Education Use Only
For educational use only — Not for clinical decision-making without independent verification
Medical Associates  ·  Department of Nephrology ← urinenephrology.org
Nephrology Education Series

Psychotropic and ADHD Medications in Hypertension and CKD: Comprehensive Medical Review

Andrew Bland, MD, FACP, FAAP UICOMP · UDPA · Butler COM 2026-08-13 28 min read

Psychotropic and ADHD Medications in Hypertension and CKD: Comprehensive Medical Review

Honest evidence-based assessment. Depression, anxiety, and ADHD are common in the hypertension and CKD clinic, and the drugs that treat them act on the same sympathetic pathways we spend the clinic day trying to quiet. The blood-pressure effect tracks norepinephrine transporter blockade, not serotonin — and that single fact predicts the entire class hierarchy. For most patients the interaction risk exceeds the direct pressor risk.

Cross-references: - Hypertension Hub - 2025 AHA expanded hypertension guideline analysis - comprehensive management of hypertension - drug dosing in CKD quick reference

Executive Summary (TL;DR)

  1. The pressor effect is a norepinephrine transporter phenomenon. A pharmacoepidemiologic-pharmacodynamic study in VigiBase found the arterial-hypertension reporting signal correlated with each drug’s NET/SERT binding ratio (R² = 0.68, Pearson r = 0.82). SNRIs outranked SSRIs consistently. The same axis runs through the ADHD agents, where atomoxetine and viloxazine are NET inhibitors by design.

  2. SSRIs are blood-pressure neutral and remain first-line in hypertension and cardiovascular disease. The 2025 Lancet network meta-analysis (151 studies, 58,534 participants, 30 drugs) found greater than 11 mmHg separating nortriptyline from doxepin in systolic change — the spread within the antidepressant class exceeds any mean class effect.

  3. Venlafaxine’s risk lives above 300 mg/day, not at baseline blood pressure. Thase’s meta-analysis of 3,744 patients found the supine diastolic effect clinically significant only above 300 mg/day, and explicitly reported that venlafaxine did not worsen control in patients with preexisting hypertension. The widely quoted “3% vs 13%” figures do not appear in that paper.

  4. Amphetamines raise blood pressure with high-certainty evidence. The 2025 Cochrane review (56 RCTs, 10,583 participants) found SBP +1.93 mmHg, DBP +1.84 mmHg, and heart rate +3.71 bpm, all high-certainty, all sustained beyond eight weeks. Withdrawal for adverse effects: RR 2.69, absolute risk increase 4.3%, NNH approximately 23.

  5. “Non-stimulant” does not mean blood-pressure safe. In the 2025 Lancet Psychiatry network meta-analysis, amphetamines, lisdexamfetamine, and methylphenidate were not associated with larger haemodynamic increments than atomoxetine or viloxazine. Atomoxetine produced the largest pediatric pulse increase (+5.58 bpm) of any agent studied. Switching to atomoxetine to protect a blood pressure is mechanistically backwards.

  6. Guanfacine lowers blood pressure and may treat both conditions. Mean decrease against placebo in adults was SBP −10.1 mmHg (95% CI −13.76 to −6.44) — very low CINeMA confidence, but directionally certain. Guanfacine is a licensed antihypertensive; in an adult with ADHD and hypertension the regimen may need to come down, not up.

  7. Long-term ADHD medication is associated with incident hypertension. In 278,027 Swedish patients, hypertension risk rose with cumulative exposure — AOR 1.72 at 3–5 years and 1.80 beyond 5 years — even though large short-term cohorts show no excess myocardial infarction, stroke, or sudden death.

  8. The highest-yield interaction is a one-word substitution. Paroxetine and fluoxetine raise metoprolol exposure three- to five-fold, with documented bradycardia and AV block. Sertraline or escitalopram eliminates the problem.

  9. In CKD, antidepressant efficacy evidence is weak to negative. CAST (n = 201, non-dialysis CKD) found a between-group difference of 0.1 points (95% CI −1.1 to 1.3, P = .82) against a 2-point minimal clinically important difference. Treat modifiable contributors first.

The Honest Read for Clinical Practice

Three rules cover most of the clinical value here. (1) Choose by transporter, not by class label — if it blocks norepinephrine reuptake it raises pressure, whether it is sold as an antidepressant, a non-stimulant, or a stimulant. (2) Reconcile the beta-blocker before choosing the SSRI — the CYP2D6 collision between paroxetine/fluoxetine and metoprolol/carvedilol is more clinically consequential than any 2 mmHg mean effect. (3) In CKD, the drug is rarely the first answer — correct anemia, dialysis adequacy, sleep apnea, and opioid burden before adding a psychotropic, because the randomized evidence for benefit in this population is genuinely poor while the adverse-event excess is consistent. The counterpoint: depression in CKD is associated with death, so none of this is an argument for nihilism — it is an argument for sequence.

1. Mechanism — Why NET Blockade Is the Whole Story

Norepinephrine transporter inhibition leaves more norepinephrine in the synaptic cleft, including at peripheral sympathetic terminals and central cardiovascular nuclei. Sympathetic tone rises: faster heart rate, higher systemic vascular resistance, higher pressure. Serotonin transporter inhibition does not do this to any comparable degree.

Montastruc and colleagues tested this rather than assuming it. The design was two-stage: a disproportionality analysis in VigiBase computing reporting odds ratios for arterial hypertension across individual SRIs and SNRIs, then a pharmacodynamic regression of that signal against published NET/SERT binding-affinity ratios. Every individual SRI except fluvoxamine generated a significant reporting odds ratio, as did every SNRI, with SNRI values consistently higher.

Reading the Correlation Honestly

An R² of 0.68 across a small number of drug-level points is a strong correlation with wide practical uncertainty, and disproportionality analyses measure reporting, not incidence — venlafaxine’s reputation for raising blood pressure could itself inflate its signal through notoriety bias. The mechanism is sound and the ranking reproduces in randomized data. The precise slope is not a bedside number.

1.1 The Opposing Effect: Alpha-1 and Alpha-2

The clinical picture is a net of opposing forces. Tertiary-amine tricyclics block alpha-1 receptors and the norepinephrine transporter simultaneously, which is why imipramine can raise supine pressure while dropping standing pressure in the same patient. That is not a paradox; it is two receptors. At the other end, guanfacine and clonidine are alpha-2 agonists — the same receptors clonidine hits as an antihypertensive — and therefore move pressure downward.

