Series

Women’s mental health 1

Oestrogens, prolactin, hypothalamic-pituitary-gonadal axis,
and schizophrenic psychoses
Anita Riecher-Rössler

Interest is growing in the potential effect of gonadal hormones, prolactin, and the hypothalamic–pituitary–gonadal Lancet Psychiatry 2016
axis in schizophrenic psychoses. Many studies from clinical, epidemiological, and fundamental research have Published Online
confirmed that oestradiol, the main component of oestrogens, can have protective effects in schizophrenic November 14, 2016
http://dx.doi.org/10.1016/
psychoses. Furthermore, many patients with schizophrenic psychoses—even in the untreated prodromal stages—
S2215-0366(16)30379-0
have hyperprolactinaemia and gonadal dysfunction, with oestrogen deficiency in women and testosterone deficiency
See Online/Comment
in men. The understanding of the pathogenetic mechanisms underlying these findings could contribute to a better http://dx.doi.org/10.1016/
understanding of the aetiopathogenesis of schizophrenic psychoses and improve therapeutic approaches. In this S2215-0366(16)30348-0
Series paper, we aim to review methodologically sound studies in this area, propose a theory to explain these This is the first of a Series of
findings in the context of psychosis, and suggest therapeutic strategies and implications for further research. four papers on Women’s mental
health

Introduction The oestrogen protection hypothesis Center for Gender Research and
Early Detection, University of
Schizophrenic psychoses still pose many challenges to Epidemiological and clinical studies Basel Psychiatric Clinics, Basel,
psychiatry. We do not yet fully understand the The hypothesis of oestrogens being protective against Switzerland
aetiopathogenetic mechanisms of these disorders, nor psychosis, first mentioned in the middle of the past (Prof A Riecher-Rössler MD)
can we sufficiently treat them, especially with regards century,1,2 has since gained support from many Correspondence to:
to cognitive and negative symptoms. The cognitive epidemiological and clinical studies. Prof Anita Riecher-Rössler,
Center for Gender Research and
symptoms are often the main obstacles to a full Oestrogen withdrawal can cause psychosis in formerly Early Detection, University of
recovery. Outcomes of research on oestrogens, healthy individuals, known as premenstrual psychosis, Basel Psychiatric Clinics,
prolactin, and the hypothalamic–pituitary–gonadal axis post-abortion psychosis, and psychoses associated with Basel CH-4051, Switzerland
might give us some important insights and offer the removal of a hydatidiform mole, discontinuation of anita.riecher@upkbs.ch

therapeutic possibilities. oral contraceptives, administration of clomifene or

On the one hand, numerous clinical, epidemiological, tamoxifen (both oestrogen receptor antagonists), or a
and fundamental research studies have shown that gonadorelin agonist (which blocks pituitary stimulation
oestradiol, the main component of oestrogens, has of endogenous oestrogen secretion). These psychotic
protective effects in schizophrenic psychoses.1–4 On the episodes are typically acute, short, and show an extensive
other hand, gonadal dysfunction, hyperprolactinaemia, range of psychotic and affective symptoms. Patients
and states of oestrogen deficiency in women with sometimes have a history of post-partum psychosis (ie,
schizophrenia have been reported.1–4 These observations women seem to be sensitive to withdrawal of oestrogens)
led us to formulate the hypotheses of (1) oestrogen and usually remit with oestrogen substitution.10
protection and (2) hypo-oestrogenism and gonadal With regards to schizophrenic psychoses, clear sex
dysfunction in schizophrenic psychoses.1,2 differences might partly be attributed to the influence of
Growing evidence for hyperprolactinaemia has also sex hormones. Thus, women seem to have a slightly
emerged in many patients with psychosis who are lower incidence of schizophrenic psychoses than men.11
antipsychotic-naive and even in individuals with an The incidence is particularly low in women younger than
at-risk mental state.5–9 Since prolactin suppresses the 40 years, corresponding to a delayed age of onset, but
production of gonadal hormones such as oestrogens and higher thereafter.11–13 Furthermore, the course of the
testosterone, this hyperprolactinaemia could explain disease is more favourable in younger women than in
these hormonal deficiencies. Furthermore, hyper- those older than 40 years.12,13 The ABC study14,15 was an
prolactinaemia could contribute to explaining how stress epidemiological study on a representative sample of
can trigger the outbreak of psychosis in vulnerable 392 first-time admitted patients with schizophrenia. We
individuals, as formulated in the stress–prolactin– found that women had a later peak of illness onset than
dopamine hypothesis.5 did men. Women also exhibited an additional, smaller
This Series paper will give an update on studies peak of illness onset after age 45 years (figure 1). After we
investigating three hypotheses (oestrogen protection, excluded psychosocial factors as explanations, we
hypo-oestrogenism and hypothalamic–pituitary–gonadal suggested that the production of physiologically high
dysfunction, and stress–prolactin–dopamine) with a focus oestradiol in young and fertile women delays the onset
on clinical studies. Implications for the clinic and for and improves the course of schizophrenia in women up
research will be described. until menopause. Then, before and around menopause

www.thelancet.com/psychiatry Published online November 14, 2016 http://dx.doi.org/10.1016/S2215-0366(16)30379-0 1

 Series

50·0 Men (n=187)

Women (n=205) 46·1
45·0

40·0
Incidence per year per 100 000

35·4
35·0
31·2 31·4
30·0 28·2

25·0 22·7 22·4

19·6
20·0

14·1 14·8 14·7

15·0 12·5 11·6
9·8
10·0 7·7
6·8
5·0 4·4
2·8
0·0 0·0 0·0
12–14 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59
Age groups, years

Figure 1: Sex-specific distribution at first admission for schizophrenia—broad definition, ICD-9: 295, 297, 298.3, 298.4
Data are all admissions in a defined catchment area in and around Mannheim, Germany, in 1987 and 1988. Data compiled from Häfner and colleagues.14

