The standard treatment-resistant depression (TRD) workup — TSH, B12, CBC, basic metabolic panel — is necessary but inadequate. Depression is a heterogeneous condition with multiple biological pathways, and the patient who has not responded to two or three adequate antidepressant trials warrants a broader workup that addresses the longevity-psychiatry mechanisms: inflammation, metabolic substrate, sleep architecture, hormonal axes, pharmacogenomics, and the cognitive-behavioral foundations that medications cannot replace. The workup is not paperwork; it is clinical action, and a properly executed TRD workup frequently identifies the treatable contributor that resolves the resistance.
Sleep is the first underrecognized contributor. Untreated obstructive sleep apnea is present in a substantial subset of depressed patients, particularly men over 50, patients with elevated BMI, hypertensives, and patients with non-restorative sleep. OSA suppresses REM sleep, fragments sleep architecture, drives systemic inflammation, and produces a depression that does not respond well to standard pharmacotherapy. The STOP-BANG screen takes one minute and identifies most high-risk patients; the polysomnography that follows finds the treatable contributor. Patients with TRD and undiagnosed OSA frequently experience depression remission with CPAP treatment alone. The workup that misses sleep apnea misses one of the most common treatable causes of TRD.
Thyroid evaluation beyond TSH. Subclinical hypothyroidism (TSH 4.0–10.0 with normal free T4) is associated with depression and TRD, and treatment of subclinical hypothyroidism in depressed patients frequently improves mood and treatment response. Beyond standard TSH, consider free T4, free T3 (low T3 syndrome is associated with depression and may not be reflected in TSH), thyroid antibodies (Hashimoto's can produce neuropsychiatric symptoms), and reverse T3 in selected cases. T3 augmentation (liothyronine 25–50 mcg) has long-standing evidence in TRD and remains underused. The thyroid-mood connection (Stage 19) is one of the most actionable in TRD workup.
Inflammation markers identify the inflamed-depression subtype. hs-CRP is the field-deployable measure; values above 3 mg/L in patients with TRD identify a subset that may respond preferentially to anti-inflammatory augmentation (omega-3 high-EPA, NSAIDs in select cases, minocycline in research populations, lifestyle anti-inflammatory interventions). IL-6, TNF-alpha, and other markers add depth in research settings. The inflamed-TRD subtype (covered in L10.5 and Stage 16) frequently responds poorly to serotonergic antidepressants and well to inflammation-targeted strategies; missing this distinction means missing the right treatment for a meaningful subset of patients.
Hormonal evaluation in perimenopausal women, men with hypogonadism, and patients on chronic glucocorticoids. Perimenopausal depression is biologically distinct and may respond preferentially to hormonal interventions (Stage 19) alongside or instead of standard antidepressants. Hypogonadism in men produces depression-like symptoms that respond to testosterone replacement (in appropriate candidates) more reliably than to SSRIs. Patients on chronic glucocorticoids — for inflammatory disease, asthma, autoimmune conditions — frequently develop steroid-induced depression that requires both the underlying steroid management and specific psychiatric treatment. Cortisol burden, HPA axis function, and reproductive hormones are part of the longevity TRD workup.
Pharmacogenomic testing has a defined but limited role. CYP2D6 and CYP2C19 polymorphisms predict altered metabolism of many antidepressants and can explain past treatment failures (ultrarapid metabolizers showing inadequate exposure on standard doses; poor metabolizers showing elevated side effects). HLA-B*1502 and HLA-A*3101 testing has specific use for carbamazepine in selected populations. Pharmacokinetic testing is more actionable than pharmacodynamic testing; broad multi-gene panels often produce more noise than signal in non-research settings. The patient with multiple unexplained treatment failures and a family history of similar patterns is the candidate where pharmacogenomic testing has the best yield. The workup discipline is to use the time and money on the tests that change the treatment plan, not to order panels that produce data without action.