The hormonal keys to depression

When Torbjörn Bäckström was a medical student in the 1970s, he didn’t understand why a woman who seemed to be in good mental health was being held at a psychiatric facility.

Then, Bäckström and his colleagues learnt that once a month, the woman would be admitted because she had been violent and attacked people. “This was becoming a sort of pattern, that when we came there after she had been taken in by force, she did not understand what had happened and why it had happened,” recalls Bäckström, who is now an endocrinologist and emeritus professor at Umeå University in Sweden.

Two details caught Bäckström’s attention. One was that her violent episodes seemed to coincide with menstruation. The second was a paragraph in a physiology textbook that mentioned premenstrual syndrome and the effects of hormones on the brain.

Those two observations launched his research career, in which he focused on steroid hormones and their neurological effects. During this time, he has seen the field evolve from a passing mention in a textbook to the 2019 approval of the first drug in a new class of antidepressants. This compound, brexanolone, marketed as Zulresso, is a neuroactive steroid — a hormone that interacts with brain chemistry — and is the only drug specifically developed to treat post-partum depression.

The interaction between neuroactive steroid hormones and the brain is complex, cyclical and highly individualized. It’s not simply a case of hormone levels being too high or too low, but rather the result of how the brain responds to changes in those levels. To further complicate things, the definitions of mood disorders with a hormonal trigger — such as perimenopausal depression and premenstrual dysphoric disorder (PMDD), which is the onset of a mix of physical and psychological symptoms that are tied to menstruation — are either too restrictive, not restrictive enough or non-existent. This lack of clarity makes clinical trials challenging.

Research groups worldwide are studying how fluctuating hormone levels trigger or exacerbate mood disorders such as depression, with the hope that this will lead to better diagnoses and treatment.

Targeting GABA

The first commercial output of this neuroscientific awakening revolves around a neuroactive steroid called allopregnanolone, which is produced from the breakdown of the hormone progesterone.

During pregnancy, allopregnanolone levels rise steadily to peak in the third trimester. After the baby is born, they drop drastically. This swift decline is thought to be a major contributor to post-natal depression. In many people, the levels rise again after a few days, but in some they don’t; these individuals can develop severe depression and become withdrawn.

In 2019, the biopharmaceutical company Sage Therapeutics in Cambridge, Massachusetts, released brexanolone, a formulation of allopregnanolone that is given intravenously to people with severe post-partum depression. For some, the clinical benefits were profound.

Allopregnanolone binds to a receptor in the brain called γ-aminobutyric acid type A (GABA_A). GABA is an inhibitory neurotransmitter, which means it reduces neuronal activity rather than excites it. The GABA system in the brain is implicated in several mental-health conditions, including depression.

Another compound called zuranolone that Sage Therapeutics is developing as an oral treatment for major depressive disorder also targets the GABA_A receptor — as do a class of sedatives called benzodiazepines, which have been around since the 1950s and include drugs such as Valium. But benzodiazepines interact with GABA_A receptors only in the synaptic junctions between neurons, whereas brexanolone and zuranolone also interact with them on regions of the neuron called extrasynaptic receptors. The interaction with the synaptic receptors “controls the moment-to-moment inhibitory tone in the brain, sort of like the weather”, says Jeff Jonas, Sage Therapeutics’ chief innovation officer. By contrast the interaction with other GABA_A receptors “controls the overall inhibitory tone — that’s more like the climate”.

The goal is not short-term inhibition of neuronal activity, but longer-term, sustained change. A 2021 study of a two-week course of zuranolone in women with post-partum depression showed that reductions in participants’ depression scores were maintained for 45 days after starting treatment¹.

Another unusual feature of the neuroactive steroids that target the GABA system is their speed of action, says Bernard Ravina, chief medical officer of Praxis Precision Medicines in Boston, Massachusetts. Praxis has been developing these compounds for the treatment of major depressive disorder. And unlike conventional antidepressants, which can take several weeks or months to reach their full effect, compounds that interact with the GABA_A receptor seem to work in a few days. “That’s a big part of the appeal,” Ravina says.

