and Leiber, 1980, but see also Skuse, 1984); anorexia nervosa presents a less uncommon example where the onset of fat gain appears to precipitate dieting, which in turn leads to developmental retardation with switching off of ovarian function (Condit, 1990). Finally, childhood stress (specifically, fam- ily conflict and father absence) has been sug- gested to influence both timing of menarche and attitudes toward interpersonal relationships (Be- lsky et al., 1991), though the empirical support for this notion is mixed (Moffitt et al., 1992). One might also consider whether depression itself or factors involved in risk for depression also influ- ence the timing and progression of puberty.
Let us now turn to the current evidence link- ing pubertal change with depression.
Depression and the parameters of puberty
Status effects of morphological pubertal stage Pubertal staging represents a convenient mea-
sure of maturational status which has been incor-
porated in numerous studies of puberty and de- pression. In the Isle of Wight study (Rutter et al., 1970) a rough observational measure of pubertal status predicted depressive feelings more strongly than did age. On the other hand, the large clini- cal study of Angold and Rutter (1992) found that controlling for age eliminated any significant ef- fect of pubertal status on depressive symptoma- tology. Rutter et al. (1989) have also reported results from a second clinical study that showed no effect of morphological pubertal stage when age was controlled for in hierarchical regression
models, but significant effects of age when puber- tal stage was controlled. Several studies using status questionnaire assessments of psychopathol- ogy have also failed to find effects of morphologi- cal pubertal stage (Susman et al., 1987a; Susman et al., 1987b; Brooks-Gunn and Warren, 1989a; Paikoff et al., 1991; Olweus et al., 1988). Overall, therefore, there is little evidence to support the notion that morphological pubertal stage per se is a significant risk factor for depression.
Relative timing of morphological puberty A substantial literature based upon observa-
tion or self-report of pubertal stage, but without direct hormonal measures, has addressed the ef- fects of the relative timing of puberty. In sum-
mary, early puberty has been associated with a positive psycho-social impact in boys, and a nega- tive impact in girls (reviewed in Stattin and Mag- nusson, 1990). But none of these studies used
measures of psychopathology suited to diagnosing depression. Perhaps most pertinently, Stattin and Magnusson (1990) found that early developing Swedish girls reported more psychosomatic and depressive symptomatology, particularly more suicidal thoughts, than did later maturing girls. Both early and late maturers (but especially the former) had more documented psychiatric service contacts by early adulthood than those whose puberty was normative. These analyses empha- sized that effects of puberty timing were medi- ated through the impact of early maturity on girls’ social lives. On the other hand, Paikoff, Brooks-Gunn and Warren (1991) found no effects of the age at which increased estradiol secretion occurred in their l-year follow-up study using the CES-D or in their original cross-sectional data
151
(Brooks-Gunn and Warren, 1989b). They did, however, find an effect of the degree of hormonal level change over time on depressive symptoms.
A provocative time effect has been reported by Petersen, Sargiani and Kennedy (1991), who found that in both boys and girls, peak growth velocity occurring prior to, or close to, a change in school was associated with the occurrence of depressive symptoms in both boys and girls. How- ever, the relatively early occurrence in puberty of this change in girls (usually during Tanner Stages I-III compared with occurrence in Tanner Stages IV and V in boys) meant that this combination of circumstances was more common in girls. Such findings further reinforce the idea that the bio- logical changes of puberty need to be examined
in the context of normative and non-normative changes in other aspects of the developing ado- lescent’s life.
Hormonal status, timing, and rate of change effects on depression
Associations of depression with levels of go- nadal hormones and gonadotropins have been reported. Abnormalities in gonadotropin secre-
tion and of gonadotropin releasing hormone have been found in depressed adult patients (see, e.g., Tolis and Stefanis, 1983) while follicle stimulating
hormone (FSH) levels have been found to corre- late positively with negative affect in normal men (Houser, 1979). Rapidly falling estrogen levels have been implicated in menstrual cycle-associ- ated mood changes (e.g., Bardwick, 1976) which have, in turn, been linked with increased suscep- tibility to major depression (reviewed in (Warner
et al., 1991). Increases in estradiol have been found to improve depressive symptoms in hypo- estrogenic women (Montgomery et al., 1987). However, findings from this considerable litera- ture are rather inconsistent and difficult to inter- pret (see Bancroft, 1991, for an excellent critical review).
The NIMH study of puberty and psycho- pathology (Susman et al., 1987a; Susman et al., 1987b; Nottelmann et al., 1987a; Nottelmann et al., 1987b) described a negative association of testosterone : estradiol ratio, sex hormone binding globulin, and androstenedione concentration with negative emotional tone, but only in boys. These
152
workers also report an association of early matu- ration (measured by estradiol and testoster- one : estradiol ratio) with less negative emotional tone in boys, but more negative emotional tone in girls. Adrenal androgen levels correlated with negative emotional tone in boys, but not girls, while the opposite was the case for FSH (Susman et al., 198.5). Overall, hormone concentrations were more strongly associated with affective be- havior in boys than in girls. In a recent summary (Nottelmann et al., 1990), the authors comment “not only that it is useful to examine more than one group of hormones in a study of hormone-be- havior relations but that it is important to exam- ine such relations in the context of other develop- mental markers such as chronological age and
pubertal stage.” However, these authors also rec- ognize that interpretation of their findings is compromised by the use of many statistical com- parisons with a small sample. Likewise, Brooks- Gunn and Warren (1989a) found that negative affect increased in lo-14-year-old girls during
rapid estrogen rise (early puberty), but that social factors and their interaction with estrogen change together explained more of the variance in nega- tive affect scores than did hormonal status alone. This work has now been extended to include a l-year follow-up of 72 girls from the original sample (Paikoff et al., 1991). A significant linear effect of estradiol level at time 1 on CES-D scores 1 year later was found. However, there was no such effect on time 2 scores from the De- pressed-withdrawal scale of the Youth Behavior profile or on mothers’ reports of their daughters’
depressive affect. Puig-Antich (1987) reported growth hormone
(GH) hypersecretion in response to insulin-in- duced hypoglycemia in prepubertal, but not ado- lescent girls, and suggested that rising estrogen levels had decoupled the GH response to ITT from the neuroregulatory mechanisms of depres- sion.
