Abstract

Active masculinization by fetal testosterone is believed to be a major factor behind sex differentiation of the brain. We tested this hypothesis in a 15O-H2O positron emission tomography study of 11 women with congenital adrenal hyperplasia (CAH), a condition with high fetal testosterone, and 26 controls. Two indices of cerebral dimorphism were measured—functional connectivity and cerebral activation by 2 putative pheromones (androstadienone [AND] and estratetraenol [EST]), previously reported to activate the hypothalamic networks in a sex-differentiated manner. Smelling of unscented air was the baseline condition, also used for measurements of functional connectivity from the amygdala. In CAH women and control women AND activated the anterior hypothalamus, and EST the amygdala, piriform, and anterior insular cortex. The pattern was reciprocal in the male controls. Also the functional connections were similar in CAH women and control women, but different in control men. Women displayed connections with the contralateral amygdala, cingulate, and the hypothalamus, men with the basal ganglia, the insular and the sensorimotor cortex. Furthermore, the connections were in CAH and control women more widespread from the left amygdala, in men from the right amygdala. Thus, we find no evidence for masculinization of the limbic circuits in women with high fetal testosterone.

Introduction

Sexual dimorphism has been reported in the anatomy, neurochemistry as well as function of the human brain (Cahill 2006; Oh et al. 2007). Of particular interest are recent findings in the amygdala and hypothalamus, 2 limbic structures showing sex differences not only in the relative volumes and neuronal responses, but also in functional connections (Kilpatrick et al. 2006; Savic and Lindstrom 2008). The underlying mechanisms are unknown, albeit intensely discussed. One possibility, advocated by the principal theory about sexual dimorphism of the brain, is that exposure to androgens during the developmental period has organizational effects on cerebral structures by influencing neuronal survival and connections (De Vries et al. 2002; Negri-Cesi et al. 2004). This hypothesis finds substance in animal experiments, but is difficult to test in humans, because prenatal treatment with testosterone is unethical. Certain rare conditions, so called experiments of nature, may here potentially offer unique information. One such condition is congenital adrenal hyperplasia (CAH). This autosomal recessive genetic disorder has an incidence rate of about one in 15 000 live births (Merke and Bornstein 2005), and is characterized by severe androgen excess beginning in the fetus. In about 95% of patients CAH is caused by a defect in the 21-hydroxylase gene (CYP21), leading to an impaired synthesis of cortisol and aldosterone. The low cortisol level triggers an increased production of adrenocorticotropic hormone, resulting in hyperplasia of the adrenal glands with increased synthesis of steroid precursors and elevation of androgen levels. The androgen excess is present from early embryogenesis, and causes in girls varying degrees of virilization of the external genitalia, depending on the degree of enzyme deficiency. The syndrome has traditionally been divided into 3 clinical forms, the salt wasting form (SW), the simple virilizing (SV), and the nonclassic (NC) form. The most severe form is SW, characterized by no or a few percent of normal enzyme activity and a life-threatening salt loss during the neonatal period; in the affected females it also causes a pronounced prenatal virilization of external genitalia. The SV form does not include overt SW, but the affected girls still present with prenatal genital virilization. Boys, on the other hand, are often diagnosed due to accelerated growth and precocious pseudopuberty during early childhood. The mildest form, nonclassic CAH, is usually recognized first later in childhood or adulthood when hyperandrogenic symptoms such as accelerated somatic growth, precocious pseudopuberty, hirsutism, and/or decreased fertility become evident. Genital virilization in the 2 more severe forms of CAH leads to diagnosis soon after birth, and to early postnatal treatment with corticosteroids. Although treatment regulates the hormonal milieu, girls with CAH are reported to exhibit more male-typical childhood play than their unaffected sisters (Berenbaum and Resnick 1997; Berenbaum 1999; Meyer-Bahlburg 2001; Nordenstrom et al. 2002; Hall et al. 2004), which may be an effect of the prenatally elevated testosterone. Some studies indicate that CAH women also have better spatial abilities (Hampson et al. 1998; Resnick et al. 1986), whereas other failed to detect difference from female controls (McGuire et al. 1975; Helleday et al. 1994). Furthermore, a virilized (<1.0) 2D:4D digit ratio (Buck et al. 2003), lower rates of exclusively heterosexual fantasy (Zucker et al. 1996) and higher rates of bisexual or homosexual fantasy or experience have been reported (Ehrhardt et al. 1968; Money et al. 1984; Dittmann et al. 1992). The increase in bisexuality and homosexuality is of particular interest, considering that putative markers of high prenatal androgens have in several studies been associated with a heteroflexible or lesbian orientation (MacCulloch and Waddington 1981; Dancey 1990). We recently found that limbic and paralimbic networks process signals from putative pheromones in a sex-differentiated manner, and that the response pattern is sex atypical in lesbian women. Whereas smelling of a putative male pheromone androstadienone (AND) displayed activations of the anterior hypothalamus in heterosexual women (HeW) only the classical olfactory regions (the piriform cortex, the amygdala, and the anterior insular cortex) were engaged in heterosexual men (HeM) and in lesbian women (Savic et al. 2001; Berglund et al. 2006). The response pattern was almost reciprocal during exposure to the estrogen-like steroid estratetraenol (EST). Using positron emission tomography (PET) measurements of the resting state cerebral blood flow (rCBF) we recently found that also functional connections from the right and left amygdala were sex differentiated, and that in lesbian women these connections resembled those of HeW rather than HeM (Savic and Lindstrom 2008).

The mechanisms underlying these observations are not evident. In the present study we reasoned that if fetal testosterone has organizational effects on the human brain, the elevated testosterone levels in CAH women would be expected to cause sex atypical cerebral activations with AND and EST, as well as sex atypical patterns of amygdala connections.

Eleven CAH women were investigated in activation experiments similar to those employed in several of our previous studies (Savic et al. 2001, 2005; Berglund et al. 2006, 2007). Data from 13 HeM and women were used for comparisons.

Materials and Methods

Subjects

Eleven right-handed women with CAH (age 30 ± 8 years, range 20–38 years) and 26 right-handed heterosexual controls (13 women, age 26 ± 7, range 20–36 years; 13 men, age 28 ± 6, range 21–36 years) were studied. Control women will throughout the manuscript be denoted as HeW, and control men as HeM. Controls and patients were investigated over a longer, but overlapping, period of time (2004–2007). CAH was related to 21-hydroxylase deficiency in all patients. The diagnosis was based on CYP21 mutation analysis using allele-specific PCR from genomic DNA prepared from venous blood (Wedell and Luthman 1993). Four women had a SW severe form of CAH, 6 had mutations corresponding to the SV form of CAH, and one had the NC form (Table 1). None of the patients was treated with dexametasone during the fetal development. In 7 patients the diagnosis was set at birth, in 2 at age 1–2 years, and in another 2 at 9–10 years. Glucocorticoid treatment was initiated directly in all the patients, and all except 2 (no 6 and 9 in Table 1) were operated for genital androgenization. Patients had normal testosterone (0.5 ± 0.5 nmol/L) and estrogen levels (265 ± 231 pmol/L). For demographical data, please see Table 1.

