π¨ Why This Lecture Exists
1 in 4
adults with CKD meet criteria for depression
3β5×
rise in metoprolol levels from paroxetine
5.5×
rise in buspirone levels from diltiazem
10 mmHg
systolic DROP from guanfacine in adults
These drugs are everywhere in the hypertension and kidney clinic. The interactions cause far more harm than the average blood-pressure effect β and the two most common mistakes are both avoidable with a single substitution.
π The One Rule
Do not memorize the list. Memorize the mechanism.
If a drug blocks the norepinephrine transporter (NET), it raises blood pressure. If it only blocks the serotonin transporter (SERT), it does not. This holds regardless of what the drug is marketed as β antidepressant, “non-stimulant” ADHD medication, or stimulant.
π΅ SERT blockade
Serotonin transporter. Treats depression and anxiety. Blood pressure: essentially unchanged.
SSRIs β sertraline, escitalopram, fluoxetine, paroxetine, citalopram
π΄ NET blockade
Norepinephrine transporter. More norepinephrine at sympathetic nerve endings β higher heart rate and vascular resistance. Blood pressure rises.
SNRIs, noradrenergic TCAs, bupropion, atomoxetine, viloxazine, stimulants
π’ Alpha-2 agonism
Reduces central sympathetic outflow β the same mechanism clonidine uses as a blood-pressure drug. Blood pressure falls.
Guanfacine, clonidine
π‘ Alpha-1 blockade
Blocks vasoconstriction on standing. Orthostatic hypotension β pressure drops when the patient stands.
Tertiary TCAs, trazodone
This is measured, not asserted. Researchers plotted each drugβs reported hypertension signal against its NET-versus-SERT binding ratio and found a strong correlation (R² = 0.68). The more a drug prefers NET, the more hypertension it produces.
π Antidepressants Ranked by Blood-Pressure Effect
β¬οΈβ¬οΈ Raises most
Levomilnacipran (SBP +3.4)
Venlafaxine (SBP +2.8)
Nortriptyline (highest in class)
β¬οΈ Raises modestly
Desvenlafaxine (SBP +1.9)
Duloxetine (SBP +1.6)
Bupropion (dose-related)
β‘οΈ Neutral β use these
All SSRIs
Mirtazapine
Buspirone
β¬οΈ Lowers / orthostatic
Doxepin (lowest in class)
Trazodone
MAOIs (usually)
β οΈ A correction to something you will hear repeated
You will be taught to avoid venlafaxine when blood pressure is uncontrolled. The original data do not say that. The meta-analysis of 3,744 patients found the effect became clinically significant only above 300 mg/day, and specifically reported that venlafaxine did not worsen control in patients who already had hypertension. Dose is the risk factor, not baseline blood pressure.
π¬ Bupropion β The Dual-Mechanism Outlier
It arrives through a side door: smoking cessation. The prescriber may not be the one managing the blood pressure.
Moves BP both ways
Raises pressure (blocks NE and dopamine reuptake) and causes orthostatic hypotension in cardiac patients. Check standing pressures, not just sitting.
Potent CYP2D6 inhibitor
Same danger group as paroxetine and fluoxetine for anyone on metoprolol or carvedilol β and almost nobody thinks of it that way.
The dialysis trap
Parent drug looks normal. Active metabolites accumulate and are not dialyzed.
β οΈ Bupropion in dialysis: 150 mg every 3 days, not daily
Because the parent drug level appears normal, checking it falsely reassures you. Hydroxybupropion builds up, and bupropionβs dose-related toxicities β lowered seizure threshold, agitation, higher blood pressure β follow total exposure. The only dedicated hemodialysis study concluded 150 mg every 3 days is more appropriate than the usual daily dosing. A dialysis patient on standard dosing is getting roughly 3Γ what the evidence supports.
Be honest about the evidence: n = 8, single dose. Thin β but it is the only direct evidence and it points one way.
β‘ ADHD Medications β Two Surprises
How much do stimulants actually raise blood pressure?