These studies all point to an effect of the biological changes of puberty on depressive symp- tomatology. Since the pattern of pubertal matura- tion differs hormonally and physically in boys and girls, these biological factors offer a potential basis for explanation of sex-differentiated trajec- tories in depression rates. But the degree to
which these trajectories relate to hormone changes per se, the associated visible changes of puberty, or other age-related psychological or so- cial changes (Brown et al., 1986; Weissman et al., 1987; Bird et al., 1988; Orvaschel et al., 1988; Goodyer and Altham, 1991) remains unclear. A theme that has begun to emerge from several of these studies is the need to consider the physio- logical changes of puberty within the social con- text of the adolescent.
While behavioral-affective effects of pubertal endocrine change have received some study, pos- sible effects of prepubertal endocrine change, or adrenarche, have not. Extensive epidemiological data on timing of adrenarche are not available; nearly all reports merely offer means and vari-
ance of circulating andrenal androgens by age, frequently by age groups. Unlike puberty, adrenarche is not characterized by hormone ele-
vations from minuscule concentrations; rather, there is an accelerated rate of increase in mid- childhood. Thus investigators have not estab- lished clear criteria for adrenarche as a develop- mental stage. Nevertheless, it does appear that adrenarche (reflected in dehydroepiandro- sterone-sulphate, androstenedione, and dehy- droepiandrosterone levels) is discernable on aver- age between ages 6 and 8, and is well established by ages 8 to 10. Although available evidence (DePeretti and Forest, 1976; Reiter et al., 1977; Parker et al., 1978) suggests that timing of adrenarche and levels of adrenal androgen out- put show considerable individual variation, that variation has not been well documented (in con- trast to intense interest in variable timing of puberty), and its relationship to psychosocial de- velopment or to behavior remains unexamined. Researchers in the field of depression and ado-
lescence have emphasized the gap in our knowl- edge of psycho-behavioral effects of adrenarche and prepuberty, and noted the need for earlier measures (at least at ages 7-9) to detect develop- mental continuities and discontinuities (Brooks- Gunn and Warren, 1992; Eccles and Buchanan,
1992).
Puberty and peri-menstrual mood disturbance In contrast to extensive research on adult
women (see Endicott, this volume), little atten-
153
tion has been given to the issue of perimenstrual
mood disturbance in adolescence. Epidemiologic data are scant, and information on the etiology of PMS is even more limited. In a study of older adolescents (mean age 17.6 years) attending an adolescent health service (primarily for sexual or gynecologic concerns), a significant percentage reported premenstrual changes considered to be severe (59%) or extreme (43%) (Fisher et al., 1989). ‘Major’ or ‘minor’ depressive changes were retrospectively reported by almost two thirds of respondents, impaired social function by almost one half and impulsive behavior by over one third. Nada Raja et al. (1992) used an arbitrary cutoff of 1SSD above the mean on their retro- spective Dysmenorrhoea Questionnaire, which included 16 psychological items, either before or during menstruation, to define PMS in Z-year- olds. They found that having premenstrual syn- drome (PMS) by this definition was associated with concurrently higher anxiety and inattention scores from the Diagnostic Interview Schedule for Children (DISC) (Shaffer et al., 19X9), but not with higher depression scores. However, these analyses did not distinguish between those with psychological symptoms and those with physical symptoms of PMS, and contained many fewer items than the better-known adult PMS question-
naires. Thus, further work is necessary to settle the question of whether there is an association between PMS and depression in early adoles- cence.
It is surprising that the adolescent literatures on pubertal effects on depression and the adult literature on perimenstrual mood disorders have
not led to the examination of the role of perimen- strual mood disturbance in relation to the in- crease in depression in adolescent girls. While it seems unlikely that the diagnostic epidemiologi- cal findings of increased rates of major depres- sion in adolescent girls are entirely the result of mis-diagnosis of menstrual-related affective dis- turbances, the potential contribution of such dis- turbances to changes in affective symptoms re- ports with age can hardly be ignored. Given the uncertainty associated with the findings from ret- rospective studies discussed above, prospective studies beginning prior to menarche offer an op- portunity for substantial progress in understand-
ing the relationships between depression and perimenstrual affective disorders. In particular, such studies would allow us to determine the degree to which perimenstrual affective distur- bances reflect exacerbations of pre-existing de- pressive symptomatology, de nouo phenomena that are precipitated by the onset of menstrua- tion, and/or precipitants or predictors of major depression.