Table 1

Patient data

Patent no. Age Mutation Category Severity Age at diagnosis 
46 I172N/I2splice SV At birth 
38 I172N/I2splice SV 4 months 
37 I172N/null SV At birth 
34 I2splice/I2splice SW (Kin 5) At birth 
33 I2splice/I2splice SW At birth 
27 V281L/I2spilce NC 9 years 
26 I2splice/null SW 9 months 
25 null/null SW (Kin 5) At birth 
23 P30L SV 10 years 
10 21 I172N/I2splice SV 30 months 
11 20 I172N/I2splice SV (Kin 2) At birth 
Patent no. Age Mutation Category Severity Age at diagnosis 
46 I172N/I2splice SV At birth 
38 I172N/I2splice SV 4 months 
37 I172N/null SV At birth 
34 I2splice/I2splice SW (Kin 5) At birth 
33 I2splice/I2splice SW At birth 
27 V281L/I2spilce NC 9 years 
26 I2splice/null SW 9 months 
25 null/null SW (Kin 5) At birth 
23 P30L SV 10 years 
10 21 I172N/I2splice SV 30 months 
11 20 I172N/I2splice SV (Kin 2) At birth 

Note: Mutation is categorized according to Wedell and Luthman (1993), with the minimum degree of severity 1, and maximum 5 (no CYP21 enzyme activity).

All subjects underwent physical and neurological examinations, and tests of olfactory thresholds with n-butyl alcohol (Savic et al. 2000). The controls were healthy and without any history or heredity for neuropsychiatric disorders. Handedness was assessed according to modified Edinburgh inventory (Oldfield 1971). All control subjects, and 8 CAH women rated themselves as entirely heterosexual (=0) on the Kinsey's Heterosexual/Homosexual scale (0 = maximally heterosexual, 6 = maximally homosexual), (Kinsey et al. 2003). One patient rated 2 on the scale, another rated 4, and a third 5.

Women were investigated during the second to third week of the menstrual cycle, assessed from participants’ verbal reports and estimated in relation to day 1 of the menstrual cycle.

The study was approved by the Human Subject Protection and Radiation Safety Committees at the Karolinska University Hospital.

Experimental Procedure

The Odorous Compounds

Like in our previous studies (Savic et al. 2000, 2001, 2005) the activation condition consisted of passive, birhinal, smelling (not sniffing) of AND, EST, and 4 different odors denoted as OO. AND and EST were during the PET scans presented in crystalline and odorous form (200 mg, Steraloids, Inc., Newport, RI). Their purity was tested by our doping laboratory, and assessed to be 98%. For further details about the experimental procedure, please see our previous reports (Berglund et al. 2006; Savic et al. 2000, 2001, 2005).

PET Experiments

rCBF was measured with PET during 4 conditions:

  • 1) Smelling AND.

  • 2) Smelling EST.

  • 3) Smelling 4 common odors presented on line during the same scan (butanol, cedar oil, lavendel oil, and eugenol) and denoted as OO.

  • 4) Smelling of room air (denoted as AIR), which was kept odorless by a suction devise in the scanner room.

AIR served as the baseline condition, and activations were defined as increases in rCBF during smelling of odorants compared with air. All the stimuli (odorous compounds, and room air) were presented in a glass bottle at a distance of 10 mm from the nose. There were twelve PET scans per person (3 scans per condition, balanced and randomly interleaved). Each scan lasted for 60 s (Savic et al. 2001). During the scans subjects were unaware of the identity of items, and instructed to restrain from sniffing or judging the odorants.

All the subjects were investigated in an identical way, and by the same experimenters. Thus, PET measurements were carried out at the same time of the day, with a standardized room temperature and air pressure (23 °C, 997 hPa; Savic et al. 2000, 2001). The experimental protocol and its justification has been described in detail elsewhere (Savic et al. 2000, 2001, 2005).

In summary, it included magnetic resonance imaging (MRI) scans (3D acquisition protocol of T1-weighted images, acquired at a 1.5 Tesla scanner, described in detail in our previous publications (Savic et al. 2001), and PET (full width at half maximum 3.8 mm) measurements of rCBF with 15O-H2O.

Respiratory movements were recorded continuously 2 min before, and during each scan, using a strain gauge around the lower thorax connected to a graph (Comair, Stockholm, Sweden) (Savic et al. 2001). After the PET scans subjects were asked to score each odorant for pleasantness, irritability, intensity and familiarity using a 100-mm visual analogue scale, as described previously (Savic et al. 2000).

Data Analysis

Activation.

The individual MRI and PET images were reformatted into a common space (standard brain), and filtered with 10-mm Gaussian kernel. Significant activations were determined with the SPM-statistics (SPM2, Wellcome Foundation, London, UK) (Friston and Penny 2003) using the following contrasts: AND–AIR, EST–AIR, and OO–AIR. Activations were first evaluated in each separate group with one-group random effect analysis, employing the entire brain as search space with exception of the caudal portion of the cerebellum (below the level of olives), which we failed to cover in 2 patients. Next, a 2-group random effect analysis was applied to test group differences. Finally, we used conjunctional analysis to investigate whether there were common activations between several groups. The height threshold for within group and conjunctional analyses was set at P = 0.001, with corrected P < 0.05. For group comparisons we used the uncorrected P < 0.05, assuming that group differences would be found either in the anterior hypothalamus or the odor processing regions (the amygdala and piriform cortex, the anterior cingulate cortex, and the insular cortex).

Resting state functional connectivity.