+1.9
mmHg systolic (high certainty)
+1.8
mmHg diastolic (high certainty)
+3.7
beats/min (high certainty)
1 in 23
stop the drug for side effects
56 randomized trials, 10,583 people. Small but real and sustained β still present beyond eight weeks.
β οΈ Surprise 1: “Non-stimulant” does NOT mean blood-pressure safe
The tempting move is switching a hypertensive patient from methylphenidate to atomoxetine because it is a “non-stimulant.” Across 102 trials, stimulants were not worse than atomoxetine or viloxazine β and atomoxetine produced the largest heart-rate increase of any agent studied in children (+5.6 beats/min).
Go back to the one rule and this is obvious: atomoxetine is a pure NET inhibitor. The category name says non-stimulant; the pharmacology says otherwise.
π Surprise 2: Guanfacine LOWERS blood pressure
Guanfacine is an alpha-2 agonist β a licensed antihypertensive that also treats ADHD. In adults it lowers systolic pressure by roughly 10 mmHg.
The two-birds prescription: in an adult with both ADHD and hypertension, guanfacine may treat both. The trap: in a patient already at goal it causes hypotension, bradycardia, and postural dizziness. If guanfacine starts, the other blood-pressure drugs may need to come down.
Do these drugs cause heart attacks and strokes?
No clear evidence of that. Two very large studies (1.2 million children/young adults; 150,000 adults) found no increase in heart attack, stroke, or sudden death.
But hypertension over years is a different question. In 278,027 people, the risk of being diagnosed with hypertension rose with duration of use β about 1.8× after more than 5 years.
These do not contradict. A drug adding 2 mmHg and 4 beats/min will not rupture a plaque this year, but over a decade it shifts the populationβs pressure upward. For a CKD patient, that is the outcome that matters.
β οΈ The Interactions That Change Management
This is the highest-yield part of the lecture. These cause more harm than any average blood-pressure effect.
π Never combine
- MAOI + any SSRI, SNRI, buspirone, or stimulant β hypertensive crisis and/or serotonin syndrome. 14-day washout each direction.
- Linezolid + any serotonergic drug or stimulant β linezolid is a reversible MAOI. Most people forget this.
- Methylene blue + serotonergic drug β same mechanism.
Linezolid is the one you will actually see. It treats VRE and resistant MRSA β infections common in dialysis patients β and many of those patients are already on an SSRI nobody reconciles. Check the psychiatric list before the antibiotic is ordered.
π§ The four fixes worth memorizing
1 On metoprolol or carvedilol?
Avoid paroxetine, fluoxetine, bupropion, duloxetine β they block CYP2D6 and raise metoprolol levels 3β5 fold, causing bradycardia and heart block.
Use instead: sertraline or escitalopram. This is the single highest-yield substitution in the whole topic.
2 On diltiazem or verapamil?
Avoid standard-dose buspirone β CYP3A4 inhibition raises buspirone levels 5.5-fold with diltiazem.
Do instead: start buspirone at 2.5β5 mg twice daily, not 7.5.
3 On a thiazide?
Watch SSRIs, especially citalopram β additive impairment of free-water excretion causes severe hyponatremia.
Do instead: check sodium at 2 and 4 weeks.
4 On clonidine or guanfacine?
Avoid TCAs and mirtazapine β they block the very alpha-2 receptor the blood-pressure drug is stimulating.
Use instead: an SSRI.
Good news: ACE inhibitors, ARBs, amlodipine, and spironolactone are essentially interaction-free with all of these drugs. The trouble concentrates in beta-blockers, diltiazem/verapamil, thiazides, and clonidine/guanfacine.
π₯ Serotonin Syndrome β Recognize It Fast
The acute emergency in this drug group. It matters to us because severe cases cause rhabdomyolysis β acute kidney injury.
Hunter criteria β and clonus is the finding that makes the diagnosis
More accurate than the older Sternbach criteria (84% vs 75% sensitivity; 97% vs 96% specificity). Seven features: clonus (inducible, spontaneous, ocular), agitation, diaphoresis, tremor, hyperreflexia, hypertonicity, temperature above 38 °C.