Puberty and behavior problems - the need to consider comorbidity
Space does not permit a detailed review of the associations between hormonal and morphologi- cal puberty and behavioral problems (see Buchanan et al. (1992) for a very substantial
review), but it is important to recall that the psychopathological changes of adolescence are far from being confined to the affective domain
(Rutter, 1990). The high rate of comorbidity be- tween depression and behavioral disorders (see Costello and Angold, 1993; Caron and Rutter, 1991) means that some apparent hormonal ef- fects on depression may, in fact, be a byproduct of effects on non-depressive disorders that are themselves correlated with depression.
The evidence in this area, though limited, ap- pears sufficient to suggest a need to incorporate measures of behavioral symptomatology and dis- orders in studies of the relationship between de- pression and maturation at adolescence. Such an approach is necessary for identification of effects
specific to the affective domain. An illustration of the potential importance of considering comor- bidity is offered by Angold and Rutter’s finding (1992) that the substantial age effects seen with affective disorders in a large clinical population did not apply to mixed disorders of conduct and emotions. The relative timing of maturational ef- fects may also differentially affect behavioral and emotional problems (Buchanan et al., 1992). Lon- gitudinal studies with systematic measurement of a wide range of psychopathology offer the best means of shedding further light on these prob- lems.
Conclusions
Overall, evidence linking the changing rates of depression in adolescence with the biological
154
changes of puberty is inconclusive, but studies that have addressed this question are surprisingly few in number. Several weaknesses in the current literature further contribute to uncertainty in this area:
(1) None of the child and adolescent psychi- atric epidemiologic studies have employed ade- quate measures of pubertal status (and most have not engaged the issue at all).
(2) None of the epidemiologic studies has ad- dressed the question of the degree to which the apparent rise in depression in adolescent girls might be related to perimenstrual mood disor- ders.
(3) The relevant endocrine studies have been compromised by small sample sizes and limited
scale measures of psychopathology. None has ex- amined rates of depressive diagnoses or detailed ratings of psychosocial impairment secondary to
depressive symptomatology. (4) The endocrine studies have not employed
epidemiologically adequate sampling strategies (Paikoff et al., 1991). No large-scale population studies of the hormonal changes of puberty in Western adolescents are available in the English language (or elsewhere, to our knowledge). Small sample size, assay variation across laboratories, and a focus on physiologic process (intra-popula- tion patterns) rather than comparative studies across populations has led to a wealth of excel- lent clinical studies (e.g., Winter and Faiman, 1972; Winter and Faiman, 1973; Lee and Migeon,
1975; Lee et al., 1976; Apter, 1980; Nottelmann et al., 1987a), but a dearth of normative value ranges based on large sample sizes. In particular, the literature on adrenarche is severely limited, and there is growing awareness of the need to begin the examination of the relationship be- tween puberty and psychiatric disorders much earlier in life than has been done up until now.
(5) Many studies of both adults and children relating psychiatric parameters to sex hormones have measured single hormones, with results that have ‘been consistently unhelpful’ (Bancroft,
1991). On the other hand, when multiple hor- mones have been measured, the use of multiple statistical tests with relatively small samples has led to interpretative problems.
(6) Although the multidimensional biochemi-
cal and physical changes of puberty extend over several years (commencing with adrenarche), no study has covered an age-range that spans this
period, and few have included any longitudinal component. Only two studies have addressed the critical issue of the impact of change in pubertal status as a possible precipitant of depression.
(7) Too little attention has been paid to the theoretical complexity of the idea of an associa- tion between puberty and depression. Analyses are needed that simultaneously consider the di- mensions of (a) pubertal status, (b) the relative timing of puberty, (c) the rate of change of puber- tal status, (d) the synchrony of pubertal change, (e) the salience of the events of puberty, (f) the personal meaning of the events of puberty, (g) the cultural meaning of the events of puberty and (h) whether each of these dimensions is best under- stood as a vulnerability or precipitating factor.
(8) Despite reports that depression is often associated with disruptive behavior disorders and that testosterone is implicated in the genesis of adolescent aggressive behavior, particularly in males, only one study considers co-morbidity as a possible confounder of the effects of pubertal status on depression.
(9) No study of a representative population sample has employed methods adequate to the task of simultaneously identifying relationships between hormonal status, morphological status, psychiatric status (at the levels of symptoms, diag- noses and psychosocial impairment), and social factors such as family difficulties or psychosocial impairment, longitudinally.
(10) Most studies, both of the effects of age and gender on depression rates and of the rela- tionships between puberty and depression, have considered only diagnoses or scale scores as mea- sures of depression. However, the study of rates of individual symptoms is also indicated by evi- dence that age and gender do not affect all symp- toms of depression equally. For instance, Nolen- Hoecksema, Girgus and Seligman (1991) found that in 3rd-graders followed for 18 months, boys had consistently higher levels of anhedonia and behavioral disturbance, as measured by the CDI, but that boys and girls scored equally on self-de- precation, physiological symptoms and dysphoric mood items.
