TCPR: When does perimenopausal depression tend to start?
Dr. Nonacs: It’s during the transition into menopause that women are most vulnerable to depression. And that transition can actually take quite a while, like 5–7 years.
TCPR: How does it present?
Dr. Nonacs: Often they’ve had a history of depression, but have done fairly well up to this point. Now they’re having more depression, irritability, insomnia, and anxiety. Often the symptoms occur premenstrually, but sometimes, especially if they’re not taking an antidepressant, it will be a constant depressive episode.
TCPR: What physical symptoms come along with the mood changes?
Dr. Nonacs: We call those vasomotor symptoms—night sweats, hot flashes—and they tend to worsen before the onset of menses. It’s sometimes hard to pick up the premenstrual pattern because the cycles can be a little bit irregular in perimenopausal women. However, you will see a stretch of time where their mood is fine and then it worsens 1–2 weeks before the onset of menses. They have some relief when their period comes, but then it repeats.
TCPR: Does perimenopausal depression require a unique treatment?
Dr. Nonacs: It tends to respond better to serotonergic agents, so my primary choices would be either the SSRIs or the SNRIs. Those medications help both the depression and the vasomotor symptoms. I’d also look more closely at sleep disturbance. Women with vasomotor symptoms have a lot of sleep disturbance, and I might use specific treatments for sleep after first addressing anxiety and depression.
TCPR: Just any hypnotic, or is there a specific one?
Dr. Nonacs: One medication we like to use in this population is gabapentin. It helps with vasomotor symptoms, and it also improves sleep quality without causing much in the way of dependence or hangover effects. Most studies use 300–900 mg of gabapentin (IR) at night, but if you push the dosage up to 2400 mg, it appears that gabapentin works just as well as estrogen for treating sleep disturbance and hot flashes (Shan D et al, Am J Obstet Gynecol 2020;222(6):564–579.e12). I typically dose it all at night. Granted, that’s a high dose and many women will have sedation, but I think it seems to work similarly to estrogen replacement. (Ed note: Gabapentin has non-linear pharmacokinetics because its absorption in the gut becomes saturated at doses beyond 900–1,200 mg per single dose. In terms of what actually gets absorbed (ie, bioavailability), 900 mg of gabapentin equates to 540 mg and 2400 mg equates to 816 mg.)
TCPR: Paroxetine earned FDA approval for vasomotor symptoms in 2013 as Brisdelle. Anything special there?
Dr. Nonacs: I don’t think any of the SSRIs or SNRIs have an edge here. In general, they reduce the hot flashes by about 65% and start working within the first week. If you’re using them just for hot flashes, the ones with the best evidence are paroxetine (7.5–15 mg qhs), escitalopram (10–20 mg qhs), citalopram (10–20 mg qhs), desvenlafaxine (50–150 mg qhs), and venlafaxine XR (37.5–150 mg qhs). Venlafaxine is the least tolerated of the bunch and has significant withdrawal symptoms (Stubbs C et al, J Okla State Med Assoc 2017;110(5):272–274). As psychiatrists we are usually treating depression, and there I’d start with an antidepressant that the patient has responded to the past. There are also data on some of the newer antidepressants like vortioxetine, and that probably also works; it does have some serotonergic effects as well. But I think using non-serotonergic meds like bupropion probably won’t work as well in this population.
TCPR: What are the stages of menopause?
Dr. Nonacs: Let’s start with menarche. That’s when the cycles begin at puberty. After that, women have regular cycles with monthly fluctuations of estrogen and progesterone. Next is perimenopause, which typically starts in the mid to late 40s. Here you’ll see various changes in cycles—sometimes closer together or farther apart—and with that are more irregular hormone fluctuations. Perimenopause typically lasts 4 years, but can be as short as 6 months or as long as 10 years. Finally there’s menopause, which is when women stop having periods for at least a year. In menopause, women have stable but low levels of estrogen and progesterone.