Dorsiflex the ankle. Clonus is characteristically worse in the legs than the arms.
Three look-alikes β the reflex exam separates them
Serotonin syndrome
Hyperreflexia + clonus
Onset under 24 hours
Neuroleptic malignant syndrome
Lead-pipe rigidity, bradykinesia
Onset days to weeks, after a dopamine blocker
Anticholinergic toxicity
Dry skin, absent bowel sounds
Normal reflexes
Common triggers on service: linezolid, methylene blue, fentanyl, tramadol, methadone, ondansetron β each added to a patient already on an SSRI.
Management: stop all serotonergic drugs, benzodiazepines for agitation, active cooling, consider cyproheptadine in moderateβsevere cases. Dantrolene and bromocriptine treat other syndromes and have no role.
β οΈ Two things NOT to do β both renal
1. No physical restraints. Straining against restraints produces isometric contraction, worsening fever, acidosis, and muscle breakdown β the exact process that injures the kidney. Sedate chemically.
2. No reflexive large-volume fluids in a dialysis patient. In rhabdomyolysis you were taught aggressive IV fluid. In an anuric patient there is no kidney left to protect and nowhere for the fluid to go β you cause pulmonary edema. They need urgent dialysis for potassium and acid, not volume.
π« Dosing in Kidney Disease
β
Safest defaults
Sertraline β no change at any stage
Guanfacine β no major change; non-renal clearance compensates
β οΈ Reduce
Venlafaxine β clearance falls approximately 55%
Escitalopram β max 10 mg
Citalopram β max 20 mg (QT)
Buspirone β 5 mg twice daily in stage 4β5
π Avoid / special
Duloxetine β not recommended below eGFR 30
Bupropion β 150 mg every 3 days on dialysis
Amphetamines β renally excreted; avoid in dialysis
π¬ The nephrology detail almost nobody knows
Amphetamine clearance depends on urine pH. Amphetamine is a weak base: alkaline urine β more reabsorbed β levels rise. Acidic urine β more excreted β levels fall.
Many CKD patients take sodium bicarbonate for metabolic acidosis, which alkalinizes the urine. If a patient says their ADHD medication “stopped working” or suddenly became intolerable after a bicarbonate change β this is a real and under-recognized explanation.
β οΈ Be honest about what is not known
There are no studies of ADHD medications in CKD or dialysis populations. The recommendations above are inferred from elimination pathways, not from trials in kidney patients. Say so when you present β a confident number you cannot source is worse than an acknowledged gap.
π§© Decision Points β Work Through These
A 58-year-old on metoprolol needs an antidepressant. Resident writes paroxetine.
Problem: paroxetine inhibits CYP2D6 β metoprolol exposure up 3β5 fold β bradycardia, heart block.
Fix: sertraline or escitalopram β or switch to atenolol/bisoprolol, which are not CYP2D6-dependent.
B 34-year-old with ADHD and new hypertension. Colleague suggests atomoxetine “because it is not a stimulant.”
Is that sound? No. Atomoxetine is a pure NET inhibitor β the exact mechanism that raises pressure β and had the largest heart-rate rise of any agent studied.
Better: guanfacine, which lowers blood pressure and treats ADHD.
C Hemodialysis patient on sertraline started on linezolid for VRE. Day 2: agitated, sweating, 38.9 °C, ankle clonus.
Diagnosis: serotonin syndrome β linezolid is a reversible MAOI.
Do: stop serotonergic drugs, benzodiazepines, cool, consider cyproheptadine.
Do NOT: use restraints (worsens rhabdomyolysis) or load volume in an anuric patient β dialyze for potassium and acidosis.
π§ͺ Apply It β Interactive Cases
Four interactive cases built on this material. Each turns on a decision point where the intuitive answer is wrong.
CYP2D6, metoprolol, and bradycardia from an unchanged dose
Start case →
Serotonin syndrome, rhabdomyolysis, and the anuric fluid trap
Start case →
Atomoxetine, guanfacine, and ADHD with hypertension
Start case →