the drop of oestrogens leads to the second peak of onsets the perimenstrual low-oestrogen phase of the cycle, but
in women and to the poorer course of the disease in also an inverse correlation of oestradiol concentrations in
elderly women.12,13,15 the blood with psychopathology.23,24 When oestradiol
Studies of late-onset schizophrenia are consistent with concentrations in the blood increased naturally during
these fingings.12,13,16 We have shown that there are twice as the cycle, there appeared to be an improvement in
many women as men with onset of schizophrenic psychotic and total symptomatology, and vice versa.
psychoses beyond age 40 years—8·9 per 100 000 women Many other investigators have also reported a
per year, but only 4·2 per 100 000 men per year in the age significant association between the menstrual cycle
group 40–59 years.12 Furthermore, symptomatology and phase and 17-β-oestradiol concentrations on cognition
disease course are unexpectedly severe in women with and on positive and negative symptoms.3,4
this late onset.12,13,16 Similarly, results from our long-term However, effects of oestrogens might not be specific
studies on earlier onset schizophrenia (before age for psychosis. For example, an increased number of
40 years) showed that in women the course of illness patient admissions during the perimenstrual period
tends to deteriorate rapidly after menopause.1,17 has been found independent of diagnosis for other
There might be different subgroups of schizophrenic disorders.3 In patients with schizophrenia, an
psychoses, not all of which are affected by oestrogens. exacerbation of many psychiatric symptoms, not only
Hence, the sex difference in the age of onset seems psychotic symptoms, has been observed during the
smaller in the subgroup of patients with a genetic risk or perimenstrual period.3 This lack of specificity is
perinatal complications than in patients without these expected because of the multiple effects of oestrogens
risk factors.18 on mental functioning.
Studies on the age at menarche also suggest a protective
effect of oestrogens. Thus, later menarche—ie, later rise Oestrogen effects in the brain
of oestrogen levels—seems to be associated with an The aetiopathogenesis of schizophrenic psychoses is
earlier onset of the disorder.19,20 In line with this, we widely considered to involve mainly neurodevelopmental
found a significantly later age of menarche in women abnormalities, but also neurodegenerative changes in
with schizophrenic psychoses than in healthy controls.21 the brain, as well as neurotransmitter dysfunctions.
Clinically, psychotic symptomatology in women often Oestrogens seem to be able to influence all these domains.
correlates with the oestrogenic state.1–3 Thus, chronic In animals, oestrogens and testosterone strongly
psychoses appear to improve during pregnancy, when affect brain development during late gestation, in the
oestrogen concentrations are increased by about early postnatal period, and around puberty. Along with
100 times to 200 times, whereas after delivery, which is others, they affect sexual dimorphism—ie, structural
accompanied by a sudden drop of oestrogens, there is differences between normal male and female
marked excess of psychoses.22 brains.25,26 This dimorphism might be disrupted in
In patients with schizophrenia, psychotic symptoms schizophrenic psychoses.27,28
often deteriorate premenstrually or perimenstrually (ie, Additionally, oestrogens have many neuroprotective
in the low-oestrogen phase of the cycle).3,4,21 We showed effects. The most active natural oestrogen in the
not only a significant excess of patient admissions during brain, 17-β-oestradiol, promotes neuronal sprouting and

2 www.thelancet.com/psychiatry Published online November 14, 2016 http://dx.doi.org/10.1016/S2215-0366(16)30379-0

 Series

myelination, enhances synaptic density and plasticity, suggesting that oestrogen treatment can ameliorate
facilitates neuronal connectivity, acts as an anti- extrapyramidal side-effects.3 Louza and colleagues52 did
inflammatory and as an antioxidant, inhibits neuronal not report a positive response to oestrogen treatment in
cell death, and improves cerebral blood flow and glucose their study, which might be because of their use of
metabolism.17,25,26,29–31 Furthermore, oestrogens might conjugated oestrogens rather than 17-β-oestradiol (the
mediate BDNF expression and activity.32 most active oestrogen in the brain).
Circulating oestrogens also modulate the dopaminergic In 2005, on the basis of the results of a Cochrane review53
and other neurotransmitter systems relevant to it was concluded that the effects of oestrogens were still
schizophrenic psychoses, such as the serotonergic, gluta- unclear and that further, larger clinical trials were needed.
matergic, noradrenergic, and cholinergic systems,2,33,34 However, the studies reviewed that had negative results
and can modulate the sensitivity and number of did not use 17-β-oestradiol, but used conjugated oestrogens.
dopamine receptors.2,33 Some authors have even Furthermore, to prevent endometrial hyperplasia, the
suggested that 17-β-oestradiol in the brain should be oestrogens were usually combined with different
regarded as a neurotransmitter itself.35 progestogens, which unfortunately can counteract the
In different animal models, oestrogens show similar positive effects of oestradiol in the brain.4 Since the
effects on animal behaviour to those of antipsychotics.1,2 Cochrane review, two quantitative reviews on oestrogen
They also positively influence many pathologies augmentation, which included new randomised trials and
associated with schizophrenia, such as disturbances of applied stricter inclusion criteria, reported superior
latent inhibition, prepulse inhibition, auditory processing efficacy when antipsychotics were augmented with
or mismatch negativity, and selective attention.4,36–38 oestrogens as compared with antipsychotics alone. These
Oestrogens not only act via the classic genomic path- effects were even larger when oestradiol was used in
way, but also show non-genomic, rapid interactions.29,39 At addition to antipsychotics.43,44 Additionally, in a systematic
least two subtypes of oestrogen receptors are known, review of 26 randomised trials on the efficacy of anti-
namely oestrogen receptor-α and oestrogen receptor-β.40 inflammatory drugs to improve symptoms in patients with
These receptors are expressed in several areas of the schizophrenia, Sommer and colleagues45 found oestrogens
human brain that are associated with neuroendocrine to be one of three promising drugs, acknowledging that
function and with emotion, memory, and cognition.29 oestrogens not only have anti-inflammatory, but also have
Variations in the oestrogen receptor-α (ESR1) have been other properties.
associated with schizophrenia.41,42 Some authors suggested Table 1 shows an overview on the randomised, double-
that the brain response to oestrogens in patients with blind, placebo-controlled trials on oestrogen augmen-
schizophrenia might be inadequate.4,31 More details on tation in schizophrenic psychoses.46–49,51,52,54
recent molecular and preclinical findings are given in a Overall, the oestrogen protection hypothesis is
comprehensive review by Gogos and colleagues.34 supported by these studies. In particular, the addition of
transdermally delivered 17-β-oestradiol to standard
Intervention studies with oestrogens antipsychotics seems to be associated with a significant
Several intervention studies with oestrogens, especially reduction of psychotic and other, particularly depressive,
with oestradiol, have shown positive results in women symptoms and potentially also with cognitive
with schizophrenic psychoses.1,3,43–45 improvements in women with schizophrenic psychoses.
Kulkarni and colleagues3,46–49 were the first to conduct The greatest effect of oestrogen treatment would be
systematic trials. In different studies, they found that expected to happen in postmenopausal women, who have
women with schizophrenia who received oestradiol as oestrogen deficiency. However, most studies of oestrogen
an adjunct to antipsychotics showed a more rapid and treatment have been done in young women. One
greater symptom improvement than did women taking exception is a small study by Good and colleagues55 who
antipsychotics alone. Patients receiving an oestradiol examined postmenopausal women with schizophrenia.
transdermal adjunct significantly improved in their After receiving hormone replacement therapy (HRT)
total positive, negative, and general symptoms,46,47,49 and for 6 months, these women showed a significant
also in cognition in one study.48 Cognitive improvement improvement of verbal memory compared with women
would be in line with studies suggesting a positive who did not receive HRT. Similarly, the results of a
effect of 17-β-oestradiol on cognitive function in healthy community study of postmenopausal women with
ageing women,30,50 but could not be confirmed in their schizophrenia who received HRT for gynaecological
last study.49 reasons showed less severe negative symptoms and these
Women with schizophrenia in the study by patients needed lower doses of antipsychotics than women
Akhondzadeh and colleagues51 achieved positive effects with the same diagnosis but who did not receive HRT.56
also with ethinyl-oestradiol as an adjunct to haloperidol. In women with post-partum psychosis and sustained
Furthermore, these patients needed significantly less oestrogen deficiency, the substitution of 17-β-oestradiol
anticholinergic medication than did patients who received produced a substantial antipsychotic effect within 1 week,
haloperidol alone, which is in line with other studies, without antipsychotics.57

www.thelancet.com/psychiatry Published online November 14, 2016 http://dx.doi.org/10.1016/S2215-0366(16)30379-0 3