155
A clear implication of this review is that the methods of intensive laboratory studies of rela- tively small groups need to be applied to large representative samples, observed at multiple points in time from prepuberty to early adult- hood, if we are to understand the role of puberty in the development of depression. We recognize the difficulties and complexities involved in adopting such an approach, but cannot avoid the
fact that development is not a simple process, and that all the evidence points towards multiple determination in depression. In our present state
of knowledge, it appears that large epidemiologi- cal studies are needed if further substantial progress is to be made. On the other hand, we have, in the increase in female depression in adolescence, a well documented effect of large size that is a good candidate for such investiga- tion, and technologies available for the measure-
ment of pubertal hormones in general population studies. In sum, the relationship between puberty and sex differences in rates of depression is ripe for further investigation by developmental epi- demiological studies.
References
American Psychiatric Association (1980) Diagnostic and Sta-
tistical Manual of Mental Disorders, 3rd Edn. Washing-
ton, DC: American Psychiatric Association.
Anderson, J.C., Williams, S., McGee, R. and Silva, P.A.
(1987) DSM-III disorders in preadolescent children: preva-
lence in a large sample from the general population. Arch.
Gen. Psychiatry 44, 69-77.
Angold, A., Weissman, M.M., John, K., Merikangas, K.R., Prusoff, B.A., Wickramaratne, P., Gammon, G.D. and
Warner, V. (1987) Parent and child reports of depressive
symptoms in children at low and high risk of depression. J.
Child Psychol. Psychiatry 28, 901-915.
Angold, A., Loeber, R., Costello, E.J., Van Kammen, W. and
Stouthamer-Loeber, M. (1993) Disappearing depression in
a population sample of boys. J. Abnorm. Child Psychiatry,
submitted.
Angold, A. and Costello, E.J. (1993) Depressive comorbidity
in children and adolescents: Empirical, theoretical and
methodological issues. Am. J. Psychiatry, in press.
Angold, A. and Rutter, M. (1992) The effects of age and
pubertal status on depression in a large clinical sample.
Dev. Psychopathol. 4, 5-28.
Apter, D. (1980) Serum steroids and pituitary hormones in
female puberty: a partly longitudinal study. Clin. En- docrinol. 12, 107-120.
Baker, M., Dorzab, J., Winokur, G. and Cadoret, R.J. (1971)
Depressive disease: classification and clinical characteris-
tics. Compr. Psychiatry 12, 354-365. Bancroft, J. (1991) Reproductive hormones. In: M. Rutter and
P. Casaer (Eds.), Biological Risk Factors for Psychosocial
Disorders, Cambridge: Cambridge University Press, pp.
260-310. Bardwick, J.M. (1976) Psychological correlates of the men-
strual cycle and oral contraceptive medication. In: E.J.
Sachar (Ed.), Hormones, Behavior and Psychopathology,
pp. 95-103.
Beck, W. and Wuttke, W. (1980) Diurnal variations of plasma
luteinizing hormone, follicle-stimulating hormone, and
prolactin in boys and girls from birth to puberty. J. Clin.
Endocrinol. Metab. 50, 635-639.
Belsky, J., Steinberg, L. and Draper, P. (1991) Childhood
experience, interpersonal development, and reproductive
strategy: an evolutionary theory of socialization. Child
Dev. 62, 647-671.
Bielicki, T. (1986) Physical growth as a measure of the eco-
nomic well-being of populations: the twentieth century. In:
F. Falkner and J.M. Tanner (Eds.), Human Growth, 2nd
Edn. New York: Plenum Press, pp. 283-305.
Bird, H.R., Canino, G., Rubio-Stipec, M., Gould, MS., Rib-
era, J., Sesman, M., Woodbury, M., Huertas-Goldman, S.,
Pagan, A., Sanchez-Lacay, A. and Moscoso, M. (1988) Estimates of the prevalence of childhood maladjustment in
a community survey in Puerto Rico: the use of combined
measures. Arch. Gen. Psychiatry 45, 1120-1126.
Brooks-Gunn, J. and Warren, M.P. (1985) Measuring physical status and timing in early adolescence: a developmental
perspective. J. Youth Adolesc. 14, 163-189.
Brooks-Gunn, J. and Warren, M.P. (1989a) Biological and
social contributions to negative affect in young adolescent girls. Child Dev. 60, 40-55.
Brooks-Gunn, J. and Warren, M.P. (1989b) Biological contri-
butions to affective expressions in young adolescent girls.
Child Dev. 60, 372-385.
Brooks-Gunn, J. and Warren, M.P. (1992) Revisiting theories
of storm and stress: the role of biology. Paper presented at
the Annual Meeting of the Society for Research on Ado-
lescence. Washington, DC, March.
Brown, G.W., Harris, T. and Bifulco, A. (1986) Long-term
effect of early loss of parent. In: M. Rutter, C. Izard and
P. Read (Eds.), Depression in Childhood: Developmental
Perspectives. New York: Guilford Press, pp. 251-296.
Buchanan, C.M., Eccles, J.S. and Becker, J.B. (1992) Are
adolescents the victims of raging hormones: evidence for activational effects of hormones on moods and behavior at
adolescence. Psychol. Bull. 111, 62-107.
Burr, I.M., Sizonenko, P.C., Kaplan, S.L. and Grumbach,
M.M. (1970) Hormonal changes in puberty. I. Correlations
of serum luteinizing hormone and follicle-stimulating hor-
mone with stages of puberty, testicular size and bone age in normal boys. Pediatric Res. 4, 25-35.