Functional connectivity was defined operationally as the extent to which normalized rCBF in seed regions of interest (ROIs) covaried with voxel-based rCBF values across the investigated subjects. Calculations were based on data from resting state 15O-H2O-PET (eyes closed, ears plugged, and normal breathing of the unscented air). The normalized rCBF was extracted from circular (5 mm) ROIs covering the right and left amygdala (center of gravity in Talairach's coordinates −18, 2, −18 and 15, 4, −16) with MarsBaR software. Significant covariations were, in accord with our previous studies (Ciumas et al. 2008), calculated at height threshold T = 3.0, corrected P < 0.05, and using the entire brain as search space except for the caudal portion of the cerebellum (multisubject condition and covariate analysis within SPM2). Based on previous studies of healthy controls (Kilpatrick et al. 2006; Savic and Lindstrom 2008) we assumed that the seed ROI would display greater connections with the contralateral amygdala, the hypothalamus, the anterior cingulate and the subcallosum in HeW compared with HeM, whereas the amygdala connections with the sensorimotor cortex, and the basal ganglia would be more pronounced in HeM than HeW. We also hypothesized that CAH women would display a pattern of connectivity resembling that of HeM rather than HeW. The predicted connectivity clusters were in the group comparisons considered significant if their T-value was 3.0 or more, P < 0.05 uncorrected. Connectivity clusters in unpredicted regions were considered significant if their value was P < 0.05 corrected across the entire search space.

Psychophysical parameters.

The mean respiratory amplitude and frequency was first calculated during each prescan and scan period. The percentage difference between the scan and prescan value was then compared between CAH women, HeM and HeW with respect to AIR, AND, EST, and OO using a 2-way ANOVA, factoring for subject group and stimulus type, as described earlier (Savic et al. 2005). A 2-way ANOVA was used also to test group differences in odor ratings, but the stimuli were AND, EST, and OO, because AIR was perceived as odorless. Finally, odor thresholds were compared between the 3 groups using one-way ANOVA. The significance level was 0.05 for all comparisons.

To test for other possible effects of androgenization on phenotype in our material we carried out a post hoc analysis of digit 2D:4D ratios. Low 2D:4D ratio, especially on the right hand, is reported to be a marker of high prenatal testosterone (Manning 2002). Using steel vernier calipers we measured the second and fourth fingers of both right and left hand. Measurements were carried out directly on the fingers, from the crease proximal to the palm of the digit to the tip of the digit, as well as on the photocopies of the ventral surface of the hands (Manning et al. 1998). For latter measurement participants were asked to straighten their fingers and very gently press their hand on the glass plate of the photocopier. Two investigators who were blind to each other's measurements (one also blind to the diagnosis) measured each finger independently. Correlation between measures of the 2 raters, as well as between photocopy and direct finger measurements was tested with simple regression (P < 0.05).

Results

Odor Activations

Within Group Activations

When smelling AND CAH women showed activation of the anterior hypothalamus (with the height maximum corresponding to the preoptic area (Schaltenbrand et al. 1982). In contrast, smelling EST yielded clusters in the amygdala, the piriform cortex, the anterior insular cortex, and in the left lingular gyrus, but not in the hypothalamus (Table 2 and Fig. 1). The activation pattern was similar in HeW, whereas in HeM it was reciprocal: HeM displayed a cluster in the anterior hypothalamus with EST but not AND, and showed activation of the left amygdala and piriform cortex, and the left parietal cortex when smelling AND (Table 2, and Fig. 1).

Table 2

Activations

 CAH women HeM HeW 
Region Z level Size (cm3Coordinates Z level Size (cm3Coordinates Z level Size (cm3Coordinates 
EST–AIR 
    Hypothalamus    4.6 1.4 6, −12, 2    
    R amygdala + piriform cortex 5.1 1.4 18, −22, 4    4.7 1.3 34, 0, −14 
    L amygdala + piriform + insular + cingulate cortex 3.7 1.2 −26, −10, −2 4.5 1.0 −22, −6,−24 4.1 0.9 −40, 16, −2 
    L lingular gyrus 4.1 0.8 −16, −50, 2       
AND–AIR 
    Hypothalamus 3.8 0.9 −10, −4, −8    6.0 1.0 −6, 0, −12 
    L amygdala + piriform cortex    4.2 0.5 −18, −10, −10    
    L insular cortex       4.0 0.9 −34, 18, −16 
    L parietal cortex    4.2 1.0 −18, −60, −24    
OO–AIR 
    R amygdala + piriform + insular + anterior cingulated cortex 4.5 1.0 18, −10, −4 5.1 3.3 20, 0, −10 5.2 3.2 24, 0, −8 
    L amygdala + piriform + insular cortex 4.5 2.8 −20, −2, −12 4.7 3.1 −20, 0, −10 4.0 2.0 −20, −4, −16 
 CAH women HeM HeW 
Region Z level Size (cm3Coordinates Z level Size (cm3Coordinates Z level Size (cm3Coordinates 
EST–AIR 
    Hypothalamus    4.6 1.4 6, −12, 2    
    R amygdala + piriform cortex 5.1 1.4 18, −22, 4    4.7 1.3 34, 0, −14 
    L amygdala + piriform + insular + cingulate cortex 3.7 1.2 −26, −10, −2 4.5 1.0 −22, −6,−24 4.1 0.9 −40, 16, −2 
    L lingular gyrus 4.1 0.8 −16, −50, 2       
AND–AIR 
    Hypothalamus 3.8 0.9 −10, −4, −8    6.0 1.0 −6, 0, −12 
    L amygdala + piriform cortex    4.2 0.5 −18, −10, −10    
    L insular cortex       4.0 0.9 −34, 18, −16 
    L parietal cortex    4.2 1.0 −18, −60, −24    
OO–AIR 
    R amygdala + piriform + insular + anterior cingulated cortex 4.5 1.0 18, −10, −4 5.1 3.3 20, 0, −10 5.2 3.2 24, 0, −8 
    L amygdala + piriform + insular cortex 4.5 2.8 −20, −2, −12 4.7 3.1 −20, 0, −10 4.0 2.0 −20, −4, −16 

Note: R = right; L = left. Clusters calculated with one-group random effect analysis, height threshold at P = 0.001, corrected P < 0.05. Italics denote clusters calculated at P = 0.001, corrected P < 0.1. Talairach's coordinates indicate local maxima.

Figure 1.

Illustration of group-specific activations with putative pheromones and odors. The Sokoloff's color scale illustrates Z values reflecting the degree of activation. As the same brain section is chosen, the figures do not always illustrate maximal activation for each condition. Clusters of activated regions are superimposed on the standard brain MRI, midsagittal plane.

Figure 1.

Illustration of group-specific activations with putative pheromones and odors. The Sokoloff's color scale illustrates Z values reflecting the degree of activation. As the same brain section is chosen, the figures do not always illustrate maximal activation for each condition. Clusters of activated regions are superimposed on the standard brain MRI, midsagittal plane.

Activations with OO were similar in all 3 groups of subjects, and confined to the previously described olfactory circuits (Table 2, Fig. 1).