 Series

Population Type of Number of Number of Weeks of Daily Outcome Results

oestrogen patients given patients treatment augmentation measures
oestrogen given dose
placebo
Kulkarni et al DSM-IV schizophrenia, Two groups 24 (12 in 50 μg 12 4 50 μg and PANSS Significant improvement of positive
(2001)46* schizophreniform, or receiving group and 12 in 100 μg (both psychotic and general symptoms in both
schizoaffective disorder, active oestradiol 100 μg group) transdermal) groups; in the 100 μg group also significant
phase of illness improvement of negative symptoms
Kulkarni et al DSM-IV schizophrenia, Oestradiol 56 46 4 100 μg PANSS Significant improvement of positive
(2008)47 and schizoaffective, or (transdermal) psychotic and general psychopathological
Kulkarni schizophreniform disorder, symptoms; significant improvement of
(2009)48 acute or chronic phase cognition
Kulkarni et al DSM-IV TR schizophrenia, Two groups 56 for 100 μg 62 8 100 μg and PANSS, cognitive Improvement of positive in particular, but
(2015)49* schizoaffective or receiving group and 62 for 200 μg (both tests also of general and total symptoms
schizophreniform disorder, oestradiol 200 μg group transdermal) (stronger effect with 200 μg); no effect
treatment resistant on cognition
Akhondzadeh DSM-IV chronic schizophrenia, Ethinyl 16 16 8 50 μg (oral) PANSS Significant improvement of positive,
et al (2003)51 active phase of illness oestradiol negative, and general symptoms
Louza et al DSM-IV schizophrenia, active Conjugated 21 19 4 625 μg (oral) Brief Psychiatric No response
(2004)52 phase of illness oestrogens Rating Scale,
Negative Symptom
Rating Scale
Ghafari et al DSM-IV chronic schizophrenia Conjugated 16 16 4 625 μg (oral) PANSS Significant decrease in positive, negative,
(2013)54 oestrogens general, and total PANSS scores

PANSS=Positive and Negative Syndrome Scale. *Administered two different doses in different subsamples.

Table 1: Randomised, double-blind, placebo-controlled trials on oestrogen augmentation in premenopausal women with schizophrenic psychoses

By contrast with studies of women with schizophrenic antipsychotics. However, increased prolactin concen-
psychoses, there are only a few small studies of male trations or hyperprolactinaemia have been shown in
patients. In men, adjunctive oestradiol showed a patients who are antipsychotic-naive with first-episode
significant reduction of psychotic symptoms compared psychosis or who have an at-risk mental state for
with placebo.3 However, long-term application is not psychosis, especially in female patients.5–9
possible because of the risk of feminisation. Testosterone Furthermore, hypo-oestrogenism was observed in
augmentation improved negative, but not positive, people with psychosis long before the introduction of
symptoms in one study.58 antipsychotics,1,2 and oestradiol concentrations below
normal were found irrespective of medication type
The hypothesis of hypo-oestrogenism and or prolactin status in women61,66 and men61 with
hypothalamic–pituitary–gonadal dysfunction schizophrenia. Additionally, in patients with schizophrenia
Outcomes of many studies have in the meantime who were antipsychotic-free, low concentrations of
confirmed earlier findings of oestrogen deficiency oestrogens and testosterone have been shown;60 low
and disturbed gonadal function in women with concentrations of testosterone were even present in
schizophrenia.1,3–5,25,59 These studies reported menstrual adolescent boys with prodromal symptoms.62 Women
irregularities and reduced concentrations of oestradiol, with first-episode psychosis, who were retrospectively
progesterone, and gonadotropins (follicle-stimulating interviewed regarding clinical signs of gonadal
hormone, luteinising hormone) in the blood throughout dysfunction before the onset of the disease, have reported
the menstrual cycle, as well as anovulation in many such signs as mid-cycle bleedings, loss of hair, hirsutism,
women with schizophrenic psychoses. and infertility more often than a control group of
Furthermore, in men with (emerging) psychosis, age-matched healthy women.21
blood levels of oestradiol44,59–61 and testosterone44,59–63 In accordance with this, reduced bone mineral density,
seem to be decreased. Fertility seems to be reduced in one of the long-term consequences of gonadal
both sexes.64 suppression,4 was found in women with first-episode
One of the main reasons for gonadal dysfunction and psychosis, although they had only minimal previous
hypo-oestrogenism is probably hyperprolactinaemia; exposure to antipsychotics.67
increased prolactin blood levels have been found in many Of utmost interest in the context of hyperprolactinaemia
patients with schizophrenia5 and it is well known that is the finding of enlarged pituitary volumes in patients
hyperprolactinaemia can suppress gonadal function.4,65 who were antipsychotic-free with first-episode psychosis
Traditionally, hyperprolactinaemia in patients with or who were in an at-risk mental state, especially in those
psychosis has been interpreted as a side-effect of with a later transition to psychosis.68 This finding might be

4 www.thelancet.com/psychiatry Published online November 14, 2016 http://dx.doi.org/10.1016/S2215-0366(16)30379-0

