Generalized anxiety and C-reactive protein levels: a prospective, longitudinal analysis.

BACKGROUND: Generalized anxiety disorder (GAD) is highly co-morbid with depression. Depression is associated with elevated levels of the inflammation marker C-reactive protein (CRP), cross-sectionally and over time. To date, no studies have looked at the association between CRP and GAD. METHOD: A total of nine waves of data from the prospective population-based Great Smoky Mountains Study (n=1420) were used, covering children in the community aged 9-16, 19 and 21 years old. Structured interviews were used at each assessment to assess GAD symptoms, diagnosis and cumulative episodes. Blood spots were collected and assayed for high-sensitivity CRP levels. RESULTS: GAD was associated with increased levels of CRP in bivariate cross-sectional analyses. These bivariate associations, however, were attenuated after accounting for demographic, substance-use and health-related covariates. In longitudinal models, there was little evidence that CRP predicted later GAD. Associations from GAD to later CRP were attenuated in models adjusted for health-related coavariates and there was evidence that the GAD-CRP association was mediated by body mass index (BMI) and medication use. CONCLUSIONS: Similar to depression, GAD was associated with elevated levels of CRP, but the effect of GAD on CRP levels was explained by the effect of GAD on health-related behaviors such as BMI and medication use. This study suggests differences in the association between inflammation and depression and GAD.

Nursing frequency, gonadal function, and birth spacing among !Kung hunter-gatherers.

Mothers among !Kung hunter-gatherers nurse briefly and frequently, with brief intervals between nursing bouts (mean +/- standard error, 13.19 +/- 1.28 minutes). The low levels of 17 beta-estradiol and progesterone in the serum of the mother are correlated with infant's age and with interbout interval, but not with total nursing time. Maternal gonadal function is apparently suppressed by a timing-dependent, prolactin-mediated effect of breast stimulation. Interbout interval may be a key variable in lactation infertility. If so, it solves the puzzle of !Kung birth spacing.

The Great Smoky Mountains Study of Youth. Goals, design, methods, and the prevalence of DSM-III-R disorders.

BACKGROUND: The Great Smoky Mountains Study of youth focuses on the relationship between the development of psychiatric disorder and the need for and use of mental health services. METHODS: A multistage, overlapping cohorts design was used, in which 4500 of the 11758 children aged 9, 11, and 13 years in an 11-county area of the southeastern United States were randomly selected for screening for psychiatric symptoms. Children who scored in the top 25% on the screening questionnaire, together with a 1 in 10 random sample of the rest, were recruited for 4 waves of intensive, annual interviews (n = 1015 at wave 1). In a parallel study, all American Indian children aged 9, 11, and 13 years were recruited (N = 323 at wave 1). RESULTS: The 3-month prevalence (+/-SE) of any DSM-III-R axis I disorder in the main sample, weighted to reflect population prevalence rates, was 20.3% +/- 1.7%. The most common diagnoses were anxiety disorders (5.7% +/- 1.0%), enuresis (5.1% +/- 1.0%), tic disorders (4.2% +/- 0.9%), conduct disorder (3.3% +/- 0.6%), oppositional defiant disorder (2.7% +/- 0.4%), and hyperactivity (1.9% +/- 0.4%). CONCLUSIONS: The prevalence of psychiatric disorder in this rural sample was similar to rates reported in other recent studies. Poverty was the strongest demographic correlate of diagnosis, in both urban and rural children.

Testosterone levels change with subsistence hunting effort in !Kung San men.

Although little is known empirically of the physiology of human hunting, arguments for innate biological bases of gender-dimorphic behaviors such as aggression frequently point to the role of hunting in human evolution. Study of !Kung San hunter-gatherer men demonstrated that the diurnal pattern in serum testosterone was altered during a six-day hunt, compared to pre- and post-hunt levels, due mainly to elevation of evening values. Hunting success did not correlate with any testosterone measures. The pattern of changes observed is most consistent with the known concomitants of moderate prolonged exercise.

Puberty onset of gender differences in rates of depression: a developmental, epidemiologic and neuroendocrine perspective.

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.

Ratios of plasma and salivary testosterone throughout puberty: production versus bioavailability.