2. Antidepressant and Anxiolytic Class Effects

Ranked by direct pressor risk. Point estimates are from the 2025 Lancet network meta-analysis unless noted.

Class / Agent Direct BP effect Magnitude Practical read
Levomilnacipran (SNRI) Increase, marked SBP +3.36, DBP +3.54 mmHg Highest NET selectivity in class
Venlafaxine (SNRI) Increase, marked SBP +2.78, DBP +2.55 mmHg Dose-dependent; the classic offender
Desvenlafaxine (SNRI) Increase SBP +1.93 mmHg New-onset HTN 1.9–4.8%, dose-dependent
Duloxetine (SNRI) Increase SBP +1.59 mmHg Mildest SNRI; usable in controlled HTN
Nortriptyline (TCA) Increase, marked Highest SBP in the network Also raises heart rate; least orthostasis among TCAs
Imipramine / amitriptyline (TCA) Up supine, down standing Mixed Orthostasis dominates clinically; imipramine worst
Bupropion (NDRI) Increase Dose-related Also causes orthostatic hypotension — see Section 3
MAOIs Decrease typically; crisis if triggered Bimodal Orthostatic hypotension is the common effect
SSRIs Neutral Not clinically significant Safest class in hypertension and CVD
Mirtazapine / mianserin Neutral Minimal Rare orthostatic hypotension and falls
Trazodone / nefazodone Decrease Orthostatic Alpha-1 blockade; syncope reported
Doxepin (TCA) Decrease Lowest SBP in the network Greater than 11 mmHg below nortriptyline
Buspirone (azapirone) Neutral Approximately 1% either direction Cleanest anxiolytic — see Section 5
Correction to a Widely Repeated Teaching Point

The frequently quoted “3% sustained hypertension below 300 mg, 13% above 300 mg” for venlafaxine does not come from the Thase meta-analysis. What that paper reported from 3,744 patients is that the supine diastolic effect was highly dose-dependent and clinically significant only above 300 mg/day — and that venlafaxine did not adversely affect blood pressure control in patients with preexisting high blood pressure or elevated baseline values. The practical consequence matters: preexisting hypertension is not, on this evidence, a reason to withhold venlafaxine at standard doses. Dose above 300 mg/day is the actual risk threshold.

2.1 Desvenlafaxine — Where Harm Numbers Are Computable

The pooled analysis of 11 short-term fixed-dose trials defined new-onset hypertension rigorously: three consecutive supine diastolic measures of 90 mmHg or greater with a rise of at least 10 mmHg, and/or supine systolic of 140 mmHg or greater with a rise of at least 10 mmHg.

Dose Incidence Placebo ARI NNH
50 mg/day 1.9% 0.8% 1.1% 91
200 mg/day 2.4% 0.8% 1.6% 63
400 mg/day 4.8% 0.8% 4.0% 25

ARI = absolute risk increase; NNH = number needed to harm over 8–12 weeks, computed from the reported incidences. An NNH of 91 at the standard 50 mg dose is a small signal. An NNH of 25 at 400 mg is not, and it reinforces the general SNRI rule: the pressor effect lives at the top of the dose range.

2.2 What the Network Meta-Analysis Did Not Find

Three negative findings deserve as much attention as the positive ones: no strong evidence that any antidepressant meaningfully prolongs QTc; no clinically significant effect on sodium; and none on urea or creatinine. Each conflicts with a substantial observational literature.

Source Discrepancy: Randomized Evidence vs. Real-World Signal

Friction point: whether antidepressants cause hyponatremia and QTc prolongation. The randomized evidence says no clinically significant effect across 58,534 participants. Observational and pharmacovigilance evidence says yes, consistently — an electronic-health-record study showed citalopram 20→40 mg raising QTc by a mean 10.3 ms within-subject.

Resolution: the trials are not built to see these events. Median treatment duration was 8 weeks. Antidepressant-associated hyponatremia is idiosyncratic, concentrated in older adults, and heavily potentiated by thiazides — precisely the patients acute monotherapy trials exclude. Use the randomized data to rank drugs; use the observational data to decide whom to monitor. The negative finding says these are not average effects in a healthy 8-week trial population. It does not say your 78-year-old on hydrochlorothiazide is safe.

3. Bupropion — The Dual-Mechanism Outlier

Bupropion deserves separate treatment because it behaves unlike anything else in this review, and because it reaches hypertensive and CKD patients through a side door: smoking cessation.

It moves pressure in both directions. As a norepinephrine-dopamine reuptake inhibitor it raises blood pressure, usually at higher doses — yet it also causes orthostatic hypotension, particularly in patients with established cardiovascular disease. Most agents do one or the other. Bupropion does both, which makes a single supine reading a poor way to monitor it.

It is a potent CYP2D6 inhibitor, placing it in the same interaction bucket as paroxetine and fluoxetine for anyone on metoprolol, carvedilol, propranolol, or nebivolol. This is the most commonly missed part of its profile, because bupropion is not thought of as an SSRI-type interaction risk — and it is frequently prescribed for smoking cessation by a clinician who is not managing the beta-blocker.

Its renal behavior is counterintuitive and specifically dangerous. The only dedicated study in dialysis patients gave a single 150 mg sustained-release dose to 8 hemodialysis smokers with sampling over 7 days:

  • Parent bupropion pharmacokinetics were similar to normal renal function — checking the parent drug reassures you falsely.
  • Hydroxybupropion and threohydrobupropion showed increased AUC, indicating accumulation.
  • Dialysis clearance of hydroxybupropion is unlikely — you cannot dialyze the problem away.
  • The authors concluded that 150 mg every 3 days is more appropriate in hemodialysis than the manufacturer’s recommended 150 mg daily.
The Bupropion Trap in CKD

Hydroxybupropion is pharmacologically active and accumulates in renal failure while the parent drug looks normal. Because bupropion’s dose-dependent toxicities — seizure-threshold reduction, agitation, blood-pressure elevation — track exposure, an ESRD patient on standard 150 mg daily is being dosed roughly three times higher than the only dedicated pharmacokinetic study supports. If a dialysis patient arrives on bupropion, the dose is the first thing to check, not the last.