 Series

a further indication of an increased prolactin production

during (emerging) psychosis, since an increase of Panel 1: Potential consequences of hyperprolactinaemia with gonadal suppression17,61
prolactin production can enlarge the pituitary gland.69 • Galactorrhoea, mammary engorgement, possibly breast cancer
All in all, there are many indications of a dysfunction of • Gonadal suppression with decline of oestrogen production in women and decline of
the hypothalamic–pituitary–gonadal axis with hyper- testosterone production in men
prolactinaemia and deficient sex hormone production - Early consequences: loss of libido, orgasmic dysfunction (erectile dysfunction,
in—at least a subgroup of—patients with schizophrenic disturbance of ejaculation, reduced spermatogenesis in men); reduced fertility; hot
psychoses. flushes, urogenital symptoms, dyspareunia in women; possibly depression;
emotional lability and risk of relapse
Implications for the clinic - Late consequences: osteopenia, osteoporosis, fragility fractures; cardiovascular risk,
Assessment and therapy of hyperprolactinaemia and skin ageing, possibly cognitive disturbance
gonadal dysfunction
In the future, oestrogens, prolactin, and the gonadal axis
should be taken more seriously in the assessment and these findings in large controlled studies by different
treatment of patients with schizophrenic psychoses. groups are urgently awaited.
Psychiatric history assessment should always include In women with frequent perimenstrual relapses,
questions on galactorrhoea, sexual problems, and continuous use of oral contraceptives without hormone-
infertility in both sexes, and menstrual irregularities and free intervals might be an option.21
amenorrhoea in women. If any of these symptoms or In women with declining oestrogen production during
other clinical signs (panel 1)17,61 are present (premature) perimenopause and postmenopause, hormonal replace-
gonadal insufficiency should be suspected. Gonadal ment seems promising. It might augment the effect of
insufficiency is due to hyperprolactinaemia in many the antipsychotics and thus minimise their dosage, and
cases, which might be a side-effect of certain simultaneously reduce perimenopausal complaints such
antipsychotics but can occur independently of as hot flushes, night sweats with sleep disturbances, and
medication. Prolactin serum concentrations should general irritability (panel 2),16,30,71–77 thereby contributing
therefore be tested in patients with these symptoms. to general wellbeing and possibly to prevention of
Routine testing of prolactin has even been suggested in psychotic relapses. Furthermore, other positive effects of
all patients with psychosis by some authors, preferably oestrogens, such as a positive influence on cognitive
before being given antipsychotics70 because gonadal function30,50 and on bone mineral density,67,78 could be
dysfunction can often be found even in women who especially helpful in patients with psychosis of this age
are menstruating24,70 and hyperprolactinaemia might group. Despite these potentially helpful aspects, women
be underdiagnosed.65 with schizophrenia have so far been less likely to use
In patients with schizophrenia, these concomitant HRT than mentally healthy women.79
risks of hyperprolactinaemia and gonadal dysfunction Oestrogen replacement in perimenopause and
are often further increased by additional risk postmenopause has been challenged by several
factors such as poor diet, smoking, and not enough studies,80,81 especially by the Women’s Mental Health
regular exercise.40,65,70 Initiative (WHI) study,80 because of suspected side-effects.
In the case of hyperprolactinaemia, prolactin-sparing However, the WHI study80 was criticised by many
antipsychotics (eg, clozapine, quetiapine, aripiprazole, experts82–84 for method flaws, particularly the high age of
or olanzapine) should be the preferred pharmacological the women studied (mean age 63 years) and their high
treatment option.65 Before switching to prolactin- prevalence of cardiovascular risk factors. Thus, many of
sparing antipsychotics, contraception counselling is the complications reported were probably associated
mandatory since gonadal function often normalises in with pre-existing risk factors. A reanalysis74 of the WHI
women with pre-existing gonadal suppression and data could not confirm the complications, but rather
fertility is regained, with a high risk of unplanned showed a cardiovascular benefit when oestrogen
pregnancy.17 If switching is not possible and if the replacement was started early after menopause. There
patient has oestrogen deficiency, the hormone should seems to be a window of opportunity for starting
be substituted in cooperation with a specialist.71 The hormone replacement.72,74 Furthermore, the WHI study80
same might be true for testosterone, although much had used continuous conjugated equine oestrogen rather
less research has been done on this. Furthermore, there than physiological oestradiol, although the latter is
might be cases with premature gonadal insufficiency known to have less side-effects. The patients were given
without hyperprolactinaemia. the drug orally rather than transdermally as
recommended. It was also combined with the
Oestradiol as a therapeutic drug? progestogen medroxyprogesterone acetate, which is
Evidence that oestradiol could be used as an adjunct to probably responsible for the slightly increased risk of
antipsychotic medication even without an oestrogen breast cancer if given for more than 7 years.72 A protective
deficiency state is increasing. Further replications of effect on memory has been shown when treatment starts

www.thelancet.com/psychiatry Published online November 14, 2016 http://dx.doi.org/10.1016/S2215-0366(16)30379-0 5

 Series

However, according to several studies,44,86,87 raloxifene

Panel 2: Important effects of oestrogen replacement in the menopause and does not improve psychotic symptoms but only negative
suggested guidelines16,30,71–77 and total symptoms in postmenopausal women with
Positive schizophrenia. Furthermore, a positive effect might be
Perimenopausal complaints reduced possible on cognition in patients with schizophrenia,
• Physical: hot flushes, genital discomfort, ageing of collagen (skin, joints, intervertebral although not confirmed.31,87,88 Although the use of
discs) reduced raloxifene is relatively safe, there seems to be a low risk
• Mental: depression, irritability, emotional lability, psychotic symptoms reduced of blood clots.31
Risk of osteoporosis and fragility fracture reduced A novel compound, 17-α-oestradiol, has been shown to
Possible delay of cognitive decline and Alzheimer’s disease be as equally protective as conventional 17-β-oestradiol
Possible cardiovascular protection (only if started immediately after menopause) with regards to cognitive deficits, depression-like
symptoms, and problems with motor coordination in
Negative rats and, most importantly, without increasing oestradiol
Increase in endometrial carcinoma if unopposed oestrogens are administered concentrations or markers of feminising action.89
- In women without hysterectomy combine with progesterone
Possible risk of breast cancer increased (probably due to combination with progestogen) Implications for research
- Do not give to patients with a familial or own risk, and usually not for longer than 7 years Clinical research
Risk of thrombosis, cerebral insult, and coronary heart disease increased There are many open research questions regarding the
- No prescription for patients at risk best clinical regimens for the different patient groups
- Rather use 17-β-oestradiol and transdermal application mentioned in this Series paper.
- Start only within the first 10 years after menopause and do not apply in patients aged What kind of hormone replacement is best for patients
>60 years with psychosis who have premature gonadal insufficiency
or menopausal symptoms? For women, transdermal
17-β-oestradiol seems to be a good option. However, we
early and 17-β-oestradiol is given.50,83 A meta-analysis84 on do not know whether the recommended ultra-low doses
43 RCTs found that menopausal hormone therapy did for hormone replacement in otherwise healthy women
not increase mortality, neither all cause nor cardiac who are perimenopausal or postmenopausal would still
deaths, or those from stroke or cancer. benefit the mental state of patients with psychosis.
A very important argument in the controversy about Furthermore, we know very little about the mental effects
the advantages and disadvantages of oestrogen of the progestogen progesterone, which is recommended
substitution is that we have to distinguish a preventive to be added to protect the endometrium.75
application from a therapeutic use. In women with Similar questions have to be posed with regards to
psychosis, oestradiol would be used therapeutically. adjunct oestrogen therapy in younger women. So far, the
Potential side-effects would have to be outweighed by doses applied in acutely ill patients seem too high for
the benefits and have to be compared with the side- long-term use. What could an adequate long-term
effects of antipsychotics and other adjunct medications. regimen be? Possibly the dose should be adjusted
Pros and cons have to be carefully assessed in each according to the physiological oestrogen concentrations
woman individually, and the final decision has to be still produced by the women themselves, in the context
made by the well informed woman herself. Oestrogens of individualised therapy. However, no studies have been
should only be substituted in women without risk done on this subject.
factors, in close cooperation with a gynaecologist, and Most importantly, hardly any studies have been done
with close monitoring. on men. How many have low testosterone concentrations
Psychiatrists should have the knowledge to be the and how should those concerned be treated?
advisers of their patients and the cooperation partners of Further studies are urgently needed on compounds
the gynaecologists. To this end, they should know the that have more specific and potent oestrogenic activity in
best method of hormone replacement.3 The natural the brain than in other tissues. Such compounds should
17-β-oestradiol not only seems to be the oestrogen with the both minimise the side-effects of hormonal therapy and
best benefit-risk profile, but also the most active in the permit new therapeutic strategies in men.
brain.26 Transdermal application (patches or gel) has fewer
side-effects than oral application.75 Progestogens have to Research into pathogenetic mechanisms and the stress–
be added to oestrogens for endometrial protection. In this prolactin–dopamine hypothesis
case, the natural, body identical, micronised progesterone The findings on hyperprolactinaemia and gonadal
seems to be the one with the least side-effects.75 insufficiency might give important insights into the
Promising alternatives to oestrogens are being pathogenetic mechanisms of emerging psychosis. Thus,
studied, especially so-called selective oestrogen receptor we have speculated that hyperprolactinaemia in
modulators. Raloxifene, for example, mainly acts on the prodromal and first-episode psychosis is caused by
bone, but might also act on different brain receptors.85 psychological stress associated with the (emerging)