TCPR: Is it worthwhile to test hormones when a woman presents with perimenopausal symptoms?
Dr. Nonacs: I don’t typically get hormone levels. I just treat the depression. It may play more of a role at the OB/GYN’s office if they are considering oral contraceptive hormones (OCPs) vs hormone replacement therapy (HRT). Women on the younger edge of perimenopause could benefit from OCPs, and they could also help with mood symptoms. The hormone-containing IUDs that inhibit ovulation may also help with mood in these women. Women who are closer to menopause may benefit from HRT, both for mood and hot flashes. Once you get beyond menopause, though, there really is no benefit to hormone treatment and probably more risks.
TCPR: Is your advice any different after surgical menopause?
Dr. Nonacs: Not really. The difference there is that the transition is much more dramatic, so the vasomotor symptoms are more severe. It can help to manage those pretty aggressively in the beginning. Another population that’s unique is women on tamoxifen for breast cancer prevention. That causes chemical menopause by blocking estrogen receptors. These women have a high incidence of mood changes and vasomotor symptoms but you can’t use HRT because of the breast cancer risk.
TCPR: Why don’t postmenopausal women benefit from HRT?
Dr. Nonacs: As they move past menopause, women start to lose estrogen receptors, which may explain why they are less responsive to estrogen. Also the older women get, the greater the risks of stroke, cardiovascular disease, and breast cancer with HRT (Marjoribanks J et al, Cochrane Database Syst Rev 2017;1(1):CD004143).
TCPR: Do you start HRT for depression in perimenopausal women?
Dr. Nonacs: No, I defer to OB/GYNs, and even there the indications are pretty limited. The OB/GYN will probably recommend HRT for perimenopausal symptoms like hot flashes, insomnia, and night sweats, but not usually for depression. I agree, and would start with an antidepressant, but HRT is feasible and has evidence for the treatment of perimenopausal depression, either on its own or when added to an antidepressant. The risks are generally with long-term use, exceeding 5 years, so they can be used for the perimenopausal period and then tapered off gradually after menopause. The women I refer for HRT are ones who either can’t tolerate the SSRIs or don’t respond fully to them. I’ll say, “Look, we’ve tried these conventional treatments for your symptoms; maybe we need to consider HRT,” and then I’ll consult with their OB/GYN.
TCPR: Tell us about the evidence for HRT in depression.
Dr. Nonacs: There are a lot of studies looking at the impact of estrogen on mood, but most were in healthier women where they looked at emotional well-being as opposed to treating depression. There are actually only two randomized controlled trials where they’ve looked specifically at women with clinical depression (total n = 84). About 65%–70% responded or remitted in terms of their depression. So there’s evidence that HRT works, but there haven’t been a lot of studies (Schmidt PJ et al, Am J Obstet Gynecol 2000;183(2):414–420; Soares CN et al, Arch Gen Psychiatry 2001;58(6):529–534).
TCPR: What type of HRT did they use?
Dr. Nonacs: Those two studies were done with transdermal HRT. Earlier studies used oral versions, and those require higher doses, which seem to have greater risks of heart disease and breast cancer. You can get away with lower doses with the transdermal form because you largely avoid the first-pass metabolism of the liver. The transdermal form also allows steadier serum levels.
TCPR: Does the patch include estrogen and progesterone?
Dr. Nonacs: Often the estrogen will be transdermal and the progestin is given orally. Estrogen has the mood-elevating effects, while progestin can be depressogenic, but you can’t give estrogen without some form of progesterone because estrogen causes the uterine lining to build up, which can lead to endometrial hyperplasia and ultimately cancer. The progestin decreases the growth by causing a menstrual cycle. But one thing about using HRT in perimenopausal women is that they don’t have to have cycles every month, so you could use the progesterone every 3–6 months for a withdrawal bleed and not have so much exposure to that component.
Table: Estrogen in Psychiatry
TCPR: You mentioned a lot of studies in healthy women. Does HRT improve well-being there?