State the limitation honestly: n = 8, single dose, and the authors themselves call for a multi-dose study. This is thin evidence — but it is the only evidence, and it points one direction.

4. ADHD Medications — Blood Pressure and Heart Rate

4.1 The Cochrane Amphetamine Data

This is the highest-quality evidence in the review: 56 randomized trials and 10,583 participants, adults and children, across ADHD, weight loss, and other indications.

Outcome Effect vs placebo 95% CI Certainty
Systolic BP +1.93 mmHg 1.54 to 2.31 High
Diastolic BP +1.84 mmHg 1.51 to 2.16 High
Heart rate +3.71 bpm 3.27 to 4.14 High
Withdrawal for adverse effects RR 2.69 2.13 to 3.40 High

Three features make this unusually solid. All four outcomes carry high-certainty GRADE ratings. A subgroup analysis restricted to trials of at least eight weeks found similar effects, indicating these are sustained, not transient. And the withdrawal outcome comes with an absolute figure — a 4.3% absolute risk increase over an average of one month, an NNH of approximately 23.

4.2 The Full Network Meta-Analysis

102 randomized trials, median follow-up 7 weeks, covering 13,315 children and adolescents and 9,387 adults across nine agents.

Population Parameter Range across agents CINeMA confidence
Children/adolescents SBP +1.07 (atomoxetine) to +1.81 (methylphenidate) Moderate
Children/adolescents DBP +1.93 (amphetamines) to +2.42 (methylphenidate) High to low
Children/adolescents Pulse +2.79 (viloxazine) to +5.58 (atomoxetine) Moderate to high
Adults SBP +1.66 (methylphenidate) to +2.3 (amphetamines) Very low
Adults DBP +1.60 (methylphenidate) to +3.07 (lisdexamfetamine) Very low
Adults Pulse +4.37 (methylphenidate) to +5.8 (viloxazine) Very low

Every adult estimate carries very low CINeMA confidence — the authors’ own rating. The direction is consistent; the magnitude is not reliable, and any precise adult figure quoted from this analysis should be tempered accordingly.

The “Non-Stimulant Is Safer” Assumption Is Not Supported

Amphetamines, lisdexamfetamine, and methylphenidate were not associated with larger haemodynamic increments than atomoxetine or viloxazine, in either children or adults. Atomoxetine produced the largest pulse increase in children (+5.58 bpm, 95% CI 4.67–6.49, high confidence) of any agent in the network. Switching a hypertensive patient from methylphenidate to atomoxetine to protect the blood pressure is not evidence-based — it is mechanistically backwards, because atomoxetine is a pure norepinephrine transporter inhibitor.

4.3 Guanfacine — The Agent That Lowers Pressure

Population SBP DBP Pulse
Children/adolescents −2.83 (95% CI −3.80 to −1.85) −2.08 (−3.00 to −1.17) −4.06 (−5.45 to −2.68)
Adults −10.1 (−13.76 to −6.44) −7.73 (−11.88 to −3.58) −6.83 (−10.85 to −2.81)

The adult estimates are very low confidence and derived from limited trials, so −10.1 mmHg should be read as a signal of a real and potentially large effect rather than a dosing target.

The Two-Birds Prescription, and Its Trap

In an adult with both ADHD and hypertension, guanfacine extended-release may treat both conditions — it is, after all, a licensed antihypertensive. Coordinate with psychiatry before adding it to an existing regimen, because in a patient already at goal it can produce symptomatic hypotension, bradycardia, and postural dizziness, all among the commonly reported adverse events in long-term adult use. If guanfacine is started, the antihypertensive regimen may need to come down, not up.

4.4 Do ADHD Medications Cause Hard Cardiovascular Events?

Two large cohorts said no. One large case-control study shows the picture changes with duration.

  • 1,200,438 children and young adults (2,579,104 person-years): current users were not at increased risk of serious cardiovascular events, adjusted HR 0.75 (95% CI 0.31–1.85). With only 81 events, the upper confidence limit cannot rule out a doubling of risk.
  • 150,359 adult medication users: adjusted RR 0.83 (95% CI 0.72–0.96) for current use versus nonuse. The methodologically important comparison — current versus remote use, which controls for healthy-user bias — was null at 1.03 (95% CI 0.86–1.24).

Against that, a Swedish case-control study of 278,027 individuals with ADHD found risk rising with cumulative duration:

Cumulative use Adjusted OR for CVD 95% CI
0 to 1 year 0.99 0.93–1.06
1 to 2 years 1.09 1.01–1.18
2 to 3 years 1.15 1.05–1.25
3 to 5 years 1.27 1.17–1.39
Greater than 5 years 1.23 1.12–1.36

Hypertension specifically was the standout endpoint — AOR 1.72 (95% CI 1.51–1.97) at 3–5 years and 1.80 (95% CI 1.55–2.08) beyond 5 years. Each additional year of use carried a 4% increase in cardiovascular risk.

How to Reconcile These

The short-term cohorts and the long-term case-control study are not in conflict — they measure different things. Acute myocardial infarction, stroke, and sudden death are not convincingly increased. Incident hypertension over years of continuous exposure is. That is exactly what a drug producing a sustained +2/+2 mmHg and +4 bpm should do: not rupture plaque this year, but shift a population’s pressure distribution over a decade. The nephrologic consequence — years of incrementally higher pressure in a CKD patient — is the outcome we care about most and the one shortest on direct evidence.

For bias assessment: the senior author of the study reporting harm discloses grants and personal fees from a manufacturer of ADHD medication. Industry ties that cut against the sponsor’s interest strengthen rather than weaken a finding.

5. Buspirone

Buspirone is a 5-HT1A partial agonist with no clinically meaningful alpha-adrenergic activity, regarded in the cardiovascular literature as having no significant cardiovascular effects. The label lists both hypertension and hypotension as infrequent (1 in 100 to 1 in 1,000) — the regulatory way of saying the signal is indistinguishable from noise. Its pharmacokinetics are dominated by extensive first-pass CYP3A4 metabolism, which is the source of nearly every problem it causes.

Efficacy context belongs with the safety claim. In a network meta-analysis of 89 generalized anxiety disorder trials (25,441 patients), buspirone was efficacious and well tolerated — but that finding was explicitly limited by small sample sizes, unlike duloxetine (MD −3.13, 95% CrI −4.13 to −2.13), venlafaxine (−2.69, −3.50 to −1.89), and escitalopram (−2.45, −3.27 to −1.63), which had robust evidence. Buspirone is the hemodynamically safest choice; it is not the most reliably effective one.