6 www.thelancet.com/psychiatry Published online November 14, 2016 http://dx.doi.org/10.1016/S2215-0366(16)30379-0

 Series

illness.5 This stress-induced hyperprolactinaemia could

contribute to triggering the onset of acute psychotic Paraventricular
Raphe
nucleus
symptoms since hyperprolactinaemia in a feedback loop nucleus + 5-HT
induces the production of the prolactin-inhibiting factor
(PIF),5 and PIF is mainly dopamine,90,91 a neurotransmitter PRF
well known to provoke psychotic symptoms.92 Hypothalamic neurons +
The onset of psychosis has been linked to stress on the opioid neurons
Hypothalamic
one hand,93–96 and to an increase of dopaminergic glutamatergic Mammillary
neurotransmission on the other.92,93 Furthermore, stress inputs nuclei
can lead to an increased dopamine release.93,96,97 However, Histamine
how stress can enhance dopaminergic neurotransmission Basal + –
forebrain Dopamine
has not yet been really understood.92,93 Explanations have –
neuron + –
so far discussed a role for the concomitantly enhanced ACh
Arcuate
stress hormone cortisol and inflammation markers,94,98 nucleus
or a hyper-reactivity of dopamine neurons to stress.95,97 +
But these theories neither integrate the findings
–
of hyperprolactinaemia, nor those of low levels of GABA
gonadal hormones.
Therefore, we have suggested the stress–prolactin–
Hypothalamus
dopamine hypothesis.5 According to this hypothesis, the
well known increase of prolactin due to stress90,94,99
Oestrogen
enhances dopamine release (ie, PIF) in a feedback
mechanism (figure 2).90,91,100 This increase of dopamine
Dopamine* –
can then provoke psychotic symptoms in people with a GABA
TRH
+ Oxytocin
susceptibility for psychosis. This hypothesis would mean To CNS VIP
that the tuberoinfundibular dopamine pathway might be
involved in the pathogenesis of schizophrenic psychoses,
+
and not just the mesolimbic and mesocortical pathways, Pituitary
as hypothesised prominently in the literature.
The stress–prolactin–dopamine hypothesis would
explain the well known fact that stress can trigger the
PRL
onset of psychosis. At the same time, stress-induced – Spinal
hyperprolactinaemia would explain the oestrogen afferent

deficiency in (a subgroup of) women with psychosis,

+
since enhanced prolactin suppresses ovarian oestrogen +
Breast
production. Together with the oestrogen protection Suckling
hypothesis, this stress–prolactin–dopamine hypothesis PRL receptors stimulus
Oestrogen receptors Multiple target organs
might also help to explain why schizophrenic psychoses
often occur or relapse during periods with strong
fluctuations of sex hormones and prolactin, such as in Figure 2: Control of prolactin production and secretion
*Dopamine is prolactin-inhibiting factor. PRL=prolactin. PRF=prolactin-releasing factor. TRH=thyrotropin-releasing
adolescence, post partum, or perimenopause. hormone. VIP=vasoactive intestinal polypeptide. GABA=γ-aminobutyric acid. ACh=acetylcholine.
Further support for this hypothesis came from Labad 5-HT=5-hydroxytryptamine. This figure has been adapted from the figure published in Williams Textbook of
and colleagues.94 They found increased prolactin (not Endocrinology, Low, Neuroendocrinology, 126–127, Copyright Elsevier (2008).90
cortisol) to be a predictor of the transition to psychosis.
Hyperprolactinaemia and increased pituitary volumes As to the specificity of the results for psychoses, other
have predominantly been found in women with psychosis. psychiatric disorders also seem to show some
This occurrence might be explained by oestrogens disturbances of the gonadal axis, although not to the
sensitising the pituitary to release prolactin.90,100 In line same degree, and secondary amenorrhoea is a well
with this would be functional MRI findings showing that known phenomenon in periods of stress.102
women react differently to stress than men do, especially
in their high oestrogen phases of the menstrual cycle.101 Conclusions
Unfortunately, very little research has been done in this Clinically, evidence that 17-β-oestradiol could be a useful
area in men with psychosis, apart from studies showing adjunctive therapy is emerging. It could complement and
that there is a subgroup with hyperprolactinaemia5,6 and improve the drug therapies that are traditionally used for
that some men with (emerging) psychosis have low women with schizophrenic psychoses. In particular, it
testosterone or oestrogen concentrations. More research might be used as an adjunct in therapy-resistant women.
in men is needed in this area. Further studies on its potential for enhancing cognition

www.thelancet.com/psychiatry Published online November 14, 2016 http://dx.doi.org/10.1016/S2215-0366(16)30379-0 7

