Am J Psychiatry 162:2, February 2005
Vulnerability to Mental Illnesses:
Gender Makes a Difference, and So Does
Providing Good Psychiatric Care
D edicating oneself to the understanding and treatment of mental illnesses is the life
work of the psychiatrist. How we do that on an individual basis may vary. Some of us
choose to work primarily with a specific subgroup, such as the elderly, children, or pa-
tients with psychosis or mood disorders. Others have a more general practice. Some
emphasize pharmacotherapy, others psychotherapy (sadly less and less), and still oth-
ers combined therapies.
Yet one problem that we all grapple with is understanding how and why our patients
develop the particular illnesses that they do and how we can predict and affect the fu-
ture course of their illnesses. Specifically, we seek
to understand how illness develops as a conse-
quence of a complex matrix of factors that im-
pinge on each individual patient. One of these
factors is gender, the topical theme of this issue.
Many other factors also play a role, such as early
and recent life experiences, the availability of
social supports, or genetic influences. As we lis-
ten to our patients and try to determine the na-
ture of their symptoms and the best way to treat
them, we are continuously pondering what addi-
tional information we should obtain and how we
can use this knowledge to inform our treatment
Gender is an interesting example of a complex factor that may influence the develop-
ment or course of mental illnesses. A person’s sex is genetically determined at the time
of conception and is not a matter of choice. However, the way a man or woman lives
within his or her assigned sex can play out in many different directions. Gender identity
is influenced by many things, such as socialization styles suggested by parents or
schools or peers, location and quality of housing, the availability of education and the
type provided, the kinds of toys that are given to the child, and perhaps even the kind of
clothing that a child is allowed to wear. All these factors—and no doubt many others—
help shape identity, and perhaps also ego strength and resilience. In turn, they may also
affect differential vulnerability to mental illnesses.
This issue of the Journal contains two interesting papers that provide some insights
about the relationship between gender and vulnerability to mood disorders. Kendler et
al. examine an interesting and powerful sample: nonidentical twins who also differ by
gender. As is well recognized, twins are a “natural laboratory” for studying the interac-
tions between genetic and nongenetic influences on vulnerability to mental illnesses.
This study measures the rates of depression in the twin pairs and then compares them
in terms of social support networks. As is almost always observed, the female subjects
have a higher rate of mood disorder than the male subjects. They also have more exten-
sive social support networks. When Kendler et al. examine the inter-relationship be-
tween social supports and vulnerability to depression, they find that women are more
sensitive to the depressogenic effects of having low levels of social support than men.
For the men the association between social supports and risk for depression is modest
and not significant. Paradoxically, having a lower rate of social support does not ac-
patterns of predisposition
occur will ultimately
improve our knowledge
of the pathophysiology
and etiology of various
Am J Psychiatry 162:2, February 2005
count for the higher rate of depression in the women; controlling for social supports ac-
tually augments the higher rate of depression in women. That is, women still have
higher rates of depression even when the influence of social supports is taken into ac-
count. These results suggest that there may be gender differences in the mechanisms of
depression, and that lack of social supports may be an important part of the causal
pathway in women but not in men. It is a pity that this paper does not also examine
same-sex dizygotic twins in order to help further disentangle relationships between
genetic and nongenetic influences.
A second paper by Kennedy et al. examines gender differences in incidence and age at
onset for bipolar disorder and mania. This study uses an excellent epidemiological de-
sign, examining all cases from the Camberwell catchment area in London over a 35-year
period. This study also examines the influence of psychosocial variables on gender dif-
ferences in bipolar disorder and mania. Men are found to have a slightly higher rate of
bipolar disorder during their early years (ages 16–25), but women have a higher rate
overall and a generally later age at onset. Women are also more likely to begin their bi-
polar life course with an episode of depression. Among the psychosocial variables ex-
amined, childhood antisocial behavior is found to be the strongest predictor of early-
onset bipolar disorder in male subjects. Other psychosocial variables examined are not
predictive, including developmental measures, premorbid functioning, and personal-
ity. The authors speculate that the observed gender differences in patterning of onset
may reflect hormonal influences across the lifespan and that early-onset bipolar disor-
der may represent a subtype within the bipolar spectrum that arises from a different
Overall, both studies fit within a large literature suggesting that there are gender dif-
ferences in vulnerability to mental illnesses. Men have higher rates of schizophrenia,
antisocial personality, ADHD, and learning disabilities. Women are more vulnerable to
mood disorders and borderline personality disorder. Understanding why these patterns
of predisposition occur will ultimately improve our knowledge of the pathophysiology
and etiology of these various mental illnesses. We still have much to learn, but these two
studies add modestly to our knowledge.
What are the implications of the gender studies in this issue of the Journal for the daily
practice of psychiatry? They are quite obvious, but still worth pointing out.
First, simply knowing a patient’s gender gives us useful information in making diag-
noses and planning treatment. This information, combined with other details, can help
us predict future course of illness and risk for recurrence.
Second, we need to go beyond simply doing DSM checklists of current symptoms
when we evaluate our patients. We need to take a careful social history and a compre-
hensive past psychiatric history. Unfortunately, economic constraints and time con-
straints have made this kind of “good clinical practice” increasingly more difficult. But
it is still necessary, and we should fight in every way we can to preserve and maintain
these standards of care.
Third, “good clinical practice” may also dictate the inclusion of some type of psycho-
therapeutic intervention in addition to medications. (Same problems as for implication
two, of course.)
Fourth, research studies can yield some small pearls that can improve “good clinical
practice” (i.e., it is good to read your AJP each month!)
For example, imagine having a 25-year-old married woman come to your office for
the first time, complaining of feeling depressed. Of course you will go through a review
of her symptoms, using DSM criteria. Suppose you reach a diagnosis of Major Depres-
sion. These articles remind us that we should do more than make a diagnosis and pre-
scribe an antidepressant. We need to understand her as a person living within a social
matrix. We need to determine whether she is close to her family—siblings, parents, and
husband—and whether she has a network of close friends. If you find that her social
Am J Psychiatry 162:2, February 2005
support network is limited (i.e., she does not feel close to family or friends), she is likely
to have a greater risk for relapse and perhaps also to be slower to respond to treatment.
Part of your treatment plan should focus on helping her improve her social supports.
She may also need to be seen more frequently and to need more psychotherapy. Over
the long run you need to be aware of the possibility that she may also develop episodes
of mania and become bipolar, since women often present first with depression and
then have their first episode of mania at a later age.
If the patient were a 25-year-old single man with depressive symptoms, the situation
is somewhat different. His risk for subsequently becoming bipolar is less overall, since
he is approaching the end of the high-risk age range for male subjects, but it will be
greater if he has a history of antisocial behavior. If the latter is the case, then you will
need to continue to be more vigilant for episodes of mania. You will of course inquire
about his social support network, but with the awareness that (on average) having a
good network will be less protective for him than for the aforementioned woman.
Admittedly, these studies provide only modest kernels of information that will help
clinicians improve their treatment plans. But they reinforce with large-group empirical
data what we have gleaned from observing individual cases. For vulnerability to mental
illnesses, gender makes a difference. Knowing some more specific details about this vul-
nerability can improve prediction of course and outcome and ultimately improve our
understanding of pathophysiology and etiology.
This illustrates, however, another important point. Being a good psychiatrist makes a
difference too. Psychiatry is the only medical specialty that treats the whole person. We
must see each patient as a unique person and understand and treat that person not only
in terms of presenting symptoms and diagnosis but also as a human being who lives
within a complex psychosocial matrix and whose present may be influenced by the
past. This is perhaps the most important implication of these two papers.