Interacting drug Effect on buspirone AUC
Diltiazem Increased 5.5-fold (Cmax 4.1-fold)
Verapamil Increased 3.4-fold (Cmax 3.4-fold)
Grapefruit juice Substantial increase
Itraconazole Marked increase
Rifampicin Markedly reduced — near-total loss of effect
The Interaction Hides in the Antihypertensive Regimen

A patient on diltiazem started on standard buspirone 7.5 mg twice daily receives exposure equivalent to roughly 40 mg twice daily. The presenting complaint will be dizziness, lightheadedness, and dysphoria — and the reflex is to blame the anxiety and increase the dose. Start at 2.5–5 mg twice daily when a non-dihydropyridine calcium channel blocker is on board. Dihydropyridines such as amlodipine do not carry this interaction.

Renal impairment. At steady state, patients with renal impairment had significantly higher Cmax and AUC than normal volunteers — but the intensity and frequency of adverse experiences were not significantly different. Exposure rises; measured toxicity did not. Standard dosing through stage 3; reduce to 5 mg twice daily in stage 4–5. Not meaningfully dialyzable (approximately 86% protein bound, large volume of distribution), so no post-dialysis supplement.

Titration and the usual failure mode. Start 7.5 mg twice daily, increase by 5 mg every 2–3 days, target 20–30 mg/day divided, maximum 60 mg/day. Onset is 2–4 weeks — the single most common cause of apparent treatment failure, because patients expecting benzodiazepine-like relief stop at day five. Say so at the first visit.

6. Drug–Drug Interactions with Antihypertensives

6.1 Absolute — Do Not Co-Prescribe

Combination Mechanism Consequence
MAOI + any serotonergic agent or buspirone Excess synaptic monoamine Hypertensive crisis, serotonin syndrome. 14-day washout each direction
MAOI + any stimulant or atomoxetine Unopposed noradrenergic surge Hypertensive crisis — labeled contraindication
Linezolid + SSRI / SNRI / buspirone / stimulant Linezolid is a reversible non-selective MAOI Hypertensive crisis with serotonin syndrome — reported
Methylene blue + serotonergic agent Potent MAO-A inhibition Same mechanism as linezolid
MAOI + sympathomimetics (OTC decongestants, tyramine foods) Unopposed pressor response Hypertensive crisis
Linezolid Is the One That Actually Happens on Your Service

The classic MAOI interaction is a historical curiosity in most practices. Linezolid is not. Every VRE bacteremia and every MRSA soft-tissue infection in a dialysis patient is a candidate, and a large fraction of those patients are on an SSRI, buspirone, or a stimulant that nobody reconciles. Hypertensive crisis accompanying serotonin syndrome after linezolid is documented, and pharmacovigilance work places linezolid among the highest-signal serotonergic interactants. Screen the psychiatric and ADHD medication list before the antibiotic is ordered, not after the pressure hits 220.

6.2 High-Risk — Avoid the Pairing or Change One Drug

Antihypertensive Psychotropic to avoid Mechanism Consequence
Metoprolol (CYP2D6 substrate) Paroxetine, fluoxetine Potent CYP2D6 inhibition Metoprolol AUC increased 3–5 fold; significant fall in SBP and heart rate; bradycardia and AV block in case reports
Carvedilol, propranolol, timolol, nebivolol Paroxetine, fluoxetine, bupropion, duloxetine CYP2D6 inhibition Clinically significant interaction demonstrated across the class
Any antihypertensive Venlafaxine above 300 mg/day, levomilnacipran NET blockade Direct antagonism of blood-pressure control
Any antihypertensive Amphetamines, lisdexamfetamine, methylphenidate, atomoxetine, viloxazine Sympathomimetic Sustained +2/+2 mmHg and +4 bpm
Diltiazem, verapamil Buspirone CYP3A4 inhibition Buspirone AUC increased 3.4–5.5 fold
Clonidine, guanfacine, methyldopa TCAs, mirtazapine, mianserin Central alpha-2 antagonism opposes alpha-2 agonism Loss of antihypertensive effect (mechanism-based; see note)
Beta-blockers and other bradycardic agents Guanfacine or clonidine prescribed for ADHD Additive Bradycardia, symptomatic hypotension, syncope
Thiazide diuretics SSRIs, especially citalopram Additive impairment of free-water excretion Severe hyponatremia
The Easiest Substitution in This Entire Review

A patient on metoprolol or carvedilol who needs an SSRI should receive sertraline, escitalopram, or citalopram — not paroxetine or fluoxetine. Sertraline and escitalopram are weak CYP2D6 inhibitors. This single substitution eliminates the most common clinically significant psychotropic–antihypertensive interaction in practice. If the patient must remain on paroxetine or fluoxetine, switch the beta-blocker to atenolol or bisoprolol, which are not CYP2D6-dependent.

A note on the clonidine–TCA interaction. This is well-established pharmacology — noradrenergic antidepressants antagonize the antihypertensive effect of central alpha-2 agonists, and mirtazapine is a direct alpha-2 antagonist. A dedicated modern clinical trial citation could not be identified, so it is presented here as mechanism-based reasoning supported by the general interaction literature rather than a trial-verified effect. Weigh it accordingly, but do not dismiss it: a patient whose clonidine stops working after starting amitriptyline is a recognizable clinical picture.

6.3 Moderate — Co-Prescribe with Monitoring

  • SSRI or SNRI plus a thiazide — check sodium at 2 and 4 weeks. Severe hyponatremia is documented specifically from the combination, and large contemporary datasets confirm the association across common antidepressants.
  • St John’s wort plus a calcium channel blocker — CYP3A4 induction reduces antihypertensive exposure. Ask about it; patients do not consider it a medication.
  • Citalopram or escitalopram plus a QT-prolonging adjunct such as sotalol — cap citalopram at 20 mg/day over age 60, and lower in CKD.
  • Trazodone plus any antihypertensive regimen — additive orthostasis and syncope.
  • Stimulant plus a non-dihydropyridine CCB or beta-blocker — the pressor effect partially offsets rate control; recheck pressure and pulse 2–4 weeks after any stimulant dose change.