 Series

3 Riecher-Rössler A, Kulkarni J. Estrogens and gonadal function in

Search strategy and selection criteria schizophrenia and related psychoses. Curr Top Behav Neurosci
2011; 8: 155–71.
I searched MEDLINE, PubMed, and Google Scholar for articles 4 Markham JA. Sex steroids and schizophrenia.
published in the past 5 years with the MeSH major topics Rev Endocr Metab Disord 2012; 13: 187–207.
“schizophrenia”, “estrogens”, “estradiol”, “prolactin”, 5 Riecher-Rössler A, Rybakowski JK, Pflueger MO, et al.
Hyperprolactinemia in antipsychotic-naive patients with
“pituitary” [Other Term], “testosterone”, and “selective first-episode psychosis. Psychol Med 2013; 43: 2571–82.
estrogen receptor modulators”. The date of my last search was 6 Aston J, Rechsteiner E, Bull N, Borgwardt S, Gschwandtner U,
Aug 16, 2016. I applied no language restrictions. I searched Riecher-Rössler A. Hyperprolactinaemia in early psychosis—not
only due to antipsychotics.
the reference lists of these articles for important earlier Prog Neuropsychopharmacol Biol Psychiatry 2010; 34: 1342–44.
papers. These publications were supplemented with earlier 7 González-Blanco L, Greenhalgh AMD, Garcia-Rizo C, Egea EF,
landmark papers and reviews. Because of space restrictions, Miller BJ, Kirkpatrick B. Prolactin concentrations in
antipsychotic-naïve patients with schizophrenia and related
I mainly cited review articles regarding older publications. disorders: a meta-analysis. Schizophr Res 2016; 174: 156–60.
8 Ittig S, Studerus E, Riecher-Rössler A. M58. A possible role for
prolactin in emerging psychosis. Abstracts from the 5th Biennial
would be very important since we still do not have potent SIRS Conference—poster abstracts. Florence; http://www.nature.
com/public/article-assets/npg/npjschz/abstracts/npjschz20168.
drugs here, as the results of a review103 on cognition- pdf (accessed Oct 7, 2016).
enhancing drugs in schizophrenia showed. Furthermore, 9 Petrikis P, Tigas S, Tzallas AT, Archimandriti DT, Skapinakis P,
raloxifene might be an option if cognition is the target. Mavreas V. Prolactin levels in drug-naive patients with
schizophrenia and other psychotic disorders.
Also, this treatment option needs further studies. Int J Psychiatry Clin Pract 2016; 20: 165–69.
In perimenopause and postmenopause, oestrogen 10 Mahe V, Dumaine A. Oestrogen withdrawal associated psychoses.
replacement for women has been recommended for Acta Psychiatr Scand 2001; 104: 323–31.
11 van der Werf M, Hanssen M, Kohler S, et al. Systematic review
many reasons (panel 2). Schizophrenic psychoses could and collaborative recalculation of 133,693 incident cases of
potentially be an additional indication. schizophrenia. Psychol Med 2014; 44: 9–16.
Of course the decision to apply oestradiol must always 12 Riecher-Rössler A. Die Spätschizophrenie—eine valide Entität?
be made on the basis of an individual risk–benefit Eine empirische Studie zu Risikofaktoren, Krankheitsbild und
Verlauf. Habilitationsschrift. Mannheim: Fakultät für klinische
assessment72 and in close collaboration with a Medizin Mannheim, Universität Heidelberg, 1994.
gynaecologist. The challenge here is that after the first 13 Riecher-Rössler A, Löffler W, Munk-Jorgensen P. What do we
alarming results of the WHI study,80 not all professionals really know about late-onset schizophrenia?
Eur Arch Psychiatry Clin Neurosci 1997; 247: 195–208.
and patients have kept updated with newer, contrasting 14 Häfner H, Riecher-Rössler A, Hambrecht M, et al. [Sex differences
results and now “overestimate the risks and in schizophrenic diseases]. Fortschr Neurol Psychiatr 1991; 59: 343–
contraindications, and underestimate the impact of 60 (in German).
15 Häfner H, Riecher-Rössler A, An Der Heiden W, Maurer K,
menopausal symptoms on a woman’s quality of life”.71 Fatkenheuer B, Löffler W. Generating and testing a causal
As particularly in women with psychosis the benefits explanation of the gender difference in age at first onset of
might often outweigh the risks, more research is schizophrenia. Psychol Med 1993; 23: 925–40.
16 Riecher-Rössler A. Psychotic disorders and menopause: the
urgently needed. untold story. In: Soares C, Warren M, eds. The menopausal
Other strategies should already be part of standard transition—interface between gynecology and psychiatry. Basel:
clinical care, such as assessments of symptoms of Karger, 2009: 115–26.
gonadal insufficiency, with differential diagnosis and 17 Riecher-Rössler A. Oestrogens and schizophrenia.
Curr Opin Psychiatry 2003; 16: 187–92.
therapeutic actions if indicated. 18 Könnecke R, Häfner H, Maurer K, Löffler W, an der Heiden W.
More research is also needed regarding the poorly Main risk factors for schizophrenia: increased familial
understood disturbances of the hypothalamic–pituitary– loading and pre- and peri-natal complications antagonize the
protective effect of oestrogen in women. Schizophr Res 2000;
gonadal axis, hyperprolactinaemia, and sex hormone 44: 81–93.
deficiency in patients with schizophrenic psychoses. Our 19 Cohen RZ, Seeman MV, Gotowiec A, Kopala L. Earlier puberty as
understanding of the pathogenesis of these disorders, at a predictor of later onset of schizophrenia in women. Am J
Psychiatry 1999; 156: 1059–64.
least in a subgroup of patients, might benefit substantially 20 Hayeems R, Seeman MV. Puberty and schizophrenia onset.
from further research in this area. In: Bergemann N, Riecher-Rössler A, eds. Estrogen efects in
psychiatric disorders. Vienna, New York, NY: Springer, 2005: 95.
Declaration of interests
I declare no competing interests. 21 Riecher-Rössler A. Oestrogen effects in schizophrenia and their
potential therapeutic implications–review.
Acknowledgments Arch Womens Ment Health 2002; 5: 111–18.
I would like to thank Claudine Pfister for her help in preparing this 22 Kendell RE, Chalmers JC, Platz C. Epidemiology of puerperal
manuscript. psychoses. Br J Psychiatry 1987; 150: 662–73.
23 Riecher-Rössler A, Häfner H, Dutsch-Strobel A, et al.
References
Further evidence for a specific role of estradiol in schizophrenia?
1 Riecher-Rössler A, Häfner H. Schizophrenia and oestrogens—
Biol Psychiatry 1994; 36: 492–94.
is there an association? Eur Arch Psychiatry Clin Neurosci 1993;
242: 323–28. 24 Riecher-Rössler A, Häfner H, Stumbaum M, Maurer K, Schmidt
R. Can estradiol modulate schizophrenic symptomatology?
2 Riecher-Rössler A. Estrogens and schizophrenia. In: Bergemann N,
Schizophr Bull 1994; 20: 203–14.
Riecher-Rössler A, eds. Estrogen effects in psychiatric disorders.
Vienna: Springer, 2005: 31–52. 25 Melcangi RC, Panzica G, Garcia-Segura LM. Neuroactive steroids:
focus on human brain. Neuroscience 2011; 191: 1–5.

8 www.thelancet.com/psychiatry Published online November 14, 2016 http://dx.doi.org/10.1016/S2215-0366(16)30379-0