Caron, C. and Rutter, M. (1991) Comorbidity in child psychol-
ogy: concepts, issues and research strategies. J. Child Psychol. Psychiatry 32, 1063-1080.
Chambers, W.J., Puig-Antich, J., Tabrizi, M.A. and Davies,
156
M. (1982) Psychotic symptoms in prepubertal major de-
pressive disorder. Arch. Gen. Psychiatry 39, 921-927.
Chambers, W.J., Puig-Antich, J., Hirsch, M., Paez, P., Am-
brosini, P.J., Tabrizi, M.A. and Davies, M. (1985) The
assessment of affective disorders in children and adoles-
cents by semistructured interview: test-retest reliability of
the schedule for affective disorders and schizophrenia for
school-age children. Present Episode Version. Arch. Gen.
Psychiatry 42, 696-702.
Cohen, P. and Brook, J. (1987) Family factors related to the
persistence of psychopathology in childhood and adoles-
cence. Psychiatry 50, 332-345.
Collu, R. and Ducharme, J.R. (1975) Role of adrenal steroids
in the regulation of gonadotropin secretion at puberty. J.
Steroid Biochem. 6, 869-872.
Condit, V. (1990) Anorexia nervosa: levels of causation. Hum. Nature 1, 391-413.
Costello, E.J. The Pittsburgh HMO follow-up study, unpub- lished.
Costello, E.J. and DeRosier, M. Continuity and change in
psychopathology from childhood to adolescence, unpub- lished.
Counts, D.R., Pescovitz, O.H., Barnes, K.M., Hench, K.D.,
Chrousos, G.P., Sherins, R.J., Comite, F., Loriaux, D.L.
and Cutler, G.B. (1987) Dissociation of adrenarche and
gonadarche in precocious puberty and in isolated hypogo-
nadotropic gypogonadism. J. Clin. Endocrinol. Metabolism
64, 1174-1178.
DePeretti, E. and Forest, M.G. (1976) Unconjugatyed dehy-
droepiandrosterone plasma levels in normal subjects from
birth to adolescence in human: the use of a sensitive
radioimmunoassay. J. Clin. Endocrinol. Metabolism 43,
962-969.
Digdon, N. and Gotlib, I.H. (1985) Developmental Considera-
tions in the Study of Childhood Depression. Dev. Rev. 5,
162-199.
Ducharme, J.R., Forest, M.G., DePeretti, E., Sempe, M.,
Collu, R. and Bertrand, J. (1976) Plasma adrenal and
gonadal sex steroids in human pubertal development. J.
Clin. Endocrinol. Metabolism 42, 458-467.
Dunger, D.B., Villa, A.K., Matthews, D.R., Edge, J.A., Jones,
J., Rothwell, C., Preece, M.A. and Robertson, W.R. (1991)
Pattern of secretion of bioactive and immunoreactive go-
nadotrophins in normal pubertal children. Clin. En-
docrinol. 35, 267-275.
Dunkel, L., Alfthan, H., Stenman, U-H. and Perheentupa, J.
(1990) Gonadal control of pulsatile secretion of luteinizing
hormone and follicle stimulating hormone in prepubertal
boys evaluated by ultrasensitive time-resolved immunoflu-
orometric assay. J. Clin. Endocrinol. Metabolism 70, 107- 114.
Eccles, J. and Buchanan, CM. (1992) Hormones and behavior at early adolescence: a theoretical overview. Paper pre-
sented at the Annual Meeting of the Society for Research on Adolescence. Washington, DC, March.
Eveleth, P. and Tanner, J.M. (1991) Worldwide Variation in Human Growth, 2nd Edn. Cambridge: Cambridge Univer-
sity Press.
Fisher, M., Trieller, K. and Napolitano, B. (1989) Premen-
strual symptoms in adolescents. J. Adolesc. Health Care 10, 369-375.
Friedman, R.C., Hurt, S.W., Clarkin, J.F. and Corn, R. (1983a)
Primary and secondary affective disorders in adolescents
and young adults. Acta Psychiatr. Stand. 67, 226-235.
Friedman, R.C., Hurt, SW., Clarkin, J.F., Corn, R. and
Aronoff, MS. (1983b) Symptoms of depression among
adolescents and young adults. J. Affect, Dis. 5, 37-43.
Goodyer, I.M. and Altham, P.M.E. (1991) Lifetime exit events
and recent social and family adversities in anxious and
depressed school-age children and adolescents ILJ. Affect. Disord. 21, 229-238.
Greif, E.B. and Ulman, K.J. (1982) The psychological impact of the menarche on early adolescent females: a review of
the literature. Child Dev. 53, 1413-1430.
Guyer, B., Lescohier, I., Gallagher, S.S., Hausman, A. and
Azzara, C.V. (1989) Intentional injuries among children
and adolescents in Massachusetts. N. Engl. J. Med. Dee 7.
1584-1589.
Hale, P.M., Khoury, S., Foster, C.M., Beitins, I.Z., Hopwood,
N.J., Marshall, J.C. and Keith, R.P. (1988) Increased
luteinizing hormone pulse frequency during sleep in early
to midpubertal boys. J. Clin. Endocrinol. Metab. 66, 785-
791.
Houser, B.B. (1979) An Investigation of the Correlation Be-
tween Hormone Levels in Males and Mood, Behavior and
Physical Discomfort. Horm. Behav. 12, 185-197.