6.4 Reference Table — Safest Pairings

If the patient is on… Prefer Avoid
Metoprolol, carvedilol, propranolol Sertraline, escitalopram, citalopram 20 mg or less Paroxetine, fluoxetine, bupropion, duloxetine
Diltiazem, verapamil Sertraline, escitalopram Buspirone at standard dose — start 2.5–5 mg twice daily
Clonidine, guanfacine SSRI TCAs, mirtazapine
Thiazide diuretic Any, with sodium monitoring Citalopram in the elderly without monitoring
Amlodipine, ACE inhibitor/ARB, spironolactone Any — no significant interaction —
Linezolid (any duration) Hold serotonergic and stimulant agents; discuss with ID SSRI, SNRI, buspirone, TCA, MAOI, stimulants, atomoxetine
Any regimen, patient needs ADHD treatment Guanfacine ER (lowers BP) — coordinate deprescribing Assuming atomoxetine is the blood-pressure-safe option

ACE inhibitors, ARBs, amlodipine, and mineralocorticoid receptor antagonists are essentially interaction-free with these drug classes. The problems concentrate in beta-blockers (CYP2D6), non-dihydropyridine calcium channel blockers (CYP3A4), thiazides (sodium), and central alpha-2 agonists (pharmacodynamic collision in both directions).

6.5 Serotonin Syndrome — Recognition and the Renal Consequence

Serotonin syndrome earns a place in a blood-pressure review for two reasons. It presents with hypertension, so it sits in the differential every time one of these drugs is followed by a pressure spike. And its feared complication is rhabdomyolysis with acute kidney injury, which makes it our problem, not just the intensive care unit’s.

Diagnose with the Hunter criteria, not Sternbach’s. Analysis of 2,222 consecutive serotonergic-overdose admissions derived decision rules from seven clinical features — clonus (inducible, spontaneous, or ocular), agitation, diaphoresis, tremor, hyperreflexia, plus hypertonicity and temperature above 38 °C, which were universal in life-threatening cases. The Hunter criteria are simpler, more sensitive (84% vs 75%) and more specific (97% vs 96%) than Sternbach’s.

Clonus Is the Discriminator

Of everything on that list, clonus does the diagnostic work, and it is characteristically lower-limb predominant. Dorsiflex the ankle. If neuroleptic malignant syndrome is the competing diagnosis, the distinction is mechanical: serotonin syndrome gives hyperreflexia and clonus with rapid onset, usually under 24 hours; NMS gives lead-pipe rigidity, bradykinesia, and a course over days to weeks following a dopamine antagonist. Anticholinergic toxicity gives dry skin, absent bowel sounds, and normal reflexes. Three toxidromes, one fever — the reflex exam separates them.

Where this actually happens in nephrology. The classic MAOI pairing is rare. The realistic triggers on a renal service are:

  • Linezolid started in a dialysis patient already taking an SSRI, SNRI, or buspirone — documented to produce hypertensive crisis with serotonin syndrome
  • Methylene blue given intraoperatively or for vasoplegia
  • Fentanyl, tramadol, or methadone added to an existing SSRI
  • Ondansetron in a patient on multiple serotonergic agents — common and easily missed
  • A serotonergic agent restarted before an adequate 14-day MAOI washout

Management is supportive and largely a matter of what not to do: stop every serotonergic agent, give benzodiazepines for agitation and to blunt the hyperadrenergic response, cool actively, and consider cyproheptadine, a 5-HT2A antagonist, in moderate-to-severe cases. Dantrolene and bromocriptine belong to other syndromes and have no role here.

Two Renal-Specific Traps

1. Do not use physical restraints in an agitated patient. Isometric muscle contraction against restraints worsens hyperthermia, lactic acidosis, and rhabdomyolysis — the exact triad that produces the kidney injury. Chemical sedation with benzodiazepines is the correct control measure.

2. Standard rhabdomyolysis management does not transfer to the anuric dialysis patient. The reflex is aggressive volume expansion to protect the kidney. In a patient with no residual function there is no kidney left to protect and no urine to make — volume loading buys pulmonary edema, not renal salvage. In that population the management is temperature and CK control plus urgent dialysis for hyperkalemia and acidosis, recognizing that myoglobin clearance by conventional dialysis is limited. Check CK, potassium, and a venous gas early; the potassium moves faster than the creatinine.

7. Prescribing in CKD and on Dialysis

7.1 Does Antidepressant Therapy Work in CKD?

Here nephrology diverges sharply from general psychiatry, and the honest answer is uncomfortable. Roughly one quarter of adults with CKD meet interview-based criteria for depression — 22.8% (95% CI 18.6–27.6) in stage 5D and 21.4% (11.1–37.2) across stages 1–5 — and depression in CKD is associated with death. The question is not whether to treat but what actually helps.

  • CAST randomized 201 patients with stage 3–5 non-dialysis-dependent CKD and major depressive disorder to sertraline (median achieved dose 150 mg/day) or placebo for 12 weeks. QIDS-C16 changed by −4.1 with sertraline and −4.2 with placebo — between-group difference 0.1 (95% CI −1.1 to 1.3), P = .82. The minimal clinically important difference was 2 points; the entire confidence interval excludes it. This is not an underpowered null.
  • ASCEND compared sertraline with cognitive behavioral therapy in 120 hemodialysis patients. Sertraline produced lower scores at 12 weeks (effect estimate −1.84; 95% CI −3.54 to −0.13; P = 0.035) but adverse events were more frequent. There was no randomized comparison with no treatment.
  • A 6-month placebo-controlled trial in 30 hemodialysis patients found scores improved substantially in both arms (BDI-II 29.1 ± 8.4 to 17.3 ± 12.4, P < 0.001) with no difference between sertraline and placebo, and greater dropout from adverse events on drug.
What This Evidence Actually Supports

In CKD and dialysis the placebo and attention response is large, the incremental drug effect is small to absent, and adverse events are consistently more frequent on drug. That argues for starting with non-pharmacologic treatment, correcting modifiable contributors — anemia, inadequate dialysis, sleep apnea, polypharmacy, opioid burden — and reserving antidepressants for clear, persistent, functionally impairing major depression. It is not an argument for nihilism, because depression in CKD is associated with death. It is an argument about sequence.

