Garrett Rossi, MD.
Inpatient/consult attending psychiatrist, AtlantiCare Regional Medical Center, Pomona, NJ.
Chris Aiken, MD.
Editor-in-Chief, The Carlat Psychiatry Report. Practicing psychiatrist, Winston-Salem, NC.
Dr. Rossi and Dr. Aiken, authors for this educational activity, have no relevant financial relationship(s) with ineligible companies to disclose.
When patients don’t respond to an antidepressant, it’s a good idea to step back and look for anything that might be getting in the way. Major stress, substance use, medication nonadherence, anxiety, and medical and psychiatric comorbidities are high on that list, but recent research has added more possibilities that we’ll review in this article.
Inflammation is the body’s natural response to a wound or infection. It is a normal immunologic response, but when it becomes chronic and uncoupled from its original mission of fighting off microbes, depression can result. Chronic inflammation contributes to one in three cases of depression and one in two cases of treatment-resistant depression (see The Carlat Psychiatry Report February 2020). Recent surgery, injury, or infection—including long COVID-19—are common causes, as are less obvious ones like poor diet, obesity, smoking, chronic stress, chronic medical illness, chemotherapy, insomnia, and lack of exercise.
If a patient has signs of inflammation and is not responding to an antidepressant, it may be worth checking their high-sensitivity C-reactive protein (hs-CRP). This inflammatory marker has been used in several studies to predict antidepressant response. Patients with an elevated hs-CRP tend to respond better to nortriptyline or bupropion than to SSRIs (Hashimoto K, Int J Mol Sci 2015;16(4):7796–7801; Jha MK et al, Psychoneuroendocrinology 2017;78:105–113). Elevated CRP also predicts response to a few complementary therapies: l-methylfolate (15 mg/day), n-acetylcysteine (2000 mg/day), and omega-3 fatty acids (1000–3000 mg/day, with an EPA:DHA ratio of at least 2:1; see The Carlat Psychiatry Report March 2022 for recommended products). Hs-CRP is a low-cost test ($20–$40), and the cutoff in most of these depression trials was >3 mg/dL for hs-CRP, although some used >2 or >1 mg/dL.
Lifestyle changes are particularly important for inflammatory depression because—in addition to improving mood—they treat the underlying problem by lowering CRP and other inflammatory markers. That is important because inflammation worsens morbidity and mortality, and it won’t necessarily go away by treating the depression. Exercise, Mediterranean-style diet, tai chi, mindfulness, and cognitive behavioral therapy for insomnia are treatments that both reduce inflammation and improve mood.
For the average patient with major depression, the optimal amount of exercise is 45 minutes of light aerobics three or four times a week. More intensive routines don’t tend to improve mood any further unless the patient has an elevated hs-CRP, in which case more intensive exercise brings greater remission (Trivedi MH et al, J Clin Psychiatry 2011;72(5):677–684).
About 60% of patients with depression respond to an SSRI, but that chance goes down to 40% in the presence of multiple medical comorbidities, particularly cardiovascular risk factors. Hypertension, hypercholesteremia (>200 mg/dL), smoking, and diabetes all dampen antidepressant response (Iosifescu DV et al, Am J Psychiatry 2003;160(12):2122–2127). Smoking cessation and good medical care are logical first steps, and some medical treatments may have antidepressant effects of their own. Metformin (1000 mg/day) and three of the statins (atorvastatin 20 mg/day, lovastatin 30 mg/day, and simvastatin 20 mg/day) augmented SSRIs in randomized controlled trials of patients with major depression without major medical illnesses, possibly because they have anti-inflammatory or neuroprotective effects (Abdallah MS et al, Neurotherapeutics 2020;17(4):1897–1906; Salagre E et al, J Affect Disord 2016;200:235–242).
Obesity is also a strong predictor of non-response as well as slow response to antidepressants. These problems begin in the overweight range (BMI 25–30) and worsen in obesity (BMI 30–35), severe obesity (BMI 35–40), and morbid obesity (BMI >40). However, obesity is strongly correlated with inflammation, and it’s difficult to disentangle that association. Two of the treatments that work better when inflammatory markers are high also work better when the patient’s BMI is elevated: augmentation with bupropion or augmentation with l-methylfolate (Jha MK et al, J Affect Disord 2018;234:34–37).
Trauma and isolation
It’s tempting to blame major stress when a patient doesn’t respond to an antidepressant, but how strong is that connection? Certainly, many patients do respond in spite of stress, and baseline personality traits can raise those odds. Traits associated with resilience, like grit, spirituality, and a sense of self-efficacy, are associated with a more favorable antidepressant response (Laird KT et al, Int J Geriatr Psychiatry 2018;33(12):1596–1603). On the other hand, a few specific stressors are robustly associated with a poor antidepressant response: isolation, poor social supports, low income or education level, and a history of childhood abuse or neglect.
Some of these problems are modifiable. Patients may have friends they can call on, but they may have kept to themselves because of cognitions that lead to depressive avoidance (“I’m no good,” “I’m a burden,” “It’s too stressful to be around people”). Those patterns may start to change when patients understand that they need at least a small daily dose of social interaction to give their antidepressant the best chance of working.
Antidepressants don’t work in a vacuum, and a finding from animal research can illustrate that point for skeptical patients. When antidepressants were given to mice with depression, the medication worked as long as the mice were allowed some social time with other mice. Antidepressants did nothing for mice that were kept isolated in their cages. Animal studies also teach us that social isolation interferes with the antidepressant effects of exercise (Rief W et al, Neurosci Biobehav Rev 2016;60:51–64).
Childhood trauma is less modifiable, but it can inform the direction of treatment. Patients who don’t respond to an antidepressant may do better with psychotherapy if they have a history of childhood trauma. When researchers compared psychotherapy, nefazodone, and the combination in 681 patients with chronic major depression, only the psychotherapy arms were effective for the 65% of subjects who had a history of early childhood abuse, neglect, or loss of a parent before age 15 (Nemeroff CB et al, Proc Natl Acad Sci USA 2003;100(24):14293–14296). Biological reasons may explain why these patients do not respond well to antidepressants, as early trauma is associated with volumetric shrinkage in the limbic system.
When patients don't fully respond to an antidepressant, look for modifiable factors that might be getting in the way. Addressing these is just as critical as the next step in the psychopharmacologic algorithm.
PO Box 626, Newburyport MA 01950