 Series

26 McCarthy MM. Estradiol and the developing brain. Physiol Rev 51 Akhondzadeh S, Nejatisafa AA, Amini H, et al. Adjunctive estrogen
2008; 88: 91–124. treatment in women with chronic schizophrenia: a double-blind,
27 Jimenez JA, Mancini-Marie A, Lakis N, Rinaldi M, Mendrek A. randomized, and placebo-controlled trial. Prog Neuropsychopharmacol
Disturbed sexual dimorphism of brain activation during mental Biol Psychiatry 2003; 27: 1007–12.
rotation in schizophrenia. Schizophr Res 2010; 122: 53–62. 52 Louza MR, Marques AP, Elkis H, Bassitt D, Diegoli M, Gattaz WF.
28 Mendrek A. Reversal of normal cerebral sexual dimorphism in Conjugated estrogens as adjuvant therapy in the treatment of acute
schizophrenia: evidence and speculations. Med Hypotheses 2007; schizophrenia: a double-blind study. Schizophr Res 2004; 66: 97–100.
69: 896–902. 53 Chua WL, Izquierdo de Santiago A, Kulkarni J, Mortimer A. Estrogen
29 DonCarlos LL, Azcoitia I, Garcia-Segura LM. Neuroprotective for schizophrenia. Cochrane Database Syst Rev 2005; 4: CD004719.
actions of selective estrogen receptor modulators. 54 Ghafari E, Fararouie M, Shirazi HG, Farhangfar A, Ghaderi F,
Psychoneuroendocrinology 2009; 34 (suppl 1): S113–22. Mohammadi A. Combination of estrogen and antipsychotics in the
30 Pompili A, Arnone B, Gasbarri A. Estrogens and memory in treatment of women with chronic schizophrenia: a double-blind,
physiological and neuropathological conditions. randomized, placebo-controlled clinical trial.
Psychoneuroendocrinology 2012; 37: 1379–96. Clin Schizophr Relat Psychoses 2013; 6: 172–76.
31 Weickert TW, Weinberg D, Lenroot R, et al. Adjunctive raloxifene 55 Good KP, Kopala LC, Martzke JS, et al. Hormone replacement
treatment improves attention and memory in men and women therapy in postmenopausal women with schizophrenia: preliminary
with schizophrenia. Mol Psychiatry 2015; 20: 685–94. findings. Schizophr Res 1999; 12: 131.
32 Wu YC, Hill RA, Gogos A, van den Buuse M. Sex differences and 56 Lindamer LA, Buse DC, Lohr JB, Jeste DV. Hormone replacement
the role of estrogen in animal models of schizophrenia: therapy in postmenopausal women with schizophrenia: positive
interaction with BDNF. Neuroscience 2013; 239: 67–83. effect on negative symptoms? Biol Psychiatry 2001; 49: 47–51.
33 Garcia-Segura L, Azcoitia I, Doncarlos L. Neuroprotection by 57 Ahokas A, Aito M, Rimon R. Positive treatment effect of estradiol in
estradiol. Prog Neurobiol 2001; 63: 29–60. postpartum psychosis: a pilot study. J Clin Psychiatry 2000; 61: 166–69.
34 Gogos A, Sbisa AM, Sun J, Gibbons A, Udawela M, Dean B. 58 Ko YH, Jung SW, Joe SH, et al. Association between serum
A role for estrogen in schizophrenia: clinical and preclinical testosterone levels and the severity of negative symptoms in male
findings. Int J Endocrinol 2015; 2015: 16. patients with chronic schizophrenia. Psychoneuroendocrinology 2007;
35 Balthazart J, Ball GF. Is brain estradiol a hormone or a 32: 385–91.
neurotransmitter? Trends Neurosci 2006; 29: 241–49. 59 da Silva TL, Ravindran AV. Contribution of sex hormones to gender
36 Arad M, Weiner I. Contrasting effects of increased and decreased differences in schizophrenia: a review. Asian J Psychiatry 2015;
dopamine transmission on latent inhibition in ovariectomized 18: 2–14.
rats and their modulation by 17beta-estradiol: an animal model of 60 Taherianfard M, Shariaty M. Evaluation of serum steroid hormones
menopausal psychosis? Neuropsychopharmacology 2010; in schizophrenic patients. Indian J Med Sci 2004; 58: 3–9.
35: 1570–82. 61 Huber TJ, Tettenborn C, Leifke E, Emrich HM. Sex hormones in
37 Arad M, Weiner I. Abnormally rapid reversal learning and psychotic men. Psychoneuroendocrinology 2005; 30: 111–14.
reduced response to antipsychotic drugs following ovariectomy in 62 van Rijn S, Aleman A, de Sonneville L, et al. Neuroendocrine
female rats. Psychoneuroendocrinology 2012; 37: 200–12. markers of high risk for psychosis: salivary testosterone in
38 Gogos A, Kwek P, van den Buuse M. The role of estrogen and adolescent boys with prodromal symptoms. Psychol Med 2011;
testosterone in female rats in behavioral models of relevance to 41: 1815–22.
schizophrenia. Psychopharmacology (Berl) 2012; 219: 213–24. 63 Akhondzadeh S, Rezaei F, Larijani B, Nejatisafa AA, Kashani L,
39 Vasudevan N, Pfaff DW. Non-genomic actions of estrogens and Abbasi SH. Correlation between testosterone, gonadotropins and
their interaction with genomic actions in the brain. prolactin and severity of negative symptoms in male patients with
Front Neuroendocrinol 2008; 29: 238–57. chronic schizophrenia. Schizophr Res 2006; 84: 405–10.
40 Oesterlund M. The role of estrogens in neuropsychiatric 64 Bundy H, Stahl D, MacCabe JH. A systematic review and
disorders. Curr Opin Psychiatry 2002; 15: 307–12. meta-analysis of the fertility of patients with schizophrenia and
41 Weickert CS, Miranda-Angulo AL, Wong J, et al. Variants in the their unaffected relatives. Acta Psychiatr Scand 2011; 123: 98–106.
estrogen receptor alpha gene and its mRNA contribute to risk for 65 Maguire GA. Prolactin elevation with antipsychotic medications:
schizophrenia. Hum Mol Genet 2008; 17: 2293–309. mechanisms of action and clinical consequences. J Clin Psychiatry
42 Min JA, Kim JJ, Pae CU, et al. Association of estrogen receptor 2002; 63 (suppl 4): 56–62.
genes and schizophrenia: A preliminary study. 66 Canuso CM, Goldstein JM, Wojcik J, et al. Antipsychotic
Prog Neuropsychopharmacol Biol Psychiatry 2012; 36: 1–4. medication, prolactin elevation, and ovarian function in women
43 Begemann MJ, Dekker CF, van Lunenburg M, Sommer IE. with schizophrenia and schizoaffective disorder. Psychiatry Res
Estrogen augmentation in schizophrenia: a quantitative review of 2002; 111: 11–20.
current evidence. Schizophr Res 2012; 141: 179–84. 67 Maric N, Popovic V, Jasovic-Gasic M, Pilipovic N, van Os J.
44 Heringa SM, Begemann MJ, Goverde AJ, Sommer IE. Cumulative exposure to estrogen and psychosis: a peak bone mass,
Sex hormones and oxytocin augmentation strategies in case-control study in first-episode psychosis. Schizophr Res 2005;
schizophrenia: A quantitative review. Schizophr Res 2015; 73: 351–55.
168: 603–13. 68 Nordholm D, Krogh J, Mondelli V, Dazzan P, Pariante C,
45 Sommer IE, van Westrhenen R, Begemann MJ, de Witte LD, Nordentoft M. Pituitary gland volume in patients with
Leucht S, Kahn RS. Efficacy of anti-inflammatory agents to schizophrenia, subjects at ultra high-risk of developing psychosis
improve symptoms in patients with schizophrenia: an update. and healthy controls: a systematic review and meta-analysis.
Schizophr Bull 2014; 40: 181–91. Psychoneuroendocrinology 2013; 38: 2394–404.
46 Kulkarni J, Riedel A, de Castella AR, et al. Estrogen–a potential 69 MacMaster FP, El-Sheikh R, Upadhyaya AR, Nutche J,
treatment for schizophrenia. Schizophr Res 2001; 48: 137–44. Rosenberg DR, Keshavan M. Effect of antipsychotics on pituitary
gland volume in treatment-naive first-episode schizophrenia: a pilot
47 Kulkarni J, de Castella A, Fitzgerald PB, et al. Estrogen in severe
study. Schizophr Res 2007; 92: 207–10.
mental illness: a potential new treatment approach.
Arch Gen Psychiatry 2008; 65: 955–60. 70 Smith S, Wheeler MJ, Murray R, O’Keane V. The effects of
antipsychotic-induced hyperprolactinaemia on the
48 Kulkarni J: Oestrogen—a new treatment approach for
hypothalamic-pituitary-gonadal axis. J Clin Psychopharmacol 2002;
schizophrenia? Med J Aust 2009; 190: S37–38.
22: 109–14.
49 Kulkarni J, Gavrilidis E, Wang W, et al. Estradiol for
71 National Institute for Health and Care Excellence. Menopause:
treatment-resistant schizophrenia: a large-scale
diagnosis and management. NICE guidelines 23. London: NICE, 2015.
randomized-controlled trial in women of child-bearing age.
https://www.nice.org.uk/guidance/NG23 (accessed Jan 16, 2016).
Mol Psychiatry 2015; 20: 695–702.
72 Santen RJ, Allred DC, Ardoin SP, et al. Postmenopausal hormone
50 Sherwin BB. Estrogen and memory in women: how can we
therapy: an Endocrine Society scientific statement.
reconcile the findings? Horm Behav 2005; 47: 371–75.
J Clin Endocrinol Metab 2010; 95 (suppl 1): s1–66.