7.2 Renal Dosing — Antidepressants and Anxiolytics

European Renal Best Practice reviewed 28 pharmacokinetic studies covering 24 antidepressants in CKD 3–5. Clearance was markedly reduced for selegiline, amitriptylinoxide, venlafaxine, desvenlafaxine, milnacipran, bupropion, reboxetine, and tianeptine. The data were too sparse and heterogeneous for meta-analysis — a limitation worth stating rather than papering over.

Agent CKD 3–4 ESRD / dialysis
Sertraline No adjustment No adjustment; not dialyzable
Escitalopram No adjustment Maximum 10 mg/day; caution
Citalopram Maximum 20 mg/day Maximum 20 mg/day (QT)
Venlafaxine Reduce 25–50% if CrCl below 30 Clearance reduced approximately 55%; poorly dialyzable
Desvenlafaxine Reduce 50 mg every other day
Duloxetine No adjustment if CrCl 30 or above Not recommended below CrCl 30 — Cmax and AUC approximately 2-fold higher; conjugated metabolites up to 9-fold higher
Bupropion Reduce 150 mg every 3 days — parent PK looks normal while active metabolites accumulate and are not dialyzable
Mirtazapine Reduce Reduce
Buspirone 5 mg twice daily in stage 4–5 Not dialyzable; label advises against in severe impairment
TCAs No adjustment (hepatic) No adjustment; monitor levels

7.3 Renal Dosing — ADHD Medications

The Evidence Base Here Is Thin

A targeted literature search returned no studies of ADHD pharmacotherapy specifically in CKD or dialysis populations, and none on methylphenidate pharmacokinetics in renal impairment. The guidance below is derived from elimination pathways and product labeling, not from renal-population trials. Say that out loud to patients and to psychiatry colleagues rather than implying a firmer basis than exists.

Agent Elimination CKD 3–4 ESRD / dialysis
Methylphenidate Hepatic de-esterification to inactive ritalinic acid No adjustment expected Ritalinic acid accumulates; significance unknown
Amphetamine / lisdexamfetamine Renal, urine-pH dependent Caution; reduce Lisdexamfetamine maximum 30 mg/day below CrCl 30 per label; avoid in dialysis
Atomoxetine CYP2D6 hepatic No adjustment No adjustment expected
Viloxazine Hepatic (CYP2D6, UGT) Limited data Limited data
Guanfacine Mixed — non-renal elimination compensates No major adjustment No major adjustment
Clonidine Approximately 50% renal, active drug Reduce Reduce; monitor for bradycardia

Guanfacine is the pharmacokinetic surprise, and a favorable one. Across the GFR spectrum, renal clearance falls from 233 to 34 to 18 mL/min and cumulative urinary excretion from 57% to 14% to 7.5% — yet total body clearance is largely preserved and the elimination half-life stays at 14 hours independent of renal function, because non-renal elimination takes over. Note the wide standard deviations in these small studies: “no major adjustment” means start low and titrate to effect, not that monitoring is unnecessary.

Amphetamines and Urine pH — An Under-Recognized CKD Interaction

Amphetamine is a weak base excreted renally, and its clearance is strongly urine-pH dependent: alkaline urine promotes tubular reabsorption and raises systemic exposure, acidic urine promotes excretion. Mechanistic modeling reproduced substantial changes in renal clearance and exposure across simulated urine-pH scenarios. This is simulation work rather than a clinical trial, but the underlying physiology is standard.

The nephrology implication is concrete: a CKD patient on sodium bicarbonate for metabolic acidosis, or on acetazolamide, has systematically altered urine pH and therefore altered amphetamine exposure. A patient whose stimulant “stopped working” or suddenly became intolerable after a bicarbonate dose change is a plausible and under-recognized presentation — and chronic bicarbonate therapy is common in exactly the population where nobody thinks about it.

7.4 The Adherence Problem

A multinational cohort of 12,174,321 adults initiating antihypertensive therapy, of whom 320,691 (2.6%) had ADHD, found ADHD associated with increased antihypertensive discontinuation at 5 years (HR 1.14; 95% CI 1.02–1.27), concentrated in middle-aged and older adults.

The effect size is modest and the design observational — confounding by socioeconomic factors, comorbid substance use, and health-system contact is not fully addressable. But the direction is clinically intuitive and actionable: executive dysfunction is a medication-adherence problem, and antihypertensive adherence is the whole ballgame in CKD progression. In a patient with ADHD and hypertension, simplify the regimen, use combination pills, and consider that treating the ADHD may improve blood-pressure control through adherence even as the stimulant nudges pressure up by 2 mmHg.

8. Monitoring Schedule

Timepoint Check Why
Baseline BP supine and standing, HR, sodium, ECG if citalopram/escitalopram plus risk factors Establishes the comparator
2 weeks BP, orthostatics, sodium if on thiazide or SSRI in the elderly Hyponatremia peaks early
4 weeks BP, HR, sodium, symptom response Drug effect emerging
8–12 weeks BP, HR, full metabolic panel Dose escalation is when SNRI and stimulant pressor effects appear
After any dose increase BP and HR within 2 weeks Effects are dose-dependent, not class-dependent
Annually on long-term ADHD medication BP, and reassess indication Hypertension risk rises with cumulative years of exposure
Do Not Skip Orthostatics

In the hypertension clinic the instinct is to watch pressure go up. In the CKD and geriatric clinic, the event that lands the patient in the hospital is pressure going down on standing. Check orthostatics before and two weeks after starting any of these drugs in a patient over 70, on three or more antihypertensives, or on dialysis. The supine pressure you record in clinic tells you nothing about the pressure at 3 a.m. on the way to the bathroom.