Group Comparisons

The pattern of activation in CAH women differed significantly from the pattern in HeM. This difference was related to AND, as well as EST. Smelling AND caused significant activations in CAH women compared with HeM in the anterior hypothalamus and the subcallosum, whereas HeM activated the olfactory circuits (the amygdala, piriform and anterior insular cortex, and the cingulate), significantly more than CAH women (Table 3, Fig. 2). The difference was almost reciprocal for EST–AIR; HeM displayed significantly higher activation in the anterior hypothalamus compared with CAH women, whereas CAH women activated the anterior cingulate significantly more than HeM (Table 3, Fig. 2). No differences were observed between CAH women and HeW, and no differences between the 3 investigated groups were found in OO–AIR.

Table 3

Group differences

 HeM–CAH women CAH women–HeM 
Region Z level Size (cm3Coordinates Z level Size (cm3Coordinates 
EST–AIR 
    Hypothalamus 4.4 0.7 −2, −16, 2    
    Anterior cingulate    4.0 1.3 12, 36, −4 
AND–AIR 
    Hypothalamus    4.4 0.4 −12, –4, −8 
    Subcallosum    3.3 1.2 14, 24, −10 
    L amygdala + piriform + anterior insular cortex 4.3 4.1 −44, −18, −10    
    Posterior cingulate 5.9 2.0 2, −64, 6    
OO–AIR  ns   ns  
 HeM–CAH women CAH women–HeM 
Region Z level Size (cm3Coordinates Z level Size (cm3Coordinates 
EST–AIR 
    Hypothalamus 4.4 0.7 −2, −16, 2    
    Anterior cingulate    4.0 1.3 12, 36, −4 
AND–AIR 
    Hypothalamus    4.4 0.4 −12, –4, −8 
    Subcallosum    3.3 1.2 14, 24, −10 
    L amygdala + piriform + anterior insular cortex 4.3 4.1 −44, −18, −10    
    Posterior cingulate 5.9 2.0 2, −64, 6    
OO–AIR  ns   ns  

Note: Clusters calculated with one-group random effect analysis, height threshold at P = 0.001, uncorrected P < 0.05 (because the differentiated hypothalamic activation was predicted in the basis of our previous studies). Talairach's coordinates indicate local maxima. No differences were found between CAH women and heterosexual female controls, and no group differences were observed for OO–AIR.

Figure 2.

Group differences. Shown are the clusters calculated with 2-group random effect analysis, superimposed on standard brain MRI. The Sokoloff's color scale illustrates Z values reflecting the degree of activation. No significant clusters were observed when comparing CAH and HeW.

Figure 2.

Group differences. Shown are the clusters calculated with 2-group random effect analysis, superimposed on standard brain MRI. The Sokoloff's color scale illustrates Z values reflecting the degree of activation. No significant clusters were observed when comparing CAH and HeW.

As reported previously, differences between HeM and HeW were detected in the hypothalamus, and in the amygdala, piriform and insular cortex (Fig. 2). The local maxima for hypothalamus clusters corresponded to the Talairach's coordinate 8, −12, −4 (Z level = 4.1, cluster size 1.1 cm3) for AND–AIR in HeW–HeM, and 2, −8, 2, (Z level = 3.3, cluster size 1.1 cm3) for EST–AIR in HeM–HeW.

Not surprisingly, conjunctional analysis showed that independently of the type of stimulus (OO, AND, or EST), all 3 groups shared clusters in the amygdala, the piriform and the anterior insular cortex (Table 4, Fig. 3). In addition, CAH women shared a hypothalamic cluster with HeW when smelling AND. Thus, independently of the type of analysis CAH women expressed a “female” type of cerebral processing, and differed in this respect from HeM.

Table 4

Conjunctional clusters

 CAH women and HeM HeM and HeW CAH women and HeW 
Region Z level Size (cm3Coordinates Z level Size (cm3Coordinates Z level Size (cm3Coordinates 
EST–AIR 
    Periaqueductal gray 4.9 0.6 −4, −38, −28       
    R amygdala + piriform cortex    4.0 4.0 34, −10, −6 3.9 0.8 20, −10, −22 
    L amygdala + piriform + insular + cingulate cortex 4.7 2.0 −22, −6, −16 4.9 7.7 −24, 0, −16 3.9 0.9 −22, −2, −12 
    L inferior temporal cortex 4.7 0.8 −50, −22, 14       
AND–AIR 
    Hypothalamus       4.2 3.2 −8, 0, −10a 
    L amygdala + piriform cortex 4.7 1.8 −22, 8, −10 4.0 3.9 26, 4, −8 4.2  −26, 4, −12a 
    R amygdale + piriform cortex 4.0 0.8 22, 2, −22       
OO–AIR 
    R amygdale + piriform + insular + anterior cingulated cortex 7.0 4.4 20, 0, −16 4.6 3.2 22, −4, −14 7.1 4.8 18, 0, 8 
    L amygdale + piriform + insular cortex 5.6 8.8 −20, −2, −12 4.6 3.3 −20, −4, −14 5.7 5.3 −24, 0, −10 
 CAH women and HeM HeM and HeW CAH women and HeW 
Region Z level Size (cm3Coordinates Z level Size (cm3Coordinates Z level Size (cm3Coordinates 
EST–AIR 
    Periaqueductal gray 4.9 0.6 −4, −38, −28       
    R amygdala + piriform cortex    4.0 4.0 34, −10, −6 3.9 0.8 20, −10, −22 
    L amygdala + piriform + insular + cingulate cortex 4.7 2.0 −22, −6, −16 4.9 7.7 −24, 0, −16 3.9 0.9 −22, −2, −12 
    L inferior temporal cortex 4.7 0.8 −50, −22, 14       
AND–AIR 
    Hypothalamus       4.2 3.2 −8, 0, −10a 
    L amygdala + piriform cortex 4.7 1.8 −22, 8, −10 4.0 3.9 26, 4, −8 4.2  −26, 4, −12a 
    R amygdale + piriform cortex 4.0 0.8 22, 2, −22       
OO–AIR 
    R amygdale + piriform + insular + anterior cingulated cortex 7.0 4.4 20, 0, −16 4.6 3.2 22, −4, −14 7.1 4.8 18, 0, 8 
    L amygdale + piriform + insular cortex 5.6 8.8 −20, −2, −12 4.6 3.3 −20, −4, −14 5.7 5.3 −24, 0, −10 

Note: Clusters calculated at P =0.001, uncorrected P <0.05. Talairach's coordinates indicate local maxima.

a

Same cluster.

Figure 3.

Group-sharing activations. Shown are conjunctional clusters in different groups of subjects, superimposed on the standard brain. All images show horizontal sections at Z = −8 according to Talairach's atlas. The Sokoloff's color scale illustrates Z values (0.0–3.5 for AND and EST and 0.0–5.0 for OO). The OO clusters were large and covered several sections. Because the same brain section is chosen, figures do not always illustrate maximal activation for each condition. Subject's right side is to the left.