Because diffusion of testosterone (T) into the salivary gland is thought to be largely limited to the free, biologically active fraction, salivary testosterone is expected to provide a better measure of testosterone bioavailability in the body than is plasma testosterone. Matched saliva and blood spot samples were collected from 218 Zimbabwean males (age 11-23) who were at different stages of puberty, as assessed by self-reported Tanner genital stage ratings. Testosterone concentrations in these matched samples were highly correlated (r = 0.83). Both salivary and plasma testosterone (converted from blood spot value) showed expected significant increases across puberty. However, plasma testosterone distinguished among subjects at different stages of genital development more effectively than did salivary testosterone, suggesting the former to be a better marker of testosterone bioavailability. Sex hormone-binding globulin (SHBG) levels were also measured in a subgroup of 93 of these subjects. After controlling for plasma T concentrations, we found a small but significant inverse correlation between blood spot SHBG levels and the proportion of plasma testosterone recovered in salvia, supporting the hypothesis that SHBG-related changes in T bioavailability are detectable in saliva. We conclude that salivary testosterone accurately reflects testicular production of testosterone, but that neither salivary testosterone nor plasma testosterone is clearly superior to the other as a measure of testosterone bioavailability.

Comparative study of flex heart rate in three samples of Nepali boys.

Flex heart rate (HR, beats per minute) has assumed increasing importance in studies of energy expenditure and physical activity. Flex HR is defined as the mean of maximum rest and minimum exercise HR recorded during a standard test. This report examines methodological and substantive issues regarding the measurement and interpretation of population variation in flex HR values. Flex HR was determined for 80 Nepali 10-14-year-old boys living in contrasting physical and socioeconomic environments (31 village boys, 24 urban middle-class schoolboys, and 25 homeless street boys). The three populations exhibited significant differences in flex HR, with villagers averaging lower values than school or homeless boys (91, 100, and 103 bpm, respectively, P < 0.0001). They also differed in mean resting HR (P < 0.0002), but not in the percentage increase of flex over resting HR. To evaluate reliability, flex HR measurements were repeated on 14 school boys after a 3-month interval. Mean initial and repeat values were not significantly different, but discrepancies were large for individual subjects (-15 to +24 bpm). This suggests that flex HR is reliable at the population level, but not necessarily stable for individuals over time. The range of factors contributing to variation in flex HR between and within populations have implications for the use of HR monitoring to estimate levels of physical activity. © 1996 Wiley-Liss, Inc.

Heart rate monitoring of physical activity among village, school, and homeless Nepali boys.

Continuous HR monitoring, in conjunction with self-reports and direct observation of physical activity, was undertaken in three samples of 10-13-year-old Nepali boys living in different physical and socioeconomic environments. A total of 134 daytime HR profiles were recorded from 67 boys (76 from 31 villagers, 39 from 20 urban middle-class schoolboys, and 19 from 16 homeless street boys). Habitual levels of physical activity were compared using mean daytime HR (bpm) and the following indicators: percentage time during which boys are active (HR> flex, defined as an individual's average between resting and exercising HR), moderately active (HR> flex + 20%, also HR> 119 bpm), and vigorously active (HR> 139 bpm) over 10 hours of daylight. Mean daytime HR (102 bpm) and percentage time spent vigorously active (4%) did not differ between samples, despite obvious differences in lifestyles; thus better indicators are needed to characterize levels of physical activity. Percentage time spent active (HR> flex) was significantly higher (P < 0.0001) for villagers participating in subsistence activities, as a result of low flex HR values, indicating a higher level of physical fitness. Evaluation of moderate physical activity was sensitive to the choice of indicator, either an individual calibrated (above flex) or an absolute threshold of HR elevation. The former may be more appropriate than the latter to compare physical activity levels in populations with different lifestyles and levels of physical fitness. © 1996 Wiley-Liss, Inc.

The weanling's dilemma reconsidered: a biocultural analysis of breastfeeding ecology.