9. Limitations and Evidence Quality

  • The strongest antidepressant evidence — the 2025 Lancet network meta-analysis — is academically funded (NIHR, Maudsley Charity, Wellcome Trust, MRC), open access, and measures direct physiological endpoints rather than surrogates. Its median treatment duration of 8 weeks cannot capture chronic or idiosyncratic effects, and trial populations are healthier than clinic populations.
  • The Cochrane amphetamine review carries high-certainty ratings on all four outcomes, but 13 of 56 studies (23%) had high risk of bias in at least one domain, mostly attrition, and the pooled population mixes ADHD, weight-loss, and other indications — so the effect is an amphetamine class effect, not an ADHD-treatment effect specifically.
  • The ADHD network meta-analysis rates every adult haemodynamic estimate as very low confidence, which is why the guanfacine figure must be read as a signal rather than a number. Median follow-up of 7 weeks cannot address long-term hypertension risk.
  • The long-term hypertension signal comes from a case-control design that cannot establish causation; confounding by indication is the obvious threat, since sicker patients stay on medication longer.
  • The venlafaxine meta-analysis is 28 years old, pools manufacturer-controlled trials, and used imipramine as comparator. Its central dose-threshold claim has been reproduced in direction and rough magnitude by contemporary randomized data, which is why it survives here.
  • There is essentially no renal-population evidence for ADHD medications, and the CKD antidepressant trials are negative or modest. Both facts are stated inline rather than smoothed over.

10. Bottom Line

  1. Blood-pressure effect tracks NET blockade, not SERT, across all three drug families.
  2. SSRIs are blood-pressure neutral and remain first-line in hypertension and cardiovascular disease.
  3. SNRIs raise blood pressure dose-dependently: levomilnacipran, then venlafaxine, desvenlafaxine, duloxetine.
  4. Venlafaxine’s clinically significant effect appears above 300 mg/day; preexisting hypertension was not worsened at standard doses.
  5. Desvenlafaxine NNH for new-onset hypertension: 91 at 50 mg, 25 at 400 mg.
  6. Amphetamines: SBP +1.93, DBP +1.84 mmHg, HR +3.71 bpm — all high-certainty, all sustained; NNH approximately 23 for withdrawal.
  7. “Non-stimulant” does not mean blood-pressure safe. Atomoxetine had the largest pediatric pulse increase.
  8. Guanfacine lowers pressure and may treat ADHD and hypertension together.
  9. Long-term ADHD medication is associated with incident hypertension despite no excess short-term hard events.
  10. Paroxetine and fluoxetine raise metoprolol exposure 3–5 fold; diltiazem raises buspirone AUC 5.5-fold.
  11. In CKD, duloxetine is contraindicated below CrCl 30; bupropion needs 150 mg every 3 days on dialysis; sertraline and guanfacine are the safest defaults.
  12. In CKD and dialysis, randomized antidepressant efficacy evidence is weak to negative — treat modifiable contributors first.

References

  1. Pillinger T, Arumuham A, McCutcheon RA, et al. The effects of antidepressants on cardiometabolic and other physiological parameters: a systematic review and network meta-analysis. Lancet. 2025;406(10515):2063-2077. PMID 41135546

  2. Montastruc JL, Rousseau V, de Canecaude C, Roussin A, Montastruc F. Role of serotonin and norepinephrine transporters in antidepressant-induced arterial hypertension. Eur J Clin Pharmacol. 2020;76(9). PMID 32483650

  3. Calvi A, Fischetti I, Verzicco I, et al. Antidepressant drugs effects on blood pressure. Front Cardiovasc Med. 2021;8:704281. PMID 34414219

  4. Thase ME. Effects of venlafaxine on blood pressure: a meta-analysis of original data from 3744 depressed patients. J Clin Psychiatry. 1998;59(10). PMID 9818630

  5. Thase ME, Larsen KG, Reines E, Kennedy SH. Effects of desvenlafaxine on blood pressure in patients treated for major depressive disorder: a pooled analysis. Curr Med Res Opin. 2015;31(4). PMID 25758058

  6. Wernicke J, Lledó A, Raskin J, Kajdasz DK, Wang F. An evaluation of the cardiovascular safety profile of duloxetine. Drug Saf. 2007;30(5). PMID 17472422

  7. Licht CM, de Geus EJ, Seldenrijk A, et al. Depression is associated with decreased blood pressure, but antidepressant use increases the risk for hypertension. Hypertension. 2009;53(4). PMID 19237679

  8. Piña IL, Di Palo KE, Ventura HO. Psychopharmacology and cardiovascular disease. J Am Coll Cardiol. 2018;71(20). PMID 29773162

  9. Roose SP, Glassman AH, Siris SG, et al. Comparison of imipramine- and nortriptyline-induced orthostatic hypotension: a meaningful difference. J Clin Psychopharmacol. 1981;1(5). PMID 6277997

  10. Flockhart DA. Dietary restrictions and drug interactions with monoamine oxidase inhibitors: an update. J Clin Psychiatry. 2012;73 Suppl 1. PMID 22951238

  11. Shahbazi F, Mirzaei S, Mehrpooya M. Hypertensive crisis associated with serotonin syndrome following linezolid administration. Curr Drug Saf. 2025;20(1). PMID 38847376

  12. Elli C, Rovera C, Gervasoni C, et al. Serotonin syndrome: a pharmacovigilance comparative study of drugs affecting serotonin levels. Eur J Clin Pharmacol. 2024;80(3). PMID 38032391

  13. Lamberg TS, Kivistö KT, Neuvonen PJ. Effects of verapamil and diltiazem on the pharmacokinetics and pharmacodynamics of buspirone. Clin Pharmacol Ther. 1998;63(6):640-645. PMID 9663178

  14. Lilja JJ, Kivistö KT, Backman JT, Lamberg TS, Neuvonen PJ. Grapefruit juice substantially increases plasma concentrations of buspirone. Clin Pharmacol Ther. 1998;64(6). PMID 9871430

  15. Barbhaiya RH, Buch AB, Greene DS, Gammans RE. Disposition kinetics of buspirone in patients with renal or hepatic impairment. Eur J Clin Pharmacol. 1994;46(4). PMID 7911763

  16. Slee A, Nazareth I, Bondaronek P, Liu Y, Cheng Z, Freemantle N. Pharmacological treatments for generalised anxiety disorder: a systematic review and network meta-analysis. Lancet. 2019;393(10173):768-777. PMID 30712879

  17. Bahar MA, Kamp J, Borgsteede SD, Hak E, Wilffert B. The impact of CYP2D6 mediated drug-drug interaction: metoprolol and paroxetine/fluoxetine. Br J Clin Pharmacol. 2018;84(12). PMID 30248178

  18. Shin J, Hills NK, Finley PR. Combining antidepressants with β-blockers: evidence of a clinically significant CYP2D6 drug interaction. Pharmacotherapy. 2020;40(6). PMID 32342526