Inamdar, S.C., Siomopoulos. G., Osborn, M. and Bianchi,
E.C. (19791 Phenomenology associated with depressed
moods in adolescents. Am. J. Psychiatry 136, 156-159. Judd, H.L., Parker, D.C. and Yen, S.S.C. (1977) Sleep-wake
patterns of LH and testosterone release in prepubertal
boys. J. Clin. Endocrinol. Metab. 44, 965-969.
Kashani, J.H., Beck, N.C., Hoeper, E.W., Fallahi, C., Corco-
ran, CM., Mcallister, J.A., Rosenberg, T.K. and Reid, J.C.
(1987) Psychiatric disorders in a community sample of
adolescents. Am. J. Psychiatry 144, 5844589.
Kazdin, A.E. and Petti, T.A. (1982) Self-report and interview
measures of childhood and adolescent depression. J. Child Psychol. Psychiatry 23, 437-457.
Korth-Schutz, S., Levine, L.S. and New, M.I. (1976) Serum
androgens in normal prepubertal and pubertal children and in children with precocious adrenarche. J. Clin. En-
docrinol. Metab. 42, 117-124.
Kulin, H.E., Moore, R.W. and Satner, S.J. (1976) Circadian
rhythms in gonadotropin excretion in prepubertal and
pubertal children. J. Clin. Endocrinol. Metab. 42, 770-773.
Landy, H., Beopple, P.A., Mansfield, M.J., Charpie, P..
Schoefeld, D.I., Link, K., Romero, G., Crawford, J.D.,
Crigler, J.F., Blizzard, R.M. and Crowley, W.F. (1092) Sleep modulation of neuroendocrine function: develop- mental changes in gonadotropin-releasing hormone secre-
tion during sexual maturation. Pediatr. Res. 213, 217. Lashansky, G., Sainger, P., Fishman, K., Gautler. T., Mayes,
D., Berg, G., diMartino-Nardi, J. and Reiter, E. (1991)
Normative data for adrenal steroidogenesis in a healthy
pediatric population: age-and sex related changes after
157
adrenocorticotropin stimulation. J. Clin. Endocrinol.
Metab. 73, 674-688. Lee, P.A., Xenakis, T., Winer, J. and Matsenbaugh, S. (1976)
Puberty in girls: correlation of serum levels of go-
nadotropins, prolactin, androgens, estrogens, and pro-
gestins with physical changes. J. Clin. Endocrinol. Metab.
43, 775-784.
Lee, P.A. (1980) Normal ages of pubertal events among
american males and females. J. Adolesc. Health Care 1,
26-29.
Lee, P.A. and Migeon, C.J. (1975) Puberty in boys: correlation
of plasma levels of gonadotropins (Ih,fsh), androgens
ttestoserone, androstenedione, dehydroepiandrosterone
and its sulfate), estrogens (estrone and estradiol), and
progestins (progesterone and 17.hydroxyprogesterone) with
physical changes. J. Clin. Endocrinol. Metab. 41, 556-562.
Marshall, J.C., Dalkin, A.C., Haisenieder, D.J., Paul, S.J.,
Ortolano, G.A. and Kelch, R.P. (1991) Gonadotropin-re-
leasing hormone pulses: regulators of gonadotropin syn-
thesis and ovulatory cycles. Rec. Progr. Horm. Res. 47,
155-187. Mascie-Taylor, C.G.N. (1991) Ciosocial influences on stature:
a review. J. Biosoc. Sci. 23, 113-128.
McConville, B.J., Boag, L.C. and Purohit, A. (1973) Three
types of childhood depression. Can. Psychiatr. Assoc. J. 18,
133-137.
McGee, R., Feehan, M., Williams, S., Partridge, F., Silva,
P.A. and Kelly, J. (1990) DSM-IIIM disorders in a large
sample of adolescents. J. Am. Academy Child Adolesc.
Psychiatry 29, 611-619.
McGee, R. and Williams, S. (1988) A longitudinal study of
depression in nine-year-old children. J. Am. Acad. Child
Adolesc. Psychiatry 27, 342-348.
Mitchell, J., McCauley, E., Burke, P.M. and Moss, S.J. (1988)
Phenomenology of depression in children and adolescents.
J. Am. Acad. Child Adolesc. Psychiatry 27, 12-20.
Moffitt, T.E., Caspi, A., Belsky, J. and Silva, P.A. (1992)
Childhood experience and the onset of menarche: a test of
a sociobiological model. Child Dev. 63, 47-58.
Montgomery, J.C., Brincat, M., Tapp, A., Appleby, L., Versi,
E., Fenwick, P.B.C. and Studd, J.W.W. (1987) Effect of
estrogen and testosterone implants on psychological disor-
ders in the climacteric. Lancet 7, 297-299.
Nada Raja, S., Feehan, M., Stanton, W.R. and McGee, R.
(1992) Prevalence and correlates of the premenstrual syn-
drome in adolescence. J. Am. Acad. Child Adolesc. Psy-
chiatry 31, 783-789.
Nolen-Hoeksema, S., Girgus, J.S. and Seligman, M.E.P. (1991)
Sex differences in depression and explanatory style in
children. J. Youth Adolesc. 20, 233-245.