Dr. Nonacs: Yes. I think women feel better on estrogen, especially in energy, weight gain, and cognition.
TCPR: Do those symptoms cause problems in perimenopausal depression as well?
Dr. Nonacs: Yes, particularly cognitive symptoms, and this is where serotonergic antidepressants don’t do a full job. Sometimes I augment with bupropion, atomoxetine (Strattera), or stimulants if the problem is severe. We were hopeful that vortioxetine would help, but its cognition benefits aren’t as clear in the perimenopausal population as they are in other groups. It has two small trials and improved cognition in only one (Freeman MP et al, Ann Clin Psychiatry 2017;29(4):249–257; Callegari C et al, Psychopharmacol Bull 2019;49(1):28–43).
TCPR: Do you always refer out for HRT?
Dr. Nonacs: Yes. I don’t feel comfortable that I can ensure adequate monitoring: blood pressure, excessive bleeding, fibroid growth, as well as the breast cancer and cardiovascular risks. I’ll say to the OB/GYN, “We’ve tried these antidepressants without a good response. Would you consider HRT for this patient?” If the OB/GYN agrees, I let them choose the HRT they feel most comfortable with. Usually they choose an oral contraceptive for younger perimenopausal women and estrogen patches in older women.
TCPR: Does the patch provide contraception as well?
Dr. Nonacs: There are a couple of different patches. There are contraceptive patches, which do confer protection, but HRT does not confer contraceptive benefit. The main difference is the estrogen dose, which is higher in HRT than contraception.
TCPR: How long do OB/GYNs usually continue HRT?
Dr. Nonacs: Usually they try to use it for less than 5 years. What they will typically do is withdraw the HRT gradually. And if women experience the return of menopausal symptoms like hot flashes and night sweats, they might consider going beyond the 5 years. We may also see a relapse into depression then, in which case we’ll put the patch back on and try to beef up the antidepressant treatment.
TCPR: What about a woman who is 10 years past menopause and has depression? Would HRT even work there?
Dr. Nonacs: It probably doesn’t work as well in that population. I do see a lot of women who say, “I haven’t felt well since I stopped menstruating. I need to go back on hormones.” I generally respond, “There are risks associated with it. It might not work as well. Can we explore other treatments for depression?”
TCPR: What about bipolar disorder?
Dr. Nonacs: Mood can become more unstable during the perimenopausal period. Here the antidepressant approach may backfire with cycling and mixed states, so I’ll raise the mood stabilizers or add an atypical. Some bipolar patients go on to HRT and it seems to help, but we don’t have much research on that.
TCPR: How is it that these hormonal therapies can help perimenopausal depression but make mood worse in adolescents?
Dr. Nonacs: You do see a small uptick in mood symptoms and suicidality in adolescents who have started OCPs. But there they are getting higher levels of estrogen, not replacement, which is why hormonal treatments seem to help mood in one population and may be harmful in another. I think it’s a small subset of adolescents who are vulnerable to this, but we can’t predict it in advance.
TCPR: Are some oral contraceptives less likely to disrupt mood?
Dr. Nonacs: We’ll often start with the lower dosages and prefer Yaz or Yasmin (drospirenone and ethinyl estradiol), which seem to cause fewer mood problems (Kelly S et al, Clin Drug Investig 2010;30(5):325–336). I’d avoid triphasic preparations, where the hormone levels fluctuate more. If the patient is having a lot of premenstrual symptoms while taking OCPs, we might recommend spacing out their periods; instead of going 21 days on hormone and 7 days off, we skip the placebo week. Then every 6 months they’ll do a placebo week so they have a period, then they’ll go back on continuous treatment.
TCPR: Thank you for your time, Dr. Nonacs.
Listen to our 10/12/20 podcast, “Menopause and Mood: An Interview with Ruta Nonacs, MD.” Search for “Carlat” on your podcast store.
PO Box 626, Newburyport MA 01950