  19. Rosner MH. Severe hyponatremia associated with the combined use of thiazide diuretics and selective serotonin reuptake inhibitors. Am J Med Sci. 2004;327(2). PMID 14770031

  20. Mo H, Zhang Y, Wang Y, et al. Hyponatremia associated with the use of common antidepressants in the All of Us Research Program. Clin Pharmacol Ther. 2025;117(3). PMID 39540435

  21. Nagler EV, Webster AC, Vanholder R, Zoccali C. Antidepressants for depression in stage 3-5 chronic kidney disease: ERBP systematic review. Nephrol Dial Transplant. 2012;27(10). PMID 22859791

  22. Troy SM, Schultz RW, Parker VD, Chiang ST, Blum RA. The effect of renal disease on the disposition of venlafaxine. Clin Pharmacol Ther. 1994;56(1). PMID 8033490

  23. Lobo ED, Heathman M, Kuan HY, et al. Effects of varying degrees of renal impairment on the pharmacokinetics of duloxetine. Clin Pharmacokinet. 2010;49(5). PMID 20384393

  24. Worrall SP, Almond MK, Dhillon S. Pharmacokinetics of bupropion and its metabolites in haemodialysis patients who smoke. Nephron Clin Pract. 2004;97(3). PMID 15292684

  25. Behlke LM, Lenze EJ, Carney RM. The cardiovascular effects of newer antidepressants in older adults and those with or at high risk for cardiovascular diseases. CNS Drugs. 2020;34(11). PMID 33064291

  26. Palmer S, Vecchio M, Craig JC, et al. Prevalence of depression in chronic kidney disease: systematic review and meta-analysis. Kidney Int. 2013;84(1). PMID 23486521

  27. Palmer SC, Vecchio M, Craig JC, et al. Association between depression and death in people with CKD: a meta-analysis of cohort studies. Am J Kidney Dis. 2013;62(3). PMID 23623139

  28. Hedayati SS, Gregg LP, Carmody T, et al. Effect of sertraline on depressive symptoms in patients with CKD without dialysis dependence: the CAST randomized clinical trial. JAMA. 2017;318(19). PMID 29101402

  29. Mehrotra R, Cukor D, Unruh M, et al. Comparative efficacy of therapies for treatment of depression for patients undergoing maintenance hemodialysis. Ann Intern Med. 2019;170(6). PMID 30802897

  30. Friedli K, Guirguis A, Almond M, et al. Sertraline versus placebo in patients with major depressive disorder undergoing hemodialysis. Clin J Am Soc Nephrol. 2017;12(2). PMID 28126706

  31. Palmer SC, Natale P, Ruospo M, et al. Antidepressants for treating depression in adults with end-stage kidney disease treated with dialysis. Cochrane Database Syst Rev. 2016;(5). PMID 27210414

  32. Castro VM, Clements CC, Murphy SN, et al. QT interval and antidepressant use: a cross sectional study of electronic health records. BMJ. 2013;346:f288. PMID 23360890

  33. Chan M, Musini VM, Wright JM. Effect of amphetamines on blood pressure. Cochrane Database Syst Rev. 2025;3:CD007896. PMID 40152309

  34. Farhat LC, Lannes A, Del Giovane C, et al. Comparative cardiovascular safety of medications for attention-deficit hyperactivity disorder in children, adolescents, and adults. Lancet Psychiatry. 2025;12(5). PMID 40203844

  35. Zhang L, Li L, Andell P, et al. Attention-deficit/hyperactivity disorder medications and long-term risk of cardiovascular diseases. JAMA Psychiatry. 2024;81(2). PMID 37991787

  36. Hennissen L, Bakker MJ, Banaschewski T, et al. Cardiovascular effects of stimulant and non-stimulant medication for children and adolescents with ADHD. CNS Drugs. 2017;31(3). PMID 28236285

  37. Cooper WO, Habel LA, Sox CM, et al. ADHD drugs and serious cardiovascular events in children and young adults. N Engl J Med. 2011;365(20). PMID 22043968

  38. Habel LA, Cooper WO, Sox CM, et al. ADHD medications and risk of serious cardiovascular events in young and middle-aged adults. JAMA. 2011;306(24). PMID 22161946

  39. Fay TB, Alpert MA. Cardiovascular effects of drugs used to treat ADHD: part 1. Cardiol Rev. 2019;27(3). PMID 30365404

  40. Fay TB, Alpert MA. Cardiovascular effects of drugs used to treat ADHD: part 2. Cardiol Rev. 2019;27(4). PMID 30531411

  41. Kirch W, Köhler H, Braun W. Elimination of guanfacine in patients with normal and impaired renal function. Br J Clin Pharmacol. 1980;10 Suppl 1:33S-35S. PMID 6994776

  42. Kiechel JR. Pharmacokinetics of guanfacine in patients with impaired renal function and in some elderly patients. Am J Cardiol. 1986;57(9). PMID 3513525

  43. Huang W, Isoherranen N. Mechanistic PBPK modeling of urine pH effect on renal and systemic disposition of methamphetamine and amphetamine. J Pharmacol Exp Ther. 2020;373(3). PMID 32198137

  44. Yao H, Chang Z, et al. ADHD and adherence to antihypertensive medication treatment: a multinational cohort study. BMC Med. 2026;24. PMID 41721349

  45. Iwanami A, Saito K, Fujiwara M, Okutsu D, Ichikawa H. Safety and efficacy of guanfacine extended-release in adults with ADHD. BMC Psychiatry. 2020;20(1). PMID 33008345

  46. Jones DW, Ferdinand KC, Taler SJ, et al. 2025 AHA/ACC guideline for the prevention, detection, evaluation and management of high blood pressure in adults. Circulation. 2025. PMID 40811497

  47. Boyer EW, Shannon M. The serotonin syndrome. N Engl J Med. 2005;352(11). PMID 15784664

  48. Dunkley EJ, Isbister GK, Sibbritt D, Dawson AH, Whyte IM. The Hunter Serotonin Toxicity Criteria: simple and accurate diagnostic decision rules for serotonin toxicity. QJM. 2003;96(9). PMID 12925718

  49. Szałański F, Kozera G, et al. Serotonin syndrome: understanding pathophysiological bases and managing a growing clinical challenge. Neurol Neurochir Pol. 2025. PMID 41263645