Figure 3.

Group-sharing activations. Shown are conjunctional clusters in different groups of subjects, superimposed on the standard brain. All images show horizontal sections at Z = −8 according to Talairach's atlas. The Sokoloff's color scale illustrates Z values (0.0–3.5 for AND and EST and 0.0–5.0 for OO). The OO clusters were large and covered several sections. Because the same brain section is chosen, figures do not always illustrate maximal activation for each condition. Subject's right side is to the left.

Amygdala Connectivity

As reported by Kilpatrick et al., and recently also by our group (Savic and Lindstrom 2008), HeW and HeM differed in their functional connections. First of all, functional connections were in HeW more widespread from the left amygdala, in HeM from the right amygdala (Tables 5, 6). Secondly, in HeW significant connections were displayed with the contralateral amygdala, the anterior cingulate and the hypothalamus, whereas in HeM connections were observed primarily with the ipsilateral piriform cortex, and the ipsilateral caudate and putamen, (Tables 5, 6, Fig. 4). Like the activation pattern, the amygdala connectivity in CAH women was congruent with that of HeW, but not HeM. Thus, connections were more extensive from the left amygdala, and detected primarily in the hypothalamus, the anterior cingulate and subcallosum, and in the contralateral amygdala (including the piriform cortex). CAH women also showed connections between the left amygdala and the precentral gyrus.

Table 5

Significant covariations with the left amygdala

 HeW HeM CAH women 
Region Z level Size (cm3Coordinates Z level Size (cm3Coordinates Z level Size (cm3Coordinates 
L amygdala Inf 9.6 −18, 2, −18a Inf 4.6 −18, 2, −18 Inf 16.0 −18, 2, −18a 
   −26, 2, −2       
R amygdala 4.2 2.0 22, −16, −12b    5.8 4.6 30, 2, −14 
       5.7 3.4 36, −22, 4 
Cingulate 5.7 1.4 −16, 45, −4    4.0 1.2 −30, 14, 2c 
 5.2 2.6 −18, 32, −12    3.6 0.8 −2, 44, −12 
Superior collicle 3.9 2.5 −14, −32, −10       
Right precentral gyrus       5.0 1.6 42, −6, 26 
 HeW HeM CAH women 
Region Z level Size (cm3Coordinates Z level Size (cm3Coordinates Z level Size (cm3Coordinates 
L amygdala Inf 9.6 −18, 2, −18a Inf 4.6 −18, 2, −18 Inf 16.0 −18, 2, −18a 
   −26, 2, −2       
R amygdala 4.2 2.0 22, −16, −12b    5.8 4.6 30, 2, −14 
       5.7 3.4 36, −22, 4 
Cingulate 5.7 1.4 −16, 45, −4    4.0 1.2 −30, 14, 2c 
 5.2 2.6 −18, 32, −12    3.6 0.8 −2, 44, −12 
Superior collicle 3.9 2.5 −14, −32, −10       
Right precentral gyrus       5.0 1.6 42, −6, 26 

Note: R = right; L = left. Clusters detected at T = 3.0, corrected P < 0.05, expressed in Talairach's coordinates. Inf = infinite.

a

Includes the anterior cingulate, subcallosum, and hypothalamus, and a minor portion of the superior temporal gyrus. Covers the hypothalamus, the subcallosum and the anterior cingulate cortex.

b

Covers the piriform cortex.

c

Subcallosum.

Table 6

Significant covariations from the right amygdala

 HeW HeM CAH women 
Region Z level Size (cm3Coordinates Z level Size (cm3Coordinates Z level Size (cm3Coordinates 
R amygdala Inf 5.0 14, 4, −14a Inf 2.4 16, 4, −14b Inf 5.7 18, 20, 6a 
         16, 4, −14 
L amygdala + hippocampus + piriform cortex 5.0 3.4 −26, 4, −14       
L putamen + insular + sensorimotor cortex    3.6 3.0 −38, 0, −8    
L middle temporal gyrus 5.0 2.6 −42, −22, −6 5.0 1.8 −42, −22, −6    
 HeW HeM CAH women 
Region Z level Size (cm3Coordinates Z level Size (cm3Coordinates Z level Size (cm3Coordinates 
R amygdala Inf 5.0 14, 4, −14a Inf 2.4 16, 4, −14b Inf 5.7 18, 20, 6a 
         16, 4, −14 
L amygdala + hippocampus + piriform cortex 5.0 3.4 −26, 4, −14       
L putamen + insular + sensorimotor cortex    3.6 3.0 −38, 0, −8    
L middle temporal gyrus 5.0 2.6 −42, −22, −6 5.0 1.8 −42, −22, −6    

Note: Threshold at T = 3.0 and P < 0.05 uncorrected; italics indicate P < 0.1 uncorrected. Inf = infinite.

a

Includes the anterior cingulate, subcallosum, and hypothalamus, and a minor portion of the superior temporal gyrus.

b

Includes the putamen.

Figure 4.

Covariations with the left and right amygdala seed ROI in the 3 study groups. The Sokoloff's scale indicates T-values. Clusters detected at T = 3.0 and P < 0.05 corrected, are superimposed on the standard MR of the brain.

Figure 4.

Covariations with the left and right amygdala seed ROI in the 3 study groups. The Sokoloff's scale indicates T-values. Clusters detected at T = 3.0 and P < 0.05 corrected, are superimposed on the standard MR of the brain.

Direct group comparisons revealed no difference between CAH patients and HeW. In contrast, in comparison with HeM–CAH women showed significantly more pronounced functional connections with the contralateral amygdala, the piriform and insular cortex, and the anterior cingulate cortex; HeM, on the other hand, displayed more pronounced connections than CAH women in the striatum and the sensorimotor cortex (Tables 7, 8, Fig. 5).