Infant feeding practices exert recognized effects on infant survival and command close attention from pediatricians, nutritionists, public health workers, demographers, social scientists, and parents. Despite considerable research and public health efforts, models to inform decision-making about timing of supplementation in policy and practice remain unsatisfactory. In the present review, new models are derived through analysis of the weanling's dilemma. After approximately 6 months, breast milk alone cannot meet the infant's nutritional requirements for growth, but nutritional supplementation raises risks of illness and malnutrition. Timing of supplementation is conditioned by complex short- and long-term trade-offs among infant and maternal needs and constraints. No formulaic solution can be prescribed for this dilemma, but optimal timing of supplementation can be ascertained from infant-, maternal-, and locale-specific conditions. Models for these determinants and trade-offs are presented and applied to data from longitudinal studies of breastfeeding, infant development, and interbirth interval.

Reproduction in the vervet monkey (Cercopithecus aethiops): III. The menstrual cycle.

The menstrual cycles of 17 multiparous vervet monkeys were studied. Based on estradiol, progesterone, and LH profiles, ovulation is predicted to occur on day 13 of the 32.4-day menstrual cycle. Estradiol peaked on the day preceding the LH peak in 75% of cycles. Average luteal phase length (progesterone greater than 4 nmol/l) was 18 days, with progesterone rising above 4 nmol/l on the day of the LH peak. Vaginal cytology and perianal skin coloration exhibited too much within- and among-animal variability to be reliable indicators of menstrual cycle stages. Uterine biopsies of the proliferative phase were characterized by mild pseudostratification of the columnar epithelium and absence of glandular secretion; in contrast, those of the luteal phase had marked pseudostratification of the tall columnar epithelium with glandular secretions in the lumen. A few follicular-phase samples contained structures such as tortuous uterine glands with secretions. Such structures are more characteristic of the luteal phase. It is suggested that their presence can be explained by incomplete sloughing of the endometrium at menstruation, as this is known to be light or convert in this species.

Reproduction in the vervet monkey (Cercopithecus aethiops): II. Annual menstrual patterns and seasonality.

Menstrual patterns and progesterone levels were monitored for 5 years from a cohort of 28 female vervet monkeys that were individually caged indoors. Three distinct cycle types (short, normal, and prolonged) were defined according to cycle length. Mean length of the normal cycle (32.5 days) and menses duration (4.8 days) are in agreement with previous reports. Prolonged cycles (> 50 days) contributed 20% of the total, with a decreased incidence during the natural peak breeding period. Weekly progesterone measurements indicated that many prolonged cycles were associated with an extended luteal phase, while others were probably due to lack of ovulation. From these data it would appear that the vervet monkey, although not strongly seasonal, does favor a particular time of year for breeding in a colony housed indoors.

Effects of ketamine anaesthesia, stress and repeated bleeding on the haematology of vervet monkeys.

Haematology values are presented for the vervet monkey (Cercopithecus aethiops), and the relative effects of high dose ketamine anaesthesia, stress of capture and repeated bleedings assessed. Anaesthesia resulted in decreased WBC and RBC values, attributed to depression of cardiovascular function. These effects were the reverse of those of alarm and strenuous exercise (leukocytosis and polycythaemia) during capture. Stress resulted in relatively high white and low red blood cell counts. Opposing effects of stress and anaesthesia led to comparable haematological values for trained, non-anaesthetized vervets and stressed, anaesthetized vervets. Effects of repeated bleedings were opposite in anaesthetized and non-anaesthetized animals. These effects, however, along with those of ketamine anaesthesia and stress, were relatively insignificant compared with the wide variation in haematological values found among individuals. The biological importance of these effects thus appeared to be slight. The concept of 'normal values' is discussed.

Social change in adolescent sexual behavior, mate selection, and premarital pregnancy rates in a Kikuyu community.

PIP: This report documents an example of interactions of cultural change with adolescent fertility and marriage patterns in an East African community. Between 1950 and 1980 the rate of unwed motherhood in Ngeca, Kenya, showed a marked increase from 0% in the 1940s to 11.4% in the 1960s. The authors present evidence of recent changes in Kikuyu culture that may account for this change. Traditional Kikuyu culture structured adolescence through status and role changes bounded and reinforced by ritual and instruction. Abandonment of traditional initiation rites and attenuation of the age-set system have most markedly altered the structure of adolescent experience by shifting the content and context of socialization. Major agents for change in this process have been the school, church, and modern economy. Responsibility for mate selection has remained with young people, but the determinants of partner desirability and gender ratios in partner availability have shifted considerably. Traditional criteria of male desirability included ability to pay bridewealth and to provide the wife with land; diligence and demeanor measured female attractiveness. At present, education and wage earning capacity affect partner attractiveness of each sex. The decline of polygyny has both shifted the balance of competition for spouses toward females, and has had significant repercussions in the marital and reproductive histories of males. Decreases in brideprice and reversals in direction of transfers of wealth at marriage are tangible signs of change in the marriage market. Deritualization of genital operations and attendant weakening of the age-set system have interrupted the flow of information on sex behavior and reproduction, controlled physical intimacy, and partner selection reinforced by peer pressure. Denial of contraception, the continued importance of marriage and fertility, and ambivalence toward sexual activity in adolescence all support adolescent sexual experimentation and increasing rates of premarital pregnancy.^ieng