But allopregnanolone is just the start of the neurosteroid story. “We also have to think about the interplay between the hormones and what they are attaching to,” says Lauren Osborne, director of the Johns Hopkins Center for Women’s Reproductive Mental Health in Baltimore, Maryland. Osborne points out that the changing levels of allopregnanolone and other hormones during pregnancy might also change the structure and function of the GABA_A receptor.

One clue that the trigger for post-partum depression is not just a straightforward drop in hormone levels after birth comes from the observation that lower than expected allopregnanolone levels during the second trimester of pregnancy correlate with an increased likelihood of developing this condition². “What’s happening at the second trimester that’s predicting something months later?” Osborne asks. “There are so many possibilities for what those links could be.”

And in many cases, people who experience these mood disorders don’t show abnormal hormone levels, says Tory Eisenlohr-Moul, a clinical psychologist at the University of Illinois in Chicago. She thinks the problem lies in the GABA_A receptor.

Eisenlohr-Moul’s research group is investigating whether certain subunits of the GABA_A receptor might be reacting differently to the changes in hormone and neurosteroid levels during the menstrual cycle in people with PMDD. The researchers’ theory is that the receptor is not adapting or responding appropriately to the hormonal fluctuations, leading to either an under-reaction or over-reaction to the hormonal shifts.

Definition and diagnosis

Between the start of menstruation and the end of menopause, some individuals experience more symptoms of hormone changes than others. Around 5% of women of childbearing age develop PMDD³, and estimates of post-natal depression⁴ range from 4% to nearly 64%. As many as 40% of women report depressive symptoms around the time of menopause⁵.

But despite the prevalence of these hormonal mood disorders, they can be difficult to diagnose; the clinical presentation encompasses a broad range of symptoms and timings. However, the clinical definition of PMDD in the latest Diagnostic and Statistical Manual of Mental Disorders, for example, is extremely precise, says psychiatrist Jayashri Kulkarni, director of the Monash Alfred Psychiatry Research Centre in Melbourne, Australia. “It must be a very clear, seven-day deterioration in mood exactly one week before menstruating, and it’s based on the 28-day cycle,” she says. “Well, that only exists in textbooks.”

Eisenlohr-Moul’s work with her own patients also suggests there are differences in the timing of PMDD symptoms relative to the menstrual cycle. Some people show symptoms such as irritability soon after ovulation, whereas others experience depression and fatigue around the onset of their period (typically 14 days before ovulation).

Similarly, researchers have not identified objective and accurate biological markers of brain-chemistry changes that can be used to diagnose perimenopausal depression. Even hormone levels in the blood don’t translate that well to what’s going on in the brain. A perimenopausal-depression diagnosis is further complicated by the fact that physical symptoms of menopause can develop up to five years after the psychological ones⁶.

There is also the question of how much of the depression and anxiety experienced by people going through their first period, pregnancy or menopause is related to non-biological factors. These life passages tend to coincide with various stressors: childbearing, raising a family and, in later years, perhaps caring for ageing parents, or dealing with retirement.

Ravina argues, however, that there is a compelling case for a biological underpinning to mood disorders. That case is strengthened by evidence that hormones influence mood disorders in everyone. “The data suggest people — regardless of sex — who have depression have lower endogenous levels of these neuroactive steroids and changes in extrasynaptic receptors,” Ravina says. “There’s clear biology driving [depression] in those hormone fluctuation reproductive states”.

The fact that hormones and their neuroactive metabolites are associated with mood disorders in both men and women could lessen the concern that linking hormones to mood will fuel the inaccurate perception that women are captive to hormones. Indeed, Eisenlohr-Moul thinks research into the link between hormones and the brain has helped to destigmatize these hormonally linked disorders — offensive jokes about premenstrual syndrome, for example, are nowhere near as common in popular culture as even 20 years ago.

“It’s an abnormal brain response to normal hormone changes and it can be treated,” Eisenlohr-Moul says. “It’s not a joke. It’s somebody’s life and it’s debilitating.”