Nottelmann, E.D., Susman, E.J., Darn, L.D., Inoff-Germain,
G., Loriaux, D.L., Cutler, Jr., G.B. and Chrousos, G.P. (1987a) Developmental processes in early adolescence: I.
Relations among chronologic age, pubertal stage, height, weight, and serum levels of gonadotropins, sex steroids,
and adrenal androgens. J. Adolesc. Health Care 8, 246- 260.
Nottelmann, E.D., Susman, E.J., Inoff-Germain, G., Cutler,
G.B., Loriaux, D.L. and Chrousos, G.P. (1987b) Develop-
mental processes in early adolescence: relationships be-
tween adolescent adjustment problems and chronological
age, pubertal stage, and puberty-related serum hormone
levels. J. Pediatr. 110, 473-480.
Nottelmann, E.D., Inoff-Germain, G., Susman, E.J. and
Chrousos, G.P. (1990) Hormones and behavior in puberty.
In: J. Bancroft and J.M. Reinisch (Eds.), Adolescence and
Puberty. New York: Oxford Press, pp. 888123.
Olweus, D., Mattsson, A., Schalling, D. and Low, H. (1988)
Circulating testosterone levels and aggression in adoles-
cent males: a causal analysis. Psychosom. Med. 50, 26l-
272.
Orvaschel, H., Walsh-Allis, G., Ye, W. and Walsh, G.T (1988)
Psychopathology in children of parents with recurrent
depression. J. Abnorm. Child Psychol. 16, 17-28.
Paikoff, R.L., Brooks-Gunn, J. and Warren, M.P. (1991) Ef- fects of girls’ hormonal status on depressive and aggressive
symptoms over the course of one year. J. Youth Adolesc.
20, 191-215.
Parker, L.N., Sack, J., Fisher, D.A. and Odell, W.D. (1978)
The adrenarche: prolactin, gonadotropins, adrenal andro-
gens, and cortisol. J. Clin. Endocrinol. Metab. 46, 396-404.
Petersen, A.C., Sarigiani, P.A. and Kennedy, R.E. (1991)
Adolescent depression: Why more girls? J. Youth Adolesc.
20. 247-271. Pollitt, E. and Leiber, R. (1980) Biological and social corre-
lates of failure to thrive. In: L.S. Greene and F.E. John-
ston (Eds.), Social and Biological Predictors of Nutritional
Status, Physical Growth, and Neurological Development.
New York: Academic Press, pp. 174-200.
Puig-Antich, J. (1986) Psychobiological markers: effects of age
and puberty. In: M. Rutter, C. Izard and P. Read (Eds.),
Depression in Young People - Issues and Perspectives.
New York: Guildford Press, pp. 341-381.
Puig-Antich, J. (1987) Affective disorders in children and
adolescents. In: H.Y. Meltzer (Ed.), Psychopharmacology:
the Third Generation of Progress. New York: Raven Press,
pp. 843-859.
Reiter, E.O., Fuldauer, V.G. and Root, A.W. (1977) Secretion
of the adrenal androgen, dehydoreplandrosterone sulfate,
during normal infancy, childhood, and adolescence, in sick infants. and in children with endocrinologic abnormalities.
J. Pediatr. 90, 76-80.
Rutter, M., Tizard, J. and Whitmore, K. (1970) The selection
of children with psychiatric disorder. Education, Health
and Behaviour. London: Longman.
Rutter, M., Graham, P., Chadwick, O.F.D. and Yule, W.
(1976) Adolescent turmoil: fact or fiction? J. Child Psy-
chol. Psychiatry 17, 35-56.
Rutter, M., Angold, A.. Taylor, E.J., Harrington, E.J. and
Nicholls, J. (1989) Age and sex effects on depressive and
other symptomatology. Presentation at the Annual Meet-
ing of the Society for Research in Child Development - May, Kansas City.
Rutter, M. (1990) Changing Patterns of Psychiatric Disorders
1.58
During Adolescence. In: J. Bancroft and J.M. Reinisch
(Eds.), Adolescence and Puberty. New York: Oxford Uni- versity Press, pp. 124-145.
Ryan, N., Puig-Antich, J., Ambrosini, P.J., Rabinovich. H.,
Robinson, D.. Nelson, B., lyengar, S. and Twomey, J.
t 1987) The clinical picture of major depression in children
and adolescents. Arch. Gen. Psychiatry 44, 854-861.
Saugstad, L.F. (1989a) Social Class, Marriage, and Fertility in
Schizophrenia. Schizophr. Bull. 15, Y-43.
Saugstad, L.F. (lY89b) Age at puberty and mental illness. Br. J. Psychiatry 155, 536-544.
Shaffer, D. (1982) Diagnostic considerations in suicidal behav-
ior in children and adolescents. J. Am. Acad. Child Ado-
lest. Psychiatry 21, 414-415.
Shaffer, D., Fisher, P., Piacentini, J., Schwab-Stone, M. and
Wicks, J. (1989) Diagnostic interview schedule for children (DISC-2.1C) Child Version, unpublished.
Simmons, R.G. and Blyth, D.A. (1992) Moving into adoles-
cence: the impact of pubertal change and school context.
In: P.H. Rossi, M. Useem, and J.D. Wright (Eds.), Social
Institutions and Social Change. New York: Aldine De Gruyter.