www.thelancet.com/psychiatry Published online November 14, 2016 http://dx.doi.org/10.1016/S2215-0366(16)30379-0 9

 Series

73 Gurney EP, Nachtigall MJ, Nachtigall LE, Naftolin F. The Women’s 88 Huerta-Ramos E, Iniesta R, Ochoa S, et al. Effects of raloxifene on
Health Initiative trial and related studies: 10 years later: a clinician’s cognition in postmenopausal women with schizophrenia:
view. J Steroid Biochem Mol Biol 2014; 142: 4–11. a double-blind, randomized, placebo-controlled trial.
74 Rossouw JE, Prentice RL, Manson JE, et al. Postmenopausal Eur Neuropsychopharmacol 2014; 24: 223–31.
hormone therapy and risk of cardiovascular disease by age and 89 Kaur SP, Bansal S, Chopra K. 17α-Estradiol: a candidate neuroserm
years since menopause. JAMA 2007; 297: 1465–77. and non-feminizing estrogen for postmenopausal neuronal
75 L’Hermite M. HRT optimization, using transdermal estradiol plus complications. Steroids 2015; 96: 7–15.
micronized progesterone, a safer HRT. Climacteric 2013; 90 Low MJ. Neuroendocrinology. In: Kronenberg HM, Melmed S,
16 (suppl 1): 44–53. Polonsky KS, Larsen PR, eds. Williams textbook of endocrinology,
76 de Villiers TJ, Gass MLS, Haines CJ, et al. Global consensus 11th edn. Philadelphia, PA: Saunders Elsevier Co, 2008: 85–154.
statement on menopausal hormone therapy. Maturitas 2013; 91 Egli M, Leeners B, Kruger THC. Prolactin secretion patterns:
74: 391–92. basic mechanisms and clinical implications for reproduction.
77 Rozenberg S, Vandromme J, Antoine C. Postmenopausal hormone Reproduction 2010; 140: 643–54.
therapy: risks and benefits. Nat Rev Endocrinol 2013; 9: 216–27. 92 Howes OD, Kapur S. The dopamine hypothesis of schizophrenia:
78 van der Leeuw C, Peeters S, Domen P, et al. Bone mineral density version III–the final common pathway. Schizophr Bull 2009;
as a marker of cumulative estrogen exposure in psychotic disorder: 35: 549–62.
a 3 year follow-up study. PLoS One 2015; 10: e0136320. 93 van Winkel R, Stefanis NC, Myin-Germeys I. Psychosocial stress
79 Lindamer LA, Buse DC, Auslander L, Unutzer J, Bartels SJ, and psychosis. A review of the neurobiological mechanisms and the
Jeste DV. A comparison of gynecological variables and service use evidence for gene-stress interaction. Schizophr Bull 2008;
among older women with and without schizophrenia. Psychiatr Serv 34: 1095–105.
2003; 54: 902–04. 94 Labad J, Stojanovic-Perez A, Montalvo I, et al. Stress biomarkers as
80 Rossouw JE, Anderson GL, Prentice RL, et al. Risks and benefits of predictors of transition to psychosis in at-risk mental states: roles
estrogen plus progestin in healthy postmenopausal women: for cortisol, prolactin and albumin. J Psychiatr Res 2015; 60: 163–69.
principal results from the women’s health initiative randomized 95 Myin-Germeys I, van Os J. Stress-reactivity in psychosis: evidence
controlled trial. JAMA 2002; 288: 321–33. for an affective pathway to psychosis. Clin Psychol Rev 2007;
81 Hlatky MA, Boothroyd D, Vittinghoff E, Sharp P, Whooley MA. 27: 409–24.
Quality-of-life and depressive symptoms in postmenopausal women 96 Aiello G, Horowitz M, Hepgul N, Pariante CM, Mondelli V.
after receiving hormone therapy: results from the Heart and Stress abnormalities in individuals at risk for psychosis: a review of
Estrogen/Progestin Replacement Study (HERS) trial. JAMA 2002; studies in subjects with familial risk or with “at risk” mental state.
287: 591–97. Psychoneuroendocrinology 2012; 37: 1600–13.
82 Birkhäuser MH, Panay N, Archer DF, et al. Updated practical 97 Mizrahi R, Addington J, Rusjan PM, et al. Increased stress-induced
recommendations for hormone replacement therapy in the dopamine release in psychosis. Biol Psychiatry 2012; 71: 561–67.
peri- and postmenopause. Climacteric 2008; 11: 108–23. 98 Mondelli V. From stress to psychosis: whom, how, when and why?
83 Azcoitia I, Arevalo MA, De Nicola AF, Garcia-Segura LM. Epidemiol Psychiatr Sci 2014; 23: 215–18.
Neuroprotective actions of estradiol revisited. 99 Prabhakar VK, Davis JR. Hyperprolactinaemia.
Trends Endocrinol Metab 2011; 22: 467–73. Best Pract Res Clin Obstet Gynaecol 2008; 22: 341–53.
84 Benkhadra K, Mohammed K, Al Nofal A, et al. Menopausal hormone 100 Fitzgerald P, Dinan TG. Prolactin and dopamine: what is the
therapy and mortality: a systematic review and meta-analysis. connection? A review article. J Psychopharmacol 2008;
J Clin Endocrinol Metab 2015; 100: 4021–28. 22 (suppl 2): 12–19.
85 Landry M, Levesque D, Di Paolo T. Estrogenic properties of 101 Goldstein JM, Jerram M, Abbs B, Whitfield-Gabrieli S, Makris N.
raloxifene, but not tamoxifen, on D2 and D3 dopamine receptors in Sex differences in stress response circuitry activation dependent on
the rat forebrain. Neuroendocrinology 2002; 76: 214–22. female hormonal cycle. J Neurosci 2010; 30: 431–38.
86 Usall J, Huerta-Ramos E, Labad J, et al. Raloxifene as an adjunctive 102 Huber TJ, Rollnik J, Wilhelms J, von zur Muhlen A, Emrich HM,
treatment for postmenopausal women with schizophrenia: Schneider U. Estradiol levels in psychotic disorders.
a 24-week double-blind, randomized, parallel, placebo-controlled Psychoneuroendocrinology 2001; 26: 27–35.
trial. Schizophr Bull 2016; 42: 309–17. 103 Keefe RSE, Buchanan RW, Marder SR, et al. Clinical trials of
87 Kulkarni J, Gavrilidis E, Gwini SM, et al. Effect of adjunctive potential cognitive-enhancing drugs in schizophrenia: what have we
raloxifene therapy on severity of refractory schizophrenia in women: learned so far? Schizophr Bull 2013; 39: 417–35.
a randomized clinical trial. JAMA Psychiatry 2016; 73: 947–54.

10 www.thelancet.com/psychiatry Published online November 14, 2016 http://dx.doi.org/10.1016/S2215-0366(16)30379-0