Measurement of gonadotropins in dried blood spots.

We describe direct immunofluorometric assays for luteinizing hormone (hLH) and follicle-stimulating hormone (hFSH) in fingerstick blood spots dried on filter paper, based on modifications of commercially available kits. Determinations are made from 2.5-mm-diameter discs (3 microL of dried blood) punched out from blood spot standards and samples. Sample dose detection limits of the assays (IU/L) are 0.26 for LH and 0.13 for FSH, with mean interassay CVs of 11.6% (LH) and 7.8% (FSH) at low concentrations. Analytical recoveries of added hormone averaged 100% for LH and 95% for FSH. Clinical studies showed that values for blood spots (x) and directly assayed plasma (y) are highly correlated, so that results from blood spots can be converted directly to plasma equivalents, as follows: yLH = 0.07 + 1.90 xLH, and yFSH = 0.424 + 2.207 xFSH. These gonadotropins are stable in blood spots for at least a year under refrigeration; LH for at least 8 weeks and FSH 6 weeks at 22 degrees C; and both hormones for a week at 37 degrees C. These methods thus allow self-sampling, serial sampling, and mailing of specimens.

Hormone measures in finger-prick blood spot samples: new field methods for reproductive endocrinology.

Comparative endocrine studies have notably advanced understanding of ecological factors that contribute to variation in human reproductive function. Such research has relied on methodological advances that permit hormone determinations in samples that are easily and safely collected, stored, and transported, most recently on measurement of steroids in saliva. This report seeks to further expand the scope of endocrine research by demonstrating the value of blood spot samples collected by finger prick. As a sampling strategy, finger-prick blood spot collection offers the advantages of short collection time, low invasiveness, repeatability, absence of postcollection processing, low biohazard risk, and ease of sample storage and transport. We document good sample stability and present sensitive assay methods for a range of steroids and proteins (FSH, LH, PRL, T, E2, DHEAS, androstenedione, cortisol, SHGB) in blood spots that require sample volumes of 3-12 microliters and display good reliability, specificity, precision, accuracy, and convertibility of results to plasma/serum equivalent concentrations. Laboratory evaluation was augmented by a feasibility study at a remote site in Papua New Guinea that confirmed validity and stability of blood spot collections under field conditions. Research applications of blood spot sampling are illustrated with a series of studies, including cross-sectional surveys for developmental and life span endocrinology, a longitudinal, population-based developmental epidemiologic study of puberty, and serial sampling in a dynamic study of neuroendocrine response to suckling. We conclude that the sampling features and wide range of measurable biomolecules of blood spots do constitute a methodological advance for endocrine research.

Sensitive salivary estradiol assay for monitoring ovarian function.

Measurement of steroids in saliva has excited interest because of the numerous potential clinical applications; noninvasive, convenient sampling; and apparently accurate reflection of the concentrations of physiologically active unbound steroid in the circulation. Although assays of saliva for several steroid hormones are available and widely used, assays for salivary estradiol are not, primarily because of methodological limitations. By modifying a commercially available kit for serum estradiol, our laboratory has developed a procedure that is sensitive, highly specific, and reliable for measuring salivary estradiol. Assay sensitivity is 0.5 fmol (0.14 pg; sample concentration 1.3 pmol/L) with a mean interassay CV of 10.8% at low concentrations. Clinical studies showed that values for serum and saliva are highly correlated (P less than 0.001), and demonstrated reliable detection of estradiol peaks during normal ovulatory cycles in serial samples from 15 women. Salivary estradiol peaked at 5.4 (SD 1.9) pmol/L on cycle day 14.4 (SD 3.2), 1.2 (SD 0.8) days before ovulation detected by ultrasound. This assay may be particularly helpful in investigating ovarian function and free estradiol in women at various stages of the reproductive cycle.