Sizonenko, P.C., Burr, I.M., Kaplan, S.L. and Grunbach,
M.M. 1970) Hormonal changes in puberty. II. Correlation of serum luteinizing hormone and follicle stimulating hor-
mone with stages of puberty and bone age in normal girls.
Pediatr. Res. 4, 36-45.
Sizonenko, P.C. and Paunier, L. (1975) Hormonal changes in
puberty III: correlation of plasma dehydroepiandros-
terone, testosterone, FSH, and LH with stages of puberty
and bone age in normal boys and girls and in patients with
Addison’s disease or hypogonadism or with premature or
late adrenarche. J. Clin. Endocrinol. Metab. 41, 894-904.
Skuse, D. (1984) Extreme deprivation in early childhood. J.
Child Psychol. Psychiatry 25, 543-572.
Stattin, H. and Magnusson, D. (1990) Paths Through Life.
Vol. 2: Pubertal Maturation in Female Development,
Hillsdale, NJ: Lawrence Erlbaum Associates.
Strober, M., Green, J. and Carlson, G.A. (1981) Phenomenol-
ogy and subtypes of major depressive disorder in adoles-
cence. J. Affect. Disord. 3, 281-290.
Susman, E.J., Nottelmann, E.D., Inoff-Germain, G.E., Dorn,
E.D.. Cutler, G.B., Loriaus, D.L. and Chrousos, G.P. (1985) The relation of relative hormonal levels and physi-
cal development and social-emotional behavior in young
adolescents. J. Youth Adolesc. 14, 245-254.
Susman. E.J., Inoff-Germain, G., Nottelmann, E.D., Loriaux,
D.L., Cutler, G.B. and Chrousos, G.P. (1987a) Hormones,
emotional dispositions, and aggressive attributes in young
adolescents. Child Dev. 58. 1114-l 134.
Susman, E.J., Nottelmann, E.D., Inoff-Germain, G., Dorn.
L.D. and Chrousos, G.P. (1987b) Hormonal influences on aspects of psychological development during adolescence.
J. Adolesc. Health Care 8, 492-504. Susman, E.J., Dorn, L.D. and Chrousos, G.P. (1991) Negative
affect and hormone levels in young adolescents: concur-
rent and predictive perspectives. J. Youth Adolesc. 20. 1677lYO.
Tolis, G. and Stefanis, C. (1983) Depression: biological and
neuroendocrine aspects. Biomed. Pharmacother. 37, 316- 322.
Velez, C.N., Johnson, J. and Cohen, P. (1989) A longitudinal
analysis of selected risk factors of childhood psychopathol-
ogy. J. Am. Acad. Child Adolesc. Psychiatry 28, X61-864.
Vihko, H. and Apter, D. (1980) The role of androgens in
adolescent cycles. J. Steroid Biochem. 12, 369-373.
Warner, P., Bancroft, J., D&on, A. and Hampson, M. (1991)
The relationship between perimenstrual depressive mood and depressive illness. J. Affect. Disord. 23, 9-23.
Weissman, M.M., Gammon, G.D., John, K., Merikangas, K.R.,
Warner, V., Prusoff, B.A. and Scholomskas, D. (1987) Children of depressed parents: increased psychopathology
and early onset of major depression. Arch. Gen. Psychiatry
44, 847-853.
Weissman, M.M. and Klerman, G.L. (1977) Sex differences and the epidemiololy of depression. Arch. Gen. Psychiatry
34,98-111.
Wennink, J.M.B., Delemarre-van de Waal, H.A., Schoe-
maker, R., Schoemaker, H. and Schoemaker, J. (1989)
Luteinizing hormone and follicle stimulating hormone se-
cretion patterns in boys throughout puberty measured
using highly sensitive immunoradiometric assays. Clin. En-
docrinol. 31, 551-564.
Wennink, J.M.B., Delemarre-van de Waal, H.A., Schoe-
maker, R., Schoemaker, H. and Schoemaker, J. (1990)
Luteinizing hormone and follicle stimulating hormone se-
cretion patterns in girls throughout puberty measured
using highly sensitive immunoradiometric assays. Clin. En-
docrinol. 33, 333-344.
Wierman, M.E., Beardsworth, D.E., Crawford, J.D., Crigler,
Jr.,J.F., Mansfield, M.J., Bode, H.H., Boepple, P.A., Kush-
ner, D.C. and Crowley, Jr., W.F. (1986) Adrenarche and
skeletal maturation during luteinizing hormone releasing
hormone analogue suppression of gonadarche. J. Clin.
Invest. 77, 121-126.
Winter, J.S.D. and Faiman, C. (1972) Pituitary-gonadal rela-
tions in male children and adolescents. Pediatr. Res. 6,
126-135. Winter, J.S.D. and Faiman, C. (1973) Pituitary-gonadal Rela-
tions in Female Children and Adolescents. Pediatr. Res. 7,
948-953.
Worthman, CM. (1992) Bio-cultural Interactions in Human
Development. In: M. Pereira and L. Fairbanks (Eds.),
Juvenile Primates: Life History, Development and Behav-
ior. Oxford: Oxford University Press.
Wu, F.C., Butler, G.E., Kelnar, C.J.H. and Sellar, R.E. (1990) Patterns of pulsatile luteinizing hormone secretion before
and during the onset of puberty in boys: a study using an
immunoradiometric assay. J. Clin. Endocrinol. Metab. 70, 629-637.