Table 7

Group differences in connectivity pattern, left amygdala

 HeW–HeM CAH women–HeM HeW–CAH women 
Region Z level Size (cm3Coordinates Z level Size (cm3Coordinates Z level Size (cm3Coordinates 
Anterior cingulate (subcallosum) 3.8 2.4 −16, 32, −1 4.9 5.8 −2, 38, 14   No difference 
   22, 0, −18       
R amygdala (+ insular and piriform cortex)    6.7 7.9 20, 2, −18    
    4.8 5.6 38, 4, −2    
    4.7 2.2 34, −16, 0    
R amygdala + hippocampus 3.1 0.3 36, −20, −14       
Cerebellum 3.2 1.4 −23, −50, −24       
 HeM–HeW HeM–CAH women CAH women–HeW 
Superior frontal gyrus 4.0 2.0 0, 22, 52      No difference 
Parietal cortex 4.4 4.4 38, −66, 40       
   56, −26, 13       
Precentral gyrus + putamen + insular cortex    3.9 7.2 −36, 20, 10    
 HeW–HeM CAH women–HeM HeW–CAH women 
Region Z level Size (cm3Coordinates Z level Size (cm3Coordinates Z level Size (cm3Coordinates 
Anterior cingulate (subcallosum) 3.8 2.4 −16, 32, −1 4.9 5.8 −2, 38, 14   No difference 
   22, 0, −18       
R amygdala (+ insular and piriform cortex)    6.7 7.9 20, 2, −18    
    4.8 5.6 38, 4, −2    
    4.7 2.2 34, −16, 0    
R amygdala + hippocampus 3.1 0.3 36, −20, −14       
Cerebellum 3.2 1.4 −23, −50, −24       
 HeM–HeW HeM–CAH women CAH women–HeW 
Superior frontal gyrus 4.0 2.0 0, 22, 52      No difference 
Parietal cortex 4.4 4.4 38, −66, 40       
   56, −26, 13       
Precentral gyrus + putamen + insular cortex    3.9 7.2 −36, 20, 10    

Note: R = right; L = left; Inf = infinite. Clusters detected at T = 3.0, uncorrected P < 0.05 (bold) and corrected P < 0.05 (ordinary text), expressed in Talairach's coordinates.

Table 8

Group differences in connectivity pattern, right amygdala

 HeW–HeM CAH women–HeM HeW–CAH women 
Region Z level Size (cm3Coordinates Z level Size (cm3Coordinates Z level Size (cm3Coordinates 
Anterior cingulated (subcallosum)    4.4 1.5 16, 46, −12   No difference 
      −6, 2, −2    
Hypothalamus + portion of L + R amygdala 3.8 1.6 −6, −26, 2       
 3.3 1.4 −10, 2, 2       
Superior collicle 3.8 1.6 0, −27, 2       
Occippital cortex 4.0 5.4 20, −70, 8       
 HeM–HeW HeM–CAH women CAH women–HeW 
R amygdala      No difference   No difference 
Frontopolar cortex 3.4 1.1 −14, 62, 0       
 HeW–HeM CAH women–HeM HeW–CAH women 
Region Z level Size (cm3Coordinates Z level Size (cm3Coordinates Z level Size (cm3Coordinates 
Anterior cingulated (subcallosum)    4.4 1.5 16, 46, −12   No difference 
      −6, 2, −2    
Hypothalamus + portion of L + R amygdala 3.8 1.6 −6, −26, 2       
 3.3 1.4 −10, 2, 2       
Superior collicle 3.8 1.6 0, −27, 2       
Occippital cortex 4.0 5.4 20, −70, 8       
 HeM–HeW HeM–CAH women CAH women–HeW 
R amygdala      No difference   No difference 
Frontopolar cortex 3.4 1.1 −14, 62, 0       

Note: R = right; L = left; Inf = infinite. Clusters detected at T = 3.0, uncorrected P < 0.05 (bold) and corrected P < 0.05 (ordinary text), expressed in Talairach's coordinates; italics denote values at T = 3.0, P < 0.1 corrected.

Figure 5.

Between-group differences in covariations with the left (a) and right (b) amygdala seed region. The Sokoloff's scale indicates T-values. Clusters detected at T = 3.0 are superimposed on the standard MR of the brain. Brain views illustrating typical group differences are shown. The clusters calculated from HeM to CAH were not significant.

Figure 5.

Between-group differences in covariations with the left (a) and right (b) amygdala seed region. The Sokoloff's scale indicates T-values. Clusters detected at T = 3.0 are superimposed on the standard MR of the brain. Brain views illustrating typical group differences are shown. The clusters calculated from HeM to CAH were not significant.

Psychophysical Data

No significant group differences were detected in odor thresholds (7 × 10−5 M ± 9 × 10−5 [HeM], 3 × 10−5 ± 3 × 10−5 [HeW], and 4 × 10−5 ± 5 × 10−5 [CAH]; P = 0.14, F = 2,07, power 0.39). Neither was there any difference in odor ratings; (P = 0.087 F = 2,5, power 0.48 for familiarity, P = 0.79, F = 0.22, power 0.08 for pleasantness, P = 0.07, F = 2,8, power 0.50 for irritability, and P = 0.07, F = 2.7, power 0.52 for intensity), see figure, in supplemental web information. We did not find any group by stimulus interaction in respiratory amplitude (F = 0.5, P = 0.74), or frequency (F = 1.2, P = 0.32). The baseline respiratory amplitude and frequency is shown in Table, supplemental web information.

Finger Ratios

There was a high correlation between the 2 ways to measure finger ratios, as well as between measures of the 2 raters (r = 0.9; P < 0.001 for both). Evaluation of 2D:4D in the CAH group was based on mean values from the 2 raters. The results presented here were based on direct measurements on the fingers because they are regarded more reliable than those based on photocopies (Manning et al. 1998). There was a significant overall group difference on the right, but not left hand (one-way ANOVA, P = 0.0074, power 0.878, df = 2, F = 6.07 right hand; P = 0.178, power 0.338, F value 1.9, df = 2, left hand). CAH women and HeM had significantly lower ratios than HeW on the right hand (P = 0.003 for HeW vs. HeM, and P = 0.014 for CAH vs. HeW). The right hand's ratio was 0.956 ± 0.024 in CAH women, 0.985 ± 0.016 in HeW, and 0.945 ± 0.011 in HeM. The corresponding values on the left hand were 0.979 ± 0.027 (CAH), 1.005 ± 0.033 (HeW), and 0.973 ± 0.047 (HeM). The ratios in controls were, thus, in accordance with those reported previously (Manning et al. 1998).

Discussion

Despite the genetically verified diagnosis and parental reports about boy-typical play behavior during childhood, the pattern of activation in the presently investigated CAH women was remarkably similar to that of female controls, and different from the pattern of male controls. CAH women and HeW activated the anterior hypothalamus with AND, whereas HeM activated this region with EST. Furthermore, whilst the amygdala connectivity differed between the male and female controls, no difference was observed between control females and CAH females. Thus, both with respect to aspects of functional organization and functional activation of the limbic circuits CAH women showed a pattern congruent with their biological sex, and different from the opposite sex. Our hypothesis that these specific aspects of cerebral dimorphism would have masculine features in CAH women was thereby rejected.