Puberty and depression: the roles of age, pubertal status and pubertal timing.

BACKGROUND: Previous work has indicated that the 2:1 female:male sex ratio in unipolar depressive disorders does not emerge until some time between ages 10 and 15. METHODS: Data from four annual waves of data collection from the Great Smoky Mountains Study (GSMS) involving children aged nine to 16 were employed. RESULTS: Pubertal status better predicted the emergence of the expected sex ratio than did age. Only after the transition to mid-puberty (Tanner Stage III and above) were girls more likely than boys to be depressed. The timing of this transition had no effect on depression rates. Before Tanner Stage III, boys had higher rates of depression than girls, and the prevalence of depression appeared to fall in boys at an earlier pubertal stage than that at which it began to rise in girls. In addition, recent transition to Tanner Stage III or higher had a transient effect in reducing the prevalence of depression in boys. CONCLUSIONS: The period of emergence of increased risk for depression in adolescent girls appears to be a relatively sharply demarcated developmental transition occurring in mid-puberty. Previously reported effects of the timing of puberty (which have tended to be transient) appeared less important in increase of risk for depression than pubertal status.

Attenuation of nursing-related ovarian suppression and high fertility in well-nourished, intensively breast-feeding Amele women of lowland Papua New Guinea.

Intense, sustained nursing lengthens inter-birth intervals and is causally linked with low natural fertility. However, in traditional settings, the effects of such nursing on fertility are difficult to disentangle from those of nutrition. Results from a prospective, direct observational study of reproductive function in well-nourished Amele women who nurse intensively and persistently but who also have high fertility are here presented. Endocrine measures show that ovarian activity resumes by median 11.0 months postpartum. Median duration of postpartum amenorrhoea is 11.3 months, time to next conception is 19.0 months, and the inter-birth interval is 28.0 months. Average life time fertility is 6.8. High fertility in Amele women is due both to refractoriness of reproductive function to suckling stimuli, and to maintenance of equivalent age-specific fertility rates across the reproductive life span.

PIP: Evidence is presented, from a study of 52 breast-feeding women with an infant under 5 years of age from 12 Amele villages in lowland Papua New Guinea, that women with good nutritional status experienced less variation in ovarian suppression than reported elsewhere among less well-nourished populations. Direct observations of nursing behavior were recorded during 660 hours. The average number of observations per child, which were conducted at different developmental stages, was 1.9. 1549 nursing events were recorded. 38 women had serum samples drawn to determine prolactin levels. Saliva samples were collected from 51 women for measurement of gonadal steroids. Women reported menstrual patterns in the preceding month. Measures were also taken of weight, skinfolds of the triceps and subscapular area, and mid-upper-arm circumference. Birth interval, completed fertility, and age at introduction of solid foods and denial of breast were obtained from ongoing demographic surveillance on nutrition in 1982-84 and 1982-83. Laboratory methods were described for serum prolactin and progesterone analysis. The results showed skewed results for the hormonal and some nursing variables, which were transformed by logarithmic functions. All Amele mothers practice indulgent demand feeding schedules. Supplementary foods are introduced at the median age of 7.4 methods. Early supplementation is liquids or semiliquids, followed by mashed starchy staples, and finally a standard diet. Cessation of breast feeding occurs at a subsequent pregnancy of the advanced age of the child (5 years). The median age of cessation was 36.3 months. The breast feeding pattern during the first 18 months was 1.5-3 times per hour for 1-3 minutes. Infant age was not linearly related to nursing frequency declines or feeding intervals. Nursing frequency was unrelated to morbidity. Prolactin declined strongly when bout frequency and interval were stable. Multiple regression findings were that duration of postpartum amenorrhea was explained by mother's parity and age, feeding interval, and time postpartum. Time of supplementary feeding was unrelated to duration of amenorrhea, nursing frequency, or feeding duration. 16% of interval variance was explained by length of postpartum amenorrhea, which was correlated with length of subsequent birth interval. 45% of the variance in prolactin was explained by triceps skinfolds and bout length.

Prolactin response to suckling and maintenance of postpartum amenorrhea among intensively breastfeeding Nepali women.

The aim of the study was to determine the association between PRL responses to suckling and maintenance of postpartum amenorrhea among breastfeeding mothers. Three blood spot samples (5, 30, and 50 min following a timed nursing bout) were collected from 71 intensively breastfeeding Nepali women for PRL determination. Maternal age, BMI (weight/height2), menstrual status, caste, infant age, nursing bout length, and duration of supplementation were recorded at time of sample collection. Independent and paired t tests, linear regression analyses, and general linear models were used to evaluate differences between cycling (n = 36) and amenorrheic (n = 35) women and associations among variables. Logistic regression analyses were used to relate PRL measures to the odds of maintaining lactational amenorrhea. Amenorrheic breastfeeding mothers had higher (P < .001) PRL levels at all 3 collection times than cycling breastfeeding mothers, and PRL levels declined with time since birth (P < 0.05). The odds (OR) of having ceased lactational amenorrhea was significantly higher (OR = 5.0, 95% Cl = 1.3-19.9) among mothers with lower PRL levels (< or = 10 ng/mL) at 50 min post-sucking, and PRL at 50 min showed a significant dose response relationship with menstrual status. The association between 50 min PRL levels and lactational amenorrhea appears to be independent of time postpartum, maternal age, BMI, nursing bout length, and duration of supplementation. Among intensively nursing women, maintenance of elevated PRL levels across the interbout interval increases the odds of maintaining lactational amenorrhea.

PIP: In August 1991, in Nepal, all women aged 19-45 from the Tamang caste (agro-pastoralists of Tibetan origin) and from the Kami caste (low-caste blacksmiths of Aryan origin) who were intensively breast feeding and lived in remote villages in the foothills of the Himalayas were included in a study aiming to examine the association between post-suckling prolactin (PRL) and menstrual status. The study also aimed to determine what PRL response time after suckling best predicts the odds of having resumed menses. There were 36 women whose menses had returned and 35 women who remained amenorrheic. 50-minute post-suckling PRL levels were more linked to menstrual status than 5-minute levels (odds ratio [OR] = 5 vs. 2.1). They (but not 5-minute post-suckling PRL levels) also had a significant dose relationship with menstrual status. During the first year postpartum, PRL levels were significantly higher at all three collection times (5, 30, or 50 minutes) in amenorrheic women than cycling women (p 0.001). PRL levels fell as infant age increased (16 ng/ml for 11 months or less, 13.7 ng/ml for 11-15 months, and 7.4 ng/ml for 15 months; p 0.05). At 50 minutes post-suckling, mothers with PRL levels at or below 10 ng/ml were much more likely to have returned to menses than those with higher levels (OR = 5; p = 0.041). The association between 50-minute PRL levels and lactational amenorrhea were independent of time postpartum, maternal age, body mass index, nursing bout length, and duration of supplementation. The maintenance of high PRL levels across the interbout interval increased the odds of maintaining lactational amenorrhea (p 0.05). In conclusion, intensively nursing women whose PRL levels remain high across the interbout interval are most likely to maintain lactational amenorrhea.

Pubertal changes in hormone levels and depression in girls.

BACKGROUND: Throughout their reproductive years, women suffer from a higher prevalence of depression than men. Before puberty, however, this is not the case. In an earlier study, we found that reaching Tanner Stage III of puberty was associated with increased levels of depression in girls. This paper examines whether the morphological changes associated with puberty (as measured by Tanner stage) or the hormonal changes underlying them are more strongly associated with increased rates of depression in adolescent girls. METHODS: Data from three annual waves of interviews with 9 to 15-year-olds from the Great Smoky Mountains study were analysed. RESULTS: Models including the effects of testosterone and oestradiol eliminated the apparent effect of Tanner stage. The effect of testosterone was non-linear. FSH and LH had no effects on the probability of being depressed. CONCLUSIONS: These findings argue against theories that explain the emergence of the female excess of depression in adulthood in terms of changes in body morphology and their resultant psychosocial effects on social interactions and self-perception. They suggest that causal explanations of the increase in depression in females need to focus on factors associated with changes in androgen and oestrogen levels rather than the morphological changes of puberty.