The methodology applied has been discussed in detail in several of our previous publications, and will not be further commented here. With the present group sizes this method is regarded to provide data, which are representative for the entire group (Friston et al. 1999; Friston and Penny 2003). Application of the same methodology in study groups of similar size has in several previous investigations shown significant group differences, with for example, a sex atypical pattern in homosexual men and lesbian women (Berglund et al. 2006; Hillert et al. 2007; Savic 2002; Savic et al. 2005). Taken together, these arguments suggest that the present data have enough power for the results to be reliable, and that the failure to detect signs of masculinization in CAH women cannot be attributed to a methodological bias. Neither can it easily be attributed to patient selection. The investigated CAH women had a clear CYP21 mutation (and 21-OH deficiency), which in 9 was moderate or severe.

M Hines proposed that prenatal testosterone seems to have little or no influence on mental rotations ability, although this brain function shows a marked sex difference (Hines 2006). In a large epidemiological survey Hines and col. also found that the increase in prevalence of male typed childhood play behavior in CAH girls was much more pronounced than the increase in homosexual orientation (Hines et al. 2004), indicating that testosterone-related influences on the phenotype may vary from one feature to another. The presently observed discrepancies, with the masculine finger ratios, the reported tom-boy behavior during childhood and the diagnosed genital androgenization as opposed to the female patterns of cerebral activation and connectivity, is in accordance with this notion.

Explanations to these discrepancies are not evident from the present data. We can only conclude that intrauterine virilization of genitalia is not necessarily paralleled by a masculinization of the limbic brain, at least not with respect to signal response to AND and EST, and the baseline amygdala connectivity, which are 2 indices of sex-dimorphism. It is theoretically possible that various sex dimorphic features are affected by fetal testosterone in a dose dependent manner. Whilst such a scenario could be attributed for the differences between HeM and HeW (with extremely high testosterone levels in male fetuses), it is less likely to explain the “male” like AND and EST activation and functional connectivity in lesbian women described in our previous studies (Berglund et al. 2006; Savic and Lindstrom 2008). None of our lesbian participants in these studies had genital masculinization, which is expected already at moderate elevations of fetal testosterone. An alternative possibility is that various sex dimorphic features may have different etiological factors; in this respect recent studies by Arnolds group at UCLA are of particular interest as they indicate existence of early, and testosterone-independent chromosomal effects on the brain (Arnold 2004). Finally, several different etiological factors could contribute to a same sexually dimorphic cerebral feature, for example, psychosexual outcome. The 3 alternatives are not mutually exclusive. Although presently speculative, in the view of present results they all seem relevant to address in the near future.

Supplementary Material

Supplementary material can be found at: http://www.cercor.oxfordjournals.org/

Funding

The Swedish Medical Research Council; and The Center for Gender Medicine at Karolinska Institute.

We thank Erik Samen for tracer synthesis, Johan Mohlin and Julio Gabriel for technical assistance, and Peter Fransson for methodological advice. We specially acknowledge Dr Per Lindstrom and Dr Hans Berglund for injecting the tracer during experiments. Conflict of Interest: None declared.

References

Arnold
AP
Sex chromosomes and brain gender
Nat Rev Neurosci.
 , 
2004
, vol. 
5
 (pg. 
701
-
708
)
Berenbaum
SA
Effects of early androgens on sex-typed activities and interests in adolescents with congenital adrenal hyperplasia
Horm Behav.
 , 
1999
, vol. 
35
 (pg. 
102
-
110
)
Berenbaum
SA
Resnick
SM
Early androgen effects on aggression in children and adults with congenital adrenal hyperplasia
Psychoneuroendocrinology.
 , 
1997
, vol. 
22
 (pg. 
505
-
515
)
Berglund
H
Lindstrom
P
Dhejne-Helmy
C
Savic
I
Male-to-female transsexuals show sex-atypical hypothalamus activation when smelling odorous steroids
Cereb Cortex.
 , 
2007
, vol. 
18
 (pg. 
1900
-
1908
)
Berglund
H
Lindstrom
P
Savic
I
Brain response to putative pheromones in lesbian women
Proc Natl Acad Sci USA.
 , 
2006
, vol. 
103
 (pg. 
8269
-
8274
)
Buck
JJ
Williams
RM
Hughes
IA
Acerini
CL
In-utero androgen exposure and 2nd to 4th digit length ratio-comparisons between healthy controls and females with classical congenital adrenal hyperplasia
Hum Reprod.
 , 
2003
, vol. 
18
 (pg. 
976
-
979
)
Cahill
L
Why sex matters for neuroscience
Nat Rev Neurosci.
 , 
2006
, vol. 
7
 (pg. 
477
-
484
)
Ciumas
C
Lindstrom
P
Aoun
B
Savic
I
Imaging of odor perception delineates functional disintegration of the limbic circuits in mesial temporal lobe epilepsy
Neuroimage.
 , 
2008
, vol. 
39
 (pg. 
578
-
592
)
Dancey
CP
Sexual orientation in women: an investigation of hormonal and personality variables
Biol Psychol.
 , 
1990
, vol. 
30
 (pg. 
251
-
264
)
De Vries
GJ
Rissman
EF
Simerly
RB
Yang
LY
Scordalakes
EM
Auger
CJ
Swain
A
Lovell-Badge
R
Burgoyne
PS
Arnold
AP
A model system for study of sex chromosome effects on sexually dimorphic neural and behavioral traits
J Neurosci.
 , 
2002
, vol. 
22
 (pg. 
9005
-
9014
)
Dittmann
RW
Kappes
ME
Kappes
MH
Sexual behavior in adolescent and adult females with congenital adrenal hyperplasia
Psychoneuroendocrinology.
 , 
1992
, vol. 
17
 (pg. 
153
-
170
)
Ehrhardt
AA
Epstein
R
Money
J
Fetal androgens and female gender identity in the early-treated adrenogenital syndrome
Johns Hopkins Med J.
 , 
1968
, vol. 
122
 (pg. 
160
-
167
)
Friston
KJ
Holmes
AP
Worsley
KJ
How many subjects constitute a study?
Neuroimage.
 , 
1999
, vol. 
10
 (pg. 
1
-
5
)
Friston
KJ
Penny
W
Posterior probability maps and SPMs
Neuroimage.
 , 
2003
, vol. 
19
 (pg. 
1240
-
1249
)
Hall
CM
Jones
JA
Meyer-Bahlburg
HF
Dolezal
C
Coleman
M
Foster
P
Price
DA
Clayton
PE
Behavioral and physical masculinization are related to genotype in girls with congenital adrenal hyperplasia
J Clin Endocrinol Metab.
 , 
2004
, vol. 
89
 (pg. 
419
-
424
)
Hampson
E
Rovet
J
Altmann
D
Spatial reasoning in children with congenital adrenal hyperplasia due to 21-hydroxylase deficiency
Dev Neuropsychol.
 , 
1998
, vol. 
14
 (pg. 
299
-
320
)
Helleday
J
Bartfai
A
Ritzen
EM
Forsman
M
General intelligence and cognitive profile in women with congenital adrenal hyperplasia (CAH)
Psychoneuroendocrinology.
 , 
1994
, vol. 
19
 (pg. 
343
-
356
)
Hillert
L
Musabasic
V
Berglund
H
Ciumas
C
Savic
I
Odor processing in multiple chemical sensitivity
Hum Brain Mapp.
 , 
2007
, vol. 
28
 (pg. 
172
-
182
)
Hines
M
Prenatal testosterone and gender-related behaviour
Eur J Endocrinol.
 , 
2006
, vol. 
155
 
Suppl. 1
(pg. 
S115
-
S121
)
Hines
M
Brook
C
Conway
GS
Androgen and psychosexual development: core gender identity, sexual orientation and recalled childhood gender role behavior in women and men with congenital adrenal hyperplasia (CAH)
J Sex Res.
 , 
2004
, vol. 
41
 (pg. 
75
-
81
)
Kilpatrick
LA
Zald
DH
Pardo
JV
Cahill
LF
Sex-related differences in amygdala functional connectivity during resting conditions
Neuroimage.
 , 
2006
, vol. 
30
 (pg. 
452
-
461
)
Kinsey
AC
Pomeroy
WR
Martin
CE
Sexual behavior in the human male. 1948
Am J Public Health.
 , 
2003
, vol. 
93
 (pg. 
894
-
898
)
MacCulloch
MJ
Waddington
JL
Neuroendocrine mechanisms and the aetiology of male and female homosexuality
Br J Psychiatry.
 , 
1981
, vol. 
139
 (pg. 
341
-
345
)
Manning
JT
The ratio of 2nd to 4th digit length and performance in skiing
J Sports Med Phys Fitness.
 , 
2002
, vol. 
42
 (pg. 
446
-
450
)
Manning
JT
Scutt
D
Wilson
J
Lewis-Jones
DI
The ratio of 2nd to 4th digit length: a predictor of sperm numbers and concentrations of testosterone, luteinizing hormone and oestrogen
Hum Reprod.
 , 
1998
, vol. 
13
 (pg. 
3000
-
3004
)
McGuire
LS
Ryan
KO
Omenn
GS
Congenital adrenal hyperplasia. II. Cognitive and behavioral studies
Behav Genet.
 , 
1975
, vol. 
5
 (pg. 
175
-
188
)
Merke
DP
Bornstein
SR
Congenital adrenal hyperplasia
Lancet.
 , 
2005
, vol. 
365
 (pg. 
2125
-
2136
)
Meyer-Bahlburg
HF
Gender and sexuality in classic congenital adrenal hyperplasia
Endocrinol Metab Clin N Am.
 , 
2001
, vol. 
30
 (pg. 
155
-
171
viii
Money
J
Schwartz
M
Lewis
VG
Adult erotosexual status and fetal hormonal masculinization and demasculinization: 46,XX congenital virilizing adrenal hyperplasia and 46,XY androgen-insensitivity syndrome compared
Psychoneuroendocrinology.
 , 
1984
, vol. 
9
 (pg. 
405
-
414
)
Negri-Cesi
P
Colciago
A
Celotti
F
Motta
M
Sexual differentiation of the brain: role of testosterone and its active metabolites
J Endocrinol Invest.
 , 
2004
, vol. 
27
 (pg. 
120
-
127
)
Nordenstrom
A
Servin
A
Bohlin
G
Larsson
A
Wedell
A
Sex-typed toy play behavior correlates with the degree of prenatal androgen exposure assessed by CYP21 genotype in girls with congenital adrenal hyperplasia
J Clin Endocrinol Metab.
 , 
2002
, vol. 
87
 (pg. 
5119
-
5124
)
Oh
JS
Song
IC
Lee
JS
Kang
H
Park
KS
Kang
E
Lee
DS
Tractography-guided statistics (TGIS) in diffusion tensor imaging for the detection of gender difference of fiber integrity in the midsagittal and parasagittal corpora callosa
Neuroimage.
 , 
2007
, vol. 
36
 (pg. 
606
-
616
)
Oldfield
RC
The assessment and analysis of handedness: the Edinburgh inventory
Neuropsychologia.
 , 
1971
, vol. 
9
 (pg. 
97
-
113
)
Resnick
SM
Berenbaum
SA
Gottesman
II
Bouchard
T
Early hormonal influences on cognitive functioning in congenital adrenal hyperplasia
Dev Psychol.
 , 
1986
, vol. 
22
 (pg. 
191
-
198
)
Savic
I
Brain imaging studies of the functional organization of human olfaction
Neuroscientist.
 , 
2002
, vol. 
8
 (pg. 
204
-
211
)
Savic
I
Berglund
H
Gulyas
B
Roland
P
Smelling of odorous sex hormone-like compounds causes sex-differentiated hypothalamic activations in humans
Neuron.
 , 
2001
, vol. 
31
 (pg. 
661
-
668
)
Savic
I
Berglund
H
Lindstrom
P
Brain response to putative pheromones in homosexual men
Proc Natl Acad Sci USA.
 , 
2005
, vol. 
102
 (pg. 
7356
-
7361
)
Savic
I
Gulyas
B
Larsson
M
Roland
P
Olfactory functions are mediated by parallel and hierarchical processing
Neuron.
 , 
2000
, vol. 
26
 (pg. 
735
-
745
)
Savic
I
Lindstrom
P
PET and MRI show differences in cerebral asymmetry and functional connectivity between homo- and heterosexual subjects
Proc Natl Acad Sci USA.
 , 
2008
, vol. 
105
 (pg. 
9403
-
9408
)
Schaltenbrand
G
Walker
AE
Andy
OJ
Schaltenbrand
G
Schaltenbrand
G
Stereotaxy of the human brain: anatomical, physiological and clinical applications
 , 
1982
Stuttgart (Germany)
G. Thieme Verlag, Thieme-Stratten
Wedell
A
Luthman
H
Steroid 21-hydroxylase deficiency: two additional mutations in salt-wasting disease and rapid screening of disease-causing mutations
Hum Mol Genet.
 , 
1993
, vol. 
2
 (pg. 
499
-
504
)
Zucker
KJ
Bradley
SJ
Oliver
G
Blake
J
Fleming
S
Hood
J
Psychosexual development of women with congenital adrenal hyperplasia
Horm Behav.
 , 
1996
, vol. 
30
 (pg. 
300
-
318
)